Professional Documents
Culture Documents
Release 6.6.1
The software described in this book is furnished under a license agreement and may be used only in accordance with the terms of the agreement. Documentation version: 6.6.1, revision 1
Legal Notice
Copyright 2010 Symantec Corporation. All rights reserved. Symantec, the Symantec Logo, NetBackup, and PureDisk are trademarks or registered trademarks of Symantec Corporation or its affiliates in the U.S. and other countries. Other names may be trademarks of their respective owners. This Symantec product may contain third party software for which Symantec is required to provide attribution to the third party (Third Party Programs). Some of the Third Party Programs are available under open source or free software licenses. The License Agreement accompanying the Software does not alter any rights or obligations you may have under those open source or free software licenses. Please see the Third Party Legal Notice Appendix to this Documentation or TPIP ReadMe File accompanying this Symantec product for more information on the Third Party Programs. The product described in this document is distributed under licenses restricting its use, copying, distribution, and decompilation/reverse engineering. No part of this document may be reproduced in any form by any means without prior written authorization of Symantec Corporation and its licensors, if any. THE DOCUMENTATION IS PROVIDED "AS IS" AND ALL EXPRESS OR IMPLIED CONDITIONS, REPRESENTATIONS AND WARRANTIES, INCLUDING ANY IMPLIED WARRANTY OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE OR NON-INFRINGEMENT, ARE DISCLAIMED, EXCEPT TO THE EXTENT THAT SUCH DISCLAIMERS ARE HELD TO BE LEGALLY INVALID. SYMANTEC CORPORATION SHALL NOT BE LIABLE FOR INCIDENTAL OR CONSEQUENTIAL DAMAGES IN CONNECTION WITH THE FURNISHING, PERFORMANCE, OR USE OF THIS DOCUMENTATION. THE INFORMATION CONTAINED IN THIS DOCUMENTATION IS SUBJECT TO CHANGE WITHOUT NOTICE. The Licensed Software and Documentation are deemed to be commercial computer software as defined in FAR 12.212 and subject to restricted rights as defined in FAR Section 52.227-19 "Commercial Computer Software - Restricted Rights" and DFARS 227.7202, "Rights in Commercial Computer Software or Commercial Computer Software Documentation", as applicable, and any successor regulations. Any use, modification, reproduction release, performance, display or disclosure of the Licensed Software and Documentation by the U.S. Government shall be solely in accordance with the terms of this Agreement.
Technical Support
Symantec Technical Support maintains support centers globally. Technical Supports primary role is to respond to specific queries about product features and functionality. The Technical Support group also creates content for our online Knowledge Base. The Technical Support group works collaboratively with the other functional areas within Symantec to answer your questions in a timely fashion. For example, the Technical Support group works with Product Engineering and Symantec Security Response to provide alerting services and virus definition updates. Symantecs support offerings include the following:
A range of support options that give you the flexibility to select the right amount of service for any size organization Telephone and/or Web-based support that provides rapid response and up-to-the-minute information Upgrade assurance that delivers software upgrades Global support purchased on a regional business hours or 24 hours a day, 7 days a week basis Premium service offerings that include Account Management Services
For information about Symantecs support offerings, you can visit our Web site at the following URL: www.symantec.com/business/support/ All support services will be delivered in accordance with your support agreement and the then-current enterprise technical support policy.
Hardware information Available memory, disk space, and NIC information Operating system Version and patch level Network topology Router, gateway, and IP address information Problem description:
Error messages and log files Troubleshooting that was performed before contacting Symantec Recent software configuration changes and network changes
Customer service
Customer service information is available at the following URL: www.symantec.com/business/support/ Customer Service is available to assist with non-technical questions, such as the following types of issues:
Questions regarding product licensing or serialization Product registration updates, such as address or name changes General product information (features, language availability, local dealers) Latest information about product updates and upgrades Information about upgrade assurance and support contracts Information about the Symantec Buying Programs Advice about Symantec's technical support options Nontechnical presales questions Issues that are related to CD-ROMs or manuals
Contents
Chapter 2
Single-port communication
.............................................. 49
About single port communication ................................................... 49 Configuring single-port communication ........................................... 50 Configuring your Domain Name Server (DNS) and firewall ............ 51
Contents
Adding FQDNs to each service .................................................. Creating a new department with single-port settings .................... Specifying port number 443 as the default port in the configuration file template ................................................ (Conditional) Configuring port 443 in replication policies .............. Installing agent software on the clients or moving clients ..............
53 53 56 57 58
Chapter 3
Chapter 4
Chapter 5
Replication ............................................................................ 77
About replication ......................................................................... About replication and PureDisk release levels ................................... About replication policies .............................................................. Creating or editing a replication policy ............................................. Completing the General tab for a Replication policy ...................... Completing the Data Selections tab for a Replication policy ........... Completing the Scheduling tab for a Replication policy ................. Completing the Parameters tab for a Replication policy ................. Replication jobs ........................................................................... 77 78 79 80 81 82 83 83 85
Contents
Copying and deleting a replication policy .......................................... Managing replicated data selections ................................................ Viewing replicated data ........................................................... Working with replicated agents and data selections ...................... Copying replicated data to clients on the destination storage pool ............................................................................... Restoring replicated data back to clients on the source storage pool ............................................................................... Restoring replicated Oracle data ............................................... Tuning replication .......................................................................
85 86 86 87 87 87 88 88
Chapter 6
10
Contents
Chapter 7
Chapter 8
Contents
11
DR_Restore_all script parameters .................................................. Example 1 ........................................................................... Example 2 ........................................................................... Example 3 ...........................................................................
Chapter 9
Disaster recovery for NetBackup 5000 appliances (deferred implementation) ........................................ 169
About recovering a NetBackup 5000 appliance ................................. Performing a metadata-only disaster recovery restore ...................... Performing a full disaster recovery restore ..................................... Redeploying the NetBackup 5000 ............................................ Reconfiguring the appliance and restoring the data .................... Recovering from corruptions in the databases or in the data store .................................................................................. 169 170 170 171 172 173
Chapter 10
Chapter 11
12
Contents
Statistics tab for a Job Details report ........................................ Files tab for a Job Details report .............................................. Errors tab for a Job Details report ............................................ Job log tab for a Job Details report ........................................... About Data mining reports ........................................................... Enabling a data mining policy ....................................................... Completing the General tab for a data mining policy ................... Completing the Scheduling tab for a data mining policy ............... Completing the Parameters tab for a data mining policy .............. Running a data mining policy manually .......................................... Obtaining data mining policy output - the data mining report ............. Interpreting the storage pool data reduction factor ..................... Effect of compression on data reduction ................................... Effects of segmentation on data reduction ................................ Obtaining data mining policy output - the Web service report ............. Web service reports .................................................................... Job status Web service reports ................................................ Dashboard Web service reports ............................................... Importing report output into a spreadsheet ............................... About Dashboard reports ............................................................. Displaying the Capacity dashboard .......................................... Displaying the Activity dashboard ........................................... Displaying the Server agent dashboard ..................................... Displaying the Client agent dashboard ..................................... Central storage pool authority reports ............................................ Displaying the Central Reporting dashboard .............................. Updating the Central Reporting dashboard ...............................
195 206 207 207 210 211 211 212 212 212 213 215 215 215 216 219 220 223 227 227 227 228 229 230 231 231 236
Chapter 12
Contents
13
Chapter 13
14
Contents
(Optional) Gathering metabase engine capacity information ................................................................... Preparing clients for rerouting ................................................ Preparing the old metabase engine for rerouting ........................ Adding the new metabase engine and recording its address .......... Rerouting the agents on the metabase engine ............................ Restarting the agent ............................................................. Verifying a metabase engine rerouting ..................................... Troubleshooting ................................................................... About clustered storage pool administration ................................... Changing the PDLinux administrators password ............................. Changing the PureDisk internal database and the LDAP administrator passwords ........................................................................... Increasing the number of client connections ................................... Changing a storage pool's time zone .............................................. Adjusting the clock on a PureDisk node .......................................... Adjusting the administrative Web UI timeout interval ....................... Stopping and starting processes on one PureDisk node (unclustered) ....................................................................... Stopping all services ............................................................. Starting all services .............................................................. Starting all services without rebooting ..................................... Stopping and starting individual services .................................. Stopping and starting processes on one PureDisk node (clustered) ........................................................................... Stopping and starting processes in a multinode PureDisk storage pool ................................................................................... Restarting the Java Runtime Environment .....................................
289 290 291 292 292 293 294 294 295 295 295 296 297 298 299 300 300 300 300 301 302 303 305
Chapter 14
Contents
15
Chapter 15
Appendix A
16
Contents
Chapter
About external directory service authentication Obtaining directory service information (Optional) Adding PureDisk groups to your directory service (Optional) Verify TLS and copy the CA certificate Linking PureDisk to the external directory service Enabling the PureDisk system policy that synchronizes PureDisk with an external directory service About maintaining synchronization between PureDisk and an external directory service Adding, changing, or deleting users or groups Changing the youruserclass, yourloginattrib, or yournameattrib variables in your directory services ldap.xml file Changing the yourdescriptionattrib variable or the yourmailattrib variable in your directory services ldap.xml file Disabling external authentication Changing the TLS specification Modifying the base search path
18
OpenLDAP, version 2.3.27 Active Directory, Microsoft Server 2008, R2 Enterprise Active Directory, Microsoft Server 2008, Service Pack 1 Active Directory, Microsoft Server 2003, Service Pack 1 Active Directory, Microsoft Server 2000, Service Pack 4
The following process explains the tasks you need to complete to configure external directory service authentication. To configure external directory service configuration
Obtain directory service information. See Obtaining directory service information on page 20.
(Optional) Add PureDisk groups to your directory service. See (Optional) Adding PureDisk groups to your directory service on page 29.
(Optional) Verify transport layer security (TLS) and copy the certificate authoritys certificate. See (Optional) Verify TLS and copy the CA certificate on page 30.
19
Link PureDisk to the external directory service. See Linking PureDisk to the external directory service on page 34.
Enable the PureDisk policy that synchronizes PureDisk with the directory service. See Enabling the PureDisk system policy that synchronizes PureDisk with an external directory service on page 41.
Assumptions
The procedures you need to perform to configure external authentication assume that you are familiar with how your sites OpenLDAP or Active Directory service is organized. The procedures also assume that your sites directory service administrator can provide you with information about how the directory service is configured.
User accounts
The following information pertains to user accounts when external authentication is enabled:
The Edit LDAP Server Configuration screen in the administrative Web UI includes a checkbox labeled Enable LDAP Authentication. When this box is checked, PureDisk authenticates through an external directory service. When this box is unchecked, PureDisk authenticates through its internal OpenLDAP directory service. You cannot merge these directory services. PureDisk can use either its internal directory service or your external directory service, but it cannot use both at the same time. When PureDisk is configured to authenticate through its internal directory service, only its local user accounts are valid. However, when PureDisk is configured to use an external directory service, only the accounts from that external directory service are valid. If the external service is down, you can authenticate from PureDisks internal OpenLDAP service. For example, the external directory service may be down. If you try to synchronize the external directory service, the job that runs the system policy for syncing external ldap users fails. If you want to add PureDisk users and groups, add them in your directory service and import them into PureDisk. When authentication through an external directory service is enabled, you cannot create users and groups directly in PureDisk. After you import users and groups from the external directory service, you need to grant PureDisk permissions to those users and groups.
20
You cannot import a user with the root logon property from an external directory service. The root users for both PureDisk and for the external directory service are always present and are always unique. By default, the PureDisk root users permissions and privileges are always the same. They remain the same regardless of whether authentication is through PureDisks internal directory service or through an external directory service.
21
Obtain general information from your sites external directory service administrator. The following table summarizes the information that you need to obtain. You can use the right column of the table to make notes about the requirements.
Item needed Directory service host fully qualified domain name (FQDN) (preferred), host name, or IP address. Site-speecific value __________________________________________
Port number to the server that hosts the __________________________________________ directory service The ldapsearch(1) command and its __________________________________________ parameters that PureDisk needs to obtain a listing. Copy of the certificate authority file. __________________________________________
The directory service administrator needs __________________________________________ to put a copy of this file on the storage pool authority. Common name of the certificate authority __________________________________________ file.
(Conditional) Obtain TLS information from your sites external directory service administrator. Perform this step if your site requires TLS. If your site has an Active Directory service, you need to convert the certificate file you receive to PEM format. Instructions later in this process explain when and how to perform the conversion.
Examine a listing from your directory service. Each OpenLDAP or Active Directory service is unique. The directory services themselves, their structures, and their schemas are site-specific and unique to their purpose. For this reason, you must examine the object classes and attributes in your directory service and map them to the PureDisk configuration information screens. The following examples show directory service listings:
22
Save your directory listing. You need to use this listing later in the configuration process.
If you want to add PureDisk user groups to your directory service at this time, proceed to the following topic: See (Optional) Adding PureDisk groups to your directory service on page 29. If TLS is required at your site, proceed to the following topic: See (Optional) Verify TLS and copy the CA certificate on page 30. Otherwise, proceed to the following topic: See Linking PureDisk to the external directory service on page 34.
This directory service has two organizational units: users and groups. You can use the ldapsearch(1) command to obtain a listing of this directory service. The command to obtain a listing of users and groups is as follows:
23
If more directory entries exist in the same directory subtree, a command such as the preceding example returns information about more than users and groups. The command writes its output to file example.txt. In the example file that follows, characters in bold represent definitions from this file that you need later in the configuration process:
# extended LDIF # # LDAPv3 # base <dc=acme,dc=com> with scope subtree # filter: (objectClass=*) # requesting: ALL # # Alice Munro, users, acme.com dn: CN=Alice Munro,OU=users,DC=acme,DC=com objectClass: top objectClass: person objectClass: organizationalPerson objectClass: user objectClass: inetOrgPerson cn: Alice Munro sn: Munro description: Alice's Description givenName: Alice distinguishedName: CN=Alice Munro,OU=users,DC=acme,DC=com displayName: Alice memberOf: CN=chicago,OU=groups,DC=acme,DC=com uSNChanged: 21751 name: Alice Munro sAMAccountName: alice.munro userPrincipalName: alice.munro@acme.com mail: alice.munro@acme.com # Bob Cratchit, users, acme.com dn: CN=Bob Cratchit,OU=users,DC=acme,DC=com objectClass: top objectClass: person objectClass: organizationalPerson # yourmailattrib # yourloginattrib # yourdescriptionattrib # yournameattrib # youruserclass # base search path
24
objectClass: user objectClass: inetOrgPerson cn: Bob Cratchit sn: Cratchit description: Bob's Description givenName: Bob distinguishedName: CN=Bob Cratchit,OU=users,DC=acme,DC=com displayName: Bob Cratchit memberOf: CN=chicago,OU=groups,DC=acme,DC=com name: Bob Cratchit sAMAccountName: bob.cratchit userPrincipalName: bob.cratchit@acme.com mail: bob.cratchit@acme.com # Claire Clairmont, users, acme.com dn: CN=Claire Clairmont,OU=users,DC=acme,DC=com objectClass: top objectClass: person objectClass: organizationalPerson objectClass: user objectClass: inetOrgPerson cn: Claire Clairmont sn: Clairmont description: Claire's Description givenName: Claire distinguishedName: CN=Claire Clairmont,OU=users,DC=acme,DC=com displayName: Claire Clairmont memberOf: CN=atlanta,OU=groups,DC=acme,DC=com name: Claire Clairmont sAMAccountName: claire.clairmont userPrincipalName: claire.clairmont@acme.com mail: claire.clairmont@acme.com # Dave Bowman, users, acme.com dn: CN=Dave Bowman,OU=users,DC=acme,DC=com objectClass: top objectClass: person objectClass: organizationalPerson objectClass: user objectClass: inetOrgPerson cn: Dave Bowman sn: Bowman description: Dave's Description
25
givenName: Dave distinguishedName: CN=Dave Bowman,OU=users,DC=acme,DC=com displayName: Dave Bowman memberOf: CN=atlanta,OU=groups,DC=acme,DC=com name: Dave Bowman sAMAccountName: dave.bowman userPrincipalName: dave.bowman@acme.com mail: dave.bowman@acme.com # chicago, groups, acme.com dn: CN=chicago,OU=groups,DC=acme,DC=com objectClass: top objectClass: group sAMAccountName: chicago member: CN=Bob Cratchit,OU=users,DC=acme,DC=com member: CN=Alice Munro,OU=users,DC=acme,DC=com distinguishedName: CN=chicago,OU=groups,DC=acme,DC=com name: chicago sAMAccountName: chicago # atlanta, groups, acme.com dn: CN=atlanta,OU=groups,DC=acme,DC=com objectClass: top objectClass: group sAMAccountName: atlanta member: CN=Dave Bowman,OU=users,DC=acme,DC=com member: CN=Claire Clairmont,OU=users,DC=acme,DC=com distinguishedName: CN=atlanta,OU=groups,DC=acme,DC=com name: atlanta sAMAccountName: atlanta # search result search: 7 result: 0 Success # numResponses: 7 # numEntries: 6 # yourusergroupclass # yournameattrib # yourmemberattrib
26
Table 1-2
cn=Florence Leeds cn=Mary Evans cn=Diana Goyer cn=Adam Smith cn=Eric Meyer cn=Joe McKinley ou=groups cn=bikers cn=drivers
This directory service has two organizational units: commuters and groups. You can use the ldapsearch(1) command to obtain a listing of this directory service. The command to obtain a listing of the users and groups is as follows:
# ldapsearch -H ldap://100.100.100.100:389/ -x \ -D "cn=Diana Goyer,ou=commuters,dc=marlins,dc=com" -W \ -b "dc=marlins,dc=com" "(objectClass=*)">/tmp/example.txt
This example writes its output to file example.txt. In the example that follows, characters in bold represent the definitions that you need later in the configuration process. The external directory service authentication configuration procedures use examples from this listing. File example.txt is as follows:
# extended LDIF # # LDAPv3 # base <dc=marlins,dc=com> with scope subtree # filter: (objectClass=*) # base search path
27
# requesting: ALL # # marlins.com dn: dc=marlins,dc=com dc: marlins objectClass: domain # commuters, marlins.com dn: ou=commuters,dc=marlins,dc=com ou: commuters objectClass: organizationalUnit # groups, marlins.com dn: ou=groups,dc=marlins,dc=com ou: groups objectClass: organizationalUnit # Florence Leeds, commuters, marlins.com dn: cn=Florence Leeds,ou=commuters,dc=marlins,dc=com mail: Florence.Leeds@marlins.com uid: fleeds objectClass: inetOrgPerson sn: Leeds cn: Florence Leeds userPassword:: cGFzc3dvcmQ= # Mary Evans, commuters, marlins.com dn: cn=Mary Evans,ou=commuters,dc=marlins,dc=com mail: Mary.Evans@marlins.com uid: mevans objectClass: inetOrgPerson sn: Evans cn: Mary Evans userPassword:: cGFzc3dvcmQ= # Diana Goyer, commuters, marlins.com dn: cn=Diana Goyer,ou=commuters,dc=marlins,dc=com mail: Diana.Goyer@marlins.com uid: dgoyer objectClass: inetOrgPerson sn: Goyer cn: Diana Goyer # yournameattrib # yourmailattrib # yourloginattrib # youruserclass
28
userPassword:: cGFzc3dvcmQ= # Adam Smith, commuters, marlins.com dn: cn=Adam Smith,ou=commuters,dc=marlins,dc=com mail: Adam.Smith@marlins.com uid: asmith objectClass: inetOrgPerson sn: Smith cn: Adam Smith userPassword:: cGFzc3dvcmQ= # Eric Meyer, commuters, marlins.com dn: cn=Eric Meyer,ou=commuters,dc=marlins,dc=com mail: Eric.Meyer@marlins.com uid: emeyer objectClass: inetOrgPerson sn: Meyer cn: Eric Meyer userPassword:: cGFzc3dvcmQ= # Joe McKinley, commuters, marlins.com dn: cn=Joe McKinley,ou=commuters,dc=marlins,dc=com mail: Joe.McKinley@marlins.com uid: jmckinley objectClass: inetOrgPerson sn: McKinley cn: Joe McKinley userPassword:: cGFzc3dvcmQ= # bikers, groups, marlins.com dn: cn=bikers,ou=groups,dc=marlins,dc=com objectClass: groupOfNames cn: bikers member: cn=Florence Leeds,ou=commuters,dc=marlins,dc=com member: cn=Mary Evans,ou=commuters,dc=marlins,dc=com member: cn=Diana Goyer,ou=commuters,dc=marlins,dc=com # drivers, groups, marlins.com dn: cn=drivers,ou=groups,dc=marlins,dc=com objectClass: groupOfNames cn: drivers member: cn=Adam Smith,ou=commuters,dc=marlins,dc=com member: cn=Eric Meyer,ou=commuters,dc=marlins,dc=com # yourusergroupclass # yournameattrib # yourmemberattrib
External directory service authentication (Optional) Adding PureDisk groups to your directory service
29
member: cn=Joe McKinley,ou=commuters,dc=marlins,dc=com # search result search: 2 result: 0 Success # numResponses: 12 # numEntries: 11
Edit your directory service and add one or more of the following typical PureDisk user groups:
Administrators. Users with full administrative privileges. Users. Users who can back up and restore data. Reporters. Users who can run reports. Installers. Users who can install PureDisk agents on client systems. Backup operators. Users who can back up files for one storage pool but cannot restore or delete.
The following manual includes information about the different types of permissions that you can grant to users and user groups: See the PureDisk Client Installation Guide.
30
External directory service authentication (Optional) Verify TLS and copy the CA certificate
If TLS is required at your site, proceed to the following topic: See (Optional) Verify TLS and copy the CA certificate on page 30. Otherwise, proceed to the following topic: See Linking PureDisk to the external directory service on page 34.
Verify the server name. See Verifying the server name on page 30.
Write the certificate to the storage pool authority. See Writing the certificate to the storage pool authority on page 31.
Proceed to the following topic: See Linking PureDisk to the external directory service on page 34.
1 2
Obtain the certificates common name from your directory service administrator. Log into the storage pool authority node (unclustered) or storage pool authority service (clustered) as root.
External directory service authentication (Optional) Verify TLS and copy the CA certificate
31
Verify the common name by issuing a ping(8) command against the certificates common name. For example:
# ping mn.north.stars PING mn.north.stars (100.100.100.101) 56(84) bytes of data. 64 bytes from mn.north.stars (100.100.100.101): icmp_seq=1 ttl=64 time=4.71 ms 64 bytes from mn.north.stars (100.100.100.101): icmp_seq=2 ttl=64 time=0.353 ms
(Conditional) Edit the /etc/hosts file. Perform this step only if the previous steps ping(8) command was unsuccessful. Add a line in the following format to the end of the /etc/hosts file:
ip_addr_of_external_directory_services common_name
For example, if the common name of your directory service certificate is mn.north.stars, the line should look like the following:
100.100.100.101 mn.north.stars
Proceed to the following: See Writing the certificate to the storage pool authority on page 31.
1 2
Log into the storage pool authority node (unclustered) or storage pool authority service (clustered) as root. Copy one of the following certificate files to the appropriate directory on the storage pool authority:
The cacert.pem file from an OpenLDAP server. cacert.pem is the default name for this file when an OpenLDAP server generates the file. This file name can be different at your site. The generated certificate from an Active Directory server.
32
External directory service authentication (Optional) Verify TLS and copy the CA certificate
In an unclustered storage pool, Symantec recommends that you write the certificate file to the /var/ldap/certstore/ directory, but you can write this file to any directory. In a clustered storage pool, Symantec recommends that you log into the storage pool authority and write the certificate file to /Storage/var/keys. When you write the certificate file to /Storage, you ensure that the certificate moves with the storage pool authority when a failover occurs.
You can use any file transfer program to copy the files.
(Conditional) Use the openssl(1) command and the x509(1) utility to convert the Active Directory certificate to PEM format. Perform this step only if an Active Directory service generated the certificate. You do not have to perform this step if an OpenLDAP directory service generated the certificate. The openssl(1) command and the x509(1) utility make the file compatible with OpenSSL. Type the following command:
# /usr/bin/openssl x509 -inform DER -outform PEM -in file.cer -out file.pem
file.pem
Use the openssl(1) command and the s_client(1) program to test the port connections and to verify that the SSL certificate operates correctly. Type the following command:
External directory service authentication (Optional) Verify TLS and copy the CA certificate
33
FQDN: port
The directory service servers FQDN and the port from which you imported the certificate. This variable takes the format FQDN:port . Specify the following values:
For FQDN, specify the FQDN of the directory service server. For port, specify the port that the directory service server uses for incoming communication. By default, the value is 636.
For example blink.acme.com:636. cert_loc Specify the absolute path to the certificate file. This file is the one that you copied in step 2. For example: /var/ldap/certstore/mycertfile.pem.
34
External directory service authentication Linking PureDisk to the external directory service
Use the ldapsearch(1) command to test the connection between the storage pool authority and the directory service server. The connection needs to be open to allow continued authentication activities. Type the command as follows: This command has the following format:
For example:
# /usr/bin/ldapsearch -H ldaps://100.100.100.101:636 -x \ -D "cn=Alice Munro,ou=users,dc=acme,dc=com" -W \ -b ou=groups,dc=acme,dc=com "(objectClass=group)"
port
If the connection is open, it displays a listing of directory service server information. If the command fails or generates messages, fix the connectivity problems before you proceed.
Proceed to the following topic: See Linking PureDisk to the external directory service on page 34.
External directory service authentication Linking PureDisk to the external directory service
35
Configure communication in the PureDisk administrative Web UI. See Configuring communication on page 35.
Configuring communication
The following procedure explains how to configure communication in the PureDisk administrative Web UI. To configure communication in the PureDisk administrative Web UI
Display the storage pool authority opening screen. Open a browser window and type the following URL:
https://URL
For URL, specify the URL to the storage pool authority. For example, in an all-in-one environment, this value is the URL of the PureDisk node upon which you installed all the PureDisk software. For example, https://acme.mnbe.com.
2 3 4 5
Type your user name and password at the prompts on the logon screen. Click Settings > Configuration. In the left pane, click the plus (+) sign to the left of LDAP server. Select External LDAP. The LDAP Server Configuration properties appear in the right pane.
Complete the Connection tab. See Completing the Connection tab on page 35.
Complete the Mapping tab. See Completing the Mapping tab on page 38.
Enable user group management. See Managing user groups on page 40.
36
External directory service authentication Linking PureDisk to the external directory service
(Optional) Check the Enable LDAP Authentication box. Perform this step if you want to enable an external OpenLDAP service or an external Active Directory service. A check in this box determines whether PureDisk uses the default PureDisk OpenLDAP directory service or whether PureDisk uses an external directory service, as follows:
If you want to enable PureDisk's internal OpenLDAP directory service, clear this box. By default, this box is clear. For example, if the external directory service goes down, clear this box. When this box is clear, PureDisk uses its internal authentication mechanism, and users can continue to use PureDisk while the external directory service is down. If you want to enable an external OpenLDAP service or an external Active Directory service, check this box.
Specify the LDAP Server Host Name. The content of this field differs depending on whether or not you enabled transport layer security (TLS), as follows:
When TLS is enable, specify the FQDN of the external OpenLDAP server. Note that this FQDN must match the common name that you specified when you created the certificate of your external OpenLDAP server. If you do not have this information yet, obtain this name from your directory service administrator. When TLS is not enabled, specify the FQDN or the IP address of the server upon which your external OpenLDAP or Active Directory service resides. If you want to enable single port communication between the nodes and clients in your PureDisk environment, Symantec recommends that you use an FQDN for this field. For example: blinkie.acme.com.
Verify, and respecify if needed, the Port number that connects PureDisk to the external OpenLDAP or Active Directory service. The port to specify depends on whether TLS is enabled, as follows:
When TLS is enabled, the default security port is 636. When TLS is not enabled, the default port is 389.
Check or clear the Enable TLS for LDAP Communication box. A check in this box determines whether TLS is enabled. TLS encrypts data transactions between your external directory service and PureDisk.
External directory service authentication Linking PureDisk to the external directory service
37
(Conditional) Specify the CA Certificate Location. Perform this step if you checked Enable TLS for LDAP Communication. Specify the full path to the file that contains the certificate authority that PureDisk can use to verify the connection to the external directory service server. That is the file you copied over from the external directory service server to the storage pool authority, and typically, this file resides in /var/ldap/certstore/. Specify the location you configured in the following procedure: See (Optional) Verify TLS and copy the CA certificate on page 30. This file must reside on the storage pool authority in PEM format.
Specify the User Distinguished Name. The distinguished name of an external directory service administrator account. This administrator account does not need to have full privileges on the directory service server. PureDisk requires only the search privilege. Contact your sites directory service administrator to obtain this value.
In the Password field, specify the password for the User Distinguished Name's account. Contact your site's directory service administrator to obtain this password.
8 9
In the Service Type pull-down menu, select either Active Directory or OpenLDAP. Specify the external directory service's Base Search Path. This path specifies a directory in the external directory service beneath which PureDisk beneath which PureDisk can find the entries that define all user groups. Type the distinguished name into this field. Specify the highest entry that is a common ancestor of the groups that you need. Note that in a later procedure you specify the individual groups that define the users that you want to authenticate as PureDisk users. This later procedure is as follows: See To manage user groups on page 41. For example, assume that you want PureDisk to use the groups from the OpenLDAP marlins.com directory service. To specify all groups, examine the ldapsearch(1) command output and complete the following steps:
Type the following into the box under the Base Search path label:
dc=marlins,dc=com
Click Add.
38
External directory service authentication Linking PureDisk to the external directory service
1 2
Click the Mapping tab. Gather the directory service listing that you obtained when you completed the following procedure: See Obtaining directory service information on page 20. If you have not yet obtained a listing, generate one now. Keep this output available to you. You might need the information in the listing to complete the Mapping tab. For example, for a small configuration, obtain a copy of the ldapsearch(1) output. For a large configuration, you can use a directory browser such as the one at the following location: http://www.anl.gov/techtransfer/Software_Shop/LDAP/LDAP.html The Symantec Corporation does not endorse, guarantee, or recommend any particular LDAP browser.
External directory service authentication Linking PureDisk to the external directory service
39
Complete the fields under the User attributes heading. The following table explains how to complete these fields.
User object class Specify the attribute that defines the object class for users on your external directory service. Active Directory example: user. OpenLDAP example: inetOrgPerson. Login attribute Specify the attribute that uniquely identifies a user. The user ID for each user is unique. Active Directory example: sAMAccountName. OPenLDAP example: uid. Name attribute Specify the attribute that defines a user's real name. The Name attribute is the login ID of a user. Active Directory example: cn. OpenLDAP example: cn. Description attribute (Optional) Specify the attribute for the descriptive field. Some directory services do not have this attribute. Active Directory example: description. OpenLDAP example: description. Email address attribute (Optional) Specify the attribute that describes the user's email address. Active Directory example: mail. OpenLDAP example: mail.
40
External directory service authentication Linking PureDisk to the external directory service
Complete the fields under the Group attributes heading. The following table explains how to complete these fields.
Group object class Specify the attribute that defines the object class for groups on your external directory service. Active Directory example: group. OpenLDAP example: groupOfNames. Membership attribute Specify the attribute that defines each member in the group. Active Directory example: member. OpenLDAP example: member. Name attribute Specify the attribute that defines a user group. Active Directory example: sAMAccountName. OpenLDAP example: cn.
5 6
Click Save. (Conditional) On the storage pool authority, edit the /etc/hosts file. Add a line that allows the storage pool authority service to resolve the directory server using the name of this certificate. Perform this step if you enabled TLS. For example, assume that the common name of the certificate file is blinkie.acme.com. Note that this string is not the FQDN of the server upon which the directory resides. Add the following entry:
100.100.100.101 blinkie.acme.com
External directory service authentication Enabling the PureDisk system policy that synchronizes PureDisk with an external directory service
41
1 2 3 4 5
Click Settings > Configuration. In the left pane, click the plus (+) sign to the left of LDAP Server. Select External LDAP. In the right pane, click Manage User Groups. Complete the Manage User Groups pane. In this screen, complete the following steps:
In the left-most box, under the Distinguished Name label, type the distinguished name of a user group into this field. For example: CN=atlanta,OU=groups,DC=acme,DC=com or cn=bikers,ou=groups,dc=marlins,dc=com. Click Add. Repeat the preceding steps to add distinguished names for all the user groups that PureDisk needs to authenticate. Click Save. If the external directory service is down at the time you click Save, PureDisk generates the following message:
Error: Invalid Search Path
Proceed to the following: See Enabling the PureDisk system policy that synchronizes PureDisk with an external directory service on page 41.
Enabling the PureDisk system policy that synchronizes PureDisk with an external directory service
PureDisk lets you create several different policies for backups, data removal, and many other purposes. However, for other tasks you only need one generic policy. PureDisk includes several system policies for these generic tasks. In most cases, you need to enable a system policy before PureDisk can run the policy. The procedure in this topic describes how to enable the system policy that can synchronize the directory services. This synchronization ensures that PureDisk recognizes the users and the groups that you add to the external directory service.
42
External directory service authentication Enabling the PureDisk system policy that synchronizes PureDisk with an external directory service
During the synchronization, PureDisk does not synchronize user passwords. The passwords reside only in the external directory service files. The PureDisk administrative Web UI does not accept empty, bank passwords. Make sure that all users you want to authenticate through an external directory service have a nonblank password. A user with a blank password cannot log into the PureDisk administrative Web UI. To enable the system policy for synchronization
1 2 3 4
Click Manage > Policies. In the left pane, under Miscellaneous Workflows, click the plus sign (+) next to External LDAP server synchronization. Select System policy for Syncing external LDAP users. Complete the General tab. Note: The General tab and the Scheduling tab include a Save option on each tab. Do not click Save until you complete the fields on each tab. If you click Save before you complete each tab, it saves the specifications you made up to that point and closes the dialog box. To complete more fields, open the dialog again in edit mode. See Completing the General tab on page 42.
Complete the Scheduling tab. See Completing the Scheduling tab on page 43.
(Optional) Type a new name for this policy in the Name field. You do not need to rename this policy. The default name is System policy for Syncing external LDAP users.
Select Enabled or Disabled. This setting lets you control whether or not PureDisk runs the policy according to the schedule you specify in the Scheduling tab.
If you select Enabled, PureDisk runs the policy according to the schedule. If you select Disabled, PureDisk does not run the policy according to the schedule. Disabled is the default.
External directory service authentication About maintaining synchronization between PureDisk and an external directory service
43
For example, you might want to stop running this policy during a system maintenance period. If you select Disabled, you do not need to enter information in the Scheduling tab to suspend and then re-enable this policy.
Specify an escalation action. PureDisk can notify you if replication does not complete within a specified time. For example, you can configure PureDisk to send an email message to an individual if a replication policy workflow does not complete in an hour. Select times in the Escalate warning after or the Escalate error and terminate after drop-down boxes to specify the elapsed time. If the policy workflow does not complete before the warning timeout, PureDisk generates a warning event. If the policy workflow does not complete before the error timeout, PureDisk generates an error event. If you select either option, you must create a policy and an event escalation action. These escalation actions define the email message, define its recipients, and associate the escalation action with the policy. For information about policy and event escalations, see the following manual: See the PureDisk Backup Operators Guide.
Proceed to the following: See Completing the Scheduling tab on page 43.
1 2
Accept the system defaults or use the drop-down lists to specify when the policy is to run. Click Save.
44
See Enabling the PureDisk system policy that synchronizes PureDisk with an external directory service on page 41. The following topics describe other changes you might need to make to your authentication configuration: See Adding, changing, or deleting users or groups on page 44. See Changing the youruserclass, yourloginattrib, or yournameattrib variables in your directory services ldap.xml file on page 45.
You can use the same procedure to add or delete user groups. The procedure is as follows: See Managing user groups on page 40.
External directory service authentication Changing the youruserclass, yourloginattrib, or yournameattrib variables in your directory services ldap.xml file
45
Changing the youruserclass, yourloginattrib, or yournameattrib variables in your directory services ldap.xml file
If you change the youruserclass, yourloginattrib, or yournameattrib variables in your directory service, you need to update PureDisks /etc/puredisk/ldap.xml file. To change the youruserclass, yourloginattrib, or yournameattrib variables
1 2 3 4 5 6 7 8
Use your directory services methods to change the object class names in your directory service source files. Log into the storage pool authority node as root. Edit file /etc/puredisk/ldap.xml to reflect the changes that you made to your directory service. Log into the PureDisk administrative Web UI. Click Manage > Policies. Under Miscellaneous Workflows, click the plus sign (+) next to External LDAP server synchronization. Select System policy for Syncing external LDAP users. In the right pane, click Run policy.
Changing the yourdescriptionattrib variable or the yourmailattrib variable in your directory services ldap.xml file
If you change the yourdescriptionattrib variable or the yourmailattrib variable in your directory service, you need to update PureDisks /etc/puredisk/ldap.xml file. To change the yourdescriptionattrib variable or yourmailattrib variable
1 2 3
Use your directory services methods to change the object class names in your directory service source files. Log into the storage pool authority node as root. Edit file /etc/puredisk/ldap.xml to reflect the changes that you made to your directory service.
46
4 5 6 7 8 9
Log into the PureDisk administrative Web UI. Click Settings > Configuration tab. In the left pane, expand LDAP Server. Select External LDAP. In the right pane, click Manage User Groups. Select each group.
10 Click Remove. 11 Click Save. 12 Run the system policy for synchronizing external directory service users.
Complete the following steps:
Click Manage > Policies. In the left pane, under Miscellaneous Workflows, click the plus sign (+) next to External LDAP server synchronization. Select System policy for Syncing external LDAP users. In the left pane, click Run policy.
13 Use the procedure in the following topic to add the user groups back:
See Managing user groups on page 40.
14 Run the system policy for synchronizing external directory service users
again. The instructions for how to run this policy are in step 12.
1 2
Click Settings > Configuration. In the left pane, click the plus (+) sign to the left of LDAP Server.
47
3 4
Select External LDAP. In the right pane, clear the Enable LDAP Authentication box. When this box is clear, PureDisk uses its internal directory service to authenticate users.
5 6
Click Save. Run the system policy for synchronizing external directory service users. Complete the following steps:
Click Manage > Policies. In the left pane, under Miscellaneous Workflows, click the plus sign (+) next to External LDAP server synchronization. Select System policy for Syncing external LDAP users. In the right pane, click Run policy.
1 2 3 4
Click Settings > Configuration. In the left pane, click the plus (+) sign to the left of LDAP server. Select External LDAP. Change the specification, as follows:
To enable TLS, check the Enable TLS for LDAP Communication box. To disable TLS, clear the Enable TLS for LDAP Communication box.
5 6 7
Click Save. Log on to the storage pool authority as root. Type the following command to restart pdweb:
# /etc/init.d/puredisk restart pdweb
48
1 2 3 4 5 6
Click Settings > Configuration. In the left pane, click the plus (+) sign to the left of LDAP server. Select External LDAP. In the right pane, in the Base Search Path field, change the search path. Click Save. Run the system policy for synchronizing external directory service users. Complete the following steps:
Click Manage > Policies tab. In the left pane, under Miscellaneous Workflows, click the plus sign (+) next to External LDAP server synchronization. Select System policy for Syncing external LDAP users. In the right pane, click Run policy.
Chapter
Single-port communication
This chapter includes the following topics:
50
Content router 1 Controller 1 Metabase engine 1 Metabase server NetBackup export engine Storage pool authority
IP-inx is an IP address behind the firewall and in a private range. For example, 100.100.100.100 through 100.100.100.024. IP-outx is an IP address outside the firewall somewhere on the Internet.
Analyze the ports currently used and configure the firewall. See Configuring your Domain Name Server (DNS) and firewall on page 51.
Add an FQDN to each PureDisk service. See Adding FQDNs to each service on page 53.
51
Create the department for which you want to use single-port communication. See Creating a new department with single-port settings on page 53.
Edit the configuration and specify the single-port number. See Specifying port number 443 as the default port in the configuration file template on page 56.
(Conditional) Configure the single port in replication policies. Perform this step if other storage pools replicate data to this storage pool. See (Conditional) Configuring port 443 in replication policies on page 57.
Assign clients to the new department that uses single ports. See Installing agent software on the clients or moving clients on page 58.
On your DNS, configure the ports that the storage pool uses. Use your firewall softwares documentation to help you determine the inbound ports and outbound ports that are in use at this time. Perform the following tasks:
Determine which services are behind the firewall. Make sure that the DNS can resolve the correct IP address regardless of whether the client is on the Internet or behind the firewall.
PureDiskCR1.acme.com IP-out3 PureDiskCR2.acme.com IP-out4 PureDiskCTRL1.acme.com IP-out5 PureDiskCTRL.acme.com2 IP-out6 PureDiskSPA.acme.com IP-out2 PureDiskMBS.acme.com IP-out2
52
PureDiskDebug.acme.com IP-out7
IP-in1
Configure the firewall to translate the IP addresses and outside ports to inside addresses and inside ports. Use your firewall softwares documentation to help you translate the ports. In the example storage pool, the translations are as follows for the outside ports:
Outside IP IP-out2 IP-out3 IP-out4 IP-out5 IP-out6 IP-out7 Outside port 443 443 443 443 443 443
In the example storage pool, the translations are as follows for the inside ports:
Inside IP IP-in1 IP-in1 IP-in2 IP-in1 IP-in2 IP-in1 Inside port 443 10082 10082 10101 10101 10087
Proceed to the following: See Adding FQDNs to each service on page 53.
53
1 2 3
Click Settings > Topology. Expand all the items in the left pane so that PureDisk displays all storage pool services. Select a service and add the FQDN information. For example, select the first content router that appears and complete the following steps:
Type the FQDN for this particular service into the Host Name (FQDN) field. This field accepts a host name, but to enable this feature, specify the FQDN for this particular service. Click Save.
Repeat step 3 for each content router, metabase engine, metabase server, storage pool authority, and NetBackup export engine service on every node. Do not perform these steps for the metabase engines. For the metabase engines, PureDisk updates the FQDN information automatically when you update the storage pool authority and the controller, respectively.
Proceed to the following: See Creating a new department with single-port settings on page 53.
54
To implement the single-port feature for a new installation OR To implement the single-port feature for an existing installation in which the clients you want to enable as single-port clients reside in different departments Perform the following tasks:
Create a new department. See (Conditional) Creating a new department on page 54. Create a new configuration template. See Creating a new configuration template on page 55. Assign the new template to the new department. See Assigning a new configuration template to a new department on page 55.
To implement the single-port feature in an existing storage pool when the clients to be enabled as single-port clients reside in the same department, perform the following tasks:
Create a new configuration template. See Creating a new configuration template on page 55. Assign the new template to the new department. See Assigning a new configuration template to a new department on page 55.
To implement the single-port feature for a new installation To implement the single-port feature for an existing installation in which the clients you want to enable as single-port clients reside in different departments
1 2 3 4 5
Click Settings > Configuration. In the left pane, expand User management > Departments. Select Departments. In the right pane, click Add Department. In the Department Name field, type in a name for this new department. For example, assume that you want to name this department for single port use. You can call this department ATOP Dept, which means All Through One Port Department.
55
6 7
Click Add. Proceed to the following: See Creating a new configuration template on page 55.
1 2 3 4 5 6
Click Settings > Configuration. In the left pane, expand Configuration file templates > PureDisk Client Agent. Select Default Value Set for PureDisk Client Agent. In the right pane, click Copy Value Set. In the left pane, select Copy of Default Value Set for PureDisk Client Agent. In the right pane, in the Value Set Name (new) field, type a new name. For example, ATOP Template.
7 8
Click Save. Proceed to the following: See Assigning a new configuration template to a new department on page 55.
1 2 3
Click Settings > Configuration. In the left pane, expand Configuration file templates > PureDisk Client Agent. Select the new template you created in the following procedure: See Creating a new configuration template on page 55. For example, select ATOP Template.
4 5
In the right pane, click Assign Template. Click the box to the left of the name of the new department. For example, click to the left of ATOP Dept.
56
6 7
Click Assign. Proceed to the following: See Specifying port number 443 as the default port in the configuration file template on page 56.
Specifying port number 443 as the default port in the configuration file template
The following procedure explains the configuration sections that you need to edit to specify that PureDisk use single port 443 for client communication. This feature supports single-port communication only through port 443. After you specify port 443, you can install the agent software on your clients. When you install the agent software, specify the new department that you created that uses the single port. These new clients assume the specifications in the new configuration template that you assigned to the new department. To specify port 443 as the default port
1 2 3
Click Settings > Configuration. In the left pane, expand Configuration file templates > PureDisk Client Agent. Expand the new template that you created in the following procedure: See Creating a new configuration template on page 55. For example, expand ATOP Template.
5 6 7 8
Under the expanded contentrouter > port, select All OS: portnumber. In the Value field, type 443. Click Save. Repeat the following steps to change the port value to 443 for the ctrl > tcport field and the debug > dldport field: Step 6 Step 7
57
(Conditional) Configure port 443 in replication policies See (Conditional) Configuring port 443 in replication policies on page 57. Install the agent software on the clients or move the clients. See Installing agent software on the clients or moving clients on page 58.
Log into the storage pool authority's administrative Web UI on the storage pool that replicates data. For example, if you replicate data from storage pool Remote to storage pool Central, log into storage pool Remote.
2 3 4
Click Manage > Policies. In the left pane, expand the tree to expose the replication policies that send data to the other storage pool. Complete the following steps to edit a replication policy:
Select a replication policy. Click the Parameters tab. Type 443 into the Communication Port field. Click Save.
5 6
Repeat step 3 for each replication policy. Proceed to the following: See Installing agent software on the clients or moving clients on page 58.
58
Use the procedures in the following manual to install the PureDisk agent software on each new client: See the PureDisk Client Installation Guide. When the software prompts you for a department name, specify the name of the new department you created that uses the single port.
1 2 3
Click Manage > Agent. In the left pane, expand the tree to expose the clients that you need to move. Complete the following steps to edit each client and add the new department:
Select a client. In the right pane, in the Properties: Agent panel, use the Department drop-down list to select the department you created for single-port communication. Click Save.
Repeat step 3 for each client that you need to move to the new department.
Chapter
Applying patches
This chapter includes the following topics:
About applying patches and updating PureDisk software Preparing the storage pool and obtaining the software Downloading and verifying a patch Uploading a patch Installing a patch Committing a patch Rolling back a patch Removing a software patch About the Software Details page About checkpoints
EEBs. Hotfixes . For example, PureDisk 6.6.1.1. Release updates. For example, PureDisk 6.6.1.
The PureDisk 6.6.1 release includes a feature that lets you install patches either by using commands or by using the storage pool configuration wizard. Patches
60
Applying patches Preparing the storage pool and obtaining the software
that require you to provide information at installation time can be installed only by using commands. However, many patches can be installed through the wizard interface. The patch descriptions contain information about specific installation options. In order for you to use the storage pool configuration wizard to install patches, your storage pool must use the VxFS file system. The following contain information about how to install PureDisk software patches:
See Preparing the storage pool and obtaining the software on page 60. See Downloading and verifying a patch on page 61. See Uploading a patch on page 62. See Installing a patch on page 63. See Committing a patch on page 64. See Rolling back a patch on page 65. See Removing a software patch on page 65. See About the Software Details page on page 66. See About checkpoints on page 66.
Log into the PureDisk administrative Web UI and make sure that no PureDisk jobs are running or are scheduled to run. Stop the jobs that are running currently. Note: The patch installation process stops all jobs that are running. If you want to stop any jobs gracefully, make sure to stop them before you proceed. For information about how to log into the administrative Web UI and how to stop jobs, see the following: See the PureDisk Backup Operator's Guide.
61
(Conditional) Use the Cluster Manager Java Console to freeze the PureDisk service groups. Perform this step if the storage pool is enabled for high availability. For information about how to freeze and unfreeze clustered storage pools, see the Symantec NetBackup PureDisk Storage Pool Installation Guide.
Proceed to the following: See Downloading and verifying a patch on page 61.
Download the software patch you want to install. You can download patches in one of the following ways:
If Symantec notified you that an EEB or other patch is available, follow the instructions in that notification to download the software. If you want to download a hotfix or a release update, you can download these patches from the following Web site: http://www.symantec.com/business/support/overview.jsp?pid=52672
Example 1. You can use WinSCP or an FTP program to download the patch to a Windows laptop computer or to a Windows desktop computer. Later, you can use the storage pool configuration wizard to browse for the downloaded file. Example 2. You can log into the storage pool authority node as root and use ftp(1) to get the patch from Symantec.
(Conditional) Move the downloaded file to the computer from which you typically start the storage pool configuration wizard. For example, you can use scp(1) to move the file. Alternatively, if you used a Windows system to download the patch, and you typically start the configuration wizard from that same Windows system, you can use the wizard to browse for the patch on the Windows system.
62
For tarname, type the file name of the downloaded patch. Include the .tar file suffix. This command computes the MD5 checksum of patchname. The MD5 checksum of the patch must match the MD5 checksum in the patch's README file on the following Web site: http://entsupport.symantec.com/docs/347536 If the checksums do not match, download the patch again.
Uploading a patch
The following procedure explains how to upload the patch to your storage pool authority. To start the storage pool configuration wizard and upload the patch
1 2 3 4
Log into the storage pool configuration wizard as root. On the storage pool configuration wizard's Welcome page, click Manage Patches. On the Upgrade and Roll Back page, click the Available Software tab. Specify a patch to upload. Use one of the following methods:
Click Browse, and select the patch you want to install from your browser's upload menu.. For example, navigate to the C: drive and select a patch. Type the path to the patch you want into the Choose file to upload: field.
5 6
Verify that the file in the Choose file to upload field is correct and click OK. Proceed to the following to install the patches you downloaded: See Installing a patch on page 63.
63
Installing a patch
The following procedure explains how to install PureDisk software patches. To install PureDisk software patches
Complete the Upgrade and Roll Back - Patch List page. Perform the following steps on this page:
Click the Available Software tab, and verify that the patch you want to install appears in the list of patches. If the patch you want to install does not appear, perform the following procedures:
See Downloading and verifying a patch on page 61. See Uploading a patch on page 62.
In the Patch Version column, click the patch number of the patch you want. Read the Software Details pop-up page for this patch, and verify the prerequisites. The Software Details page explains the patch purpose, explains the rollback possibilities for this patch, and contains other information. Verify that your storage pool meets the patch prerequisites, and click Close. For more information about the Software Details page, see the following: See About the Software Details page on page 66. Click Install.
Complete the Upgrade and Roll Back - Installation Options page. Perform the following steps on this page:
(Optional) Select installation options. Depending on the nature of the patch you want to install, the Upgrade and Roll Back - Installation Options page displays installation options. Select one or more of the following possible options:
Create a complete checkpoint for this patch Most patches let you create a system checkpoint before the installation starts. A checkpoint is a record of your storage pool at a specific time. Later, if the patch has a negative effect on your storage pool, you can remove the patch and reinstate the checkpoint. For information about checkpoints, see the following: See About checkpoints on page 66. Upgrade agents automatically after patch installation completes
64
Some patches let you specify whether or not to upgrade the agents automatically. Other patches upgrade the agents without asking you to confirm this action.
Complete the Upgrade and Roll Back - Installation Progress page. Perform the following steps on this page:
Wait for the installation to complete. Click Go To Installed Software or Close. If you click Go To Installed Software, the wizard returns to the Upgrade and Rollback - Patch List page. If you click Close, the upgrade and roll back pages close. The configuration wizard remains open in another window.
Enter a test period and commit the patch after the test period ends. See Committing a patch on page 64. Roll back the patch. See Rolling back a patch on page 65.
If you did not create a checkpoint, you can commit or remove the patch. See Committing a patch on page 64. See Removing a software patch on page 65.
Committing a patch
The following procedure explains how to commit a patch. To commit a patch
1 2 3 4
Log into the storage pool configuration wizard as root. On the storage pool configuration wizard's Welcome page, click Manage Patches. On the Upgrade and Roll Back page, click the Installed Software tab. Click Commit.
65
1 2 3 4 5 6 7
Log into the storage pool configuration wizard as root. On the storage pool configuration wizard's Welcome page, click Manage Patches. On the Upgrade and Roll Back - Patch List page, click the Installed Software tab. Click Roll back. On the pop-up that appears, click OK to confirm that you want to perform a rollback. Wait for the rollback to complete. On the Upgrade and Roll Back - Rollback Progress page, click Go To Installed Software or Close. If you click Go To Installed Software, the wizard returns to the Upgrade and Rollback - Patch List page. If you click Close, the upgrade and roll back pages close. The configuration wizard remains open in another window.
1 2 3 4
Log into the storage pool configuration wizard as root. On the storage pool configuration wizard's Welcome page, click Manage Patches. On the Upgrade and Roll Back - Patch List page, click the Available Software tab. Click Remove.
66
5 6
On the pop-up that appears, click OK to confirm that you want to remove the patch. Click Close.
The purpose of the patch. Fixes included in the patch. System prerequisites. This statement describes the required base release and lists incompatibilities with other patches. The patch's installation options. You can use the storage pool configuration wizard to install most patches. Some patches cannot be installed through the storage pool configuration wizard. These patches need to be installed by typing commands. Rollback options. The Software Details page states whether the patch can be rolled back. If you used the storage pool configuration wizard to install a patch, you can always use the storage pool configuration wizard to roll back the patch.
About checkpoints
PureDisk considers a storage pool to be in checkpoint mode if you specified that you wanted to create a checkpoint during the installation of a patch. You can have more than one checkpoint on your system at a time. After you commit a patch, PureDisk removes the checkpoint for the patch. If one or more checkpoints exist, the PureDisk administrative Web UI displays the message Storage pool is in checkpoint mode. The storage pool configuration wizard also displays a checkpoint mode message. Depending on how full your system is, you might be able to leave a storage pool in checkpoint mode for a few days. The length of time also depends on the change rate of the clients being backed up. Garbage collection does not occur when a storage pool is in checkpoint mode. Monitor your system closely and remove the checkpoint as soon as you can. Do not attempt to remove a checkpoint manually. Always use the storage pool configuration wizard to remove checkpoints. The following contain more information about checkpoints:
67
See Managing multiple checkpoints on page 67. See Storage pool functionality while in checkpoint mode on page 67.
You can start the storage pool configuration wizard, but its available functions are severely reduced. You can manage patches, but you cannot perform other functions. For example, you cannot add a new node while the storage pool is in checkpoint mode because the addition would invalidate the checkpoint. To prevent performance degradation and data loss, PureDisk prevents certain storage pool operations from occurring when the storage pool is in checkpoint mode. PureDisk deduplication option (PDDO) operations, rerouting, replication, disaster recovery restores, and garbage collection cannot run when the storage pool is in checkpoint mode.
68
Chapter
About dedicated networks within a storage pool Configuring dedicated networks Configuring a dedicated replication network Configuring a dedicated backup and restore network Configuring addressing
See Configuring dedicated networks on page 70. See Configuring addressing on page 73.
70
1 2
Make sure that PureDisk 6.6.1 is installed correctly and is known to be operational. Make sure that additional network interface cards (NICs) are installed in the appropriate computers. The additional NICs you need are as follows:
An additional NIC in the storage pool authority node. Install an additional NIC in the storage pool authority node when you configure a dedicated network for backups, restores, or replication. If you want to enable a dedicated network for replication, make sure to install the additional NICs in both the source storage pool authority node and in the destination storage pool authority node. An additional NIC in each client that you want to include on the dedicated backup or restore network. You can install additional NICs in one or more clients. When you install the NICs in clients, make sure that the IP address on the additional NIC is on the same subnet as the additional NIC in the storage pool authority node. To configure an additional IP address on a client, follow the instructions for that client's operating system. (Conditional) An additional NIC on the NetBackup media server. If you have the PureDisk deduplication option (PDDO) enabled, install an additional NIC in the NetBackup media server. This additional NIC can increase the performance of PDDO backups and restores. Make sure all NICs on the backup network are on the same subnet. To configure an additional IP address on a NetBackup media server, follow the instructions for that media server's operating system. (Optional) An additional NIC in the metabase engine node and in the content router node. For best performance in multinode storage pools, install additional NICs in the metabase engine nodes and in the content router nodes. Make sure that the additional NICs in the metabase engine nodes and content router nodes are on the same subnet as the dedicated NIC in the storage pool authority node.
Many computers come equipped with additional NICs. If you need to install additional NICs, use the NIC manufacturer's instructions to perform the NIC
Networks for backups, restores, and replication Configuring a dedicated replication network
71
installation. To configure an IP address on a NIC in a storage pool authority node, see the following topic: See Configuring addressing on page 73.
See Configuring a dedicated replication network on page 71. See Configuring a dedicated backup and restore network on page 72.
Start the storage pool configuration wizard on the source storage pool. Type the following URL into a browser window:
http://addr/Installer
For addr, type the IP address or FQDN of the storage pool authority node. Type your site's user name and password, and click Log on.
2 3
On the Welcome page, click Specify networks. On the Specify networks page, in the Networks for replication field, type the subnet(s) that you configured for replication. If you configured more than one subnet, you can specify a range of subnet(s). PureDisk can detect a correctly configured NIC. The configuration wizard displays the subnet(s) that it detects on the Specify networks page. Specify the subnet(s) as the wizard shows them. If the configuration wizard does not detect the subnet(s) you want to use, specify the subnet(s) manually. Use classless inter-domain routing (CIDR) notation in your specification, as follows:
If the subnets you want to use are 10.20.30.01/24 and 10.20.30.02/24, specify 10.20.30.0/24 in the Networks for replication field. If you have more than one subnet, specify a comma-separated list of subnets. For example, specify 10.20.30.0/21,10.20.30.0/24
4 5 6
Click Save. Start the storage pool configuration wizard on the destination storage pool. On the Welcome page, click Specify networks.
72
Networks for backups, restores, and replication Configuring a dedicated backup and restore network
On the Specify networks page, specify the replication subnet in both the Networks for backup/restore field an in the Networks for replication field. On the destination storage pool, PureDisk interprets the replication job as a backup from another storage pool. For this reason, you need to specify the subnet(s) in both fields of the Specify networks page on the destination storage pool. Specify the same subnet(s) you specified on the Specify networks page for the source storage pool, but on the destination storage pool specify the subnet(s) in both the Networks for backup/restore field and in the Networks for replication field. For information about how to specify the subnet data in the format that is accepted in these fields, see the following step: Step 3
Click Save.
Start the storage pool configuration wizard. Type the following URL into a browser window:
http://addr/Installer
For addr, type the IP address or FQDN of the storage pool authority node. Type your site's user name and password, and click Log on.
2 3
On the Welcome page, click Specify networks. On the Specify networks page, in the Networks for backup/restore field, type the subnet(s) that you configured for backups and restores. If you configured more than one subnet, you can specify a range of subnets or a comma-separated list of subnets. PureDisk can detect a correctly configured NIC. The configuration wizard displays subnet(s) that it detects on the Specify networks page. Specify the subnet(s) as the wizard shows them. If the configuration wizard does not detect the subnet(s) you want to use, specify the subnet(s) manually. Use classless inter-domain routing (CIDR) notation in your specification, as follows:
73
If the subnets you want to use are 10.20.30.01/24 and 10.20.30.02/24, specify 10.20.30.0/24 in the Networks for replication field. If you have more than one subnet, specify a comma-separated list of subnets. For example, specify 10.20.30.0/21,10.20.30.0/24
Click Save.
Configuring addressing
This procedure explains how to configure static addressing on a network interface card in a PureDisk node. PureDisk does not support dynamic addressing. PureDisk supports Dynamic Host Configuration Protocol (DHCP) only if resolves to a static address. This procedure assumes that you do not use a DHCP server. If your DHCP server is configured to use DHCP reservations, use this procedure to verify your configuration settings. Note: Because of a PDLinux limitation, you cannot combine static addressing with DHCP. When you use static addressing, the Update Name Servers and search list via DHCP option is grayed out in the Hostname and Name Server Configuration page. If the Update Name Servers and search list via DHCP option is grayed out while it is selected, you cannot disable it. The procedures in this topic guide you through the configuration process in a way that avoids this problem. The following topics explain how to configure addressing:
See Gathering NIC information on page 73. See Configuring the IP address on a NIC on page 75.
Gather information about the additional NIC(s) in this computer. Contact your network administrator for the information in the following table:
Information needed NIC data
74
________________________________________
The IP address of the DNS server or name ________________________________________ server. The domain for this network. The IP address of the default gateway. ________________________________________ ________________________________________
You can obtain some of the information you need from the ip a(8) command. The following is example ip a(8) command output:
# ip a 1: lo: <LOOPBACK,UP> mtu 16436 qdisc noqueue link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00 inet 127.0.0.1/8 scope host lo inet6 ::1/128 scope host valid_lft forever preferred_lft forever 2: eth0: <BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast qlen 100 link/ether 00:11:43:e4:0b:2a brd ff:ff:ff:ff:ff:ff inet 10.80.92.102/21 brd 10.80.95.255 scope global eth0 inet6 fec0::80:211:43ff:fee4:b2a/64 scope site dynamic valid_lft 2591999sec preferred_lft 604799sec inet6 fe80::211:43ff:fee4:b2a/64 scope link valid_lft forever preferred_lft forever 3: eth1: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast qlen 1000 link/ether 00:11:43:e4:0b:2b brd ff:ff:ff:ff:ff:ff 4: eth2: <BROADCAST,MULTICAST> mtu 1500 qdisc noop qlen 1000 link/ether 00:04:23:b0:4f:86 brd ff:ff:ff:ff:ff:ff 5: eth3: <BROADCAST,MULTICAST> mtu 1500 qdisc noop qlen 1000
75
link/ether 00:04:23:b0:4f:87 brd ff:ff:ff:ff:ff:ff 6: sit0: <NOARP> mtu 1480 qdisc noop link/sit 0.0.0.0 brd 0.0.0.0
Proceed to the following topic to configure static addressing: See Configuring the IP address on a NIC on page 75.
Type the following command to launch the SUSE Linux YaST configuration tool:
# yast
You can type yast or YaST to start the interface. Do not type other combinations of uppercase and lowercase letters.
2 3 4
In the YaST Control Center main page, select Network Devices > Network Card and press Enter. On the Network Setup Method page, select Traditional Method with ifup and select Next. On the Network Card Configuration Overview page, select an unconfigured NIC. In the bottom pane, the Device Name field identifies each NIC by its MAC address.
5 6
Tab to Hostname and Name Server and press Enter. (Conditional) On the pop-up that appears, select Modify, and press Enter. Perform this step if the pop-up appears.
On the Hostname and Name Server Configuration page, perform the following steps:
In the Hostname field, type a unique host name for this NIC. The default PDLinux host name is PureDisk, but each node must have a unique host name.
76
Press Tab to position the cursor at the Domain Name field, and type the domain name for this network. Press Tab to position the cursor at Change Hostname via DHCP, and press Enter to deselect it. Press Tab to position the cursor at Name Server 1, and type the IP address of the DNS server or the name server. (Optional) Type additional IP addresses for the Name Server 2 and Name Server 3 fields, as necessary. Press Tab to position the cursor at Update Name Servers and search list via DHCP, and press Enter to deselect it. Tab to OK and press Enter.
8 9
On the Network Address Setup page, select Static Address Setup. Press the Tab key until you select the IP Address field, and type the IP address that you want to assign to the nodes primary NIC. own subnet mask. Do not use Classless Interdomain Routing (CIDR) notation.
10 Tab to the Subnet Mask, and accept the default subnet mask or type your
11 On the Network Address Setup page, select Routing and press Enter. 12 On the Routing Configuration page, specify the IP address of the default
gateway and select OK.
13 On the Network Address Setup page, select Next. 14 Select Finish. 15 Select Quit. 16 At the command line prompt, type the following command to restart the
configuration wizard:
# /etc/init.d/pdinstgui restart
Chapter
Replication
This chapter includes the following topics:
About replication About replication and PureDisk release levels About replication policies Creating or editing a replication policy Replication jobs Copying and deleting a replication policy Managing replicated data selections Tuning replication
About replication
When PureDisk runs a replication policy, it copies the backed up PureDisk data selections from one storage pool to another storage pool. The replication process copies file data and file metadata. First, PureDisk takes a snapshot of the data selection. Next, it copies the data within the snapshot to a different data selection in a different storage pool. For example, you can create a policy to replicate backed-up data selections from a small storage pool in a small office's location to a large storage pool in a central location. If you have a work group in a small office that is connected to your headquarters over a relatively slow WAN, you can replicate data from that small office to the headquarters location if you have a storage pool in both locations. The PureDisk storage pool in the small office is an all-in-one storage pool; all of the PureDisk services are installed on a single node. Your headquarters has other, larger storage pools. For safety and security reasons, a copy of all data must be
78
available at headquarters at all times. You can create a replication policy to copy data from the remote storage pool to the headquarters storage pool at regular intervals, such as nightly. The replication process does not copy system data such as data selections or policies that you configured. You can preserve this system data through storage pool authority replication. For more information about storage pool authority replication, see the following: See About storage pool authority replication (SPAR) on page 175. You can run a replication policy at any time. However, if the policy includes any data selections that PureDisk has not yet backed up, the replication policy does not replicate those data selections. A replication policy copies only backed up data selections. Additionally, replication jobs and content router rerouting jobs cannot run simultaneously. If you start a replication job and then start a rerouting job, PureDisk stops the replication job. Note: If you experience replication job performance degradation and you have a high-latency communication network between the two storage pools, you can possibly improve performance by changing some default TCP/IP settings. For more information, see "About changing TCP/IP settings to improve replication job performance" in the PureDisk Best Practices Guide, Chapter 5: Tuning PureDisk. The following contain more information about replication:
See About replication and PureDisk release levels on page 78. See About replication policies on page 79. See Creating or editing a replication policy on page 80. See Replication jobs on page 85. See Copying and deleting a replication policy on page 85. See Managing replicated data selections on page 86. See Tuning replication on page 88.
79
You can replicate between storage pools if each storage pool is at the same release level. You can replicate between storage pools when the destination storage pool is at a release level that is higher than the source storage pool. However, Symantec recommends that you install all your storage pools with the same PureDisk release. For example, Symantec supports replication between a source storage pool at the PureDisk 6.5.x release level and a destination storage pool at the PureDisk 6.6 release level. Symantec does not guarantee data integrity when you replicate between storage pools with other nonidentical release levels. You cannot replicate between storage pools when the destination storage pool is at a release level that is lower than the source storage pool.
80
Note: You can delete data from a data selection when you run a removal policy. PureDisk does not remove this data automatically from a replicated data selection. PureDisk does not replicate delete actions. If you want to keep the source and the replicated data selection identical, define similar removal policies for both data selections. After you create and run a policy to replicate data from a source storage pool to a destination storage pool, you can do the following:
View the replicated data on the destination storage pool. For more information about how to view replicated data, see the following: See Replication jobs on page 85. Use restore functions to copy the replicated data to a client that is attached to the destination storage pool. For more information about how to copy replicated data, see the following: See Copying replicated data to clients on the destination storage pool on page 87. Restore the replicated data back to the original client or another client on the source storage pool. For more information about how to restore replicated data, see the following: See Restoring replicated data back to clients on the source storage pool on page 87.
1 2 3
Select Manage > Policies. In the left pane, under Data Management Policies, click Replication. Complete one of the following steps:
To create a policy, in the right pane click Create Policy. To edit a policy, expand Replication and click a policy name.
Complete the General tab. See Completing the General tab for a Replication policy on page 81.
Complete the Data Selections tab. See Completing the Data Selections tab for a Replication policy on page 82.
81
Complete the Scheduling tab. See Completing the Scheduling tab for a Replication policy on page 83.
Complete the Parameters tab. See Completing the Parameters tab for a Replication policy on page 83.
Click Add.
1 2
Type a name for this policy in the Name field. Select Enabled or Disabled. This setting lets you control whether PureDisk runs the policy according to the schedule you specify in the Scheduling tab, as in the following situations:
If you select Enabled, PureDisk runs the policy according to the schedule. If you select Disabled, PureDisk does not run the policy according to the schedule. Disabled is the default. For example, you might want to stop running this policy during a system maintenance period. If you select Disabled, you do not need to enter information in the Scheduling tab to suspend and then reenable this policy.
(Optional) Specify an escalation action. PureDisk can notify you if replication does not complete within a specified time. For example, you can configure PureDisk to send an email message to an individual if a replication policy workflow does not complete in an hour. Select times in the Escalate warning after or the Escalate error and terminate after drop-down boxes to specify the elapsed time. If the policy workflow does not complete before the warning timeout, PureDisk generates a warning event. If the policy workflow does not complete before the error timeout, PureDisk generates an error event. If you select either option, you must create a policy and an event escalation action. These escalation actions define the email message, define its recipients, and associate the escalation action with the policy. For more information about policy and event escalations, see the following: See the PureDisk Backup Operators Guide.
82
1 2 3
Click Data Selection. Select a data selection type from the Data Selection Type drop-down list. Expand the tree and select one or more data selections. To make a selection, click in the box to the left of the label. To select a number of data selections, select a storage pool or department. This action selects all data selections that exist under the storage pool or department. A data selection is associated with a client. PureDisk applies the policy to all selected data selections and the associated clients when you save the policy. If a replication policy includes any data selections that belong to an inactive client, PureDisk replicates that inactive clients data selections. If you do not select a data selection in the tree, PureDisk uses all data selections in the storage pool.
Decide whether you want to include all the data selections you checked in step 3. Proceed as follows:
If you want to include all the data selections in the box, make sure that Include all data selections selected above is selected. Proceed to See Completing the Scheduling tab for a Replication policy on page 83. If you want to exclude, or filter out some data selections, select Apply all inclusion rules below to dataselections selected above. Proceed to step 5.
Fill in the Data selection name or Data selection description fields. Use wildcard characters (* and ?) to filter the data selections. You can filter on the data selection name field or on the data selection description field.
83
For example, assume that the following data selection names exist and that they were selected under a department in the tree:
U_*.jpg_files W_*.jpg_files W_*.xls_files
If you type *files in the Data selection name field, the policy backs up all three data selections. If you type W* in the field, the data selections for this policy include only the data selections that are named W_*.jpg_files and W_*.xls_files. For more information on filtering, see the following: See the PureDisk Backup Operators Guide.
Select a template from the Data selections based on template drop-down list to specify an existing data selection template to use. PureDisk uses all data selections that it linked to the template. This action assumes that the template was previously applied to the client.
1 2 3
Click Scheduling. Select hourly, daily, weekly, or monthly to specify how frequently you want this policy to run. Select schedule details. The detail options that PureDisk displays depend on whether you selected an hourly, daily, weekly, or monthly schedule in the preceding step.
84
1 2
Click Parameters. In the IP of the remote SPA field, type the IP address or the FQDN for the destination (remote) storage pool. Symantec recommends that you specify an FQDN.
3 4 5
In the Login remote SPA field, type the login ID for the destination (remote) storage pool. In the Password remote SPA field, type the password for the destination (remote) storage pool. (Optional) In the Communication Port field, type the open communication port of the storage pool that is used for replication. By default, this port is 10082. If you enabled single port communication, specify port number 443. For more information about single ports, see the following: See the PureDisk Storage Pool Installation Guide.
Select Normal to include images of any new files or changed files since the full replication. Typically, you can select Normal. Select Reverify all to provide a complete copy of all the files that were specified in the data selection. Select Reverify all if you suspect that a problem exists on your destination storage pool. For example, content router corruption or a complete storage pool crash.
(Optional) Type a bandwidth size to use when files are transferred. If you leave this field blank, PureDisk by default uses the maximum bandwidth for the network. This field lets you specify the maximum network bandwidth to use during the backup. For example, if you specify 50 Kbps, the agent uploads data at that rate to the content routers. This number is an average. The peak bandwidth usage can be higher than the number specified, but the average bandwidth usage cannot.
85
Make sure that the Force encryption box is checked. When checked option, PureDisk encrypts and compresses the data over the wire during transmission between the source storage pool and the destination storage pool. Symantec recommends that you configure replication policies with this box checked. If you want the data to arrive at the destination storage pool in an encrypted state, specify encryption in the backup policy on the source storage pool.
(Optional) In the Define backup window field, select a start time and end time for the backup window from the drop-downs. PureDisk queues the job at the time you specify in the Scheduling tab. Other jobs in the queue might prevent PureDisk from running the job immediately. When you specify a start time and an end time, you ensure that the job does not start outside of this period. In addition, PureDisk stops the job if the job does not end before the end time that you specify. Make sure that the start time you specify falls within this replication window.
Replication jobs
PureDisk runs replication jobs on the source storage pool. PureDisk creates virtual agents on the destination (remote) storage pool when you implement replication. PureDisk runs the following types of jobs on virtual agents:
86
1 2
Select Manage > Policies. In the left pane, under Data Management Policies, expand Replication. To expand Replication, click the plus sign (+) to the left of the Replication label.
3 4
Select the replication policy that you want to copy or delete. In the right pane, click one of the following:
See Viewing replicated data on page 86. See Working with replicated agents and data selections on page 87. See Copying replicated data to clients on the destination storage pool on page 87. See Restoring replicated data back to clients on the source storage pool on page 87. See Restoring replicated Oracle data on page 88.
87
1 2 3
Log on to the administrative Web UI of the destination storage pool. Click Manage > Agent. In the left pane, select the replicated client. The icon for a replicated client includes [R] and the name of the client. When you click a client name, client information appears in the right pane.
1 2 3
Log on to the administrative Web UI of the destination storage pool. Click Manage > Agent. In the left pane, select the replicated client. The icon for a replicated client includes [R] and the name of the client. When you click a client name, client information appears in the right pane.
Use the PureDisk restore functions to copy the data. For more information about how to restore data, see the following: See the PureDisk Backup Operators Guide.
88
You can replicate an MS Exchange, MS SQL, Oracle, Oracle UDJ, or System State and Services data selection to another storage pool. To restore data, treat the client to which you want to restore the data as if it were an alternate client. For more information about alternate client restores of application data, see the PureDisk Backup Operators Guide. The following procedure assumes that the original client system can no longer be used for some reason or the original data is corrupted. To restore replicated data back to clients on the source storage pool
Install PureDisk on the client where you want to restore data. See the PureDisk Client Installation Guide.
Log on to the administrative Web UI of the destination storage pool. That is the storage pool that received the replicated data from the source storage pool.
3 4
Click Manage > Agent. In the left pane, select the replicated client. The icon for a replicated client includes [R] and the name of the client. When you click a client name, client information appears in the right pane.
In the right pane, select Restore Files. When you restore a data selection, select the new client of the source storage pool as the destination client. See the PureDisk Backup Operator's Guide for information about how to perform the restore.
Tuning replication
PureDisk includes configuration parameters that you can manipulate to tune replication performance. For information about tuning, see the following: See Tuning replication performance on page 323.
Chapter
About exporting data to NetBackup Configuring PureDisk and NetBackup for export capability Creating or editing an export to NetBackup policy Running an export to NetBackup policy Performing a point-in-time export to NetBackup Troubleshooting export job failures Copying or deleting an export to NetBackup policy Restoring from NetBackup Restoring to a PureDisk client that is not a NetBackup client Restoring to a PureDisk client that is also a NetBackup client
90
After you export the PureDisk files to NetBackup, you can treat these files as if they were native NetBackup files. From the NetBackup Administration Console , you can generate NetBackup reports, browse the files, and manage the files. To restore the data that you exported to NetBackup, use the NetBackup procedures that are described in the NetBackup administration guides. The following provide an overview of the NetBackup export engine:
See Export limitations on page 90. See Requirements for exporting data to NetBackup on page 90. See Requirements for restoring data from NetBackup on page 91. See Enabling and using the NetBackup export engine on page 91.
Export limitations
The PureDisk NetBackup export engine lets you export backed up PureDisk Files and Folders data selections to NetBackup. The NetBackup export engine does not export other PureDisk data selection types. When you choose data selections for export, a tree structure appears in the administrative Web UI, and you make your selection from the tree. Be aware that PureDisk exports only the Files and Folders data selections in the tree. For example, if you select a storage pool that has many types of data selections, PureDisk exports only the Files and Folders data selections. In addition, if you choose to export replicated data from a target storage pool, PureDisk exports only the replicated Files and Folders data selections.
NetBackup 6.0 MP5 or later NetBackup client A NetBackup DataStore license and NetBackup client license When you export PureDisk data to NetBackup, you use a NetBackup DataStore policy. This feature requires that you install a NetBackup DataStore license on the NetBackup server. The PureDisk license key includes a NetBackup DataStore license.
Contact your Symantec sales representative to obtain the required software and license.
Exporting data to NetBackup Configuring PureDisk and NetBackup for export capability
91
See Configuring PureDisk and NetBackup for export capability on page 91. See Creating or editing an export to NetBackup policy on page 102. See Running an export to NetBackup policy on page 108. See Performing a point-in-time export to NetBackup on page 108. See Troubleshooting export job failures on page 109. See Copying or deleting an export to NetBackup policy on page 111. See Restoring from NetBackup on page 111. See Restoring to a PureDisk client that is not a NetBackup client on page 113. See Restoring to a PureDisk client that is also a NetBackup client on page 114.
The PureDisk NetBackup export engine. To configure this engine, install the PureDisk nbu service on this node. When you install the nbu service, you enable this node to export data from this PureDisk storage pool to a NetBackup environment. You can install the nbu service at initial installation time, or you can add it later. The NetBackup client software. These clients are distributed with NetBackup.
92
Exporting data to NetBackup Configuring PureDisk and NetBackup for export capability
You can configure the required software on its own dedicated node, or you can configure this software on a node with other PureDisk services. Figure 6-1 shows the software that you need to configure to enable PureDisk exports to NetBackup.
Exporting data to NetBackup Configuring PureDisk and NetBackup for export capability
93
Figure 6-1
6 7
The figure shows a multinode PureDisk environment that is attached to a NetBackup environment, as follows:
94
Exporting data to NetBackup Configuring PureDisk and NetBackup for export capability
Label Object
1 2 3 PureDisk storage pool NetBackup environment PureDisk node_1, which hosts the following services:
Storage pool authority Metabase server Metabase engine NetBackup export engine NetBackup client
4 5
PureDisk node_2, which hosts a content router service PureDisk node_3, which hosts the following services:
6 7
Both node_1 and node_3 host a PureDisk NetBackup export engine and the NetBackup client software. In this storage pool, node_3 can be a low-end computer because it only serves to transfer data. If you had an all-in-one PureDisk environment, you would have to install the NetBackup client on that one node. The figure shows two clients: kwiek and speedy. The NetBackup export engine on node_1 exports data from kwiek. The NetBackup export engine on node_3 exports data from speedy. To perform a direct restore of files from NetBackup to speedy, install the NetBackup client software on speedy. Configure the PureDisk environment first, and then configure NetBackup. To configure PureDisk and NetBackup to export PureDisk files and folders data selections
Configure NetBackup to receive data that is exported from a PureDisk storage pool. See Configuring NetBackup to receive files and folders data selections exported from PureDisk on page 95.
(Conditional) Configure NetBackup Access Control on the PureDisk nodes. See (Conditional) Configuring NBAC on the PureDisk nodes on page 97.
Exporting data to NetBackup Configuring PureDisk and NetBackup for export capability
95
Complete the NetBackup configuration. See Completing the NetBackup configuration on page 99.
Configure a PureDisk policy that exports data to NetBackup. See Configuring PureDisk to export data to NetBackup on page 100.
Configuring NetBackup to receive files and folders data selections exported from PureDisk
The procedure in this topic explains how to configure NetBackup to accept PureDisk files and folders data selections. To configure NetBackup for PureDisk export capability
Install the NetBackup Linux SUSE 2.6 client on each PureDisk node. Both the PureDisk NetBackup export engine service and the NetBackup client need to be running together on the same PureDisk node or nodes. When you install a NetBackup Linux client, the following message might appear:
No [x]inetd process found.
Ignore this message. A later step in this procedure starts the xinetd daemon. For information about how to install the NetBackup client, see the following: See the NetBackup Installation Guide for UNIX and Linux.
(Conditional) For each PureDisk node, create a file for the host FQDN and another file for the service FQDN in the altnames directory on the NetBackup master server. Perform this step if the storage pool you want to back up is clustered. This step is needed because the bp.conf file on each node contains the physical host address. However, the backup process and the restore process use the service address. If necessary, create the altnames directory itself. Within the altnames directory, use the touch(1) command to create a file for each node's host FQDN and each node's service FQDN. Example 1. To create the altnames directory on a UNIX master server, type the following command:
# mkdir /usr/openv/netbackup/db/altnames
96
Exporting data to NetBackup Configuring PureDisk and NetBackup for export capability
Example 2. Assume that you want to create file names in the altnames directory of a UNIX NetBackup master server for the nodes in the following two-node cluster:
Node 1 = allinone.acme.com (host FQDN) and allinones.acme.com (service FQDN) Node 2 = passive.acme.com (host FQDN) and passives.acme.com (service FQDN)
For the all-in-one node (node 1), type the following commands on the master server to create the correct files in the altnames directory:
# touch allinone.acme.com # touch allinones.acme.com
For the passive node (node 2), type the following command on the master server to create the correct files in the altnames directory:
# touch passive.acme.com # touch passives.acme.com
For more information about the altnames directory and creating files inside the altnames directory, see the following: See the NetBackup Administrators Guide, Volume I.
Determine if NetBackup Access Control (NBAC) is enabled in your NetBackup environment. One way to tell if NBAC is enabled is to examine the bp.conf file. If the USE_VXSS = AUTOMATIC or USE_VXSS = REQUIRED appear, then NBAC is enabled.
If NBAC is enabled in your NetBackup environment, proceed to the following: See (Conditional) Configuring NBAC on the PureDisk nodes on page 97. If NBAC is not enabled in your NetBackup environment, proceed to the following: See Completing the NetBackup configuration on page 99.
Exporting data to NetBackup Configuring PureDisk and NetBackup for export capability
97
Log into the NetBackup master server and complete the following steps:
Change to the directory where the bpnbat command resides. On UNIX master servers, the bpnbat command resides in /usr/openv/netbackup/bin. On Windows master servers, the bpnbat command resides in install_path\NetBackup\bin.
The bpnbat command prompts you for the computer name, prompts you to create an NBAC password, and prompts you to confirm the password. For the computer name, type the FQDN of the PureDisk node.
Change to the directory where the bpnbaz command resides. On UNIX master servers, the bpnbaz command resides in /usr/openv/netbackup/bin/admincmd. On Windows master servers, the bpnbaz command resides in install_path\NetBackup\bin\admincmd.
Type the following commands to enable the PureDisk node to perform access checks:
bpnbaz -allowauthorization pd_node_fqdn Operation completed successfully.
Repeat the preceding bulleted steps for each PureDisk node that hosts a NetBackup client.
masterserver# cd /usr/openv/netbackup/bin masterserver# bpnbat -addmachine Machine Name: potato.idaho.com Password: ***** Password: ***** Operation completed successfully.
98
Exporting data to NetBackup Configuring PureDisk and NetBackup for export capability
On the PureDisk node that hosts the NetBackup client software, install VxAT. Perform the following steps:
Log into the PureDisk node as root. Type the following commands to create NBAC credentials for the node:
puredisknode# cd /usr/openv/netbackup/bin/admincmd puredisknode# bpnbat -loginmachine
Does this machine use Dymanic Host Configuration Protocol (DCHP)? (y/n)?
Authentication Broker:
Specify the host name of the NetBackup authentication broker, and press Enter.
Press Enter to use the default authentication port. If your NetBackup environment uses a site-specific authentication port, type that port number.
Machine name:
Password:
Add entries for the NetBackup master server and the NetBackup media server.
Set the USE_VXSS = and (optionally) the AUTHENTICATION_DOMAIN = parameters as dictated by your site practices. You can set these parameters through the NetBackup administrator interface or by using commands. See your NetBackup administrator documentation for more information about how to edit the bp.conf file.
Exporting data to NetBackup Configuring PureDisk and NetBackup for export capability
99
Repeat the preceding bulleted steps for each node that hosts a NetBackup client.
For example:
puredisknode# cd /usr/openv/NetBackup/bin puredisknode# bpnbat -loginmachine Does this machine use Dynamic Host Configuration Protocol (DCHP)? Authentication Broker: colonel.flagg.com Authentication port [Enter = default]: Machine Name: potato.idaho.com Password: ***** Operation completed successfully. (y/n)? n
Proceed to the following: See Completing the NetBackup configuration on page 99.
On each PureDisk node that you want to configure, make sure that the xinetd daemon is running. Enter the following command to determine if xinetd is running:
# ps -aef |grep xinetd
To ensure that the xinetd daemon starts after you restart the system, type the following command:
# chkconfig xinetd on
From a NetBackup Administration Console , create a DataStore policy with an Application Backup Schedule.
100
Exporting data to NetBackup Configuring PureDisk and NetBackup for export capability
A DataStore policy is a specific policy type. An Application Backup Schedule is the default schedule. For information about how to create this policy, see the NetBackup System Administrators Guide, Volume I. The following are some guidelines for this policy:
When you create this policy, make sure to specify a schedule for 24 hours a day and 7 days a week. This schedule leaves the policy open and available for whenever PureDisk needs to send data. Remember the name of this policy. You need to specify this name when you create the PureDisk Export to NetBackup policy. If you enable encryption, NetBackup asks you for a passphrase. NetBackup uses this passphrase for all data that you export from the client to NetBackup. Ensure that you use the same passphrase for the following:
The NetBackup export engine The client upon which the source files reside All clients to which you might want to restore the exported data.
If you want to export data to NetBackup from a clustered PureDisk storage pool, specify the nodes upon which the NetBackup export engine service resides. In NetBackup, specify the fully qualified domain names (FQDNs) of the physical host for these nodes. These are the FQDNs of the physical nodes. Do not specify the service FQDNs.
Proceed to the following: See Configuring PureDisk to export data to NetBackup on page 100.
Make sure that the NetBackup export engine service is included in the storage pool configuration, is installed on the node that you want to designate as the NetBackup export engine, and is activated. You can specify the nbu as a service on more than one node. For information about how to add a service, see the following:
See About adding services on page 265. See Adding a service to a node on page 266.
Exporting data to NetBackup Configuring PureDisk and NetBackup for export capability
101
For information about how to activate the NetBackup export engine service, see the following: See Troubleshooting export job failures on page 109.
Use the procedure in the PureDisk Client Installation Guide to install and configure PureDisk on clients. You can configure clients with any number of features. However, you cannot enable a data lock password on a client if you want to export that clients data to NetBackup. If necessary, you can enable the data lock password for that client at install time. When the time comes to export data, you need to disable the password. Also note that client naming conventions differ between NetBackup and PureDisk. If the names of your PureDisk clients do not conform to NetBackups client naming conventions, PureDisk transforms the client name to a compatible name. PureDisk changes the name internally, and you can see the result in log files and messages. In addition, you can check the names in the following file:
/Storage/var/NbuExportClientNameChanges.txt
By convention, NetBackup client names can include only the following characters:
The uppercase English language alphabetic characters, A through Z The lowercase English language alphabetic characters, a through z The digits, 0 through 9 A period (.) A hyphen (-) An underscore (_)
Example 2: Assume that you have two clients with the following names:
my agent name is strider my agent_name*is strider
102
To avoid duplication, PureDisk adds a counter to the end of the second name it encounters and transforms the names as follows:
my_agent_name_is_strider my_agent_name_is_strider_2
Make sure that one or more data selections that you want to export have been created and backed up. If no backed up data selections exist in the storage pool, use the instructions in the PureDisk Backup Operator's Guide to create one or more data selections and back them up to the storage pool.
Create a PureDisk policy to export data to NetBackup. For information about how to create an export to NetBackup policy, see the following: See Creating or editing an export to NetBackup policy on page 102.
You can create multiple PureDisk export policies for a single NetBackup export engine. PureDisk runs one export job per export policy at a time. If you have two or more PureDisk export policies, these policies can send data to the same NetBackup DataStore policy. However, you are not limited to only one NetBackup DataStore policy. You can have multiple NetBackup DataStore policies. PureDisk can run multiple export jobs simultaneously from multiple NetBackup export engines if the data originated on two or more PureDisk clients. However, if the export jobs work with data that originated from a single PureDisk client, PureDisk runs the jobs one at a time.
Use the following procedure to create a PureDisk policy that can export Files and Folders data selections to NetBackup. To create an Export to NetBackup policy
1 2
Click Manage > Policies. In the left pane, under Data Management Policies, click Export to NetBackup.
103
To create a policy, in the right pane click Create Policy. To edit a policy, expand Export to NetBackup and click a policy name.
Complete the General tab. See Completing the General tab for an Export to NetBackup policy on page 103.
Complete the Data Selections tab. See Completing the Data Selections tab for an Export to NetBackup policy on page 104.
Complete the Scheduling tab. See Completing the Scheduling tab for an Export to NetBackup policy on page 106.
Complete the Parameters tab. See Completing the Parameters tab for an Export to NetBackup policy on page 106.
Complete the Metadata tab. See (Optional) Completing the Metadata tab for an Export to NetBackup policy on page 106.
1 2
Type a name for this policy in the Policy Name field. Select Enabled or Disabled. This setting lets you control whether PureDisk runs the policy according to the schedule that you specify in the Scheduling tab. The following situations illustrate how the settings might be used:
If you select Enabled, PureDisk runs the policy according to the schedule. This selection is the default. If you select Disabled, PureDisk does not run the policy according to the schedule. One example of how to use Disabled is when you want to prevent this policy from running during a system maintenance period. However, you
104
do not want to enter information in the Scheduling tab to suspend and reenable this policy.
(Optional) Specify escalation times and actions. PureDisk can notify you if a backup does not complete within a specified time. For example, you can configure PureDisk to send an email message to an individual if a backup policy workflow does not complete in an hour. Select times in the Escalate warning after or the Escalate error and terminate after drop-down boxes to specify the elapsed time. If the policy workflow does not complete before the warning timeout (six hours is the default value), PureDisk generates a warning message. If the policy workflow does not complete before the error timeout, PureDisk generates an error message and terminates the job. The default value for the error timeout is five days. Do not set the time values low if the backup policy pertains to agents that have to back up a large number of files. For example, on a system with 1.5 million files, the scan step of a backup job typically takes six hours to complete. The amounts of time required needed for the other steps are more difficult to predict. They depend on the actual number of files that need to be uploaded. In addition to selecting escalation times, you need to create a policy escalation action and an event escalation action. These escalation actions define the email message, define its recipients, and associate the escalation action with the policy. For information about policy escalation actions and event escalation actions, see the PureDisk Backup Operators Guide.
105
Expand the tree and select one or more data selections. PureDisk displays the Files and Folders data selections. To specify a number of data selections, select a storage pool or department. This action selects all data selections that exist under the storage pool or department. PureDisk associates a data selection with a client. When you save the policy, PureDisk applies it to all specified data selections and the associated clients. If you do not select at least one Files and Folders data selection, PureDisk backs up all data selections in the storage pool. If you do not select at least one item in the tree, PureDisk does not save the filter.
2 3
Select Include all data selections selected above. (Optional) Select Apply all inclusion rules below to data selections selected above. Use one of the following filter methods:
Fill in the Data selection name or Data selection description fields with characters and wild cards (* and ?). This method filters the data selections based on their names or their descriptions. For example, assume that the following data selection names exist and are selected under a department in the tree:
U_*.jpg_files W_*.jpg_files W_*.xls_files
If you type *files in the Data selection name field, the policy backs up all three data selections. If you type W* in the field, the data selections for this policy include only the data selections that are named W_*.jpg_files and W_*.xls_files. For more information about filtering, see the following: See the PureDisk Backup Operators Guide.
(Conditional) Select a template from the Data selections based on template drop-down box to specify an existing data selection template to use. This template applies to Files and Folders data selections only. The data selection uses the data selection rules from the template. These rules determine the files and directories to back up. You can select any data selection that you previously applied to the client.
106
Select hourly, daily, weekly, or monthly to specify how frequently you want this policy to run. PureDisk runs the policy at the time you specify in the storage pools time zone.
Select the schedule details. Base your selections on the frequency that you specified in the previous step.
In the NetBackup Policy Name field, specify the name of the NetBackup DataStore policy. For more information about configuring NetBackup to accept PureDisk data selections, see the following: See Configuring NetBackup to receive files and folders data selections exported from PureDisk on page 95.
In the PureDisk to NetBackup export engine field, select the PureDisk node where you installed the NetBackup client and the PureDisk NetBackup export engine. The node appears on the drop-down list.
107
Click Add to define a metadata inclusion rule for the policy. You can specify any or all filters. A file must fulfill all of the specified rules before PureDisk includes the file in the data selection.
In Rule name, type a name for this filter. Tip: Use a U or a W as the first character in the filter name. It helps you to identify whether the filter is for a UNIX or a Windows client.
In Folder name, type the complete path or a pattern of the folder where the files reside. Tip: You can use characters and wildcards to specify both absolute folder patterns and relative folder patterns.
In File name field, type a pattern that describes the files. For more information and examples about filtering, see the PureDisk Backup Operator's Guide.
In Size, use the drop-down lists to specify a filter. Base the choice on the size of the file. For example:
You can select a file size that is greater than or equal to 100 bytes and less than or equal to 500 bytes. In this case, PureDisk includes the files that are from 100 bytes to 500 bytes. You can select a file size that is less than or equal to 500 bytes. In this case, PureDisk includes the files that are from 1 byte to 500 bytes in length. You can select a file size that is greater than or equal to 100 bytes. In this case, PureDisk includes the files that are 100 bytes or longer.
In Last Modification, use the drop-down lists to specify a filter. Base this choice on the time the file was last modified. Select a date-based or time-based boundary. For example, if you specify January 20, 2009, in the Before field, the filter includes all files that were modified before 12:01 A.M. on January 20, 2009. The filter includes the files that were modified any time of day on January 19, 2009.
Click OK.
108
Caution: The filters in the Metadata tab of an Export to NetBackup policy let you narrow the list of files that you want PureDisk to export. If you do not define filters, PureDisk exports all the files. When you specify filters in an Export to NetBackup policy, you might encounter occasional problems when you browse files in NetBackup Backup and Restore interface. The problems can occur with the NetBackup images that PureDisk creates. If you do not find your exported files when you use NetBackup's Backup and Restore interface, use the bplist(1M) NetBackup command-line utility. To edit or delete a metadata inclusion rule
1 2
Select a rule. Select Edit and make changes to the filter rule, or select Remove.
1 2 3 4
Click Manage > Policies. In the left pane, under Data Management Policies, expand Export to NetBackup. Select the policy you want to run. In the right pane, click Run Policy. The PureDisk Backup Operators Guide explains how to enable policy escalation actions and how to monitor policy runs. NetBackup writes log files of its activity to /usr/openv/netbackup/logs/pdexport.
109
Verify that you have a NetBackup DataStore policy defined. See Configuring NetBackup to receive files and folders data selections exported from PureDisk on page 95.
2 3 4 5 6
Click Manage > Agents. In the left pane, expand the tree to display the data selection that you want to export. Select the data selection that you want to export. In the right pane, click Export Whole Data Selection. Specify the following information in the Export files from agent dialog box:
Date The date when the data selection to be exported to NetBackup was backed up. The time when the data selection to be exported to NetBackup was backed up. Specify the name of the NetBackup DataStore policy. Select the PureDisk NetBackup export engine in your storage pool that you want to use to perform the export.
Time
Click Export files. Caution: If the export job finds no backups that match the date specified, the job runs and shows successful completion, but PureDisk exports nothing to NetBackup. This behavior differs from the behavior for regular backups because regular backups fail if nothing is backed up.
See NetBackup export engine log files on page 110. See Problems with inactive server agents on page 110.
110
Click Settings > Configuration > Configuration File Templates > PureDisk Server Agent > Default Value Set for PureDisk Server Agent > debug > logging. Set logging to either info (default), debug, or trace.
Click Settings > Configuration > Configuration File Templates > PureDisk Server Agent > Default Value Set for PureDisk Server Agent > debug > trace. The trace field specifies the directory to which PureDisk writes the log file. The default is /Storage/log. Edit the trace field to specify an alternative directory.
You installed two or more NetBackup export engines on different PureDisk nodes. PureDisk lets you install more than one NetBackup export engine in a storage pool. However, if the following three conditions also exist, an export job fails. The server agent on one of the nodes that hosts a NetBackup export engine is inactive. You have export policies for each of the NetBackup export engines. Each export policy exports the same data selection.
In this situation, PureDisk cannot run the policies to export that specific data selection. Even the export jobs that are scheduled to run on nodes with active server agents fail to run. The server agent must always be active in order for PureDisk to run jobs for that node. However, in this scenario, an inactive server agent can affect the scheduled jobs on an active server agent.
111
To avoid this situation, ensure that each server agent on each node that hosts a NetBackup export engine is activated. To activate a server agent
1 2 3
Click Settings > Topology. Expand the tree in the left pane so that it shows all the PureDisk services. Select the service you want to start. For example, select NBU Export Engine.
1 2 3 4
Click Manage > Policies. In the tree pane, expand Data Management Policies > Export to NetBackup. Select a policy. In the right pane, click the action you want to perform. Select from one of the following:
Delete Policy. Copy Policy. A copy of the policy appears in the tree. The policy is in the disabled state.
112
The example shows an export policy on node_3 that exports data from speedy to NetBackup. In NetBackup, the client name that appears in the NetBackup catalog is speedy. Example 2. If you replicate a data selection and then export that data selection from the destination storage pool, the destination storage pool displays the source client's name. The source client's name appears in the following format:
[R] client_name (agx,stpy)
client_name is the name of the client. x is the agent identifier. y is the storage pool identifier.
When you export a replicated data selection from the destination storage pool to NetBackup, NetBackup removes the [R] characters at the beginning, the agent ID at the end, and the storage pool ID at the end. For example, assume the following series of events:
You have two storage pools: my_spa and your_spa. A client named clientA is attached to my_spa. You replicate clientA's backed-up data selections from my_spa to your_spa. If you look at your_spa's administrative Web UI, clientA appears as follows:
[R] clientA (ag5,stp123).
You configure your_spa as a NetBackup client. You export clientA's data selections from your_spa to NetBackup. You restore clientA's data selections from NetBackup. When you want to restore the data selections from clientA, look for a client named clientA in the NetBackup interface.
Consider the example configuration that is shown in the following topic: See Figure 6-1 on page 93. In the example, the NetBackup job monitor displays the name of the node that hosts the PureDisk NetBackup export engine and the NetBackup client. In , the name that appears in the job monitor depends on which export engine did the export. The name is node_1 or node_3. In the NetBackup catalog, the NetBackup policy that exports data from PureDisk to NetBackup is called a PureDisk-Export policy. The policy type number is 38. For more information about how to restore files from NetBackup, see the following:
Exporting data to NetBackup Restoring to a PureDisk client that is not a NetBackup client
113
Log on to the PureDisk node that hosts the NetBackup client software. If more than one node hosts NetBackup client software, log on to the node to which you want to write the files.
From the NetBackup Administration Console , use the NetBackup Backup, Archive, and Restore interface to perform a client-redirected restore. The client-redirected restore is needed for the following reasons:
You can restore only to a system that hosts a NetBackup client. The name of the NetBackup client appears in the NetBackup catalog. Note that it is not the PureDisk client. Refer to the example in the following topic: See Figure 6-1 on page 93. The example shows that the clients speedy and kwiek appear in the NetBackup catalog.
More information on how to enable a client-redirected restore is available. See the NetBackup Administrators Guide, Volume I. For example, assume the following:
The PureDisk environment is depicted in the following topic: See Figure 6-1 on page 93. The name of the NetBackup policy that performed the export was PDExport. The name of the NetBackup master server is NBUMasterServer.
114
Exporting data to NetBackup Restoring to a PureDisk client that is also a NetBackup client
You want to write the files to their original location on the PureDisk node that is defined as the PureDisk NetBackup export engine (node_3). This location was /bin/myfiles. More information about restoring files to an alternate directory. See the NetBackup Administrators Guide, Volume I
Use a network method to move the files from the PureDisk node with the NetBackup client software to the PureDisk client that needs the files. For example, you can use FTP to transfer the file. This step writes the files to the client, but it does not put the files under PureDisk control. Perform the next step if you want to use PureDisk to back up the files again, which puts them under PureDisk control.
(Optional) Use PureDisk to back up the files. This step puts the files back into the PureDisk environment. More information on how to perform a backup is available. See the PureDisk Backup Operators Guide.
1 2 3
Log on to NetBackup. From the NetBackup Administration Console , use the NetBackup Backup, Archive, and Restore interface to perform a restore directly to the client. (Optional) Use PureDisk to back up the files. This step puts the files back into the PureDisk environment. More information on how to perform a backup is available. See the PureDisk Backup Operators Guide.
Chapter
About creating a disaster recovery strategy About backing up your PureDisk environment using NetBackup Configuring PureDisk disaster recovery backup policies About backing up your PureDisk environment using scripts Troubleshooting a disaster recovery backup
See Disaster recovery backup methods on page 116. See About performing disaster recovery backups with a policy or script on page 119.
116
See Planning your disaster recovery backups on page 120. See Storage pool disaster recovery strategy examples on page 122.
For more general information on disaster recovery backups, see the following topics:
See About backing up your PureDisk environment using NetBackup on page 124. See Configuring PureDisk disaster recovery backup policies on page 130. See About backing up your PureDisk environment using scripts on page 139. See Troubleshooting a disaster recovery backup on page 143. See About storage pool authority replication (SPAR) on page 175.
Disaster recovery backups. Protects the PureDisk environment in the even of a complete hardware failure on one or more nodes. The failure can consist of data loss, data corruption, a disk crash, a flood, or another complete disaster. By default, a disaster recovery backup policy includes all storage pool data, including the backup data on the content routers. When you run a default disaster recovery backup policy, all data is protected. But the backups themselves can require a lot of space on the backup media (tape). In addition, these backups can take a long time to complete. As an alternative, you can configure a disaster recovery backup policy to back up only storage pool metadata. Metadata-only backups include the storage pool services ' databases, the topology files, and the configuration files. When the disaster recovery backup policy runs, it does not back up the following directories: /Storage/spool, /Storage/queue, and /Storage/data. These backups do not include the actual file backup data on the content routers. So they take less space on tape and complete more quickly than a full data backup. You can restore a metadata backup relatively quickly and be ready to back up data again. If you perform full disaster recovery backups that include both metadata and backup data, you have some flexibility during a disaster recovery restore. You can specify to restore only the metadata, or you can specify to perform a complete restore of both metadata and file data. However, if your disaster recovery backups include only metadata, then you can restore only metadata. Checkpointing.
117
Typically enabled with an upgrade, the goal of a checkpoint is to protect the PureDisk environment against data loss or data corruption. Checkpoints do not protect a storage pool from hardware failures because the checkpoint files reside on storage pool disks. PureDisk can create a checkpoint quickly, and a checkpoint is suitable protection for your system at the time apply an upgrade or a patch. Note: Symantec recommends that you use checkpointing only with upgrades and patch installations. Checkpoint files consume needed disk space on storage pool nodes. In addition, when checkpoints are created, PureDisk disables rerouting, PureDisk Deduplication Option (PDDO) backups, and disaster recovery. Failover. Protects a specific, clustered PureDisk node against a complete crash. Failover can occur only in highly available, clustered PureDisk storage pools. If a complete node failure occurs, the clustering software enables a passive node in the storage pool to assume the role of the failed active node. All the PureDisk services that were on the original, active node move to the passive node. Later, you can dispense with or replace the failed node. Both the active node and passive node must be attached to the same shared storage. Storage pool authority replication (SPAR). Uses the PureDisk replication feature to copy one storage pool's metadata to another storage pool. In a SPAR backup, the storage pool's metadata and configuration data is backed up. SPAR enables you to deploy a new storage pool with the same capabilities as the failed storage pool. You can recover a storage pool's backup capabilities and begin backing up clients quickly. SPAR is limited in that SPAR backups do not include any file data. SPAR backups can be enabled successfully in remote office environments where failover cannot be used. The storage pool that runs the SPAR backup must be an all-in-one storage pool.
Table 7-1 summarizes the storage pool disaster recovery backup methods.
118
Drawbacks
Disaster Yes. recovery backup policy, including backup data and metadata
Protects your storage pool configuration and all your file backups from a full system failure.
No.
Protects your Consumes storage pool space on configuration backup media. from a full system failure. Works for both all-in-one (single-node) storage pools or multinode storage pools. Protects your storage pool from data loss and data corruption during an upgrade. Limits storage pool activity when in checkpoint mode. Cannot retain the checkpoints over a long time period. Requires an expensive infrastructure, including a storage area network (SAN) and an extra (passive) node.
Checkpointing No.
Yes.
Failover
Yes.
No.
119
Drawbacks
SPAR
Implementation Reenables a requires two functioning storage pools. storage pool as soon as possible.
File data on the content routers. The content routers are the repository for backup files. Storage pool databases and topology files. These files are as follows:
The storage pool authority database. The content router databases. The metabase engine databases. The metabase server databases The topology files.
Data in the spool area. The spool area is the buffer in which PureDisk stores data before it writes the data to the content routers. During an incremental backup, PureDisk creates a list of changed files and new files on the content router. An incremental backup backs up new or changed content router data from the last full backup. Your last full backup must be available for incremental backups.
120
Storage pool configuration files. PureDisk configuration files reside on the content routers, on the metabase servers, on the metabase engines, and on the storage pool authority. Client data. Client data includes the spool area, the new files, and the changed files on the content routers.
As an alternative to backing up all the preceding data, you can specify that you only want to back up the storage pool metadata. The storage pool metadata includes the services' databases, the topology files, and the configuration files. When you back up only this smaller dataset, disaster recovery backups complete more quickly, but your content router data is not protected from a disaster. When you perform a disaster recovery restore, you cannot recover the backed up file data. That is because that data was never included in a disaster recovery backup. If you do not want to configure a disaster recovery backup policy, you can use scripts to protect your PureDisk environment. The scripts back up all the data that PureDisk includes in a default disaster recovery backup policy. That is, the scripts back up file data, storage pool metadata, configuration files, and so on. The following procedure and examples can help you plan your disaster recovery strategy:
See Planning your disaster recovery backups on page 120. See Storage pool disaster recovery strategy examples on page 122.
Plan your disaster recovery backup policies. When you configure disaster recovery backup policies, consider the following options:
You can configure complete storage pool backups or you can configure metadata-only backups. You can configure a disaster recovery backup policy to include all storage pool data. That includes file backup data, storage pool metadata, the spool buffer, and so on. Alternatively, you can back up only your storage pool's metadata. If you back up only the metadata, you cannot restore the backup data itself. If you configure the backup policy to back up all storage pool data, you have some flexibility when you need to perform a recovery. The DR_Restore_all script lets you specify the type of data you want to restore. You can choose to restore either all the data or only the metadata. If you
121
need to recover the storage pool quickly, you can specify to restore only the metadata.
You can configure only full disaster recovery backups, or you can also configure incremental disaster recovery backups. A full disaster recovery backup includes all the data present in the storage pool at the time the backup starts to run. An incremental backup includes only the information that changed since the last full backup ran. If you select Backup only metadata (databases & configuration) on a full disaster recovery backup policy, you do not need to configure an incremental disaster recovery backup policy. When you select this option, the full and the incremental disaster recovery policies back up the same data.
Note: Symantec recommends that you configure your site's disaster recovery backup policies to include either default, full disaster recovery backups or metadata-only backups. Do not edit your disaster recovery backup policy between policy runs. If you alter your policy between runs, your restore becomes more difficult because of inconsistencies in the backed up file data and metadata. For more information, see the following:
See About creating a disaster recovery strategy on page 115. See About performing disaster recovery backups with a policy or script on page 119.
Plan your disaster recovery backup schedules. You might need to experiment with your disaster recovery backup policy schedules for both full backups and incremental backups. A full backup takes longer to complete than an incremental backup. The exact schedule can depend on the following factors:
How much data is in your storage pool How frequently you need to restore data
122
Plan your NetBackup policies, Samba repository, or third-party backup program. PureDisk lets you configure full disaster recovery backups, incremental disaster recovery backups, or both full backups and incremental backups. In addition, you can specify to include all storage pool data or to include only storage pool metadata in a backup. If you plan to send your backups to a NetBackup repository, make sure that the NetBackup Standard policy and the NetBackup DataStore policy you create are unique and that the repositories that the policies target are unique. The following explains how to configure these policies: See About backing up your PureDisk environment using NetBackup on page 124. If you plan to send your backups to a Samba repository, make sure to move the disaster recovery backups to a server outside of the PureDisk storage pool after the backups complete. The following contains disaster recovery strategy examples: See Storage pool disaster recovery strategy examples on page 122. If you plan to use a third-party product or scripts to perform the disaster recovery backup, make sure that backing up /DRdata to at least one other computer system, outside of your storage pool, is part of your disaster recovery strategy. As part of a disaster recovery restore, you need to move /DRdata back to the restored storage pool.
See Example 1 - Storage pool BLACK on page 122. See Example 2 - Storage pool YELLOW on page 123. See Example 3 - Storage pool RED on page 124.
A PureDisk System policy for full DR backup that includes all storage pool data. This policy runs once each month.
123
A PureDisk System policy for incremental DR backup that includes all storage pool data. This policy runs once every three days.
If you need to recover storage pool BLACK, you can recover all the backup data or you can recover only the storage pool metadata. This strategy requires the following two NetBackup policies:
A NetBackup Standard policy. For example, you can call the NetBackup Standard policy Standard_full. When you enable the PureDisk System policy for full DR backup, specify the name of this policy on the Parameters tab. A NetBackup DataStore policy. For example, you can call the NetBackup DataStore policy DataStore_inc. When you enable the PureDisk System policy for incremental DR backup, specify the name of this policy on the Parameters tab.
When you create the PureDisk full disaster recovery policy and the PureDisk incremental disaster recovery policy for storage pool BLACK, remember to specify both Standard_full and DataStore_inc on both PureDisk policies.
A PureDisk full disaster recovery backup policy that includes all storage pool data. This policy runs once each month. A PureDisk incremental disaster recovery backup policy that includes only the storage pool metadata. This policy runs once every day.
If you need to recover storage pool YELLOW, your options for data recovery are as follows:
You can fully recover the storage pool. That is, you can recover all configuration information, all databases, and all the backup data. To accomplish this full recovery, use the last full disaster recovery backup that ran. You can recover only the storage pool metadata and keep the existing backup data in the data partition. Perform this kind of recovery if a database error occurred. At the end of the recovery process, make sure to run the recoverCR tool. To accomplish this recovery, use the last incremental metadata backup that ran. You can reinstall PureDisk and recover the metadata from the last incremental metadata backup. This kind of a recovery results in a new storage pool with the old storage pool's configuration. You can begin to take backups again very quickly.
124
Disaster recovery backup procedures About backing up your PureDisk environment using NetBackup
In this example, the PureDisk System policy for full DR backup creates a full backup of all the storage pool data. The System policy for incremental DR backup creates a backup of only the storage pool metadata. These backups contain different content. When you configure PureDisk backup policies with different content, make sure that the NetBackup policies to which PureDisk writes the data are not the same policies. This strategy requires the following four NetBackup policies:
For the full backups, create a NetBackup Standard policy and a NetBackup DataStore policy. For example, you can call these policies Standard_full and DataStore_full. When you enable the PureDisk System policy for full DR backup, specify the names of these policies on the Parameters tab. For the incremental backups, create a NetBackup Standard policy and a NetBackup DataStore policy. For example, you can call these policies Standard_inc and DataStore_inc. When you enable the PureDisk System policy for incremental DR backup, specify the names of these policies on the Parameters tab.
A NetBackup Standard policy. For example, you can call it Standard_full. When you enable the PureDisk System policy for full DR backup, specify the name of this policy on the Parameters tab. A NetBackup DataStore policy. For example, you can call it DataStore_inc. When you enable the PureDisk System policy for full DR backup, specify the name of this policy on the Parameters tab.
Disaster recovery backup procedures About backing up your PureDisk environment using NetBackup
125
The following sections describe how to use NetBackup to back up your PureDisk environment:
See Prerequisites for NetBackup disaster recovery backups on page 125. See Configuring the NetBackup client software on page 125. See Enabling NetBackup for PureDisk backups on page 127. See About NetBackup policy names on page 130.
A separately mounted partition named /Storage. The separate mounting enables high performance and is a requirement for disaster recovery backups. A network connection between every PureDisk node and a NetBackup environment. The NetBackup environment must be running NetBackup server software at the 6.0 MP5 release level or greater. A NetBackup client software package at the 6.0 MP5 release level or greater. Install this client software on every PureDisk node in the storage pool. Note: Make sure that the NetBackup client software version number is the same as the NetBackup environment version number. For more information about how to install the software, see the following: See Configuring the NetBackup client software on page 125.
Install the NetBackup Linux SUSE 2.6 client on each node in your PureDisk storage pool. If the storage pool is clustered, install the client on all nodes, including the passive node. For general information about how to install the NetBackup client, see the NetBackup Installation Guide for UNIX and Linux. When you install the client software, use the fully qualified domain name (FQDN) for the client name. For example, answer n to the following short name prompt:
126
Disaster recovery backup procedures About backing up your PureDisk environment using NetBackup
Would you like to use "my_pdnode" as the configured name of the NetBackup client? [y,n] (y) n Enter the name of the NetBackup client: my_pdnode.my_domain.com
If you accept the short name during the install, edit the /usr/openv/netbackup/bp.conf file on each node and change the line that identifies the client. For example:
CLIENT_NAME=my_pdnode.my_domain.com
(Conditional) For each PureDisk node, create a file for the host FQDN and another file for the service FQDN in the altnames directory on the NetBackup master server. Perform this step if the storage pool you want to back up is clustered. This step is needed because the bp.conf file on each node contains the physical host address. However, the backup process and the restore process use the service address. If necessary, create the altnames directory itself. Within the directory, create a file of the following format for each node:
xxxxnode1.symc.be
Create these files for each node in the PureDisk storage pool. Example 1. To create the altnames directory on a UNIX master server, type the following command:
# mkdir /usr/openv/netbackup/db/altnames
Example 2. Assume that you want to create file names in the altnames directory for the nodes in the following two-node cluster:
Node 1 = allinone.acme.com (host FQDN) and allinones.acme.com (service FQDN) Node 2 = passive.acme.com (host FQDN) and passives.acme.com (service FQDN)
To create a file in the altnames directory of a UNIX master server, you type the following commands:
Disaster recovery backup procedures About backing up your PureDisk environment using NetBackup
127
For information about the altnames directory and creating files inside the altnames directory, see the NetBackup Administrators Guide, Volume I.
Verify that the xinetd daemon is running on each node. This service ensures proper communication between the NetBackup master server and the Linux client. Type the following command to determine if xinetd is running:
# ps -aef |grep xinetd
If you restart the system, type the following command to ensure that the xinetd daemon starts:
# /sbin/insserv /etc/init.d/xinetd
Make sure that you have accurate topology and node identification information for this storage pool. This information might be needed during a disaster recovery. Make sure that the information on the cluster planning spreadsheet (for clustered storage pools) or on the installation worksheets (for unclustered storage pools) is accurate.
128
Disaster recovery backup procedures About backing up your PureDisk environment using NetBackup
Plan your NetBackup policies. For planning considerations, see the following: See About creating a disaster recovery strategy on page 115.
Create one or two NetBackup Standard policies with a user backup schedule. This policy is for the PureDisk content router and spool area data. Create policies as follows:
Create one policy if your PureDisk System policy for full DR backup and your PureDisk System policy for incremental DR backup both back up all storage pool data or both back up only storage pool metadata. Create two policies if your PureDisk System policy for full DR backup and your PureDisk System policy for incremental DR backup back up different content. For example, create two policies if your System policy for full DR backup backs up all storage pool data and your System policy for incremental DR backup backs up only storage pool metadata.
For more information about the number of policies to create, see the following: See Storage pool disaster recovery strategy examples on page 122. When you create the NetBackup policies, note the following rules:
Observe the NetBackup policy naming rules. For information about NetBackup policy names, see the following: See About NetBackup policy names on page 130. Remember the name of these NetBackup policies. You specify these names again when you create the PureDisk disaster recovery backup policies. When you create the NetBackup Standard policy, use the same FQDN for the client that you specified when you installed the client. For example, use my_pdnode.my_domain.com to identify the client. The client name must appear identically in the NetBackup policy and in the bp.conf file on the PureDisk node. Make sure that all PureDisk nodes are included in the client list in the NetBackup Standard policy. If the storage pool is clustered, specify the node service FQDN in the NetBackup Standard policy. Do not specify the host address. If your PureDisk environment is set up with /Storage mounted on an NFS share, make sure to check the option Follow NFS in the NetBackup Standard policy definition. Entries are not required on the NetBackup Backup Selections tab.
Disaster recovery backup procedures About backing up your PureDisk environment using NetBackup
129
Make sure that the schedule allows backups 24 hours a day and seven days a week. This method allows PureDisk to send data to NetBackup at any time.
Create one or two NetBackup DataStore policies with an application backup schedule. This policy is for the PureDisk metadata. Create policies as follows:
Create one policy if your PureDisk System policy for full DR backup and your PureDisk System policy for incremental DR backup both back up all storage pool data or both back up only storage pool metadata. Create two policies if your PureDisk System policy for full DR backup and your PureDisk System policy for incremental DR backup back up different content. For example, create two policies if your System policy for full DR backup backs up all storage pool data and your System policy for incremental DR backup backs up only storage pool metadata.
For more information about the number of policies to create, see the following: See Storage pool disaster recovery strategy examples on page 122. Observe the same naming rules regarding NetBackup policies, FQDNs, and so on as described in the following: See About NetBackup policy names on page 130. Make sure that all PureDisk nodes are included in the client list in the NetBackup DataStore policy.
The System policy for full DR backup The System policy for incremental DR backup
PureDisk includes these policies by default. When you edit these policies, enable them, specify information specific to your site, and optionally, create policy escalation actions for them. Make sure that all PureDisk nodes are included in the client list in the NetBackup DataStore policy. For more information about how to create PureDisk disaster recovery backup policies, see the following: See Configuring PureDisk disaster recovery backup policies on page 130.
130
Disaster recovery backup procedures Configuring PureDisk disaster recovery backup policies
Policy names cannot start with a dash (-). Policy names cannot include space characters. Policy names cannot include the characters / @ # * & ^
A full disaster recovery backup policy, which backs up the entire storage pool. An incremental disaster recovery backup policy, which backs up the information that changed since the last full or incremental backup ran.
Both disaster recovery policies can include either all storage pool data or the storage pool metadata only. The following procedure explains how to enable the backup policies. To enable PureDisk disaster recovery backup policies
Plan your disaster recovery strategy. For planning considerations, see the following: See About creating a disaster recovery strategy on page 115. For general disaster recovery information, see the following: See About performing disaster recovery backups with a policy or script on page 119.
Log on to the storage pool and start the administrative Web UI.
Disaster recovery backup procedures Configuring PureDisk disaster recovery backup policies
131
Verify that you configured NetBackup or scripts to back up your PureDisk storage pool. Make sure that one of these backup structures is in place before you enable the PureDisk disaster recovery backup policies. For more information about how to create backups, see the following: See About backing up your PureDisk environment using NetBackup on page 124. See About backing up your PureDisk environment using scripts on page 139.
4 5 6
Click Manage > Policies. In the left pane, under Storage Pool Management Policies, expand Disaster Recovery Backup. Select one of the following policies:
System policy for full DR backup System policy for incremental DR backup
The procedures for enabling these policies are identical. In addition, the tabs and fields in the administrative Web UI are the same for each of these policy types. These policies differ only in the type of backup that PureDisk runs when you enable them.
Complete the General tab. See Completing the General tab for a disaster recovery backup policy on page 132.
Complete the Scheduling tab. See Completing the Scheduling tab for a disaster recovery backup policy on page 132.
Complete the Parameters tab. See Completing the Parameters tab for a disaster recovery backup policy on page 133.
10 Click Save when done. 11 (Optional) Create a policy escalation action for this policy.
The policy escalation action defines the event escalation email message and its recipients. Make sure to associate the escalation action with the policy. For more information, see the PureDisk Backup Operator's Guide.
12 (Optional) Repeat the preceding steps to configure the other policy type.
For example, after you configured a System policy for full DR backup, repeat the preceding steps but select System policy for incremental DR backup.
132
Disaster recovery backup procedures Configuring PureDisk disaster recovery backup policies
(Optional) Type a new name for this policy in the Name field. Perform this step only if you want to rename this policy.
Select Enabled or Disabled. This setting lets you control whether PureDisk runs the policy according to the Scheduling tab, as in the following situations:
If you select Enabled, PureDisk runs the policy according to the schedule. After you enable a policy, you can run the policy on a schedule or manually. If you select Disabled, PureDisk does not run the policy according to the schedule. If a policy is disabled, PureDisk cannot run the policy according to a schedule, and you cannot run the policy manually. This selection is the default. For example, if you want to stop running the policy during a system maintenance period select Disabled. You do not need to enter information in the Scheduling tab to suspend and later reenable the policy.
(Optional) Select times in the Escalate warning after or the Escalate error and terminate after drop-down lists. These times specify the elapsed time before PureDisk sends an email message. PureDisk can notify you if a backup does not complete within a specified time. These fields let you define the times for escalation actions. For example, you can configure PureDisk to send an email message to an administrator if the policy does not complete in eight hours.
Select hourly, daily, weekly, or monthly schedule to specify how frequently you want this policy to run. You can also specify an exact start time to run the schedule.
Disaster recovery backup procedures Configuring PureDisk disaster recovery backup policies
133
Symantec recommends that you run disaster recovery backups when other backups are not running. You can run a disaster recovery backup at the same time that regular system backups run. However, you cannot restore the data you backed up during the regular system backup. When you run a full disaster recovery backup, content router performance degrades. This issue is due to increased file system activity. In extreme cases, a full backup can cause regular backup jobs to fail. Schedule full backups during a time when other backups are not running. You can customize the schedule to suit your sites needs. You can experiment with different schedules. You want to balance the frequency with which this policy runs and system resource usage. If you want to run the policy only one time, select an active policy and click Run Policy from the left pane. After the policy runs, open the General tab again and disable the policy. If you do not disable the policy, it runs again according to the schedule that you specified in the Scheduling tab. If you use NetBackup as your backup tool, make sure that you schedule the PureDisk backup policies to run during NetBackups open window. Symantec recommends that you specify the NetBackup schedule to allow backups 24 hours a day and seven days a week.
Choose one of the following disaster recovery backup methods and follow the procedure that is associated with that backup method:
NetBackup. For information about how to complete the Parameters tab for a NetBackup backup, see the following: See Completing the Parameters tab on a Disaster recovery policy to back up PureDisk to a NetBackup environment on page 134.
134
Disaster recovery backup procedures Configuring PureDisk disaster recovery backup policies
Samba. For information about how to complete the Parameters tab for a NetBackup backup, see the following: See Completing the Parameters tab on a Disaster recovery policy to back up PureDisk to a Samba file system on page 135. Third party. For information about how to complete the Parameters tab for a NetBackup backup, see the following: See Completing the Parameters tab on a Disaster recovery policy to back up PureDisk to a third-party product on page 138.
Completing the Parameters tab on a Disaster recovery policy to back up PureDisk to a NetBackup environment
The following procedure explains how to complete the Parameters tab to back up PureDisk to a NetBackup environment. Make sure that you are familiar with how to create NetBackup policies. For information about how to create NetBackup policies, see the NetBackup Administrators Guide, Volume I. The following procedure describes how to use NetBackup to back up PureDisk data. To use NetBackup to back up PureDisk
1 2 3 4
Select Use NetBackup. In the Standard Policy field, type the name of the NetBackup standard policy that you configured to back up this data. In the DataStore Policy field, type the name of the NetBackup DataStore policy that you configured to back up this data. In the Number of Parallel streams drop-down list, select the number of parallel streams that you specified when you created the NetBackup Standard policy.
Disaster recovery backup procedures Configuring PureDisk disaster recovery backup policies
135
(Optional) Check Backup only metadata (databases & configuration). By default, a disaster recovery backs up all of a storage pool's file data and all configuration data. If you check Backup only metadata (databases & configuration), PureDisk backs up only the storage pool metadata. The storage pool metadata includes the services' databases, the topology files, and the configuration files. When you back up only this smaller dataset, disaster recovery backups complete more quickly. But your content router data is not protected from a disaster because PureDisk does not back up any file content on the content routers. When you perform a disaster recovery restore, you cannot recover the backed up file data because that data was never included in a disaster recovery backup. When you back up only metadata, the disaster recovery backup completes in a shorter amount of time. If you need to perform a recovery, you can recover the storage pool metadata quickly, and you can resume backups quickly. Be aware that if you back up only metadata and you need to recover your system, the backup data that was on the content routers before the disaster is not recoverable. If you have a metadata-only backup, you can use this backup to restore the databases quickly after you see error messages that describe corruptions in the Postgresql database, the metabase engine database, or the storage pool authority database. If you select Backup only metadata (databases & configuration) on a full disaster recovery backup policy, you do not need to configure an incremental disaster recovery backup policy. When you select this option, the full and the incremental disaster recovery policies back up the same data.
Click Save.
Completing the Parameters tab on a Disaster recovery policy to back up PureDisk to a Samba file system
The following procedure describes how to complete the Parameters tab to back up PureDisk data to a Samba file system on another computer. To use a Samba share to back up PureDisk
1 2 3
Make sure that Samba is configured on the computer to which you want to write the PureDisk backups. Select Use Samba Share. In the Full path of data backup program field, specify the full path and the name of the backup script that you want to use.
136
Disaster recovery backup procedures Configuring PureDisk disaster recovery backup policies
The /opt/pdconfigure/scripts/support/DR_BackupSampleScripts/ directory contains sample disaster recovery backup scripts. Symantec recommends that you use the following scripts:
full_DR_backup.sh incremental_DR_backup.sh
In the Directory Path Name field, specify the full path (mount point) to a directory in which to write the backed up files. Specify /DRdata in this field if the following are both true:
You used the full_DR_backup.sh script or the incremental_DR_backup.sh script. You did not modify the scripts. These scripts write to /DRdata. The write occurs even if the directory is mounted to another disk or partition. The script mounts to the directory you specify and writes to it.
Specify your own directory in this field if either of the following are true:
You do not use the full_DR_backup.sh script or the incremental_DR_backup.sh script. You modified the scripts to write to a different directory. Make sure that your backup scripts write to the directory you specify. PureDisk does not mount this directory.
In the Share Name field, specify the name of a remote Samba shared file system. Use the following format for the shared file system:
//hostname/sharename
sharename
6 7
In the Workgroup / Domain field, specify the domain name. In the User Name field, specify the Samba user name.
Disaster recovery backup procedures Configuring PureDisk disaster recovery backup policies
137
8 9
In the Password field, specify the Samba password. Select Use Encryption to have PureDisk encrypt the configuration data before it writes the data. Use Encryption does not cause segment data to be automatically encrypted.
11 Click Save. 12 (Conditional) Update or verify the storage pools topology information.
Perform this step if you selected Use Encryption in the previous step. If you perform the backup with encryption enabled, make sure that you have accurate topology and node identification information. This information is needed during a disaster recovery. Make sure that the information on the cluster planning spreadsheet (for clustered storage pools) or on the installation worksheets (for unclustered storage pools) is accurate. Click Settings > Topology to examine the storage pool's topology.
138
Disaster recovery backup procedures Configuring PureDisk disaster recovery backup policies
Completing the Parameters tab on a Disaster recovery policy to back up PureDisk to a third-party product
The following procedure describes how to complete the Parameters tab to back up PureDisk data to a third-party product. Caution: If you plan to use a third-party product or scripts to perform the disaster recovery backup, make sure that backing up /DRdata to at least one other computer system, outside of your storage pool, is part of your disaster recovery strategy. As part of a disaster recovery restore, you need to move /DRdata back to the restored storage pool. If you choose this method, be aware that you need to copy your backups to a secondary host. If the primary host fails, you are likely to lose both the original files and the backed up files that are written to the local directory. To use a third-party product to back up PureDisk
1 2
Select Use Third Party Product. In the Full path of data backup program field, specify the full path and the name of the backup script you want to use. The /opt/pdconfigure/scripts/support/DR_BackupSampleScripts/ directory contains sample disaster recovery backup scripts. Symantec recommends that you use the following scripts:
full_DR_backup.sh incremental_DR_backup.sh
In the Directory Path Name field, specify the full path to the directory in which to write the backed up files. If you modified or did not use the full_DR_backup.sh script or the incremental_DR_backup.sh script, specify your own directory in this field . If you used the full_DR_backup.sh script or the incremental_DR_backup.sh script and did not modify them, specify /DRdata in this field. These scripts write to /DRdata. The write occurs even if the directory is mounted to another disk or partition. The script mounts to the directory you specify and writes to it.
Select Use Encryption to have PureDisk encrypt the configuration data before it writes the data. Use Encryption does not cause segment data to be automatically encrypted.
Disaster recovery backup procedures About backing up your PureDisk environment using scripts
139
(Optional) Check Backup only metadata (databases & configuration). By default, a disaster recovery backs up all of a storage pool's file data and all configuration data. If you check Backup only metadata (databases & configuration), PureDisk backs up only the storage pool metadata. The storage pool metadata includes the services' databases, the topology files, and the configuration files. When you back up only this smaller dataset, disaster recovery backups complete more quickly. But your content router data is not protected from a disaster because PureDisk not back up any file content on the content routers. When you perform a disaster recovery restore, you cannot recover the backed up file data because that data was never included in a disaster recovery backup. When you back up only metadata, the disaster recovery backup completes in a shorter amount of time. If you need to perform a recovery, you can recover the storage pool metadata quickly, and you can resume backups quickly. Be aware that if you back up only metadata and you need to recover your system, the backup data that was on the content routers before the disaster is not recoverable. If you select Backup only metadata (databases & configuration) on a full disaster recovery backup policy, you do not need to configure an incremental disaster recovery backup policy. When you select this option, the full and the incremental disaster recovery policies back up the same data.
6 7
Click Save. (Conditional) Update or verify the storage pools topology information. Perform this step if you selected Use Encryption in the previous step. If you perform the backup with encryption enabled, make sure that you have accurate topology and node identification information. This information is needed during a disaster recovery. Make sure that the information on the cluster planning spreadsheet (for clustered storage pools) or on the installation worksheets (for unclustered storage pools) is accurate. Click Settings > Topology to examine the storage pool's topology.
Use one of the sample scripts that PureDisk provides. Alternatively, you can customize these scripts for your own sites use.
140
Disaster recovery backup procedures About backing up your PureDisk environment using scripts
Write your own backup script. This script can invoke a third-party backup tool.
See Prerequisites for script-based disaster recovery backups on page 140. See PureDisks disaster recovery backup or restore script examples on page 140. See Creating a backup script on page 141.
The server to which you want to write the PureDisk backup is connected to the network. A Samba shared file system is mounted on the computer to which you want to write the backup. The /Storage partition is mounted as a separate partition. The separate mounting enables high performance and is a requirement for disaster recovery backups.
Table 7-2 lists the scripts that are located in this directory and describes their functions.
Disaster recovery backup procedures About backing up your PureDisk environment using scripts
141
full_DR_backup.sh
incremental_DR_backup.sh
DRrestore.sh
If you modify the scripts that PureDisk provides, the scripts are not protected. During a restore procedure, PureDisk overwrites the scripts if they remain in the default installation directory (/opt). You must place them in another directory for protection (for example, in /usr or /tmp).
Use a text editor to create a backup script that backs up the data directly or calls a backup product. The backup script must perform the actual backup of the data. When you create the script, include the following options:
--new listfile Used for an incremental backup. Specifies that you want to back up the new container directories that were created since the last disaster recovery backup was performed.
142
Disaster recovery backup procedures About backing up your PureDisk environment using scripts
--changed listfile
Used for an incremental backup. Specifies that you want to back up the container directories that have changed since the last disaster recovery backup ran.
--full listfile
Used for a full backup. Specifies that you want to back up all container directories.
--spool spool_files
The spool area on a content router is a buffer area. It holds file data until PureDisk writes the data to the content router. Specifies the node which is to be backed up. Each node that runs this script supplies its own agent identifier.
--agentid agent_id
Do not create actual files for listfile or spool_files. The disaster recovery workflow creates these files and provides them to the script. If you run an incremental backup, and no full backup exists, PureDisk performs a full backup. The disaster recovery backup policy calls the script and runs it each time with a different option, in the following order:
--new --changed Backs up the newly created container directories Backs up the existing container directories that have changed Backs up both the new container directories and the existing container directories that have changed Backs up any files in the content routers spool area Specifies the agent identification of the node that is being backed up
--full
--spool
--agentid
Copy the script you created to every content router in your environment. Write this script to the same location on each content router. For example, /opt/external_scripts.
143
PureDisk overwrites the scripts during a restore, so write a copy to /usr or /tmp.
See Missing pdkeyutil file on page 143. See Content router modes set incorrectly on page 144.
To enable the pdkeyutil command, enter the following command on all active nodes:
# /opt/pdag/bin/pdkeyutil -insert
The preceding command initiates a dialog session with the pdkeyutil utility. The utility prompts you to specify a password for the encryption utility to use during disaster recovery backups and restores. Remember the password that you type. You need this password to restore PureDisk storage pool authority configuration files in the event of a disaster. If you do not remember this password, you cannot complete the restore. When you perform a restore, make sure to use the password that was in effect when your disaster recovery backup ran. If you have changed the password, ensure
144
that the password you specify is the same password that was used when the disaster recovery backup ran. To determine if the key is enabled, enter the following command:
# /opt/pdag/bin/pdkeyutil -display
1 2
Log into one of the content router nodes as root. Type the following command:
# /opt/pdcr/bin/crcontrol --getmode
145
Examine the crcontrol command's output. If the modes are set correctly, the output indicates that all modes =YES except for REROUTE mode, which should be set to =NO. The following output indicates that the modes are set correctly:
Mode : GET=Yes PUT=Yes DEREF=Yes SYSTEM=Yes STORAGED=Yes REROUTE=No
If the crcontrol command output for your content router is not set correctly, type the following command to set one or more modes manually:
# /opt/pdcr/bin/crcontrol -mode mode=Yes
For mode, type one of the following: GET, PUT, DEREF, SYSTEM, or STORAGED. For example, if DEREF=Hold in your output, type the following command:
# /opt/pdcr/bin/crcontrol -mode DEREF=Yes
Repeat this procedure for each content router node in your storage pool.
146
Chapter
About disaster recovery restores Recovering from corruptions in the databases Recovering from corruptions in the data store Recovering data and metadata from a complete storage pool disaster DR_Restore_all script parameters
See Recovering from corruptions in the databases on page 148. See Recovering from corruptions in the data store on page 149. See Recovering data and metadata from a complete storage pool disaster on page 150.
Note: To recover PureDisk when you have enabled the PureDisk deduplication option (PDDO), see the PureDisk Deduplication Option Guide. It contains PDDO-specific information, which includes how to avoid a potential data loss situation.
148
Possible signs of a corrupted database are messages such as the following in the log file:
FATAL: could not open file "/Storage/databases/pddb/data/global/1262": No such file or directory LOG: could not open temporary statistics file "/Storage/databases/pddb/data/global/pgstat.tmp.4391": No such file or directory
The following procedure explains how to recover corrupt databases. To recover from corruption in the /Storage/databases directory
(Conditional) Freeze the cluster. Perform this step if the storage pool is clustered.
When the DR_Restore_all script runs, specify that you want to restore all the nodes in the storage pool. The DR_Restore_all command initiates a dialog with you and requires responses from you. The questions that the script asks differ depending on the disaster recovery backup type you used. For information about the DR_Restore_all command's dialog, see one of the following:
See Running DR_Restore_all to restore a disaster recovery backup performed by NetBackup on page 152. See Running DR_Restore_all to restore a disaster recovery backup written to a Samba shared file system on page 156. See Running DR_Restore_all to restore a disaster recovery backup performed by a third-party product on page 161.
Disaster recovery restore procedures Recovering from corruptions in the data store
149
Log into the storage pool authority node as root. Type the following command to extract the recoverCR tool:
(Conditional) Unfreeze the cluster. Perform this step if the storage pool is clustered.
(Conditional) Freeze the cluster. Perform this step if the storage pool is clustered.
When the DR_Restore_all script runs, specify that you want to restore all the nodes in the storage pool. The DR_Restore_all command initiates a dialog with you and requires responses from you. The questions that the script asks differ depending on the disaster recovery backup type you used. For information about the DR_Restore_all command's dialog, see one of the following:
See Running DR_Restore_all to restore a disaster recovery backup performed by NetBackup on page 152. See Running DR_Restore_all to restore a disaster recovery backup written to a Samba shared file system on page 156. See Running DR_Restore_all to restore a disaster recovery backup performed by a third-party product on page 161.
150
Disaster recovery restore procedures Recovering data and metadata from a complete storage pool disaster
Log into the storage pool authority node as root. Type the following command to extract the recoverCR tool:
(Conditional) Unfreeze the cluster. Perform this step if the storage pool is clustered.
Other methods to restore a storage pool have failed. One or more of your PureDisk nodes is down. That is, the hardware does not function and cannot be repaired. Disks have crashed or other hardware has failed.
The disaster recovery restore procedure has two main phases. In phase 1, you restore all the PureDisk environment that was destroyed in the disaster. In phase 2, you run the PureDisk DR_Restore_all.sh script. The DR_Restore_all.sh script retrieves all your storage pool metadata and all your backup data from the latest disaster recovery backup.
Disaster recovery restore procedures Recovering data and metadata from a complete storage pool disaster
151
1 2
Repair or replace any broken hardware. Reinstall PureDisk. Use the instructions in the PureDisk Storage Pool Installation Guide to reinstall the storage pool software. That is, install PDLinux, install the PureDisk application, upgrade to PureDisk 6.6.1, and apply any hot fixes or EEBs that you had installed on the storage pool. Your goal is to recreate exactly the same PureDisk environment that existed before the disaster occurred. When you reinstall PureDisk, make sure to specify the same passwords, pdkeyutil information, and so on. Refer to the installation worksheets or the cluster planning spreadsheet that you completed when you originally configured the storage pool. These worksheets contain topology information, the storage pool ID number, and other information that can help you recreate the environment.
Use the change_password.sh script to correct the PureDisk database password and the LDAP administrator password on the newly installed PureDisk environment. Perform the following steps:
Log onto the storage pool authority node as root. Type the following command to change to the install scripts directory:
# cd /opt/pdinstall
db to change the PureDisk database password ldap to change the LDAP administrator password all to change both passwords
Respond to the script's prompts. The script prompts you once for the old password and twice for the new password. When it prompts you for the old passwords, type the passwords that are in file /Storage/etc/topology.ini at this time. When it prompts you for the new passwords, type the passwords that were in effect at the time the last good disaster recovery backup ran. These passwords might be the passwords you specified and then recorded on
152
Disaster recovery restore procedures Recovering data and metadata from a complete storage pool disaster
the planning worksheets or spreadsheets when you originally configured the storage pool. However, if you reset any passwords before the last disaster recovery backup ran, reset the passwords to those that existed at the time that backup ran.
(Conditional) Freeze the cluster. Perform this step if the storage pool is clustered.
(Conditional) Install the NetBackup client software on all nodes that failed. Perform this step if you write your disaster recovery backups to a NetBackup environment.
(Conditional) From the NetBackup interface, manually expire any NetBackup images that are newer than the date of the last successful backup. Perform this step only if a previous full disaster recovery backup failed and you write this storage pool's disaster recovery backups to NetBackup. Symantec highly recommends that you purge any corrupted backup images from any unsuccessful backups before you try the restore. You can search the catalog and use the PureDisk server as the client name. Look for the images that have the Standard and DataStore disaster recovery policies. For information about how to search the catalog for images, see the NetBackup Administrators Guide, Volume I.
Run the DR_Restore_all.sh script. The DR_Restore_all.sh script differs depending on the type of disaster recovery backup you need to restore. For information, see one of the following:
See Running DR_Restore_all to restore a disaster recovery backup performed by NetBackup on page 152. See Running DR_Restore_all to restore a disaster recovery backup written to a Samba shared file system on page 156. See Running DR_Restore_all to restore a disaster recovery backup performed by a third-party product on page 161.
Disaster recovery restore procedures Recovering data and metadata from a complete storage pool disaster
153
Run the DR_Restore_all disaster recovery script from the PDLinux command line on the storage pool authority node. To start the script, type the following:
# /opt/pdinstall/DR_Restore_all.sh --no-install [--metadaonly]
Specify the --metadataonly parameter only if you do not want to restore the file backup data. All disaster recovery backup policies back up your storage pool metadata, so you can always run DR_Restore_all to recover your metadata. Do not specify the --metadataonly parameter if you want to restore backup data. Your storage pool backup file data is available to be recovered if you did not check Backup only metadata (databases & configuration) on the Parameters tab of the disaster recovery backup policy.
Affirm whether you installed the NetBackup client on the node. If you answer n, the script stops and you can restart the script after you install the NetBackup client software. For example:
Is the NetBackup client installed on all nodes and pointing to the NBU Server that was used to do the backups? [Yn]: y
Respond to the prompts regarding the topology files. Directory /Storage/etc must contain the following files:
154
Disaster recovery restore procedures Recovering data and metadata from a complete storage pool disaster
If these files are not present, the script retrieves them. If topology.ini.enc is present, the script issues the following prompt for the password:
topology.ini file needs to be decrypted before proceeding enter aes-256-cbc decryption password:
Type the password that you use for the storage pool configuration wizard.
Examine the topology information the script displays and specify that you want to restore all the nodes. The script reads the topology file and presents a display like the following example:
STORAGE POOL TOPOLOGY Node IP Address Services ---- -----------------1 10.80.62.1 spa mbe mbs cr nbu 2 10.80.62.2 cr Node number(s): 1,2
The preceding example shows the topology that the script can restore in this disaster recovery operation. Examine this information for accuracy and specify that you want to restore all the nodes. Use commas to separate the node numbers.
Respond to queries from the script regarding passwords. The restore process can take hours to complete. As the script runs, you might be asked to specify the system passwords for the remote (or local) nodes. This prompt occurs early in the process during secure shell (SSH) authentication.
When the restore is complete, answer the prompts about encryption of the topology.ini file. For example:
Would you like to encrypt the topology.ini file? [Yn]:y Encrypting /opt/pdinstall/topology.ini enter aes-256-cbc encryption password: xxxx Verifying - enter aes-256-cbc encryption password:
Type the password that you use for the storage pool configuration wizard.
Disaster recovery restore procedures Recovering data and metadata from a complete storage pool disaster
155
Observe the completion message. When the operation completes successfully, the script displays the following message:
Disaster recovery complete
(Conditional) Run the following script on the storage pool authority node to upgrade the security protocol:
# /opt/pdinstall/disable_sslv2.sh
Perform this step if you ran the disable_sslv2.sh script on this storage pool at any time. The disaster recovery restore does not enable this script automatically. Symantec recommends that you run the script unless PureDisk 6.5.x storage pools need to replicate to this storage pool.
Log into the storage pool authority node as root. Type the following command to extract the recoverCR tool:
156
Disaster recovery restore procedures Recovering data and metadata from a complete storage pool disaster
13 (Conditional) Re-enable the NetBackup export engine on any nodes that hosted
only a NetBackup export engine service. Perform this step only if you have a node that hosted only a NetBackup export engine service. For information about how to enable a NetBackup export engine, see the following: See About exporting data to NetBackup on page 89.
Running DR_Restore_all to restore a disaster recovery backup written to a Samba shared file system
The DR_Restore_all script initiates a dialog with you. Answer the questions that the script displays. The following procedure explains how to recover a PureDisk storage pool that was backed up to Samba.
Disaster recovery restore procedures Recovering data and metadata from a complete storage pool disaster
157
Run the DR_Restore_all disaster recovery script from the PDLinux command line on the storage pool authority node. To start the script, type the following:
# /opt/pdinstall/DR_Restore_all.sh --no-install [--metadaonly]
Specify the --metadataonly parameter only if you do not want to restore the file backup data. All disaster recovery backup policies back up your storage pool metadata, so you can always run DR_Restore_all to recover your metadata. Do not specify the --metadataonly parameter if you want to restore backup data. Your storage pool backup file data is available to be recovered if you did not check Backup only metadata (databases & configuration) on the Parameters tab of the disaster recovery backup policy.
Provide the information that PureDisk needs to mount the Samba shared file system. For example:
Please enter remote samba share (i.e. //11.88.77.33/remoteSambaShare): //rmns1.min.boston.com/PD_DRdata
158
Disaster recovery restore procedures Recovering data and metadata from a complete storage pool disaster
Provide information about the local mount point. The local mount point is the path name of a directory on which to mount the share. If you used the DR_Restore_all script in the default PureDisk location, the mount point must be /DRdata.
Please enter local mount point (default: /DRdata):
Provide authentication information. This information includes the user name, password, and work group for the Samba share on the remote file server. For example:
Please enter samba user name: pduser Please enter samba password: pdpwd Please enter samba workgroup: pdwgroup Please enter location to restore CR data and spool area from (default: /DRdata):
Type the full system path name to the disaster recovery script used to save your PureDisk data. If you used the DR_Restore_all script in the default PureDisk location, press return. If you supplied your own restore script, PureDisk does not protect it. The scripts are overwritten during a restore procedure if they remain in the default installation directory (/opt). You must write them to another directory for protection (for example, in /usr or /tmp). For example:
Please enter full path of customized DR restore script (default: /opt/pdconfigure/scripts/support/DR_BackupSampleScripts/DRrestore.sh):
Disaster recovery restore procedures Recovering data and metadata from a complete storage pool disaster
159
Respond to the prompts regarding encryption. If you used a pdkeyutil password to enable encryption of your disaster recovery data during backup, supply the password that you specified. If you did not use pdkeyutil, type no; the script does not ask you for a password. For example:
Was encryption used to do the Disaster Recovery Backup [Yn]: y Please provide the pdkeyutil pass phrase: *******
Respond to the prompts regarding the topology file. Directory /Storage/etc must contain the following files:
If these files are not present, the script retrieves them. If topology.ini.enc is present, the script issues the following prompt for the password:
topology.ini file needs to be decrypted before proceeding enter aes-256-cbc decryption password:
Type the password that you use for the storage pool configuration wizard.
Examine the topology information the script displays and specify that you want to restore all the nodes. The script reads the topology file and presents a display like the following example:
STORAGE POOL TOPOLOGY Node IP Address Services ---- -----------------1 10.80.62.1 spa mbe mbs cr nbu 2 10.80.62.2 cr Node number(s): 1,2
The preceding example shows the topology that the script can restore in this disaster recovery operation. Examine this information for accuracy and specify that you want to restore all the nodes. Use commas to separate the node numbers.
160
Disaster recovery restore procedures Recovering data and metadata from a complete storage pool disaster
11 When the restore is complete, answer the prompts about encryption of the
topology.ini file.
For example:
Would you like to encrypt the topology.ini file? [Yn]:y Encrypting /opt/pdinstall/topology.ini enter aes-256-cbc encryption password: xxxx Verifying - enter aes-256-cbc encryption password:
Type the password that you use for the storage pool configuration wizard.
13 (Conditional) Run the following script on the storage pool authority node to
upgrade the security protocol:
# /opt/pdinstall/disable_sslv2.sh
Perform this step if you ran the disable_sslv2.sh script on this storage pool at any time. The disaster recovery restore does not enable this script automatically. Symantec recommends that you run the script unless PureDisk 6.5.x storage pools need to replicate to this storage pool.
Log into the storage pool authority node as root. Type the following command to extract the recoverCR tool:
Disaster recovery restore procedures Recovering data and metadata from a complete storage pool disaster
161
Run recoverCR.
17 (Conditional) Re-enable the NetBackup export engine on any nodes that hosted
only a NetBackup export engine service. Perform this step only if you have a node that hosted only a NetBackup export engine service. For information about how to enable a NetBackup export engine, see the following: See About exporting data to NetBackup on page 89.
(Conditional) Remove corrupted files from an incomplete full disaster recovery backup. Perform the following steps to remove corrupted files if any of the files in your disaster recovery backup are corrupted or incomplete:
Examine your backup repository. Locate the last successful backup. Remove any subsequent failed backups from the system.
(Conditional) Restore the /DRdata directory on the storage pool authority node. Perform this step if the storage pool authority node was not affected by the disaster. Use a file transfer program, such as FTP, to move the /DRdata directory from its backup location back to the storage pool authority.
162
Disaster recovery restore procedures Recovering data and metadata from a complete storage pool disaster
Run the DR_Restore_all disaster recovery script from the PDLinux command line on the storage pool authority node. To start the script, type the following:
# /opt/pdinstall/DR_Restore_all.sh --no-install [--metadaonly]
Specify the --metadataonly parameter only if you do not want to restore the file backup data. All disaster recovery backup policies back up your storage pool metadata, so you can always run DR_Restore_all to recover your metadata. Do not specify the --metadataonly parameter if you want to restore backup data. Your storage pool backup file data is available to be recovered if you did not check Backup only metadata (databases & configuration) on the Parameters tab of the disaster recovery backup policy.
Provide information about where the data to be restored is located. Press Enter to accept the defaults for the following prompts:
Please enter location to restore metadata from (default: /DRdata): Please enter location to restore CR data and spool area from (default: /DRdata):
Disaster recovery restore procedures Recovering data and metadata from a complete storage pool disaster
163
Type the full system path name to the disaster recovery script used to save your PureDisk data. If you used the DR_Restore_all script in the default PureDisk location, press return. If you supplied your own restore script, remember that the scripts are not protected. The scripts are overwritten during a restore procedure if they remain in the default installation directory (/opt). To prevent this problem, you must place them in another directory for protection, such as in /usr or /tmp. For example:
Please enter full path of customized DR restore script (default: /opt/pdconfigure/scripts/support/DR_BackupSampleScripts/DRresto re.sh):
Respond to the prompts regarding encryption. If you used a pdkeyutil password to enable encryption of your disaster recovery data during backup, supply the password that you specified. If you did not use pdkeyutil, type no; the script does not ask you for a password.
Was encryption used to do the Disaster Recovery Backup [Yn]: y Please provide the pdkeyutil pass phrase: *******
Respond to the prompts regarding the topology file. Directory /Storage/etc must contain the following files:
If these files are not present, the script retrieves them. If topology.ini.enc is present, the script issues the following prompt for the password:
topology.ini file needs to be decrypted before proceeding enter aes-256-cbc decryption password:
Type the password that you use for the storage pool configuration wizard.
164
Disaster recovery restore procedures Recovering data and metadata from a complete storage pool disaster
Examine the topology information the script displays and specify that you want to restore all the nodes. The script reads the topology file and presents a display like the following example:
STORAGE POOL TOPOLOGY Node IP Address Services ---- -----------------1 10.80.62.1 spa mbe mbs cr nbu 2 10.80.62.2 cr Node number(s): 1,2
The preceding example shows the topology that the script can restore in this disaster recovery operation. Examine this information for accuracy and specify that you want to restore all the nodes. Use commas to separate the node numbers.
11 When the restore is complete, answer the prompts about encryption of the
topology.ini file.
For example:
Would you like to encrypt the topology.ini file? [Yn]:y Encrypting /opt/pdinstall/topology.ini enter aes-256-cbc encryption password: xxxx Verifying - enter aes-256-cbc encryption password:
Type the password that you use for the storage pool configuration wizard.
165
13 (Conditional) Run the following script on the storage pool authority node to
upgrade the security protocol:
# /opt/pdinstall/disable_sslv2.sh
Perform this step if you ran the disable_sslv2.sh script on this storage pool at any time. The disaster recovery restore does not enable this script automatically. Symantec recommends that you run the script unless PureDisk 6.5.x storage pools need to replicate to this storage pool.
Log into the storage pool authority node as root. Type the following command to extract the recoverCR tool:
17 (Conditional) Re-enable the NetBackup export engine on any nodes that hosted
only a NetBackup export engine service. Perform this step only if you have a node that hosted only a NetBackup export engine service. For information about how to enable a NetBackup export engine, see the following: See About exporting data to NetBackup on page 89.
166
entire storage pool, overwriting all data and metadata. Type the DR_Restore_all command without any parameters when a node is completely unfunctioning. The parameters are as follows:
/opt/pdinstall/DR_Restore_all.sh --metadataonly --no-install --parallel --help
167
--parallel
Restores the backup data in the background and in parallel on all selected content router nodes. By default, the restore process starts with one content router, moves to another, then to another, and so on. Parallel restores can help a disaster recovery restore to complete faster. Displays the information about the DR_Restore_all script parameters.
--help
Example 1
Assume that you have a multinode storage pool. One node hosted only a metabase engine. You need to restore the metabase engine database on one node. When you perform the restore, specify the following:
Specify the --no-install parameter and the --metadataonly parameter on the DR_Restore_all script. When the DR_Restore_all script runs, it asks a series of questions. One question asks you to specify the nodes you want to restore. Specify only that you want to restore the metabase engine node, so type the node number that corresponds to the node that hosted the corrupted metabase engine database.
168
Example 2
Assume that an entire multinode storage pool has crashed. You want to get the storage pool back up and running as soon as possible so you can start to back up clients again. You are not interested in restoring the backup data that was on the storage pool at the time of the crash. The steps are as follows:
Reinstall PureDisk. Use the information in the PureDisk Storage Pool Installation Guide to completely reinstall the storage pool. Log into the storage pool authority node and run the DR_Restore_all script. When you run the DR_Restore_all script, specify the following:
# /opt/pdinstall/DR_restore_all.sh --no-install --metadataonly
When the DR_Restore_all script runs, it prompts you to specify the nodes that you want to restore. Respond to the script by specifying that you want to restore all the nodes. The DR_Restore_all script's parameters have the following effects:
PureDisk restores the storage pool configuration files, the services' databases, the policies you configured, and the user permissions you configured. This minimal restore can complete relatively quickly. The storage pool can back up clients again after the restore is complete. Because the file data from previous backups is not restored, you can never restore the data that was backed up before the disaster.
Example 3
Assume that a file system error occurred on the /Storage/databases partition in an all-in-one storage pool. To restore only the databases and the storage pool configuration, type the following to start the restore:
/opt/pdinstall/DR_Restore_all.sh --no-install --metadataonly
After the DR_Restore_all script finishes, run the recoverCR tool to restore database consistency on the storage pool. The procedure that explains how to run the DR_Restore_all script also explains when to run the recoverCR tool. The DR_Restore_all command in this example leaves all file data intact. Backups and restores can continue for this storage pool after the databases are recovered. The disaster recovery backup can restore only the data it has backed up. If a backup updated the database between the time of the last disaster recovery backup and the time the database corruption occurred, that data backup is lost.
Chapter
About recovering a NetBackup 5000 appliance Performing a metadata-only disaster recovery restore Performing a full disaster recovery restore Recovering from corruptions in the databases or in the data store
See Performing a metadata-only disaster recovery restore on page 170. See Performing a full disaster recovery restore on page 170. See Recovering from corruptions in the databases or in the data store on page 173.
Note: To recover PureDisk when you have enabled the PureDisk deduplication option (PDDO), see the PureDisk Deduplication Option Guide. It contains PDDO-specific information, which includes how to avoid a potential data loss situation.
170
Disaster recovery for NetBackup 5000 appliances (deferred implementation) Performing a metadata-only disaster recovery restore
A Postgresql database corruption Missing tables in the storage pool authority database or the metabase engine database
For information about additional parameters you can specify, see the following: See DR_Restore_all script parameters on page 165. The script prompts you for information. The prompts that the script issues depend on the method you used to back up your storage pool. For information about how to respond to the DR_Restore_all.sh script prompts, see one of the following:
See Running DR_Restore_all to restore a disaster recovery backup performed by NetBackup on page 152. See Running DR_Restore_all to restore a disaster recovery backup written to a Samba shared file system on page 156. See Running DR_Restore_all to restore a disaster recovery backup performed by a third-party product on page 161.
Run the recoverCR tool. See the following topic for more information: See Recovering from corruptions in the databases or in the data store on page 173.
Disaster recovery for NetBackup 5000 appliances (deferred implementation) Performing a full disaster recovery restore
171
Redeploy the storage pool. See Redeploying the NetBackup 5000 on page 171.
Reconfigure the appliance and restore the data. See Reconfiguring the appliance and restoring the data on page 172.
For each node that you want to recover, attach the following:
An external DVD drive. You can plug the drive into a USB port on the NetBackup 5000. A keyboard. A monitor.
2 3 4 5
Insert the PureDisk DVD into the DVD drive. Turn on (boot) the node. On the opening screen, use the arrow keys on the keyboard to select Install and press Enter. Wait for the installation to complete. This step can take about 20 minutes.
At the system login: prompt, log on as root. The following is an example of a system logon prompt:
Linux-pd1 login:
172
Disaster recovery for NetBackup 5000 appliances (deferred implementation) Performing a full disaster recovery restore
7 8
At the password: prompt, type root. After you successfully reinstall PureDisk, perform the following procedure to reconfigure the appliance and restore the data from a disaster recovery backup: See Reconfiguring the appliance and restoring the data on page 172.
Make sure that you successfully reinstalled the PureDisk software on the failed nodes. See Redeploying the NetBackup 5000 on page 171.
Specify network information for each appliance in the storage pool. Perform the following steps for each appliance:
Connect a laptop computer to the NetBackup 5000 appliance. On the laptop computer, start a browser, and type the following URL into the browser:
http://192.168.1.1/ChangeIP
More information is available. See the following topics in the NetBackup 5000 Getting Started Guide:
Connecting a laptop computer to the NetBackup 5000 appliance (Conditional) Specifying network information
(Conditional) Log into the storage pool authority node as root and type the following command:
# /opt/pdappliance/scripts/push-ssh-pubkey.exp
Perform this step only if your storage pool consists of two or more appliance nodes. This command pushes the authentication (ssh) key to all non-storage pool authority nodes.
Disaster recovery for NetBackup 5000 appliances (deferred implementation) Recovering from corruptions in the databases or in the data store
173
Run the following script on each failed node to prepare the nodes for a disaster recovery restore:
/opt/pdappliance/scripts/prepare_DR_Restore_All.sh
For multiple failed nodes, run the script on the failed storage pool authority node first, and then run the script for the remaining failed nodes.
To recover the storage pool, run the script with the following parameters:
parameters
Specify the --metadataonly parameter if you checked Backup only metadata (databases & configuration). For information about additional parameters you can specify, see the following: See DR_Restore_all script parameters on page 165. For information about how to respond to the DR_Restore_all.sh script prompts, see the following:
See Running DR_Restore_all to restore a disaster recovery backup performed by NetBackup on page 152. See Running DR_Restore_all to restore a disaster recovery backup written to a Samba shared file system on page 156. See Running DR_Restore_all to restore a disaster recovery backup performed by a third-party product on page 161.
(Conditional) Run the recoverCR tool. Perform this step if you specified the --metadataonly parameter when you ran DR_Restore_all. See the following topic for more information: See Recovering from corruptions in the databases or in the data store on page 173.
174
Disaster recovery for NetBackup 5000 appliances (deferred implementation) Recovering from corruptions in the databases or in the data store
inconsistencies, contact Symantec technical support to help you identify and solve the cause of the inconsistency before you run the recoverCR tool. Note: Use the recoverCR tool only in conjunction with assistance from Symantec technical support. The recoverCR tool also fixes inconsistencies that occur after a partial disaster recovery restore. For example, when you use the --metadataonly option, you perform a partial restore. The opposite of a partial restore is a full restore; in a full restore, all components, all data, all metadata, and all nodes are recovered. The following procedure explains how to use the recoverCR tool to fix PureDisk databases and data store objects that become inconsistent after a metadata-only recovery. To recover from corruptions in databases or data store
For version, specify the version number of the recoverCR tool that appears in the directory. For example:
tar -C /Storage/tmp -xzf /opt/pdinstall/var/packages/recoverCR-v3.27-rev41791.tgz
Review and follow instructions in the README_FIRST file. The README_FIRST file contains the instructions that explain how to use the recoverCR tool.
Run recoverCR-version.tgz.
Chapter
10
About storage pool authority replication (SPAR) Activating the local storage pool Enabling SPAR backups Running a SPAR policy manually Restoring from a SPAR backup
See Disaster recovery strategies on page 177. See About creating a disaster recovery strategy on page 115.
176
Storage pool authority replication (SPAR) About storage pool authority replication (SPAR)
Note: The main storage pool can be configured as a clustered storage pool. However, Symantec does not support SPAR for clustered local storage pools. When SPAR runs under cluster control, a failover moves all node functions to a passive node. However, the failover does not move the SPAR feature that you enabled on the original local storage pool authority node. Figure 10-1 shows an example PureDisk environment with two storage pools. Figure 10-1 SPAR example
SP_local is a small, local storage pool in Duluth and SP_main is in a main office
in Minneapolis. SPAR is implemented to back up system information from SP_local to SP_main. The information in this section uses this example environment. SPARs main benefit is that SPAR enables you to restart a storage pool and begin backing up data soon after a disaster. A SPAR recovery is best performed in the following circumstance:
You have an all-in-one local storage pool that is down completely. You want to restore all your user information, data selection definitions, backup policies, and system policies. This data includes all the user data and storage pool data that enables client backups. This data does not include the backup data or backup metadata. You want to be able to start backing up data again very quickly.
SPAR differs from the other disaster recovery methods because SPAR does not recover your backup data or metadata. A full disaster recovery can take several hours or days, depending on how much data you backed up. SPAR recoveries are faster. After a SPAR recovery, PureDisk sees the local storage pool as if it were a newly configured storage pool. The backups you perform immediately after a SPAR recovery are all full backups. When you enable both comprehensive disaster recovery backups and SPAR, you can choose the recovery method you want to use. If you perform a SPAR recovery, you can use full disaster recovery methods to restore your file data and metadata
Storage pool authority replication (SPAR) About storage pool authority replication (SPAR)
177
to an alternate storage pool. You need an alternate storage pool in this circumstance. A full disaster recovery to a local storage pool destroys all the data that you backed up between the SPAR recovery and the time you performed the full disaster recovery. Note: If you experience replication job performance degradation and you have a high-latency communication network between the two storage pools, you can possibly improve performance by changing some default TCP/IP settings. For more information, see "About changing TCP/IP settings to improve replication job performance" in the PureDisk Best Practices Guide, Chapter 5: Tuning PureDisk.
Complete disaster recovery. This method uses a PureDisk policy that sends data to NetBackup or scripts to back up and restore storage pool specifications, file data, and file metadata. For more information about complete disaster recovery, see the following: See About creating a disaster recovery strategy on page 115. Storage pool authority replication (SPAR). This method backs up and restores storage pool specifications, but not file data or metadata.
Along with disaster recovery backups, you can replicate all your data selections to another storage pool. You can use the replicated data selections to restore a PureDisk storage pool. If you replicate all data selections to another storage pool, you can limit data loss after a disaster. For information about data replication, see the following: See About replication on page 77. The method that is best for your site depends on your configuration, practices, and disaster recovery goals. Table 10-1 shows the characteristics of these two methods.
178
Storage pool authority replication (SPAR) Activating the local storage pool
SPAR
Storage pool metadata.
Estimated restore Depends on the amount of data. time This step can take hours. Storage pool type Any type of storage pool.
SPAR restores take much less time than a complete disaster recovery. The protected storage pool must be an all-in-one, single-node, unclustered storage pool. You want to restore the storage pool and back up the clients as quickly as possible.
Restore goal
Restores your storage pool to the Restores your storage pool users, state it was in when the last accounts, data selections, policies, disaster recovery backup was run. and all other storage pool configuration information. This method does not restore any file data or file metadata. After you restore, you need to run backups for all your clients. Old or changed data is no longer available.
1 2 3
In the local storage pools administrative Web UI, click Settings > Topology. In the left pane, select a local storage pool. In the right pane, click Activate SPA Replication.
Storage pool authority replication (SPAR) Activating the local storage pool
179
Complete the display that appears. The display requests the same information that you need to supply when you install an agent on the storage pool authority. These fields are as follows:
Login The root users logon on the main storage pool authority node. The root users password on the main storage pool authority node. The fully qualified domain name (FQDN) of the main storage pool. Example:
Password
SP_Main.acme.com
The local storage pools host name. Type this name as you want it to appear in the main storage pools administrative Web UI. The path to the Linux agent software on the main storage pool. This agent software is the same as the agent software that you use for all other Linux PureDisk clients. You can specify an IP address or a host name to identify the main storage pool. In this field, do not specify the actual file in which the Linux agent installation software resides. You specify the file name in the next field, Binary. Example: Assume that you installed the agent software packages in the default location on SP_Main. The default location is as follows: /opt/pdweb/htdocs/download/Linux_Clients For the Binary Location field, use the fully qualified domain name (FQDN) and specify the following path: https://SP_Main.acme.com/download/Linux_Clients
Binary Location
Binary
The name of the file that includes the Linux agent installation software on the main storage pool. The Binary Location fields content points to this file name. For example, pdagent-Linux_2.6_x86-6.2.0.5.run.
180
Path
The path to which you want to install the Linux agent on the local storage pool. For example, /opt/SPAR.
Record the information you specified and keep it in a safe place. If the local storage pool goes down, you need the information you specified on this screen to perform a restore. All the arguments to these fields appear on the main storage pool, except for the Path field. If a disaster occurs, you can restore a downed local storage pool faster if you have this information recorded and stored safely.
Click Save. After you click Save, PureDisk runs a job to activate the local storage pool as a client to the main storage pool. After activation, the status of the two storage pools is as follows:
The local storage pool appears in the main storage pools list of clients. Two agents reside on the local storage pool. The first is the local storage pools primary server agent. The second is an agent that connects the local storage pool to the main storage pool.
181
1 2 3 4
On the local storage pool, click Manage > Policies. In the right pane, under the Storage Pool Management Policies category, click the plus (+) sign to the left of SPA Replication. Select System policy for SPA replication. Complete the required information in the General tab. See Completing the General tab for a SPA Replication on page 181.
Complete the required information in the Scheduling tab. See Completing the Scheduling tab for a SPA Replication policy on page 182.
Complete the required information in the Parameters tab. See Completing the Parameters tab for a SPA Replication policy on page 182.
Click Save.
(Optional) Type a new name for this policy in the Name field. You do not have to rename this policy.
If you select Enabled, PureDisk runs the policy according to the schedule in the Scheduling tab. If you select Disabled, PureDisk does not run the policy according to the schedule in the Scheduling tab. This selection is the default. You might use Disabled to stop this policy from running this policy during a system maintenance period. Then, you would not need to enter
182
information in the Scheduling tab to first suspend and later reenable this policy.
Select times in the Escalate warning after or the Escalate error and terminate after drop-down boxes. These times specify the elapsed time before PureDisk sends a message. PureDisk can notify you if a policy does not complete its run within a specified time. For example, you can configure PureDisk to send an email message to an administrator if a policy does not complete in an hour. If you select either of these options, create a policy escalation action. The action defines the email message, defines its recipients, and associates the escalation action with the policy. For more information about policy escalations, see the PureDisk Backup Operator's Guide.
Specify the schedule details that define how frequently you want the policy to run.
(Conditional) Correct the information in these fields. Perform this step only if the fields contain incorrect information. For example, if the URL, logon, or password for the main storage pool ever change, correct the information in this tab.
Click Save.
183
1 2 3 4
On the local storage pool, click Manage > Policies. In the left pane, under Storage Pool Management Policies, click the plus (+) sign to the left of SPA Replication. Select system policy for spa replication. (Conditional) Enable the policy. You have to enable a policy before you can run it. In the right pane, click Enabled and Save.
5 6
1 2
Gather the information you need to perform a SPAR restore. Install the PureDisk operating system (PDLinux) on the local storage pool. For more information about how to install PDLinux, see the PureDisk Storage Pool Installation Guide.
Use the storage pool configuration wizard to configure the PureDisk storage pool. Configure the storage pool software on the local storage pool. For more information about how to use the storage pool configuration wizard, see the PureDisk Storage Pool Installation Guide Perform the following steps if the storage pool software does not function:
Remove the storage pool softwares previous upgrade package. The following is the directory you need to remove:
/etc/puredisk
Configure new storage pool software. During this reconfiguration, the installer proposes again a (new) random storage pool ID. Do not accept this proposed storage pool ID. Instead, specify the original storage pool ID. If you do not specify the original
184
storage pool ID, the storage pool becomes inoperable after you perform the SPAR restore. When you configure the new storage pool software, specify the same passwords that you specified during the previous configuration.
Deactivate the agent in the main storage pool that performed the SPAR. Perform the following steps:
Log on to the main storage pools administrative Web UI. In the left pane, click Manage > Agents. Select the PureDisk agent that represents the PureDisk storage pool. In the right pane, click Deactivate Agent.
Retrieve the information you used to configure SPAR initially. See To register the local storage pool on page 178. This procedure advises you to record the configuration information and store it in a safe place. Retrieve this information now. You need this information and some additional information to create the restore command. A later step in this procedure directs you to use the RestoreSPAAIO.php command to perform the restore. You can obtain information about many of this commands arguments from the main storage pool. Other information, however, such as the install path to the SPAR client on the local storage pool is not recorded anywhere in PureDisk. If you have this information before you begin, the restore command is easier to specify.
185
6 7
Log on to the local storage pool as root. Use the RestoreSPAAIO.php command to perform the restore. The restore command restores the local storage pools system information and re-establishes the local storage pools client relationship to the main storage pool. Refer to Table 10-2. The figure shows the arguments to the RestoreSPAAIO.php command. You can type the arguments in any order, but the arguments must match your original configuration. The following example shows the command with all the required arguments:
# /opt/pdag/bin/php /opt/pdspa/cli/RestoreSPAAIO.php --ip SP_Main.acme.com --login root --password root --hostname SP_local-Duluth --binary pdagent-Linux_2.6_x86-6.2.0.5.run --binaryloc https://SP_Main.acme.com/download/Linux_Clients/ --agentlocation /opt/SPAR --agentid 2 --dsid 2 --dumpdir /Storage/SPAR/
1 2 3
--agentlocation
Click Manage > Agent. In the right pane, select the local storage pools agent icon. Note the number in the Storage Pool ID field.
Full path to the directory in which the agent resided for the previous SPAR backup. You specified this information in the Path field when you configured SPAR.
186
--binaryloc
--dsid
1 2 3 4
--dumpdir
Click Manage > Agent. In the left pane, click the plus sign (+) to the left of the local storage pools agent icon. Select the SPAR data selection. Note the number in the ID field.
Full path to the restore directory. Specify the same dump directory that you used for the previous SPAR backup. Host name of the local storage pool as it appeared in the main storage pools administrative Web UI for the previous SPAR backup. FQDN, host name, or IP address of the main storage pool. Specify a resolvable identifier. Root user logon to the main storage pool. Root user password for the main storage pool. (Optional) Runs command in debug mode. (Optional) Runs command in silent mode. (Optional) Runs command in verbose mode.
--hostname
--ip
187
First, upgrade the storage pool from which you replicate the storage pool authority. Second, upgrade the storage pool to which you replicate the storage pool authority.
For more information about upgrades, see the PureDisk Storage Pool Installation Guide.
188
Chapter
11
Reports
This chapter includes the following topics:
About reports Permissions and guidelines for running and viewing reports Reports for a running job About policies and workflows Obtaining detailed job reports About Data mining reports Enabling a data mining policy Running a data mining policy manually Obtaining data mining policy output - the data mining report Obtaining data mining policy output - the Web service report Web service reports About Dashboard reports Central storage pool authority reports
About reports
The following explain how to run and display PureDisk reports:
See Permissions and guidelines for running and viewing reports on page 190. See Reports for a running job on page 191. See About policies and workflows on page 192.
190
See Obtaining detailed job reports on page 194. See About Data mining reports on page 210. See Enabling a data mining policy on page 211. See Running a data mining policy manually on page 212. See Obtaining data mining policy output - the data mining report on page 213. See Obtaining data mining policy output - the Web service report on page 216. See Web service reports on page 219. See About Dashboard reports on page 227. See Central storage pool authority reports on page 231.
Permission to create reports. A user needs Report permission and View permission to create reports.
A user in the reporters group and a user with Report permission can run reports and can run data mining policies. You can assign users to the reporters group only at the World level. A user needs View permission at the storage pool level to retrieve data mining reports.
Permission to view reports. View permission determines how much information that user can see in a data mining report. For example, users with View permission at the storage pool level can view information for the entire storage pool in data mining reports. If a user has View permission only at the client level, they can see only that client's information in a data mining report. Permission to create reports about central storage pools. A user needs Central Report permission at the storage pool level. Permission to access the Reports tab. PureDisk displays the Reports tab for the following types of users:
Users with root permission Users with Report permission at the world level
191
Other factors can affect the availability of reporting data. For example, if you restored a storage pool, wait about 15 minutes before accessing report data. For more information about permissions, see the PureDisk Client Installation Guide.
See Examining a running job on page 191. See Restarting a backup job on page 192.
1 2 3
Click Manage > Agents. In the left pane, select the storage pool. In the right pane, select More Tasks and pull down Job Steps Report. PureDisk displays the Job Steps Report.
Click one of the tabs to see different aspects of a jobs progress. For example, click Job Steps with Problems to see job steps. Click one of the numbers in the Job ID column to display more details about a specific job. For more information about the job details displays, see the following: See Obtaining detailed job reports on page 194. If a data lock password is enabled on this agent, the Files and Errors tabs prompt you for the password when you attempt to view them. For more information about the data lock password, see the PureDisk Client Installation Guide.
192
1 2
Click Monitor > Jobs. (Optional) Narrow your search for the job you are interested in. If there are many jobs in the right pane, you can narrow your search through one of the following:
Specify information in the Look for field, pull down an object type from the in field, and click Find now to display only a subset of the default information. You can specify a string of characters to search for in the Look for field. For example, you can specify the full name of a client in the Look for field, and then you can select Agent Name from the in pull-down menu. In the left pane, select a category from the View Jobs By pull-down menu.
See Types of workflows on page 192. See Workflows in policies on page 193.
Types of workflows
A workflow is a collection of steps that PureDisk completes to accomplish a task. A policy is a special kind of workflow. To create a policy manually, or to edit a policy, click Manage > Policies. The PureDisk administrative Web UI categorizes policies and workflows as follows:
Backup Policies Data Management Policies Storage Pool Management Policies Restore Workflows
193
Miscellaneous Workflows
If you upgraded from a previous PureDisk release, you might also see the Legacy Workflows category with one or more workflows beneath it. For example, this category might contain the following workflows:
6.5 Data selection removal workflow 6.5 Rerouting Workflow 6.5 MBDataMining workflow
Whether the administrative Web UI displays any legacy workflows depends on the presence of existing workflows at the time of your upgrade. If you ran a data mining policy before you applied an upgrade, the workflow appears in the administrative Web UI after the upgrade is installed. You can examine the outcomes of these workflows, or you can delete them.
Workflows in policies
A workflow step defines a PureDisk action. PureDisk accomplishes its work by running a series of workflow steps. The individual workflow steps are predefined, and each performs a specific action. When you use PureDisk to perform a backup, a restore, or any other kind of task, PureDisk completes that task by running several workflows. A policy defines a data management or maintenance action. Within a backup policy, for example, the schedule determines when the policy runs, the agents and the data selections to back up, and other various parameters. PureDisk can stop processing after a timeout . A timeout can occur in two different ways:
In a workflow step. PureDisk permits internal workflow steps to run only for a limited time. In a policy. The General tab of a backup policy lets you specify the amount of time a policy can run before PureDisk terminates the policy run.
PureDisks internal watchdog monitors workflow steps. In the case of a backup policy, the watchdog issues a message if the backup does not complete within the specified backup window. The watchdog also issues messages for individual workflow steps or the policies that terminate. You can configure event monitoring to notify you of these occurrences. For more information about how to configure events, see the PureDisk Backup Operators Guide.
194
1 2
Click Monitor > Jobs. (Optional) Specify information in the Look for field, pull down an object type from the in field, and click Find now to display only a subset of the default information. You can specify a string of characters to search for in the Look for field. For example, you can specify the full name of a client in the Look for field, and then you can select Agent Name from the in pull-down menu.
In the right pane, in the Job ID column, click the number that corresponds to the job you want to examine. An informational window appears with the several tabs. For example, the pop-up window includes the following tabs for a backup job:
General Includes the jobs execution status, whether there were any errors during the jobs run, and when the job commenced. Shows the status for each specific part of a jobs run. On this tab, you can see how PureDisk breaks a job apart for processing. Provides the information on the number of processed files and the number of bytes of data that was transferred between the client and the content router. Lists the files that the job backed up. Includes whether PureDisk backed up the files successfully, the client upon which the file resided, and the name of the file. Lists the files with the errors that PureDisk encountered when it processed the job. Shows the job step output.
Details
Statistics
Files
Errors
Job log
195
Delete job
If you see this message, perform the procedure in the following section: See Examining lengthy job logs on page 210.
196
For efficiency reasons, PureDisk always uploads files smaller than 16 KB to the content router, even if they are already stored on the content router. Consequently, the backup statistics can be different from what you expect if you back up many files smaller than 16 KB. For example, the data reduction factor can be lower than expected, or the number of bytes transferred can be higher than expected. Table 11-1 contains information about how to interpret the statistics in a backup job. Table 11-1 Lines in the Statistics tab for a backup job Meaning
Statistic or heading
Data Reduction: Global data reduction savings
The percentage of source data bytes that did not have to be transmitted to the content routers because of data reduction. Higher numbers correlate to more efficiency. The total number of bytes for the files that PureDisk backed up divided by the amount of bytes transferred to the content routers. Higher numbers correlate to more efficiency.
Data Uniqueness: Unique files and folders The number of backed up files that were globally unique, after backed up global data reduction, before segmentation, and before compression. This statistic is the number of files that are unique in the group of data selections under consideration. The files themselves are considered, but optimization through segmentation is not considered. For example, if a file resides on three different clients, PureDisk stores the file only once and counts it only once in this number. At the segment level, however, PureDisk performs more optimization. A file segment can be present in more than one file, and PureDisk stores that segment only once. Unique bytes backed up The total number of bytes in the backed up files that were globally unique. This statistic is the accumulated size of the unique files that were transferred to the content routers. When encryption or compression are enabled, it is the accumulated size of the encrypted or compressed unique files. The reported value also includes all overhead bytes necessary for headers, alignment, and so on. The values in the Source bytes backed up and Unique bytes backed up fields are not always identical even if all files backed up are unique.
197
Table 11-1
Statistic or heading
Source selection: Files selected on source
The number of files that meet the data selection inclusion and exclusion rules. Pertains to regular files only. This number does not include the number of special files, such as symbolic links or device special files.
Bytes selected on source The total number of bytes for the files that meet the data selection inclusion and exclusion rules. Pertains to regular files only. This number does not include the volume of special files, such as symbolic links or device special files. Files new on source The number of selected files that are new compared to the previous backup run. Pertains to regular files only. This number does not include the number of special files, such as symbolic links or device special files. The total number of bytes for the selected files that are new compared to the previous backup run. Pertains to regular files only. This number does not include the volume of special files, such as symbolic links or device special files.
Files modified on source The number of selected files that were modified compared to the previous backup run. Pertains to regular files only. This number does not include the number of special files, such as symbolic links or device special files. Bytes modified on source The total number of bytes for the selected files that were modified compared to the previous backup run. This number does not include the volume of special files, such as symbolic links or device special files. Files not modified on source Bytes not modified on source Files deleted on source Bytes deleted on source The number of files that were not modified since the last backup ran. The total number of bytes for the files that remained unchanged since the last backup ran. The number of files that were deleted since the last backup ran. The total number of bytes for the files that were deleted since the last backup ran.
Network:
198
Table 11-1
Statistic or heading
Backup speed
Bytes transferred
Protected Data: Source files backed up The number of selected files that were backed up correctly. That is the sum of new, modified, and nonmodified files that are correctly backed up and do not contain errors. The total number of bytes for the selected files that were backed up correctly. For information about the relationship of this field to the Unique bytes backed up field, see the Unique bytes backed up field description. Source files with errors The number of selected files that PureDisk cannot back up.
Source bytes with errors The total number of bytes for the selected files that PureDisk cannot back up. Time: Start date/time Stop date/time Backup time duration The date and time that the job started. The date and time that the job ended. The amount of time that elapsed between when the job started and when the job ended.
Notes:
Table 11-1 shows the statistics for one job. However, the data mining reports, when run at the storage pool level, show the data reduction factor for the storage pool. The storage pool data reduction factor in the data mining reports represents the volume of all data ever backed up to that storage pool, in bytes, that is retained and currently available for restores versus the number of bytes consumed on the content routers.
199
The storage pool data reduction factor differs from the statistics because the statistics in the table are generated for only one job. More information is available about the data mining reports. See About Data mining reports on page 210.
Several factors can affect the Bytes transferred statistic. The data selection may contain a huge number of small files or have very small segment sizes. In these cases, the bytes transferred can be much larger than the on-source values. The following additional information applies to this statistic:
If the backup includes only special files, the "...on source" statistics show zero files selected because there were no regular files to back up, but the Bytes transferred statistic can be a large number. Compression has the greatest effect on the Bytes transferred statistic. If you enable compression, the Bytes transferred statistic is usually lower than the Bytes selected on source statistic. The Bytes transferred statistic might be higher if the data being transferred cannot be compressed. Data that cannot be compressed includes data that is already compressed such as movies, JPEG files, MP3 files, or files in ZIP format. For the files that are already compressed, the compression is ineffective and might result in a slight increase of the data to be transferred. For a repeated backup, the Bytes transferred statistic should be much lower than Bytes selected on source. The Bytes selected on source statistic is the sum of all bytes present in the data selection. For an initial backup, if you disable compression, the Bytes transferred statistic is usually higher than the Bytes selected on source statistic because of the overhead in the internal data format. The rate of data change on the client affects the Bytes transferred statistic. The Bytes selected on source represents the sum of all bytes in the entire data selection. If the data change rate is 100% (for example, if all files changed or it is a first-time backup) and you disable compression, the Bytes transferred statistic is always higher. If the change rate is less than 100%, Bytes transferred statistic is lower. The file size affects the Bytes transferred statistic. As a performance enhancement, PureDisk always transfers files for which the content of files is smaller than the segment size. In this case, PureDisk does not perform a prior-existence check on the content routers. That is in contrast to backups for the file content that is larger than the segment size. PureDisk always performs a prior-existence check for the files that are larger than the segment size.
200
Consequently, if a data selection consists mainly of files smaller than the segment size, the number of bytes transferred can be higher than expected. For example, if PureDisk backs up an identical set of files on two different clients, it would be logical to expect that the bytes transferred would be low for the second client because the files already exist on the content router. That is not the case, however.
Segmentation size and file size affect the Bytes transferred statistic. PureDisk uses a special data format to store the data on its content routers. This data format has a per-segment overhead of a 22-byte header, a 12-byte trailer, 16 bytes per block of 32 KB of data, and up to 7 padding bytes. The padding bytes enable PureDisk to align the data according to its internal data format, which requires data to be aligned on an 8-byte boundary. For example, if you have a segment of exactly 128 KB, the total data overhead is 34 + (4 * 16) = 96 bytes. As another example, if the segment is 128 KB - 1 byte long, the total data overhead is 96 + 1 padding byte = 97 bytes. If the segment is smaller than 32 KB, the number of overhead bytes can vary between (32 + 16) = 48 and (32 + 16 + 7) = 55 bytes. If you disable encryption, the header is 14 bytes long instead of 22.
For Linux and UNIX backups that contain duplicates of small files, PureDisk correctly stores only one copy of the file on the storage pool's content routers. But PureDisk reports contain errors for the following statistics: Unique files and folders backed up and Unique bytes backed up. The erroneous statistics appear in both the job log and in the Statistics tab for a backup job. If a backup contains a duplicate of a file that is smaller than the small-file threshold, neither the job log nor the job statistics treat the second file as a duplicate. For example, when PureDisk backs up two identical files that are smaller than 16 KB in length, the following occur:
The backup job statistics erroneously report that two unique files have been backed up. Only one unique file has been backed up. The backup job's log file incorrectly reports that two new files have been stored. The log file should report one new file and one unchanged file.
The backup job statistics calculate the Global data reduction savings and the Global data reduction factor incorrectly. PureDisk calculates these statistics as if the files are different and no deduplication was possible. If a backup contains a duplicate of a file that is smaller than the segment size but larger than the small-file threshold, neither the job log nor the job statistics treat the second file as a duplicate. For example, when PureDisk backs up two identical files that are smaller than 128 KB in length, the following occur:
201
The backup job statistics erroneously report that two unique files have been backed up. Only one unique file has been backed up. The backup job's log file incorrectly reports that two new files have been stored. The log file should report one new file and one unchanged file.
Bytes total
Target: Files new on target The number of new files that reside on the client after the restore is complete. If you restore to the original directory and overwrite the original files, PureDisk reports that there are no new files. If you restore the files to a different directory for the first time, PureDisk reports that all the files you restored are new files on the client. The number of bytes occupied by the restored files. This statistic is the number of bytes consumed by the files that are noted in the Files new statistic.
Files modified on target The number of files that are different on PureDisk storage when compared to the target directory for the restore. This number counts the number of files on the client source that have different content when compared to the files you restored. Bytes modified on target The number of bytes occupied by the files in the Files modified on target statistic. Files unmodified on target The number of files that are identical on both PureDisk storage and on the target directory. For example, if this value is 0, this means that all the files you restored have changed since they were backed up.
202
The number of bytes actually restored. If nothing has been replaced, the value is 0. The average transfer rate during the transmission of unique data.
Data Uniqueness: Unique items restored The number of items that PureDisk wrote to the target computer. This statistic is a count of the number of files, directories, and special files. It includes only the items that were different on the target computer as compared to PureDisk storage. If nothing has changed, this value is 0. The number of unique items that were included in this restore job. The count excludes directories and special files.
Restore Failures: Error count Files with errors The number of errors that were generated during the restore. The number of files that generated errors during the restore and could not be restored. The total number of bytes represented in the files that had an error and could not be restored. For example, if a 1-MB file could not be restored due to an error, this statistic is 1 MB. The total number of errors encountered when the job attempted to restore ACLs. This value can be nonzero for a variety of reasons. For example, the following conditions, and others, can cause ACL restore errors:
ACL errors
An ACL could not be found on storage A parent directory does not allow restore of ACLs.
Verification failures
The number of files for which verification failed. This field is applicable only if you backed up the files with verification enabled.
Restore Successes:
203
/a/b/c/file1 /a/b/d/file2
In this case, the Directory count is 4. Devices (Linux and UNIX systems only) The number of block and character device files restored. This value is always 0 on Windows systems. (Linux and UNIX systems only) The number of symbolic links restored. This value is always 0 on Windows systems. (Linux and UNIX systems only) The number of hard links restored. This value is always 0 on Windows systems. (Windows systems only) The number of ACLs restored. This value is always 0 on Linux and UNIX systems. The number of files for which verification succeeded. This field is applicable only if you backed up the files with verification enabled.
Symbolic links
Hard links
ACL
Verification successes
Time: Start date/time Stop date/time Restore time duration The date and time that the job started. The date and time that the job ended. The amount of time that elapsed between when the job started and when the job ended.
204
Items new in source data The number of data objects replicated to the target storage pool selection that were not included in a previously replicated PureDisk backup. Bytes new in source data The number of bytes replicated to the target storage pool that selection have not been included in a previous PureDisk backup. Items modified in source The number of data objects replicated that have been modified data selection since the previous replication. This number counts the number of data objects on the source that have different content when compared to the files you replicated at an earlier time. Bytes modified in source The number of bytes occupied by the data objects in the Items data selection modified in source data selection statistic. Items deleted in source data selection Bytes deleted in source data selection The number of data objects that were deleted from the source data selection since the last replication. The number of bytes occupied by the data objects in the Items deleted in source data selection statistic. This statistic is the total number of bytes deleted from the source data selection. Not included in PDDO replication statistics. Errors: Items with replication errors The number of data objects that generated errors during the replication process. Not included in PDDO replication statistics. Bytes with replication errors The number of bytes of data in the Items with replication errors statistic that generated errors during the replication process. Not included in PDDO replication statistics. Replicated Data: Items replicated The number of files, directories, or data items that are replicated to the target storage pool. The number of bytes replicated to the target storage pool.
205
Time: Start date/time Stop date/time Replication time duration The date and time that the job started. The date and time that the job ended. The amount of time that elapsed between when the job started and when the job ended.
Media The percentage of backup data that PureDisk found in the media server's server cache. cache hit percentage Network:
206
Bytes The number of bytes of new, nondeduplicated data that PureDisk sent to the transferred content router for storage. to content router Time: Start date/time Stop date/time Backup time duration The date and time that the job started.
The amount of time that elapsed between when the job started and when the job ended.
207
Download
A link you can click to restore the file. Tip: You can restore a file by clicking a Download link in the Download column. You can download junction point files that PureDisk backed up on a Windows system. Individually, these files are of limited use. To restore files of this type, use the restore feature as described in the following manual: PureDisk Backup Operator's Guide
This screen contains no information when PureDisk does not back up any files. This situation is possible for an incremental backup if files have not changed.
Remark
The job is queued. The job step completed successfully.
208
Remark
The job step failed. The job step is running. A job step is preparing to run. The job step ran successfully but encountered nonfatal errors. The user stopped the job step. The PureDisk watchdog stopped the job step after the job step timed out. The job is running, but the current job step is on hold. Not all required fields were calculated. This local file type is not supported. File not found. Access denied. An ACL does not permit read access. On a restore operation, this message means that the disk is full. Otherwise, this means that a parent directory does not allow write access. The file is locked by another process. Applies to Windows clients only. Could not connect to the content router. PureDisk could not find the metadata for this file in the metabase engine.
7 8
ABORTED_BY_USER ABORTED_BY_WATCHDOG
RUNNING_HOLD
10
INCOMPLETE
103
UNWRITABLE_LOCALLY
104
LOCKED_LOCALLY
200
CR_CONNECTION_ERROR
201
NONEXISTENT_ON_SP
209
Remark
One of the following conditions is present: A read error occurred on the storage pool. The data selection does not allow read permission. This agent does not have read permission.
203
UNWRITABLE_TO_SP
One of the following conditions is present: A write error occurred on the storage pool. The data selection does not allow write permission. This agent does not have write permission.
204
UNKNOWN_CR_ERROR
This error is the generic content router error. For more information, see the agent or the server logs. Unused. See 201. See 202. See 203. Pdweb not running. Generic error log. For more information, see the agent or the server logs.
The value of 0 indicates that no limit is set. It does not mean that no data is transferred.
210
1 2
Click the Job log tab. Click Download Whole Job Log. This link appears above the upper-right corner of the right column.
On the pop-up window that appears, specify the interface you want to use to display the file. For the display method you choose, you might need to insert return characters because the return characters might not appear correctly in the display. In some cases, the job log might exceed the display length for this window. The job log includes log information at the beginning and at the end, which enables you to see what happened when the job started and when the job finished. In this case, PureDisk deletes repetitive information from the middle of the job log report.
211
See Running a data mining policy manually on page 212. See Obtaining data mining policy output - the data mining report on page 213. See Obtaining data mining policy output - the Web service report on page 216.
1 2 3 4
Click Manage > Policies. In the left pane, under Storage Pool Management Policies, click the plus (+) sign to the left of Data Mining. Select System policy for data mining. Complete the General tab. See Completing the General tab for a data mining policy on page 211.
Complete the Scheduling tab. See Completing the Scheduling tab for a data mining policy on page 212.
Complete the Parameters tab. See Completing the Parameters tab for a data mining policy on page 212.
Click Save.
(Optional) Type a new name for this policy in the Policy name field. You do not have to rename this policy.
If you select Enabled, PureDisk runs the policy according to the schedule in the Scheduling tab. If you select Disabled, PureDisk does not run the policy according to the schedule in the Scheduling tab. This value is the default. For example, you can use Disabled if you want to stop running this policy during a system maintenance period, but you do not want to enter
212
information in the Scheduling tab to suspend, and then reenable, this policy.
(Optional) Specify an escalation procedure. Select times in the Escalate warning after or the Escalate error and terminate after drop-down boxes to specify the elapsed time before PureDisk sends a message. PureDisk can notify you if an update does not complete within a specified time. For example, you can configure PureDisk to send an email message to an administrator if a policy does not complete in an hour. If you select either of these options, create a policy escalation action that defines the email message, defines its recipients, and associates the escalation action with the policy. For more information, see the PureDisk Backup Operators Guide.
Specify the schedule details that define how frequently you want the policy to run.
Choose a report format of Light (default) or Full. These settings specify whether PureDisk includes file extension information in the reports. File extensions include .mp3, .doc, .txt, and so on. When the Light setting is in effect, the report does not contain file extensions. When the Full setting is in effect, the report includes all file extension information. The Full setting also increases the load on the storage pool, particularly the metabase engines.
Reports Obtaining data mining policy output - the data mining report
213
1 2 3 4
Click Manage > Policies. In the left pane, under Storage Pool Management Policies, click the plus sign (+) to the left of Data Mining. Select System policy for data mining. (Conditional) Enable the policy. You must enable a policy before you can run it. If the policy is disabled, on the General tab, click Enabled and click Save.
5 6
In the right pane, click Run Policy. Examine the output. For information about how to examine the output, see one of the following: See Obtaining data mining policy output - the data mining report on page 213. See Obtaining data mining policy output - the Web service report on page 216.
Make sure that you have the correct permissions to create and view reports. The data mining report shows only the data selections a user is entitled to view. See Permissions and guidelines for running and viewing reports on page 190.
Make sure that a data mining policy has been run. You can run the policy manually or you can configure PureDisk to run the policy on a schedule. For information about how to run a data mining policy, see the following: See Running a data mining policy manually on page 212.
3 4
Click Manage > Agents. In the left pane, select the scope of the data mining report. You can obtain a data mining report on one of the following levels:
214
Reports Obtaining data mining policy output - the data mining report
5 6
Click Data Mining Report in the right pane. (Optional) Click Select in history in the upper right corner. Perform this step if you want to view a data mining report from an earlier data mining workflow. By default, PureDisk displays data mining information from the most recent run of the data mining workflow. The following list explains some of the information in the data mining report:
Total Storage Pool Volume Used The volume of backup data, in bytes, on the content routers in this storage pool. The information in this field is updated every 15 minutes. The output in this field might not account for the data that was added to storage during the last 15 minutes. During installation, the agents are stored on the storage pool. Consequently, if you run a data mining policy before any backups have run, the report indicates that a small amount of storage is already in use. Total Storage Pool Data Reduction Factor The volume of all data ever backed up to this storage pool, in bytes, that is retained and currently available for restores divided by the global storage pool volume. See Interpreting the storage pool data reduction factor on page 215. Total size on source The volume of files, in bytes, in this data selection on the source client. This number includes all versions of all files. The estimated data volume, in bytes, stored on the storage pools content routers for this data selection. This statistic is the source size of this data selection divided by the storage pool data reduction factor.
Reports Obtaining data mining policy output - the data mining report
215
If this factor is 1, the backed up source files consume the same capacity on the storage pool content routers as on the source clients. If this factor is greater than 1, this value is the factor by which PureDisk has reduced the source volume through data reduction before it writes to the content routers. If this factor is less than 1, the backed up data consumes more space on the content routers compared to the backed up volume of the source files. That can be due to compression and to encryption overhead.
For example, assume that PureDisk backed up eight 10-MB files within a data selection. The eight files had identical content, and this particular content is new to the storage pool. PureDisk determines that these identical files all have the same fingerprint. Because they all have the same fingerprint, PureDisk stores only one copy on the content routers. The statistics are as follows:
The volume on source for this data selection is 8 X 10 MB = 80 MB (source size). The volume on the storage pool for this data selection is 10 MB (storage pool size). The storage pool data reduction factor is 80/10 MB = 8.
216
Reports Obtaining data mining policy output - the Web service report
A smaller segment size can yield better data reduction rates. However, performance can degrade because of the higher maintenance costs that are involved in managing a larger number of segments. A larger segment size can yield better performance, but the data reduction rate can degrade. Larger segments can also use a higher amount of disk space. PureDisk considers the following factors when it segments the file:
The default segment size for the data selection type or the segment size you specify. The maximum number of segments allowed, which is 5,120 segments. The maximum segment size allowed, which is 16 MB.
Table 11-6 shows the arguments in the URL. Table 11-6 Argument
url
login pwd
Reports Obtaining data mining policy output - the Web service report
217
When you run the Web service report to obtain data mining output, you retrieve information on all data selections in the storage pool. You cannot narrow the report to include information for only one data selection. Information about how to report on only one data selection is available. See Obtaining data mining policy output - the data mining report on page 213. For example, assume that you type the following URL:
https://valhalla.minnesota.com/spa/ws/ws_datamining.php?login=root&passwd=root&action= getReport&runid=1
218
Reports Obtaining data mining policy output - the Web service report
-<ACCESSRANGE> -<item id="-1 day"> <amountoffiles>13</amountoffiles> <totalfilesize>33017267</totalfilesize> </item> -<item id="1 day-1 week"> <amountoffiles>16</amountoffiles> <totalfilesize>60604429</totalfilesize> </item> -<item id="1 month-1 year"> <amountoffiles>45</amountoffiles> <totalfilesize>56263445</totalfilesize> </item> -<item id="1 week-1 month"> <amountoffiles>4</amountoffiles> <totalfilesize>3520124</totalfilesize> </item> </ACCESSRANGE> -<MODRANGE> -<item id="+1 year"> <amountoffiles>10</amountoffiles> <totalfilesize>1814927</totalfilesize> </item> -<item id="-1 day"> <amountoffiles>2</amountoffiles> <totalfilesize>29874649</totalfilesize> </item> -<item id="1 day-1 week"> <amountoffiles>1</amountoffiles> <totalfilesize>207</totalfilesize> </item> -<item id="1 month-1 year"> <amountoffiles>61</amountoffiles> <totalfilesize>70561729</totalfilesize> </item> -<item id="1 week-1 month"> <amountoffiles>4</amountoffiles> <totalfilesize>51153753</totalfilesize> </item> </MODRANGE> -<SIZERANGE> -<item id="0-10KB"> <amountoffiles>20</amountoffiles>
219
<totalfilesize>26942</totalfilesize> </item> -<item id="100KB-1MB"> <amountoffiles>19</amountoffiles> <totalfilesize>7200946</totalfilesize> </item> -<item id="10KB-100KB"> <amountoffiles>29</amountoffiles> <totalfilesize>1745269</totalfilesize> </item> -<item id="10MB-100MB"> <amountoffiles>4</amountoffiles> <totalfilesize>129892970</totalfilesize> </item> -<item id="1MB-10MB"> <amountoffiles>6</amountoffiles> <totalfilesize>14539138</totalfilesize> </item> </SIZERANGE> -<TYPES> -<item id="0"> <amountoffiles>78</amountoffiles> <totalfilesize>153405265</totalfilesize> </item> </TYPES> </dataselection> </mbe> </mbe_range_statistics> -<dataselectionlist_SIS_reporting> <global_storagepool_VOL unit="bytes">4176478208</global_storagepool_VOL> <global_storagepool_SIS>0.03673077108511</global_storagepool_SIS> </dataselectionlist_SIS_reporting> <MBDataminingHistory/> </MBDatamining>
220
format. You can import the XML output to a spreadsheet. See the following section for more information: See Importing report output into a spreadsheet on page 227. Caution: For security reasons, use a Web browser that uses POST requests, not GET requests, when retrieving Web service reports. For example, Microsoft Internet Explorer does not use POST requests and is not secure. You can also follow your spreadsheets instructions for importing the XML data. For example, the following URL contains logon information, password information, and a request for information about successful job runs:
https://100.100.100.100/spa/ws/ws_getsuccessfuljobs.php?login=root&passwd=root
Note: The administrative Web UI's URL parameters are case-sensitive . Make sure that you type them exactly as shown in this chapter. The ampersand (&) character acts as a separator for the fields in the URL. The bracket characters in the following sections [ ] represent optional URL fields. The following sections describe the reports that you can obtain through the Web services:
See Job status Web service reports on page 220. See Dashboard Web service reports on page 223. See Obtaining data mining policy output - the Web service report on page 216.
ws_getfailedjobs.php
The URL format for a Web service report on job statuses is as follows:
https://url/spa/ws/web_service?login=login&passwd=pwd[&filter][&filter]
221
Table 11-7 shows the arguments in the URL. Table 11-7 Argument
url
web_service
login
The storage pool authority administrator logon. For example: root. The storage pool authority administrator password. For example: mypwd. (Optional) One or more filters. If you specify a filter, the report displays only the data that matches the filter. If you specify more than one filter, use the ampersand (&) character to separate each filter. The filter names are case-sensitive . See Table 11-8 on page 221.
pwd
filter
Table 11-8 shows the filters you can specify on a Web service URL for the job status reports. Table 11-8 Filter
locationName=name
departmentName=name
222
fromDate=mm-dd-yyyy
Returns only the jobs that started on or after the specified date. For example: 06-30-2007. Returns only the jobs that ended on or before the specified date. For example: 10-05-2007. Returns only the jobs for a particular workflow. To see the list of possible values for name, click Manage > Policies and observe the left pane of the administrative Web UI. This pane shows the list of possible policies and workflows. Specify the policy or the workflow name as shown in the administrative Web UI. Examples:
toDate=mm-dd-yyyy
workflowName=name
For example, assume that you want to examine statistics for restore jobs. You can enter the following URL:
https://100.100.100.100/spa/ws/ws_getsuccessfuljobs.php?login=root&passwd=root&workflo wName=Files and Folders Restore
223
<executionStatusID>2</executionStatusID> <executionStatusName>SUCCESS</executionStatusName> <workflow>Restore Workflow</workflow> <scheduledStartTime>1151762912</scheduledStartTime> <startDate>1151762916</startDate> <finishDate>1151762970</finishDate> <dataselectionID>2</dataselectionID> <dataselectionName>reroute</dataselectionName> <statistics /> </job> - <job> <jobID>4</jobID> <agentID>1000000</agentID> <agentName>SPA</agentName> <locationName>my location</locationName> <departmentName>my department</departmentName> <executionStatusID>2</executionStatusID> <executionStatusName>SUCCESS</executionStatusName> <workflow>Restore Workflow</workflow> <scheduledStartTime>1151763438</scheduledStartTime> <startDate>1151763439</startDate> <finishDate>1151763545</finishDate> <dataselectionID>2</dataselectionID> <dataselectionName>reroute</dataselectionName> <statistics /> </job> . . .
The preceding output has been truncated at the end for inclusion in this manual. If you run a report that contains information about backup jobs, the information PureDisk returns contains the same statistics that you can obtain from clicking Data Mining Report in the left pane after a data mining workflow was run.
224
You can write an application to parse or extract pieces of information from this reports output and show that in a Web page as a custom dashboard. The information for this report is not generated in real time. PureDisk refreshes the data every 15 minutes. The timestamp is shown at the beginning of the XML report. The report is formatted in XML. The URL format is as follows:
https://url/spa/ws/ws_dashboard.php?login=login&passwd=pwd&filterType=type&filterID=id &action=getDashBoard
Note: Type the preceding URL on one continuous line. Table 11-7 shows the arguments in the URL. Table 11-9 Argument
url
login pwd
type
225
In the left pane, select an agent, a department, a location, or the storage pool. Visually inspect the right pane to obtain the identification number from the Agent ID, Department ID, Location ID, or Storage Pool ID fields. Tip: Obtain that id number before you start to type the URL for the Web service report. If you begin to type the report URL into a browsers address field, and have to click in the PureDisk administrative Web UI to retrieve that id information, you lose the information you typed into the address field. Alternatively, you can also retrieve the id in a different window.
For example, assume that you want to obtain a dashboard Web service report for an agent. You can enter the following URL:
https://valhalla.minnesota.com/spa/ws/ws_dashboard.php?login=root&passwd=root&filterTy pe=agent&filterID=3&action=getDashBoard
226
<Status id="2">ACTIVE</Status> <IPAddress>10.80.139.49</IPAddress> <OS id="20">Linux</OS> <Version>6.5.0.8987</Version> <Department id="1">qe</Department> <Location id="1">mn</Location> <MetabaseEngine id="1" agentid="33000000">10.80.139.49</MetabaseEngine> <Controller id="1" agentid="33000000">10.80.139.49</Controller> <ConnectionStatus>Connected</ConnectionStatus> <ConnectionDetails> <FromIP>10.80.139.49</FromIP> <SessionID>pdagent</SessionID> <Sent unit="MB">45.11</Sent> <Received unit="MB">1.34</Received> <Version/> </ConnectionDetails> <Jobs> <Job id="77"> <JobID>77</JobID> <AgentID>33000000</AgentID> <Workflow id="13500">Maintenance</Workflow> <Policy id="8">System policy for Maintenance</Policy> <PolicyRunID>28</PolicyRunID> <Scheduled>2008-03-15 06:20:01</Scheduled> <Start>2008-03-15 06:20:03</Start> <Stop>2008-03-15 06:20:23</Stop> <Status id="2">SUCCESS</Status> </Job> </Jobs> <JobSteps/> <Statistics id="33000000" TimeStamp="2008-03-15 09:30:01" xml:base="/Storage/var/stats_33000000.xml"> . . .
The preceding output has been truncated at the end for inclusion in this manual.
227
1 2
Use your Web browser to save the output as an XML file. Import the data into your spreadsheet. For example, in Microsoft Excel, specify Data > Import External Data > Import Data. When it prompts you, specify the file to which you saved the XML output.
See Displaying the Capacity dashboard on page 227. See Displaying the Activity dashboard on page 228. See Displaying the Server agent dashboard on page 229. See Displaying the Client agent dashboard on page 230.
Dashboard reports are available when a central reporting storage pool authority is installed. See Central storage pool authority reports on page 231.
228
backup completes, the information in the display might not reflect the conditions that result from that backup. To display the capacity dashboard
1 2 3
Click Settings > Topology. In the left pane, select the storage pool. In the right pane, click Capacity Dashboard. The following figure shows an example capacity dashboard.
229
1 2 3
Click Settings > Topology. In the left pane, select the storage pool. In the right pane, click Activity dashboard. The following figure shows an example activity dashboard.
230
1 2 3
Click Settings > Topology. In the left pane, select the storage pool. In the right pane, click Server agent dashboard. The following figure shows an example server agent dashboard.
1 2
Click Manage > Agent. Expand the tree view in the left pane until the department or agent you want displays. Click the plus sign (+) next to each entity to expand the tree.
231
3 4
Select a department or agent. Click Agent Dashboard in the right pane. The following figure shows an example agent dashboard.
See Displaying the Central Reporting dashboard on page 231. See Updating the Central Reporting dashboard on page 236. See About central reporting on page 285.
232
Verify that you have the Central Report permission. Only users with Central Report permissions have rights to view and update the reports. For more information about permissions, see the PureDisk Client Installation Guide.
2 3 4 5
Click Settings > Central SPA. In the left pane, click Storage Pool Management. In the right pane, click Central SPA Dashboard. Click one of the following tabs in the dashboard display:
Enterprise License Report (default view). See Enterprise License Report tab on page 232. Storage Pools. See Storage Pools tab on page 234. Licenses / Features. See Licenses \ Features tab on page 234. Capacity Usage Report. See Capacity Usage Report tab on page 235.
233
Licensed
For a feature license row, this column lists the number of clients that can use this feature. For the PureDisk edition row, this column lists the amount of front-end data, on the client, that you can protect with PureDisk backups.
Used
For a feature license row, this column lists the number of clients that currently use this feature. For the PureDisk edition row, this column displays the amount of front end, client storage that this storage pool currently protects.
Alerts
Displays alerts under the following conditions: When the used capacity is greater than the licensed capacity. When the number of features used is greater than the number of features licensed.
234
235
Locations
To view a licenses and features report for a particular license type or all types
Use the Filter on feature pull-down menu to select a license type. Your choices are as follows:
All Premium Infrastructure Windows Application & Database Pack Standard Agent
Used capacity
Standard Agents
236
Chapter
12
About the log file directory Audit trail reporting Setting debugging mode
For each seven-day interval, PureDisk retains up to 1000 lines of logging messages in the active log file in /Storage/log. Note that log files from PureDisk services are often greater than 5 MB in length, but PureDisk does not retain job log files that are greater than 5 MB in length. PureDisk uses the standard Linux log rotation mechanism to rotate the audit log every seven days. Log rotation ensures that the log files do not become too large. PureDisk moves old logging information into separate files and compresses the files to save space. PureDisk does not remove old log files. You can examine the old log files in /Storage/log. The old files are named /Storage/log/audit.log.1.bz2, /Storage/log/audit.log.2.bz2, and so on. The last 1000 lines of every log file are always accessible in the /Storage/log directory. The following describe log files:
See Content router log files on page 238. See Metabase engine log file on page 241. See Workflow engine log file on page 244.
238
See Server agent log files on page 247. See About international characters in log files on page 248.
See The spoold.log file on page 238. See The storaged.log file on page 239. See Logging and debugging options on page 240.
January 17 16:10:11 INFO [1076910400]: Task Manager: started task 0 [thread 1079552320] for 192.168.163.1:1636 January 17 16:10:11 INFO [1079552320]: Remote is using libcr Version 6.5.0.6792, Protocol Version 6.1 running on WIN32. Agent pdbackup.exe requesting access for DataSelection ID 7
Example 2. The following shows the metabase engine (192.168.163.132 = MBE IP) requesting a POList (MBE-CLI application) from system data selection 1:
January 17 16:10:16 INFO [1076910400]: Task Manager: started task 0 [thread 1079552320] for 192.168.163.132:51050 January 17 16:10:16 INFO [1079552320]: Remote is using libcr Version 6.5.0.6792, Protocol Version 6.1 running on Linux-x86_64. Agent MBE-CLI requesting access for DataSelection ID 1
If you want an overview of all incoming single-stream backups (PutFiles), you can search the spoold.log file, as follows:
239
January 17 14:10:12 INFO [1076910400]: Task Manager: started for 192.168.163.1:3738 January 17 14:10:12 INFO [1079552320]: Remote is using libcr Version 6.1 running on WIN32. Agent PutFiles requesting access for DataSelection ID -January 17 15:10:10 INFO [1076910400]: Task Manager: started for 192.168.13.41:4438 January 17 15:10:10 INFO [1079552320]: Remote is using libcr Version 6.1 running on WIN32. Agent PutFiles requesting access for DataSelection ID -January 17 16:10:11 INFO [1076910400]: Task Manager: started for 192.168.163.14:1638 January 17 16:10:11 INFO [1079552320]: Remote is using libcr Version 6.1 running on WIN32. Agent PutFiles requesting access for DataSelection ID -January 17 17:10:13 INFO [1076910400]: Task Manager: started for 192.168.163.1:2201 January 17 17:10:13 INFO [1079552320]: Remote is using libcr Version 6.1 running on WIN32. Agent PutFiles requesting access for DataSelection ID
task 0 [thread 1079552320] Version 6.5.0.6792, Protocol 1 task 0 [thread 1079552320] Version 6.5.0.6792, Protocol 7 task 0 [thread 1079552320] Version 6.5.0.6792, Protocol 4 task 0 [thread 1079552320] Version 6.5.0.6792, Protocol 9
In the preceding grep(1) command, the -B 1 parameter specifies to show the line before the match, so the connecting client IP address is also displayed.
August 25 14:58:45 INFO [1077967168]: Queue processing triggered by external request. August 25 14:58:45 INFO [1077967168]: Starting sort of tlog file range 521 - 525. August 25 14:58:45 INFO [1077967168]: Finished sort of tlog file range 521 - 525 in 0 seconds.
240
August 25 14:58:45 INFO [1077967168]: Preparing to process transaction log /Storage/queue/sorted-521-525.tlog August 25 14:58:45 INFO [1077967168]: Synchronization for transaction log /Storage/queue/sorted-521-525.tlog started, 14306 transactions pending. August 25 14:58:49 INFO [1077967168]: Number of data store commits: 225 August 25 14:58:49 INFO [1077967168]: Time required to build index on objects2 table: 0.211361 August 25 14:58:49 INFO [1077967168]: Time required to drop objects table: 0.013812 August 25 14:58:49 INFO [1077967168]: Time required to rename objects2 table to objects: 0.000773 August 25 14:58:49 INFO [1077967168]: Transaction log 521-525 Completed. Expect: 14306 (0.82MB) Commit: 14306 (0.00MB) Retry: 0 Log: /Storage/queue/sorted-521-525.tlog SO: Add 0, Ref Add 14298, Ref Add Fail: 0, Ref Del 0 DO: Add 0, Ref Add 0, Ref Add Fail: 0, Ref Del 4 TASK: Add 2, End 2, End All 0, Del 0 DCID: SO 0, SO Fail 0, DO 0, DO Fail 0 MARKER: 0, Fail 0 August 25 14:58:49 INFO [1077967168]: Update last committed tlogid from 520 to 525 August 25 14:58:49 INFO [1077967168]: Start processing delayed operations of '/Storage/queue/sorted-521-525.delayed'. August 25 14:58:49 INFO [1077967168]: Completed processing of delayed operations of '/Storage/queue/sorted-521-525.delayed'.
241
In the PureDisk administrative Web UI, click Settings > Configuration > Configuration File Templates > PureDisk Content Router > Default Value Set for PureDisk ContentRouter > Logging > Logging. Change the All OS: value to full, thread. Type the following command to restart the content router:
2 3
If you want to specify that the log files include more information, include the --trace parameter when you restart the content router. For example:
If you specified the --trace parameter, later you can specify the following to disable tracing:
# /etc/init.d/puredisk restart pdcr
Sat Jan 12 22:51:33 CST 2008 <INFO> [Task [2-1200199892625]] (DISPATCHER) Dispatcher has set a task to approved. Sat Jan 12 22:51:34 CST 2008 <INFO> [Task [2-1200199892625]] (DOWNLOAD 0) Download: A job has arrived! Sat Jan 12 22:51:40 CST 2008 <INFO> [Task [2-1200199892625]] (DOWNLOAD 0) Download: finished! Sat Jan 12 22:51:41 CST 2008 <INFO> [Task [2-1200199892625]] (DISPATCHER) Dispatcher has set a task to approved.
242
Sat Jan 12 22:51:42 CST 2008 <INFO> [Task [2-1200199892625]] (SORT0) Starting sort for /Storage/tmp/pre57295.tmp Sat Jan 12 22:51:43 CST 2008 <INFO> [Task [2-1200199892625]] (DISPATCHER) Dispatcher has set a task to approved. Sat Jan 12 22:51:44 CST 2008 <INFO> [Task [2-1200199892625]] (SPLIT0) Starting to split Sat Jan 12 22:51:47 CST 2008 <INFO> [Task [2-1200199892625]] (SPLIT0) Splitting file Sat Jan 12 22:51:48 CST 2008 <INFO> [Task [2-1200199892625]] (SPLIT0) Generating dirfile Sat Jan 12 22:51:49 CST 2008 <INFO> [Task [2-1200199892625]] (SPLIT0) Done converting Sat Jan 12 22:51:49 CST 2008 <INFO> [Task [2-1200199892625]] (DISPATCHER) Dispatcher has set a task to approved. Sat Jan 12 22:51:50 CST 2008 <INFO> [Task [2-1200199892625]] (IMPORT0) ImportThread has work to do. Sat Jan 12 22:51:50 CST 2008 <INFO> [Task [2-1200199892625]] (IMPORT0) ImportThread going to import /Storage/tmp/bulkInsert57297 Sat Jan 12 22:51:50 CST 2008 <INFO> [Task [2-1200199892625]] (IMPORT0) ImportThread going to import a raw POlist Sat Jan 12 22:51:51 CST 2008 <INFO> [Task [2-1200199892625]] (IMPORT0) ImportThread going to import a raw POlist Sat Jan 12 22:51:51 CST 2008 <INFO> [Task [2-1200199892625]] (IMPORT0) Import done Sat Jan 12 22:51:51 CST 2008 <INFO> [Task [2-1200199892625]] (DISPATCHER) Dispatcher has set a task to approved. Sat Jan 12 22:51:52 CST 2008 <INFO> [Task [2-1200199892625]] (EVAL0) Evaluating DataSelection 2. Sat Jan 12 22:51:53 CST 2008 <INFO> [Task [2-1200199892625]] (EVAL0) No duplicate or minor PO's were detected. Sat Jan 12 22:51:53 CST 2008 <INFO> [Task [2-1200199892625]] (EVAL0) DataSelection 2 succesfully evaluated. Sat Jan 12 22:51:55 CST 2008 <INFO> (DEPARTSERVLET) Task has completed:Task [2-1200199892625] Sat Jan 12 22:55:38 CST 2008 <INFO> [Task [4-1200200137663]] (DISPATCHER) Dispatcher has set a task to approved. Sat Jan 12 22:55:38 CST 2008 <INFO> [Task [4-1200200137663]] (DOWNLOAD 0) Download: A job has arrived! Sat Jan 12 22:55:39 CST 2008 <INFO> [Task [4-1200200137663]] (DOWNLOAD 0) Download: finished! Sat Jan 12 22:55:40 CST 2008 <INFO> [Task [4-1200200137663]] (DISPATCHER) Dispatcher has set a task to approved. Sat Jan 12 22:55:40 CST 2008 <INFO> [Task [4-1200200137663]]
243
(SORT1) Starting sort for /Storage/tmp/pre57303.tmp Sat Jan 12 22:55:42 CST 2008 <INFO> [Task [4-1200200137663]] (DISPATCHER) Dispatcher has set a task to approved. Sat Jan 12 22:55:42 CST 2008 <INFO> [Task [4-1200200137663]] (SPLIT0) Starting to split Sat Jan 12 22:55:42 CST 2008 <INFO> [Task [4-1200200137663]] (SPLIT0) Splitting file Sat Jan 12 22:55:42 CST 2008 <INFO> [Task [4-1200200137663]] (SPLIT0) Generating dirfile Sat Jan 12 22:55:42 CST 2008 <INFO> [Task [4-1200200137663]] (SPLIT0) Done converting Sat Jan 12 22:55:43 CST 2008 <INFO> [Task [4-1200200137663]] (DISPATCHER) Dispatcher has set a task to approved. Sat Jan 12 22:55:43 CST 2008 <INFO> [Task [4-1200200137663]] (IMPORT0) ImportThread has work to do. Sat Jan 12 22:55:43 CST 2008 <INFO> [Task [4-1200200137663]] (IMPORT0) ImportThread going to import /Storage/tmp/bulkInsert57305 Sat Jan 12 22:55:43 CST 2008 <INFO> [Task [4-1200200137663]] (IMPORT0) ImportThread going to import a raw POlist Sat Jan 12 22:55:43 CST 2008 <INFO> [Task [4-1200200137663]] (IMPORT0) ImportThread going to import a raw POlist Sat Jan 12 22:55:43 CST 2008 <INFO> [Task [4-1200200137663]] (IMPORT0) Import done Sat Jan 12 22:55:44 CST 2008 <INFO> [Task [4-1200200137663]] (DISPATCHER) Dispatcher has set a task to approved. Sat Jan 12 22:55:44 CST 2008 <INFO> [Task [4-1200200137663]] (EVAL0) Evaluating DataSelection 4. Sat Jan 12 22:55:44 CST 2008 <INFO> [Task [4-1200200137663]] (EVAL0) No duplicate or minor PO's were detected. Sat Jan 12 22:55:44 CST 2008 <INFO> [Task [4-1200200137663]] (EVAL0) DataSelection 4 succesfully evaluated.
If your log file is large, you can search for the information you want. For example, type the following command to display all imports for data selection 7:
PureDisk:/Storage/log # grep 'Task \[7-' mbe.log
The metabase engine disk evaluator logs disk usage every 5 minutes. For example:
Thu Jan 17 17:28:57 CST 2008 <INFO> (DISKEVALUATOR0) Evaluating left disk space Thu Jan 17 17:28:58 CST 2008 <INFO> (DISKEVALUATOR0) Diskspace used on partition with the metabase database is: 28.0%.
244
Thu Jan 17 17:29:00 CST 2008 <INFO> (DISKEVALUATOR0) Diskspace used on partition with metabase tmp folder is: 28.0%.
See The pdwfe.log file on page 244. See Logging and debugging options on page 246.
The pdwfe.log file contains information about the following common workflow engine actions:
245
Thu Jan 17 2008 14:00:47.645449 INFO Thu Jan 17 2008 14:00:47.649075 INFO
About agents when they request the next job step (nextJobStep web service):
Thu Jan 17 2008 14:05:16.667128 INFO (1074796864): Agent 'PureDisk' (id: 379000000): no jobstep found
17 2008 14:10:02.782948 INFO (1075325248): Return Jobstep 'PrepareBackup.php' (id: 155) to Agent 'ros2pc00' (id: 2) 17 2008 14:10:07.152421 INFO (1074268480): Update status of jobstep 155 from RUNNING to SUCCESS
You can retrieve log information that is related to a single job. For example, to obtain workflow engine log information that is related to job ID 24, type the following command:
PureDisk:/Storage/log # grep 'Job 24' pdwfe.log Thu Jan 17 2008 14:10:02.178834 INFO (1075853632): Job 24: Created 'Files and Folders Backup' for Agent 'ros2pc00' (id: 2) Thu Jan 17 2008 14:10:02.782948 INFO (1075325248): Job 24: Return Jobstep 'PrepareBackup.php' (id: 155) to Agent 'ros2pc00' (id: 2) Thu Jan 17 2008 14:10:07.152421 INFO (1074268480): Job 24: Update status of jobstep 155 from RUNNING to SUCCESS Thu Jan 17 2008 14:10:07.703998 INFO (1074796864): Job 24: Return Jobstep 'ScanFilesystem.php' (id: 156) to Agent 'ros2pc00' (id: 2) Thu Jan 17 2008 14:10:09.478683 INFO (1075325248): Job 24: Update status of jobstep 156 from RUNNING to SUCCESS Thu Jan 17 2008 14:10:09.980596 INFO (1074268480): Job 24: Return Jobstep 'PutFiles.php' (id: 157) to Agent 'ros2pc00' (id: 2) Thu Jan 17 2008 14:10:11.406297 INFO (1075853632): Job 24: Update Variables Thu Jan 17 2008 14:10:11.411698 INFO (1075853632): Job 24: Update Variables Thu Jan 17 2008 14:10:11.739674 INFO (1074268480): Job 24: Update Variables Thu Jan 17 2008 14:10:12.770142 INFO (1074796864): Job 24: Update Variables Thu Jan 17 2008 14:10:12.974254 INFO (1075325248): Job 24:
246
Update Variables Thu Jan 17 2008 14:10:12.998051 INFO (1075325248): Job 24: Update Variables Thu Jan 17 2008 14:10:14.314949 INFO (1074268480): Job 24: Update status of jobstep 157 from RUNNING to SUCCESS_WITH_ERRORS Thu Jan 17 2008 14:10:14.674194 INFO (1074796864): Job 24: Return Jobstep 'MBImportAction.php' (id: 158) to Agent 'PureDisk' (id: 379000000) Thu Jan 17 2008 14:10:24.548564 INFO (1074796864): Job 24: Update Variables Thu Jan 17 2008 14:10:24.732565 INFO (1075325248): Job 24: Update Variables Thu Jan 17 2008 14:10:25.044541 INFO (1075853632): Job 24: Update status of jobstep 158 from RUNNING to SUCCESS Thu Jan 17 2008 14:10:25.428563 INFO (1074268480): Job 24: Return Jobstep 'ProcessJobStatistics.php' (id: 159) to Agent 'PureDisk' (id: 379000000) Thu Jan 17 2008 14:10:26.186136 INFO (1075325248): Job 24: Update Variables Thu Jan 17 2008 14:10:26.602925 INFO (1075853632): Job 24: Update status of jobstep 159 from RUNNING to SUCCESS Thu Jan 17 2008 14:10:27.153563 INFO (1074796864): Job 24: Return Jobstep 'FinishBackup.php' (id: 160) to Agent 'ros2pc00' (id: 2) Thu Jan 17 2008 14:10:29.298115 INFO (1075853632): Job 24: Update status of jobstep 160 from RUNNING to SUCCESS Thu Jan 17 2008 14:10:29.504469 INFO (1085315392): Job 24: Process Workflow Engine Job Step 'markexit' (id :161) Thu Jan 17 2008 14:10:29.514389 INFO (1085315392): Job 24: Workflow Engine: Step 161 returns SUCCESS
247
See The Agent.log file on page 247. See The job step log on page 248. See Logging and debugging options on page 248.
Thu Jan 17 2008 16:10:23.899687 INFO (1074796864): Incoming request: kick Thu Jan 17 2008 16:10:24.512531 INFO (1080609088): Jobstep: ProcessJobStatistics.php Thu Jan 17 2008 16:10:24.512951 INFO (1080609088): Logfile path: '/Storage/tmp/workflow.172' Thu Jan 17 2008 16:10:25.053758 INFO (1080609088): Updating status for job step #172 Thu Jan 17 2008 16:10:25.054380 INFO (1080609088): Upload logfile /Storage/tmp/workflow.172 (1382 bytes) using SPA webservice. Thu Jan 17 2008 16:10:25.575934 INFO (1080609088): Jobstep 172 successfully set to status 2 Thu Jan 17 2008 17:10:15.702997 INFO (1074268480): Incoming request: kick Thu Jan 17 2008 17:10:16.433983 INFO (1077438784): Jobstep: MBImportAction.php Thu Jan 17 2008 17:10:16.434432 INFO (1077438784): Logfile path: '/Storage/tmp/workflow.178' Thu Jan 17 2008 17:10:25.277339 INFO (1077438784): Updating status for job step #178 Thu Jan 17 2008 17:10:25.278032 INFO (1077438784): Upload logfile /Storage/tmp/workflow.178 (1445 bytes) using SPA webservice. Thu Jan 17 2008 17:10:25.806503 INFO (1077438784): Jobstep 178 successfully set to status 2
248
In this example, there are two job steps processes: ProcessJobStatistics and MBImportAction. All log lines that relate to these job steps have the same thread ID: 1080609088 for ProcessJobStatistics and 1077438784 for MBImportAction.
1 2 3 4 5
Click Monitor > Jobs. In the right pane, click the number in the Job Id column that corresponds to the job that contains the job step that you want to examine. In the pop-up that appears, click the Details tab. On the Details tab, click on the row that describes the job step you want to examine. On the left pane of the Details tab, note the jobid information. If necessary, use the pull-down menu to select Normal (the default), Verbose, Very Verbose, or Show All to display differing amounts of information.
If you use international characters to specify names, descriptions, and other labels in the storage pool If the PureDisk agent is installed on a localized client
249
PureDisk displays these characters correctly when you view log files, such as job logs, through the administrative Web UI. You can view the server logs stored in /Storage/log on a PureDisk node. However, you might need to make some configuration changes depending on where and how you want to view these log files. These configuration changes are as follows:
If you log on to the PureDisk node with a Secure Shell connection (SSH) on Linux or UNIX, make sure that you use a UTF-8 locale. For example, use en_US.UTF-8. You can log on to the PureDisk node with a Windows terminal client such as Putty. Ensure that the terminal client uses the UTF-8 character set and a font that contains the international characters that you need to display. If you log on to the PureDisk node directly through the console, PureDisk does not display international characters properly. Use one of the previous methods to view log files with international characters.
250
Figure 12-1
1 2 3
Click Manage > Agent. In the left pane, select World. In the right pane, click Show Audit Trail or Download Audit Trail. Alternatively, you can click Show Audit Trail and then click Download Audit Trail if you decide later that you want to download the information. The following information pertains to the output formats available to you from the right pane:
If you click Download Audit Trail, PureDisk downloads the report to a compressed file that ends in .tgz.
251
If you click Show Audit Trail, the audit trail appears in the right pane. The following three icons appear above the Object Name column:
The printer icon. If you click the printer icon, follow the instructions in the dialog boxes that appear to select a printer and send the report to that printer. The spreadsheet icon. If you click the spreadsheet icon, a dialog box appears. Click OK. Follow the instructions in the next dialog box to write or save these files. The refresh icon.
See Enabling debugging mode on page 251. See Disabling debugging mode on page 252. See Removing temporary debugging files on page 253.
1 2 3 4
Click Manage > Agent. In the left pane, select the agent for which you want to enable debug mode. In the right pane, pull down More Tasks, and select Set Debug Mode. Run the job that you want to troubleshoot.
252
Analyze the log files. The location of the temporary debugging files depends on your platform, as follows:
On Windows platforms, PureDisk writes debugging files to install_dir\Symantec\NetBackup PureDisk Agent\tmp. For example, C:\ Program Files\Symantec\NetBackup PureDisk Agent\tmp. Linux, UNIX, or MacOS, PureDisk writes debugging files to /opt/pdagent/tmp.
Disable debugging mode. For information about how to disable debugging mode, see the following: See Disabling debugging mode on page 252.
1 2
Click Settings > Topology. In the left pane, select the server agent for which you want to enable debug mode. The node identifiers appear under the storage pool name. For example, if there are multiple server agents, multiple node identifiers appear under the storage pool name in the left pane. The node identifier can be an FQDN, host name, or IP address.
3 4 5 6
In the right pane, select Set Debug Mode. Run the job that you want to troubleshoot. Analyze the log files in /Storage/tmp. Disable debugging mode. For information about how to disable debugging mode, see the following: See Disabling debugging mode on page 252.
1 2
Click Manage > Agent. In the left pane, select the agent for which you want to disable debug mode.
253
3 4
In the right pane, pull down More Tasks, and select Reset Debug Mode. Remove the temporary debugging files. For information about how to remove the temporary files, see the following: See Removing temporary debugging files on page 253.
1 2
Click Settings > Topology. In the left pane, select the server agent for which you want to disable debugging mode. The node identifiers appear under the storage pool name. For example, if there are multiple server agents, multiple node identifiers appear under the storage pool name in the left pane. The node identifier can be an FQDN, host name, or IP address.
3 4
In the right pane, select Reset Debug Mode. Remove the temporary debugging files. For information about how to remove the temporary files, see the following: See Removing temporary debugging files on page 253.
Change to the directory that contains the temporary files. The location of the temporary debugging files depends on your platform, as follows:
On Windows platforms, PureDisk writes debugging files to install_dir\Symantec\NetBackup PureDisk Agent\tmp. For example C:\ Program Files\Symantec\NetBackup PureDisk Agent\tmp. Linux, UNIX, or MacOS, PureDisk writes debugging files to /opt/pdagent/tmp.
254
1 2 3 4
Click Manage > Policies. In the left pane, under Storage Pool Management Policies, expand Maintenance. Select System policy for Maintenance. In the right pane, click Run Policy. This action runs the System policy for Maintenance on the entire storage pool, not just on the agent you debugged. The policy removes the scripts and the temporary files that PureDisk writes to the agent during debugging.
Chapter
13
About storage pool management Enabling trust between a storage pool and another Symantec application Changing node IP addresses to FQDNs or changing node FQDNs to new FQDNs (unclustered storage pools) About adding services Adding a service to a node Activating a new service in the storage pool Rerouting a content router and managing content routers Deactivating a service Managing license keys About central reporting Rerouting a metabase engine About clustered storage pool administration Changing the PDLinux administrators password Changing the PureDisk internal database and the LDAP administrator passwords Increasing the number of client connections Changing a storage pool's time zone Adjusting the clock on a PureDisk node
256
Adjusting the administrative Web UI timeout interval Stopping and starting processes on one PureDisk node (unclustered) Stopping and starting processes on one PureDisk node (clustered) Stopping and starting processes in a multinode PureDisk storage pool Restarting the Java Runtime Environment
See Enabling trust between a storage pool and another Symantec application on page 257. See Changing node IP addresses to FQDNs or changing node FQDNs to new FQDNs (unclustered storage pools) on page 258. See About adding services on page 265. See Adding a service to a node on page 266. See Activating a new service in the storage pool on page 272. See Rerouting a content router and managing content routers on page 273. See Deactivating a service on page 280. See Managing license keys on page 284. See About central reporting on page 285. See Rerouting a metabase engine on page 289. See About clustered storage pool administration on page 295. See Changing the PDLinux administrators password on page 295. See Changing the PureDisk internal database and the LDAP administrator passwords on page 295. See Increasing the number of client connections on page 296. See Changing a storage pool's time zone on page 297. See Adjusting the clock on a PureDisk node on page 298. See Adjusting the administrative Web UI timeout interval on page 299. See Stopping and starting processes on one PureDisk node (unclustered) on page 300.
Storage pool management Enabling trust between a storage pool and another Symantec application
257
See Stopping and starting processes on one PureDisk node (clustered) on page 302. See Stopping and starting processes in a multinode PureDisk storage pool on page 303. See Restarting the Java Runtime Environment on page 305.
Determine the host name of computer upon which the other application's authentication broker resides. Generally, the authentication broker host is the server system upon which the program is installed. If you want to enable trust between PureDisk and a clustered application, the host name is the name that was provided as the remote authentication broker host during installation. For example, the OpsCenter authentication broker host name is stored in the vxss.hostname parameter in the following file:
On Windows:
C:\Program Files\Symantec\OpsCenter\server\config\security.conf
On UNIX:
/opt/SYMCOpsCenterServer/config/security.conf
2 3
Log into the PureDisk storage pool authority node as root. Type the following command:
# su www-data
258
Storage pool management Changing node IP addresses to FQDNs or changing node FQDNs to new FQDNs (unclustered storage pools)
For hostname, specify the host name of computer upon which the authentication broker host for the other application resides. For example, specify a name in the following form: OpsCenterABhost:2821.
When the system displays Do you want to trust the above?(y/n), type y and press Enter. After trust is successfully enabled, PureDisk issues a message similar to the following:
setuptrust ------------------------------------------Setup Trust With Broker: OpsCenterABhost
Changing node IP addresses to FQDNs or changing node FQDNs to new FQDNs (unclustered storage pools)
The procedure in this topic explains how to use the changeFQDN.sh script to change a node's address from an IP address to a fully qualified domain name (FQDN). This topic also explains how to change an existing FQDN to a new FQDN. This procedure can be performed on unclustered storage pools only. Caution: For a clustered environment, do not try to change FQDNs, host names, or IP addresses. Contact technical support for assistance with these changes. The full path to the script is as follows:
/opt/pdconfigure/scripts/support/changeFQDN.sh
This documentation uses term FQDN to mean FQDN or host name. For information about how to use the conversion scripts to perform reconfiguration tasks, see the following:
See Cautions about changes to FQDNs and IP addresses on page 259. See (Optional) Creating a mapping file on page 260. See Changing a node's addressing information on page 261.
Storage pool management Changing node IP addresses to FQDNs or changing node FQDNs to new FQDNs (unclustered storage pools)
259
See (Conditional) Updating other storage pools when replication is enabled on page 263.
See Cautions for specific situations on page 259. See DNS resolution on page 259.
You administer multiple storage pools in your PureDisk environment. You replicate data from one storage pool to another. In this situation, make sure to update the node definitions on all storage pools that take part in replication. You want to change the storage pool authority node's address.
DNS resolution
If you migrate a storage pool from a configuration with IP addresses to a configuration with FQDNs, note the following:
Ensure that your DNS server resolves the FQDNs used to the IP addresses used. Otherwise, PureDisk services fail to communicate with each other during or after the switch. The entire environment must be DNS aware. All PureDisk agents must be able to resolve the FQDNs being used. For example, assume that your storage pool authority has the IP address 10.0.0.1. You want to switch to the FQDN myspa.acme.com. Ensure that myspa.acme.com resolves to 10.0.0.1 on all clients and nodes.
260
Storage pool management Changing node IP addresses to FQDNs or changing node FQDNs to new FQDNs (unclustered storage pools)
1 2
Log on to the storage pool authority as root. Change to the directory in which you want to create the mapping file. You can create this file in any directory. When you run the conversion scripts, you specify the full path to this file.
Use a text editor to open a new file. List the PureDisk nodes current addresses and new addresses in the following two-column format:
current_address_for_node1 current_address_for_node2 new_FQDN_for_node1 new_FQDN_for_node2
Create one line in the mapping file for each node in the storage pool. For each line in the mapping file, specify the current IP address or the current FQDN in column 1. Specify the new FQDN in column 2. Use one or more tabs or one or more space characters to separate the current addresses from the new FQDNs on each line. For example, assume that a storage pool has two nodes. The storage pool authority node is acme2.rr.com. An extra content router is acme2cr.rr.com. acme2.mapping.file is as follows:
acme2.rr.com beep2.rr.com acme2cr.rr.com beep2cr.rr.com
Storage pool management Changing node IP addresses to FQDNs or changing node FQDNs to new FQDNs (unclustered storage pools)
261
Use the chmod(1) command to set the following permissions on the mapping file: For example:
# chmod 664 acme2.mapping.file
Prompts you to decrypt the topology_nodes.ini file, if necessary. Performs some system verification checks. Changes the remote storage pool entries. Updates the records in the agent table of your storage pool authority database. It updates all server agent entries and backs up all SQL statements. You can use the SQL statements to go back to the databases original state. The SQL file path is as follows:
/opt/pdconfigure/scripts/support/backup/DNSIPChange_timestamp
Updates the record in the storage pool table of the storage pool authority database. It backs up the old table to the following file:
/opt/pdconfigure/scripts/support/backup/DNSIPChange_timestamp
Pushes the new routing tables to all server agents and reloads them on the content routers. Updates all configuration files on all server nodes with the new FQDNs. The script starts a workflow to write the new configuration files. Updates the topology_nodes.ini file with the new FQDNs. It writes the old topology_nodes.ini file to the following file:
262
Storage pool management Changing node IP addresses to FQDNs or changing node FQDNs to new FQDNs (unclustered storage pools)
/opt/pdconfigure/scripts/support/backup/DNSIPChange_timestamp
Updates the firewall rules. Prompts you to encrypt the topology_nodes.ini file again.
Use the following procedure to perform the conversion to FQDNs. To change a node's addressing information
1 2
Log on to the storage pool authority as root. Run the changeFQDN.sh script in one of the following forms:
# /opt/pdconfigure/scripts/support/changeFQDN.sh mapping_file
OR:
# /opt/pdconfigure/scripts/support/changeFQDN.sh --build
If a mapping file exists, use the first form. Specify the full path to the mapping file. A mapping file exists because you created one manually or because you edited a mapping file that PureDisk created. See (Optional) Creating a mapping file on page 260. If you ran a conversion script at an earlier time, PureDisk updates the mapping file before it converts the storage pool. If no mapping file exists, use the second form. If you specify --build, the script creates a mapping file for you. The mapping file lists IP addresses, FQDNs, and remote references. If reverse DNS lookups are allowed, it looks up the FQDNs for you.
Respond y to continue the conversion or n to stop the conversion. This decision is as follows:
If you want to verify the mapping file before you continue, respond n to the prompt and proceed to the next step. Symantec recommends that you verify the mapping file before you continue with the conversion. This verification is particularly important if you run changeFQDN.sh as part of the procedure to configure single-port communication. If you want to continue with the conversion, respond y to the prompt and skip the next step.
Storage pool management Changing node IP addresses to FQDNs or changing node FQDNs to new FQDNs (unclustered storage pools)
263
Perform this step if you need to verify the mapping file. Complete the following tasks:
Open the file with a text editor. Edit the file, if necessary. If the script cannot determine a particular FQDN, it issues messages. It writes the mapping file with placeholder information between less-than (<) and greater-than (>) characters. For example: <enter hostname here>. Supply the correct names at this time. Close the file. Return to step 2.
Use the Web UI to update the agent configuration files. Perform the following steps:
Click Manage > Agent. Select your storage pool. In the right pane, click Update Agent Configuration Files. Wait until the Update Agent Configuration Files job finishes.
6 7
Log on to the storage pool authority node as root. Type the following command to restart PureDisk:
# /etc/init.d/puredisk restart
If this storage pool is used for replication, update other storage pools. For information about how to update other storage pools, see the following: See (Conditional) Updating other storage pools when replication is enabled on page 263.
See Converting storage pools when you use replication on page 264. See Example of converting storage pools with replication on page 265.
264
Storage pool management Changing node IP addresses to FQDNs or changing node FQDNs to new FQDNs (unclustered storage pools)
Log on to the storage pool authority on the destination storage pool as root. This server hosts the storage pool authority in the storage pool that receives the replicated data.
Run the appropriate conversion script and use the --build option. The script generates a mapping file with information about remote references. The following is an example:
# /opt/pdconfigure/scripts/support/changeFQDN.sh --build
3 4
Examine the mapping file in /opt/pdinstall/mapping.txt and edit it if necessary. Run the appropriate conversion script and use the mapping_file option. Following are examples:
# /opt/pdconfigure/scripts/support/changeFQDN.sh /opt/pdinstall/mfile
Log on to the storage pool authority on the source storage pool as root. This server hosts the storage pool authority in the storage pool that transfers its data.
Run the appropriate conversion script and use the --build option. The script generates a mapping file with information on remote references. Following are examples:
# /opt/pdconfigure/scripts/support/changeFQDN.sh --build
Run the appropriate conversion script with the mapping_file and --remoteonly options. Following are examples:
265
On DESTINATION, run changeFQDN.sh with the --build option to build a mapping file. On DESTINATION, run changeFQDN.sh again. Specify the mapping file. This script completes the conversion on DESTINATION. On SOURCE, run changeFQDN.sh with the --build option. This script generates a mapping file that contains the remote references. On SOURCE, run changeFQDN.sh again. Specify the --remoteonly option and the mapping file. When you specify these options, the script changes the references to DESTINATION. The script does not change references to the rest of the SOURCE storage pool. It does not modify the local storage pool.
Examine events from the system monitor script. The system monitor script monitors system activity. By default, it runs every five minutes and sends a status message. To see the messages, click Monitor > Alerts & Notification. In the right pane, pull down Application, and type MonitorStatistics in the Look for: field. For more information, see the following: The PureDisk Backup Operator's Guide.
Display capacity dashboards. For more information, see the following: See About Dashboard reports on page 227.
For example, you might need to add the following additional services:
Metabase engine. One metabase engine service can support 1,000 clients. Add an additional metabase engine service if your site needs to support more than 1,000 clients.
266
As you add new clients, PureDisk assigns them to the new metabase engine. It does not move clients from one metabase engine to another metabase engine.
Content router. Add an additional content router service if the /Storage/data partition fills. The system monitor scripts report and the capacity dashboard include information on the disks that have reached their capacity. For more information about adding content routers and content router rerouting, see the following: See Rerouting a content router and managing content routers on page 273. A NetBackup export engine. This service lets you send content router data to a NetBackup storage unit.
See Adding a service to a node on page 266. See Activating a new service in the storage pool on page 272. See Rerouting a content router and managing content routers on page 273. See Deactivating a service on page 280. See Managing license keys on page 284. See About central reporting on page 285. See Rerouting a metabase engine on page 289. See About clustered storage pool administration on page 295. See Changing the PDLinux administrators password on page 295. See Changing the PureDisk internal database and the LDAP administrator passwords on page 295. See Increasing the number of client connections on page 296. See Adjusting the clock on a PureDisk node on page 298. See Adjusting the administrative Web UI timeout interval on page 299.
See Adding a new service on an existing node on page 267. See Adding a new node and at least one new service on the new node on page 268.
267
See Verifying and specifying content router capacity on page 269. See Adding a new passive node to a cluster on page 270.
For information about how to add services to an unclustered storage pool, see the following:
(Conditional) Freeze the service group on the node to which you want to add the new service. Perform this step if the storage pool is clustered. This action prevents VCS from failing over the node when it restarts the server process. To freeze the node, use the Cluster Manager Java console.
In a browser window, type the following to start the storage pool configuration wizard:
http://URL/Installer
For URL, type the FQDN of the node that hosts the storage pool authority service.
3 4
Click Next on the wizard's pages until you arrive at the Services Configuration page. On the Service Configuration page, perform the following steps:
Click Change. Select the service you want to add. Click Next when the Services Configuration page is complete.
5 6 7
Click Next until you arrive at the Implementation page. On the Implementation page, click Finish. (Conditional) Visually inspect the Cluster Manager Java Console Web UI and check for fault conditions. Perform this step if the storage pool is clustered. VCS might have detected that the service is down. In this case, the resource might appear as faulted in the PureDisk administrative Web UI. If the resource appears as faulted, complete the following steps:
268
Right-click the resource and select Clear Fault - Auto. After you clear the fault, the resource appears as Offline. Although it is started again, specify to VCS that you want it to monitor the resource. To enable monitoring again, right-click the resource and select probe node-name.
(Conditional) Visually inspect the Cluster Manager Java Console and make sure that all resources now appear with a status of Online. Perform this step if the storage pool is clustered.
(Conditional) In the Cluster Manager Java Console, right-click the service group and select Unfreeze. Perform this step if the storage pool is clustered.
If you added a content router, see the following: See Verifying and specifying content router capacity on page 269. If you added a metabase engine or a NetBackup export engine, see the following: See Activating a new service in the storage pool on page 272.
Adding a new node and at least one new service on the new node
The following procedure explains how to add a new node and a service. To add a service to a new node
Install PDLinux on the computer that you want to configure as a new node. Use the instructions in the PureDisk Storage Pool Installation Guide to install PDLinux.
In a browser window, type the following to start the storage pool configuration wizard:
http://URL/Installer
For URL, type the FQDN of the node that hosts the storage pool authority service.
Click Next until you arrive at the Storage Pool Node Summary page.
269
Visually inspect the Storage Pool Node Summary page and determine if the new node appears. If the new node does not appear, click Add Node and add the node. Use the instructions in the PureDisk Storage Pool Installation Guide to add the node.
Click Next until you arrive at the Storage Selection pages. Use the instructions in the PureDisk Storage Pool Installation Guide to configure storage for this node. If no disks appear in the wizard, it might be because your disks need to be formatted or repartitioned.
Click Next until you arrive at the Services Configuration page. Use the instructions in the PureDisk Storage Pool Installation Guide to configure services on this node.
7 8 9
Click Next until you arrive at the Implementation page. On the Implementation page, click Finish. (Conditional) If TCP/IP settings on the other nodes have been changed to improve replication job performance, run the following script on the new node:
# /opt/pdconfigure/scripts/support/tcp_tune.sh modify
If you added a content router, see the following: See Verifying and specifying content router capacity on page 269. If you added a metabase engine or a NetBackup export engine, see the following: See Activating a new service in the storage pool on page 272.
You added a content router to an existing storage pool. You configured a new storage pool and your content routers have different capacities.
You do not need to perform this procedure if you added a metabase engine or NetBackup export engine.
270
PureDisk assumes that all content routers have the same storage capacity. However, you may have content routers each with different capacities. If you do not specify the content router capacities explicitly, the content router with the smallest capacity fills up first. As a result, you must add another content router sooner. To verify and specify content router capacity
1 2 3 4 5
Click Settings > Topology. In the left pane, expand the tree until you see all the content routers. Select a content router. In the right pane, visually inspect the Storage size (GB) field. (Conditional) Specify the content routers capacity in the Storage size (GB) field. Perform this step if the displayed capacity is incorrect. PureDisk uses the information in this field when it determines the fingerprint range to assign to each content router. After you change this value, PureDisk redistributes the fingerprint ranges relative to the new capacity specifications. If the content routers already contain data, PureDisk redistributes the data, too.
6 7 8
Click Save to save your changes. Perform step 3 through step 6 for each content router in the storage pool. Your next action depends on which type of service you changed or added, as follows: If you edited the information for an active content router, perform the procedure in the following section: See Rerouting a content router and managing content routers on page 273. If you edited the information for the content routers that you installed as part of a new storage pool, perform the procedure in the following section: See Rerouting a content router and managing content routers on page 273.
271
Install PDLinux on the computer that you want to configure as a new node. Use the instructions in the PureDisk Storage Pool Installation Guide to install PDLinux.
In a browser window, type the following to start the storage pool configuration wizard:
http://URL/Installer
For URL, type the FQDN of the node that hosts the storage pool authority service.
3 4
Click Next until you arrive at the Storage Pool Node Summary page. Visually inspect the Storage Pool Node Summary page and determine if the new node appears. If the new node does not appear, click Add Node and add the node. Use the instructions in the PureDisk Storage Pool Installation Guide to add the node. When the new node appears in the node summary, perform the following steps:
Select the node you want to configure as a passive node. Click Edit Node. In the Node Type pull-down menu, select Passive. Click OK.
5 6 7
Click Next until you arrive at the Implementation page. On the Implementation page, click Finish. (Conditional) If TCP/IP settings on the other nodes have been changed to improve replication job performance, run the following script on the node you added:
# /opt/pdconfigure/scripts/support/tcp_tune.sh modify
Use the Cluster Manager Java Console to perform a manual failover to the new node. Symantec recommends that you test a manual failover to this new node at a time that is convenient in your schedule. When you perform a manual failover, your storage pool will be temporarily offline. See the instructions on how to perform a manual failover in the Veritas Cluster Server (VCS) documentation.
272
1 2 3 4
Click Settings > Topology. In the left pane, expand the tree view until you see the new service. Select the new service. In the right pane, click one of the following:
Activate Content Router Activate Metabase Engine Activate NetBackup Export Engine
If you added a new metabase engine or a new NetBackup export engine, you are finished. Do not complete the rest of this procedure. If you added a new content router, proceed to the next step.
(Conditional) If you added a new content router, confirm whether to reroute the storage pool at this time. Select one of the following options:
Yes, reroute now. This selection starts the rerouting process. Select this choice only after you activate all the content routers you intend to activate. You want to reroute only on time, and you want to reroute only after you activate all the new content routers. No, I want to continue making changes. If you select this choice, make your changes, activate additional content routers, and proceed to the following section: See Rerouting a content router and managing content routers on page 273.
Note: When you reroute the storage pool, PureDisk moves data between content routers. This process requires some free storage space on each of the content routers. If a content router has no more storage available, your rerouting might take much longer. Determine whether to run your data selection removal policies and data removal policies to free some storage space before you start the rerouting process.
Storage pool management Rerouting a content router and managing content routers
273
See Planning for a new content router on page 273. See Permissions for rerouting on page 274. See Disaster recovery backups and rerouting on page 274. See Data replication policies and rerouting on page 274. See Activating and deactivating content routers on page 275. See Alleviating content router congestion on page 275. See Parallel and serial rerouting examples on page 276. See Rerouting the content routers on page 278. See Troubleshooting a content router rerouting job on page 279.
When parallel rerouting is performed, all content routers actively redistribute data simultaneously. Parallel rerouting is faster than serial rerouting, but parallel rerouting is not always possible. When serial rerouting is performed, only one content router at a time redistributes its data.
The current capacity of the content routers in the storage pool. If your content routers have very little excess capacity, PureDisk performs serial rerouting. PureDisk employs this method because it requires excess capacity on each content router when it moves the data between content routers.
274
Storage pool management Rerouting a content router and managing content routers
If you receive warning messages from PureDisk about your content router capacity, you can assume that the routers are near capacity. If your content routers are near capacity, PureDisk is more likely to perform serial routing. If your content routers are at a low capacity, PureDisk is more likely to perform parallel rerouting. The excess capacity on the content routers allows PureDisk to move the data between content routers more efficiently.
The number of content routers you add to the storage pool. If possible, always add enough content routers to make parallel rerouting possible. Generally, if you can double the number of content routers in a storage pool, you request parallel rerouting. For example, assume that you add only one content router to a storage pool that contains three or more content routers. Each router is near capacity. In that case, PureDisk reroutes by using the serial method. However, if you add two or three content routers, you can request parallel rerouting. For more information about rerouting, see the following: See Parallel and serial rerouting examples on page 276.
For information about permissions, see the PureDisk Client Installation Guide.
Storage pool management Rerouting a content router and managing content routers
275
Activation requested You have requested activation of this content router, but have not yet started rerouting. Deactivation requested You have requested deactivation of this content router, but have not yet started rerouting. Activation pending During rerouting, content routers that you have activated change from the state "Activation requested" to "Activation pending" as soon as the actual rerouting of data starts. Deactivation pending During rerouting, content routers that you have deactivated change from the state "Deactivation requested" to "Deactivation pending" as soon as the actual rerouting of data starts. Active This content router is active. Inactive This content router is inactive.
In this case, you can still make changes, either to activate or to deactivate the content router, before you start the rerouting process again. However, try to avoid such situations because they result in unnecessary data movement between content routers.
276
Storage pool management Rerouting a content router and managing content routers
Second level. Backups stop when the first content router in a storage pool reaches this level. This level is the warning threshold.
Message The Content Router has insufficient disk space to accept new data. No new data will be accepted until more disk space becomes available. Rerouting, data removal and/or garbage collection is needed urgently.
Third level. Data from the spool area can continue to fill the content routers even after the backups stop. The content routers are full.
Message The Content Router has insufficient disk space to accept new data. No new data will be accepted until more space becomes available. Rerouting, data removal and/or garbage collection is needed urgently. Manual intervention to temporarily free disk space may become necessary.
If your content routers fill up, perform one or more of the following actions:
First, run a data removal policy. If you know that you have a lot of unneeded data on the content router, this process frees up needed space. For information on data removal policies, see the PureDisk Backup and Restore Guide. Second, add another content router and reroute your data. Because you have full content routers, this process is very slow. Use the procedures in this chapter, and perform this action if the data removal policy did not free up enough space. Third, call Symantec Technical Support .
Storage pool management Rerouting a content router and managing content routers
277
If you have little excess capacity in your content routers, PureDisk performs serial rerouting, regardless of how much more capacity you add. If you have much excess capacity in your content routers and much new capacity in the new content routers, Symantec recommends that you select parallel rerouting.
While it is not possible to guarantee circumstances under which parallel rerouting always occurs, the following provide example scenarios:
See Example 1 - Serial rerouting scenario on page 277. See Example 2 - Parallel rerouting scenario on page 277. See Other rerouting examples on page 277.
You have three 4-TB content routers. Their total maximum capacity is 12 TB. Each content router is at 80% capacity. 9.6 TB of data reside on these three content routers. The routers have 2.4 TB of excess capacity today. You want to add one more 4-TB content router.
In this case, the capacity you want to add (4 TB) is not significantly more than the excess capacity that exists today (2.4 TB). PureDisk performs serial rerouting.
You have four 500-GB content routers. Their total maximum capacity is 2000 GB. Each content router is at 50% capacity. 1000 GB of data reside on these four content routers. The routers have 1000 GB of excess capacity today. You want to add four more 500-GB content routers. These routers are to increase the capacity of the storage pool by 2000 GB.
In this case, the capacity you want to add (2000 GB) is significantly more than the excess capacity that exists today (1000 GB). Symantec recommends parallel rerouting.
278
Storage pool management Rerouting a content router and managing content routers
Table 13-1
1 2 3 4 5 6
Click Settings > Topology. In the left pane, select the storage pool. Expand the storage pool until the PureDisk administrative Web UI displays the content router you added. Select the content router you added. In the right pane, click Activate Content Router. Specify whether you want to reroute now or later. For example, respond Yes, reroute now if you have added the last new content router or the only new content router to this storage pool. Respond No, reroute later if you added more than one content router, and you need to activate other new content routers.
(Conditional) In the left pane, select the storage pool. Perform this step if the correct storage pool is not already selected.
(Conditional) In the right pane, click Reroute Content Router. Perform this step if the correct storage pool was not already selected.
Storage pool management Rerouting a content router and managing content routers
279
Wait for the rerouting process to complete. This action redistributes the data across the content routers in the storage pool. PureDisk messages indicate whether you can perform parallel rerouting or serial rerouting. If possible, always choose parallel rerouting. Rerouting can take some time to complete. For example, it can take several hours to even several days, depending on the volume of data that needs to be rerouted. If possible, keep other system activity, such as backup and removal jobs, low during this activity. Each content router receives a rerouting job. Look in the job table to make sure that all rerouting jobs complete successfully. If a job fails, analyze the problem through the job log, correct the problem if possible, and retry rerouting as soon as possible. If the second try is unsuccessful, contact your Symantec Technical Support representative.
Retrieve the job log for the failed rerouting job. Perform the following steps:
Click Monitor > Jobs. Click the job number of the failed rerouting job. On the job details display, click Job Log.
Examine the job log for network errors or other environmental factors.
280
(Conditional) Fix external conditions. If the job log noted network errors or environmental factors that contributed to the job's failure, remedy those conditions.
Rerun the rerouting job. Even if there were no external conditions for you to remedy, run the job again.
If repeated efforts to remedy environmental problems and to run the job successfully have failed, contact Symantec Technical Support .
Deactivating a service
The following procedure explains how to deactivate a content router or a NetBackup export engine in a storage pool. Other services cannot be deactivated.
281
Before you deactivate a content router, check the capacity of the other content routers in your storage pool. For information about how to check this capacity, see the following section: Preparing to deactivate a content router
Display the capacity dashboard. See Displaying the Capacity dashboard on page 227.
Review the following examples: See Example 1 - Content routers that can be rerouted on page 281. See Example 2 - Content routers that cannot be rerouted on page 282.
Examine the dashboard to determine whether remaining content routers have sufficient excess capacity to ensure that the rerouting can complete.
Content router 1 has 1-TB capacity and is 60% filled. In other words, 600 GB are used, and 400 GB are free. Content router 2 has 2-TB capacity and is 60% filled. In other words, 1200 GB are used, and 800 GB are free. Content router 3 has 1-TB capacity and is 60% filled. In other words, 600 GB are used, and 400 GB are free.
282
Option 1. Assume that you want to remove content router 2 by deactivating it and rerouting the storage pool. To reroute the storage pool, you must reroute the 1200 GB of data on content router 2 to content routers 1 and 3. However, this approach cannot work because together content routers 1 and 3 have only 800 GB of free space. Option 2. Assume that you want to remove content router 3 or content router 1. For each of these content routers, ample free space exists on the remaining content routers to hold the 600 GB of data from the source.
Content router 1 has 1-TB capacity and is 70% filled. In other words, 700 GB are used, and 300 GB are free. Content router 2 has 1-TB capacity and is 70% filled. In other words, 700 GB are used, and 300 GB are free. Content router 3 has 1-TB capacity and is 60% filled. In other words, 600 GB are used, and 400 GB are free.
Assume that you want to remove content router 3 by deactivating it and rerouting the storage pool. You must reroute and redistribute the 600 GB of data on content router 3 to content router 1 and content router 2. Together, content router 1 and content router 2 have 600 GB free. The deactivation appears feasible. However, this plan is not feasible because the rerouting would fill each content router to 100% capacity. The rerouting process requires that the host that receives the data has a margin of excess capacity. A content router always has an internal soft limit and an internal hard limit on capacity. A content router requires a margin of excess capacity to function. Another reason for maintaining a margin is that the rerouting process is not always even. Content router 1 might receive 300 GB of data and reach its limit before content router 2 received 100 GB of data. The rerouting process would fail even though content router 2 had excess capacity. For more information about soft limits and hard limits, see the PureDisk Backup Operator's Guide.
283
To deactivate a service
1 2 3
Click Settings > Topology. Expand the tree in the left pane until you see the content router or NetBackup export engine that you want to deactivate. Select the service. Warning: Ensure that sufficient capacity exists on the remaining content routers in your storage pool before you deactivate your content router and try to reroute. Failure to ensure that the rerouting can complete might render your storage pool inoperable. Before you proceed to the next step, ensure that you properly prepared to deactivate the content router. See Preparing to deactivate a content router on page 281.
In the right pane, click Deactivate Content Router or Deactivate NetBackup Export Engine. For a content router, the status changes to Deactivation requested. For a NetBackup export engine, the status changes to Inactive. If you deactivated a NetBackup export engine, proceed to the following step:
Step 8.
5 6 7
In the right pane, respond to the question about whether to reroute now or whether you want to make more changes. Select the storage pool. In the right pane, click Reroute Content Router. This selection starts the rerouting process, which redistributes data over all active content routers. The process moves data from the content router in Deactivation requested status to the content routers that you want to remain active. Early in the rerouting process, the state of the content router changes from Deactivation requested to Deactivation pending. At the end of the rerouting process, PureDisk sets its state to Inactive. Wait for rerouting to complete successfully before proceeding to the next step. For more information about the rerouting process, see the following: See Rerouting a content router and managing content routers on page 273.
(Conditional) Take offline the cluster group to which the active service belongs.
284
The storage pool is clustered. This service is the only remaining active service on the node.
From the Cluster Manager Java Console, right-click the cluster group, and select Offline > All Systems.
1 2
Click Settings > Configuration. In the left pane, select License Management. The right panel displays license key information. Continue with this procedure if you want to add a license key.
3 4 5
In the right pane, click Add Key. In the Key field, type the license key. Click Add. If you try to add an expired license key, PureDisk generates the following message:
Key not registered: No such file or directory
Check the expiration date for the key. To delete a license key
1 2 3 4 5
Click Settings > Configuration. In the left panel, expand License Management. Select the license key that you want to delete. In the right panel, click Delete Key. To confirm the deletion, click OK on the window that appears.
285
See Enabling a storage pool as a central storage pool on page 285. See Adding a remote storage pool to a central storage pool on page 286. See Disabling central reporting on page 287. See Managing storage pools configured in the central storage pool on page 287.
Make sure that the storage pool you want to designate as central is not already designated as a central storage pool. Log on to the administrative Web UI, click Settings and visually inspect the administrative Web UI. If Central SPA appears under Settings, this storage pool is already designated as a central storage pool. Do not perform the remaining steps of this procedure. If Central SPA does not appear under Settings, the storage pool is an independent storage pool or a central storage pool manages it. If you have a managed storage pool and you want to make it a central storage pool, you have the following choices:
You can disable the managed-to-central reporting relationship that is currently in effect. For the current central storage pool and for all of its managed storage pools, perform the procedure in the following section: See Disabling central reporting on page 287.
286
You can continue with this procedure and make this storage pool a central storage pool, too. A storage pool can be both a central storage pool and be a storage pool that is included in another central storage pools reports.
2 3
Log on to the storage pool authority node as root. Type the following command:
# /opt/pdinstall/add_central_reporting.sh
Add one or more storage pools to this new central storage pool. See Adding a remote storage pool to a central storage pool on page 286.
Make sure that the storage pool you want to use as a central storage pool is enabled as a central storage pool. For information about how to designate a storage pool as a central storage pool, see the following: See Enabling a storage pool as a central storage pool on page 285.
2 3 4 5
Click Settings > Central SPA. In the left pane, click Storage Pool Management. In the right pane, click Add Remote SPA Entry. Complete the following fields in right pane:
Storage pool name. The name of a storage pool that you want to manage. The storage pool you name here will be managed by the central storage pool. Host name (FQDN). The FQDN (recommended) or host name of the storage pool authority service of the storage pool that you want to manage. That is the address of the other storage pool authority service. Login. The logon name of a valid user of the managed storage pool. The user must have Central Report rights.
287
Password. The password of that valid user of the managed storage pool.
Click Add. When you add a storage pool, PureDisk queries for a list of other storage pools that are known (through replication) to that storage pool. PureDisk adds these linked storage pools to the central storage pool list.
1 2
Log on to the central storage pool authority node as root. Type the following command:
# /opt/pdinstall/del_central_reporting.sh
See Editing or deleting a storage pool in the central storage pool list on page 287. See Starting the PureDisk administrative Web UI for another storage pool in another window on page 288. See Testing connections between storage pools on page 288.
1 2 3 4
Click Settings > Central SPA. In the left pane, select a storage pool. Edit the fields. Click Save.
288
1 2
Click Settings > Central SPA. In the left pane, select a storage pool. You cannot delete the central storage pool.
Starting the PureDisk administrative Web UI for another storage pool in another window
The following procedure explains how to start the PureDisk administrative Web UI for another storage pool. The two storage pools must be in a managed-to-central relationship. To start the PureDisk administrative Web UI for another storage pool in another window
1 2 3 4
Click Settings > Central SPA. In the left pane, select a storage pool. In the right pane, select Manage Storage Pool. Type the logon and password to the other storage pool.
1 2 3
Click Settings > Central SPA. In the left pane, select a storage pool. In the right pane, click Test Connection. The administrative Web UI displays a status message for the connection to the storage pool.
289
See (Optional) Gathering metabase engine capacity information on page 289. See Preparing clients for rerouting on page 290. See Preparing the old metabase engine for rerouting on page 291. See Adding the new metabase engine and recording its address on page 292. See Rerouting the agents on the metabase engine on page 292. See Restarting the agent on page 293. See Verifying a metabase engine rerouting on page 294. See Troubleshooting on page 294.
1 2 3 4
Log on to the storage pool authority's administrative Web UI. Click Settings > Topology. In the left pane, select the storage pool. In the right pane, click Capacity Dashboard.
290
Examine the information in the lower half of the Capacity Dashboard report under MetaBase Engine Capacity Report. If Current Usage is nearing 90%, consider whether to add another metabase engine. At 90% of capacity, the metabase engine shuts down. After you add an additional metabase engine, you can reroute the metabase engines.
Prepare the clients for rerouting. See Preparing clients for rerouting on page 290.
1 2 3
Decide which agents you want to move to a new metabase engine. Click Manage > Agent. For each agent that you want to move to the new metabase engine, perform the following steps:
In the left pane, select an agent. In the right pane, under More Tasks, click Agent Dashboard.
For each agent that you want to move, record the agent ID information from the Agent Dashboard display. For example, you can record the information below:
The ID field for the first agent The ID field for the second agent The ID field for the third agent The ID field for the fourth agent _________________________ _________________________ _________________________ _________________________
5 6 7
Click Monitor > Jobs. Select an agent that you want to move. Examine the jobs for this agent and terminate any that are still running. During the rerouting process, make sure that no jobs are running on the agents you want to move, including the following types of job workflows:
291
Files and Folders Backup Full System Backup MS Exchange Backup MS SQL Backup System State and Services Backup UNC Path Backup Data Removal Export To NetBackup
If any jobs are running that use data selections that reside on the agents you want to move, click Stop job gracefully to end them.
8 9
Click Manage > Agent. In the left pane, select the agent that you want to move.
10 In the right pane, under More Tasks, click Deactivate Agent. 11 Repeat step 5 through step 10 for each agent that you want to move. 12 Prepare the old metabase engine for rerouting.
See Preparing the old metabase engine for rerouting on page 291.
1 2
Click Monitor > Jobs. In the left pane, use the View jobs by pull-down menu to select Topology. This action lets you select the metabase engine and display its jobs.
3 4
In the left pane, expand the tree and select the old metabase engine. In the right pane, examine the running jobs and terminate any that are still running. During the rerouting process, no jobs can run on the old metabase engine. These jobs include the following types of workflows:
292
If any jobs are running, click Stop job gracefully to end them.
Proceed to the following section: See Adding the new metabase engine and recording its address on page 292.
Add a new metabase engine to the storage pool. See About adding services on page 265.
2 3 4 5
Click Settings > Topology. Expand the tree in the left pane until it displays all the storage pool services. Select the new metabase engine server agent. In the right pane, note the Agent Address field, and record the FQDN of the new metabase engine. Metabase engine nodes identification __________________________
Reroute the agents on the metabase engine. See Rerouting the agents on the metabase engine on page 292.
1 2
Log on to the storage pool authority as root. Type the following command to change to the PureDisk commands directory:
# cd /opt/pdspa/cli
293
Answer the prompts from the script. For example, the script prompts you to confirm that you want to continue with the rerouting process.
Observe the completion messages. Make sure that the rerouting completes before you start any new jobs for the agents you moved. As part of its work, the script performs the following actions:
Activates the agents that now reside on the new metabase engine. Updates the agent configuration files. Copies the data selections from the original metabase engine to the new metabase engine.
Restart the agent Information about how to restart the agent is available. See Restarting the agent on page 293.
294
1 2 3
Log on to the client system as an administrator. Click Start > Settings > Control Panel > Administrative Tools > Services. In the Services window, locate and select the Veritas NetBackup PureDisk Client Agent service. Click Restart to restart the service.
1 2
1 2 3 4
Click Manage > Agent. In the left pane, select an agent. In the right pane, look for Metabase Engine: and make sure that the agent is attached to the new metabase engine. Test the new configuration by running a manual backup from this agent.
Troubleshooting
You might need to abort a metabase engine rerouting job or a metabase engine rerouting job might fail. The following procedure returns a storage pool to the state it was in before you started a metabase engine rerouting job. To troubleshoot a failed metabase rerouting job
1 2
Log on to the storage pool authority as root. Type the following command to change to the PureDisk commands directory:
# cd /opt/pdspa/cli
295
Log on to the node as root. If the storage pool is clustered, log on to the physical computer hardware.
Type the following command to reset the password in the operating system:
# passwd
When the command issues prompts, type the old and new passwords.
Changing the PureDisk internal database and the LDAP administrator passwords
Depending on your site practices, or your security status, you might need to regenerate passwords for the PureDisk internal database or for the LDAP administrator. These passwords are internal. At configuration time, the storage pool configuration wizards sets them to a random, internal value. The following procedure explains how to change the following passwords:
296
(Conditional) Freeze the cluster. Perform this step if the storage pool is clustered. This action prevents VCS from failing over the node when it restarts the server process. To freeze the node, use the Cluster Manager Java console.
In a browser window, type the following to start the storage pool configuration wizard:
http://URL/Installer
For URL, type the FQDN of the node that hosts the storage pool authority service.
3 4 5 6 7
Click Next until you arrive at the Regenerate Passwords page. Click Regenerate Passwords. Wait for the process to complete. Click Cancel. (Conditional) In the Cluster Manager Java Console, right-click the service group and select Unfreeze. Perform this step if the storage pool is clustered.
Keep a copy of the new topology.ini file and the new topology.nodes.ini file in a safe place. Encrypt these files if that is your site practice. You need to supply these new passwords if you perform a disaster recovery restore.
297
1 2
Click Settings > Configuration. In the left pane, expand Configuration File Templates > PureDisk ContentRouter > Default Value Set for PureDisk ContentRouter > ContentRouter > MaxConnections > All OS:number. Select All OS:number. In the right pane, in the Value field, increase the present number to the number of clients + 5. Five slots are reserved. The maximum value you can specify is 8192.
3 4
5 6 7 8
Click Save. In the left pane, select TaskThreadStackSize. In the right pane, select Add Configuration File Value. On the Properties: Configuration File Value screen, change the Value field to 256 and click Add. This value is the stack size for client threads.
Restart the content router. For information about how to restart the content router or other processes see the following:
See Stopping and starting processes on one PureDisk node (unclustered) on page 300. See Stopping and starting processes on one PureDisk node (clustered) on page 302. See Stopping and starting processes in a multinode PureDisk storage pool on page 303.
298
In a browser window, type the following to start the storage pool configuration wizard:
http://URL/Installer
For URL, type the FQDN of the node that hosts the storage pool authority service.
2 3
Click Next on the wizard's pages until you arrive at the Storage Pool Details page. On the Storage Pool Details page, in the Time Zone field, use the pull-down menu to select the correct time zone for this storage pool. You can select only one time zone for a storage pool. PureDisk does not support multiple time zones in a single storage pool.
4 5
Click Next until you arrive at the Implementation page. On the Implementation page, click Finish.
Stop all PureDisk services on the node. Typically, this action causes running jobs to fail. For information about how to stop and start processes, see the following:
See Stopping and starting processes on one PureDisk node (unclustered) on page 300. See Stopping and starting processes on one PureDisk node (clustered) on page 302.
299
See Stopping and starting processes in a multinode PureDisk storage pool on page 303.
(Conditional) Make sure that the NTP server works properly and can be reached from the node. Perform this step if the node you want to fix hosts the storage pool authority service.
3 4
Adjust the time on the node. Start the PureDisk processes on the node. For information about how to stop and start processes, see the following:
See Stopping and starting processes on one PureDisk node (unclustered) on page 300. See Stopping and starting processes on one PureDisk node (clustered) on page 302. See Stopping and starting processes in a multinode PureDisk storage pool on page 303.
To adjust the clock on a PureDisk node if the time difference is more than one day
1 2 3
Log into the storage pool authority node as root. Open file /opt/pdgui/tomcat/webapps/PureDisk/WEB-INF/web.xml. Search for the following string:
<session-timeout>30</session-timeout>
300
Storage pool management Stopping and starting processes on one PureDisk node (unclustered)
For example, assume that you had to abort the installation of a PureDisk environment. Before trying to install the software again, you need to stop all the services on the host. You can enter the preceding command to stop all services correctly before you try to reinstall the environment.
Storage pool management Stopping and starting processes on one PureDisk node (unclustered)
301
start. Starts the specified services. stop. Stops the specified services.
restart. Stops and then starts specified services. The system restarts the services, but it does not enumerate in its messages which services it restarts. reload. Reloads the configuration files for the specified services and starts them if they are not already started. status. Displays the status of the specified services.
302
Storage pool management Stopping and starting processes on one PureDisk node (clustered)
service
Optional. Use caution when you specify individual service specifications. If you do not specify any services, the command affects all services. This syntax is the preferred way to use this command. If you want to specify individual services, specify them in the order that they appear in the following lists. If you want to stop more than one service, enter a space character between each service. Stop them in the following order:
pdworkflowd. The PureDisk workflow daemon. pdcr. The PureDisk content router. pdagent. The PureDisk server agent. pdctrl. The PureDisk controller. pdweb. The PureDisk web server. pddb. The PureDisk database service. pdmemcached. The memcache daemon.
If you want to start more than one service, enter a space character between each service. Start them in the following order:
pdmemcached. The memcache daemon. pddb. The PureDisk database service. pdweb. The PureDisk web server. pdctrl. The PureDisk controller. pdagent. The PureDisk server agent. pdcr. The PureDisk content router. pdworkflowd. The PureDisk workflow daemon.
Storage pool management Stopping and starting processes in a multinode PureDisk storage pool
303
For example, assume that you want to stop PureDisk processes on node1 in your storage pool. The services on node1 appear in the Cluster Manager Java Console as pd_group1, and you want to restart one or more services on that node. To stop and start services on example node 1, which is known to VCS as pd_group1
In the Cluster Manager Java Console, right-click pd_group1 and select Freeze Temporary. This action prevents pd_group1 from failing over when VCS detects that some services are down.
Use a PureDisk procedure to stop or restart the services on that node. For more information, see: See Stopping and starting processes on one PureDisk node (unclustered) on page 300.
Visually inspect the Cluster Manager Java Console Web UI and check for fault conditions. Depending on how long ago the PureDisk service went down, VCS might have detected that the service is down. In this case, the resource might appear as faulted in the PureDisk administrative Web UI. If that is the case, complete the following steps:
Right-click that resource and select Clear Fault - Auto. After you clear the fault, the resource appears as Offline. Although it starts again, specify to VCS that you want it to monitor the resource. To enable monitoring again, right-click the resource and select probe - node1. This assumes that node1s other services are online currently.
4 5
Visually inspect the display and make sure that all resources of pd_group1 now appear with a status of Online. Right-click the resource group (pd_group1) and select Unfreeze. This action unfreezes the node.
304
Storage pool management Stopping and starting processes in a multinode PureDisk storage pool
(Conditional) Freeze the cluster. Perform this step if the storage pool is clustered. This action prevents VCS from failing over the node. To freeze the cluster, use the Cluster Manager Java console, and freeze all the service groups.
Stop the services on all the non-storage pool authority nodes first. The order in which you stop these nodes does not matter, but do not stop the storage pool authority node at this time. See Stopping and starting processes on one PureDisk node (unclustered) on page 300.
Stop the services on the storage pool authority node. See Stopping and starting processes on one PureDisk node (unclustered) on page 300.
4 5
Perform the maintenance operations that you need to perform. Restart services. See To start services in a multinode PureDisk storage pool on page 304.
Start the services on the storage pool authority node. See Starting all services without rebooting on page 300.
Start the services on all the non-storage pool authority nodes. See Starting all services without rebooting on page 300. If the storage pool is not clustered, stop here. Do not perform the rest of the steps in this procedure. If the storage pool is clustered, complete the rest of the steps in this procedure.
Visually inspect the Cluster Manager Java Console Web UI and check for fault conditions. Depending on how long ago the PureDisk service went down, VCS might have detected that the service is down. In this case, the resource might appear as faulted in the PureDisk administrative Web UI. If that is the case, complete the following steps:
Right-click that resource and select Clear Fault - Auto. After you clear the fault, the resource appears as Offline. Although it starts again, specify to VCS that you want it to monitor the resource. To
305
enable monitoring again, right-click the resource and select probe - node. This assumes that nodes other services are online currently.
4 5
Visually inspect the display and make sure that all resources of resource_group now appear with a status of Online. Right-click a resource_group and select Unfreeze. This action unfreezes the resource group. Perform this step for all resource groups.
1 2
Log onto the node that hosts the storage pool authority as root. Type the following command:
# /etc/init.d/puredisk status
306
Checking Checking Checking Checking Checking Checking Checking Checking Checking Checking Checking Checking Checking
for for for for for for for for for for for for for
LDAP-server VxATd daemon PureDisk Memory Cache Daemon PureDisk Database Server PureDisk WebServer PureDisk Controller PureDisk Controller Monitor PureDisk Server Agent PureDisk ContentRouter PureDisk Workflow Engine PureDisk MetabaseEngine PureDisk JAVA GUI CRON daemon
running running running running running running running running running running unused unused running
The preceding output indicates that a metabase engine is also installed on this node. The status for both the metabase engine and the JAVA GUI are set to unused. This output indicates that the JRE has failed.
Chapter
14
About the configuration files Examining configuration settings Editing the configuration files with the administrative Web UI Editing the configuration files with a text editor Updating the agent configuration files on a client
See Examining configuration settings on page 308. See Editing the configuration files with the administrative Web UI on page 308. See Editing the configuration files with a text editor on page 312. See Editing an agent configuration file to improve backup and restore performance on page 318.
308
See Editing an agent configuration file to accommodate large backups on page 319. See Updating the agent configuration files on a client on page 313. For the parameters that you can use to tune PureDisk, see the PureDisk Best Practices Guide.
1 2 3
Click Settings > Configuration. In the left pane, click the plus sign (+) to the left of Configuration file templates. Expand one of the PureDisk templates. For example, expand PureDisk Storage Pool Authority > Default Value Set for PureDisk Storage Pool Authority > watchdog > interval.
Make a copy of the default value set you want to change. See Making a copy of a value set on page 309.
Navigate to a configuration value in the copy of the value set. See Navigating to a value in the configuration file copy on page 309.
Reconfiguring your PureDisk environment Editing the configuration files with the administrative Web UI
309
Changing a configuration value or deleting a configuration value. See Changing a configuration file value or deleting a configuration file value on page 310.
Assign the new template and, optionally, push the configuration file changes. You can assign the template and wait to push the changes, or you can assign the template and push the changes in one operation. See Assigning the template and, optionally, pushing the configuration file changes on page 311.
1 2 3
Click Settings > Configuration. In the left pane, expand Configuration File Templates. Expand an agent or service. For example, expand PureDisk Client Agent.
Select the default value set. For example, select Default Value Set for PureDisk Client Agent.
5 6 7
In the right pane, click Copy Value Set. In the left pane, select the copy of the value set. (Optional) In the right pane, in the Value set name (new) field, type a new name for the copy of the value set. Alternatively, you can decide to accept the default name.
8 9
Click Save. Visually inspect the left panel to check that the copy appears.
310
Reconfiguring your PureDisk environment Editing the configuration files with the administrative Web UI
Note: Do not edit a line if its default value contains brace characters. For example All OS:{{$agentid}}.These are system variables. To navigate to the content of a configuration file field
(Conditional) Click Settings > Configuration. Perform this step if the copy of the default value set does not already appear in the left pane.
(Conditional) In the left pane, click the plus sign (+) to the left of Configuration file templates. Perform this step if the copy of the default value set does not already appear in the left pane.
(Conditional) Expand the configuration file template so that you can see the copy of the default value set. Perform this step if the copy of the default value set does not already appear in the left pane.
4 5
Select the copy of the default value set you want to change. Proceed to the following: See Changing a configuration file value or deleting a configuration file value on page 310.
Reconfiguring your PureDisk environment Editing the configuration files with the administrative Web UI
311
In the left pane, expand the value set and select a line. For example, expand PureDisk Client Agent > Copy of Default Value Set for PureDisk Client Agent > progress > showicon > All OS:1.
In the right pane, edit the Value field or click Delete Configuration File Value. If you clicked Delete Configuration File Value, the administrative Web UI asks you to confirm your action. After you delete a value, you cannot add it back. For example, by default, the showicon value is set to 1, which is enabled. It shows the icon on the client while PureDisk does backups. To disable the icon, change the value to 0.
3 4
Click Save. Proceed to the following: See Assigning the template and, optionally, pushing the configuration file changes on page 311.
Assigning the template and, optionally, pushing the configuration file changes
After you edit the configuration, you can perform operations to test the new configuration. After the tests finish, however, you need to push the configuration file changes you made to the PureDisk system services in order for the changes to become permanent. If you do not push your configuration file changes, PureDisk the values revert to their previous settings the next time you push the configuration file changes. The following procedure explains how to propagate configuration file changes to the storage pool. To update PureDisk services
(Conditional) Click Settings > Configuration. Perform this step if the copy of the value sets does not already appear in the left pane.
In the left pane, click the plus sign (+) to the left of Configuration File Templates. Perform this step if the copy of the value set does not already appear in the left pane.
312
Reconfiguring your PureDisk environment Editing the configuration files with a text editor
(Conditional) Expand the tree in the left pane and select value set copy. Perform this step if the copy of the value set does not already appear in the left pane. For example, select PureDisk Client Agent > Copy of Default Value Set for PureDisk Client Agent.
4 5 6 7
In the right pane, click Assign template. Select the entities that you want to use this value set. Click Assign. (Conditional) Click Push Configuration Files. Perform this step if you want the changes you made to become permanent. The list of members should include the services you selected in 5. PureDisk monitors the configuration files that are pushed to each agent and checks if the value set has changed. PureDisk performs this check and creates update jobs only if the value set has changed since the last update job that ran for each agent. For example, if you push a value set to an agent twice without changing the value set, PureDisk creates only one job. If you use the Force option, the server-side change checking is ignored and an update job is always created.
(Conditional) Click Push. Perform this step if you also performed step 7. Confirm your actions in the dialog boxes that appear.
Reconfiguring your PureDisk environment Updating the agent configuration files on a client
313
Table 14-1
Configuration file
Storage pool authority Content router Metabase server Metabase engine Controller PureDisk node agent
Example 1. A client system crashes or has been reimaged. In this case, you can reregister the client on the storage pool authority. The original configuration file is completely lost. Assume the following series of events:
A user installs the PureDisk agent software on a client. The client system crashes. The user reinstalls the operating system software on the client. The user can run the pdregister command as shown in this procedure to restore the configuration file for this particular client.
314
Reconfiguring your PureDisk environment Updating the agent configuration files on a client
Example 2. You accidentally make erroneous edits to a configuration file. You might change configuration file parameters and later want to revert to the original configuration file. If the old configuration file exists and has a valid agent ID, you can obtain a new copy from the storage pool authority. Example 3 (Linux only). If you have SPAR enabled and you need to retrieve a new configuration file. SPAR enables you to replicate storage pool information from a remote storage pool to a main storage pool. The remote storage pool acts as a client to the main storage pool. For more information about how to use SPAR, see the PureDisk Administrators Guide.
In the following procedure, the format for the pdregister command is shown generically for MS Windows or UNIX clients. The .exe suffix applies only to Windows clients. To retrieve a new configuration file for a client
Invoke the PureDisk administrative Web UI and make sure that the client appears in the list of clients in the left pane when you select Manage > Agent. Do not perform this procedure if the client is not registered on the storage pool currently.
2 3
Log on to the client system as root (Linux or UNIX platforms) or as the administrator (Windows platforms). Change to the directory into which you installed the agent software. On Linux and UNIX platforms, change to install_path/pdag/bin. The default is /opt/pdag/bin. On Windows platforms, change to the directory into which you installed the agent. By default, this directory is C:\Program Files\Symantec\NetBackup PureDisk Agent\bin.
Run the pdregister command. To retrieve a new agent configuration file from the storage pool authority, type the command with the following parameters:
This set of parameters assumes that the original configuration file is completely lost. To ensure that PureDisk recognizes this client, you need to specify the agent ID.
Reconfiguring your PureDisk environment Updating the agent configuration files on a client
315
(Linux only) Specify --productname=SPARR if SPAR is enabled, and this host is a client to a main storage pool. Make sure that the logon and password belong to a user that has Agent Management permissions. To retrieve a new agent configuration file from the storage pool authority and overwrite an existing configuration file, type the command with the following parameters:
pdregister[.exe] --action=configfile --type=Agent --url=https://yourspa/spa/ws --todisk [--productname=SPARR]
This set of parameters assumes that the original configuration file still resides on the client. Your intent is to restore it to the form it has on the storage pool authority. You do not need to specify the agent ID or the logon credentials. More information about the parameters and arguments that pdregister accepts is available. See the PureDisk Client Installation Guide
(Conditional) Activate the agent. Perform this step if the agent is not activated. More information about how to activate the agent is available. See the PureDisk Client Installation Guide
Invoke the PureDisk administrative Web UI and make sure that the client appears in the list of clients when you click Manage > Agent. Do not perform this procedure if the client is not registered on the storage pool currently.
Log on to the client system as root for UNIX clients and as admin for MS Windows clients.
316
Reconfiguring your PureDisk environment Updating the agent configuration files on a client
Change to the directory into which you installed the agent software. On Linux and UNIX platforms, change to install_path/pdag/bin. The default is /opt/pdag/bin. On Windows platforms, change to the directory into which you installed the agent. By default, this directory is C:\Program Files\Symantec\NetBackup PureDisk Agent\bin.
Make sure that the logon and password belong to a user that has Agent Management permissions.
Chapter
15
See Editing an agent configuration file to improve backup and restore performance on page 318. See Editing an agent configuration file to accommodate large backups on page 319. See Multistreamed (parallel) backups on page 320. See Multistreamed (parallel) restores on page 321. See Segmentation options for backup jobs on page 322.
Any changes you make to these the configuration files are considered to be an advanced operation. You need to measure your results after each change. These changes require testing and experimentation. When you measure performance, measure the time for the putfiles job step, not total job time. If your backups and restore jobs seem slow, try increasing the number of streams and checking the resulting performance. In some cases, performance can be better if you use fewer streams.
318
To tune backup performance, you can edit the following settings under Settings > Configuration > Configuration File Templates > PureDisk Client Agent > Default Value Set for PureDisk Client Agent > backup:
DontSegmentTypes. The default value is an empty list, All OS:. DontSegmentThreshold. The default value is All OS:0. maxstreams. The default value is All OS:1. MaxRetryCount Specifies the maximum number of times that the backup job can attempt to send each file in the event of network errors. Specify any positive integer. The default is 5. The backup job stops and issues an error the first time it encounters a file that cannot be transmitted and the retry count has been exceeded.
The PureDisk Backup Operators Guide. See Segmentation options for backup jobs on page 322.
To tune restore performance, you can edit the following settings under Settings > Configuration > Configuration File Templates > PureDisk Client Agent > Default Value Set for PureDisk Client Agent > restore:
319
maxstreams. The default value is All OS:1. MaxSegmentPrefetchSize. Specifies the number of bytes to prefetch during a restore. The default is 16MiB. Valid units are B, KiB (1024), MiB, GiB, KB (1000), MB, and GB. When the value is set to zero (0), PureDisk disables prefetching. During restore, PureDisk fetches only one data segment at a time. This behavior is identical to the default restore behavior before PureDisk 6.6. If you restore to a client that is very low on memory, and you want to ensure that memory use is low during the restore job, you can set this value to zero (0) or to a value that is less than zero. SegmentChunkSize. Specifies the number of bytes of data to transfer over the network from the server to the client at one time. The default is 32KiB. Valid units are B, KiB (1024), MiB, KB (1000), and MB. The range for this setting is from 1KiB through 16 MiB. This setting has no effect if MaxSegmentPrefetchSize is set to zero (0).
/opt/pdag/var /opt/pdag/tmp
On Windows clients, the agent writes these lists to the following directory: C:\PROGRA~1\Symantec\NetBackup PureDisk Agent\var. During very large backups, these lists can grow beyond the space that you allocated to the / partition, which is typically kept relatively small. If you expect this space problem might happen on a client, use the following procedure to modify the agent configuration file on that system. Note: Repeat this procedure each time you update the agent configuration files through the administrative Web UI. The repetition is necessary because updates through the administrative Web UI overwrite all agent configuration files.
320
In the [paths] section, type new paths for the following parameters:
var temp
The new paths must be full paths, not relative paths. They must refer to a partition large enough for the backups.
4 5
Save and close the file. Type the following command to start the agent service:
# /opt/pdag/bin/pdagent
If the number of data entries is less than the number of requested streams, a single stream backup is used. If the number of data entries is less than 4096 per stream, they are balanced across the streams. If the number of data entries are more than 4096 per stream, each stream is given 4096 data entries. As each stream finishes, it is given the next 4096 data entries until no entries remain.
For multistreamed backup jobs, a distinction is made between transient and fatal error conditions. An example of a transient error condition is a temporary loss
321
of connection with remote storage. Examples of fatal error conditions are an empty routing table, an invalid data selection ID, or an out-of-memory error. If you use multiple backup streams, the behavior of the Sorting option is slightly changed. That occurs, for example, if you set Number of backup streams in the Parameters tab of a backup policy to 2 or more. If you enable both sorting and multistreaming, the PureDisk agent first sorts the complete list of files by their size. It then distributes small chunks of the sorted list over the available stream. For this reason, and because PureDisk backs up multiple files at one time, the order in which it backs up the files cannot match exactly the order of the files in the sorted list. For large backups, the trend of the backup order is still as expected (smallest files first), but for small backups of only a few files, there may be some anomalies. If you run too many parallel backup streams, you can overload the client system. If you overload the client, one of the following occurs:
The backups might fail with a status of SUCCESS_WITH_ERRORS. In this case, not all files are transferred, and there is no obvious error in the log file. PureDisk might display the following message:
Insufficient system resources exist to complete the requested service
When disk usage is between 40% and 80%, backups with a single write thread can have faster throughput than backups configured with two write threads. When the disk utilization is between 90% and 100%, backups with two write threads can have faster throughput than backups with one or four write threads. You might need to experiment with more than one approach to performance tuning. Or you might need to use different combinations of streams and segment size values. The exact values depend on the specific client and its hardware configuration. Symantec recommends that you start with a small number of streams and that you increase max_streams only if the backup performance is unacceptable. At some point, if you increase max_streams, performance does not improve. A max_streams value that is too large provides no benefit, can overload the client system, and can cause backups to fail.
322
length of the data entries list and the number of streams requested, PureDisk determines and initializes the optimal number of threads. PureDisk validates the data list. After the validation, if it determines that the number of data entries is fewer than three per stream, it uses a single stream restore.
See Segmentation threshold value on page 322. See File-type segmentation on page 322.
By editing these segmentation values, you might be able to improve performance. A difference in backup performance is more noticeable if there is a high latency on the connection between the agent and the content router(s). Note: If these two options are not set identically on all agents, the effectiveness of PureDisk deduplication option (PDDO) can be reduced. If you enable PDDO, the MATCH_PDRO parameter is enabled by default. When you enable the MATCH_PDRO parameter, it specifies that the PureDisk calculate the segment size based on the file size, which is the same method by which PureDisk calculates the segment size for a typical backup.
File-type segmentation
A comma-separated list of file types (identified by file name suffixes) can be set in the administrative Web UI or in the agent configuration file.
323
As a PureDisk agent transfers a file to a content router, it checks whether the files suffix is contained in this list. If the suffix is in the list, the file is transmitted in only one segment. If the file is larger than the maximum segment size, it is still transmitted in multiple segments. However, each segment is the maximum segment size , except possibly the last segment.
Unexpected results
Multiple backup streams or restore streams can have the following unexpected results:
Multistreaming can increase the load on content routers. With some fast network or hardware configurations, multistreaming can slow the performance of other backup or restore jobs. For example, if the PureDisk metabase is on the same node as a content router, the metabase import step can take longer. The content router might not have completed its processing, because the spool is filled faster than the content router empties it. This delay of the metabase import can lead to a longer total backup time, but during that time, the client and the wire are idle. Multistreaming can stop other agents from connecting to a content router because each job stream needs one of the limited connections (limited to 300). Client backup and restore performance might be affected if all parallel streams do central-processor-bound fingerprint calculations. If you run too many parallel backup streams, you can overload the client system. When that occurs, the system generates the following error: Unable to resolve host name. PureDisk generates this error because each backup stream has its own content router context. Because of the unique contexts, each stream makes DNS address lookups for the content router address(es). If the content router does not get a response from DNS lookups in one minute, it reports this error. If backups have many streams, DNS lookups can fail when the central processor usage is high and some streams are not rescheduled fast enough. If you add multistreaming after the CPU/wire is full, you see no performance increase.
324
Parallelism. To increase parallelism in PDDO replication, set the maxstreams parameter to a larger value. The higher the value of the maxstreams parameter, the more parallelism is obtained. The valid range is from 1 to 10 streams. Bandwidth. If you need to limit the bandwidth that is used between storage pools during replication, you can set the bandwidthlimit parameter. PureDisk uses only the amount of bandwidth that is specified in this parameter. Every stream uses the specified bandwidth value. For example, if you use four streams and a bandwidth of 200 KB, PureDisk uses 800 KB of bandwidth between storage pools during replication. Network errors. If you want to minimize the effect of network errors on replication, you can use the following parameters: maxretrycount, sleeptime, and maxsleeptime. For example, the replication process might not be able to complete because of network errors. In this case, you can specify the number of times that the source storage pool needs to wait before trying to send the data again to the destination storage pool. Depending on network conditions, you might specify to wait a few seconds or wait a few minutes.
1 2 3
Log into the source storage pool's administrative Web UI. Click Settings > Configuration > Configuration File Templates > PureDisk Server Agent > Default Value Set for PureDisk Server Agent > replication. Change the values for one or more of the following configuration parameters under replication:
bandwidthlimit. Specifies the bandwidth limit, between the storage pools, for the replication activity, in KB/sec. If you set this parameter to zero (0), you specify unlimited bandwidth. The default is 0 (disabled). maxretrycount. Specifies the maximum number of times that the PureDisk replication workflow can attempt to send data. Specify any positive integer. The default is 10. maxsleeptime. Specifies the maximum amount of time, in seconds, that the replication workflow is permitted to sleep between retries. Specify any positive integer. The default is 60. maxstreams. Specifies the maximum number of PDDO replication streams per replication job. The maximum value is 10; if you set this parameter to a value greater than 10, PureDisk uses 10 streams per replication job. Specify any positive integer. The default is 4.
325
sleeptime. Specifies the number of seconds that the replication workflow can sleep between two retries. Specify any positive integer. The default is 10.
The delayeddosize and maxsocachesize parameters also affect replication, but Symantec recommends that you do not reset these parameters without the assistance of Symantec Technical Support staff. For information about how to edit the configuration file parameters, see the following: See About the configuration files on page 307.
326
Appendix
Third-party legal notices for the Symantec NetBackup PureDisk product family Third-party trademarks for the Symantec NetBackup PureDisk product family
Third-party legal notices for the Symantec NetBackup PureDisk product family
Third-party software may be recommended, distributed, embedded, or bundled with this Symantec product. Such third-party software is licensed separately by its copyright holder. All third-party copyrights associated with this product are listed in the following document, which is accessible from the PureDisk storage pool configuration wizard's landing page: See PureDisk Third-party Legal Notices
328
Third-party legal notices Third-party trademarks for the Symantec NetBackup PureDisk product family
Products bearing SPARC trademarks are based upon an architecture developed by Sun Microsystems, Inc. AMD is a trademark of Advanced Micro Devices, Inc. Firefox and Mozilla are registered trademarks of the Mozilla Foundation. Intel, Itanium, Pentium, and Xeon are trademarks or registered trademarks of Intel Corporation or its subsidiaries in the United States and other countries. Java, Sun, and Solaris are trademarks or registered trademarks of Sun Microsystems, Inc., in the United States and other countries. Linux is a registered trademark of Linus Torvalds in the United States and other countries. Mac OS is a trademark of Apple Inc., registered in the U.S. and other countries. Nessus is a trademark of Tenable Network Security, Inc. NetApp is a registered trademark of Network Appliance, Inc. in the U.S. and other countries. Novell and SUSE are registered trademarks of Novell, Inc., in the United States and other countries. OpenLDAP is a registered trademark of the OpenLDAP Foundation. Oracle is a registered trademark of Oracle and/or its affiliates. Red Hat and Enterprise Linux are registered trademarks of Red Hat, Inc., in the United States and other countries. UNIX is a registered trademark of The Open Group. VMware, vSphere, and the VMware "boxes" logo and design are trademarks or registered trademark of VMware, Inc., in the United States and other countries.
Index
A
Accessing a managed storage pool 288 ACL errors restore job statistic 201 ACL restore job statistic 201 activating a new PureDisk component 272 Activating a server agent 109 adding PureDisk components 265 Administrative Web UI Starting an additional 288 Average restore rate restore job statistic 201
C
central storage pool disabling 287 enabling 285 managing 287 testing connections 288 changeFQDN.sh script 258 clustering administering a storage pool 295 configuration files editing agent configuration files for large backups 319 editing ASCII files 312 editing through the administrative Web UI 308 pushing changes 311 updating agent configuration files 313 configuring see reconfiguring PureDisk 307 configuring SPAR 177178 copying a replication policy 85
B
Backup job statistics 195 Backup networks, specifying 69 Backup speed backup job statistic 195 Backup time duration backup job statistic 195 Backup time duration PDDO backup job statistic 205 Bytes deleted in source data selection replication job statistic 203 Bytes deleted on source backup job statistic 195 Bytes modified in source data selection replication job statistic 203 Bytes modified on source backup job statistic 195 Bytes modified on target restore job statistic 201 Bytes new in source data selection replication job statistic 203 Bytes new on source backup job statistic 195 Bytes new on target restore job statistic 201 Bytes not modified on source backup job statistic 195 Bytes received by agent restore job statistic 201 Bytes replicated replication job statistic 203 Bytes scanned during backup PDDO backup job statistic 205 Bytes selected on source backup job statistic 195 Bytes total restore job statistic 201 Bytes transferred backup job statistic 195 Bytes transferred replication job statistic 203 Bytes transferred to content router PDDO backup job statistic 205 Bytes unmodified on target restore job statistic 201 Bytes with errors restore job statistic 201
D
dashboard reports 227 data lock password export to NetBackup 100 files tab 191 data mining reports 210 data replication effect when changing to host names 263 effect when changing to IP addresses 263 data selection removal policies scheduling 43 Dedicated networks for backups, restores, and replication 69 deleting a replication policy 85 Devices restore job statistic 201 Directory count restore job statistic 201
330
Index
disaster recovery backup procedures creating backup or restore scripts 141 policies 130 troubleshooting 143 using NetBackup 119, 124 using scripts 119, 139 restoring PureDisk 150 strategies 177
I
Internal networks for backups, restores, and replication 69 international characters 248 Items deleted in source data selection replication job statistic 203 Items modified in source data selection replication job statistic 203 Items new in source data selection replication job statistic 203 Items replicated replication job statistic 203 Items with replication errors replication job statistic 203
E
EEBs, applying 59 Error count restore job statistic 201 export engine see NetBackup - export engine 89 external directory service authentication adding PureDisk groups 20 changing TLS specification 47 disabling 46 introduction 18 linking with PureDisk 34 maintaining PureDisk groups 44 modifying the base search path 47 synchronizing with PureDisk 41, 43 verifying TLS 30
L
license keys adding and removing 284 report generation 288 required 90 viewing in reports 236 log files 237
M
Managed storage pool Accessing 288 mapping file 260 Media server cache hit percentage PDDO backup job statistic 205 metabase engine rerouting 289 multistreamed (parallel) backups 320 multistreamed (parallel) restores 321 multistreamed replication 323
F
file-type segmentation 322 Files deleted on source backup job statistic 195 Files modified on source backup job statistic 195 Files modified on target restore job statistic 201 Files new on source backup job statistic 195 Files new on target restore job statistic 201 Files not modified on source backup job statistic 195 Files selected on source backup job statistic 195 Files unmodified on target restore job statistic 201 Files with errors restore job statistic 201
N
NetBackup disaster recovery 119, 124 export engine 89 adding the service 265 configuring 91 exporting Files and Folders data selections 102 job failures 109 point-in-time export 108 restoring 111 server agents and export jobs 109 Networks for backups, restores, and replication 69
G
Global data reduction factor backup job statistic 195 Global data reduction saving PDDO backup job statistic 205 Global data reduction savings backup job statistic 195
H
Hard links restore job statistic 201 Hot fixes, applying 59
Index
331
O
optimization 317, 323
P
Patches, applying 59 PDDO backup job statistics 205 pdkeyutil utility 143 PDLinux changing the password 295 performance optimization file-type segmentation 322 multistreamed (parallel) backups 320 multistreamed (parallel) restores 321 segmentation options for backup jobs 322 segmentation threshold values 322 policies data mining 210 disaster recovery backup procedures 130 exports to NetBackup 102 NetBackup DataStore 108 replication 79 run once 83, 132 processes stopping and starting on a multinode PureDisk storage pool 303 stopping and starting on a PureDisk node (clustered) 302 stopping and starting on a PureDisk node (unclustered) 300 PureDisk administrative Web UI Starting an additional 288
R
reconfiguring PureDisk changing to host names 258 changing to IP addresses 258 changing topology.ini 258 editing agent configuration files for large backups 319 editing configuration files 308 overview 307 updating agent configuration files 313 using scripts 258 recovery from a disaster 150 Release updated, applying 59
removing a content router 280 replication copying replicated data 87 managing replicated data selections 86 policies copying and deleting 85 creating 79 full replication 83 incremental replication 83 scheduling 83 viewing replication jobs 85 restoring replicated data 87 see storage pool authority replication (SPAR) 175 understanding 77 using replicated data 86 viewing replicated data selections 86 Replication job statistics 203 Replication networks, specifying 69 replication policy creating 80 editing 80 Replication time duration replication job statistic 203 replication tuning 323 reports central storage pool 231, 285 central storage pool test connection 288 dashboards 227, 231 data mining 210 finished job 194 for a running job 191 overview 190 permissions 190 web service 219 rerouting a metabase engine 289 Restore job statistics 201 Restore networks, specifying 69 Restore time duration restore job statistic 201 RestoreSPASIO command 185 restoring from NetBackup 111 retrieving information from a data mining policy 213 Rollback 59
S
segmentation options for backup jobs 322 segmentation threshold value 322 server agents activating 109 single port communication 49
332
Index
single port communications 83 SIS see data deduplication 215 Source bytes backed up backup job statistic 195 Source bytes with errors backup job statistic 195 Source files backed up backup job statistic 195 Source files with errors backup job statistic 195 Specify networks link in configuration wizard 69 Start date/time backup job statistic 195 Start date/time PDDO backup job statistic 205 Start date/time replication job statistic 203 Start date/time restore job statistic 201 Starting another PureDisk administrative Web UI 288 Statistics for a backup job 195 Statistics for a PDDO backup job 205 Statistics for a replication job 203 Statistics for a restore job 201 Stop date/time backup job statistic 195 Stop date/time PDDO backup job statistic 205 Stop date/time replication job statistic 203 Stop date/time restore job statistic 201 stopping and starting processes multinode PureDisk storage pool 303 PureDisk node (clustered) 302 PureDisk node (unclustered) 300 storage pool authority replication (SPAR) disaster recovery strategies 177 enabling backups 180 example 175 RestoreSPASIO command 185 restoring 183 running a SPAR policy manually 182 upgrading PureDisk with SPAR enabled 187 storage pool management activating a new PureDisk service 272 adding and removing license keys 284 adding PureDisk components 265 adjusting the clock on a PureDisk node 298 administering a clustered storage pool 295 changing a storage pool's time zone 297 changing database or LDAP admin passwords 295 changing the PDLinux password 295 converting storage pools 263 creating a mapping file 260 disabling central reporting 287 enabling central reporting 285 increasing the number of client connections 296
storage pool management (continued) license key report generation 288 managing in the central storage pool 287 reconfiguring PureDisk 258 removing a content router 280 rerouting a metabase engine 289 testing connections 288 storage pools stopping and starting processes 303 Symbolic links restore job statistic 201
T
time zone, changing 297 Time-out interval, adjusting 299 Total files restore job statistic 201 troubleshooting NetBackup export engine jobs 109 tuning 317, 323
U
Unique bytes backed up backup job statistic 195 Unique files and folders backed up backup job statistic 195 Unique items received restore job statistic 201 Unique items restored restore job statistic 201 user authentication root broker 231
V
Verification failures restore job statistic 201 Verification successes restore job statistic 201
W
web service reports 219 Web UI See Administrative Web UI 288
Glossary
absolute path
The location of a given file or directory on a file system regardless of the current working directory. for a particular file or directory.
ACL (access control list) The information that identifies specific users or groups and their access privileges
active agent
The PureDisk software that is enabled for use within a PureDisk environment. The PureDisk agent (the software) is installed on clients (the hardware). The nodes in a high availability cluster on which services are running. data selection automatically when you perform a full system backup. A PureDisk storage pool with all PureDisk services installed on one node. A client other than the source client that receives restored files. If the data is not restored to the original client, the client that is designated to receive the data is the alternate client. See single-port communication. An installation that requires active interaction with a person. The process that communicates with a root broker to verify user identities. A process where selected files on a computer drive are copied and stored on a reliable form of media. A user or a group of users with the rights to initiate client backups. A connection between the PureDisk client and the PureDisk content router through which data is sent. In case of multistreaming, the client establishes multiple connections to the same content router and distributes the total volume of backup data over all available connections. Multistreamed backups (vs. single streamed) increase the aggregated throughput from client to content router, thus allowing backups to finish sooner. The timeframe in which backups are permitted. A trusted third-party organization or company that issues digital certificates that are used to create digital signatures and public-private key pairs. The role of the CA in this process is to guarantee that the entity granting the unique certificate
active node
Agent Files and Folders A collection of files, folders, or directories to be backed up. PureDisk creates this data selection All-in-one storage pool alternate client
ATOP (All Through One Port) attended installation authentication broker backup
334
Glossary
is, in fact, who it claims to be. This means that the CA usually has an arrangement with the requesting entity to confirm a claimed identity. CAs are a critical component in data security and electronic commerce because they guarantee that the two parties exchanging information are really who they claim to be.
central reporting
A PureDisk feature in which one or more storage pools send reporting data to another storage pool. different platforms and computers can share files. A set of hosts (each termed a node) that share a set of disks and are connected by a set of redundant heartbeat networks. A shared directory structure. Common root is a concept applicable to data restore. When the user restores the data, there is a Do not restore common root option. The user can enable or disable the feature. The files that define PureDisks methods and assumptions. A custom configuration file for each component is created automatically by the storage pool authority. It is pushed out to the component for which this file is created. If you need to tune any of the default PureDisk processes, you can edit the configuration files. A service that stores and retrieves file content. PureDisk breaks larger files into segments and distributes the segments across the available content routers. This workflow removes files and objects that cannot be removed during the normal data removal process.
CIFS (Common Internet A protocol that defines a standard for remote file access. CIFS allows users with File System) cluster
common root
configuration files
content router
content router garbage The process of removing unneeded data objects and files from the content router. collection
CSV (comma-separated A text file that uses commas as data delimiters. variable) file data lock password
An option that allows the administrator to require users to enter a password before they perform certain operations. These operations include any operation that exposes directory names, directory content, file names, or file content. The process of collecting information about all the files in a PureDisk storage pool. The process of removing old and unneeded versions of the files that PureDisk previously backed up. A list of files, directories, or other data objects that you want PureDisk to back up. They can be used to specify files and folders, databases, system information, and other types of data. After a data selection is created, it can be backed up automatically (through a policy) or on demand (initiated by a backup operator).
data mining
data removal
data selection
Glossary
335
The process of removing the entire data selection (including file content data) from the content router. This process also removes the associated metadata information from the metabase engine. for PureDisk to back up. Templates can be developed that include or exclude certain file types, or that back up a specific directory. PureDisk includes some default data selection templates.
data selection template A pattern that is used for creating the list of files, directories, or other data objects
deduplication
The process of dividing a file into segments, comparing each segment with the previously stored file segments, and then storing only the unique segments. Deduplication significantly reduces the amount of data that is stored because redundant data is replaced with a pointer to the unique data copy. A logical collection of client systems. a policy includes deactivated data objects, the policy does not include them when the policy runs.
department
disabled data selection A data selection that PureDisk ignores when it performs policy-based actions. If
disabled policy
A policy that exists in the PureDisk environment but that is currently not activated. No jobs are created for this policy. The process of restoring information from a backup after the original data was lost (due to a disaster) or deleted. An input/output (I/O) enhancement technique that balances I/O across many available paths from the computer to the storage device to improve performance and availability. An activated list of files, directories, or other data objects that PureDisk uses for back ups and other policy-based actions. A policy that exists in the PureDisk environment that is currently activated and run according to a schedule. PureDisk automatically creates jobs to run this policy. A defined procedure that takes place when an event occurs. A significant occurrence in a system or application that a program detects. Events typically trigger actions, such as sending a user notification or adding a log entry. place.
disaster recovery
enabled policy
event escalation action A defined procedure that takes place when a specific notable occurrence takes
A listing of files or file patterns that are not included in a data selection. The means by which PureDisk determines the files or folders that should not be part of a given data selection. See also inclusion rules. The process of loading the PDLinux software onto a computer with multiple hard drives where none of the options is predefined.
336
Glossary
external authentication A credential verification authority that resides on a computer that is not part of
The process of moving services from the active node in a cluster to a passive one. The frequency with which files on a client system are modified. A character sequence that includes wild cards and instructs PureDisk to select multiple files based on the character sequence. The ability to search through a graphic representation of a computer systems file structure. A data selection that is used to back up files, folders, and directories. A unique sequence of digits identifying a file or a file segment. The fingerprint of a file or segment is computed from the file or segments content and is unique for that file or segment. An absolute path that may contain wild cards and instructs PureDisk to select folders based on the character sequence. An unambiguous domain name that specifies the exact location of a computer within the domain's hierarchy. The process of removing stale data or records from PureDisk that cannot be removed during the normal data removal process. A document that indicates the various components that are known to work with a given software product. A signal sent at regular intervals to indicate that a host and its connections are operating normally. A system or a resource that is continuously operational. The TCP/IP address of a computer. A client computer that is registered to the storage pool authority, but which is not yet acknowledged as part of the PureDisk environment. A listing of files or file patterns that are included in a data selection. The means by which PureDisk determines the files to consider as part of a given backup. See also exclusion rules. The process of receiving attributes from a parent object, such as a template. An operation that has been scheduled for processing. Jobs contain source or destination information, settings, and a schedule.
folder pattern
inheritance job
Glossary
337
A software protocol that enables anyone to locate organizations, individuals, and other resources such as files and devices in a network, whether on the Internet or on a corporate intranet. LDAP is a lightweight (smaller amount of code) version of Directory Access Protocol (DAP), which is part of X.500, a standard for directory services in a network. A logical grouping of one or more departments. A location does not need to be a physical or a geographic location. A persistent connection within Microsoft Windows to a shared directory on a remote server that is assigned a drive letter. The drive mapping may or may not survive a restart, depending on how it is configured. A two-column file that lists both IP addresses and fully qualified domain names (FQDNs) for each node. Mapping files are used when redefining a PureDisk environment to use either FQDNs or IP addresses. A distributed database that contains all the metadata about the files that are backed up by PureDisk. The service that maintains and manages file metadata information in the metabase database. During backup, the PureDisk agent records all relevant metadata information (the file attributes) of each file it backs up. File attributes include the file name, its location, its size, its type, and so on. The metabase engine stores these metadata records in its database. The metabase engine manages an inventory of all files that are backed up. The process of removing the stale, unneeded information from the metabase that cannot be removed during the normal data removal process. The service that redirects metabase queries to the correct metabase engine(s). Each metabase engine in the storage pool is responsible for managing the metadata records from a group of agents. Queries for a file or files do not go directly to the metabase engine but instead are directed to the metabase server. The metabase server redirects the query to the metabase engine that manages the metadata records of the agent that made the query or to which the query is associated. In some cases the metabase server may have to redirect the incoming query to multiple metabase engines. Structural data describing the attributes of files on a disk. A predefined data selection within PureDisk that backs up Microsoft Exchange server databases. A predefined data selection within PureDisk that backs up Microsoft SQL server databases. clustered.
location
mapping file
metabase
metabase engine
multinode storage pool A storage pool that includes more than one PDLinux server node and can be
338
Glossary
multistreaming
The process of establishing multiple connections to the content router for the purposes of backing up data faster. An optional service that allows backed up PureDisk Files and Folders data selections to be exported to NetBackup. The NetBackup export engine does not export other PureDisk data selection types. A directory on a remote server that is designated as shared. A computer in a storage pool that hosts PDLinux and other PureDisk services. A network protocol that is designed to work on TCP/IP stacks. It extracts information from a hierarchical directory such as X.500. This software gives users a single tool to find a particular piece of information. For example, a user can find a user name, an email address, security certificate, or other contact information. A predefined data selection within PureDisk that backs up Oracle databases. The process of redistributing data among the content routers in a PureDisk storage pool. All content routers are actively involved in the redistribution at the same time. Any node in a highly available, clustered environment that is not running PureDisk services. The directory location of a given file or directory on a file system. Paths can be either relative or absolute. A plug-in that uses the NetBackup OpenStorage API to enable NetBackup to write backups to a PureDisk storage pool. All NetBackup data that is written to a PureDisk storage pool is deduplicated. The operating system that hosts the PureDisk application. Symantec developed PDLinux based on SUSE Linux Enterprise Server. A method for managing backup jobs and strategies. Policies contain settings for jobs. backup failure) takes place.
passive node
path
policy escalation action A procedure that is defined to take place when a specific policy event (such as a
private network
A computer network that is accessible only by other servers, not by the general user base. A network interface card that communicates to a private network. A computer network that is accessible to the general user base. A network interface card that communicates to a public network. The process of enrolling with the storage pool authority.
Glossary
339
relative path
The directory location of a given file or directory on a file system that depends on the current working directory. The process of copying backed up data selections from one storage pool to another storage pool. The process of redistributing data over all available content routers. When the process finishes, each content router stores a volume of data proportional to its relative capacity. Rerouting is necessary when a new content router is activated, or an existing content router is deactivated. The unique identifier for a service on a PureDisk node. The authentication authority in the network. A root broker is local when it resides on the same physical computer as the PureDisk storage pool authority. A root broker is remote when it resides on a different PureDisk computer in the network. A root broker is external when it does not reside on any PureDisk computer within the storage pool. A piece of a file. The process of breaking a file down into smaller pieces for backup.
replication
rerouting
segment segmentation
segmentation threshold The maximum allowable size for a file fragment. serial rerouting
The process of redistributing data among the content routers in a PureDisk storage pool. Only one content router redistributes data at a time. A PureDisk software component. The possible services are as follows: content router, metabase engine, metabase server, storage pool authority, and NetBackup export engine. The TCP/IP address associated with a service group. A collection of PureDisk services. A physical hard drive on a computer that can be remotely accessed from another computer. In a highly available cluster, the shared disk is normally a drive that does not physically reside in any of the cluster nodes. Any resources that can failover among cluster nodes must reside on a shared disk. A network directory, to which multiple users have read and write access, used to exchange files with other users. An installation in which the user sees no indication that the installation is occurring. The user is not prompted to enter any information and the user does not see status messages. See also attended installation. A PureDisk feature that directs all network communication through one port. Storage pools that implement single-port communication require fewer firewall ports to be open between PureDisk service agents and clients.
service
shared folder
silent installation
340
Glossary
snapshot
A consistent point-in-time view of a volume that is used as the reference point for the backup operation. After a snapshot is created, the primary data can continue being modified without affecting the backup operation. The service that manages a storage pool. The replication of storage pool authority configuration information from an all-in-one local storage pool to a main storage pool. The main data repository in PureDisk. PureDisk writes backup copies of content and metadata to the disk storage that is associated with a storage pool. A storage pool consists of one or more PureDisk nodes. A sequence of digital data. A distribution of Linux software. The name is an acronym for the German phrase Software-und System-Entwicklung (Software and system development). A data selection that is used to back up Microsoft system data on Windows platforms. The process of receiving attributes from a parent template. An encrypted protocol that provides secure communications in the PureDisk environment. The types of PureDisk services that a storage pool includes. All storage pool topologies include one or more of the following services: storage pool authority, content router, metabase engine, metabase server. A controller is installed on a metabase engine. Optionally, a storage pool can also include a NetBackup export engine. An installation that does not require human interaction. Also use this data selection to indicate the path for a NetApp Filer.
SPA (storage pool authority) SPAR (storage pool authority replication) storage pool
stream SUSE
System State and Services data selection template inheritance TLS (transport security layer) topology
unattended install
UNC path data selection A data selection that backs up data on a CIFS network drive on a Windows client.
user
An individual with rights to access your protected network resources. Users are defined by creating a user account that consists of a unique user name and authentication method. A collection of users with identical permissions. These users can perform common functions within a PureDisk environment. The process of cleaning up and optimizing a database. Vacuuming removes the records that are no longer needed and results in better database performance. High-availability cluster software developed by Symantec for UNIX, Linux, and Windows platforms.
user group
vacuuming
Glossary
341
VEA (Veritas Enterprise A separate middleware server used by the SAN Access Layer and other processes Administrator)
to provide client-server communication. The VEA infrastructure enables software components to share information about objects, manage those objects, and effect change on those objects. A graphical user interface that displays reports and other information for users of CommandCentral Service through a standard Web browser. The Console provides a central point to manage cost analysis and chargeback for services, managing workflow, displaying and managing reports, and other tasks. Symantec software that backs up open files. PureDisk uses VSP on Windows 2000 clients. A set of application programming interfaces (APIs) that creates a framework. Within this framework, volume backups and application write can occur at the same time. PureDisk uses Microsofts VSS technology to back up open files on Windows 2003 and Windows XP clients. A Symantec product installed on storage clients that enables management of physical disks as logical devices. It enhances data storage management by controlling space allocation, performance, data availability, device installation, and system monitoring of private and shared systems. A symbol that enables multiple matching values to be returned based on a shared feature. Two wildcards are available: the question mark (?) and the asterisk (*). The question mark stands for any single character, and the asterisk stands for any character string of any length. For example, the file specification *.* returns all files, regardless of their file names; the file specification *.sc? would return all file names that have a three-character extension beginning with sc (such as compusrv.scr, compusrv.scx, and so on). A collection of steps that the software completes to accomplish a task. A journaling file system that you can configure on a PureDisk node.
CommandCentral Console
wildcard character
YaST (Yet another Setup The operating system installation tool for SUSE Linux. Tool)
342
Glossary