Professional Documents
Culture Documents
V
MEMORY
LEGAL NOTICE
Copyright 2010-2013 Violin Memory, Inc. All rights reserved.
Violin Memory, Violin Technologies, Violin, vSHARE, vCACHE, Flash Forward, and Violin and Design are trademarks,
registered trademarks or service marks of Violin Memory, Inc. (Violin) in the United States and other countries.
All other brands, product names, company names, trademarks, and service marks are the properties of their respective
owners.
This document and the associated software product are protected by copyright and international treaties, and are
distributed under license from Violin, including restrictions on their use, copying, redistribution and reverse engineering.
Unless otherwise agreed by Violin in writing, Violins standard end user license agreement shall apply, which may be
reviewed at www.violin-memory.com/legal. No part of this document may be reproduced, adapted or translated without
prior written permission of Violin, except as permitted under applicable copyright law. The associated software product may
include, access or otherwise operate, interface or be delivered with third party software or other applications or copyrighted
materials, which are copyrighted and licensed by Violin suppliers. Such third party materials and licenses are identified in
this document and/or at www.violin-memory.com/legal.
Violin assumes no responsibility for any typographical, technical or other error or omission in this document. Violin reserves
the right to periodically change the information contained in this document, but Violin makes no commitment to provide any
such changes, updates, enhancements or other additions in a timely manner or at all.
The only warranties for Violin products and services are set forth in the express warranty statements accompanying such
products and services. Nothing herein should be construed as constituting an additional warranty. THIS DOCUMENT
(INCLUDING ANY EXAMPLES AND OTHER INFORMATION CONTAINED HEREIN) IS MADE AVAILABLE AS IS
WITHOUT REPRESENTATION OR WARRANTY OF ANY KIND. VIOLIN MAKES NO REPRESENTATION OR
WARRANTY IN THIS DOCUMENT REGARDING ANY ASSOCIATED SOFTWARE OR ANY OTHER VIOLIN OR THIRD
PARTY HARDWARE, SOFTWARE OR OTHER PRODUCTS OR SERVICES REFERENCED HEREIN. TO THE FULLEST
EXTENT PERMITTED BY LAW, VIOLIN (FOR ITSELF AND ITS LICENSORS AND OTHER THIRD PARTIES IDENTIFIED
HEREIN) HEREBY DISCLAIMS ALL REPRESENTATIONS AND WARRANTIES, WHETHER EXPRESS OR IMPLIED,
ORAL OR WRITTEN, WITH RESPECT TO THE FOREGOING, INCLUDING WITHOUT LIMITATION, ALL IMPLIED
WARRANTIES OF TITLE, NON-INFRINGEMENT, QUIET ENJOYMENT, ACCURACY, INTEGRATION,
MERCHANTABILITY OR FITNESS FOR ANY PARTICULAR PURPOSE.
IN NO EVENT SHALL VIOLIN (OR ITS LICENSORS OR ANY OTHER THIRD PARTY IDENTIFIED HEREIN) BE LIABLE
CONCERNING ANY USE OF THIS DOCUMENT, REGARDLESS OF THE FORM OF ANY CLAIM OR ACTION
(WHETHER IN CONTRACT, NEGLIGENCE, STRICT LIABILITY OR OTHERWISE), FOR ANY DIRECT, INDIRECT,
PUNITIVE, INCIDENTAL, RELIANCE, SPECIAL, EXEMPLARY OR CONSEQUENTIAL DAMAGES, INCLUDING
WITHOUT LIMITATION, ANY LOSS OF DATA, LOSS OR INTERRUPTION OF USE, COST OF PROCURING
SUBSTITUTE TECHNOLOGIES, GOODS OR SERVICES, OR LOSS OF BUSINESS, REVENUES, PROFITS OR
GOODWILL, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGES.
Violin Memory, Inc.
685 Clyde Avenue
Mountain View, CA 94043
USA
ii
535-0008-00 Rev 10
Table of Contents
Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
CHAPTER 1. Violin 6000 Series Memory Array Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
New in This Release. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
About the Memory Array. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Violin 6000 Series Architecture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
7
8
8
9
Product Models. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
Violin 6000 Series Platform Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
Chassis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
System LEDs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .15
Power Supplies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .16
Fans . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .16
CHAPTER 2. Manage, Monitor, and Administer with the Violin Web Interface . . . . . . . . . . . . . . . . . . 31
Introduction to the Violin Web Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
User Interface Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .32
System and Browser Requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .33
Abbreviations for Large Numbers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .34
Web Interface Help . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .34
535-0008-00 Rev 10
39
40
41
42
Home Dashboard . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
Dashboard . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
System View. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
LUN Status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Front and Rear Panel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
43
48
51
54
Manage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
Manage LUNs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
Manage Initiators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
Manage Targets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
Manage LUN Groups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
Data Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
Thin Provision LUNs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
Manage Snapshots . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
Space Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
Monitor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
Monitor Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Monitor LUNs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Monitor LUN Groups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Monitor Targets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
106
110
111
111
Administration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
Cluster Admin. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Network . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Notification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Tools. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
114
122
128
137
139
CHAPTER 3. Manage the Memory Array with the Web Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143
Introduction to the Memory Array Web Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143
High-Level Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Web Interface Requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Web Interface Help. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Configuring the Violin Web Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Printing Screens . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
144
145
145
146
146
149
150
151
152
154
158
163
168
170
171
Monitoring the Memory Array Using the XTree Browser Interface . . . . . . . . . . . . . . . . . . . . . . . . . . 173
ii
535-0008-00 Rev 10
iii
233
233
235
236
238
239
239
240
240
242
243
244
244
245
249
249
250
250
Snapshots . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 263
Creating Snapshots . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 264
Deleting Snapshots . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 264
Viewing Snapshot Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265
iv
535-0008-00 Rev 10
319
320
320
320
325
326
326
326
332
332
332
332
333
333
333
334
335
335
336
336
337
338
338
338
339
340
341
344
346
350
353
356
535-0008-00 Rev 10
535-0008-00 Rev 10
vii
varray . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 405
vcounts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 406
veeprom . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 409
vincident -a . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 410
vinfo . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 410
vinventory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 412
vmesg. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 413
vpartial . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 413
vring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 415
vspeedtest . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 415
vstat . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 416
vvimms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 419
vzero . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 419
perf_test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 420
433
434
435
436
Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 438
viii
467
467
468
469
469
535-0008-00 Rev 10
472
473
474
474
475
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 477
535-0008-00 Rev 10
ix
535-0008-00 Rev 10
Preface
Intended Audience
This guide is intended for experienced systems administrators. Violin Memory
assumes that you are experienced in installing and servicing high-performance
storage systems.
Contact Violin Memory Customer Support for any assistance with installing and
servicing this system. See Contacting Violin Memory on page 5 for contact
information.
535-0008-00 Rev 10
Document Organization
This guide is organized into the following sections:
Chapter 1, Violin 6000 Series Memory Array OverviewIntroduces the Violin 6000 Series
Memory Array and its features, architecture, and specifications.
Chapter 2, Manage, Monitor, and Administer with the Violin Web InterfaceDescribes Web
interface tasks specific to the Memory Gateways that are internal to the Violin 6000 Series
Memory Array.
Chapter 3, Manage the Memory Array with the Web InterfaceDescribes using the Violin
Web interface to complete the basic Memory Array monitoring and administrative tasks and
the tasks specific to the Memory Gateways that are internal to the Violin 6000 Series Memory
Array.
Chapter 4, vSHARE ConfigurationDescribes vSHARE, its features and functions, and the
steps required to configure a Memory Gateway running vSHARE to manage block storage.
Chapter 6, Manage the Memory Array with the CLIProvides detailed procedures for
configuring and showing status of the Violin 6000 Series Memory Array and its internal
modules using the Command Line Interface (CLI).
Appendix A, CLI ReferenceDescribes the Violin 6000 Series Memory Array command line
interface (CLI) and how to use it; describes the commands available in the Violin 6000 Series
Memory Array.
Appendix B, Alarm ReferenceDescribes the various system alarms that may occur when
using the Violin 6000 Series Memory Array.
Appendix C, Violin DriversDescribes how to install, use and remove the Linux drivers on a
host computer directly attached to the Violin 6000 Series Memory Array.
Appendix D, Violin UtilitiesDescribes the Violin 6000 Series Memory Array utility commands
and how to use them.
Appendix F, SNMP TrapsDescribes using SNMP Traps with the Violin 6000 Series Memory
Array.
Appendix G, Optimizing Connectivity for Initiator HostsDemonstrates ways in which you can
optimize the connectivity of initiator (client-side) hosts with Violin Memory Arrays.
535-0008-00 Rev 10
Reference Documents
In addition to this guide, the following Violin Memory documents comprise the documentation suite
that will assist you with setting up, using and servicing the Violin 6000 Series Memory Array. These
guides are available for download from the Violin Memory Support site at http://www.violinmemory.com/support/
This document...
Release Notes
Reference Documents
Document Conventions
Safety Icons
The table below summarizes warning, caution, and note icons used in this document and includes
sample text.
Safety Icons
Icon
WARNING!
535-0008-00 Rev 10
Sample Text
Sample Text
Caution:
Note:
Typographical Conventions
The following typographic conventions are used in this guide:
Format
Meaning
Bold
Italic
Courier
Courier bold
<Courier italic>
[ ]
{ }
Typographical Conventions
Security
Violin Memory, Inc., cannot be responsible for unauthorized use of equipment and will not make
allowance or credit for unauthorized use or access.
535-0008-00 Rev 10
Model and serial number of the system for which you are requesting support.
Software version.
A brief description of the problem.
535-0008-00 Rev 10
535-0008-00 Rev 10
CHAPTER 1
This chapter introduces the Violin 6000 Series Memory Array in the following
sections:
535-0008-00 Rev 10
Data Management
vMOS Data Management is introduced on Violin 6000 Series flash Memory Arrays with the
release of vMOS-6. vMOS Data Management allows you to manage and protect critical data,
as well as simplify backup and disaster recovery using the following features:
For information on Web interface procedures, see Manage LUN Groups on page 76. For indepth information on CLI procedures and commands, see CHAPTER 8, Data Management
on page 253.
LUN groups
This release introduces the ability to group LUNs for simplified administration and
management. It also provides the ability to monitor a group of LUNs as a combined entity. For
information on Web interface procedures, see Manage LUN Groups on page 76.For in-depth
information on CLI procedures and commands, see Working with LUN Groups on page 197.
SCSI-3 ALUA
This release introduces an enhancement for Asymmetric Active-Active ALUA for enhanced
Data Management performance. For more information, see Data Management on page 253.
535-0008-00 Rev 10
SAN-attached Storage (vSHARE): connects to a storage area network (SAN) via the
integrated or external vSHARE Memory Gateways using Fibre Channel, InfiniBand or iSCSI/
Ethernet. This configuration allows multiple hosts, including clustered applications, to share
the same storage.
Direct-attached Storage: connects via PCI Express (PCIe) to one or two host computers
running Linux and Windows. This configuration provides low latency and high bandwidth.
The Violin 6000 is ideal for any application that requires active processing or rapid access to large
amounts of data, including applications for transaction processing (Oracle, DB2, SQL Server,
Sybase), analytics and data warehousing, messaging systems (e-mail), active file storage and
metadata (GPFS, Lustre).
535-0008-00 Rev 10
The Memory Array includes four vRAID Controllers. These act as the roots of the VXM networks
and can be configured for complete redundancy.Three vRAID controllers out of four can fail without
loss of connectivity to all of the VIMMs. These vRAID controllers can also be hot-swapped as
required.
vRAID
The Memory Array employs a vRAID technology to optimize key flash storage attributes including
latency under load, bandwidth, storage efficiency and reliability.
vRAID provides full RAID data protection across the array while simultaneously guaranteeing that
any read will not be blocked by an erase providing for spike-free latency under load and sustained
performance.
Unlike existing RAID 5 and RAID 6 solutions, which were designed for hard disk drives and rely on
Read-Modify-Write operations that are unsuited to flash memory, vRAID was specifically designed
to enhance flash storage systems. It uses an efficient 4+1 Parity model and writes any 4kB block
without a requirement for Read-Modify-Write.
PCIe Switching
The core of the Memory Array includes redundant PCIe switches that connect the vRAID controllers
to the external PCIe ports or the integrated Memory Gateways. PCIe switches are high bandwidth
and low latency and are an ideal complement to the VXM technology used for flash aggregation.
Unlike the PCIe used in standard servers, the PCIe switches are implemented to support hot-swap
of all modules for full redundancy. Each vRAID Controller is dual-homed to two different PCIe
switches.
Product Models
This Violin 6000 Memory Array is available in thee following models.
Model #
Configuration
6606
24 SLC VIMMs
6212
24 MLC VIMMs
6611
44 SLC VIMMs
6222
44 MLC VIMMs
6616
64 SLC VIMMs
6232
64 MLC VIMMs
10
535-0008-00 Rev 10
Chassis
The Memory Array is a 3U high chassis and mounts on adjustable sliding rails into equipment racks
from 19" to 36" deep.
The chassis itself mounts on slide-out rails. The lid can be removed without taking the system out
of service.
Caution: To avoid overheating and the generation of system alarms, the lid must not be off of
the system for more than 5 minutes when the ambient temperature is 35 C (95 F) and above.
The front panel LEDs and the Violin Web interface can indicate which units are faulty and
require replacement
535-0008-00 Rev 10
11
Top View
Memory Array components are hot-swappable from the top of the chassis.
Network
Interfaces
Power
Controllers
B
VIMMs
Internal
Memory
Gateways
B
A
Array
Controllers
D
vRAID
Controllers
A1
B1
C1
Fans
A0
B0
C0
12
Fans
Six fans are arranged in two rows of three each. Each fan may be
replaced independently of others.
Array Controllers
Internal Memory
Gateways
Power Controllers
535-0008-00 Rev 10
Network Interfaces
VIMMs
vRAID Controllers
The Memory Array has four vRAID controllers for redundancy and
additional bandwidth.
Front View
The front panel contains two USB ports (A and B), the ID button, and six LEDs (AC-A, AC-B, System
Status, Port/Interface Status, VIMM Status, ID).
USB Port A
USB Port B
AC-A LED
AC-B LED
System Status
ID LED/Button
Port/Interface Status
VIMM Status
The USB ports are used for performing software upgrades. See Upgrading the Memory Array
Firmware on page 236. For more information on LEDs, including the ID LED/button, see System
LEDs on page 15.
535-0008-00 Rev 10
13
Rear View
Network Interface
Cards and Ports
A
A
B
D
A
B
n
D
n
C
n
B
n
A
A
B
ID
A
B
ID LED
Power Supply A
Power Supply B
Ground
Ethernet Ports
Serial Console
Ports
Power Supplies
14
Each power supply has two LEDs. The bottom LED is reserved
for alarms. The power supply must be replaced if the bottom
LED is solid amber. Blinking amber indicates overcurrent. Power
supplies are also hot-serviceable. For more information on
power supplies, see Power Supplies on page 16.
535-0008-00 Rev 10
The PCIe and SAN/LAN connections may be removed and reinstalled without rebooting the system.
Each interface port on the rear of the system has an associated LED indicating if it is in use or not.
System LEDs
The Violin 6000 is designed for highly reliable data center operation featuring hot-swappable
modules and no single point of failure for high availability. The VIMMs, controllers, and fans are hotserviceable from the top. The System Status and VIMM Status LEDs alert you of system and VIMMspecific alarms respectively. Service LEDs indicate which modules are faulty and enable you to
replace the modules with no downtime.
Six system LEDs are located on the front of the Memory Array.
Front LEDs
Six LEDs (AC-A, AC-B, System Status, VIMM Status, and Interface Status) are displayed on the
front of the chassis.
AC-A
AC-B
Interface Status
ID LED
Button
System Status
VIMM Status
LED
LED State
Meaning
AC-A (USB0)
Green (solid)
AC-B (USB1)
Green (solid)
System Status
Green (solid)
Amber (solid)
Grey (solid)
535-0008-00 Rev 10
15
LED
LED State
VIMM Status
Meaning
Green (solid)
Amber (solid)
ID
Blue (solid)
Active system alarms may be viewed by running the show alarms command on the Memory
Array. See Alarm Reference on page 369 for more information.
Power Supplies
The chassis contains two power supplies that provide an AC power source for the Memory Array.
Power Supply B
Power Supply A
The 90% efficient power supplies convert the incoming AC power from individual AC feeds to DC
power used by the system. These 2000-watt c20 receptacles are redundant in their configuration.
The power supplies are hot-swappable from the rear of the chassis.
16
535-0008-00 Rev 10
Fans
Six fans are located in the front of the chassis and provide forced airflow ventilation for the Memory
Array. When the system is first powered on, the fans come up at high speed. Built-in temperature
sensors automatically lower the fan speed if lower temperature conditions are detected.
The fans are arranged in two rows of three each. Every fan may be replaced independently of one
another. Multiple fans may fail and the system will remain operational. See Figure 1.2, Memory
Array: Top View, on page 12 for fan layout and numbering. Multiple temperature sensors within the
system provide indication, alarm and automated system control functions.
The fans switch to full speed in the event of overheating conditions, high ambient temperature and
other fault conditions, including failure of a VIMM, vRAID controller module or the fans themselves.
There is roughly a 20-25C temperature rise inside the system when air is flowing freely in and out of the
system. Removal of a fan appears like a fault condition, causing all fans to switch to full speed until
the removed fan is replaced.
Caution: The Memory Array is designed to tolerate a single fan failure and work with inlet
temperatures between 0C and 35C. Unobstructed airflow is expected. Any obstruction to airflow causes
unnecessary temperature rise of components inside the system. The fans operate in high-speed mode when
the inlet temperature rises above 30C.
The fans plug into connectors inside the front of the chassis. For more information, see Replacing
a Fan in the Violin 6000 Series Memory Array Service Guide.
535-0008-00 Rev 10
17
The vRAID Controllers, Array Controllers, Power Controllers, internal Memory Gateways, and
network interface cards can all be provisioned redundantly. Any single failure is automatically
handled until the faulty unit is replaced.
Memory
Gateway
Fan
VIMM
Array
Controller
vRAID
Controller
Power
Controller
Network
Interface
Power
Supply
All active components are hot-swappable and can be replaced without powering down the system.
Memory Plane
The base system has a set of boards that act as the memory plane and provide interconnection
between all modules and controllers. The memory plane is passive to ensure an extremely high
mean time between failures.
Power Controllers
The redundant Power Controllers provide power to all controllers, modules and fans. Each slot or
socket can be individually powered up or down to simplify remote control. The Power Controllers
are designed so that no single failure will impact the ability of the system to provide storage
services.
18
535-0008-00 Rev 10
Array Controllers
The Array Controllers provide PCI Express (PCIe) interfaces, switching and management services.
These include serial management of the system including power supplies, fans and memory
infrastructure. Command line, Web, e-mail, Syslog and SNMP interfaces are provided.
ACM Connection for vMOS-6
vMOS-6 requires that the ACMs be externally connected to each other to interconnect the internal
Memory Gateways. This cable connection enables a Non-Transparent Bridge (NTB), which allows
for high-bandwidth, low latency inter-cluster coordination between the two Memory Gateways for
effective management of thin provisioned LUNs and snapshots. For more information, see the Violin
6000 Series Memory Array Installation Guide for Release V6.0.0.
Memory Gateways
The internal Memory Gateways are integral components of the Memory Array, as they manage
connectivity to SAN systems. Block storage (vSHARE) capabilities are provided.
Violin Memory Gateways with vSHARE provide value-added functionality such as flexible Logical
Unit Number (LUN) provisioning, LUN masking and reservations, and full-speed LUNs. Each LUN
can use the full IOPS/bandwidth of the Violin 6000 platform if the other LUNs are idle.
Network Interfaces
The network interfaces provide the physical connectivity to the SAN or LAN. Eight 8Gbit Fibre
Channel, InfiniBand or eight 10GbE ports are available via four separate Network Interfaces.
vRAID Controllers
A vRAID Controller Module (VCM) manages the Violin Switched Memory Network (VXM) of VIMMs
and implements the vRAID algorithms.
vRAID Controllers enable the VIMMs to be hot swapped and automatically rebuild RAID groups on
the failure of a single VIMM.
The Memory Array has four vRAID Controllers for redundancy and additional bandwidth.
535-0008-00 Rev 10
19
Data is written to VIMMs using the logical block address. The flash controller on the VIMM assigns
the logical block to a physical page within the flash of the VIMM. Metadata is used to map between
logical addresses and physical addresses. Each VIMM includes:
Violin Memory Arrays achieve high sustained performance by distributing a complex task called
grooming to each VIMM and performing this task in hardware. Grooming is the process by which
flash with stale data is recovered and the flash memory is prepared to be used for new data. This
process involves reading a flash address space (block), writing the still-current data to a new flash
address space, and then erasing the previous flash address space. Erasing, reading, and writing is
a time- and resource-consuming task, which the VIMM is optimized to complete efficiently.
20
535-0008-00 Rev 10
VIMMs are designed to enable the scalability of large arrays of flash memory. The advantages of
this architecture are:
Unlike most solid-state drives and PCIe cards, a failed flash device does not cause a VIMM to lose
data or be taken out of service. ECC and RAID protection are used to protect the data from errors.
Data is read from other VIMMs, rebuilt using the RAID algorithm, and written back to the failed
VIMM. This VIMM writes the same data to a different flash device, avoiding the need to be replaced.
The Violin 6000 also allows the VIMM to be replaced without powering the system down or
disrupting access to user data. Each VIMM can be independently powered down and powered up.
A fully populated system contains 48 data VIMMs, 12 parity VIMMs and four spare VIMMs for failure
protection. All memory modules are hot-swappable with automated rerouting that will not interrupt
data service.
The normal operating temperature range of VIMMs is 40-69C. As air flows into the Memory Array,
the first row of VIMMs preheats air passing over the second row of VIMMs, which then preheats air
flowing over the third row of VIMMs. This results in the third row of VIMMs usually operating hotter
than the first two rows. VIMMs in this third row typically record higher temperatures.
In the case of a module failure where the module must be replaced, the Memory Array does not
have to be powered down. See the Violin 6000 Series Memory Array Service Guide for VIMM
replacement instructions.
VIMM Types
The Memory Array performance depends on the type of memory used and the number of modules.
VIMMs are available in the memory types listed in the following table.
VIMM Type
VIMM Capacity
SLC
256 GiB
MLC
512 GiB
Flash-Based VIMMs
Flash systems store up to 32 TiB of raw flash storage per system. The amount of usable memory
in a system is the number of VIMMs used for data storage, which is four per RAID group (up to 64
VIMMs), multiplied by the amount of memory per module, multiplied by the usable percentage for
which the system is formatted. For example, in a fully loaded system with 64 VIMMs, the usable
memory is 48 VIMMs multiplied by 512 GiB VIMMs, multiplied by the usable percentage, such as
84%, which is 20.7 TiB.
535-0008-00 Rev 10
21
VIMM Configurations
The Memory Array contains up to 64 flash-based VIMMs, which operate in the VIMM configurations
listed in the following table.
VIMMs
Active
Spares
24
20
44
40
64
60
Any outstanding writes are committed to flash on the VIMMs, with no loss of data. Once the failed
VCM is replaced and fully booted, its original VIMMs return (failback) to the replacement VCM.
If a RAID rebuild is in progress when a VCM fails, the rebuild process stops and then restarts once
the VCM failover process completes.
The System Status LED on the front of the Violin 6000 is amber if a VCM or other system
component is faulty. See the Violin 6000 Series Memory Array Service Guide for replacement
procedures.
Utilizing Floating Spares
Each Violin 6000 Memory Array model has four spare VIMMs. Two spares are available and ready
in the event of a VIMM failure within another RAID group. The spare VIMMs available to VCMs A
and B reside in slots 3 and 6; the spare VIMMs available to VCMs C and D reside in slots 13 and
16. See VIMM Slot Locations on page 23 for more information, along with a top-level view of where
the spares reside in each configuration.
Each VCM can support up to two VIMM failures in a single RAID group without loss protection from
a third failure.
22
535-0008-00 Rev 10
The VIMM Status LED on the front of the Memory Array is amber when an alarm has been raised
on a VIMM. Failed VIMMs should be replaced with new spares as soon as possible. See the Violin
6000 Series Memory Array Service Guide for replacement procedures.
Number
of VIMMs
Slot Numbers
24
0, 1, 3, 6, 8, 9, 10, 11, 13, 16, 18, 19, 43, 44, 45, 46,
47, 48, 55, 56, 57, 58, 59, 60
44
64
Spare Locations
Slots 3, 6, 13, 16
All Violin 6000 Series Memory Array models contain four spare VIMMs.
535-0008-00 Rev 10
23
24
535-0008-00 Rev 10
535-0008-00 Rev 10
25
26
535-0008-00 Rev 10
Item
Size / Quantity
VIMM Types
Number of VIMMs
24, 44 or 64
1, 2 or 4
Network Interfaces
See the Violin 6000 Series Memory Array Installation Guide for diagrams of the configurations
supported by Memory Arrays, along with external and internal Memory Gateways.
Management Interfaces
The Memory Array is accessible through physical port interfaces on the rear of the chassis and can
be accessed remotely using a number of other interfaces, all briefly described in the following table.
Interface
Usage
SNMP
Syslog
535-0008-00 Rev 10
27
The CLI provides a full set of commands and information for configuring and operating the platform.
It uses a Cisco-style CLI interface with command completion and integrated help.
For more information, see Manage the Memory Array with the CLI on page 221 and CLI Reference
on page 287.
The most common browsers (Firefox, Internet Explorer, Safari, Chrome) are supported. For more
information, see Supported Web Browsers on page 33.
E-mail Interface
The e-mail interface allows reports and alerts to be sent through the local e-mail server to both local
administrators and Violin Memory Customer Support (or its partners). These reports can be sent
daily or when specific events occur that require attention.
For more information, see Notification on page 168 or Event Notification on page 302.
Syslog Interface
Logging of events via Syslog is supported. This includes all CLI and configuration events as well as
any internal warnings or alarms.
Using an external Syslog server, such as Splunk, allows syslog events from many Violin Memory
Arrays and hosts to be consolidated and correlated.
See Event Logging Configuration and Viewing on page 294 for more information.
28
535-0008-00 Rev 10
Management Interfaces
SNMP Traps
Simple Network Management Protocol (SNMP) enabled management systems can use SNMP to
track the status of the Memory Array. Both SNMP Traps and SNMP polling can be performed via a
generic SNMPv2 interface.
The Memory Array Management Information Base (MIB) is available for integration into 3rd party
management systems such as HP Network Node Manager and IBM Tivoli.
For more information, see SNMP Traps on page 433.
535-0008-00 Rev 10
29
30
535-0008-00 Rev 10
CHAPTER 2
This chapter introduces the Violin Web Interface and covers the following topics:
535-0008-00 Rev 10
31
Chapter 2 - Manage, Monitor, and Administer with the Violin Web Interface
Main Menu BarDisplays clickable icons for the major functional categories: Home, Manage,
Monitor, and Administration.
Function Menu BarChanges to display the options for the currently active category. Click an
icon to display the associated page and submenu options.
Software VersionDisplays the version of the software currently running on the Memory
Gateways.
SessionShows the user name of the current user at the right, and exits the session when
clicked.
Board Status TabExpands to show board status information when clicked. Click the tab
again to hide the board status page.
Copyright
32
Online Help
Session
Software Version
535-0008-00 Rev 10
Linux
Supported Browsers
Mac
Windows
Note: For Windows Internet Explorer, Compatibility Mode should be turned OFF.
The following table lists the recommended software and settings for optimum performance of the
Violin Web Interface.
Recommendation
Description
Display resolution
JavaScript
Cookies
535-0008-00 Rev 10
33
Chapter 2 - Manage, Monitor, and Administer with the Violin Web Interface
Meaning
bytes
KB
kilobytes
MB
megabytes
GB
gigabytes
TB
terabytes
PB
petabytes
For E (exabytes), Z (zettabytes), and Y (yottabytes), a single letter omitting the B is used for units
other than bytes, or to save space in a display.
Context-sensitive online help for any page in the Web InterfaceClick the Page Help icon in
the upper right corner of the Web Interface window.
A full PDF version of this guideSelect Administration > Information > Documentation and
click the PDF link.
Description
Admin
The Admin user can access the Web Interface pages for
managing, monitoring, and administrating.
Monitor
34
535-0008-00 Rev 10
http://<master_gateway_ip_address>
http://<master_gateway_hostname>
http://<cluster_node_ip_address>
http://<cluster_node_hostname>
Note: If you specify the IP address or hostname of a cluster node (other than the master),
the connection is automatically redirected to the master Memory Gateway.
following example.
535-0008-00 Rev 10
35
Chapter 2 - Manage, Monitor, and Administer with the Violin Web Interface
The Logout page appears acknowledging that you successfully logged out.
3. To log back in, click Login Again and re-enter your credentials.
36
535-0008-00 Rev 10
http://<master_gateway_ip_address>
http://<master_gateway_hostname>
http://<cluster_node_ip_address>
http://<cluster_node_hostname>
Note: If you specify the IP address or hostname of a cluster node (other than the master),
the connection is automatically redirected to the Master Gateway.
following example.
535-0008-00 Rev 10
37
Chapter 2 - Manage, Monitor, and Administer with the Violin Web Interface
3. Choose
4. Enter a title for the shortcut icon, and click Add. A shortcut icon appears on your home screen.
5. Close the Web Interface session, then click the home screen icon to launch the optimized
version.
6. Continue with Customizing the Dashboard on page 44.
38
535-0008-00 Rev 10
Home on page 39
Manage on page 40
Monitor on page 41
Administration on page 42
Home
Function
Dashboard
System View
LUN Status
Front and Rear Panel
Tasks
535-0008-00 Rev 10
39
Chapter 2 - Manage, Monitor, and Administer with the Violin Web Interface
Manage
vMOS-6 provides the following Manage features and functions:
Manage LUNs, Initiators, Targets, and LUN Groups available on v5.x and v6.x platforms.
vMOS Data Management available only on v6.x platforms.
Manage LUNs
Manage Initiators
Manage Targets
Manage LUN Groups
Tasks
40
535-0008-00 Rev 10
Manage Snapshots
Space Management
Tasks
Monitor
Function
Monitor Overview
Monitor LUNs
Monitor LUN Groups
Monitor Targets
Tasks
535-0008-00 Rev 10
41
Chapter 2 - Manage, Monitor, and Administer with the Violin Web Interface
Administration
Function
Cluster Admin
Tasks
Tools
Information
Network
Notification
42
535-0008-00 Rev 10
Home Dashboard
Home Dashboard
The Home dashboard is a landing page that provides a selection of metrics for tracking system
health and performance. The customizable dashboard allows you to easily add, remove, and
rearrange the gadgets for at-a-glance system assessment. Likewise, the System View page
provides a visual overview of Violin Cluster Status for health and performance. The LUN Status
page provides an easy access list of all LUNs and their statistics, while the Front and Rear Panel
page
Note: The dashboard supports Internet Explorer 9 (IE 9) and higher. If you are using an earlier
version of Internet Explorer, the dashboard does not display. For more information, see Supported
Web Browsers on page 33.
The following table provides a list of tasks and their associated functions. Click a hyperlink to jump
directly to the task.
Function
Dashboard
System View
LUN Status
Front and Rear Panel
Tasks
Dashboard
You can select the gadgets that appear on the landing page and arrange them to maximize
efficiency for your style of work. You can populate your dashboard with any of the following gadgets.
BandwidthShows reads and writes performed in Megabytes per second (MB/s). You can
display this information at the container level or LUN level. Click the settings icon in the upper
right corner to customize the display.
IOPSShows read, write, and total statistics in real-time, per hour, or per day values. Click
the settings icon in the upper right corner to customize the display. You can display this
information at the container level or LUN level.
535-0008-00 Rev 10
43
Chapter 2 - Manage, Monitor, and Administer with the Violin Web Interface
LatencyShows read, write, and total latency for a container or LUN in real-time, per hour, or
per day values. Click the settings icon in the upper right corner to customize the display.
Live StatsShows real-time values for bandwidth, IOPS and latency on a LUN level or
container level. Click the settings icon in the upper right corner to customize the display.
IndicatorsShows the LED, temperature and fan status for a container on an array. Click the
settings icon in the upper right corner to customize the display.
Flash/VIMM StatusShows the VIMM status (number active, down, etc.) and the remaining
length of flash life for a container on an array. Click the settings icon in the upper right corner to
customize the display.
Container StatusShows the amount of space used and available on the container, the
number of LUNs, and provides an HA status indicator for a container on an array. Click the
settings icon in the upper right corner to customize the display.
Port StatsShows the Received and Sent statistics for InfiniBand and Fibre Channel ports in
real-time, per hour, or per day values for a port within a container. Click the settings icon in the
upper right corner to customize the display.
Snapshots SummaryShows the number of snapshots, the space used, and the percentage
of total space consumed by snapshots. A list of the days schedules for snapshots is shown.
Click the arrow (>) on the right to jump to the Snapshot Management page and view detailed
information on the selected schedule. This gadget is only available with vMOS-6.
Note: For instructions on how to change the information shown on a gadget, see Customizing
44
535-0008-00 Rev 10
Home Dashboard
The toolbar appears at the bottom of the window, as shown in the following example.
3. Click an icon on the toolbar to add it to the dashboard, then select and drag icons to rearrange.
Click the X icon in the upper right corner of the gadget to remove it from the dashboard
display.
535-0008-00 Rev 10
45
Chapter 2 - Manage, Monitor, and Administer with the Violin Web Interface
4. To display detailed data for a gadget, move your cursor over an area in the graph. A pop-up
5. To hide the toolbar, click the X icon in the lower right corner of the window.
46
535-0008-00 Rev 10
Home Dashboard
Customizing Gadgets
Each dashboard gadgets allows you to customize the display.
To customize the information shown on a gadget, do the following:
1. Open the gadget on the dashboard, as described in Customizing the Dashboard on page 44.
2. Click the Settings icon in the upper right corner of the gadget, as shown circled in red in the
following example.
The gadget flips over to show a back panel with options appropriate for that gadget.
3. Make the desired selections and click Save to apply, or Cancel to revert to the previous
535-0008-00 Rev 10
47
Chapter 2 - Manage, Monitor, and Administer with the Violin Web Interface
System View
The System View tab brings up the Violin Cluster Status page, displaying information depicting the
health and performance of both Memory Gateways. Easy to read graphics allow you to quickly
assess if there is a problem with the system. An expandable Board Status page provides an at-aglance view of VIMM statistics.
Note: A status message appears in the header when a RAID rebuild is in progress
48
535-0008-00 Rev 10
Home Dashboard
Home > System View. The Violin Cluster Status page appears.
yellow. A green box with a white check mark represents a healthy, fully functioning fan.
5. Check the VIMM, Controller, and ambient temperatures. Green is healthy, yellow a caution,
Home > System View, to display the Violin Cluster Status page.
2. Click the
535-0008-00 Rev 10
49
Chapter 2 - Manage, Monitor, and Administer with the Violin Web Interface
3. For general status information about specific VIMMs, move the cursor over the VIMM number.
4. To view Alarm, Temp, or Lifetime information, move the cursor over the appropriate indicator
box.
Home > System View to display the Violin Cluster Status page.
2. Click the
3. To view VIMMs by RAID group, select a RAID Group number at the top-center of the Board
Status panel.
Selecting a RAID Group highlights the VIMMs belonging to that RAID Group in the Board Status
panel, as shown in the following example.
50
535-0008-00 Rev 10
Home Dashboard
6. To view admin down VIMMs, select the Admin Down link in the Show area of the Board Status
panel.
7. To view spare VIMMs, select the
LUN Status
The LUN Status page allows you to view information about the LUNs within a container. Every LUN
is created within a storage container (array).
2. Select the
535-0008-00 Rev 10
51
Chapter 2 - Manage, Monitor, and Administer with the Violin Web Interface
3. Select the LUN icon to view real-time statistics about the LUN.
4. Click the container name to view detailed information about the container and the LUNs it
contains. The LUN Management page appears, with the container information in the upper left
panel.
Note: For more information on managing LUNs, see "Manage LUNs" on page 59.
2. Select a
52
535-0008-00 Rev 10
Home Dashboard
3. Select the LUN icon next to the name to view real-time statistics about that LUN.
4. Click the LUN name link to view detailed information about that LUN.
The LUN Management page appears, with the selected LUN highlighted in the LUN panel on
the left. In the top right panel, details of the selected LUN are shown.
Note: For more information on managing LUNs, see Manage LUNs on page 59.
535-0008-00 Rev 10
53
Chapter 2 - Manage, Monitor, and Administer with the Violin Web Interface
The Front and Rear Panel page displays, with an image of the back of the Memory Array shown
at the top.
2. Rest the mouse pointer over a port (power supply, Ethernet connector, Network Interface port)
going through the Memory Gateways (NIC ports). Since the Web interface is run from the Memory
Gateways, PCIe port information is unavailable in the Web interface.
54
535-0008-00 Rev 10
Home Dashboard
The active port area selected is displayed in the lower left panel of the page. The following
example shows the lower left and right panels for an InfiniBand configuration.
4. If specific port information is available, click a port in the lower left panel to view the
Power Supplies A and B: The power supplys status, and current voltage and amps are
shown.
Ethernet Ports A and B: View the Memory Gateway Management IP address; host names
and IP addresses of the Master and Standby Gateways; the ACM Management IP address;
host names and IP addresses of the Master and Standby ACMs.
Network Interface Ports: View the Memory Gateway that manages the port (mg-a or mg-b)
and the ports state and speed. Also available is an Port Stats gadget, which shows the
Received and Sent statistics for InfiniBand and Fibre Channel ports in real-time, per hour, or
per day values for a port within a container.
If you select the active Network Interface Ports area on the Rear Panel image, the lower left
panel that displays shows the slot and port numbers for the cards installed in the system,
along with the representative CLI values for each port.
Note: The physical port numbers on the network interface cards (NICs) are not aligned with their
corresponding port names in the CLI. For example, in a Fibre Channel environment, NIC port A1 is
shown as hba-a1 in the CLI, and NIC port C1 is also shown as hba-a1 in the CLI. This is because
the two Memory Gateways, shown with their CLI representations in the lower left panel (mg-a and
mg-b), separately manage two NICs each. To verify the actual port numbering shown in the CLI, log
in to the Memory Gateway master and then run the show targets command.
To switch the view to the front panel, Click the Front Panel button in the upper right of the Rear
Panel View.
535-0008-00 Rev 10
55
Chapter 2 - Manage, Monitor, and Administer with the Violin Web Interface
The Front and Rear Panel page displays, with an image of the back of the Memory Array shown
at the top.
2. Click the Front Panel button in the upper right of the Rear Panel View image to switch the
3. Rest the mouse pointer over a USB port or LED to view the name of the port or LED. The
USB ports: The USB ports are used for performing software upgrades.
AC-A and AC-B LEDs: The Master ACM (AC-A or AC-B) is shown in green. Use its USB
port when upgrading the software from a USB drive.
SYS (System Status) LED: This LED is green when all system modules are operating
normally, amber when an alarm has been raised, and grey when the Memory Array is off
or starting up.
VIMM LED: This LED is green when all VIMMs are functioning correctly, and amber when
an alarm has been raised for at least one VIMM, or when a VIMM is not powered or
present.
ID LED/Button: This LED is blue when the button has been pressed or when it has been
activated remotely using the Violin Web interface. See Remotely Activating the ID LED of
a Memory Array on page 56.
To switch the view to the rear panel, Click the Rear Panel button in the upper right of the Front Panel
View.
Remotely Activating the ID LED of a Memory Array
The ID LED can be remotely activated from both the front- and rear-panel views. The ID LED is used
to locate a particular Memory Array in an equipment rack, which is helpful when a Memory Array
needs attention and the equipment rack holds more than one Array.
56
535-0008-00 Rev 10
Home Dashboard
To activate the ID LED from the Rear Panel View, do the following:
1. Select
Home > Front and Rear Panel. The Front and Rear Panel page appears.
2. Click the lower left of the Rear Panel image to remotely activate the ID LED of this Memory
Array.
The LED is activated when the LED is blue on the rear panel image.
3. Click the blue LED to turn off the ID LED.
To activate the ID LED from the Front Panel View, do the following:
1. Select
Home > Front and Rear Panel. The Front and Rear Panel page appears.
2. Click the Front Panel button in the upper right of the Rear Panel View image to switch the
The LED is activated when the LED is blue on the front panel image.
4. Click the blue ID button to turn off the ID LED.
535-0008-00 Rev 10
57
Chapter 2 - Manage, Monitor, and Administer with the Violin Web Interface
Manage
You can manage LUNs, initiators, targets, and LUN groups, as well as utilizing data management
functionality. The following tables provide a list of tasks and their associated functions. Click a
hyperlink to jump directly to the task.
vMOS-6 provides the following Manage features and functions:
Manage LUNs, Initiators, Targets, and LUN Groups available on v5.x and v6.x platforms.
vMOS Data Management available only on v6.x platforms.
Manage LUNs
Manage Initiators
Manage Targets
Manage LUN Groups
Tasks
58
535-0008-00 Rev 10
Manage
Manage Snapshots
Space Management
Tasks
Manage LUNs
The following panels comprise the LUN Management Page:
ContainerDisplays the container name, the percentage of space used, the number of LUNs
in the container, Gateway A and B names, HA status, active online ports, and a bar graph
depicting container space usage.
LUNsShows information about the LUNs within the selected container, including name,
serial number (exported LUNs only), provisioned size, allocated size, and total number of
snapshots for the LUN, if any. You can create LUNs, search for a LUN, reorder the list, and
perform actions on LUNs by selecting one or more check boxes. When you highlight a LUN
(selecting it so that it changes color), detailed information about the LUN appears in a panel on
the upper right.
Note: Only one LUN can be highlighted at a time.
Highlighted LUNShows detailed information on the highlighted LUN (in the LUNs panel),
including if it is read-only, and/or has Port A or Port B offline. You can view and add or remove
Exports for a selected LUN.
535-0008-00 Rev 10
59
Chapter 2 - Manage, Monitor, and Administer with the Violin Web Interface
Caution: You can change the online status of Port A and Port B (Gateway A and Gateway B) in
the highlighted LUN information panel. However, this is generally not recommended.
Note: You can select multiple LUNs and perform an operation on them simultaneously. A check
mark in the box next to the LUN name shows that a LUN is selected.
2. To search for a LUN, enter a string in the Search field. A list of LUNs that contain the string
the following example), serial number (for exported LUNs), provisioned size, allocated size, or
number of snapshots. Click the up arrow in any of the columns to view the list in ascending
order (the default display).
60
535-0008-00 Rev 10
Manage
5. To change the number of entries shown, choose one of the following from the Show entries
2. In the LUNs panel, select the check boxes next to the desired LUNs. A check mark appears in
In the following example, Export Checked LUNs is selected. For more information on how to
export LUNs, see Exporting LUNs, Initiator Groups, and Initiators on page 67.
535-0008-00 Rev 10
61
Chapter 2 - Manage, Monitor, and Administer with the Violin Web Interface
Verify the block size that is acceptable for the clients (initiators) on your network.
Disk storage systems previously assumed a 512-byte sector or block size as the unit to store data.
Newer devices, such as Violin Memory Arrays, have a native storage unit sized at 4096 bytes but
can also emulate 512 bytes. Not all client operating systems are able to understand accessing
anything other than a 512-byte based disk drive.
To be safe, choose a 512 block size if you are not sure what type of clients will be connected to an
exported LUN. For example, RHEL5 and Windows Server 2003 are not be able to properly use a
4096 block-sized disk, and this selection leads to unexpected behavior. For newer operating
systems which support a 4096 block size, the emulation layer can be avoided to provide optimal
performance.
Note: VMware ESX 4.1 does not support 4K block size. For VMware ESX 4.1, use a 512 byte
Note: When creating more than one LUN, the name of each new LUN is appended with a number
beginning with 1, and continuing through the number of specified LUNs. This ensures that each
LUN has a unique name. Creating a large number of LUNs at once (50 LUNs or more), the
operation may take some time.
62
535-0008-00 Rev 10
Manage
To define the size of the LUN based on the space available, select Use all available
space equally.
To manually define the size of each LUN, select Specific size per LUN and enter the size
in GB.
Note: The size of the LUNs may be restricted by the space available in the container.
5. Select one of the following block sizes for the LUN:
512 bytes.
4096 bytes.
Caution: Not all systems can handle 4096-byte blocks. Use 512-byte blocks, if you are not sure what type
of clients will be connected to an exported LUN.
6. Specify the following settings, as appropriate:
Thin ProvisioningCreates a thin LUN. For more information, see Thin Provisioning on
page 260.
7. Click
2. Search for the LUN, as described in Searching, Sorting, and Performing LUN Operations on
page 60.
535-0008-00 Rev 10
63
Chapter 2 - Manage, Monitor, and Administer with the Violin Web Interface
3. Select the LUN so that it highlights, then remove the LUN exports in the following way:
a. In the Exports panel on the right, click the check boxes for each of the exports associated
Unexport icon. Then click OK in the pop-up dialog that appears to confirm the
action.
Note: All exports must be removed before the LUN can be deleted.
4. Click the check box of the LUN you want to delete. A check mark appears.
Note: The check box must display a check mark for the LUN to be deleted. Simply
highlighting the selection does not include it in the operation.
5. In the LUNs panel, click the Delete icon. Then click
Editing LUNs
You can change the name of an existing LUN and increase the size of the logical volume.
Decreasing the size of an existing LUN is not supported.
For a smaller size LUN, create a new LUN of the desired size. For more information, see Creating
and Deleting LUNs on page 62.
WARNING! Existing file systems on the client side are not extended as part of the resize
operation. File systems must be extended from the client initiator. Not all client systems will detect
the LUN size change until they reboot.
2. Select a LUN so that it highlights. Details appear in the top-right LUN information panel.
64
535-0008-00 Rev 10
Manage
ALUAEnables or disables Asymmetric Logical Unit Access. For more information, see
Using Asymmetric Logical Unit Access (ALUA) on page 66.
Read OnlyActivating read-only access on an existing LUN can cause data loss for
connected clients with cached data.
Online PortsUnchecking these options takes the LUN offline for the respective Memory
Gateway nodes.
NACAprovides a performance enhancement for IBM AIX systems. For more
information, see AIX Compatibility NACA Bit Support on page 179.
Caution: After the NACA option is enabled, you should only export the LUN to an AIX host.
535-0008-00 Rev 10
65
Chapter 2 - Manage, Monitor, and Administer with the Violin Web Interface
Resize LUN.
Note: You can increase the size of a LUN. However, decreasing the size is not allowed.
4. Enter the
5. Click
OK.
66
535-0008-00 Rev 10
Manage
Note: ALUA is not enabled by default. For optimum performance, it is recommended that you
2. In the LUNs panel, highlight the LUN for which you want to enable ALUA. Information about
Changing the LUN ID after a LUN has been exported is not recommended, as traffic is
disrupted during the change. The effect on the client depends on how it handles LUN ID
changes.
You can export LUNs to different igroups on different targets, even when they have initiators in
common. This is an enhancement for vMOS-6.0 only.
Note: After you create an export for a specific iSCSI initiator using a specific target, the initiator
is excluded from seeing LUN exports from that target to all initiators.
2. In the LUNs panel, select the check box next to the LUN(s) you want to export.
535-0008-00 Rev 10
67
Chapter 2 - Manage, Monitor, and Administer with the Violin Web Interface
Note: The check box next to the LUN name must show a check mark for the LUN to be
included in the export.
3. Click Export Checked LUNs. The Add Export dialog box appears.
4. Do one of the following to select the initiators to which the LUNs will be exported.
To export to select initiator groups or initiators, select Specific Initiator Groups and/or
Initiators. Select one or more Initiator Groups from the list, and enter and select initiator
names (each on their own line) in the Individual Initiator panel.
5. Specify the target ports through which the LUNs are exported in one of the following ways:
Select Specific Ports, to export through target ports selected from the list.
Select Value and enter a value in the text field to manually assign a LUN ID.
User-defined LUN IDs can make it easier to differentiate between LUNs. If you do assign a
user-defined LUN ID, Violin Memory recommends that you specify a number below 255 as
some operating systems (for example, Windows) only discover LUN IDs between 0 and
254. This section is automatically disabled when you create multiple LUNs at one time.
68
Select Auto to automatically assign an ID to the exported LUNs. This option is mandatory
when you export multiple LUNs at once.
535-0008-00 Rev 10
Manage
Caution: Re-exporting a LUN with an auto-assigned LUN ID may cause the system to assign a
new LUN ID to the export, which may disrupt the existing I/O traffic. The same holds true for manual
LUN ID changes.
7. Click
OK. Initiator and target information for the exported LUNs appears in the Exports panel
on the right side of the window.
Manage Initiators
In a Violin Memory environment, hosts (such as database servers or application servers) that
access LUNs are initiators and the arrays are the targets. You can control initiator access to target
LUNs in the following ways:
The use of igroups streamlines initiator management by allowing you to define a group of initiators
(igroup), and then assign that group access to select LUNs. So instead of specifying access for
each individual initiator, you specify access for the igroup and all the initiators that are members of
the igroup are granted access to the LUNs.
The Web Interface allows you to easily create and modify initiator groups, as described in Creating
and Deleting Initiator Groups on page 69 and Adding or Deleting Initiators from a Group on page
71. You can also combine initiator groups within another initiator group, as described in Creating,
Modifying, and Deleting Nested Initiator Groups on page 72.
Note: Check with your network administrator before adding initiator hosts in the Advanced panel,
535-0008-00 Rev 10
69
Chapter 2 - Manage, Monitor, and Administer with the Violin Web Interface
Manage > Manage Initiators. The Manage Initiators and igroups page appears.
Create igroup.
Note: Adding an existing igroup to a new igroup creates a nested igroup. For more information,
see Creating, Modifying, and Deleting Nested Initiator Groups on page 72.
2. Select the check box for one or more initiator groups you want to delete.
A check mark appears next to the name of a selected igroup. Only igroups with a check mark
next to the name will be deleted.
70
535-0008-00 Rev 10
Manage
3. Click
Delete.
OK.
A prompt appears verifying the success of the operation, and the selected initiator groups are
removed from the list.
Caution: Adding an initiator to multiple igroups can create conflicts, and cause problems in
export behavior.
Prerequisites
When adding initiators, verify the validity and availability of the address. Check with your system
administrator regarding the availability status of the initiators.
Each initiator address has a protocol-specific identifier:
535-0008-00 Rev 10
71
Chapter 2 - Manage, Monitor, and Administer with the Violin Web Interface
2. Highlight the initiator group in list that you want to modify. The initiators and igroups that are
4. To remove an initiator host or igroup from the group, deselect a check box.
5. To add initiator hosts that are not on the list, under Custom Initiators, do the following:
a. Click
More.
b. In the text box at the bottom of the window, enter a valid initiator host address.
c. Press
Enter to add another address. Each address must be on its own line.
Modify.
Caution: Custom initiator hosts appear in the group even if the host has not yet been added to
the network. Check with your network administrator for status on the availability of these hosts.
Note: You can create up to three hierarchy levels within a nested igroup.
72
535-0008-00 Rev 10
Manage
Prerequisite
You must have created igroups, as described in Creating and Deleting Initiator Groups on page 69.
To create a nested igroup, do the following:
1. Select
Manage > Manage Initiators. The Manage igroups and Initiators page appears.
2. In the All igroups and Initiators panel, highlight the igroup in which you will nest other igroups,
Modify.
5. A prompt appears to tell you that the operation was successful. Click
OK.
The results of the modification appear in the hierarchy tree on the right, as well as in the igroups
and Initiators columns to the left.
535-0008-00 Rev 10
73
Chapter 2 - Manage, Monitor, and Administer with the Violin Web Interface
Manage > Manage Initiators. The Manage igroups and Initiators page appears.
2. In the All igroups and Initiators panel, highlight the nested igroup you want to modify, then click
Edit igroup.
3. Select a check marked box to remove the item from the igroup. Select other check boxes to
Modify.
5. Click
OK to verify the success of your actions. The results appear in the Manage igroups and
Initiators panel.
Manage > Manage Initiators. The Manage igroups and Initiators page appears.
2. In the All igroups and Initiators panel, highlight the nested igroup you want to deconstruct (i.e.
Modify.
4. Click
OK to verify the success of your actions. The results appear in the Manage igroups and
Initiators panel.
Manage > Manage Initiators. The Manage igroups and Initiators page appears.
2. In the All igroups and Initiators panel, select the check box for the nested igroup you want
removed.
Note: Be sure to deselect the check marked boxes of items you do not want removed.
3. Click
74
Delete. The nested igroup is removed, and disappears from the list.
535-0008-00 Rev 10
Manage
Manage Targets
In a cluster, each Memory Gateway operates as a SAN (Fibre Channel/InfiniBand), or LAN (iSCSI)
target that provides access to the LUNs stored on its attached memory arrays. Fibre Channel,
InfiniBand, and iSCSI target ports are all are supported. Every target is either a specific port on a
hardware Fibre Channel host bus adapter (HBA), InfiniBand host channel adapter (HCA), or a
software-configured target for iSCSI.
Manage > Manage Targets. The Target Management page appears, showing
information for all available target ports in one table.
2. To view details on a specific target, select (highlight) the target in the Available Targets panel
on the left.
Fibre Channel target ports are automatically configured when you create the storage
containers on Violin Memory Arrays.
InfiniBand target ports are automatically configured when you create the storage containers on
the Violin Memory Arrays.
535-0008-00 Rev 10
75
Chapter 2 - Manage, Monitor, and Administer with the Violin Web Interface
2. In the Manage Target Protocols panel, click the appropriate Disable protocol button.
For information on LUN group performance monitoring, see Monitor LUN Groups on page 111.
2. To view LUN group information, select a name in the LUN Group List so that it changes color
(highlights). A list of LUNs belonging to the group appears in the right panel.
3. (Optional) Change the order of the list (ascending/descending) by clicking the Up or Down
76
535-0008-00 Rev 10
Manage
A LUN group can contain a minimum of one LUN and a maximum of eight LUNs.
A LUN group name must start with an alphanumeric character, and may contain dashes,
underscores, and periods.
Note: Deleting a LUN group does not effect the LUNs belonging to the group, it only deletes the
Manage > Manage LUN Groups. The LUN Group Management page appears.
2. Click the
3. In the Create a new LUN Group dialog, enter a unique LUN Group Name in the text field, then
LUNs list. Likewise, to remove a LUN from the group, click the Remove radio button in the
Selected LUNs list.
6. Click OK to create the LUN group. The new LUN group appears in the LUN Group List.
535-0008-00 Rev 10
77
Chapter 2 - Manage, Monitor, and Administer with the Violin Web Interface
Manage > Manage LUN Groups. The LUN Group Management page appears.
2. In the LUN Group List, select the LUN group so that it highlights, changing color.
3. Click the
Note: Deleting a LUN group does not effect the LUNs belonging to the group, it only deletes the
procedure.
Note: A LUN group can contain a minimum of one LUN and a maximum of eight LUNs.
2. In the LUN Group List, select the LUN group you want to modify so that it highlights, changing
color.
3. Click the
78
535-0008-00 Rev 10
Manage
4. (Optional) To search for a LUN to add or delete, enter a text string in the appropriate search
Click the Add arrow in the Available LUNs list to add the associated LUN to the group.
Click the Remove radio button in the Selected LUNs list to delete the associated LUN from
the group.
6. Click OK to accept the changes, or Cancel to abort the operation without making any changes.
Data Management
vMOS Data Management allows you to manage and protect sensitive data, as well as simplify
backup and disaster recover with the following features.
For a list of data management terminology and use case scenarios, see CHAPTER 8, Data
Management on page 253.
535-0008-00 Rev 10
79
Chapter 2 - Manage, Monitor, and Administer with the Violin Web Interface
The following table provide a list of tasks and their associated functions. Click a hyperlink to jump
directly to the task.
Function
Manage Snapshots
Space Management
Tasks
80
535-0008-00 Rev 10
Manage
Thin LUNs and thick LUNs can be easily distinguished by their icons. Thick LUNs have a solid icon
to the left of the name, while thin LUNs have a partially transparent icon, as shown in the following
example.
Thick LUN
Thin LUN
Figure 2.2 Thick LUN and Thin LUN Icons.
Note: A LUN can be created as thin or thick. However, once a LUN is created as a certain type
(thick or thin), the type cannot be changed. A thick LUN remains thick and a thin LUN remains thin
until deleted.
535-0008-00 Rev 10
81
Chapter 2 - Manage, Monitor, and Administer with the Violin Web Interface
The amount of allocated space (in Gigabytes) for each LUN, or use all available space
equally across the number of LUNs you are creating
4. Select the
5. Click
Thin Provisioning check box, along with other appropriate Settings options.
OK. The newly created thin provision LUN appears at the top of the LUN list with a thin
Note: An exported LUN cannot be deleted. You must unexport the LUN first.
82
535-0008-00 Rev 10
Manage
the procedure.
If you confirm the deletion, the thin LUN disappears from the LUN list.
You perform these procedures the same way you would for a thick LUN. For more information, see
Editing LUNs on page 64.
Caution: Changing the name of an exported LUN (thin or thick) is not recommended.
Manage Snapshots
The Snapshot Management page allows you to view and manage snapshots for LUNs and LUN
groups. You can also view and manage the schedules used to create snapshots and snap groups.
This section covers the following topics:
535-0008-00 Rev 10
83
Chapter 2 - Manage, Monitor, and Administer with the Violin Web Interface
Snapshot Rules
The following rules apply for snapshots:
Caution: Thin provisioned LUNs and LUNs with snapshots should always use ALUA, to prevent
Schedules View: Displays a list of schedules for LUNs and LUN groups, along with the type of
schedule, and the number of snapshots created to date. You can select a LUN or LUN group
from the list to view a list of snapshots in a table, or a calendar depicting the snapshot
schedule.
LUN View: Displays a list of LUNs that have snapshots, along with the size of the LUN and the
number of snapshots to date for each LUN. You can select a LUN and view a list of its
snapshots, or a calendar depicting the snapshot schedule.
LUN Group View: Displays a list of LUN groups, along with the size of the LUN group and
current number of snapshots. You can select a LUN group and view a list of its snapshots, or a
calendar depicting the snapshot schedule.
Note: A snap group is an object that represents a set of snapshots taken across a collection
of LUNs that belong to a LUN group. Snapshots of each LUN in a LUN group are taken, then
grouped as a snap group.
Table View: Displays a list of LUNs or LUN groups with snap shots or snap groups. All LUNs
on the container are not shown.
Calendar View: Displays a calendar with showing scheduled snapshots by the month, week,
day, or in a list.
Note: The LUN View panel (Manage > Manage Snapshots) only shows the LUNs and LUN
groups that have snapshots and snap groups. It does not show all LUNs. To view all available LUNs,
select Manage > Manage LUNs. To view all available LUN groups, select
Manage > Manage LUN Groups.
84
535-0008-00 Rev 10
Manage
For more information on snapshots and how to create and manage them using the CLI, see
CHAPTER 8, Data Management on page 253.
Creating Schedules
This section demonstrates how to create immediate and scheduled snapshots of a LUN or LUN
group.
To create a one-time immediate snapshot, do the following:
1. Select
2. Click
3. Select
Snapshot NOW, and enter a unique Snapshot Name in the text field followed by a
description in the Description text field.
535-0008-00 Rev 10
85
Chapter 2 - Manage, Monitor, and Administer with the Violin Web Interface
OK to create the snapshot, or Cancel to leave the operation without creating a snapshot.
3. Select
4. Enter a unique
Note: A schedule is enabled is by default. If you disable a schedule, you must manually
enable it again before it will begin taking automatic snapshots. Click the check box to disable
the schedule; the check mark disappears.
5. Select one of the following:
One-time snap. A check mark appears, designating that one snapshot will be taken at
86
535-0008-00 Rev 10
Manage
9. Click
OK to create the snapshot, or Cancel to leave the operation without creating a snapshot.
4. Enter a unique
Note: A schedule is enabled is by default. If you disable a schedule, you must manually
enable it again before it will begin taking automatic snapshots. Click the check box to disable
the schedule; the check mark disappears.
5. Select one of the following:
535-0008-00 Rev 10
87
Chapter 2 - Manage, Monitor, and Administer with the Violin Web Interface
6. Select the LUN or LUN Group name from the drop-down list, leave the
box unchecked, and deselect the No End Date option. The dialog box expands.
7. Specify the following:
Enter a Start Date in the text field, or select the date from the pop-up calendar. Then, set
the Start Time using the drop-down menus.
Select the End Date (calendar) and End Time.
8. Specify the snapshot interval by entering a number in the Repeat every text box, and select a
frequency interval from the drop-down menu: Minute(s), Hour(s), Day(s) (default), Week(s),
Month(s).
9. Specify the maximum number of snapshots to keep by entering a number in the Max. number
to keep text box. Then select the following check boxes, as necessary:
Protected: When the number of snapshots exceeds a specified threshold, reclamation
begins with the oldest unprotected snapshot. When this option is checked, a snapshot or
snap group is protected from automatic free space reclamation.
Writable: By default, snapshots and snap groups are read-only. When this option is
checked, the snapshot or snap group is writable.
10.Click
snapshot.
Schedules, then in the Schedules View select a schedule in the list. It highlights,
changing color.
3. Click
Edit.
The Edit Schedule dialog appears, showing information for the selected schedule.
88
535-0008-00 Rev 10
Manage
5. Click
OK to apply the changes, or Cancel to leave the task without changing to the existing
schedule.
To delete a schedule, do the following:
1. Select
2. Click
Schedules, then in the Schedules View select a schedule in the list. It highlights,
changing color.
3. Click
Delete.
default to delete only the schedule. Then, click OK to make the deletion or Cancel to abort the
task.
535-0008-00 Rev 10
89
Chapter 2 - Manage, Monitor, and Administer with the Violin Web Interface
Viewing Schedules
This section demonstrates how to view a list of schedules, associated snapshots, and exported
snapshots.
To view a list of schedules and associated snapshots, do the following:
1. Select
2. Click
Schedules.
The Schedule View panel on the left displays a list of schedules. The One-Time-Grouped
schedule is selected by default, and is a collection of one-time scheduled snapshots. The
schedule is highlighted, and a list of snapshots and snap groups is shown on the right.
Note: The One-Time-Grouped schedule can contain snapshots and snap groups. All other
schedules contain either snapshots or snap groups, but not both.
3. To view the snapshots and snapgroups for another schedule, select the schedule in the
Schedule View.
The schedule changes color (highlights). The associated snapshots and snapgroups appear in
the panel on the right.
The schedule changes color (highlights). The associated snapshots and snap groups appear
in the panel on the right.
3. Select a snapshot from the list on the right. The selected snapshot changes color.
90
535-0008-00 Rev 10
Manage
4. Click the number in the Export column on the far right. A pop-up dialog appears, showing a list
of exported snapshots.
Note: The only LUNs listed in the LUN View panel (Manage
2. Click
LUNs to display the LUN View panel, and select a LUN from the list. The selected LUN
highlights, changing color. The snapshots associated with the LUN appear in the panel to the
right.
3. Click
535-0008-00 Rev 10
91
Chapter 2 - Manage, Monitor, and Administer with the Violin Web Interface
4. Leave
Snapshot NOW selected, and enter a Snapshot Name. You can also enter an
optional Description.
5. Leave
LUN selected, and select the LUN Name from the drop-down list, if not already
selected.
6. Select the following check boxes, as necessary:
OK to create the snapshot, or Cancel to leave the task without creating a snapshot.
92
535-0008-00 Rev 10
Manage
Note: The only LUN Groups listed in the LUN Group View panel (Manage > Manage
Snapshots) are those that have snapshots, not all the available LUN Groups in the container. To
view all LUN groups in the container, select Manage > Manage LUN Groups. For information on
Description.
4. Leave
LUN Group selected and choose a name from the LUN Groups Names drop-down
list.
5. Select the following check boxes, as necessary:
6. Click
535-0008-00 Rev 10
OK to create the snapshot, or Cancel to leave the task without creating a snapshot.
93
Chapter 2 - Manage, Monitor, and Administer with the Violin Web Interface
Schedules, then select a a LUN group schedule from the list on the left.
The LUN group schedule changes color (highlights) and the snap groups appear in the panel
on the right.
3. Select a snap group from the list on the right. The selected snapshot changes color.
4. In the # column on the far right, click the number link. The snapshots belonging to the snap
5. Click
Hide in the upper right corner of the snap group snapshots panel to hide the list of
snapshots from view.
94
535-0008-00 Rev 10
Manage
Note: The only LUNs or LUN Groups listed in the LUN View or LUN Group View panels (Manage
> Manage Snapshots) are those that have snapshots, not all the available LUN Groups in the
container. To view all available LUNs and LUN groups in the container, select Manage > Manage
LUNs or Manage > Manage LUN Groups. For information on editing snapshot schedules, see
2. Click
LUN or LUN Groups, and then select the LUN or LUN Group name from the list on the
left so that it highlights, changing color.
A list of snapshots associated with the selected LUN or LUN Group appears in the panel on the
right.
3. In the right panel, select the snapshot you want to edit so that it highlights, changing color.
535-0008-00 Rev 10
95
Chapter 2 - Manage, Monitor, and Administer with the Violin Web Interface
Description
Protected
Writable
5. Click
OK to apply the changes, or Cancel to leave the operation without making any changes.
Note: The only LUNs or LUN Groups listed in the LUN View or LUN Group View panels (Manage
> Manage Snapshots) are those that have snapshots, not all the available LUN Groups in the
container. To view all available LUNs and LUN groups in the container, select Manage > Manage
LUNs or Manage > Manage LUN Groups.
3. Select a LUN or LUN Group name in the list on the left so it highlights, changing.
A list of snapshots associated with the selected LUN or LUN Group appears in the panel on the
right.
4. In the right panel, select the check box for the snapshot you want to export. A check mark
appears.
96
535-0008-00 Rev 10
Manage
5. Click the
Export icon.
The Add Export dialog appears. All Initiators and All Ports are the defaults.
To export the snapshot or snap group to All Initiators and All Ports, go to step 5.
To export to selected initiators and initiator groups, click Specific Initiator Groups and/
or Initiators, and select the Initiator Groups and/or Individual Initiators from the lists.
Press Ctrl to select multiple initiators or initiator groups.
To export to selected ports, click Specific Ports, and select the ports from the list. Press
Ctrl to select ports.
LUN ID: Auto (automatic assignment of a LUN ID) is selected by default. To specify a
unique LUN ID, click Value and enter a number in the text box.
Note: LUN IDs must be unique within the container.
7. Click
OK to export the snapshot to the selected initiators and ports, Or click Cancel to exit the
operation without facilitating an export.
535-0008-00 Rev 10
97
Chapter 2 - Manage, Monitor, and Administer with the Violin Web Interface
Note: The only LUNs or LUN Groups listed in the LUN View or LUN Group View panels (Manage
> Manage Snapshots) are those that have snapshots, not all the available LUN Groups in the
container. To view all available LUNs and LUN groups in the container, select Manage > Manage
LUNs or Manage > Manage LUN Groups.
To remove all exports from a LUN or LUN group snapshot, do the following:
1. Select
2. Click
3. Click the check box to the left of a LUN or LUN Group snapshot (that you want to remove the
5. Click
OK to remove all exports, or Cancel to leave the operation without removing the exports.
Note: A LUN or LUN group has to be unexported prior to performing a rollback. For more
information, see Exporting and Unexporting LUNs in a LUN Group on page 79.
3. Select the LUN or LUN group (you want rolled back) from the list of snapshots on the left.
98
535-0008-00 Rev 10
Manage
4. Select the LUN or LUN Group snapshot (whose contents you want to roll back to the LUN or
LUN group) from the list on the right, so that it changes color (highlights).
5. Click
6. Click
OK to remove all exports, or Cancel to leave the operation without rolling back the LUN
or LUN group.
Snapshots that belong to a snap group cannot be deleted individually. However, you can
delete the entire group of snapshots.
Snapshots that have been exported cannot be deleted. You must unexport the snapshot first,
then it can be deleted.
Protected snapshots cannot be deleted, unless you have Admin user privileges. However, it is
recommended that you unprotect the snapshots first, then delete them.
Note: The only LUNs or LUN Groups listed in the LUN View or LUN Group View panels (Manage
> Manage Snapshots) are those that have snapshots, not all the available LUN Groups in the
container. To view all available LUNs and LUN groups in the container, select Manage > Manage
LUNs or Manage > Manage LUN Groups.
3. Select a LUN or LUN Group from the list on the left. A list of associated snapshots appears in
535-0008-00 Rev 10
99
Chapter 2 - Manage, Monitor, and Administer with the Violin Web Interface
5. Click the
Delete icon.
6. Click OK to delete the selected snapshot, or Cancel to leave the operation without deleting the
snapshot.
monthshows a current month planner, with schedules appearing on days snapshots are
taken
weekshows a current week planner, with schedules appearing on days snapshots are taken
dayshows a current day planner, with schedules appearing at the time they are taken
listshows a list of scheduled snapshots for the current day
Note: Calendar View reflects current schedules only. When you change or delete a schedule, the
100
535-0008-00 Rev 10
Manage
3. To display a week-planner schedule view, click week. Schedules appear in the time slot of the
day or days on which the snapshots are taken. The current day is highlighted.
535-0008-00 Rev 10
101
Chapter 2 - Manage, Monitor, and Administer with the Violin Web Interface
list.
Table View.
Space Management
You can monitor and manage the available space on containers for regular (thick) LUNs, thin LUNs,
LUN groups, snapshots, and snap groups. This section covers the following topics:
2. Review the left-most panel for the following specific container information:
Container Name
Number of LUNs (in the container)
HA (High Availability) Status
3. In the center panel, review container space usage in the following areas:
102
535-0008-00 Rev 10
Manage
Provision Thresholds:
Alert at: A warning occurs when the provisioned space reaches or goes above this
percentage
Limit: Prevents any space provisioning above this percentage
Used Space Thresholds
Alert at: A warning occurs when space usage reaches or goes above this level
Reclaim: Space reclamation starts when space usage goes above this level
Container View.
535-0008-00 Rev 10
103
Chapter 2 - Manage, Monitor, and Administer with the Violin Web Interface
3. Review the top-most pane for information on maximum usable capacity, allocated size,
Gigabytes) used for LUNs, snapshots, and snap groups, as well as the amount available free
space. The graph in the bottom half of the pane illustrates the data.
5. Review the right Provisioned pane for a breakdown of the logical space (Terabytes and
Gigabytes) provisioned for LUNs and snapshots, as well as the remaining available free
space. The graph in the bottom half of the pane illustrates the data.
Note: The LUN View panel only shows LUNs with snapshots, not all LUNs in the container.
LUN View, then select a LUN name from the list on the left so that it highlights.
The details of the selected LUN appear in the panel on the right, as shown in the following
example.
104
535-0008-00 Rev 10
Monitor
3. View a list of LUN names, the Type, Provisioned Size (Gigabytes), and Size Used (Gigabytes)
5. Sort the contents of a column by clicking the up arrow (ascending) or down arrow
(descending).
6. To view details of a specific LUN, select the LUN in the list (in the left pane) so it highlights,
changing color. The details for the LUN appear in the LUN Details pane (on the right).
7. Enter a text string in the Search field to search for a particular snapshot of the selected LUN.
Monitor
The Monitor function provides an at-a-glance view of the health and performance of cluster nodes,
and the ability to analyze historical data and real time trends. LUN monitoring provides container
level statistics both on a per-node basis, as well as aggregated. Target monitoring provides port
level statistics.
The following table provide a list of tasks and their associated functions. Click a hyperlink to jump
directly to the task.
Function
Monitor Overview
Monitor LUNs
Monitor LUN Groups
Monitor Targets
535-0008-00 Rev 10
Tasks
105
Chapter 2 - Manage, Monitor, and Administer with the Violin Web Interface
Monitor Overview
The Overview function provides a high-level view of real-time or historical data. You can customize
how you view the data, choosing between line and area graphs, time trends, as well as the type of
metrics used to formulate the data. You also have the option to view data for a container, LUN, or
port.
Real Time Performance TrendShows the last five minutes of performance trends for the
selected metric. New real time values are appended into the graph every ten seconds.
Current ValueShows real time data at this time. The graph time is listed at the bottom of the
panel.
Settings & Download. The panel expands to show the following options.
3. Select the
For more information, see Monitoring Container, LUN, and Port Data on page 108.
4. Select a Metric display type from the drop-down list, as shown in the following example.
A only shows data for the Gateway master node. B only shows data from the standby
node.
535-0008-00 Rev 10
Monitor
5. Choose an option from the drop-down list to customize the graph display. The Real-Time
6. To specify the real-time performance trend, at the top of the left-most panel choose one of the
following options:
7. To view detailed information for a point in time, move the cursor over a point on the graph. A
8. Click
535-0008-00 Rev 10
107
Chapter 2 - Manage, Monitor, and Administer with the Violin Web Interface
5. (Optional) View detailed information on the container by clicking the blue information (i) icon in
6. (Optional) Modify the output display, as described in Customizing the Output Display on page
106.
To monitor a specific LUN, do the following:
1. Select
2. Click
on the right.
108
535-0008-00 Rev 10
Monitor
4. (Optional) View detailed information on the LUN by clicking the blue information (i) icon.
5. Select a Metric from the drop-down list. Click the blue information (i) icon for a description of
each option.
The Current Value panel changes to display the data for the selected metric.
6. (Optional) Modify the output data display, as described in Customizing the Output Display on
page 106.
To monitor a port, do the following:
1. Select
2. Click
Type drop-down list, then choose a port name from the drop-down list
port name.
5. Choose a Metric from the drop-down list, as available. Some port selections only provide one
option.
6. (Optional) Modify the output data display, as described in Customizing the Output Display on
page 106.
535-0008-00 Rev 10
109
Chapter 2 - Manage, Monitor, and Administer with the Violin Web Interface
Monitor LUNs
You can monitor the performance of selected LUNs in a container. The LUN Monitoring page allows
you to search for and view real-time statistics or historical trends for selected LUNs.
3. Select the LUN so that it highlights. Real-time data for the LUN appears in the panels on the
right.
Read/Write Bandwidth
Read/Write IOPs
Read/Write Latency
2. To find a specific LUN, follow the instructions described in Searching for a LUN on page 110,
point in time.
110
535-0008-00 Rev 10
Monitor
Monitor > Monitor LUN Groups. The LUN Monitoring page appears.
Read/Write Bandwidth
Read/Write IOPs
Read/Write Latency
2. To find a specific LUN, follow the instructions described in Searching for a LUN Group on page
point in time.
Monitor Targets
You can analyze performance data for target ports in time statistics, as well as using historical
trends.
Note: Fibre Channel data is shown in frames per second. InfiniBand data is shown in packets per
second.
535-0008-00 Rev 10
111
Chapter 2 - Manage, Monitor, and Administer with the Violin Web Interface
Monitor > Monitor Targets. The Target Port Monitoring page appears. Statistics for
each target port are shown in the left panel, with real time data in a graph format on the right.
2. Select a port name from the list on the left so that it highlights. The graphs on the right side of
Monitor > Monitor Targets. The Target Port Monitoring page appears.
Line
Area
Stacked Line
Stacked Area
The following example shows port data using a Stacked Area graph.
112
535-0008-00 Rev 10
Administration
Administration
The Administration page of the Violin Web Interface provides a single point of contact where you
can perform a wide range of Memory Gateway administration tasks.
The following table provide a list of tasks and their associated functions. Click a hyperlink to jump
directly to the task.
Function Page
Cluster Admin
Tasks
Tools
Information
Network
Notification
535-0008-00 Rev 10
113
Chapter 2 - Manage, Monitor, and Administer with the Violin Web Interface
Cluster Admin
Cluster administration includes configuration and management tasks for the nodes in a cluster. A
cluster consists of the two Memory Gateway nodes, a master node and a standby.
The Master Memory Gateway is the current master node in the cluster. The master node works
together with a standby node to manage the availability of the cluster.
2. To view the version of software running on the Memory Gateway, look at the
Software
ShutdownStops a Memory Gateway node or removes the node from the cluster for planned
maintenance.
RebootRestarts a Memory Gateway node, clearing its data but retaining the existing
configuration settings. In a cluster, the standby Memory Gateway takes over while the
previously active master node reboots.
Caution: Rebooting a Memory Gateway node may cause a temporary decline in overall performance.
The Cluster Administration page displays information on the current software version, the
installed software image, and configuration files. Two boot images are installed for each
Memory Gateway in separate partitions.
Note: Rebooting the system will take several minutes to complete.
2. Do one of the following:
114
535-0008-00 Rev 10
Administration
To reboot the system, click Reboot. In a cluster configuration, the standby Memory
Gateway takes over while the master reboots.
To shut down the system, click Shutdown. In a cluster configuration, the standby Memory
Gateway takes over when the master is shut down.
A pop-up dialog appears.
3. Click
OK to continue with the reboot process, or Cancel to quit without rebooting or shutting
down the system.
WARNING! After the system is shut down, you will have to manually power it back on.
2. To apply a previous configuration to the Memory Gateway, select the check box next to the
3. To save the current Memory Gateway configuration, enter a unique name in the
Save
535-0008-00 Rev 10
115
Chapter 2 - Manage, Monitor, and Administer with the Violin Web Interface
2. To change the number of nodes displayed per page, select a number from the
Display drop-
down list and click Apply. The range of display alerts appears between the Prev and Next
hyperlinks.
3. To view detailed information on a particular node, click the Gateway node link in the list.
The View Gateway page shows the internal and external IP addresses, the nodes role and
status, uptime, software version, and the name of its installed image. The page also shows the
amount of DRAM installed on the node.
4. Click
Viewing Alerts
The Alerts page shows a listing of current and past system events that triggered an alert. Select the
types of alerts to display (informational, warning, or error) and click the APPLY button.
To view cluster alerts, do the following:
1. Select
2. To change the types of alerts that appear in the list, deselect or select the check box next to
the Show icon: Info, Warning, Error. A check mark indicates that the alert type is selected to
appear in the list.
3. Click the
4. To change the number of alerts displayed on a page, choose a number from the Display drop-
116
535-0008-00 Rev 10
Administration
Note: You can only view logs for the current master gateway using the Violin Web Interface.
2. Select a
Log File from the drop-down list, and click Apply. The current log is the default.
Click Next (>) to page forward through the log files and Prev (<) to go back a page.
Click the double greater-than arrows (>>) to jump to the last page, and the double lessthan arrows (<<) to jump back to the first page.
Enter a log page number in the Go to Page text field and press Enter to go directly to a
page.
4. To go to the most recent log entries, click the double arrow icon (>>) last available page in the
display the name of the log file that you are viewing, which appears directly below the Log File field.
535-0008-00 Rev 10
117
Chapter 2 - Manage, Monitor, and Administer with the Violin Web Interface
2. In the Download the Combined Log Files panel, choose one of the following options from the
Start.
The generation generally takes a few minutes to complete. The Generation Status line apprises
you of the progress. When the combined log file is generated, the Download button becomes
active.
4. Click
Download to save the newly generated combined log file to the local machine.
When active, the Download button has a mouse-over ToolTip that shows the log file name,
which includes a time stamp showing when the file was generated. The combined log files are
contained in a single .tar file.
118
535-0008-00 Rev 10
Administration
A users assigned role determines the level of privileges they have in accessing and administering
the cluster. Some of parameters are only visible or configurable by users with certain roles
(privileges). The available roles and their privileges are as follows:
User Role
Privileges
admin
monitor
An Admin user can add new users and define their account privileges.
To add a new user account and define the access privileges, do the following:
1. Select
4. Click
OK to add the new account, Cancel to abort the action, or Reset to clear the fields and
start over.
535-0008-00 Rev 10
119
Chapter 2 - Manage, Monitor, and Administer with the Violin Web Interface
An Admin user can set or reset the passwords and define access privileges for other user accounts.
To add a new user account and define the access privileges, do the following:
1. Select
User Name: Enter the user name for the login credential.
Password: Enter the new password.
Confirm: Re-enter the new password.
4. Click
OK to accept the change, Cancel to abort the action, or Reset to clear the fields and
start over.
An Admin user can change an existing user account, modifying personal and contact information.
To edit an existing user account, do the following:
1. Select
2. Click
120
Role: Select a role from the drop-down list, admin or monitor. For more information, see
Understanding User Roles and Privileges on page 119.
First Name: Enter the users first name.
535-0008-00 Rev 10
Administration
4. Click
OK to accept the changes, Cancel to abort the action, or Reset to clear the fields and
start over. If you accept the changes, another dialog appears.
5. Click
OK in the confirmation dialog to save the changes. The results appear in the User
Management panel.
Delete next to the account name you want to remove. A confirmation dialog appears
3. Click
Yes to confirm you choice, or No to cancel the action. If you click Yes, the user account
is deleted from the list.
535-0008-00 Rev 10
121
Chapter 2 - Manage, Monitor, and Administer with the Violin Web Interface
2. View more details on a node, click the Gateway ID link on the left. The View Gateway page
Network
The Network Web Interface pages provide network configuration information for the cluster, and
allows you to manage the cluster address and global default gateway.
The tasks you can perform on these pages are intended for network administrators, and require
Admin user access privileges. The cluster management address (also known as the virtual IP
address, or VIP) is the IP address for the Memory Gateway master.
Administration > Cluster Admin > Network > Network Settings. The Network
The IP address for cluster management is set as part of the initial out-of-the box configuration.
You can change the cluster management IP address and network mask length in this section,
but it is not recommended. Consult with your network architect before changing these values.
122
535-0008-00 Rev 10
Administration
3. Review the
This refers to Cluster Interface Bonding. For more information on bonded interfaces, see
Network Bond Commands on page 315.
4. Review the
Routing settings.
This is the address for the default global gateway. You can enable and disable the gateway and
set a new address for the default gateway.
5. Review the cluster
This is the routing information for the cluster, including destination, mask, gateway, whether the
node is active or static, and the interface.
Note: Consult with your network administrator before changing the cluster management address.
If you change this address through the Web Interface, you lose connectivity with the cluster.
Administration > Cluster Admin > Network > Network Settings. The Network
Apply.
Administration > Cluster Admin > Network > Network Settings. The Network
click Enable.
3. To change the global default gateway, enter the new default gateway address in the text field
535-0008-00 Rev 10
123
Chapter 2 - Manage, Monitor, and Administer with the Violin Web Interface
Note: If one or more IP addresses for DNS servers were set up as part of the initial out-of-the-
2. Go to the DNS Servers panel at the top of the window to view information on the DNS servers
2. To modify a configured name server, go to the Add or Modify Name Servers panel and change
box.
4. To add a third name server, enter an IP address in the
5. Click
Apply.
124
535-0008-00 Rev 10
Administration
3. Click
Remove Selected.
4. To add a new domain name, enter the fully qualified domain name in the text field in the Add
Synchronization check box. The check mark disappears to show that it is disabled. Then click
Apply.
To enable, disable, or remove an NTP server, do the following:
1. Select
2. To enable an NTP server, click the check box next the server name in the NTP Servers panel.
535-0008-00 Rev 10
125
Chapter 2 - Manage, Monitor, and Administer with the Violin Web Interface
3. To disable an NTP server, click the check box next server name in the NTP Servers panel.
2. Go to the Add New NTP Server panel and enter the new
box.
3. Select a Version from the drop-down list:
Note: Ask your network administrator, if you are unsure which version to choose.
4. Specify whether or not you want the new NTP server enabled or disabled by choosing
Yes or
Caution: Disabling the Violin Web Interface terminates your Web session, after which you can
only access the cluster through the command line interface. For information of the CLI, see
APPENDIX A, CLI Reference on page 287.
126
535-0008-00 Rev 10
Administration
Administration > Network > Web Admin. The Network Settings page appears.
Note: Web Management must be enabled by default, or you wouldnt be able to access the
2. To disable Web Management, click the Enable Web Management check box to deselect this
Note: After you disable the Violin Web Interface, the only way to access the Web Interface again
is to use the web enable command in the CLI. For more information, see APPENDIX A, CLI
Reference on page 287.
Administration > Network > Web Admin. The Network Settings page appears.
HTTP PortPort 80 is the default. Change this to a different free and unused port to
access the Violin Web Interface in the following format:
http://<Management IP address or vCLUSTER name> <port#>/
Web UI Inactivity Timeout (secs)Enables automatic logout and disconnect of the Web
Interface connection to the cluster after a specified period of inactivity. The default auto
logout timeout is 9000 seconds (15 minutes).
Web Session Renewal (secs)Time in seconds before the Violin Web Interface session
cookie expires.
Web Session Timeout (secs)Time in seconds before the Violin Web Interface session
cookie times out.
535-0008-00 Rev 10
127
Chapter 2 - Manage, Monitor, and Administer with the Violin Web Interface
3. Click
Notification
Web Interface notifications provide a means of delivering messages to a set of recipients when an
event or alert occurs. For example, an e-mail can be sent to the administrator when the chassis
internal temperature exceeds a specified threshold.
The Notification function allows you to manage the following:
Call Home
Email Settings
Email Events
SNMP Settings
SNMP Events
Auto SupportEnables the cluster to report alerts and critical events to a specified email
address. Auto Support requires an SMTP server and email domain name.
Fault ReportingSpecifies the SMTP server and mail domain name used for call-home
messages. These settings are also used for alerts.
2. Select the Enable Auto Support Notifications check box. A check mark appears. To turn this
feature off, select the check box again. The check mark disappears.
3. In the Fault Reporting section, it is recommended that you keep the default
callhome.vmem.com. This ensures that Violin Memory Customer Support is notified in the
case of a system critical issue.
4. Enter the Mail Domain Name in the text field, and then click
128
Apply.
535-0008-00 Rev 10
Administration
You can specify whether or not to receive alerts from the vSHARE cluster. Call Home Settings
entries are used to send the alerts. For information on how to enable Call Home settings, see
Specifying Call Home Settings on page 128.
To enable email alerts, do the following:
1. Complete the task as described in To enable auto support and fault reporting, do the following:
on page 128.
2. Select
535-0008-00 Rev 10
129
Chapter 2 - Manage, Monitor, and Administer with the Violin Web Interface
4. Click
You can configure the email server that handles event notifications in the
Note: Go to Managing E-mail Notifications on page 132 for information on how to specify the
Mail Hub: Enter the mail hub fully qualified domain name (FQDN), such as
callhome.example.com.
Mail Hub Port: Enter the port number for the mail hub.
Mail Return Address: Enter an email return address. Use do-not-reply, if you do not
want to receive responses to alert notifications.
Mail Return Domain Name: Enter a domain name for return mail, or leave blank to use
the system domain name.
Include Host Name to Return Address: Clear this check box to remove the host name
from the return email address.
SMTP Authentication (SMTP AUTH) is an extension of the Simple Mail Transfer Protocol (SMTP).
When enabled, an SMTP client logs in using an authentication mechanism supported by the SMTP
server.
The following task demonstrates how to set up SMTP authentication for the e-mail server from
which email alerts will be sent.
To configure SMTP authentication, do the following:
1. Select
130
535-0008-00 Rev 10
Administration
check box. A check mark appears, showing that SMTP Authentication turned on.
3. Enter the following information in the appropriate text fields:
4. Click
Apply to save your configuration entries, or Cancel to delete them and start over.
The Alert Recipient Management section shows information about the current e-mail notification
recipient, if one is configured.
You can edit the account for an existing alert recipient, add new recipients, delete recipients, and
send test e-mails to recipients.
To edit, delete, add and verify alert recipients, do the following:
1. Select
Administration > Notification > Email Settings, and go to the Alert Recipient
Management panel.
Recipient Email Address: Enter a new email address for the recipient in the text field, if
needed.
Get Details? Select this check box to receive alerts as specified by the Call Home
Get Infos? Select this check box to receive less urgent informational events, or deselect
Settings page, or deselect the check box to turn off this option.
535-0008-00 Rev 10
131
Chapter 2 - Manage, Monitor, and Administer with the Violin Web Interface
3. To add a new recipient click Add a New Recipient, then in the dialog that appears, specify the
following:
Recipient Email Address: Enter an email address for the recipient in the text field, if
needed.
Get Details? Select this check box to receive alerts as specified by the Call Home
Get Infos? Select this check box to receive less urgent informational events, or deselect
Get Failures? Select this check box to receive notifications on failure events (such as file
Settings page, or deselect the check box to turn off this option.
the check box to turn off this option.
system errors, process crashes, and unexpected shutdowns), or deselect the check box to
turn off this option.
Click OK to activate the request, Cancel to abort the request, or Reset to clear the fields
and start again.
You can intuitively select notification events to be sent to alert recipients as specified in Managing
Alert Recipients on page 131.
132
535-0008-00 Rev 10
Administration
2. Select the desired events from the list using the following actions:
Expand and collapse a category by clicking the arrow next to its name, or click Show All to
view all events or Hide All to collapse all categories.
Select individual events in a category by clicking the check box next to the name. A check
mark appears when the event is selected. Select the event again to deselect the event
and remove the check mark.
Select all the events in a category by clicking the check mark icon on the right side of the
header.
Deselect all selected events in a category by clicking the X icon on the right side of the
header.
Reset the selections for a category to the default selections by clicking the star icon on the
right side of the header.
3. Click
535-0008-00 Rev 10
133
Chapter 2 - Manage, Monitor, and Administer with the Violin Web Interface
You can interactively enable and disable SNMP Management from the SNMP Settings page.
To enable or disable SNMP management on your system, do the following:
1. Select
2. In the upper panel, select the Enable SNMP Management check box. A check mark appears
3. To disable SNMP Management, select the check box again. The check mark disappears.
4. Click
Configuring SNMP
You can configure SNMP, including settings for the Read-only Community and SNMP Traps
Community.
134
An SNMP TRAP is a message that is sent out when a significant system event is detected. A
trap sink community defines a host and a community string pair. The community string is the
most basic form of SNMP security.
SNMP community strings are like passwords for network elements. One community string is
often used for read-only access with the default set to public. The read-only string is used for
the transfer of data between devices.
A read-write community string is used less often, with a default value of private. The readwrite community string provides the ability to change MIB variables on the device.
535-0008-00 Rev 10
Administration
2. In the SNMP Configuration panel, enter the following information the appropriate text fields:
3. Enable or disable the following options by selecting the check box next to the option, then from
Read-Only Community
SNMP Traps Community
4. Click
A trap sink defines a host and a community string pair. When a trap event occurs the SNMP agent
sends out a trap to each host in the trap sink list using the community string. This section shows
you how to configure a trap sink list, by removing, enabling, disabling, and adding SNMP servers
(hosts) to be included in the list.
The SNMP Trap Sinks panel lists configured SNMP servers, their type-version (type of security
implemented as related to the version of SNMP), if SNMP management is enabled or not, and the
community configuration. For more information on the available SNMP traps, see Available Traps
on page 438.
To remove, enable, or disable SNMP trap sink servers, do the following:
1. Select
535-0008-00 Rev 10
135
Chapter 2 - Manage, Monitor, and Administer with the Violin Web Interface
2. Go to the SNMP Trap Sinks panel and under Add New SNMP Trap Sinks, enter the
SNMP
Yes or No.
5. From the Community drop-down list, choose Public or Private. For more information on these
6. Click
Note: The notifications are sent to the email address specified on Administration > Notification >
Email Settings page. For more information, see Configuring SNMP on page 134.
The SNMP Traps Events page allows you to interactively select SNMP trap events for which
notifications are sent. For information on the available SNMP traps, see Available Traps on page
438.
136
535-0008-00 Rev 10
Administration
2. Select the desired SNMP traps events using the following actions:
Expand and collapse a category by clicking the arrow next to its name, or click Show All to
view all events or Hide All to collapse all categories.
Select individual events in a category by clicking the check box next to the name. A check
mark appears when the event is selected. Select the event again to deselect the event
and remove the check mark.
Select all the events in a category by clicking the check mark icon on the right side of the
header.
Deselect all selected events in a category by clicking the X icon on the right side of the
header.
Reset the selections for a category to the default selections by clicking the star icon on the
right side of the header.
3. Click
535-0008-00 Rev 10
137
Chapter 2 - Manage, Monitor, and Administer with the Violin Web Interface
Tools
Upgrade
Feature Licenses
Note: Before attempting to upgrade the system, always review the latest Release Notes for the
most up-to-date instructions, information on the necessary software image, and required upgrade
paths.
2. Check the status of the nodes in the cluster. Resolve any problems, before continuing with the
upgrade.
Note: If the node count does not match the expected number during an upgrade, you are
prompted with an option to force the upgrade. Warnings of an unexpected node count also
appear on the header status, cluster summary page, index page, versions page, and upgrade
page.
138
535-0008-00 Rev 10
Administration
Enter a URL or the name of an image in the Upgrade from URL or image text box.
Select a file name from the Available Images drop-down list. When you select an image
from this list the name of the image automatically appears in the Upgrade from URL or
image text box.
Note: If you select a URL image for the upgrade and have a slow Internet connection, your
login session may time out before the image download is complete. As a workaround, you can
temporarily extend the timeout period on the Administration > Web Admin page.
4. Select one of the following upgrade modes by clicking the radio button:
Immediate reboot: Upgrades the ACMs and the Memory Gateways immediately. This IS
NOT a non-destructive upgrade.
Staged reboot: Upgrades only the Memory Gateways at specified intervals, or stages. This
is a non-destructive upgrade.
Note: You are logged out of your browser session when the master Violin Memory Gateway
reboots.
5. Click
WARNING! Once an upgrade has begun, aborting the upgrade is not recommended. Doing so
Adding a new feature requires a license key. You must obtain from your Violin Memory Customer
Support representative before you begin the following procedure.
To add a new feature license, do the following:
1. Select
2. In the Add New Licenses panel, enter the license key for the new features in the text box.
Note: If you are adding more than one license, enter each key on a new line.
535-0008-00 Rev 10
139
Chapter 2 - Manage, Monitor, and Administer with the Violin Web Interface
3. Click
Add Licenses. The new feature license appears in the Installed Licenses panel at the
You can easily remove installed feature licenses, either individually or all at once, using the following
procedure.
To remove a license, complete the following steps:
1. Select
2. In the Installed Licenses panel at the top of the window, select the check box of the licenses
3. Click
Information
The Information function provides easy access to the following:
AboutShows which version of the Memory Gateway is running and provides basic
information about trademarks and patents.
140
535-0008-00 Rev 10
Administration
The help for that page appears, as shown in the following example.
3. Click
Hide Page Help to close the Page Help panel, as shown in the previous example.
To view the searchable PDF version of this guide, you must have Adobe Acrobat Reader installed.
You can download Adobe Acrobat Reader free from the Adobe Reader Web page.
To access a full PDF version of this guide, do the following:
1. Select
535-0008-00 Rev 10
141
Chapter 2 - Manage, Monitor, and Administer with the Violin Web Interface
A new tab opens in the browser, displaying a full PDF version of this guide.
3. To save a copy of the guide on your local system, click the Save icon in the upper left corner of
Administration >
142
535-0008-00 Rev 10
CHAPTER 3
This chapter provides basic Memory Array monitoring and administrative tasks in
the following sections, focusing on completing the tasks using the Violin Web
interface. If CLI commands are available for a task, references to those sections
are provided.
See Manage, Monitor, and Administer with the Violin Web Interface on page 31 for
instructions on managing the Memory Gateways that are internal to the Memory
Array.
535-0008-00 Rev 10
143
High-Level Overview
The streamlined Web interface includes the following interactive components, and is shown in
Figure 3.1.
Main Menu BarDisplays clickable icons for the major functional categories: Home, Monitor
and Administration.
Function Menu BarChanges to display the options for the currently active category. Click an
icon to display the associated page and submenu options.
Software VersionDisplays the version of the software currently running on the Memory
Array.
SessionShows the role of the current user at the right, and exits the session when clicked.
Board Status TabExpands to show status of all the Memory Array modules when clicked.
Click the tab again to hide the Board Status page.
Online Help
Session
Software Version
144
535-0008-00 Rev 10
Operating System
Linux
Mac
Windows
Supported Browser
The following table lists the recommended software and settings for optimum performance of the
Violin Web interface.
Recommendation
Supported Browser
Display resolution
JavaScript
Cookies
Context-sensitive online help for any page in the Web interfaceClick the Page Help button in
the upper right corner of the Web interface.
A full PDF version of this guideSelect Administration > Information > Documentation and
then click the guide cover image to view the PDF.
535-0008-00 Rev 10
145
Printing Screens
You can print any screen in the Violin Web Interface by typing CTRL+P and then selecting Print from
the menu.
http://<ACM_management_IP_address>
http://<system_hostname>
If you specify the IP address or host name of a cluster node (other than the master), the
connection is automatically redirected to the Master ACM (Array Controller Module).
The Web Interface log in page appears.
146
535-0008-00 Rev 10
3. Enter a valid user name and password in the text fields of the Login page, as shown in the
following example.
The Logout page appears acknowledging that you successfully logged out.
535-0008-00 Rev 10
147
To log back in, click Login Again and re-enter your credentials.
Web UI
CLI
page 148
page 321
Getting Help
page 172
page 310
page 154
page 239
N/A
page 240
148
535-0008-00 Rev 10
The system name and ID, with the total number of VIMMs in the cluster.
how many VIMMs are in various states (active, spare, booting, etc.)
temperatures of ambient air, controller, and VIMMs (for each Array)
fan status indicators
estimated lifetime of flash storage
status lights: Alarm, Status, and Ready
yellow. A green box with a white check mark represents a healthy, fully functioning fan.
5. Check the VIMM, Controller, and ambient temperatures. Green is healthy, yellow a caution,
535-0008-00 Rev 10
149
All of the modules are color-coded with indicator boxes for alarm status, temperature, and
remaining lifetime.
You can obtain general information about a module by placing the cursor over the module on the
Board Status screen. Click an individual module for more detailed information.
2. Click the
150
535-0008-00 Rev 10
3. Place the cursor over the VIMM for general information. Click the VIMM for more detailed
Above the VIMMs, you can click a RAID Group number to focus on the VIMMs within the selected
RAID group. Below the VIMMs, you can click a Show category to highlight the following types of
VIMMs:
failed
booting
admin down
spares
2. Click the
3. Click the letter next to VIMM Controller on which you want to focus (A, B, C or D).
535-0008-00 Rev 10
151
2. Click the
3. Click the Power Controller letter (A or B) on the bottom left of the Board Status screen.
152
Real Time Performance TrendShows the last five minutes of performance trends for the
selected metric. New real time values are appended into the graph every ten seconds.
Current ValueShows real time data at this time. The graph time is listed at the bottom of the
panel.
535-0008-00 Rev 10
2. To specify the real-time performance trend, at the top of the left-most panel choose one of the
following options:
3. To view detailed information for a point in time, move the cursor over a point on the graph. A
Download button.
b. Choose a Time Interval from the drop-down list.
c. Click Download as CSV. Data for the specified time interval is saved as a CSV (comma
separated value) file. Once saved, the file can be opened using Microsoft Excel.
535-0008-00 Rev 10
153
Administration
The Administration page of the Violin Web Interface provides a single point of contact where you
can perform a wide range of Memory Array administration tasks.
You must have Admin privileges to perform the following tasks using the Administration function
pages.
Function
Page
Array Admin
Tasks
Tools
Information
Network
Notification
Array Admin
Array administration includes configuration and management tasks for the Memory Array.
Rebooting and Shutting Down the Memory Array
The Cluster Administration page displays information on the current software version, the installed
software image, and configuration files. Two boot images are installed for each ACM in separate
partitions.
The Software section displays the software version number and allows you to choose which boot
image the ACMs should use (the Switch Boot Partition button). Two boot images are installed for
each ACM in separate partitions.
154
535-0008-00 Rev 10
Administration
Administration > Array Admin > Reboot. The Cluster Administration page appears.
OK to continue with the reboot process, or Cancel to quit without rebooting or shutting
2. To apply a previous configuration to the Memory Array, select the check box next to the Config
3. To save the current Memory Array configuration, enter a unique name in the
Save
To delete a configuration, select the check box next to the Config Name and click Delete.
535-0008-00 Rev 10
155
Viewing ACMs
View ACMs to see a list of all nodes in the cluster. The status of each ACM is normally online
although transition states also occur while the ACM is joining or leaving the cluster.
Select Administration > Array Admin > ACMs. The ACM List page appears.
The Master ACM is the current master control node in the cluster. It works together with a Standby
to manage the availability of the cluster.
To view more detail about an ACM:
Click View to show more details about an individual ACM. Alternatively, you can show the
details by clicking the ACM ID in the ACM column.
This screen shows the software version running on the ACM, its IP addresses, status and role. It
also shows how long the ACM has been running (Uptime).
Viewing Alerts
The Alerts page shows a listing of current and past system events that triggered an alert. Select the
types of alerts to display (informational, warning, or error) and click the APPLY button.
To view alerts, do the following:
1. Select
156
535-0008-00 Rev 10
Administration
2. To change the types of alerts that appear in the list, deselect or select the check box next to
the Show icon: Info, Warning, Error. A check mark indicates that the alert type is selected to
appear in the list.
3. Click the
4. To change the number of alerts displayed on a page, choose a number from the Display drop-
Administration > Array Admin > Logs. The Memory Array Logs page appears.
Click Next (>) to page forward through the log files and Prev (<) to go back a page.
Enter a log page number in the Go to Page text field and press Enter to go directly to a
page.
3. To view an archived log, or to return to the current log after viewing another one, click the Log
File field, select a file from the drop-down list, then click Apply.
The Log File field displays the top file in the drop-down list (Current log). It does not display the
name of the log file that you are viewing, which appears directly below the Log File field.
Administration > Array Admin > Versions. The Versions page appears with a list of
535-0008-00 Rev 10
157
To view more details on a ACM, click the ACM ID link on the left. The View ACM page appears, as
described in Viewing ACMs on page 156.
The system administrator can permit others to access the Memory Array Web interface and the CLI.
This section provides an overview of the two user roles and the privileges available to each.
User Roles
The Memory Array has two built-in system accounts:
Caution: Initially these accounts have no password. Setting a password for admin is highly
recommended.
A users assigned role determines the level of privileges they have in accessing and administering
the Memory Array. Some of parameters are only visible or configurable by users with certain roles
158
535-0008-00 Rev 10
Administration
(privileges). The available roles and their privileges are as follows. Privileges are set by assigning
one of the following roles to a user account:
admin:
The administrator role has full privileges to view anything, take any action, or
change any configuration. This role can access every screen in the Violin Web
Interface.
monitor:
The monitor role can read all data and perform some actions, such as rebooting
the system and configuring various system parameters, but cannot change the
configuration of the Violin Memory Array. This role can view some of the screens in
the Web Interface.
The CLI command modes Config, Enable, and Standard correspond to the user roles admin,
monitor, and unpriv, respectively.
You can add user accounts and set their privileges and passwords in the Administration section of
the Violin Web Interface.
Adding a User and Specifying a Role
An Admin user can add new users and define their account privileges.
To add a new user account and define the access privileges, do the following:
1. Select
4. Click
OK to add the new account, Cancel to abort the action, or Reset to clear the fields and
start over.
535-0008-00 Rev 10
159
An Admin user can set or reset the passwords and define access privileges for other user accounts.
To add a new user account and define the access privileges, do the following:
1. Select
User Name: Enter the user name for the login credential.
Password: Enter the new password.
Confirm: Re-enter the new password.
4. Click
OK to accept the change, Cancel to abort the action, or Reset to clear the fields and
start over.
An Admin user can change an existing user account, modifying personal and contact information.
To edit an existing user account, do the following:
1. Select
2. Click
Edit for the desired user. The Edit Existing User dialog appears.
160
Role: Select a role from the drop-down list, admin or monitor. For more information, see
Understanding User Roles and Privileges on page 158.
First Name: Enter the users first name.
535-0008-00 Rev 10
Administration
4. Click
OK to accept the changes, Cancel to abort the action, or Reset to clear the fields and
start over. If you accept the changes, another dialog appears.
5. Click
OK in the confirmation dialog to save the changes. The results appear in the User
Management panel.
Delete next to the account name you want to remove. A confirmation dialog appears.
3. Click
Yes to confirm your choice, or No to cancel the action. If you click Yes, the user account
is deleted from the list.
535-0008-00 Rev 10
161
Recipient E-mail Address: Enter a new email address for the recipient in the text field, if
needed.
Get Details? Select this check box to receive alerts as specified by the Call Home
Settings page, or deselect the check box to turn off this option.
Get Infos? Select this check box to receive less urgent informational events, or deselect
the check box to turn off this option.
Get Failures? Select this check box to receive notifications on failure events (such as file
system errors, process crashes, and unexpected shutdowns), or deselect the check box to
turn off this option.
Click OK to activate the request, Cancel to abort the request, or Reset to clear the fields
and start again.
3. To add a new recipient click Add a New Recipient, then in the dialog that appears, specify the
following:
Recipient Email Address: Enter an email address for the recipient in the text field, if
Get Details? Select this check box to receive alerts as specified by the Call Home
Settings page, or deselect the check box to turn off this option.
Get Infos? Select this check box to receive less urgent informational events, or deselect
the check box to turn off this option.
needed.
162
535-0008-00 Rev 10
Administration
Get Failures? Select this check box to receive notifications on failure events (such as file
system errors, process crashes, and unexpected shutdowns), or deselect the check box to
turn off this option.
Click OK to activate the request, Cancel to abort the request, or Reset to clear the fields
and start again.
To delete an alert recipient click Delete next to the recipients name, then in the dialog that appears
click Yes to confirm or No to cancel the action.
Network
The Network Web interface pages provide network configuration information for the Memory Array,
and allows you to manage the cluster address and global default gateway.
The tasks you can perform on these pages are intended for network administrators, and require
Admin user access privileges. The cluster management address is the IP address for the Violin
Memory Array cluster.
Administration > Network > Network Settings. The Network Settings page appears.
2. Verify the
The IP address for cluster management is set as part of the initial out-of-the box configuration.
You can change the cluster management IP address and network mask length in this section,
but it is not recommended. Consult with your network architect before changing these values.
3. Review the
Routing settings.
This is the address for the global default gateway. You can enable and disable the gateway and
set a new address for the default gateway.
5. Review the cluster
This is the routing information for the Memory Array, including destination, mask, gateway,
whether the node is active or static, and the interface.
network administrator before changing the cluster management address. If you change this
address through the Web Interface, you lose connectivity with the cluster.
535-0008-00 Rev 10
163
You can change the cluster management address from the Network Settings page. In most cases,
it is recommended that you change the cluster management address through a serial console using
the CLI. However, there are few situations where changing the cluster management address
through the Web Interface is advised, one being in preparation for a physical move of the devices.
To change the cluster management address, do the following:
1. Select
Administration > Network > Network Settings. The Network Settings page appears.
Apply.
Administration > Network > Network Settings. The Network Settings page appears.
click Enable.
3. To change the global default gateway, enter the new default gateway address in the text field
Note: If one or more IP addresses for DNS servers were set up as part of the initial out-of-the-
2. Go to the DNS Servers panel at the top of the window to view information on the DNS servers
164
535-0008-00 Rev 10
Administration
2. To modify a configured name server, go to the Add or Modify Name Servers panel and change
box.
4. To add a third name server, enter an IP address in the
5. Click
Apply.
3. Click
Remove Selected.
535-0008-00 Rev 10
165
2. To enable an NTP server, click the check box next to the server name in the NTP Servers
3. To disable an NTP server, click the check box next to the server name in the NTP Servers
2. Go to the Add New NTP Server panel and enter the new
box.
3. Select a Version from the drop-down list:
Note: Ask your network administrator, if you are unsure which version to choose.
166
535-0008-00 Rev 10
Administration
4. Specify whether or not you want the new NTP server enabled or disabled by choosing
Yes or
Caution: Disabling the Violin Web interface terminates your Web session, after which you can
only access the Memory Array through the command line interface. For information of the CLI, see
APPENDIX A, CLI Reference on page 287.
Administration > Network > Web Admin. The Network Settings page appears.
Note: Web Management must be enabled by default, or you wouldnt be able to access the
2. To disable Web Management, click the Enable Web Management check box to deselect this
Note: After you disable the Web management, the only way to access the Web Interface again
is to use the web enable command in the CLI. For more information, see APPENDIX A, CLI
Reference on page 287.
Administration > Network > Web Admin. The Network Settings page appears.
Enable HTTPEnables (or disables) HTTP on the specified port. Port 80 is the default.
Change this to a different free and unused port to access the Violin Web Interface in the
following format:
http://<Management IP address or vCLUSTER name> <port#>/
Enable HTTPSEnables (or disables) HTTPS on the specified port. The default is 443.
Change it to another available port as described in HTTP Port setting above.
535-0008-00 Rev 10
167
Session RenewalTime in seconds before the Web interface session cookie expires.
Session TimeoutTime in seconds before the Web interface session cookie times out.
Click Apply to save the configurations, or Cancel to return to the previous settings.
Notification
Web interface notifications provide a means of delivering messages to a set of recipients, when an
event or alert occurs. For example, an e-mail can be sent to the administrator when the chassis
internal temperature exceeds a specified threshold.
The Notification function allows you to manage the following:
Call Home
Email Settings
Auto SupportEnables the Memory Array to report alerts and critical events to a specified
email address. Auto Support requires an SMTP server and email domain name.
Fault ReportingSpecifies the SMTP server and mail domain name used for call-home
messages. These settings are also used for alerts.
2. Select the Enable Auto Support Notifications check box. A check mark appears. To turn this
feature off, select the check box again. The check mark disappears.
3. In the Fault Reporting section, enter an SMTP Server address in the text field, such as
callhome.example.com.
4. Enter the Mail Domain Name in the text field.
168
535-0008-00 Rev 10
Administration
5. Click
Apply.
Recipient E-mail Address: Enter a new email address for the recipient in the text field, if
Get Details? Select this check box to receive alerts as specified by the Call Home
needed.
535-0008-00 Rev 10
Settings page, or deselect the check box to turn off this option.
Get Infos? Select this check box to receive less urgent informational events, or deselect
the check box to turn off this option.
Get Failures? Select this check box to receive notifications on failure events (such as file
system errors, process crashes, and unexpected shutdowns), or deselect the check box to
turn off this option.
Click OK to activate the request, Cancel to abort the request, or Reset to clear the fields
and start again.
169
3. To add a new recipient click Add a New Recipient, then in the dialog that appears, specify the
following:
Recipient Email Address: Enter an email address for the recipient in the text field, if
needed.
Get Details? Select this check box to receive alerts as specified by the Call Home
Get Infos? Select this check box to receive less urgent informational events, or deselect
Get Failures? Select this check box to receive notifications on failure events (such as file
Settings page, or deselect the check box to turn off this option.
the check box to turn off this option.
system errors, process crashes, and unexpected shutdowns), or deselect the check box to
turn off this option.
Click OK to activate the request, Cancel to abort the request, or Reset to clear the fields
and start again.
Tools
The Tools pages provide the functions to update or remove feature licenses and upgrade the
Memory Array software.
Adding a new feature requires a license key. You must obtain a license key from your Violin Memory
Customer Support representative before you begin the following procedure.
1. Select
2. In the Add New License(s) panel, enter the license key for the new features in the text box.
Note: If you are adding more than one license, enter each key on a new line.
170
535-0008-00 Rev 10
Administration
3. Click
Add Licenses. The new feature license appears in the Installed Licenses panel at the
You can easily remove installed feature licenses, either individually or all at once, using the following
procedure.
1. Select
2. In the Installed Licenses panel at the top of the window, select the check box of the licenses
3. Click
Information
The Information function provides easy access to the following:
AboutShows which version of the Memory Array software is running and provides basic
information about trademarks and patents.
535-0008-00 Rev 10
171
page 146.
2. On the Web interface page for which you want Help, click the
The help for that page appears, as shown in the following example.
3. Click
Hide Help to close the Page Help panel, as shown in the previous example.
To view the searchable PDF version of this guide, you must have Adobe Acrobat Reader installed.
You can download Adobe Acrobat Reader free from the Adobe Reader Web page.
To access a full PDF version of this guide, do the following:
1. Select
172
535-0008-00 Rev 10
A new tab opens in the browser, displaying a full PDF version of this guide.
3. To save a copy of the guide on your local system, click the Save icon in the upper left corner of
Administration >
page 146.
2. Once logged in to the Memory Array, place your pointer in the address bar at the end of the
Memory Array host name, and then replace what is to the right of the host name with the
following: /xtree/*
535-0008-00 Rev 10
173
For example:
You are now logged into the XTree interface. From here you can view values for and gain
information about many of the Memory Array components, and click a path name to narrow your
search.
To return to the Violin Web Interface, remove what is to the right of the host name from the address
bar and then press Return.
174
535-0008-00 Rev 10
CHAPTER 4
vSHARE Configuration
This chapter describes vSHARE, its features and functions, and the steps required
to configure a Memory Gateway running vSHARE to manage block storage. The
following topics are covered:
Understanding vSHARE
vSHARE is a solution for block storage management. vSHARE runs as software on
a Memory Gateway enabling host systems (for example, database servers) to use
the following transport protocols to access logical units of data (LUNs) stored within
Memory Arrays:
iSCSI
Fibre Channel (FCP)
InfiniBand (SRP)
On the host system, the LUN appears as a local SCSI disk. The host formats and
partitions the LUN. The storage system sees the contents of a LUN as a set of
blocks of arbitrary data.
Administrators may use vSHARE to define rules for controlling access to LUNs
based on the containers, initiators and initiator groups, and target ports configured
within the system.
535-0008-00 Rev 10
175
Storage Containers: A storage container comprises a Violin Memory Array. Every LUN
created and managed by vSHARE is created within a storage container. You can create many
LUNS within a container. However, currently there is a one-to-one (1:1) mapping between an
array and a container (one container per array).
Initiators and Initiator groups: An initiator is a host or client that requires access to the LUNs
stored in the containers. Multiple initiators may be grouped together in an initiator group
(igroup). Access to the LUNs in a container may be restricted to specific initiators or initiator
groups.
Targets: A target identifies the port on the Memory Gateway through which the initiator may
access the LUNs stored in a container. Access to the LUNs in a container may be restricted by
target port.
The vSHARE application runs on Memory Gateways and allows you to manage LUN and initiatortarget relationships. The Memory Gateways operate as the SAN targets and the hosts (such as a
database server or application server) are the initiators.
The Memory Gateway provides network connectivity, utilizing Fibre Channel, InfiniBand, and iSCSI
protocols. The Memory Gateway provides high performance processing, high bandwidth DRAM,
and high bandwidth access to the Violin Memory Arrays. All LUNs created and managed by
vSHARE. are stored in containers, each container comprises a Violin Memory Array that utilizes
flash memory storage.
For maximum performance, Single Level Cell (SLC) flash is used. For capacity to support readcentric applications, Multiple Level Cell (MLC) should be used. The Memory Gateways and Memory
Arrays are connected via PCIe.
176
535-0008-00 Rev 10
To use the SRP protocol with vSHARE InfiniBand, clients must have the SRP initiator software and
networking hardware listed in this section.
Hardware
Quad Data Rate (QDR) InfiniBand (IB) Host Channel Adapters (HCA) and switches are supported.
In particular, the following QDR IB switches are supported:
Mellanox IS5000 family 36-port QSFP 40Gb/s IB switch (IS5025, IS5030 or IS5035)
Qlogic 12200 Series QDR 36-port Edge Switch
Software
The following software SRP initiators are supported:
VMware SRP initiator, (Mellanox OFED Drivers for VMware vSphere 4 and vSphere 5).
Linux SRP initiator, available since November 2005.
Windows SRP initiator, available through the WinOF InfiniBand stack.
Target Ports
You can view information for InfiniBand target ports using the show targets command. For more
information, see show targets Command on page 186. You can view target statistics using the show
stats ib command. For more information, see iSCSI Commands on page 325.
535-0008-00 Rev 10
177
Note: If you are using Fibre Channel, the target ports are automatically configured when you
create the storage container on the Memory Array. If using iSCSI, the target ports must be
configured manually.
3. Configure initiator groups and initiators. For more information, see Manage Initiators on page
69.
4. Configure LUNs. For more information, see Manage LUNs on page 59.
5. Export LUNs. For more information, see Manage LUNs on page 59.
The following flowchart provides a visual overview of the process for configuring the Memory
Gateway for vSHARE block storage.
Note: Configuring target ports and initiator groups are optional steps in the vSHARE
configuration process. Organizations that do not want to use LUN-masking can control access to
LUNs through ports or switches. LUN masking is an authorization process used to make a logical
unit number available to some hosts and unavailable to others.
1. Configuring Storage Containers on page 180. This step initializes the Violin Memory Array to
manage block storage and create one or more storage containers (partitions) to manage the
LUNs. To format the Violin Memory Array from vSHARE, the media init command is
required. This command enables you to initialize one or more Memory Arrays to support block
storage. Once a Memory Array has been initialized to support block storage, you may create
one or more storage containers on each Memory Array, which may be used to store and
manage LUNs.
2.
(Optional) Configuring Target Ports on page 182. This step configures the target ports so they
can control access to LUNs. Configuring target ports is useful for security and bandwidth
management.
178
535-0008-00 Rev 10
3.
(Optional) Configuring Initiator Groups on page 187. This step configures the initiator groups,
adding one or more initiators to each initiator group. You can restrict access to LUNs by
initiator group, or on a per-initiator basis.
LUNs are created within a storage container that comprises a particular Violin Memory Array. A
LUN inherits the attributes of the container (Violin Memory Array) in which it is created.
initiators via target ports. Only those initiator groups or initiators to which the LUN is exported
may access the LUN. Access may be further restricted to a specific target port.
Caution: This feature is not recommended for use with operating systems other than AIX.
You can turn on (1) this NACA bit support when you create a new LUN, as demonstrated in the
following task. Optionally, you can turn on the NACA option after a LUN is created. This feature is
turned off (0) by default.
For instructions on how to turn on NACA through the Violin Web interface, see described in Creating
and Deleting LUNs on page 62. You have the option of modifying this feature through the Web
interface, as described in Editing LUNs on page 64.
To enable NACA bit support when creating a new LUN, do the following:
1. Connect to the Violin Memory Gateway through a terminal port, such as PuTTY.
2. Enter Configure mode, using the following commands.
> enable
# configure terminal
3. Create a new LUN and enable NACA.
535-0008-00 Rev 10
179
When a LUN is created, it is assigned to a specific container and inherits the attributes of that
container.
Configuring a Storage Container uses the following process:
1. Initializing Violin Memory Arrays on page 180
2. Viewing Containers on page 182
180
535-0008-00 Rev 10
Parameters
device <device>
type
name
force
> enable
# configure terminal
(config) #
3. Enter the media init command to initialize the Violin Memory Array.
535-0008-00 Rev 10
181
Viewing Containers
Once you have initialized the Violin Memory Array, you can see the container that was created. This
section covers how to use the show containers command to view existing containers.
Before a block media device can be used for block storage, it must be enabled. In the following task,
you use the media block id <id> enable command to enable the device.
To view existing containers, complete the following steps:
1. Log on as an administrator to the Memory Gateway attached to Port 1 of the Violin Memory
Array.
2. Use the configure terminal command to enter Configure mode.
# configure terminal
(config) #
3. Enter the media block id all enable command to enable all block containers.
182
535-0008-00 Rev 10
Fibre Channel
InfiniBand SRP
iSCSI
Every target is a specific port on a hardware Fibre Channel host bus adapter (HBA), InfiniBand host
channel adapter (HCA), or a software-configured target for iSCSI.
For Fibre Channel or InfiniBand, the target ports are present when the Memory Gateway
boots, if the HCAs or HBAs are installed on the system. Creating a container is not necessary.
Note: The SCSI RDMA Protocol (SRP) is an industry standard protocol for utilizing block storage
A LUN may be exported through multiple target ports of different types. Active-active multipathing
is supported.
Note: Mellanox ConnectX-2 InfiniBand HCA is supported in release G5.2.0. Be advised that the
two ports on Mellanox HCAs are not entirely separate entities with respect to target port masking.
Port steering for hca-a1 also uses hca-a2.
535-0008-00 Rev 10
183
Parameters
[hostname <hostname>]
[protocol <fc/ib/iscsi>]
[sessions]
[detail]
Displays in detail.
Note: In the case of an HA pair, when you create a target and bind it to an interface on one node,
the target must also be a bound to the (same) interface on the other node of the HA pair.
184
535-0008-00 Rev 10
Typically, you should configure one of the four 10GbE interfaces identified as eth3, eth4, eth5,
or eth6 to act as the iSCSI target port. In Step 6 of this procedure, you will bind the iSCSI target
to the IP address of the configured interface.
4. Run the iscsi enable command to enable the iSCSI target ports.
configured.
(config) # show targets
Node Hostname
Protocol Target
Enab Address
---- --------------- -------- ---------- ---- -------4
<hostname> iscsi
<target_name> yes
iqn.200402.com.vmem:<hostname>:<target_name>
The address of the iSCSI target consists of three elements: the IQN (iqn.200402.com.vmem), the <hostname>, and the <target_name>.
The IQN consists of the literal (iqn), the date (yyyy-mm), and the reversed domain name of the
vendor, Violin Memory. A colon prefixes the node_name. The default node_name is generated from
the hostname. A second colon prefixes the user-defined target_name.
iscsi enable Command
The iscsi enable command may be used to enable or disable iSCSI target ports.
Syntax
535-0008-00 Rev 10
185
<target_name>
[hostname <hostname>]
[protocol <fc/ib/iscsi>]
[sessions]
[detail]
Note: For an HA pair, when you create a target and bind it to an interface on one node, the target
must also be a bound to the (same) interface on the other node of the HA pair.
Syntax
186
535-0008-00 Rev 10
Syntax Description
<target_name>
<ip_address>
[to <ip_address>]
iSCSI initiator names have several formats: iqn..., naa..., eui... of which iqn names are
most prevalent. For example, iqn.2004-02.com.vmem.
Fibre Channel initiators are identified by World-Wide Names (WWN). Fibre Channel initiators
are fixed by the appropriate HBA port.
You can include an initiator group within another initiator group to create a nested initiator group
(nested igroup). A nested igroup can contain one or more igroups. For more information on nested
initiator groups, see Working with Nested Initiator Groups on page 194 and Managing Nested
Initiator Groups on page 213.
Initiator Group Configuration
If you intend on using nested initiator groups, it is recommended that you plan the nested igroups
hierarchy following the guidelines described in Working with Nested Initiator Groups on page 194
and Managing Nested Initiator Groups on page 213.
To configure an initiator group, complete the following steps:
1. From the CLI, use the enable command (to enter Enable mode), and then the configure
example.
(config) # igroup create name <my_igroup>
3. Add one or more initiators to the igroup, using the igroup addto command. The
initiator_name must be a valid iqn (iSCSI) or wwn (Fibre Channel), or guid (InfiniBand) name,
such as, iqn.2004-02.com.vmem:initiator1.
535-0008-00 Rev 10
187
name
188
535-0008-00 Rev 10
Examples
The following examples demonstrate how the igroup create command can be used. to manage
initiator groups and initiators.
The following command creates an initiator group named FINANCE:
igroup create name FINANCE
The Cisco-style no prefix may be used to delete initiator groups or initiators. In the following
example, the command deletes the igroup FINANCE.
no igroup create name FINANCE
igroup addto Command
The igroup addto command adds one or more initiators to a group.
Syntax
igroup_name
initiator_name
The following command adds three initiators to the initiator group named FINANCE:
igroup addto FINANCE initiators wwn.20:20:20:20:20:20:20:20
wwn.20:20:20:20:20:20:20:21 wwn.20:20:20:20:20:20:20:22
The following command removes the initiator wwn.20:20:20:20:20:20:20:20 from igroup FINANCE
and preserves the other two initiators (wwn.20:20:20:20:20:20:20:21
wwn.20:20:20:20:20:20:20:22).
no igroup addto FINANCE initiators wwn.20:20:20:20:20:20:20:20
535-0008-00 Rev 10
189
Creating LUNs
On the host system (initiator), the LUN appears as a local SCSI disk. The host may format and
partition the LUN. The target, or storage system, sees the contents of the LUN as a set of blocks of
arbitrary data.
Every LUN is created within a container, and each container comprises a Violin Memory Array.
vSHARE implements LUNs at a minimum of 1GB increments with the ability to configure up to 1024
LUNs per Memory Gateway.
LUN Create Command
The lun create command enables you to create a LUN within a specified storage container.
Command Syntax
[no] lun create container <container_name> name <LUN_name> size [<size GB>
| equal] [quantity <number>] [nozero] [readonly] [startnum <unsigned
integer>] [blksize 512 | 4096] [offline] alua naca
Arguments and Options
size
Defines the size of each LUN in GB, or equal to divide the free
space of the container equally among the new LUNs. A value of
0 (zero) may be used in place of equal.
container
quantity
190
535-0008-00 Rev 10
nozero
readonly
startnum
blksize
offline
alua
naca
535-0008-00 Rev 10
191
container
name
count
reserved
serial
sessions
Viewing Containers
The show containers command displays all available containers.
Command Syntax
show containers
Resizing LUNs
You can increase the size of an existing LUN by specifying a the larger size in Gigabytes (GB), or
by specifying an amount (GB) to add to the current size.
Decreasing the size of an existing LUN is not supported. For a smaller size LUN, create a new LUN
of the desired size.
192
535-0008-00 Rev 10
WARNING! Existing client side file systems that might be stored on the LUN are not extended
as part of the resize operation. Increasing the size of an existing client side file system must be done
on the client. Some client systems do not detect a LUN size change until they reboot.
Exporting LUNs
The fifth step in configuring the vSHARE Memory Gateway is to export the LUNs to an initiator. The
LUN must be exported before it may be accessed over block protocols.
Note: vSHARE LUNs are defined and saved within the Arrays themselves. vSHARE LUN export
records are, however, managed and stored within the vCLUSTER configuration database owned
by the master node.
Access to LUNs may be restricted to a specific initiator group, initiator, or target port when the LUN
is exported using the lun export command.
The optional [igroup | initiator] parameter identifies the initiator group or initiator that
may access the LUN.
The optional [port] parameter identifies the target that may access the LUN.
A LUN ID is a SCSI identifier which may be used to differentiate between devices on the same
target port. By default, vSHARE automatically assigns LUN IDs to LUNs when they are exported to
an initiator group or initiator.
Note: After you create an export for a specific iSCSI initiator using a specific target, the initiator
is excluded from seeing LUN exports from that target to all initiators.
vSHARE optionally enables you to assign a special, user-defined LUN ID to a vSHARE LUN when
you export LUNs to an initiator group or initiator. User-defined LUN IDs may make it easier
differentiate between LUNs. If you do assign user-defined LUN ID, Violin Memory recommends that
you specify a number below 255 as some operating systems (for example, Windows) will only
discover LUN IDs between 0 and 254. LUNs cannot be exported if they are assigned an existing
LUN ID; an error message is returned and the export fails.
535-0008-00 Rev 10
193
Each LUN ID (LUN name) must be unique to avoid the chance of conflicts.
Changing the LUN ID (LUN name) after a LUN has been exported is not recommended, as
traffic is disrupted during the change. The affect on the client depends on how it handles LUN
ID changes.
Re-exporting a LUN with an auto-assigned LUN ID may cause the system to assign a new
LUN ID to the export, which may disrupt the existing I/O traffic. The same holds true for
manual LUN ID changes.
You can export LUNs to different igroups on different targets, even when they have initiators in
common.
Exporting a LUN to different targets using a different LUN ID (LUN name) for each target can
result in conflicts.
194
535-0008-00 Rev 10
OpenVMS Support
lun_name
container <container_name>
name <lun_name>
lunid
igroup
initiator
port
OpenVMS Support
OpenVMS (Open Virtual Memory System) is a multi-user, multiprocessing virtual memory-based
operating system (OS). OpenVMS servers require that LUNs have a valid Unique Device Identifier
(UDID) for discovery and usage by the operating environment.
Support for OpenVMS provides the following capabilities:
Specifying the OpenVMS UDID at the time a LUN is created, or after a LUN is created
Removing a LUN UDID from being reported
Listing LUN UDID information
OpenVMS requires that a target array provide a LUN 0, which is seen as Disk Device Type (0x0).
By default vSHARE provides LUN 0 as the SCSI Enclosure Services Device Type (0x0D).
OpenVMS specifies LUN 0 as a controller device, with other LUNs appearing on other LUN
identifiers.
535-0008-00 Rev 10
195
PuTTY.
2. Enable configure mode, using the following commands.
> enable
# configure terminal
(config) #
3. Create a 1 Gb LUN in a container.
196
55
535-0008-00 Rev 10
PuTTY.
2. Enable configure mode, using the following commands.
> enable
# configure terminal
(config) #
3. To set the UDID of a LUN in a container (array), use the following command.
Note: A LUN group is a management object only, and does not impact the data path.
535-0008-00 Rev 10
197
Note: You can create a LUN group with a maximum of eight LUNs.
Substitute LUN names for the <lun#> variables in the command above. You must have one LUN to
create a LUN group, and optionally can specify seven other LUNs as members.
# lungroup create container <container_name> name <lungroup_name> lun
<lun1> lun [<lun2> lun <lun3> lun <lun4> ... lun <lun8>]
Note: You must delete the snap groups (snapshots of a LUN group) associated with a LUN group
before you are allowed to delete the LUN group. Snapshots that belong to a group cannot be
deleted individually. You must delete all the snapshots in a snap group.
Note: If a LUN is deleted, it is automatically removed from any LUN groups of which it is a
member. If deleting a LUN results in a LUN group with zero members, then that LUN group is
automatically deleted.
198
535-0008-00 Rev 10
the group. If you delete all LUNs from a LUN group, the LUN group is automatically deleted.
Note: Renaming a LUN group with exported LUNs is not recommended, as traffic can be
disrupted. For more information, see Guidelines for Exporting LUNs on page 194.
# lungroup rename container <container-id> name <old-lungroup-name> to
<new-lungroup-name>
535-0008-00 Rev 10
199
# show lungroups
Container: MyContainer
Container: 41238F00608
LUN Group
LUNs
---------------------------------------------------------------------lung1
lun1,lun2,lun3
200
535-0008-00 Rev 10
CHAPTER 5
vSHARE Management
201
The session and detail options are particularly useful for understanding target port activity and
status.
The sessions option displays the number of SCSI sessions active on a target and the initiator and
the assigned LUN ID for each session.
SM075 [MY-vSHARE00: master] (config) # show targets protocol fc node 3
hostname SM076 sessions
Node Hostname
Protocol Target
Enab Address
---- --------------- -------- ---------- ---- -----------------------3
SM076
fc
hba-a1
yes
wwn.21:00:00:1b:32:8f:24:5d
Connected Sessions
SM076
fc
hba-a2
Connected Sessions
: wwn.21:00:00:1b:32:82:1e:50
: wwn.21:00:00:1b:32:94:13:32
: wwn.21:01:00:1b:32:a2:1e:50
: wwn.21:01:00:1b:32:af:24:5d
: wwn.21:01:00:1b:32:b4:13:32
: wwn.21:fd:00:05:1e:09:ed:48
yes
wwn.21:01:00:1b:32:af:24:5d
:
:
:
:
:
wwn.21:00:00:1b:32:82:1e:50
wwn.21:00:00:1b:32:94:13:32
wwn.21:01:00:1b:32:a2:1e:50
wwn.21:01:00:1b:32:b4:13:32
wwn.21:fd:00:05:1e:09:ed:48
The detail option returns the network bindings for each target.
SM075 [MY-vSHARE00: master] (config) # show targets protocol fc node 3 hostname
SM076 detail
Target: hba-a1
Node
: 2
Hostname
: qa-stein136-mga
Protocol
: fc
Address
: wwn.21:00:00:24:ff:38:54:b2
Enabled
: yes
State
: Online
Port Type Setting
:
Port Type
: NPort (fabric via point-to-point)
Speed Setting
:
Speed
: 8 Gbit
For detailed information about this command, see Show Targets Command on page 216.
Viewing Interface Status
show interfaces [<ifname>] [configured | brief]
The show interfaces command may be used to view information about a specific interface or,
if one is not named, all interfaces. By default, detailed information about the interface and its runtime
202
535-0008-00 Rev 10
state is given. The configured and brief options enable the user to specify which information
is displayed in the CLI:
535-0008-00 Rev 10
203
The show media command may be defined by the following options, which may be used in various
combinations:
all
block
detail
freelist
Display media allocation freelist information for all devices including flash
memory devices not allocated or formatted.
global
health
Display status information and lifetime remaining for media devices. For
more information on the health option, see Show Media Health
Commands on page 209.
id
Display information for a specified media device or all media (id all).
The following show commands described in this section can be used in Configure or Enable mode.
All other media management commands require Configure mode.
show media
show media all
show media global
show media block id all
show media block id <id>
show media block detail
show media
The show media command returns all media devices that can be used as block storage on the
current Memory Gateway (node or module). This command shows a summary line for each device
location, giving the size and status of that device.
Location
Media ID
Model
Size
Status
--------------------------------------------------------------------unknown-00
Fender1002
6895.75G
online
unknown-01
Strad130
432.00G
online
The show media command is an alias for show media block id all command. Media used
by the system for purposes other than block storage are not shown.
204
535-0008-00 Rev 10
The show media all command displays all media devices installed in a Memory Gateway. This
command shows a summary line for each device location, giving the size and status (online/offline)
of the installed media device.
Location
Media ID
Model
Size
Status
---------------------------------------------------------------------Boot Disk
---
---
PCI 4
41148F00147
74.53G
online
2671.00G
online
The Media ID and Model values are displayed only for media devices that are usable as block
storage; otherwise, a string of three hyphens (---) is displayed instead.
show media global
To view information about storage media devices on a cluster, connect to the master node of the
cluster and add the global option to the show media commands.
SM075 [MY-vSHARE00: master] # show media global
Module 2: SM075 (10.10.0.37)
Location
Media ID
Model
Size
Status
---------------------------------------------------------------------unknown-00
Fender1002
6895.75G
online
unknown-01
Strad130
432.00G
online
Media ID
Model
Size
Status
---------------------------------------------------------------------PCI 2
Strad134
399.75G
online
The show media all global command returns information about all media devices on the
cluster including non-storage devices.
535-0008-00 Rev 10
205
The show media all block id all command displays the location, media ID, model, size,
and status (online/offline) of all block storage media.
SM075 [MY-vSHARE00: master] # show media block id all
Location
Media ID
Model
Size
Status
---------------------------------------------------------------------unknown-00
Fender1002
Violin Memory Array
6895.75G
online
unknown-01
Strad130
Violin Memory Array
432.00G
online
Information about a particular block storage media device can be displayed by identifying the device
with the block id <id> option of the show media command:
SM075 [MY-vSHARE00: master] # show media block id Strad130
Location
Media ID
Model
Size
Status
---------------------------------------------------------------------unknown-01
Strad130
Violin Memory Array
432.00G
online
The <id> of specific media devices can be found in the output of the show media all or show
media block id all commands.
show media block detail
The detail option of the show media command returns detailed information about storage
media. The detail option may be combined with any other show media command option in any
order.
SM076 [MY-vSHARE00: standby] (config) # show media block id all detail
Location PCI 4:
206
Status:
online
Size:
2867964411904
Type:
block device
Firmware version:
6.0.0
Manufacturer model:
VIOLIN_MEMORY_ARRAY
41148F00147
Device ID:
ata-VIOLIN_MEMORY_ARRAY_41148F00147
Violin model:
535-0008-00 Rev 10
41148F00147
V6000
Violin Revision:
Media Block Enable Commands
Before a media device can be used for block storage, it must be enabled. This section describes
commands for enabling and disabling block storage media devices on a vSHARE Memory
Gateway.
The enable commands described in this section can be used in Configure mode only. For more
information, see Command Modes on page 225.
Commands include:
The media block id <id> enable command enables a specified block storage media device.
The block option specifies a vSHARE block media device.
In the following example, the media block id <id> enable command is used to enable a
single block storage media device:
* gateway02 [cluster1: master] (config) #
all
Change block media
<id>
VTMS0101-2
VTMS0101-4
* gateway02 [cluster1: master] (config) #
enable
* gateway02 [cluster1: master] (config) #
media block id ?
state for all block devices
To enable all block storage media devices on a local Memory Gateway, use the all option as
described in media block id all enable on page 208.
535-0008-00 Rev 10
207
To disable a media device, use the no form of the media block id <id> enable command.
* gateway02 [cluster1: master] (config) # no media block id VTMS0101-2
enable
Warning: this command will cause users of this media to reset and lose
all content.
Confirm command? [no] yes
* gateway02 [cluster1: master] (config) #
Caution: When a block media device is disabled, all applications using that device immediately stop using
it, and will have space from other media allocated for them. This may be a disruptive operation.
To disable all block storage media devices on a local Memory Gateway, use the all option as
described in no media block id all enable on page 208.
media block id all enable
The media block id all enable command enables all block storage media devices on the
local Memory Gateway.
In the following example, the media block id all enable command enables all storage media
devices on the local Memory Gateway:
SM076 [MY-vSHARE00: standby] (config) # media block id all enable
Enabling 23108R00000043
Enabling Strad134
no media block id all enable
To disable the use of all media devices in the Memory Gateway, use the no form of the above
command in Configure mode.
Caution: When all media devices on a Memory Gateway are disabled, applications using those devices
immediately stop using them and will have space allocated for them on other Memory Gateways. This may be
a disruptive operation.
208
535-0008-00 Rev 10
In the following example, the no media block id all enable command disables all block
storage media devices on the local Memory Gateway:
* gateway02 [cluster1: master] (config) # no media block id all enable
Warning: this command will cause users of this media to reset and lose
all content.
Confirm command? [no] yes
Disabling VTMS0101-2
Disabling VTMS0101-4
* gateway02 [cluster1: master] (config) #
Show Media Health Commands
The health option of the show media command returns information about the status and
expected lifetime of media devices. The option can be included with other options in various forms
in any order.
The show media health commands display the location, media ID, model, life remaining, and
status (or health) of media devices on the local Memory Gateway or on all of the Memory Gateways
in the cluster. The commands can be used in Enable and Configure modes.
Understanding Media Health Status
The device is running correctly and has plenty of expected life remaining.
unknown
warning
critical
Status is determined for each individual S.M.A.R.T attribute by comparing its current normalized
value to a manufacturer-specified usage threshold. The detail option lists all of the S.M.A.R.T
attributes and gives the status of each attribute. Without the detail option, show media health
gives the worst-case status among all the devices attributes.
Normalized values begin at their maximum (usually 100 or 254) and decrease with use of the
device, so they approach their specified thresholds from above. A warning status indicates that the
value is close to its threshold, and a critical status indicates that it has reached or passed the
threshold.
Note: Warning and critical status do not necessarily indicate any current problems; they only
show that devices are near the manufacturers thresholds for expected lifetime performance. When
a device reaches warning status, its detailed status should be monitored carefully.
535-0008-00 Rev 10
209
Caution: Critical status indicates that failures are likely in the near future. It is recommended that devices
that reach critical status be disabled and replaced before they fail.
The show media health command returns information about the health of storage media
devices on the local Memory Gateway.
SM075 [MY-vSHARE00: master] # show media health
Location
Media Id
Model
Life remain
Health
---------------------------------------------------------------------PCI 4
Violin123
Violin Memory Array
100.0%
ok
To view information about storage media on other nodes in the cluster, use the global option.
The show media health command returns information about the block storage media only. To
include non-storage media devices, use the show media all health command.
show media health detail
The show media health detail command returns detailed information about block storage devices
on the local Memory Gateway including all of the S.M.A.R.T attributes and provides the status of
each attribute.
In the following example, the block id <id> option is used to identify a specific media device.
SM075 [MY-vSHARE00: master] # show media health block id Strad130
detail
Location unknown-01:
Model:
Violin Memory Array
Est. life remaining:
96.0%
Serial no: Strad130
Health status:
ok
SMART Attribute
Type
Raw value
Norm Thres
Status
------------------------------------------------------------------Reallocated Sector Ct
Old age
0
100
0
ok
Power On Hours
Old age
0
100
0
ok
Power Cycle Count
Old age
0
100
0
ok
Unsafe Shutdown Count
Old age
0
100
0
ok
Temperature Celsius
Old age
54
54
0
ok
Host Write Count
Old age
0
200
0
ok
Avail Reserved Space
Old age
100
100
10
ok
Media Read and Write Statistics
The show stats media commands enable you to view read and write rates (in MB per second)
for both the current ten-second sample and an average of the samples taken over the last five
minutes.
210
535-0008-00 Rev 10
The show stats media command displays read and write statistics for block storage media
devices in one or more nodes in the cluster. The scope of each command is the same as the
corresponding show media command, described above.
show stats media
The structure of the show stats media commands is identical to the show media commands
described above, except that these commands also support a continuous option to display a
continually updating view of the statistics.
SM075 [MY-vSHARE00: master] # show stats media
Media Stats Summary (sampled @10 secs)
Read (bytes/s)
Write (bytes/s)
Read Latency (us)
Write
Latency (us)
Media Current
Last 5m Current
Last 5m
Current
Last 5m Current
Last 5m
----------------------------------------------------------------------------Violin123
0
21k
0
31k
112us
121us
74us
69us
If the continuous option is specified, the display of statistics is updated continually. Use CTRL+
C to exit the display.
The command show stats media is an alias for show stats media block id all. The
command returns information about storage media devices on the local Memory Gateway and
omits all non-block storage media.
Managing Containers
The show containers command returns information about the storage containers on a Memory
Gateway cluster including the node, container name, the number of LUNs, the space allocated, and
the space free.
HR-mga [cluster: master] # show containers
Container LUNs Total Free Prov Ports HostA HostB Status Encrypted
--------------------------------------------------------------------HR2013
30
10356G 9745G 459G A,B
HR-mga HR-mgb
HA
No
535-0008-00 Rev 10
211
Full CopyEnables the disk arrays to make deep copies of data without requiring ESX to
access the data.
Block ZeroingAllows ESX to send commands to zero blocks on the array, avoiding the need
to send WRITE commands containing zeros to the array.
Note: By default VAAI is enabled on ESX. To disable VAAI, see the VMware knowledge base
212
535-0008-00 Rev 10
Note: After you create an export for a specific iSCSI initiator using a specific target, the initiator
is excluded from seeing LUN exports from that target to all initiators.
Each initiator is a Fibre Channel WWN, iSCSI IQN or InfiniBand GUID address.
535-0008-00 Rev 10
213
Note: You can create up to three hierarchy levels within a nested igroup.
Use a unique LUN ID (LUN name) for each LUN to avoid the chance of conflicts.
Refrain from exporting a LUN to different targets using a different LUN ID (LUN name) for the
LUN for each target, as this practice can result in conflicts.
Plan the nested igroups hierarchy first. Then export the LUNs, allowing the system to autoassign LUN IDs to ensure they are each unique.
Managing LUNs
This section covers the following topics:
214
535-0008-00 Rev 10
The show luns session command returns information about the sessions connected to LUNs
including the target port and the session ID. Use the name option to identify a specific LUN.
# show luns name Fender1002 sessions
Container: HA-1
LUN
Size RW Bksz Status
Ports
------------------------------------------------------------------Fender1002
1024M rw 512 Single
A
Connected Sessions
1
1
1
1
1
1
1
1
1
1
1
1
1
LUN ID: 2
: Port: Procurve, Session: iqn.1994-05.com.redhat:Bow7_1,
LUN ID: 2
: Port: Procurve, Session: iqn.1994-05.com.redhat:Bow7_2,
LUN ID: 2
: Port: Procurve, Session: iqn.1994-05.com.redhat:Bow7_3,
LUN ID: 2
To view a list of containers, enter show luns container ?.
SM075 [MY-vSHARE00: master] (config) # show luns container ?
Violin123
The name option enables you to view only those LUNs which are prefixed by a particular name.
LUN Create Commands
[no] lun create container <container_name> name <LUN_name> size [<size GB>
| equal] [quantity <number>] [nozero] [readonly] [startnum <unsigned
integer>] [blocksize 512 | 4096] [offline]
The lun create command enables you to create a LUN within a specified storage container.
LUN Set Commands
[no] lun set container <id> name <name> readonly
535-0008-00 Rev 10
215
The lun set command enables you to set a LUN as read-only. The lun set command is the
only command usable after a LUN has been created except the no lun create readonly
command.
LUN Statistics Command
LUN statistics are accessed via the CLI with the following command:
show stats lun [container <container_id>] [name <lun_name>] [on global|on
host <hostname>] [continuous] [detail]
Use the container <container_id> option to restrict statistics display to a single container. If
it is omitted, statistics are shown for all the containers accessible to the hosts.
Use the name <lun_name> option to get statistics for a single LUN. If it is committed, statistics for
all LUNs are specified.
Note: Select shell wildcard patterns can be used in the <lun_name> field to match a set of LUNs.
For example, you can use the asterisk (*) to match all characters that follow. Such as xyz* to match
all LUN names that begin with xyz. Or, you can specify a range of LUN names using the [a-e]
wildcard. Such as xyz-lun[2-4][0-9] to match all LUNs from xyz-lun20 to xyz-lun49.
Use the on global option to get statistics for data flowing to LUNs through all nodes in a cluster.
Use the on host <hostname> option to get statistics for data flowing to LUNs on a specific host.
Use the continuous option to refresh displayed statistics every ten seconds.
Use the detail option to display statistics in a verbose format.
Managing Targets
This section covers the show targets command.
Show Targets Command
show targets [node <cluster node id>] [hostname <hostname>] [protocol <fc/
iscsi>] [id <target id>] [sessions] [detail]
The show targets command displays all Fibre Channel or iSCSI targets. Use the hostname,
protocol, and id parameters to filter the targets returned.
The command returns the node, hostname, target port, status (enabled or not), and address (WWN
or IQN). The detail parameter returns the network bindings for each target.
216
535-0008-00 Rev 10
Section
Manage
Monitor
If a system begins shutdown or if a maximum number of emails are staged (default 5), they are
sent out automatically.
Emails are consolidated on a per-host basis. Emails from different hosts (acm-a, acm-b, mg-a,
or mg-b) cannot be combined.
Log in to the ACM master or MG master node to configure email consolidation. The default is
disabled.
217
218
535-0008-00 Rev 10
535-0008-00 Rev 10
219
220
535-0008-00 Rev 10
CHAPTER 6
This chapter provides detailed procedures for configuring and showing status of the
Memory Array and its internal modules. Most of these procedures can only be
performed using the Command Line Interface (CLI). If the tasks in this chapter can
also be performed using the Violin Web interface, references to those sections are
provided.
The following sections are included in this chapter:
See APPENDIX A, CLI Reference on page 287 for details on using the CLI
commands.
535-0008-00 Rev 10
221
show clock
sh h
show hosts
Additional letters can be included but none can be skipped; for example, the show clock
command can be typed as sho clo or sh clock or various other combinations, but not shw clk.
Other commands that are frequently typed in shorthand include:
en
enable
conf t
configure terminal
ex
exit
If the command is shortened too much, an error message appears and help is offered. For example,
the abbreviation sh cl could mean show clock or show cluster so it generates this error
message:
> sh cl
% Ambiguous command "cl".
Type "sh cl?" for help.
Getting Help
In any mode of the CLI you can query for help by using the help command or typing a question
mark. See Command Modes on page 225 for descriptions of the three CLI modes that are available.
The mode determines the set of commands that can be executed.
Enter help at the prompt for a summary of how to use question marks to obtain context-sensitive
help, as described here. Just entering a question mark ? by itself provides a list of available
commands corresponding to the current mode.
You can also query for options of a specific command by typing in the command, following it with a
space, and adding a question mark. After displaying a list of options, the command line echoes the
string and puts the cursor after it, ready for more input.
222
535-0008-00 Rev 10
For example, in Standard mode you can enter cli ? at the command line and see the following
output.
> cli ?
clear-history
session
> cli
If the command is complete without further options or values, <cr> (carriage return) is displayed
on a separate line and the command is echoed at the prompt. Pressing the Enter key (also known
as carriage return, <cr>) will then issue the command if no values are required, or <value
required> will be displayed.
For example:
> cli session ?
auto-logout Configure keyboard inactivity timeout for automatic logout
paging Configure the ability to view text one screen at a time
prefix-modes Configure the CLI's prefix modes feature for this session
progress Configure progress updates for lon operations
terminal et terminal parameters
> cli session paging ?
enable Enable paging
> cli session paging enable ?
<cr>
> cli session paging enable
>
> cli session terminal ?
length Set the number of lines for this terminal
resize Resize the CLI terminal settings (to match with real terminal)
type Set the terminal type
width Set the width of this terminal in characters
> cli session terminal width ?
535-0008-00 Rev 10
223
<number of characters>
> cli session terminal width 60
>
When <value required> is displayed and only specific values can be used (such as interface
names or a VLAN identifier), those values will be displayed on new lines after <value required>.
Similarly, when the command is complete but could include additional options, <cr> and the
options are displayed, each on a separate line.
sh<tab>
Completes the show keyword (but s<tab> does not, because more than
one command starts with s).
show <tab>
Lists options of the show command that can immediately follow the show
keyword.
sh<tab><tab>
Completes the show keyword and lists options of the show command.
s<tab><tab>
For a list of all commands currently available, press the Tab key twice at the prompt.
In Standard mode, for example, press the Tab key twice to list these commands:
> <tab><tab>
224
cli
help
show
enable
no
slogin
exit
ping
telnet
terminal
535-0008-00 Rev 10
Command Modes
The CLI can be in one of three modes, which determine the set of commands that can be executed.
Commands that are not currently available do not show in help or completion, and generally behave
as if they do not exist.
Standard Mode
When the CLI is launched, it begins in Standard mode. This is the most restrictive mode and only
has commands to query a restricted set of state information. In this mode you cannot take any
actions that would directly affect the system, nor can you change any configuration.
User accounts with the unpriv role are restricted to Standard mode.
Enable Mode
The enable command moves the CLI to Enable mode. This mode has commands to view all state
information and take certain kinds of actions, such as rebooting the system or configuring some
system parameters, but it excludes commands that configure the cluster. Its commands are a
superset of those in Standard mode. The disable command moves the CLI from Enable mode
back to Standard mode.
The exit command (in Enable mode) closes the CLI.
User accounts with the monitor role can use all Enable mode commands.
Config Mode
The configure terminal command moves the CLI from Enable mode to Config mode.
On the clusters master node, Config mode has a full unrestricted set of commands to view
anything, take any action, or change any configuration. Its commands are a superset of those
in Enable mode.
On nodes other than the master, Config mode only includes commands that operate on the
local node. Using a global command on a standby or normal node either has a temporary local
effect (which is overridden as soon as the node synchronizes with the master node) or
produces an error message, which identifies the master node where the command can be
used.
The exit command moves the CLI from Config mode to Enable mode. Using the exit command
twice closes the CLI, or you can use the quit command to close the CLI directly.
To return to Standard mode from Config mode, first exit Config mode and then disable the Enable
mode.
User accounts with the admin role can use all Config mode commands.
535-0008-00 Rev 10
225
Enable mode:
Config mode:
Configuration commands that succeed in doing what was asked do not print any response, so the
next thing you see after pressing <Enter> is another command prompt. You can verify the effect
of a configuration command by using its corresponding show command to display current settings.
If an error occurs in executing a command, the response begins with % followed by some text
describing the error.
226
bytes
kB
MB
GB
535-0008-00 Rev 10
TB
PB
A
B
A
B
n
D
n
C
n
B
n
A
A
B
ID
A
B
Console Port B
Console Port A
Array.
2. Connect the other end of the cable to a computer.
3. When you are ready to connect to the CLI, start a terminal application, and then use the
535-0008-00 Rev 10
227
4. Log in to the Master ACM as admin. The Master ACM is set up during initial configuration of
the Memory Array. Enter the system host name or the ACM management IP address to
ensure a direct connection.
PuTTY.
The New Connection - PuTTY1 window opens.
3. In the Connection Description window:
a. Type the name of the new connection in the Name field.
b. Select the appropriate icon from the list of available ones.
c. Click OK when done.
You are now connected to the Master ACM of the Memory Array and can access the various
CLI commands. Closing the PuTTY window terminates the connection.
.kermrc file in your home directory so that the Kermit application recognizes
it; otherwise, use the -y option followed by the location and file name of the configuration file.
To connect to the CLI through the serial connection:
1. Log into the host computer as root.
2. At the host computers command line prompt, type:
# kermit
228
535-0008-00 Rev 10
or
# vtkermit
3. Specify the line of the Kermit connection. At the Kermit prompt, type:
C-Kermit> set line /dev/ttyS0
Note: The device name may vary depending on the host and the Linux operating system that is
system.
9. At this point, you are connected to the CLI of the Memory Array and can access the various
535-0008-00 Rev 10
229
configuration write
write memory
An asterisk (*) before the command prompt is a reminder that changes were made but have not yet
been saved; therefore after saving the configuration, the asterisk no longer appears.
Example of configuration write command:
Either command (or the Commit Changes button) stores all current configuration parameters to
persistent storage in the active configuration file.
Web: In the Web Interface, the Commit Changes button appears in the message bar only when
the current configuration includes changes that have not yet been saved; otherwise the
message bar displays a No unsaved changes message.
You can switch to another configuration file by saving the current configuration to a specified
filename.
This option is at the bottom of the Cluster Administration screen. First enter a filename, then
click the SAVE button.
For example, the following command saves the configuration to a file named cf1 and makes it the
active configuration file:
230
535-0008-00 Rev 10
If you want to save the current configuration to a different file without making it active, use the noswitch option of the configuration write to command. This option can be used to save
an interim configuration which you will finish later. Saving with the no-switch option does not
change a configurations unsaved status (indicated by an asterisk in the CLI), because recent
changes are still not stored in the active configuration file.
For example:
The display shows all configuration files associated with this system and identifies the active file,
which is backed up as <filename>.bak. Backups of previously active configuration files also appear
in the display.
For example, after switching from the original configuration file (initial) to cf1 and saving to cf2
with the no-switch option, as shown above, the display lists these five configuration files:
Showing Configuration
The CLI commands that would bring the state of a fresh system up to match a given configuration
can be displayed by using one of the following commands:
show configuration - display commands to reproduce the state saved in the active
configuration file.
show configuration running - display commands to reproduce the state of the current
running configuration, which could include some unsaved configuration settings.
535-0008-00 Rev 10
231
show configuration full - display all commands, including any default settings, to
reproduce the state saved in the active configuration file.
show configuration running full - display all commands, including any default
settings, to reproduce the state of the current running configuration.
The output of each of these commands begins with a short header that contains the name and
version number of the configuration, in a comment. Except for the show configuration full
and show configuration running full commands, the output omits commands that are not
required because they set parameters to their default values.
Web: In the Violin Web Interface, you can select any available configuration file in the Cluster
Administration screen and use the APPLY button to apply its configuration to the system.
An active configuration file may not be deleted or renamed, nor may it be the target of a move or
copy. It may be the source of a copy, in which case the original remains active.
Web: In the Current Configurations section of the Cluster Administration screen, you can select
one or more configuration files to delete. First click the check box for each file you want to
delete, then use the DELETE button.
232
535-0008-00 Rev 10
This section provides the basic steps required to upgrade the Memory Array firmware and Memory
Gateway software using the CLI.
Before attempting to upgrade the system, always review the latest Release Notes for the most
up-to-date instructions, information on the necessary software image and required upgrade
paths.
To upgrade Memory Array firmware or Memory Gateway software using the Web interface,
see Upgrading the Violin Cluster on page 138.
Note: You must register as a Violin Memory customer to download software updates. If not
WARNING! Make sure that all alarms are cleared before starting an upgrade process.
535-0008-00 Rev 10
233
the Master Memory Gateway node: the left USB port is for MG A; the right USB port is for
MG B.
3. Start a terminal application and log in to the Master Memory Gateway as admin.
4. At the command prompt, type enable and then configure terminal:
image file:
a. To upgrade the image from a USB drive, type the following commands to mount the
Note: The upgrade process may take from 20 minutes to 1 hour to complete. Do not power off
Once all modules are updated with the new software, both Memory Gateways shut down. The
ACMs then reboot. The VCMs and VIMMs are updated, if necessary. The Memory Gateways then
boot up with the new version of software.
To monitor the upgrade process:
Wait 1-2 minutes for the system to reboot and then log in to the Master ACM to monitor the upgrade
process, if desired.
234
Once the VCM upgrade process is completed, type vdiag to show the VIMM upgrade
progress.
535-0008-00 Rev 10
Note: Only the Memory Gateway software is upgraded using these instructions. Ensure that the
Memory Array is running compatible versions of Memory Gateway software and Memory Array
firmware. See the latest Release Notes or contact Violin Memory Customer Support for more
information.
There are two methods available for upgrading the Memory Gateways: staged (non-disruptive) and
immediate. Both are described in this section.
Skip to step 3 if you are not upgrading the software from a USB drive.
1. If upgrading from a USB drive, make sure that the software image is loaded on the USB drive.
2. Insert the USB drive into the USB port on the front of the Memory Array that is associated with
the Master Memory Gateway node: the left USB port is for MG A; the right USB port is for
MG B.
3. Start a terminal application and log in to the Master Memory Gateway as admin.
4. At the command prompt, type enable and then configure terminal:
For a staged upgrade, the cluster is split into two clusters, A and B, with one Memory
Gateway in cluster A and the other in Cluster B. The Memory Gateway in Cluster B is
upgraded first, followed by Cluster A. While Cluster B is being upgraded, Cluster A provides
access to storage, and vice-versa.
For an immediate upgrade, the Memory Gateways are upgraded simultaneously. During this
process clients are unable to maintain connections to exported LUNs until the cluster is
restarted. Although this method of upgrading the cluster is relatively quick, you experience
some downtime while the nodes reboot.
Both Memory Gateways will upgrade and then reboot on their own. Once the Gateways reboot, the
command prompt returns to the ACM Master.
535-0008-00 Rev 10
235
6. Wait 1-2 minutes and then type show array modules type mg detail to see if the
Gateways have booted. The Gateways are powered on if their IP addresses are shown in the
output.
up during initial configuration of the Memory Array. Enter the system host name or the ACM
management IP address to ensure a direct connection.
2. At the command prompt, type enable and then configure terminal:
The command prompt returns to the ACM Master once the Memory Gateways shut down.
7. From the ACM Master command prompt, type the following to verify that the Memory
236
535-0008-00 Rev 10
The Memory Gateways power on automatically after the Memory Array firmware is upgraded.
To upgrade the Memory Array firmware:
Note: Make sure the Memory Gateways are shut down and powered off before continuing.
1. If upgrading from a USB drive, make sure the new image file is loaded on the USB drive, as
up during initial configuration of the Memory Array. Enter the system host name or the ACM
management IP address to ensure a direct connection.
3. Log in to the Memory Array as admin.
4. At the command prompt, type enable and then configure terminal:
Memory Array that is associated with the Master node: the left USB port is ACM A; the right
USB port is ACM B. The LED next to the USB port will start flashing.
7. Perform an upgrade using one of the following methods, depending on the image location:
a. To upgrade from a USB drive, type the following to mount the USB drive and then install
the image:
(config) # usb mount
(config) # cluster upgrade usb:// <path> immediate
b. To upgrade from a URL accessible to the system, type the following, where <URL> is
535-0008-00 Rev 10
237
8. If upgrading from a USB drive, log back in to the same ACM from which the upgrade was done
and then type the following at the command prompt to eject the USB drive:
violin-acma [violin: master] (config) # usb eject
Note: The upgrade process may take from 20 minutes to 1 hour to complete. Do not power off
or reboot the system during the upgrade process. Type monitor upgrades at the command
prompt to monitor the VCM upgrade process. Once the VCM upgrade process is completed, type
vdiag to show the VIMM upgrade progress. The Memory Gateways power on when the VCMs are
ready.
An asterisk (*) before the command prompt is a reminder that changes were made but have not yet
been saved; therefore after saving the configuration, the asterisk no longer appears.
See Configuration File Management on page 230 for more information.
238
535-0008-00 Rev 10
page 236.
5. Once the Memory Gateways are shut down, return to the Master ACM and then type the
The following modules are rebooted: Array Controller Modules, vRAID Controller Modules, internal
Memory Gateways, VIMMs and Host Bus Adapters.
Depending on the number (24 or 64) and type of VIMMs (SLC or MLC), the reboot process will take
at least one hour for an SLC system and as long as two hours for an MLC system.
Type the following command to view reboot progress:
violin-acma [violin: master] (config) # monitor boot
page 236.
5. Once the Memory Gateways are shut down, return to the Master ACM and then type the
535-0008-00 Rev 10
239
To power off all types of a Memory Array module, type the following at the command prompt:
violin-acma [violin: master] (config) # no array modules type <vcm | acm | fpm |
mg | hba> enable
Where:
vcm
is the vRAID Controller Modules in the Memory Array. This powers off all VCMs and
VIMMs.
acm
fpm
mg
is the internal Memory Gateways in the Memory Array. This powers off all Memory
Gateways.
hba
is the Host Bus Adapters in the Memory Array. This powers off all HBAs.
To power on a specific type of Memory Array module, type the following at the command
prompt:
violin-acma [violin: master] (config) # array modules type <vcm | acm | fpm | mg
| hba> enable
240
vimmxx
vcm-x
mg-x
hba-x
535-0008-00 Rev 10
the flag has been set for a VIMM by the system, you will see output similar to the following:
# show donotuse
vimmid system VIMM is not responding at 11-01-2012 20:41:11 UTC
If the administrator set the flag, by issuing the no array modules id vimmid command,
the output would be:
vimmid user-set VIMM is not responding at 11-01-2012 20:41:11 UTC
2. Type one of the following at the command prompt:
a. To clear the flag for a VIMM, where id is the VIMM number:
power-on sequence.
Note: If the VIMM or VCM has failed and requires replacement, see the Violin 6000 Series
Memory Array Service Guide for instructions on replacing Memory Array modules.
535-0008-00 Rev 10
241
Caution: The Memory Array must be completely shut down before proceeding.
1. Unplug the AC power cords from the power receptacles on the rear of the Memory Array
chassis.
2. Wait 30 seconds and then reconnect the power cords to their power receptacles.
The Memory Array begins to start up. During the boot up process, the green Status LEDs will flash.
The Memory Array is fully booted when the Status LED is continuously lit.
Note: If one of the balance functions is enabled (automatic or scheduled), it must be disabled
Command
Description
array balance
This starts a RAID rebuild for each VCM that requires a rebuild.
More than one rebuild may be required to balance the system.
242
535-0008-00 Rev 10
Command
Description
535-0008-00 Rev 10
243
up during initial configuration of the Memory Array. Enter the system host name or the ACM
management IP address to ensure a direct connection.
2. At the command prompt, type enable and then configure terminal:
up during initial configuration of the Memory Array. Enter the system host name or the ACM
management IP address to ensure a direct connection.
2. At the command prompt, type enable and then configure terminal:
244
535-0008-00 Rev 10
up during initial configuration of the Memory Array. Enter the system host name or the ACM
management IP address to ensure a direct connection.
2. At the command prompt, type enable and then configure terminal:
up during initial configuration of the Memory Array. Enter the system host name or the ACM
management IP address to ensure a direct connection.
2. Type the following at the command prompt:
535-0008-00 Rev 10
:
:
:
:
balance in progress
balanced
balanced
unbalanced
245
246
535-0008-00 Rev 10
CHAPTER 7
Data Security
This chapter covers available options for enhancing data security and includes the
following topics:
535-0008-00 Rev 10
247
Note: When you enable CHAP on a Memory Gateway (target) for one client (initiator), the rest of
the clients (for that target) must also use CHAP. Clients have to enable CHAP (from their side)
before they are able to log in to the target.
248
535-0008-00 Rev 10
CHAP Requirements
iSCSI CHAP configurations must meet the following requirements:
The minimum CHAP secret size is 12 (bytes) characters (IPSec is not supported).
For bidirectional CHAP, each target-initiator pair must have one target CHAP secret and one
initiator CHAP secret configured.
In bidirectional CHAP mode, the target CHAP secret cannot be identical to the initiators CHAP
secret to which it is connected.
A particular target can have bidirectional CHAP set up with multiple clients.
Note: A CHAP secret must have a minimum of 12 characters (bytes). The maximum length is 64
characters.
Use the CLI command {target-secret | initiator-secret} <CHAP secret word>, for example:
iscsi chap target iscsi target-secret Myfirstsecret to initiator
iqn.1991-05.com.microsoft:violin-w2k3r2
Enter the secret at the prompt at the CHAP secret prompt using {target-secret-prompt |
initiator-secret-prompt} format. In this case the character input is hidden, as shown in the
following example:
iscsi chap target iscsi initiator-secret-prompt to initiator
iqn.1991-05.com.microsoft:violin-w2k3r2
CHAP secret: ***************
535-0008-00 Rev 10
249
> enable
# configure terminal
(config) #
3. Establish the target CHAP secret for a specified initiator using the following command,
substituting the variables with information for your site. See CHAP Requirements on page 249
for CHAP secret recommendations and restrictions.
# iscsi chap target <target_name> {target-secret|target-secretprompt} <CHAP_secret_word> to {<initiator_name>|<igroup_name>}
Note: The igroup_name variable option, as used in this command, allows you to establish the
CHAP secret for a group of initiators all at once. However, when you view iSCSI CHAP
configurations, as described in Showing iSCSI CHAP Configurations on page 252, the individual
initiators that are members of the igroup are shown individually without reference to the
igroup_name (they belong to).
> enable
# configure terminal
(config) #
3. Establish a target (Memory Gateway) CHAP secret, as described in Configuring Unidirectional
250
535-0008-00 Rev 10
4. Establish the initiator CHAP secret using the following command. Substitute target variables
with the values used to establish the target CHAP secret, and initiator variables with the
appropriate initiator information. See CHAP Requirements on page 249 for CHAP secret
recommendations and restrictions.
# iscsi chap target <target_name> {initiator-secret|initiatorsecret-prompt} <CHAP_secret_word> to
{<initiator_name>|<igroup_name>}
Note: The igroup_name variable option, as used in this command, allows you to establish the
CHAP secret for a group of initiators all at once. However, when you view iSCSI CHAP
configurations, as described in Showing iSCSI CHAP Configurations on page 252, the individual
initiators that are members of the igroup are shown individually without reference to the
igroup_name (they belong to).
> enable
# configure terminal
(config) #
3. Remove the initiator CHAP secret for a target and initiator pair using the following command.
Substitute target and initiator variables with the values appropriate for each device.
# no iscsi chap target <target_name> to
{<initiator_name>|<igroup_name}
535-0008-00 Rev 10
251
252
535-0008-00 Rev 10
CHAPTER 8
Data Management
This chapter explains the features and functionality of vMOS Data Management,
and covers the following topics:
535-0008-00 Rev 10
253
LUN Group
A LUN Group is a set of LUNs upon which you can perform actions. All the LUNs used by a specific
application can be grouped together into a LUN group. All LUNs in a LUN group must belong to the
same array. Actions to a LUN group are applied to all LUNs that are members of the LUN group.
You can have a maximum of eight LUNs in a LUN group. For more information on LUN groups, see
Working with LUN Groups on page 197.
Protected Snapshots
Protected snapshots allow you to specify that a snapshot cannot be automatically deleted by the
system during space reclamation. An Admin user can manually delete a protected snapshot at any
time, but the system cannot automatically delete the snapshot to reclaim space.
Snapshot
A snapshot is a space efficient (exact) replica of the state of a LUN or group of LUNs at a particular
point in time. Snapshots are read-only by default, but can be made read-write if necessary.
Snap Group
A snap group is an object that represents snapshots taken across a collection of LUNs belonging
to a LUN group. To create a snap group, snapshots of each LUN in a LUN group are taken, then
grouped as a snap group.
Thin Provisioning
Thin Provisioning is a method of optimizing the utilization of available storage through the use of
on-demand allocation of blocks of data instead of allocating all the blocks of data up front. This
means that the physical storage space is only allocated when data is written to the LUN.
254
535-0008-00 Rev 10
Over Provisioning
Over provisioning is the practice of provisioning space (for thin LUNs) beyond the physical capacity
of the array. In this case, allocation of the physical space occurs as the data is a written to the LUNs.
Decision support analysis and reportingOperations that do not need real-time information
can use a replica database created from snapshots. This approach enhances performance by
eliminating unnecessary competition for access to the primary database.
Testing and trainingDevelopment and service groups can test data for applications. This
practice provides a realistic basis for testing robustness, integrity, and performance of new
applications at no impact to the production application.
Database error recoveryYou can quickly recover a database using a snapshot, and rollback
to an earlier instance of the data. This practice ensures the integrity of a database in the case
of an admin or application program error that compromised the database.
Example Scenarios
Daily backups for analytics, reports, and data protectionA crash-consistent point-in-time
snapshot of the database used primarily for read-only purposes, such as backups, running
business analytics, or off-loading the generation of reports by querying data.
Best Practices
This section covers recommended vMOS Data Management practices in the following areas:
Over Provisioning
Usage Thresholds
Performance Optimization
535-0008-00 Rev 10
255
Over Provisioning
Thin provision LUNs are allocated as the application writes to these LUNs. Consequently, there is
a difference between provisioned capacity and actual capacity for an array supporting thin LUNs.
Over provisioning is the practice of defining additional space beyond the capacity of the physical
capacity of the array.
Note: Over provisioning applies to the entire array, not individual LUNs.
The following guidelines are suggested practices for over provisioning the array:
Over provisioning is supported to 1000% (10 x) beyond the physical space on an array
SNMP trap event disk-space-low should be set
Hard and soft (alert, limit, and reclaim) space usage thresholds should be set
Snapshots and snap groups should not all be specified as protected to allow for free space
reclamation
For information on usage thresholds, see Free Space Management on page 279. For information
on SNMP traps, see Trap/Notify Event E-mail and Logged Events on page 436 and Available Traps
on page 438.
Usage Thresholds
Effective space management entails specifying maximum limits for the used space, provisioned
space, and space used by snapshots. When this limit is reached, warnings are issued until space
allocation falls below the limit. You set free space thresholds in the following ways:
usedspace-thresholdManages the physical space usage of the system, through the use of
hard and soft (alert and limit) thresholds.
provision-thresholdManages the logical space usage of the system, through the use of hard
and soft (alert and reclaim) thresholds.
Performance Optimization
Asymmetric Active-Active ALUA is recommended for use on thin provision LUNs and LUNs with
snapshots for maximum performance. ALUA optimizes performance by directing the I/O to one port
(Memory Gateway) instead of both. Asymmetric logical unit access is managed implicitly by the
target device.
Caution: Thin provisioned LUNs and LUNs with snapshots should always use ALUA to prevent
256
535-0008-00 Rev 10
Caution: Thin provisioned LUNs and LUNs with snapshots should always use ALUA to prevent
Setting the Preferred Port and ALUA Attributes for a New Thin Provision LUN on page 257
Setting ALUA for an Existing LUN on page 258
535-0008-00 Rev 10
257
Note: The preferredport setting is only applicable when the alua attribute is enabled on a
LUN. If a preferred port is not specified, it defaults to 1. To ensure optimum performance, manual
balancing of LUNs between the two preferred ports is recommended.
Note: The preferredport setting is only applicable when the alua attribute is enabled on a
LUN. If the alua attribute is not enabled, the preferredport setting has no effect.
# lun set container <container_name> name <lun_name> preferredport <1|2>
258
535-0008-00 Rev 10
You can display a list of preferredport sessions using the following command.
# show luns preferredport sessions
Container: MyContainer
LUN
Preferred port
---------------------------------------------VMEM_1
1
Connected
Sessions : vmem-1-mg-a
: Port: hba-a1, Session: wwn.21:00:00:24:ff:45:fc:9a, LUN ID: 1
: Port: hba-b1, Session: wwn.21:00:00:24:ff:45:fc:9a, LUN ID: 1
VMEM_1
1
Connected
Sessions : vmem-1-mg-b
: Port: hba-a1, Session: wwn.21:00:00:24:ff:45:fc:9a, LUN ID: 1
: Port: hba-b1, Session: wwn.21:00:00:24:ff:45:fc:9a, LUN ID: 1
VMEM_2
2
Connected
Sessions : vmem-2-mg-a
: Port: hba-a1, Session: wwn.21:00:00:24:ff:45:fc:9a, LUN ID: 2
: Port: hba-b1, Session: wwn.21:00:00:24:ff:45:fc:9a, LUN ID: 2
VMEM_2
2
Connected
Sessions : vmem-2-mg-b
: Port: hba-a1, Session: wwn.21:00:00:24:ff:45:fc:9a, LUN ID: 2
: Port: hba-b1, Session: wwn.21:00:00:24:ff:45:fc:9a, LUN ID: 2
535-0008-00 Rev 10
259
You can display a list of alua sessions using the following command.
# show luns alua sessions
Container: MyContainer
LUN
ALUA
-----------------------------------VMEM_1
1
Connected
Sessions : vmem-1-mg-a
: Port: hba-a1, Session: wwn.21:00:00:24:ff:45:fc:9a, LUN ID: 1
: Port: hba-b1, Session: wwn.21:00:00:24:ff:45:fc:9a, LUN ID: 1
VMEM_1
1
Connected
Sessions : vmem-1-mg-b
: Port: hba-a1, Session: wwn.21:00:00:24:ff:45:fc:9a, LUN ID: 1
: Port: hba-b1, Session: wwn.21:00:00:24:ff:45:fc:9a, LUN ID: 1
VMEM_2
1
Connected
Sessions : vmem-2-mg-a
: Port: hba-a1, Session: wwn.21:00:00:24:ff:45:fc:9a, LUN ID: 2
: Port: hba-b1, Session: wwn.21:00:00:24:ff:45:fc:9a, LUN ID: 2
VMEM_2
1
Connected
Sessions : vmem-2-mg-b
: Port: hba-a1, Session: wwn.21:00:00:24:ff:45:fc:9a, LUN ID: 2
: Port: hba-b1, Session: wwn.21:00:00:24:ff:45:fc:9a, LUN ID: 2
Thin Provisioning
Thin provisioning provides on-demand allocation of blocks of data, versus the traditional method of
allocating all the blocks up front. vMOS Data Management allocates storage capacity to the thin
provisioned LUNs as application writes data over time, thereby allowing the capacity to grow on
demand.
This section covers the following topics:
260
535-0008-00 Rev 10
Thin Provisioning
Note: A LUN can be created as thin or thick. However, once a LUN is created (thick or thin), the
type cannot be changed. A thick LUN remains thick and a thin LUN remains until deleted.
A thick LUN is a LUN for which the entire provisioned size of the LUN is allocated up front,
when the LUN is created. For instance, when you create a thick LUN of size 10 TB, vMOS will
allocate all the 10 TB right at the time of LUN creation. For information on how to create thick
LUNs, see Working with LUNs on page 190.
A thin LUN is a LUN for which space is allocated as the application writes to the LUN. The
provisioned size of the LUN indicates the maximum size to which this LUN can grow. For
instance, when you create a thin LUN of size 10 TB, vMOS will initially allocate no space for
this LUN at the time of creation. Later, as the application writes to the LUN, space will be
allocated.
This is a flexible approach that allows you to use space as needed. However, there is a risk of
running out of space if you over provision the array. For more information, see Thin Provision
LUNs on page 261 and Free Space Management on page 279.
vMOS provides visibility into the provisioned capacity and allocated capacity for each LUN as well
as the entire container. Free Space Management policies can be used for enforcing control over the
free space available in the container. For more information, see Free Space Management on page
279.
Caution: Thin provisioned LUNs and LUNs with snapshots should always use ALUA, to ensure
optimum performance. For more information, see Optimizing Performance with ALUA on page 257.
Prerequisite
Before you can create a thin provision LUN, log in to the Master Memory Gateway as an Admin
user, then enter Enable and Config mode as shown in the following example.
login as: admin
{cluster: master} > enable
# configure terminal
535-0008-00 Rev 10
261
The following example demonstrates the creation of a 10 GB thin provision LUN named Mylun1 in
a container named MyContainer.
# lun create container MyContainer name Mylun1 size 10 thin
LUN create: success!
Note: Once a LUN is exported, it cannot be deleted without first unexporting the LUN.
Note: You must unexport a LUN prior to deleting it. You are not allowed to delete an exported
LUN.
262
535-0008-00 Rev 10
Snapshots
Note: You must unexport a LUN prior to deleting it. You are not allowed to delete an exported
LUN.
Snapshots
A snapshot is a space-efficient, point-in-time copy of a LUN or other storage volume. By default,
snapshots are read-only. However, you can set the permissions on snapshots to read-write.
Snapshots help high-availability systems avoid down-time, because the time and I/O needed to
create a snapshot does not increase with the size of the data set. A direct backup, however, is
proportional to the size of the data set.
Note: An application has to be quiesced (on the client side) before you can take an application
consistent snapshot. The snapshots discussed in this section refer to crash consistent snapshots,
taken on the array.
535-0008-00 Rev 10
263
Creating Snapshots
Snapshot creation is instantaneous. When a snapshot is initially created it consumes no space. As
writes occur to the LUN, the snapshot grows in size.
The following rules apply for snapshots:
A container can have a maximum of 10,000 snapshots across all the LUNs in the container
Taking a snapshot of a snapshot is not allowed
Snapshots are taken one at a time (serial); creating multiple simultaneous snapshots is not
allowed at this time
Caution: Thin provisioned LUNs and LUNs with snapshots should always use ALUA, to ensure
optimum performance. For more information, see Optimizing Performance with ALUA on page 257.
By default, snapshots are created read-only. However, you can choose to create read-write
snapshots, or change a snapshot to a read-write after its created.
A snapshot name must start with alphanumeric characters, and may also include dashes and
underscores. The [description] option allows you to add a description to the snapshot. You can
allow write privileges to the snapshot with the [readwrite] option. The [protect] option
prevents the snapshot from being deleted during space reclamation.
# snapshot create container <container_name> lun <lun_name> name
<snapshot_name> [description][readwrite][protect]
Deleting Snapshots
You must have Admin user privileges to delete snapshots. When a snapshot is deleted, the amount
of space allocated to the snapshot is returned to free space.
Review the following rules apply when deleting snapshots:
Snapshots that belong to a group cannot be deleted individually. However, you can delete all
the snapshots in a group.
Snapshots that have been exported cannot be deleted. You must unexport the snapshot first,
then it can be deleted.
Protected snapshots cannot be deleted, unless you have Admin user privileges. However, it is
recommended that you unprotect the snapshots first, then delete them.
264
535-0008-00 Rev 10
Snapshots
Container: 41238F00608
Snapshot
LUN
RW Protect Status
Ports AllocSz Created
--------------------------------------------------------------------------------------snapg1_lun1_20130416_133533 lun1
ro no
HA
A,B
0G 2013/04/16_13:35:33
snapg1_lun2_20130416_133533 lun2
ro no
HA
A,B
0G 2013/04/16_13:35:33
snapg1_lun3_20130416_133533 lun3
ro no
HA
A,B
0G 2013/04/16_13:35:33
535-0008-00 Rev 10
265
Container: MyContainer
Snapshot
LUN
RW Protect Status
Ports AllocSz
---------------------------------------------------------------------s3-20-13
lun3
ro no
HA
A,B 50G
Connected Sessions : vmem-dm1-mga
: Port: hba-a2, Session: wwn.21:00:00:24:ff:26:68:08, LUN ID: 1
: Port: hba-a2, Session: wwn.21:00:00:24:ff:26:68:64, LUN ID: 1
: Port: hba-a2, Session: wwn.21:00:00:24:ff:26:68:09, LUN ID: 1
Connected Sessions
: Port: hba-a2,
: Port: hba-a2,
: Port: hba-a2,
: vmem-dm1-mgb
Session: wwn.21:00:00:24:ff:26:68:08, LUN ID: 1
Session: wwn.21:00:00:24:ff:26:68:64, LUN ID: 1
Session: wwn.21:00:00:24:ff:26:68:09, LUN ID: 1
266
535-0008-00 Rev 10
Snapshots
Container: MyContainer
snapshot
NACA
------------------------------Admin_20130329_5500
0
lun5_20130329_5700
1
lun7_20130329_5800
1
lun3_20130329_5301
0
lun9_20130329_5601
1
HR_1_20130329_5201
0
Eng1_20130329_5900
1
Container: MyContainer
LUN
Snapshot Target
Initiator
LUNID Status
--------------------------------------------------------------------Admin1_thick 20813
iscsi iqn.1994-05.com:01f775debf9 1
Active
lun05_thin
20814
iscsi iqn.1994-05.com:01f775debf9 3
Active
lun09_thin
20815
iscsi iqn.1994-05.com:01f775debf9 4
Active
HR13_thick
20816
iscsi iqn.1994-05.com:01f775debf9 2
Active
Performing a Rollback
Before you try performing a rollback, you should understand the requirements and the process.
Requirements:
a. Both ports of a LUN must be offline before issuing the snapshot rollback command.
b. If the snapshot you want to restore the LUN from is read-write, the snapshot must be
unexported first.
c. After the snapshot is unexported, the LUN you want to restore must be unexported as
well.
535-0008-00 Rev 10
267
Rollback Process:
Command Syntax
Offline Ports
Renaming Snapshots
You can change the name of a snapshot at any time, provided it has not been exported. Changing
the name of a snapshot does not impact the existing sessions for that snapshot, as long as the
snapshot has not yet been exported.
Note: Changing the snapshot name after the snapshot has been exported is not recommended,
because traffic can be disrupted. The effect on the client depends on how it handles the change.
Specify the container ID, LUN name, old snapshot name, and new snapshot name.
268
535-0008-00 Rev 10
Snapshots
Note: Existing snapshot sessions are disconnected after making this change.
You are not allowed to modify the individual properties of a snapshot that belong to a snap group.
For example, you are not allowed to change the read-only/read-write property or the protect
property of a snapshot that belongs to a snap group.
# snapshot set container <container_name> lun <lun_name> name <snapshotname> readwrite
Exporting Snapshots
You can export snapshots in the same way you can export LUNs, using the following options:
lunid: assigns a LUN ID to the exported snapshot
to: exports a snapshot to an initiator or initiator group
using: exports a snapshot using a target port
Unexporting a Snapshot
# no snapshot export container <container_name> lun <lun_name> [snapshotname]
535-0008-00 Rev 10
269
Note: If there are no unprotected snapshots and free space has gone below the threshold value,
an event is generated using syslog, SNMP trap, email and other supported notification methods.
Protecting a Snapshot
# snapshot set container <container_name> lun <lun_name> name <snapshotname> protect
Unprotecting a Snapshot
# no snapshot set container <container_name> lun <lun_name> name
<snapshot-name> protect
Snap Groups
A snap group is an object that represents a set of snapshots taken across a collection of LUNs that
belong to a LUN group. When a snapshot of a LUN group is taken, a snapshot is taken of each
individual LUN in the LUN group (serially), then all the resulting snapshots are grouped together as
a Snap Group.
Snap groups are useful for creating an application level checkpoint, when an application is deployed
across several LUNs.
This section covers the following topics:
270
535-0008-00 Rev 10
Snap Groups
For information on how to modify a snap group after its created, see Modifying a Snap Group on
page 272.
Note: Snapshots that belong to a group cannot be deleted individually. However, you can delete
535-0008-00 Rev 10
271
Note: Each LUN must have port A and port B put offline before a rollback can be executed.
272
535-0008-00 Rev 10
Snap Groups
535-0008-00 Rev 10
273
Schedule Management
You can automatically schedule the creation of snapshots and snap groups using schedule
management commands. This feature allows you to define a schedule that creates snapshots or
snap groups at specified time intervals.
Every schedule supports a maximum number of snapshots and snap groups that can be kept at one
time, through the max-keep option. When the max-keep option is implemented, if the maximum
number of stored snapshots is reached, the oldest unexported and unprotected snapshot or snap
group (created by the scheduler) is deleted when the next snapshot is created.
Note: The max-keep limit is set on a per schedule basis, not per LUN or LUN group.
Creating a Schedule
Snapshots can be taken at regular intervals to facilitate space efficient backups. Schedules
automate this process, offering a variety of options so you can tailor them to different needs. For
information on how to create a schedule using the Web interface, see Creating Schedules on page
85.
You can create more than one schedule for a given LUN or LUN group. However, each schedule
applies to only one LUN or LUN group. A cluster can support up to 1024 schedules. When multiple
schedules are set to be executed at the same time, they execute one after the other, not
simultaneously. The sequential execution of schedules may cause a delay between the specified
time of execution and the actual time the schedule is launched.
Each snapshot created by the scheduler is created using the following naming convention:
<LUN_name|LUN_group_name>_<scheduler-name>_<time-stamp>
Note: If you change the name of a LUN, any existing schedules that include the LUN will not
reflect the name change. LUN names are not automatically changed in schedules, however, the
snapshot schedule will continue un-impacted with the renamed LUNs.
274
535-0008-00 Rev 10
Schedule Management
By default, a schedule snapshot is created as read-only. However, when creating a schedule, you
can change this option to read-write.
Customizing a Schedule
Use the following command syntax to create a schedule:
# schedule snapshot create container <container-name> <options>
# write memory
Use the following <options> to customize a schedule:
description: Adds an identifying description to the schedule
disabled: Schedule will be disabled upon creation; default is enabled; it can later be
modified to be enabled
end-date: Specifies the end date for the schedule
end-time: Specifies the end time for the schedule
lun: Creates a schedule for a LUN
lungroup: Creates a schedule for a LUN group
max-keep: Specifies the number of snapshots kept before deleting; does not apply for a
one-time schedule
protect: Specifies the snapshots taken by this schedule are protected from automatic
deletion
readwrite: Specifies that the snapshots taken by this schedule have read/write access
schedule-name: Specifies a name for the schedule
start-date: Specifies the start date for the schedule
start-time: Specifies the start time for the schedule
time-of-day: Specifies the time of day at which a non-periodic schedule triggers
type-daily: Specifies a daily schedule [time-of-day]
type-monthly: Specifies a monthly schedule [day-of-month] [interval]
[day-of-month]: Specifies the day of the month on which a monthly schedule executes; use
a negative number to count backwards from the end of the month
[interval]: Specifies the number of months for the schedule to repeat; the default is 1 if this
option is not specified
type-once: Specifies a one-time schedule [date]
[date]: Specifies the execution date for a one-time schedule
type-periodic: Specifies a periodic schedule [periodicity]
[periodicity]: Specifies the number of recurring intervals for the schedule
type-weekly: Specifies a weekly schedule [days-of-week]
[days-of-week]: Specifies the days of the week on which a weekly schedule executes
Deleting a Schedule
You can delete a schedule at any time. Deleting a schedule does not affect (or impact in any way)
the snapshots or snap groups already created by the schedule. If a schedule is currently in the
process of creating a snapshot or snap group, it does not impact the creation.
When deleting a schedule, you have the option of deleting all unprotected and unexported
snapshots and snap groups created by that schedule. When a LUN or LUN group is deleted, all the
schedules associated with that LUN or LUN group are automatically deleted.
535-0008-00 Rev 10
275
For information on how to delete a schedule using the Web interface, see Modifying and Deleting
Schedules on page 88.
Modifying Schedules
Every schedule supports a maximum number of snapshots and snap groups that can be kept at one
time, through the max-keep option. When the max-keep value is decreased, unexported and
unprotected snapshots and snap groups are deleted as necessary, to meet the new max-keep
requirement. The default value of max-keep is 20 snapshots.
For information on how to modify schedules using the Web interface, see Modifying and Deleting
Schedules on page 88.
This section covers the following topics:
276
535-0008-00 Rev 10
Schedule Management
the disable action does not interrupt or abort the schedule. The schedule will be put into a disabled
state after the creation is complete.
The following sections demonstrate the basic syntax for modifying a schedule, and provides
examples of common uses as well as the options that are unique to modifying a schedule.
Enabling a Schedule
The following example enables a schedule.
# schedule snapshot modify container <container_name> lun <lun_name>
schedule-name <schedule_name> enable
Disabling a Schedule
The following example disables a schedule.
# schedule snapshot modify container <container_name> lun <lun_name>
schedule-name <schedule_name> disable
535-0008-00 Rev 10
277
Note: You can disabletemporarily pausea schedule at any time. If a schedule is in progress,
the disable action does not interrupt or abort the schedule. The schedule will be put into a disabled
state after the creation is complete.
278
535-0008-00 Rev 10
Viewing Schedules
Use the following command to view a particular schedule.
# show schedule <schedule_name>
If a snapshot is created by a schedule, it appears in the list of objects. In this case, the schedule
name appears in the snapshot list.
To monitor space through the Violin Web UI, see CHAPTER 2, Manage, Monitor, and Administer
with the Violin Web Interface on page 31.
535-0008-00 Rev 10
279
# show containers
Container LUNs Total Free Provisioned Ports HostA HostB Status
----------------------------------------------------------------41238F
8
20685G 20661G
40G
A,B
v-mga
v-mgb
HA
provisioned size: amount of logical space used, or the space provisioned for use
770G
740G
20676G
20676G
280
535-0008-00 Rev 10
# show luns
Container: MyContainer
LUN
Size Type
RW Bksz Status Ports Snapshots AllocSz TotalSz
---------------------------------------------------------------------lun5_thck
50G thick rw 512 HA
A,B
324
50G
50G
lun6_thck
75G thick rw 512 HA
A,B
25
75G
75G
lun7_thck
60G thick rw 512 HA
A,B
202
60G
60G
lun23_thn
35G thin
rw 512 HA
A,B
121
21G
23G
lun24_thn
45G thin
rw 512 HA
A,B
221
12G
15G
lun25_thn
50G thin
rw 512 HA
A,B
355
22G
27G
The following command displays information about a particular LUN on a specified container.
# show luns container <container_name> name <lun_name>
Container: MyContainer
LUN Size Type RW Bksz Status
Ports Snapshots AllocSz TotalSz
---------------------------------------------------------------------lun1
1G thin rw 512 HA
A,B
1
0G
0G
Total Allocated Size: 0G
Lun Allocated/Provisioned Size: 0G/1G
Lun Snapshot Protected/Unprotected/Shared Size: 0G/0G/0G
535-0008-00 Rev 10
281
Threshold
usedspace-threshold
provision-threshold
Soft Limit
Hard Limit
Range: 0-100
Default:
0 => no used
space alerts
Range: 0-100
Default:
0 => no space
reclamation
Range: 0 - 999
Default: 0 => no
provisioned space
warnings or alerts
Range 0 - 999
Default: 100 =>
provisioned space
equals usable
space
When provisioned
space exceeds the
soft limit, LUN
creation succeeds
but a warning and
SNMP alerts are
issued.
When provisioned
space exceeds
the hard limit,
further
provisioning is
stopped.
The hard limit
defines the
logical size of
the system.
Example
Soft limit = 50
Hard limit = 75
When 50% of the usable
space in the array is used
up, an SNMP alert is
issued.
When 75% of the usable
space in the array is used
up, space reclamation
begins by deleting viable
snapshots.
Soft limit = 125
Hard limit = 150
Assuming the usable
capacity of array is 20 TB, a
hard limit of 150% allows
the logical capacity of the
array to be 30 TB.
When creating a new LUN
causes the total
provisioned size to exceed
25 TB (125% of 20 TB), a
warning is issued but the
LUN creation succeeds.
When creating a new LUN
causes the total
provisioned size to exceed
30 TB, LUN creation fails
and an error is issued.
535-0008-00 Rev 10
You should not specify all snapshots and snapgroups as protected, otherwise automatic space
reclamation is not possible.
You can never specify a LUN size larger than the total usable capacity on the container.
If the provision-threshold is exceeded, LUN creation is prohibited until LUNs are deleted
or you increase the provisioning space threshold.
con The maximum number of snapshots that can be deleted at one time (max-delete default)
is 10. The max-delete default cannot be modified.
Snapshots and snap groups are deleted from oldest to newest, until one of the following is
achieved:
You can set the provision-thresholdthe maximum provisioned space on the arrayto
a maximum of ten times (1000%) the physical space on the array.
The free space on the array drops below the trigger value.
The maximum number (max-delete) of unprotected, read-only, unexported snapshots or
snapgroups is deleted.
If the max-delete is reached and the available free space does not fall below the trigger,
the identified snapshots and snap groups are deleted and the reclamation process is
restarted.
You can specify the maximum number of snapshots or snap groups to keep (max-keep), as
described in Modifying Schedules on page 276.
For information on the rules that apply to space reclamation, see Rules for Reclaiming Free Space
on page 283.
Note: Threshold values are in percentages (%). Threshold triggers are set at the container-level
(array).
535-0008-00 Rev 10
283
284
535-0008-00 Rev 10
REFERENCE
CLI Reference
APPENDIX A
This appendix lists and describes the available commands in the command line
interface (CLI) of the Violin 6000 Series Memory Array in the following sections.
535-0008-00 Rev 10
287
288
<cluster id>
<domain>
<hostname>
<ifname>
<IP address>
<log level>
<MAC address>
<netmask>
<network prefix>
<node id>
<port>
<regular
expression>
<TCP port>
<URL>
Either a normal URL, using any protocol that wget supports, including
http, https
535-0008-00 Rev 10
If you omit the :password part, you may be prompted for the password
in a follow up prompt, where you can type it securely (without the
characters being echoed). This prompt will occur only if the cli default
prompt empty-password setting is true; otherwise, the CLI assumes
you do not want any password.
If you include the :S character, this is taken as an explicit declaration
that the password is empty, and you will not be prompted in any case.
INTERACTIVE
A command that is only available for interactive usage from the CLI
shell. Specifically, this excludes using the command from jobs, because
the Scheduler runs commands in a batch mode with no direct user
interactivity. In most cases, it is obvious why each of these commands
has the INTERACTIVE restriction.
cli session ... commands change a setting only for the current CLI session. They do not
affect any other sessions, and can be performed by any user at any time. All of these
commands are INTERACTIVE (not available from the Scheduler).
cli default ... commands change the defaults for the specified setting for all future CLI
sessions of all users. They also change the setting for the current session from which they
were executed, but not for any other currently active sessions. Since they change
configuration, the user must be in Configure mode to run them, and hence they can only be
run by admin.
Other cli ... commands which take one-time actions, rather than change a setting, and
thus do not fall under the session or default umbrellas. For example, cli clear-history.
terminal ... commands are clones of a subset of the cli session ... commands, and
are only present for Cisco compatibility. All of these commands are INTERACTIVE (not
available from the Scheduler).
Note that some settings, such as the terminal length and width, are inherently session-specific, and
there are no corresponding commands to set the defaults in configuration.
535-0008-00 Rev 10
289
535-0008-00 Rev 10
These settings are persistent only for the current CLI session. They are also lost if the terminal is
resized and the CLI is able to auto-detect its new size.
Note that the commands without the cli session prefix are identical to those with the prefix, and
are present only for Cisco compatibility.
535-0008-00 Rev 10
291
292
535-0008-00 Rev 10
Displays the product information for the system you are logged in to, as shown in the following
example.
vmem-mga [cluster: master] # show version
Product name:
Product release:
Build ID:
Build date:
Target arch:
Built by:
Array version:
Array Build ID:
Cluster ID:
Licensed for:
vMOS
V6.0.0
#27
2013-04-24 20:13:18
x86_64
vmem-builds.int
A6.0.0
#12
99999-9999-4583
Violin Memory Gateway [ VMOS6 ]
Uptime:
7d 19h 9m 54.688s
Product model:
Host ID:
Serial number:
tt_shb
9d959bb327b4
47208T00102
535-0008-00 Rev 10
293
294
535-0008-00 Rev 10
number> command before the logging <ip-addr> command, as shown in the following
example.
myvmem-n1 [myvmem-n1: master] (config) # vsyslog port 5555
myvmem-n1 [myvmem-n1: master] (config) # logging 10.1.4.169
myvmem-n1 [myvmem-n1: master] (config) # write mem
535-0008-00 Rev 10
295
296
535-0008-00 Rev 10
Note: Files are uploaded from all gateways simultaneously. Therefore, care should be taken to
allow that many simultaneous connections on the server accepting the files (for example, increase
the maximum open ftp connections parameter).
enable
immediate
{current, <file
number>}
535-0008-00 Rev 10
297
include-dump
interval
<interval>
max-size <file
size>
remote-dir <dir
name>
remote-site
<site name>
user <user
name>
<password>
Note: Files are uploaded from all gateways simultaneously. Therefore, care should be taken to
allow that many simultaneous connections on the server accepting the files (for example, increase
the maximum open ftp connections parameter).
If files <file number> is specified, view an archived log file, where the number is from 1
up to the number of archived log files.
If matching <regular expression> is specified, the file is piped through the grep utility
to only include lines matching the provided regular expression.
298
535-0008-00 Rev 10
Note: There are two defined system accounts: one with full privileges (admin), and one with
privileges for reading all data and performing all actions, but not for changing any configuration
(monitor).
Note: Two types of password encryption are supported, SHA-1 and MD5. Secure Hash
Algorithms (SHA) are cryptographic hash functions structured differently and are distinguished by
the appended number, such as -1. Message-Digest Algorithm is a cryptographic hash function that
produces a 128 bit (16 byte) hash value.
299
The nopassword command means that no password is required to log in. The disable command
configures the account so that no one can log into that account.
The disable password command leaves the account as a whole the same, but forbids login with
a password. It is assumed that ssh key access will be used instead. To re-enable the account, the
user must un-disable it, and put a password on it.
The no username <userid> disable command prints a message to this effect; it only exists
to avoid stumping users with an apparently irreversible command.
Disabling a user account does not terminate any current sessions that user has open; it just
prevents new sessions from being established.
300
535-0008-00 Rev 10
too many of them to display in a flat list. A given city may be required to be specified in two, three,
or four words, depending on the city. The possible forms this could take include:
<continent> <city>
<continent> <country> <city>
<continent> <region> <country> <city>
<ocean> <island>
An offset from GMT. This will be in the form GMT-offset GMT, GMT-offset GMT+<1-12>, or
GMT-offset GMT-<1-14>.
UTC. This is almost identical to GMT.
The default is GMT-offset GMT.
show clock
Display the current system time, date, and timezone.
ntpdate <IP address>
Set the system clock using the specified NTP server. This is a one-time operation and does not
cause the clock to be kept in sync on an ongoing basis. It will generate an error if NTP is enabled,
as the socket it requires will already be in use.
ntp peer <IP address> [version <number>]
no ntp peer <IP address>
Add or remove an NTP peer. Allowable version numbers are 3 and 4. If no version number is
specified when adding a peer, the default is 4.
ntp server <IP address> [version <number>]
no ntp server <IP address>
Add or remove an NTP server. Allowable version numbers are 3 and 4. If no version number is
specified when adding a server, the default is 4.
[no] ntp peer <IP address> disable
[no] ntp server <IP address> disable
Disable or re-enable an NTP server or peer. Servers and peers start enabled; disabling is just a way
of making them temporarily inactive without losing their configuration.
[no] ntp enable
[no] ntp disable
Enable or disable NTP overall. Note that the latter is the Cisco command; the former is just a pair
of aliases added to increase usability, as otherwise it may be hard for a user to figure out how to
enable NTP if they are not aware of no commands and only see a way to disable it.
535-0008-00 Rev 10
301
show ntp
Display current NTP settings.
Event Notification
[no] email enable
Enable or disable the sending of e-mail when certain events occur. These events are the same as
those for which SNMP traps can be sent (see SNMP Configuration on page 365).
email mailhub <hostname>
no email mailhub
Set or clear the mail relay to use to send notification emails.
email mailhub-port <TCP port>
no email mailhub-port
Set the mail relay port to use to send notification emails. The no variant resets the port to its default,
which is 25.
email domain <domain name>
no email domain
Set the domain name from which the emails will appear to come from. This is used in conjunction
with the system hostname to form the full name of the host from which the email appears to come.
The rules are as follows:
If an email domain is specified using this command, it is always used. If the hostname has any
dots in it, everything to the right of the first dot is stripped and the email domain is appended.
302
535-0008-00 Rev 10
Note that this does not affect autosupport emails. Autosupport can be disabled overall, but if it is
enabled, all autosupport events (currently only PM process failures) are sent as emails.
[no] email notify recipient <email addr>
Add or remove an email address from the list of addresses to send email notifications of events.
[no] email notify recipient <email addr> class info
[no] email notify recipient <email addr> class failure
Enable or disable the sending of informational or failure events to the specified recipient. Each
event type is classified as either informational or failure. The specified recipient will receive the
intersection of the set of events specified by this command, and the set of events specified overall
with the [no] email notify event <event name> command.
[no] email notify recipient <email addr> detail
Specify whether the emails this recipient is sent should be detailed or summarized. Each email
potentially has both a detailed and summarized form, where the detailed form has a superset of the
information. (In practice, only PM process failure emails currently have a detailed form; for
everything else, the two are the same.)
[no] email autosupport enable
Enable or disable the sending of email to vendor autosupport when certain failures occur.
email autosupport mailhub <hostname> no email autosupport mailhub
Set the mail relay to use to send autosupport emails.
email send-test
Send a test email to all of the configured notification email recipients. This is useful to make sure
the configuration works without having to wait for an event to occur.
show email
Display notification settings. This does not include SNMP traps, which are under the snmp-server
command tree.
Diagnostic Tools
This section provides the syntax for a variety of diagnostic tool commands, along with an
explanation of the functionality.
535-0008-00 Rev 10
303
tcpdump
tcpdump [<options>]
Network diagnostic tool tcpdump. Invokes standard binary, passing command line parameters
straight through. Runs in foreground, printing packets as they arrive, until user presses CTRL+C.
slogin
slogin [<options>] <hostname>
Invokes the SSH client. The user is returned to the CLI when SSH finishes.
telnet
telnet [<options>]
Invokes the telnet client. The user is returned to the CLI when telnet finishes.
show version
show version [concise]
Display version information for the currently running system image. The basic command shows
each field with a description, one per line; the concise variant fits it all onto one line, without labels,
in a form suitable for pasting in a bug report.
show memory
show memory
Display information about system memory utilization.
reload
reload [halt] [noconfirm]
Reboot the system if reload, shut down the system if reload halt.
If there are unsaved changes to the configuration, the user may be prompted as to whether they
want to save these changes (that is, do a write memory) first before rebooting. The prompt will
be suppressed if confirmation of losing unsaved changes is disabled (with the no cli default
confirm lose-unsaved command).
The user may also be prompted to confirm the reload regardless of whether there are unsaved
changes or not. This prompt is contingent on a separate configuration setting, controlled with the
304
535-0008-00 Rev 10
[no] cli default confirm reload command. If both prompts are enabled, and the
configuration was unsaved, the user will be prompted twice.
The noconfirm parameter suppresses both of these confirmations if it is specified.
reload force
reload force
If the system is busy performing another operation requiring the management subsystem (which is
almost any management operation), the regular reload [halt]command will block until it is
finished. If rebooting the system is urgent, the reload force command will do it immediately.
This reboots the system, and there is no halt variant. There is also never any confirmation, whether
or not there are any unsaved changes to the configuration.
Configuration Commands
write memory
Save the running configuration to the active configuration file.
write memory local
Same as configuration write local, provided for Cisco compatibility.
write terminal
Same as show running-config, provided for Cisco compatibility.
configuration jump-start
Run the initial-configuration wizard. The wizard is automatically invoked whenever the CLI is
launched when the active configuration file is fresh (that is, not modified from its initial contents).
This command invokes the wizard on demand.
show configuration [full]
Display a list of CLI commands that will bring the state of the a fresh system up to match the current
persistent state of this system. A short header is included, containing the name and version number
of the configuration, in a comment.
535-0008-00 Rev 10
305
Commands which are not required because they would set something to its default are not included
thus running this command on a fresh configuration will produce no output, aside from the header.
Note that this does not include changes that have been made but not yet written to persistent
storage.
The full option will also include commands that set things to their defaults, but excludes those of
them that are instances of hidden commands.
show configuration running
Same as show configuration except that it applies to the currently running configuration, rather
than the current persisted configuration. Show running-config is for Cisco compatibility.
show configuration files [filename]
If no filename is specified, display a list of configuration files in persistent storage. If a filename is
specified, display the commands to recreate the configuration in that file. In the latter case, only
non-default commands are shown, as for the normal show configuration command; the full,
all, and subtree options are not available.
Note: These configuration commands only work when you are in Configure mode (which
you enter by using the conf t commandsee Enable and Config Modes on page 310).
configuration new
configuration new <name> [factory [keep-basic]]
Create a new configuration file under the specified name. If no additional parameters are specified,
active licenses from the current running configuration are copied over into it. If the factory
parameter is specified, nothing is copied over; it has the factory defaults. If the keep-basic
parameter is specified, local network configuration is copied over so you won't lose network
connectivity when you switch to it.
configuration revert
configuration revert saved
306
535-0008-00 Rev 10
Revert the running configuration to the latest saved version of the active configuration file.
configuration revert factory [keep-basic | keep-connect]
Revert both running and saved configurations to factory defaults. If keep-basic is specified,
preserve local network configuration. If keep-connect is specified, preserve active licenses.
These two options are mutually exclusive.
configuration merge
configuration merge <name>
Merges the shared configuration from one configuration file into the running configuration.
No configuration files are modified during this process. The <name> must be a non-active
configuration file.
configuration switch-to
configuration switch-to <name>
Load configuration from the specified file and change that to be the active configuration file. The
current running configuration is lost, and not automatically saved to the previous active
configuration file.
configuration write
configuration write
configuration write to <name>
Write the running configuration to persistent storage. The former command writes it to the currently
active file. The latter command writes it to the specified file, and changes the active file to that one.
configuration write local
On a system with clustering, only save the configuration on the local box instead of attempting to
save on all cluster members.
configuration write to usb [<name>]
Save a local configuration file to a USB.
535-0008-00 Rev 10
307
configuration fetch
configuration fetch <URL> [<name>]
Download a configuration file from a remote host.
configuration fetch usb [<name>]
Download a configuration file from a USB.
configuration upload
configuration upload <name> <URL>
configuration upload active <URL>
Download or upload a configuration file. A file may not be downloaded over the active configuration
file. If no name is specified for a configuration fetch, it is given the same name as it had on the
server. If active is specified for a configuration upload, the currently-active configuration file is
uploaded. No configuration file may have the name active.
See Key to Command Parameters on page 288 for a description of valid URLs.
Note: If downloading configuration files from another system, they can be found in the /config/
db directory. So an example command line to fetch the initial configuration database would be:
configuration fetch scp://admin:password@hostname/config/db/initial write
memory
write terminal
write terminal
Same as show running-config, provided for Cisco compatibility.
configuration jump-start
configuration jump-start
Run the initial-configuration wizard. The wizard is automatically invoked whenever the CLI is
launched when the active configuration file is fresh (that is, not modified from its initial contents).
This command invokes the wizard on demand.
308
535-0008-00 Rev 10
show configuration
show configuration [full]
Display a list of CLI commands that will bring the state of the a fresh system up to match the current
persistent state of this system. A short header is included, containing the name and version number
of the configuration, in a comment.
Commands which are not required because they would set something to its default are not
includedthus running this command on a fresh configuration will produce no output, aside from
the header.
Note that this does not include changes that have been made but not yet written to persistent
storage.
The full option will also include commands that set things to their defaults, but excludes those of
them that are instances of hidden commands.
show configuration running
Same as show configuration except that it applies to the currently running configuration, rather
than the current persisted configuration. The command show running-config is for Cisco
compatibility.
show configuration files
Displays a list of configuration files in persistent storage.
show configuration files [filename]
If a filename is specified, the commands to recreate the configuration in that file display. In the latter
case, only non-default commands are shown, as for the normal show configuration command;
the full, all, and subtree options are not available.
Image Management
This section provides the syntax for image management commands, along with an explanation of
the functionality of each command.
309
Show Images
show images
Show all image files on the system, as well as what images are installed in two locations: the active
location (which was most recently booted from) and the default location (which is the default to boot
from in the future). There may or may not be overlap between these two lists.
show bootvar
Similar to show images in that it displays what images are on the two locations, and which are the
active and default location. But not all of the show images data is displayed, and additionally this
displays whether or not a boot manager password is set.
310
535-0008-00 Rev 10
535-0008-00 Rev 10
311
312
535-0008-00 Rev 10
535-0008-00 Rev 10
313
314
535-0008-00 Rev 10
backup
link-agg
link-agglayer2+3
link-agglayer3+4
535-0008-00 Rev 10
315
Routing Commands
ip default-gateway <IP address>
no ip default-gateway
Set or remove the default route.
ip route <network prefix> <network mask> <next hop IP address>
no ip route <network prefix> <network mask> [<next hop IP address>]
Set or remove a static route. If it is called with only a network prefix and mask, it deletes all routes
for that prefix.
show ip route
Display the routing table in the system, which includes dynamic routes as well as any active static
routes.
show ip route static
Display the list of configured static routes
show ip default-gateway [static]
Display the currently active default route (or the configured one, if static is used). This is
redundant with the show ip route [static] command, but is provided to enhance usability.
Administrative Commands
This section describes the administrative commands. In general, the commands described in this
section only work when you are in Config mode.
donotuse Command
VIMMs and VCMs that are powered off or that have failed are flagged by the system as a module
that should not be used or may need to be replaced. This flagwhich can be displayed in the CLI
by typing show donotuseis a reminder to the system administrator if a VIMM or VCM fails. This
flag may also display for VIMMs and VCMs that do not require replacement.
show donotuse
The show donotuse command shows VIMMs and VCMs that are flagged as donotuse faults,
including user- and system-set faults. The following is an example of the show donotuse
command along with output for a system-set fault:
# show donotuse
vimmid system VIMM is not responding at 11-01-2012 20:41:11 UTC
User-set faults enable the donotuse flag when an administrator issues one of the following
commands:
316
535-0008-00 Rev 10
The following is an example of the show donotuse command along with output for a user-set fault:
# show donotuse
vimmid user-set VIMM is not responding at 11-01-2012 20:41:11 UTC
no array modules id vimmid donotuse
Use the above command to clear the donotuse flag for a VIMM, where id is the VIMM number.
no array modules id vcm-x donotuse
Use the above command to clear the donotuse flag for a VCM, where x is the VCM letter (a, b, c or
d).
no array modules id <component> donotuse
Use the above command to clear the flag for all system-set faults, where component is either VIMM
or VCM.
no array modules id <component> enable
Use the above command to clear the flag for all user-set faults, where component is either VIMM
or VCM.
If a system begins shutdown or if a maximum number of emails are staged (default 5), they are
sent out automatically.
Emails are consolidated on a per-host basis. Emails from different hosts (acm-a, acm-b, mg-a,
or mg-b) cannot be combined.
Log in to the ACM master or MG master node to configure email consolidation. The default is
disabled.
535-0008-00 Rev 10
317
318
535-0008-00 Rev 10
535-0008-00 Rev 10
319
Remove a static entry from ARP. Note that this cannot be used to remove dynamic entries.
Caution: The Memory Array is preformatted to specific storage capacities, depending on the
type of VIMMs in the system. A single level cell (SLC) system is formatted to 65% and a multi-level
cell system is formatted to 84%. For optimum system performance, do not change these values.
The following storage capacity percentages are supported: 50%, 65%, 78%, 84% and 87%.
violin-acma [violin: master] (config) # array format capacity <50 | 65
| 78 | 84 | 87>
320
535-0008-00 Rev 10
Use the array modules type <module type> enable command to power up a specific array
module type:
violin-acma [violin: master] (config) # array modules type <vcm | acm |
fpm | mg | hba> enable
Use the [no] prefix to power down a specific module type, for example:
violin-acma [violin: master] (config) # no array modules type vcm
enable
Use the array cooling command to adjust the allowable temperature of the Memory Array. This
command reduces the speed of the fans based on the preferred setting.
violin-acma [violin: master] (config) # array cooling <normal | eco>
The eco option allows the Memory Array to run at a higher temperature before the fans increase in
speed to lower the temperature of the system.
535-0008-00 Rev 10
321
Use the show array detail command to show more information about the array. The following
example includes information about power supplies, LEDs and fans:
violin-acma [violin: master] # show array detail
ARRAY: VIOLIN_MEMORY_ARRAY_23109R00000012
INDEX: 0
Chassis Type
: V6000
Number of VIMMs
: 64
Ambient Temp
: 22
Controller Temp
: 24
Power A
: OFF
Power B
: ON
Lid Ajar Time
: 0 secs
Alarm LED
: ON
Power A LED
: OFF
Power B LED
: ON
MG-a Status
: running
MG-b Status
: running
Fans:
Fan a1
: High
Fan b1
: Slow
Fan c1
: Slow
Fan a0
: High
Fan b0
: Slow
Fan c0
: Slow
violin-acma [violin: master] #
show array modules
Display information about all of the array modules.
show array summary
Display summary information for arrays.
Additional show commands are available by typing show ? at the command prompt.
show array modules type <all | vimm | vcm | acm | pcm |fpm | mg | hba>
Display information about a module type.
See General Configuration Commands on page 292 for additional Array configuration commands,
and the Violin 6000 Series Memory Installation Guide for the initial steps to configure the Violin 6000
Series Memory Array.
322
535-0008-00 Rev 10
Balance of the system can be restored by moving VIMMs from RAID rebuilds. If necessary,
multiple RAID rebuilds can be performed at the same time.
array balance enable
The above command enables automatic balancing. When unbalanced RAID groups are detected,
they are automatically rebuilt.
array balance
The above command starts a RAID rebuild for each VCM that requires a rebuild. More than one
rebuild may be required to balance the system.
no array balance enable
Use this command to disable (turn off) automatic rebalancing.
array balance schedule once
Use the above command to schedule an array balance for the default time of the upcoming
Saturday at 1:00 A.M. This starts a RAID rebuild for each VCM that requires a rebuild. More than
one rebuild may be required to balance the system.
no array balance schedule once
Use the above command to cancel a scheduled one-time rebalance.
show array balance
Use the above command to display the current RAID rebalance settings and view the status of all
the VCMs.
array balance schedule once [day-of-week <sun|mon|tue|wed|thu|fri|sat>]
[time <hh:mm:ss>]
Use the above command and parameters to set a one-time RAID rebalance at a specified time and
day of the week.
array balance schedule weekly [day-of-week <sun|mon|tue|wed|thu|fri|sat>]
[time <hh:mm:ss>]
Use the above command and parameters to set a specific time and day of the week on a recurring
basis. Multiple RAID rebuilds are completed until the system is balanced. The scheduled array
balance is disabled once the system is in balance.
no array balance schedule weekly
Use the above command to cancel scheduled weekly recurring rebalances.
See Rebalancing RAID Groups on page 242 for more information on using the array balance
commands.
535-0008-00 Rev 10
323
VIMM Commands
This section describes the VIMM configuration and monitoring commands that you can use in the
CLI.
Use the array modules id vimm command to power on VIMMs. Use the [no] prefix to power
off a VIMM. For example:
violin-acma [violin: master] (config) # [no] array modules id vimm<xx>
enable
Where:
[no]
id
<xx>
Use the show vimms set of commands to display all or specific VIMM information. For example:
violin-acma [violin: master] (config) #show vimms [detail | summary]
Where:
324
<detail>
<summary>
535-0008-00 Rev 10
iSCSI Commands
The following example shows output of the show vimms detail command for one VIMM:
violin-acma [violin: master] (config) #show vimms detail
VIMM01:
VCM
: vcm-a
Type
: 512G-MLC-Flash
Status
: Active
Present
: yes
Power
: yes
DoNotUse Fault
: no
Current (mA)
: 989.38
RAID
: 1
Spare
: no
Health
: health threshold: (OK)
Temp (C)
: 39
Serial
: 2610CN00002723
Model
: 620-0046-01_R03
Date
: 20110212
FpgaVersion
: 19935
SwVersion
: 19892
%-Format Capacity
: 65.00
%-DieFail
: 0.00
%-BlkFail
: 0.00
%-BlkEraseAvg
: 2.00
%-LifeTimeRemaining
: 98.00
vvimms
Use this command to display the Violin 6000 Series Memory Array VIMMs.
iSCSI Commands
This section provides the syntax for iSCSI commands, along with an explanation of the functionality
of each command.
#Bindings: 4
10.1.9.82 (unknown)
535-0008-00 Rev 10
325
192.168.1.101 (eth3)
192.168.2.100 (unknown)
192.168.3.101 (eth4)
Security Commands
This section covers the following commands:
326
535-0008-00 Rev 10
Security Commands
CHAP Requirements
iSCSI CHAP configurations must meet the following requirements:
The minimum CHAP secret size is 12 characters (bytes) (IPSec is not supported).
For bidirectional CHAP, each target-initiator pair must have one target CHAP secret and one
initiator CHAP secret configured.
In bidirectional CHAP mode, the target CHAP secret cannot be identical to the initiators CHAP
secret to which it is connected.
535-0008-00 Rev 10
327
RADIUS Configuration
radius-server timeout <seconds>
no radius-server timeout
Sets (or resets to the default) a global communication value for all RADIUS servers. Can be
overridden in a radius-server host command. The default is 3. Sets the timeout for
retransmitting a request to any RADIUS server. Range is 1-60.
radius-server retransmit <retries>
no radius-server retransmit
Sets (or resets to 0) a global communication value for all RADIUS servers. Can be overridden in a
radius-server host command. Defaults to 1. Sets the number of times the client will attempt
to authenticate with any RADIUS server. To disable retransmissions set it to zero. Range is 0-5.
radius-server key <string>
no radius-server key
Sets (or clears) a global communication value for all RADIUS servers. Can be overridden in a
radius-server host command. Sets the shared secret text string used to communicate with
any RADIUS server.
radius-server host {hostname , ip-address} [auth-port <port-number>]
[timeout <seconds>] [retransmit <retries>] [key <string>]
no radius-server host {hostname , ip-address} [auth-port <port-number>]
Add a RADIUS server to the set of servers used for authentication. Some of the parameters given
may override the configured global defaults for all RADIUS servers. The auth-port defaults to
1812 and is used for authentication requests.
The same IP address can be used in more than one radius-server host command as long as
the auth-port is different for each. auth-port is a UDP port number. auth-port must be
specified immediately after the host option (if present).
If no radius-server host {hostname , ip-address} is specified, all radius specific
configuration for this host is deleted. no radius-server host {hostname , ip-address}
328
535-0008-00 Rev 10
Security Commands
auth-port <port> may be specified to refine which host is deleted, as the previous command
will delete all RADIUS servers with the specified ip-address.
RADIUS servers are tried in the order they are configured.
show radius
Show the RADIUS configuration.
Note: We do not have CLI commands to specify the acct-port (accounting port) or retransmit
on a per server basis. Also, there is no radius-server deadtime command as there is in the
Cisco command set.
TACACS+ Configuration
This section provides the syntax for TACACS+ configuration commands, along with an explanation
of the functionality of each command.
535-0008-00 Rev 10
329
Note: By default ALUA is not enabled. To use this feature, you must enable ALUA on a per LUN
basis.
lun set container <container_name> name <lun_name> alua
Enables ALUA for a single LUN in a specified container.
lun set container <container_name> name * alua
Enables ALUA for all LUNs in the container through the use of the wildcard (*) character.
lun create container <container_name> name <LUN_name> size <size_GB> alua
Enables ALUA for the newly created LUN.
330
535-0008-00 Rev 10
This section provides the syntax for initiator group and initiator group related commands. An
explanation of the functionality of each command is also provided.
Changing the LUN ID after a LUN has been exported is not recommended. For example, if a
LUN is exported as LUN ID 1 and later the export is changed to LUN ID 2, traffic is disrupted
during the change. The affect on the client depends on how it handles LUN ID changes.
You can export LUNs to different igroups on different targets, even when they have initiators in
common.
Note: After you create an export for a specific iSCSI initiator using a specific target, the initiator
is excluded from seeing LUN exports from that target to all initiators.
535-0008-00 Rev 10
331
Use a unique LUN ID (LUN name) for each LUN to avoid the chance of conflicts.
Refrain from exporting a LUN to different targets using a different LUN ID (LUN name) for the
LUN for each target, as this practice can result in conflicts.
Plan the nested igroups hierarchy first. Then, export the LUNs to the different igroups, allowing
the system to auto-assign LUN IDs to ensure they are each unique.
create name A
create name B
addto name A initiators B
addto name A initiators wwn.50:01:43:80:18:6b:3e:f2
Now the igroup A contains two members wwn.50:01:43:80:18:6b:3e:f2 and B. Since B is an igroup,
we call the A is a nested igroup.
332
535-0008-00 Rev 10
LUN Commands
LUN Commands
This section covers the following commands:
This section provides the syntax for LUN commands, along with an explanation of the functionality
of each command.
VMware ESX 4.1 does not support 4K block size. For VMware ESX 4.1, use a 512 byte block size
LUN, which is supported.
Create LUNs
The lun create command enables you to create a LUN within a specified storage container.
Command Syntax
[no] lun create container <container_name> name <LUN_name> size [<size GB>
| equal] [quantity <number>] [nozero] [readonly] [startnum <unsigned
integer>] [blksize 512 | 4096] [offline] alua naca
535-0008-00 Rev 10
333
size
Defines the size of each LUN in GB, or equal to divide the free
space of the container equally among the new LUNs. A value of
0 (zero) may be used in place of equal.
container
quantity
nozero
readonly
startnum
blksize
offline
alua
naca
View LUNs
The show luns command enables you to view LUNs. LUNs may be filtered by container, name,
and session parameters. To view a list of containers, enter show luns container ?. The name
parameter enables you to view only those LUNs which are prefixed by a particular name. The udid
option shows the UDID associated with a LUN.
show luns [container] [name] [count] [reserved] [serial] [sessions]
The following table provides descriptions for each command argument and option.
334
535-0008-00 Rev 10
LUN Commands
container
name
count
reserved
serial
sessions
Rename a LUN
You can rename a LUN using the following command.
Note: Renaming or deleting a LUN is not allowed with actively exported LUNs. You must
535-0008-00 Rev 10
335
Resize a LUN
You can increase the size of an existing LUN by specifying a new size or specifying the amount you
want to increase the existing size in gigabytes. Decreasing the size of an existing LUN is not
supported. For a smaller size LUN, create a new LUN of the desired size.
lun resize container <container_id> name <lun_name> size <new_GB>
Increases the size of an existing LUN by specifying the new size in gigabytes.
lun resize container <container_id> name <lun_name> size +<increment_GB>
Increase the size of an existing LUN by specifying the amount (GBs) to increase the current size.
WARNING! Existing client side file systems that might be stored on the LUN are not extended
as part of the resize operation. Increasing the size of an existing client side file system must be done
on the client. Some client systems do not detect a LUN size change until they reboot.
Export LUNs
The lun export command enables you to export the LUNs within a specific container to an
initiator or initiator group. A single LUN may be exported multiple times through different target ports
(multipathing) or to different initiators (clustered access).
Caution: When working with initiator groups or nested initiator groups, review the recommended
Syntax
336
535-0008-00 Rev 10
LUN Commands
lun_name
container <container_name>
name <lun_name>
lunid
igroup
initiator
port
Note: Select shell wildcard patterns can be used in the <lun_name> field to match a set of LUNs.
For example, you can use the asterisk (*) to match all characters that follow. Such as mylun* to
match all LUN names that begin with mylun. Or, you can specify a range of LUN names using the
[a-e] wildcard. Such as my-lun[2-4][0-9] to match all LUNs from my-lun20 to my-lun49.
Use the on global option to get statistics for data flowing to LUNs through all nodes in a cluster.
Use the on host <hostname> option to get statistics for data flowing to LUNs on a specific host.
Use the continuous option to refresh displayed statistics every ten seconds.
Use the detail option to display statistics in a verbose format.
535-0008-00 Rev 10
337
A LUN group can contain a minimum of one LUN and a maximum of eight LUNs.
A LUN group name must start with an alphanumeric character, and may contain hyphens,
underscores, and periods.
Substitute LUN names for the <lun#> variables in the command above. You must have one LUN to
create a LUN group, and optionally can specify seven other LUNs as members.
# lungroup create container <container_name> name <lungroup_name> lun
<lun1> lun [<lun2> lun <lun3> lun <lun4> ... lun <lun8>]
Note: You must delete the snap groups (snapshots of a LUN group) associated with a LUN group
before you are allowed to delete the LUN group. Snapshots that belong to a group cannot be
deleted individually. You must delete all the snapshots in a snap group.
Note: If a LUN is deleted, it is automatically removed from any LUN groups of which it is a
member. If deleting a LUN results in a LUN group with zero members, then that LUN group is
automatically deleted.
338
535-0008-00 Rev 10
the group. If you delete all LUNs from a LUN group, the LUN group is automatically deleted.
Note: Renaming a LUN group with exported LUNs is not recommended, as traffic can be
disrupted. For more information, see Guidelines for Exporting LUNs on page 194.
# lungroup rename container <container-id> name <old-lungroup-name> to
<new-lungroup-name>
535-0008-00 Rev 10
339
Container: MyContainer
LUN Group
LUNs
-------------------------------------------------------------Finance_LG
2013,1stQ,2ndQ,3rdQ,4thQ
HR_LG
Benefits,Employees,Expenses
RandD_LG
Resources,Expenses,HeadCount
Eng_thn_LG
eng1_thn,eng2_thn,eng3_thn,eng5_thn
Admin_thn_LG
adm1_thn,adm2_thn,adm3_thn
Corp_thn_LG
corp5_thn,corp6_thn,corp7_thn
340
535-0008-00 Rev 10
Caution: Thin provisioned LUNs and LUNs with snapshots should always use ALUA, to prevent
There is a preferredport LUN attribute that is used in conjunction with the alua LUN attribute
that sets the preferred bit and target port asymmetric access state of each target port group.
Note: The preferredport setting is only applicable when the alua attribute is enabled on a
Note: The preferredport setting is only applicable when the alua attribute is enabled on a
LUN. If the alua attribute is not enabled, the preferredport setting has no effect.
# lun set container <container_name> name <lun_name> preferredport <1|2>
535-0008-00 Rev 10
341
342
535-0008-00 Rev 10
Container: MyContainer
LUN
Preferred port
---------------------------------------------VMEM_1
1
Connected
Sessions : vmem-1-mg-a
: Port: hba-a1, Session: wwn.21:00:00:24:ff:45:fc:9a, LUN ID: 1
: Port: hba-b1, Session: wwn.21:00:00:24:ff:45:fc:9a, LUN ID: 1
VMEM_1
1
Connected
Sessions : vmem-1-mg-b
: Port: hba-a1, Session: wwn.21:00:00:24:ff:45:fc:9a, LUN ID: 1
: Port: hba-b1, Session: wwn.21:00:00:24:ff:45:fc:9a, LUN ID: 1
VMEM_2
2
Connected
Sessions : vmem-2-mg-a
: Port: hba-a1, Session: wwn.21:00:00:24:ff:45:fc:9a, LUN ID: 2
: Port: hba-b1, Session: wwn.21:00:00:24:ff:45:fc:9a, LUN ID: 2
VMEM_2
2
Connected
Sessions : vmem-2-mg-b
: Port: hba-a1, Session: wwn.21:00:00:24:ff:45:fc:9a, LUN ID: 2
: Port: hba-b1, Session: wwn.21:00:00:24:ff:45:fc:9a, LUN ID: 2
Viewing ALUA Settings
You can display a list of LUNs with alua settings using the following command.
# show luns alua
Container: MyContainer
LUN
ALUA
-----------------------------------ALUAlun
1
bigthin
1
thick1
0
thick2
0
thick3
0
thin1
1
thin2
1
thin3
1
thin4
1
535-0008-00 Rev 10
343
You can display a list of alua sessions using the following command.
# show luns alua sessions
Container: MyContainer
LUN
ALUA
-----------------------------------VMEM_1
1
Connected
Sessions : vmem-1-mg-a
: Port: hba-a1, Session: wwn.21:00:00:24:ff:45:fc:9a, LUN ID: 1
: Port: hba-b1, Session: wwn.21:00:00:24:ff:45:fc:9a, LUN ID: 1
VMEM_1
1
Connected
Sessions : vmem-1-mg-b
: Port: hba-a1, Session: wwn.21:00:00:24:ff:45:fc:9a, LUN ID: 1
: Port: hba-b1, Session: wwn.21:00:00:24:ff:45:fc:9a, LUN ID: 1
VMEM_2
1
Connected
Sessions : vmem-2-mg-a
: Port: hba-a1, Session: wwn.21:00:00:24:ff:45:fc:9a, LUN ID: 2
: Port: hba-b1, Session: wwn.21:00:00:24:ff:45:fc:9a, LUN ID: 2
VMEM_2
1
Connected
Sessions : vmem-2-mg-b
: Port: hba-a1, Session: wwn.21:00:00:24:ff:45:fc:9a, LUN ID: 2
: Port: hba-b1, Session: wwn.21:00:00:24:ff:45:fc:9a, LUN ID: 2
Caution: Thin provisioned LUNs and LUNs with snapshots should always use ALUA, to prevent
a substantial performance hit. For more information, see Optimizing Performance with ALUA on
page 341.
344
535-0008-00 Rev 10
The following example demonstrates the creation of a 10 GB thin provision LUN named Mylun1 in
a container named apple.
# lun create container MyContainer name Mylun1 size 10 thin
LUN create: success!
Note: Once a LUN is exported, it cannot be deleted without first unexporting the LUN.
Note: You must unexport a LUN prior to deleting it. You are not allowed to delete an exported
LUN.
535-0008-00 Rev 10
345
Note: You must unexport a LUN prior to deleting it. You are not allowed to delete an exported
LUN.
Snapshots
A snapshot is a space-efficient, point-in-time copy of a LUN or other storage volume. By default,
snapshots are read-only. However, you can set the permissions on snapshots to read-write.
Note: An application has to be quiesced (on the client side) before you can take an application
consistent snapshot. The snapshots discussed in this section refer to crash consistent snapshots,
taken on the array.
Creating Snapshots
A snapshot name must start with alphanumeric characters, and may also include dashes and
underscores. The [description] option allows you to add a description to the snapshot. You can
allow write privileges to the snapshot with the [readwrite] option. The [protect] option
prevents the snapshot from being deleted during space reclamation.
# snapshot create container <container_name> lun <lun_name> name
<snapshot_name> [description][readwrite][protect]
Deleting Snapshots
The following rules apply when deleting snapshots:
346
Snapshots that belong to a group cannot be deleted individually. However, you can delete all
the snapshots in a group.
Snapshots that have been exported cannot be deleted. You must unexport the snapshot first,
then it can be deleted.
Violin 6000 Series Memory Array Users Guide
535-0008-00 Rev 10
Protected snapshots cannot be deleted, unless you have Admin user privileges. However, it is
recommended that you unprotect the snapshots first, then delete them.
Container: 41238F00608
Snapshot
LUN
RW Protect Status
Ports AllocSz Created
--------------------------------------------------------------------------------------snapg1_lun1_20130416_133533 lun1
ro no
HA
A,B
0G 2013/04/16_13:35:33
snapg1_lun2_20130416_133533 lun2
ro no
HA
A,B
0G 2013/04/16_13:35:33
snapg1_lun3_20130416_133533 lun3
ro no
HA
A,B
0G 2013/04/16_13:35:33
535-0008-00 Rev 10
347
Container: MyContainer
snapshot
ALUA
------------------------------vmem_20130319_175400
0
lun1_20130319_175700
1
lun2_20130319_180000
1
lun3_20130319_180301
0
lun4_20130319_180601
0
HR_1_20130319_181201
0
Eng1_20130319_181500
1
Container: MyContainer
LUN
Snapshot Target
Initiator
LUNID Status
--------------------------------------------------------------------Admin1_thick 20813
iscsi iqn.1994-05.com:01f775debf9 1
Active
lun05_thin
20814
iscsi iqn.1994-05.com:01f775debf9 3
Active
lun09_thin
20815
iscsi iqn.1994-05.com:01f775debf9 4 Inactive
HR13_thick
20816
iscsi iqn.1994-05.com:01f775debf9 2
Active
348
535-0008-00 Rev 10
Performing a Rollback
Before you try performing a rollback, you should understand the requirements and the process.
Requirements:
a. If the snapshot you want to restore the LUN to is read-write, the snapshot must be taken
offline first.
b. After the snapshot is taken offline, the LUN you want to restore must be taken offline.
Process:
Command Syntax
Note: Existing snapshot sessions are disconnected after making this change.
You are not allowed to modify the individual properties of a snapshot that belong to a snap group.
For example, you are not allowed to change the read-only/read-write property or the protect
property of a snapshot that belongs to a snap group.
535-0008-00 Rev 10
349
Unexporting a Snapshot
# no snapshot export container <container_name> lun <lun_name> [snapshotname]
Protecting a Snapshot
# snapshot set container <container_name> lun <lun_name> name <snapshotname> protect
Unprotecting a Snapshot
# no snapshot set container <container_name> lun <lun_name> name
<snapshot-name> protect
Snap Groups
You can create a snap group from an established LUN group. A snap group is a group of snapshots
of the LUNs belonging to the LUN group.
Creating a Snap Group
This operation creates a snapshot of each LUN in the snap group, then links the snapshots together
into a group.
# snapgroup create container <container_name> lungroup <lungroup_name>
name <snap_group_name>[readwrite] [protect] [description]
350
535-0008-00 Rev 10
Note: Snapshots that belong to a group cannot be deleted individually. However, you can delete
535-0008-00 Rev 10
351
352
535-0008-00 Rev 10
Schedule Management
You can automatically schedule the creation of snapshots and snap groups using schedule
management commands.
Creating a Schedule
Use the following command syntax to create a schedule:
# schedule snapshot create container <container-name> <schedule-options>
Use the following <options> to customize a schedule:
535-0008-00 Rev 10
353
protect: Specifies the snapshots taken by this schedule are protected from automatic
deletion
readwrite: Specifies that the snapshots taken by this schedule have read/write access
schedule-name: Specifies a name for the schedule
start-date: Specifies the start date for the schedule
start-time: Specifies the start time for the schedule
time-of-day: Specifies the time of day at which a non-periodic schedule triggers
type-daily: Specifies a daily schedule [time-of-day]
type-monthly: Specifies a monthly schedule [day-of-month] [interval]
[day-of-month]: Specifies the day of the month on which a monthly schedule executes; use
a negative number to count backwards from the end of the month
[interval]: Specifies the number of months for the schedule to repeat; the default is 1 if this
option is not specified
type-once: Specifies a one-time schedule [date]
[date]: Specifies the execution date for a one-time schedule
type-periodic: Specifies a periodic schedule [periodicity]
[periodicity]: Specifies the number of recurring intervals for the schedule
type-weekly: Specifies a weekly schedule [days-of-week]
[days-of-week]: Specifies the days of the week on which a weekly schedule executes
Deleting a Schedule
Deleting a schedule does not affect (or impact in any way) the snapshots or snap groups already
created by the schedule. If a schedule is currently in the process of creating a snapshot or snap
group, it does not impact the creation.
# no schedule snapshot create container <container-name> <scheduleoptions>
354
535-0008-00 Rev 10
Enabling a Schedule
The following example enables a schedule.
# schedule snapshot modify container <container_name> lun <lun_name>
schedule-name <schedule_name> enable
Disabling a Schedule
The following example disables a schedule.
# schedule snapshot modify container <container_name> lun <lun_name>
schedule-name <schedule_name> disable
Note: You can disabletemporarily pausea schedule at any time. If a schedule is in progress,
the disable action does not interrupt or abort the schedule. The schedule will be put into a disabled
state after the creation is complete.
535-0008-00 Rev 10
355
Viewing Schedules
Use the following command to view a particular schedule.
# show schedule <schedule_name>
If a snapshot is created by a schedule, it appears in the list of objects. In this case, the schedule
name appears in the snapshot list.
356
535-0008-00 Rev 10
535-0008-00 Rev 10
357
358
535-0008-00 Rev 10
535-0008-00 Rev 10
359
Display notification settings. This does not include SNMP traps, which are under the snmp-server
command tree.
See Event Notification on page 302 for additional e-mail-specific commands.
show hosts
Show all values configured for the following: hostname, name servers, domain name list, and static
host mappings.
show interfaces
Display detailed running state for all interfaces.
show inventory
Display basic information about the VCMs and VIMMs in the Array.
show ip <default-gateway | dhcp | route>
Display IP-related information for the commands above.
show log <continuous | files | matching | not>
Display logging configuration for the above commands. Where:
<cr>
<continuous>
<files>
<matching>
<not>
360
535-0008-00 Rev 10
labeled external a and b below. The display will show optimal for a port that is configured at its
highest speed, degraded if it is at a lower speed and no link if it cant detect a connection.
violin-acma [violin: master] (config) # show pcie
Current PCI-E config: mg
Config mg
Config direct-attach
Ports acm-a (lab-stein12-acma):
mg_out: optimal
hba-a: degraded: no link
hba-b: degraded: no link
mg_in: optimal
vcm-a: optimal
vcm-b: optimal
vcm-c: optimal
vcm-d: optimal
Ports acm-b (lab-stein12-acmb):
mg_out: optimal
hba-c: degraded: no link
hba-d: degraded: no link
mg_in: optimal
vcm-a: optimal
vcm-b: optimal
vcm-c: optimal
vcm-d: optimal
violin-acma [violin: master] (config) #
show version
Display version information for the current system image.
show vimms
Display information about the VIMMs.
See VIMM Commands on page 324 for additional VIMM-specific commands.
535-0008-00 Rev 10
361
remote-first: If a local-user mapping attribute is returned and is a valid local user name, map
the authenticated user to the local user specified in the attribute. Otherwise, if the attribute is
not present or not valid locally, use the user specified by the default-user command. (This is
the default behavior.)
remote-only: Only try to map a remote authenticated user if the authentication server sends a
local-user mapping attribute. If the attribute does not specify a valid local user, no further
mapping is tried.
local-only: All remote users will be mapped to the user specified by the aaa authorization map
default-user <user name> command. Any vendor attributes received by an authentication
server are ignored.
show aaa
Shows the current authentication and authorization settings.
SSH Configuration
SSH Server
[no] ssh server enable
Enable or disable the ssh server.
If the ssh server is disabled, the CLI is only accessible over the serial console. Note that this does
not terminate existing ssh sessions; it will only prevent new ones from being established.
ssh server host-key generate
362
535-0008-00 Rev 10
Regenerate new host keys for the SSH server. This generates three keys: RSA for sshv1, RSA for
sshv2, and DSA for sshv2. Note that the system automatically generates the host keys on its first
boot, so this only needs to be done if a security breach is suspected and the keys need to be
changed.
ssh server host-key <type> private-key <key>
ssh server host-key <type> public-key <key>
Manually set the host-key (either private or public, but should be both if changing) of the specified
key type.
[no] ssh server listen enable
Enable (or disable) the listen interface restricted list for sshd. If enabled and at least one non-DHCP
interface is specified in the list, the SSH connections are only accepted on those specified
interfaces. When disabled, SSH connections are accepted on any interface.
[no] ssh server listen interface <ifname>
Provides a mechanism to add and remove interfaces to the 'listen' list. If the interface is also running
as a DHCP client, it will be as if the interface was not added. If DHCP is later turned off on this
interface, it will be as if the interface was then added to the listen list.
show ssh server
Display information about the ssh server, including whether or not it is enabled, and the host keys.
SSH Client
ssh client user <username> identity <key-type> generate [passphrase
<phrase>]
Generate a new identity (private and public keys) for the specified user name. The given user name
must correspond to a valid local user account. When the keys are generated, the private key is
written to the users ssh directory in an appropriately named file (id_dsa).
This identity can be used when the user uses the slogin command to connect from the system to
another host.
DSA and RSA v2 keys for SSHv2 can be generated. This is specified with dsa2 or rsa2 as the
key-type parameter.
ssh client user <username> identity <key-type> public-key <key>
ssh client user <username> identity <key-type> private-key <key>
Set the public or private key (of specified type) for the specified user name. This is an alternative to
generating the key in the above command and is also used for reverse mapping generated keys.
no ssh client user <username> identity [<key-type>]
Removes the public/private keys for the specified user. Any private key file in a valid user SSH
directory is deleted.
ssh client user <username> authorized-key sshv2 <key>
535-0008-00 Rev 10
363
The specified key is added to the list of authorized sshv2 RSA or DSA public keys for this user
account. These keys can be used to log into the users account. The specified user must be a valid
account on the system. As keys are added, an implicit id is associated with the key. This is to make
key deletion easier.
Be aware that if a key is being pasted from a cut buffer and was displayed with a paging program,
it is likely that newline characters have been inserted, even if the output was not long enough to
require paging. Most likely show command output will be displayed this way, as paging is enabled
by default in the CLI. One can specify no cli session paging enable before doing the show
command to prevent the newlines from being inserted.
no ssh client user <username> authorized-key sshv2 <key id>
Remove a public key from the specified users authorized key list. The key identifier can be found
by using 'show ssh client'.
show ssh client
Display information about SSH client identities (public/private keys) and the per user list of
authorized keys for the users.
Banner
At various login points, some legal and welcome text can be displayed. This text is controlled by
bindings and can be set by the user from the CLI as follows:
banner login <string>
no banner login
Set the contents of the /etc/issue and /etc/issue.net files.
banner motd <string>
no banner motd
Set the contents of the /etc/motd file.
show banner
Display the contents of the currently configured banners.
CLI Options
There are four groups of commands relating to the CLI itself:
364
cli session ... commands change a setting only for the current CLI session. They do not
affect any other sessions, and can be performed by any user at any time. All of these
commands are INTERACTIVE (not available from the Scheduler).
cli default ... commands change the defaults for the specified setting for all future CLI
sessions of all users. They also change the setting for the current session from which they
were executed, but not for any other currently active sessions. Since they change
configuration, the user must be in Config mode to run them, and hence they can only be run by
admin.
535-0008-00 Rev 10
Other cli ... commands which take one-time actions, rather than change a setting, and
thus do not fall under the session or default umbrellas. For example, cli clear-history.
terminal ... commands are clones of a subset of the cli session ... commands, and
are only present for Cisco compatibility. All of these commands are INTERACTIVE (not
available from the Scheduler).
Note that some settings, such as the terminal length and width, are inherently session-specific, and
there are no corresponding commands to set the defaults in configuration.
cli default auto-logout <length in minutes>
cli session auto-logout <length in minutes>
no cli default auto-logout
no cli session auto-logout
Control the length of user inactivity required before the CLI will automatically log a user out. The no
... variants of this command disable the automatic logout feature.
SNMP Configuration
SNMP MIBs are available at ftp://ftp.violin-memory.com and in the Violin Web interface.
[no] snmp-server enable
Enable or disable the SNMP server. Note that this not only stops serving of SNMP variables, but
also the sending of SNMP traps.
[no] snmp-server enable traps
Enable or disable sending of SNMP traps from this system. Traps may only be enabled if the SNMP
server overall is enabled. The traps sent by the SNMP agent are:
Note that traps are only sent if there are trap sinks configured with the snmp-server host ...
command, and if these trap sinks are themselves enabled.
[no] snmp-server listen enable
Enable (or disable) the listen interface restricted list for snmpd. If enabled and at least one nonDHCP interface is specified in the list, the snmp connections are only accepted on those specified
interfaces. When disabled, snmp connections are accepted on any interface.
[no] snmp-server listen interface <ifname>
535-0008-00 Rev 10
365
Provides a mechanism to add and remove interfaces to the listen list. If the interface is also
running as a DHCP client, it will be as if the interface was not added. If DHCP is later turned off on
this interface, it will be as if the interface was then added to the listen list.
snmp-server community <community name>
no snmp-server community
Set the community name required to be supplied with SNMP requests to the system.
snmp-server contact <contact name>
no snmp-server contact
snmp-server location <system location>
no snmp-server location
Set the syscontact and syslocation variable served from the System MIB in MIB-II.
snmp-server host <hostname> traps [version <1,2c>] <community string>
no snmp-server host <hostname>
Add or remove a host to which SNMP traps should be sent. Note that this setting is only meaningful
if traps are enabled, though the list of hosts may still be edited if traps are disabled.
[no] snmp-server host <hostname> disable
Disable a trap sink without actually removing it altogether from the configuration. All trap sinks are
created enabled.
[no] snmp-server traps event <event name>
Specify which types of events should be sent as SNMP traps. By default the entire list of notifiable
events are sent as SNMP traps to any declared trap sinks. This command enables or disables a
single event for conversion to an SNMP trap.
show snmp
Display all SNMP configuration options.
366
535-0008-00 Rev 10
535-0008-00 Rev 10
367
368
535-0008-00 Rev 10
Alarm Reference
APPENDIX B
This chapter describes how you may use Memory Array tools to diagnose system
and VIMM-specific alarms. The chapter describes non-service affecting and
service affecting alarms that may occur during operation of the Violin Memory Array
and provides information on resolving those alarms in three sections.
Alarms are indicated by the state of the LEDs on the front of the Violin 6000
Series Memory Array chassis.
Information about an active alarm may then be viewed by running the show
alarms command.
Alarms for Violin 6000 Series Memory Array modules are also shown on the
Array Status screen in the Violin Web Interface. See Showing Status of the
Memory Array Modules on page 149 for more information.
This appendix provides information about how to resolve and clear these alarms
when they occur.
535-0008-00 Rev 10
369
AC-A
AC-B
Interface Status
ID LED
Button
System Status
VIMM Status
The System Status and VIMM Status LEDs alert you of System and VIMM-specific alarms by their
color and state: a solid amber light. When both LEDs are solid amber, an alarm has been registered
for one or more VIMMs. A solid amber System Status LED by itself indicates that an alarm has been
registered for an issue unrelated to a VIMM. See System LEDs on page 15 for information about
the LEDs not shown below.
LED
System Status
VIMM Status
Interface Status
LED State
Description
Green
Amber
Grey
Green
Amber
Green
Many alarms occur during bootup and automatically clear as soon as the system stabilizes. For
example, the Data Plane Unavailable alarm frequently occurs during bootup because the data
plane is not enabled until later in the process. Such alarms may be safely ignored.
370
535-0008-00 Rev 10
535-0008-00 Rev 10
371
Service-Affecting Alarms
Service-affecting alarms are alarms raised by the system when an error occurs that interrupts traffic
and indicates that the data plane is either severely impacted or completely disabled.
This section is organized into eight topics:
Note: Before performing a reboot, stop any activity that may be affected by the interruption, such
372
535-0008-00 Rev 10
Service-Affecting Alarms
Note: Before performing a reboot, stop any activity that may be affected by the interruption, such
535-0008-00 Rev 10
373
This should only occur in the unlikely event of both power supplies failing. At least one power supply
should be replaced. Contact Violin Memory Customer Support for assistance.
Note: Before performing a reboot, stop any activity that may be affected by the interruption, such
535-0008-00 Rev 10
Service-Affecting Alarms
Note: Before performing a reboot, stop any activity that may be affected by the interruption, such
535-0008-00 Rev 10
375
Note: Before performing a reboot, stop any activity that may be affected by the interruption, such
376
535-0008-00 Rev 10
535-0008-00 Rev 10
377
Caution: To avoid overheating and the generation of system alarms, the lid must not be off of
the system for more than 5 minutes when the ambient temperature is 35 C (95 F) and above.
See Replacing a Fan in the Violin 6000 Series Memory Array Service Guide.
High Temperature on Board Sensor Alarm
The High Temperature on Board Sensor alarm indicates that the specified board sensor on the
system is unreadable and that the system cannot retrieve system environment information.
To clear the alarm, check fan operation, the fan filter, and the ambient temperature of the system.
See Replacing a Fan in the Violin 6000 Series Memory Array Service Guide.
535-0008-00 Rev 10
Note: Before performing a reboot, stop any activity that may be affected by the interruption, such
379
Port Is Not Running at the Maximum Data Rate Possible Alarm on page 380
380
535-0008-00 Rev 10
RAID Rebuild of Group Completed with Errors (No Additional Information Available) Alarm on
page 381
535-0008-00 Rev 10
381
382
535-0008-00 Rev 10
The alarm indicates that the VIMM was shut down because it
became too hot.
Isolated
Dead
The alarm indicates that the amount of ECC errors has been
exceeded.
Incompatible memory
configuration
Formatted capacity is
incompatible with system
The alarm indicates that the ID of the identified VIMM could not
be set.
The alarm indicates that the number of bad blocks on the VIMM
has exceeded the safe operational threshold.
535-0008-00 Rev 10
383
The alarm indicates that the type of flash (MLC, SLC) is not
compatible with the other VIMMs in the system.
See Replacing a VIMM in the Violin 6000 Series Memory Array Service Guide.
VIMM Failed Alarm
The VIMM Failed alarm indicates that the specified VIMM has failed and is inoperable.
To clear the alarm, replace the VIMM. See Replacing a VIMM in the Violin 6000 Series Memory
Array Service Guide.
VIMM Failed: Excessive ECC Errors Alarm
The VIMM Failed: Excessive ECC Errors alarm indicates that a specific VIMM is taken out of service
because it experienced excessive ECC errors.
VIMM Hot Alarm
The VIMM Hot alarm indicates that the specified VIMM is reaching its maximum temperature within
the acceptable temperature range. The alarm specifies the VIMM ID and the current temperature
in Celsius. The VIMM shuts down if the temperature reaches the maximum acceptable temperature.
To clear the alarm:
384
535-0008-00 Rev 10
See Replacing a VIMM in the Violin 6000 Series Memory Array Service Guide.
VIMM Incompatible Formatted Capacity Alarm
The VIMM Incompatible Formatted Capacity alarm indicates that the specified VIMM has a
formatted capacity that is different from the rest of the system, and is thus not being used.
VIMM in Maintenance State Alarm
The VIMM in Maintenance State alarm indicates that the specified VIMM is in maintenance state.
To clear the alarm, Contact Violin Memory Customer Support for assistance.
VIMM Isolated Alarm
The VIMM Isolated alarm indicates that the specified VIMM is isolated and not reachable by any
other VIMM within the tree. This is caused by the failure of other VIMMs, preventing a path from the
main controller to the VIMM.
To clear the alarm, replace the failed VIMM(s). See Replacing a VIMM in the Violin 6000 Series
Memory Array Service Guide.
VIMM Temperature Unreadable, VIMM Shut Down Alarm
The VIMM Temperature Unreadable, VIMM Shut Down alarm indicates that the temperature for the
specified VIMM cannot be read and the VIMM has been shut down.
To clear the alarm, replace the VIMM. See Replacing a VIMM in the Violin 6000 Series Memory
Array Service Guide.
VIMM Unformatted Alarm
The VIMM Unformatted alarm indicates that the specified VIMM is unformatted, and is thus not
being used.
VIMM Threshold Exceeded Alarm
The VIMM Threshold Exceeded alarm indicates that a VIMM has been taken out of service due to
an exceeded monitored threshold.
VIMM Incompatible Formatted Capacity Alarm
The VIMM Incompatible Formatted Capacity alarm indicates that a VIMM has been formatted to a
capacity that is incompatible with the system. The VIMM ID is displayed in the alarm.
535-0008-00 Rev 10
385
386
535-0008-00 Rev 10
Violin Drivers
APPENDIX C
Note: See the Violin Array Controller Driver for Windows Installation Guide for
instructions on installing and using the driver for Windows. The guide is available
from the Violin Memory Customer Support site.
Topics covered include system prerequisites, the installation of the Violin drivers,
the installation of the Violin utilities, and the creation of block device partitions.
535-0008-00 Rev 10
387
The default and preferred driver mode. Allows for file system creation,
mounting, and optional host buffer cache usage.
In general, a Violin Linux driver provides block device access to the Memory Array enabling the host
to create and mount block storage.
This chapter describes the steps required to install and configure Violin Linux drivers as block
devices only. If you plan to use the Violin Linux driver to enable character or SCSI connections,
contact Violin Memory Customer Support.
Step 1: Installing Violin Linux Driver Prerequisites: Ensure that all of the required
packages are installed on the Linux host before you install and build the Violin Linux driver.
Step 2: Installing Violin Linux Drivers: Install the Violin Linux driver on the host, build the
driver, and configure the host to load the Violin Linux driver manually or as a module.
Step 3: Loading the Violin Linux drivers: Depending upon where the Linux drivers were
installed, they may be loaded manually or automatically as modules when the host boots up.
Step 4: Installing Violin Utilities: Violin utility commands enable you to monitor and fine tune
the performance of the Memory Array.
Step 5: Configuring Partitions and File Systems: If the Violin Linux driver is used as a block
device, you may create partitions and file systems.
388
535-0008-00 Rev 10
GCC Compiler
The GCC Compiler compiles the Linux kernel and modules. The GCC Compiler version should
match the version that was used to build the running kernel.
535-0008-00 Rev 10
389
the host computer before you install and load the Violin Linux driver onto the host.
Because the PCIe Hot Plug is not supported by most current BIOS versions, you must power
up the Memory Array before powering up the host computer.
2. Log into the host computer as root.
3. Download the Violin Linux install file (vtms-linux-driver<release>.run) from the Violin
./vtms-linux-driver-<release>.run
390
535-0008-00 Rev 10
The script begins installing the driver, as shown in the following example:
# ./vtms-linux-driver-D6.0.0.3.run
Copying to a temporary location...
Creating directory vtms-linux-driver-D6.0.0.3
Verifying archive integrity... All good.
Uncompressing Violin Memory, Inc. Linux Driver
source..............................
Current settings detected:
------------------------------------------------Kernel
= 2.6.32-100.28.5.el6.x86_64
Kernel Dir = /lib/modules/2.6.32-100.28.5.el6.x86_64/build
Kernel GCC = 4.4.4
GCC ver
= 4.4.4
Arch
= x86_64
View the driver README.txt? [n]:
Skipping /tmp/vtms-linux-driver-D6.0.0.3/README.txt
7. Type y if you want to view the README.txt.
535-0008-00 Rev 10
391
directory.
If you select yes, the Violin Linux driver module is added to the kernel modules.
If you select no, the Violin Linux driver is not installed to the /lib/modules directory.
A message appears stating that the Violin Linux driver installation is complete.
10.Once the installation is complete you may load the Violin Linux driver for use.
392
535-0008-00 Rev 10
If you installed the Violin Linux driver in the /lib/modules directory, see Loading Violin
Linux Drivers as Modules on page 395 for detailed instructions.
If you did not install the Violin Linux driver in the /lib/modules directory, see Loading
Violin Linux Drivers Manually on page 394.
cd /tmp/vtms-linux-driver-<release>
make uninstall
The Violin Linux driver is removed from the /lib/modules directory, but can still be loaded and
unloaded using the steps described in Loading and Unloading Linux Drivers on page 394.
Load the Violin Linux driver again if you want to use the Memory Array. See Installing Violin Linux
Drivers on page 390 for details on installing the Violin Linux driver.
535-0008-00 Rev 10
393
the host computer before installing and loading the Violin Linux driver onto the host computer.
2. Log into the host computer as root.
3. At the system prompt, navigate to the following directory:
cd /tmp/vtms-linux-driver-<release>/<kernel_version>/
<architecture_type>
Where:
<release>
<kernel_version>
<architecture_type>
./load_strad.sh
5. At any point, if you decide to install the Violin Linux driver in /lib/modules at the system
prompt, type:
cd /tmp/vtms-linux-driver-<release>
make install
depmod
394
535-0008-00 Rev 10
Where:
Indicates the release number, such as 4.5.4.1.
<release>
the host computer before installing and loading the Violin Linux driver onto the host computer.
2. Log into the Linux host as root.
3. Open a terminal window.
4.
(Optional) To update the dependency mappings of newly installed modules, at the system
prompt, type:
depmod
You may need to run depmod for modprobe to detect the new module.
[use_cdev=1]
[use_scsi=1]
After loading the Violin Linux driver, create partitions and file systems on the Memory Array or
directly access the character device.
rmmod vtms
535-0008-00 Rev 10
395
Utility
Description
perf_test
vcounts
veeprom
Displays information, such as the main board serial number and the
management MAC address.
vinfo
vpartial
vstat
vzero
Table C.1 Memory Array Commands for Viewing Violin Linux Driver Information
For the complete syntax and usage details of each command, see Violin Utilities on page 403.
On each command, specify the optional device index ( [ <device_index> ] ) to list the
information for a particular Memory Array. If you do not specify an individual device, the command
lists the information for all the Memory Arrays found.
(vtms-linux-<release>.tar).
396
535-0008-00 Rev 10
./vtms-linux-utils-<release>.run
Where <release> indicates the release number, such as D6.0.0.3
The following command is only needed when splitting the block device into multiple partitions.
2. First, open the device for adding partitions by typing the following at the command prompt:
535-0008-00 Rev 10
397
p
Where p creates a primary partition.
The command displays information similar to the following:
Command action
e
extended
p
primary partition (1-4)
p
Partition number (1-4): 1
First cylinder (1-16709, default 1):
Using default value 1
Last cylinder or +size or +sizeM or +sizeK (1-16709, default 16709): 8000
Specify the partition number if you will be creating more than one partition on the device,
or accept the default value of 1 by pressing the Enter key. The default value to set the
beginning cylinder to is 1, which can be changed if needed.
Specify the size of the block device or accept the default value of the available memory by
pressing the Enter key.
w
Where w writes the partition table to disk and exit.
The command displays information similar to the following:
Command (m for help): w
The partition table has been altered!
Calling ioctl() to re-read partition table.
Syncing disks.
#
The system command prompt is returned. Finish by creating the file system that uses the new
block device partition, creating the new mount path, and mounting the new block device file
system.
398
535-0008-00 Rev 10
/dev/vtmsa<n>
mkdir /mnt/strad
Where /mnt/strad specifies the name of the new mount path.
8. At the command prompt, mount the new file system by typing:
535-0008-00 Rev 10
399
Where:
-t ext3
/dev/vtmsa<n>
Specifies the name of the new file system, where <n> is the assigned
device number.
/mnt/strad
umount /mnt/strad
Where:
/dev/
<partition>
/mnt/
<mount_pth>
<fs_type>
<mount_opt>
400
535-0008-00 Rev 10
<dump_opt>
<fsck_opt>
<release>.tgz, where <release> indicates the release number; for example, D6.0.0.3.
The expanded file contains the following two directories:
lvm
openfiler
following:
sudo ./vtms-lvm.patch.sh
535-0008-00 Rev 10
401
402
535-0008-00 Rev 10
APPENDIX D
Violin Utilities
This chapter describes Violin utilities for configuring and performance testing the
Violin 6000 Series Memory Array system. These commands are common to the
Violin Linux driver and to the Violin Memory Gateway.
535-0008-00 Rev 10
403
do not specify an individual device, the command lists the information for all Memory Arrays found.
Running Violin Utilities on a Linux Host
Violin utilities are run on the Linux host from the command line.
Running Violin Utilities on a Windows Host
A shortcut icon is created on the Windows desktop to the Violin Utilities command shell when you
install the Violin Utilities on the Windows host.
If you are not logged on as Administrator, right-click the icon and then select Run As
Administrator to open the shell with the proper permissions.
Running Violin Utilities from the Memory Array
Violin utilities must be run from the Master Memory Gateway. The following is an example of the
steps required to log in to the internal Master Memory Gateway on a Violin 6000 Series Memory
Array.
1. Start a terminal application and log in to the Master ACM as admin. The Master ACM is set
up during initial configuration of the Violin 6000 Series Memory Array. Enter the system host
name or the ACM management IP address to ensure a direct connection.
2. Once logged in to the Master ACM, type enable at the command prompt:
404
535-0008-00 Rev 10
varray
The varray command displays array information, including number of VIMMs, temperature, fan
speed and LED states.
Syntax
varray
535-0008-00 Rev 10
405
Example
The varray utility returns information similar to the following when run on a Linux host:
violin-mga [violin: master] (config) # varray
Violin Memory, Inc.
Version: vtms-linux-utils-V6.0.0 03/13/2013
V6000
64
20C
42C
ON
ON
242745 secs
0 secs
Slow
Slow
Slow
Slow
Slow
Slow
OFF
ON
ON
ON
vcounts
The vcounts utility displays data transfer counters for the Memory Array.
Syntax
vcounts [ <device_index> ]
406
535-0008-00 Rev 10
Example
The vcounts utility returns information similar to the following when run on a Linux host:
# vcounts
Violin Memory, Inc.
Version: vtms-linux-utils-V6.0.0 03/13/2013
Device:
Index:
/dev/vtmsa
0
-- Target Counts
IO
> 128K
64K < IO <=128K
32K < IO <=64K
16K < IO <=32K
8K
< IO <=16K
4K
< IO <=8K
IO
=
4K
IO
<
4K
-:
:
:
:
:
:
:
:
IO requests
:
IO requests completed
active io sent to VCMs
IO requests failed
IO zero-size requests
0
0
49
2
52
282
315193111
2076
:
:
:
:
535-0008-00 Rev 10
315212448
315195572
0
0
16876
205897852
201347061
0
1291037701120
645633350656
645404350464
315195935
157626576
157569359
0
0
4095
4096
1971
3
3
0
157627156
157569436
2112
17
407
Returns
The utility returns the following counters:
408
IRQ calls
The total calls to the Memory Array device driver where work
was done.
Completed I/O's
The total I/O read requests from a Memory Array. This is not the
individual DMA descriptors completed, but for each of the user
requested I/Os.
The total I/O write requests to a Memory Array. This is not the
individual DMA descriptors completed, but for each of the user
requested I/Os.
The total failed I/O read requests from a Memory Array. This is
not the individual DMA descriptors failed, but for each of the
user requested I/Os.
The total failed I/O write requests to a Memory Array. This is not
the individual DMA descriptors failed, but for each of the user
requested I/Os.
The total I/O read requests from a Memory Array, but only
incremented when an unaligned host address required special
buffer byte copying to service the DMA request.
535-0008-00 Rev 10
veeprom
The veeprom utility displays the Memory Array hardware information such as the main board serial
number, MAC address of the management interface, and so on.
Syntax
veeprom [ <device_index> ]
Linux Example
The veeprom utility returns information similar to the following example:
# veeprom
Violin Memory, Inc.
Version: vtms-linux-utils-V6.0.0 03/13/2013
Device:
Proc device:
Index:
/dev/vtmsa
/proc/driver/vtms/strad0
0
ee_partnum
ee_serialnum
ee_boardver
ee_mfgdate
ee_mgmtmac
535-0008-00 Rev 10
409
vincident -a
The vincident script collects useful information from the host servers and array such as version/
timestamp of the current kernel, CPU information, partition information, Memory Array
configuration, and Memory Array logs.
Once collected, this information can be sent to Violin Memory Customer Support for analysis to
determine the source of performance issues, such as ECC errors.
The vincident script is installed on the host server as part of the Violin Utility package. For more
information, see Installing Violin Utilities on page 403 for Linux hosts, or Installing Violin Utilities on
page 403 for Windows hosts.
Syntax
vincident {-a | <tty_device> | <ip_address> } [--max-timeout secs]
vincident may be run using the IP address of either Memory Array VCM (<ip_address>) or by
specifying the VCM name (for example, "vincident vcm-a"). You can also use a serial cable
connected from the host server to the Memory Array (<tty_device>).
The vincident utility creates an incident report in the current directory with a name similar to
vincident.20110624T110149.
In the event that the full path name of the vincident utility is not in the root PATH, the full path
of vincident is /opt/violin/bin/vincident.
Example
# vincident -a
Gathering information from host...
Gathering information from target...
Target IP = vcm-a
This could take a couple of minutes, please be patient...
Target IP = vcm-b
This could take a couple of minutes, please be patient...
Target IP = vcm-c
This could take a couple of minutes, please be patient...
Target IP = vcm-d
This could take a couple of minutes, please be patient...
Incident report created in vincident.20121001T153541
vinfo
The vinfo utility displays the type and version for a flash-based Memory Array.
Syntax
vinfo [ <device_index> ]
410
535-0008-00 Rev 10
Example
The vinfo utility returns information similar to the following example when run on a Windows host:
# vinfo
Violin Memory, Inc.
Version: vtms-linux-utils-V6.0.0 03/13/2013
Device:
Index:
/dev/vtmsa
0
vtms-linux-driver-V6.0.0
Mar 03 2013 11:15:03
A5.5.0 #1-rc (d442848)
1
strad0
strad1
strad2
Unknown0 Unknown1 Unknown2
4
12288
23109R00000016
8607114461184 bytes
256GB FLASH VIMMs
12 (60-VIMM)
128 bytes
4096
30
512 bytes
0
0
0x0
Unknown
Unknown
10.1.14.140
0.0.0.0
0.0.0.0
VCM Devices:
Device:
Memory(bytes):
RAID groups:
irqtune:
debug:
Serial #:
VCM-A
strad0
8607114461184
12 (60-VIMM)
96
0x0
2710BN00000090
VCM-B
strad1
8607114461184
12 (60-VIMM)
96
0x0
2710BN00000079
strad3
Unknown3
VCM-C
strad2
8607114461184
12 (60-VIMM)
96
0x0
2710BN00000104
VCM-D
strad3
8607114461184
12 (60-VIMM)
96
0x0
2710BN00000094
Returns
The utility returns the following information.
Host Driver
Driver Date
Displays block mode only. When set to 1, allows READA read-ahead I/Os
to be accepted.
Target S/W
535-0008-00 Rev 10
411
Memory
Displays the size in bytes of usable system capacity. For flash VIMMs, this
value changes based on formatted capacity.
Memory Type
RAID groups
Displays the number of 5-VIMM RAID groups. Spare VIMMs are not
counted.
Granularity
RingSize
Displays the size of driver DMA descriptor ring per Memory Array.
Must be power of 2 with range of 2 - 4096.
IrqTune
Displays the Interrupt combining tunable with 0 = disabled and 4095 being
the highest value.
IoTimeout
Displays the time in seconds before the device driver declares an I/O as
stuck and disables Memory Array I/O access. 0 = disables timeout.
NoMSI
When set to 1, specifies that the driver will not attempt to allocate a PCIe
MSI-based interrupt vector.
Debug
Serial #
Displays the Memory Array serial number stored on its EEPROM and also
shown on the label on the back of the unit.
Mgmt. MAC
Displays the Memory Array Ethernet port MAC address, useful for adding
into a DHCP server configuration file.
vinventory
The vinventory utility scans the board and displays the main board inventory. The output displays
all modules (present or not), their state, serial and model numbers.
Syntax
vinventory
412
535-0008-00 Rev 10
vmesg
The vmesg utility displays kernel log messages, and is primarily used for debugging purposes.
Syntax
vmesg
vpartial
The vpartial utility displays the number of read/write I/O requests processed and the number of
partial 4kB flash pages.
Syntax
vpartial
535-0008-00 Rev 10
[ <device_index> ]
413
Example
The vpartial utility returns information similar to the following example when run on a Linux host:
# vpartial
Violin Memory, Inc.
Version: vtms-linux-utils-V6.0.0 03/13/2013
Device:
Index:
/dev/vtmsa
0
157570174
3
0
157570251
17
Returns
The utility returns following information.
414
535-0008-00 Rev 10
vring
The vring utility debugs the internal Memory Array I/O request ring at a low level, and checks for
unaligned flash device access.
Look for the transfer sizes under the column labeled SIZE. If most of the lines show 4096, full 4kB
accesses are being done to the Memory Array hardware, which is optimal. In an unaligned access
case, you will see lines alternate between 512 and 3584 for transfer size, since two read-modifywrite operations occur for each 4kB of data.
Syntax
vring [ -p <partition_name> ]
Example
The vring utility returns information similar to the following example when run on a Linux host:
# vring
Violin Memory, Inc.
Version: vtms-linux-utils-V6.0.0 03/13/2013
IDevice:
Proc device:
Index:
/dev/vtmsa
/proc/driver/vtms/strad0
0
HOST_ADDR
0x00000005a56c9000
0x0000000313bdc000
0x0000000313bdc000
0x000000031212b000
0x00000005a56c9000
SIZE
4096
4096
4096
4096
4096
Parameters
Parameter
Description
-p
<partition_name>
vspeedtest
The vspeedtest utility displays a quick speed test of the Array.
Syntax
vspeedtest
535-0008-00 Rev 10
415
Example
# vspeedtest
Checking speed on array device id ataVIOLIN_MEMORY_ARRAY_23109R00000016
3609.854 Read MB/s, 881311 Read IOPS
vstat
The vstat utility displays the status of the connection and the ready status of a Memory Array.
Syntax
vstat [ <device_index> ]
416
535-0008-00 Rev 10
Example
The vstat utility returns information similar to the following example when run on a Linux host:
violin-acmb-acm-b [violin: master] (config) # vstat
Violin Memory, Inc.
Version: vtms-linux-utils-V6.0.0 03/13/2013
Device:
Index:
/dev/vtmsa
0
VCM-A
OFF
1
0
0
0
1
0
0
4
4
256
1024
4096
0
0
0
0
VCM-B
OFF
1
0
0
0
1
0
0
4
4
256
1024
4096
0
0
0
0
VCM-C
OFF
1
0
0
0
1
0
0
4
4
256
1024
4096
0
0
0
0
VCM-D
OFF
1
0
0
0
1
0
0
4
4
256
1024
4096
0
0
0
0
Returns
The utility returns the following information.
Status LED
Alarm LED
Indicates whether the Alarm LED is on or not. If it is on, it indicates the status
of the LED flashing.
PWR_A LED
PWR_B LED
ready
535-0008-00 Rev 10
417
418
formatting
format_done
paused
link
lid_ajar
Indicates whether the lid is closed or not. To avoid overheating and the
generation of system alarms, the lid must not be off of the system for more
than 5 minutes when the ambient temperature is 35 C (95 F) and above.
raid_rebuild
write_buffer
linkwidth
maxlinkwidth
cur_payload
max_payload
cur_read_req
dma_active
io_pend
Indicates the number of I/O requests in the queue for a Violin 6000 Series
Memory Array.
A single I/O request may involve more than one 4kB DMA descriptor.
535-0008-00 Rev 10
vvimms
Use this command to display information about the Memory Array VIMMs.
Syntax
vvimms
Example
# vvimms
Violin Memory, Inc.
Version: vtms-linux-utils-V6.0.0 03/13/2013
Device:
Index:
/dev/vtmsa
0
VCM
A
A
A
A
A
B
B
B
B
B
C
C
C
C
C
D
D
D
D
D
vzero
The vzero utility resets the I/O counters to zero.
Syntax
vzero [ <device_index> ]
535-0008-00 Rev 10
419
Example
The vzero utility returns information similar to the following example when run on a Linux host:
# vzero
Violin Memory, Inc.
Version: vtms-linux-utils-V6.0.0 03/13/2013
Zeroed counters for Violin index 0
perf_test
The perf_test utility is a lightweight yet powerful tool for measuring the I/O performance of
storage devices.
The utility is designed to push the I/O device as hard as possible by a variety of methods including
POSIX threads, variable block sizing, asynchronous I/O, direct communication, as well as
sequential and random read/write operations to data over a user-defined address range.
Syntax
perf_test
Example
By default, the perf_test accesses the first Memory Array device using a single thread for a
repeated 4kB read of sector 0. For best results, use the -R option to randomly access a specific
range (-s) using multiple threads (-t), direct I/Os (-o), and asynchronous I/Os (-A).
The perf_test utility returns information similar to the following example when run on a Linux
host:
# perf_test -s 640 -R -A64 -t8 -o
Violin Memory, Inc.
Version: vtms-linux-utils-V6.0.0 03/13/2013
Command: perf_test -s 640 -R -A64 -t8 -o
Running with options:
threads
block_size
memory size
memory start addr
write:read ratio
random mode
=
=
=
=
=
=
duration
path
MB
no-cache mode
AIO depth
= 0 secs
= /dev/vtms-strad
= 1000000 bytes
= 1
= 64
420
8
4096
640 GB
0x0
0:1
1
535-0008-00 Rev 10
817.181
817.132
817.160
817.096
MB/s0.000
MB/s0.000
MB/s0.000
MB/s0.000
MB/s817.181
MB/s817.132
MB/s817.160
MB/s817.096
MB/s199507001995070
MB/s1994950199495
MB/s1995010199501
MB/s1994860199486
For a discussion of using the perf_test utility in Violin Memory Array performance tuning, see
Performance Testing Using Perf_Test on page 428.
Options
Option
Description
-h | -?
Displays help.
-b block-size
Sets the size of each block, in bytes, such as 128, 512, 1024, and so
on.
Recognized unit/size qualifiers include: k and K for the appropriate
binary multipliers.
-B
[low,high[,step]]
-r read-weight
Specifies the read weight, relative to the write weight. Specify 0 to not
do any reads at all.
-w write-weight
Specifies the write weight, relative to the read weight. Specify 0 to not
do any writes at all.
-s memory-size
-d duration
-p device-path
-t num-threads
-A async-io-depth
-a address
535-0008-00 Rev 10
421
Option
Description
-c char
-V
Verifies writes with a read to ensure data integrity. This option can
have a significant impact on overall I/O throughput.
-I
-R
-C
-q
-z
-L
-u
-o
422
535-0008-00 Rev 10
APPENDIX E
This chapter provides an overview of Violin Memory Array performance tuning and
testing, describes flash-layer performance tuning, and network-layer performance
testing techniques using the perf_test utility.
535-0008-00 Rev 10
File System-layer: The file system and its settings such as block size,
snapshots, and logging
423
Database-layer: The database (if used) and its configuration and optimizations
In general, performance tuning and testing is a six-step process by which each layer of the Memory
Array system is tuned before moving on to the next layer in the system.
Performance tuning and testing techniques for the higher system levels are beyond the scope of
this book. This chapter describes techniques for tuning and testing the first two layers in the Memory
Array system: the flash layer and the network layer.
Step 1: Flash-layer performance tuning: Ensure that the file system and operating system
are 4kB aligned. The vpartial and vring utilities enable you to identify 4kB alignment
issues.
After the flash and network layers have been configured and tested, you may test the performance
of the higher system layers.
While these tests are operating, use operating system tools to confirm that actual I/O activity is
similar to the levels tested using the perf_test utility. If the I/O is not similar to the baselines, then
the operating system, file system, or database may need to be tuned. Please contact Violin Memory
Customer Support for assistance if needed.
Type of VIMM: Single level cell (SLC) or multi-level cell (MLC): SLC is faster for writes
Number of VIMMs per Memory Array: More VIMMs usually deliver more performance.
Formatted Capacity: 65% format is faster than 84%
Number of Violin Memory Arrays: Like all storage media, more arrays provides higher
bandwidth.
Flash-based memory arrays are designed for 4kB block access or any multiple of 4kB blocks.
Smaller block sizes (e.g., 512 bytes) will significantly reduce performance, particularly for writes.
Therefore, it is import that you verify whether the file system and operating system are 4kB aligned.
For more information see Memory Array Flash 4kB Alignment on page 425.
Random workloads with high IOPS are well supported by flash memory. Unlike hard disk drives,
sequential accesses are not significantly faster. For this reason, it is typically better that higher
layers are tuned for 4kB-16kB block sizes and high IOPS.
Where an application needs very high bandwidths, multiple memory arrays are recommended.
Flash Memory Arrays achieve maximum performance when large numbers of flash devices are
used simultaneously, which requires a large number of simultaneous accesses by the application
and file system. Typically, this is enabled through asynchronous I/O and having many threads
simultaneously access the flash memory. Where this is not the case, this is known as singlethreaded access and severely limits performance.
424
535-0008-00 Rev 10
vtmsa
0
For more information on the vpartial utility, see vmesg on page 413.
Checking Flash 4kB Alignment and Request Ring
The vring utility debugs the internal Memory Array I/O request ring at a low level, and checks for
unaligned flash device access.
Look for the transfer sizes under the column labeled SIZE. If most of the lines show 4096, full 4kB
accesses are being done to the Memory Array hardware, which is optimal. In an unaligned access
535-0008-00 Rev 10
425
case, you will see lines alternate between 512 and 3584 for transfer size, since two Read-ModifyWrite operations occur for each 4kB of data.
# vring
Violin Memory, Inc.
Version: vtms-linux-utils-D5.2.0.4, 05/23/2012
INDX
0
1
2
3
4
...
CMD/FLAGS
0x00000000
0x00000000
0x00000000
0x00020000
0x01020000
TRGT_ADDR
0x0000001b32acc000
0x0000001b32acd000
0x0000001b32ace000
0x0000001b32acf000
0x00000033eb48c000
HOST_ADDR
0x0000000044736000
0x0000000063e77000
0x0000000022bb8000
0x0000000079639000
0x00000000865bb000
SIZE
4096
4096
4096
4096
4096
The vring utility requires access to /dev/vtms-strad, which is a symbolic link to the first
discovered Memory Array.
If more than one Memory Array is connected, the devices can be explicitly selected by using the
partition 0 control device. For example, if the second Memory Array is vtms1, then the command is:
vring -p /dev/rdsk/vtms1v0
For more information on the vring utility, see vring on page 415.
Addressing 4kB Alignment Issues on Linux Hosts
On Linux systems, the fdisk utility with the -S option enables you to override the default geometry
of 63 sectors per track and specify a different number of sectors for use by the partition tables.
Running fdisk -S56 provides 4kB alignment automatically on created partitions. For example,
fdisk -S56 /dev/vtmsa
Addressing 4kB Alignment Issues on Windows Hosts
Windows Server 2008 and Windows 7 Servers
On Windows 2003 and Windows XP Servers, 4kB addresses may be realigned using the
diskpart hard disk partitioning utility.
1. If the disk you are aligning is already blank (raw), go to Step 3. If the disk contains data, back
426
535-0008-00 Rev 10
If the disk you want to align does not appear in the list, make sure that it exists and is accessible
using the Disk Management snap-in.
5. At the diskpart command prompt, type Select Disk X.
Where x is the number of the disk as shown in the output of the List Disk command. The
diskpart utility should return a message that indicates that Disk x is the selected disk.
6. At the diskpart command prompt, type Create Partition Primary Align=64.
7. At the diskpart command prompt, type Assign Letter=<DriveLetter>.
8. After the drive letter is assigned, type exit to exit the diskpart utility.
9. Use the Disk Management snap-in or the Format command to format the partition as an
NTFS-formatted partition.
NOTE: when formatting, select an Allocation Unit Size that is a multiple of 4kBs.
Reference: Microsoft Technet article http://technet.microsoft.com/en-us/library/
aa998219.aspx
535-0008-00 Rev 10
427
428
535-0008-00 Rev 10
For example, the command perf_test -s 640 -R -A64 -t8 -o returns the following example
when run on a Linux host:
# perf_test -s10 -R -A64 -t8 -o
Violin Memory, Inc.
Version: vtms-linux-utils-D6.0.0.3, 03/07/2013
Command: perf_test -s10 -R -A64 -t8 -o
Running with options:
threads
block_size
memory size
memory start addr
write:read ratio
random mode
duration
seconds to skip
path
MB
fill pattern
no-cache mode
AIO depth
=
=
=
=
=
=
=
=
=
=
=
=
=
Time (sec)
Write BW
1
2
3
4
5
6
7
8
9
Average
Read BW
3811.958
4010.030
3809.862
3824.481
3707.888
3699.851
3991.936
3925.807
3997.503
MB/s
MB/s
MB/s
MB/s
MB/s
MB/s
MB/s
MB/s
MB/s
3864.368 MB/s
8
4096
10 GB
0x0
0:1
1
0 secs
0 secs
/dev/vtms-strad
1000000 bytes
0x0
1
64
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
MB/s
MB/s
MB/s
MB/s
MB/s
MB/s
MB/s
MB/s
MB/s
0.000 MB/s
Total BW
3811.958
4010.030
3809.862
3824.481
3707.888
3699.851
3991.936
3925.807
3997.503
MB/s
MB/s
MB/s
MB/s
MB/s
MB/s
MB/s
MB/s
MB/s
3864.368 MB/s
Read IOPS
Write IOPS
Total IOPS
0
0
0
0
0
0
0
0
0
930653
979011
930142
933711
905246
903284
974593
958448
975952
943448
930653
979011
930142
933711
905246
903284
974593
958448
975952
943448
The following table describes the options used in this command. For a complete list of perf_test
utility options, see perf_test on page 420.
-s memory-size
-t num-threads
-R
535-0008-00 Rev 10
429
-o
-A async-io-depth
For best results, take the following factors into consideration when running the perf_test utility.
Memory Access Patterns
The perf_test memory access modes include random, sequential, and single-address.
Accessing just a single address can radically impact the performance a device can deliver, as can
sequential and random access. So, know which address access mode is being used when running
tests.
Threading and Asynchronous I/O
Use perf_test to specify the number of threads to spawn and the number of outstanding I/Os
within each request.
The Linux operating system tends to deal with larger asynchronous I/O depths better than a larger
number of threads. High amounts of I/O can be achieved by understanding the operating system
behavior and using a combination of thread counts and asynchronous I/O.
Block Size
Consider the block size when testing for maximal bandwidth versus maximal IOPS. Use a larger
block size (megabytes, if necessary) when evaluating bandwidth performance and smaller block
sizes, such as 4kB, when evaluating a devices maximum IOPS potential.
Latency
Use perf_test to measure round-trip latency of an I/O request. Generally, latency measurements
should only be made using one thread and an asynchronous I/O depth of one. Any other values
skew the latency measurements.
Direct-Attached and Buffered I/O
Use perf_test defaults to use the system buffer cache. In some cases, such as sequential read
tests on systems with a large amount of RAM, this can be an advantage because data read from
the device can be cached in the systems buffer cache.
However, that does not help in qualifying the speed of the device itself. To better qualify the speed
of the device being tested, use the -o option, which tells perf_test to use the O_DIRECT flag
when attaching to the device, thus circumventing the buffer cache.
Specify less threads using the -t option for Windows and Linux drivers and specify more
asynchronous I/Os using the -A option, as it reduces CPU context switching.
On Linux, the -o option is only available if the libaio and libaio-devel packages were
installed with the Violin Utilities installation. See Libaio Development Package on page 389 for more
information.
430
535-0008-00 Rev 10
Each of these tests should be run using asynchronous I/O settings (-A) and a range (-t) of thread
counts (for example, 1- 64).
The baselines established using perf_test may enable you to identify performance issues at the
operating system-layer, file system-layer, or database-layer of the Memory Array system.
Testing Read Bandwidth
By default, the perf_test accesses the first Memory Array device using a single thread for a
repeated 4kB read of sector 0.
The following example shows a command that uses the (-s), (-R), (-b), (-d), and (-p) options to
check the read bandwidth of the Violin 6000 Series Memory Array.
# perf_test -s 384 -R -t4 -b 16384 -d5 -p /dev/rdsk/vtms2v1
Running with options:
threads
= 4
block_size
= 16384
memory size
= 393216 MB
memory start addr
= 0 MB
write:read ratio
= 0:1
random mode
= 1
duration
= 5 secs
path
= /dev/rdsk/vtms2v1
MB
= 1000000 bytes
Time (sec) Read BW
Write BW
Total BW
Read IOPS
1
1259.292 MB/s 0.000 MB/s 1259.292 MB/s 76861
2
1258.271 MB/s 0.000 MB/s 1258.271 MB/s 76798
3
1259.717 MB/s 0.000 MB/s 1259.717 MB/s 76887
4
1258.863 MB/s 0.000 MB/s 1258.863 MB/s 76834
5
1258.300 MB/s 0.000 MB/s 1258.300 MB/s 76800
Average
1258.889 MB/s 0.000 MB/s 1258.889 MB/s 76836
535-0008-00 Rev 10
Write IOPS
0
0
0
0
0
0
Total IOPS
76861
76798
76887
76834
76800
76836
431
The example shows that the read bandwidth is roughly 1.2 GBps, which is what you might expect
on a PCIe bus with the payload set to the lowest setting of 128-bytes.
Testing Write Bandwidth
To check the write bandwidth, use the -w and -r options to define the ratio of writes to reads at 1:0.
This ratio enables you to determine the random percent of 1/Os that are writes.
In the following example, a performance test is run to check write bandwidth using the -w1 -r0
options to place the write to read ratio at 1:0.
# perf_test -s 384 -R
Running with options:
threads
block_size
memory size
memory start addr
write:read ratio
random mode
duration
path
MB
Time (sec) Read BW
1
0.000 MB/s
2
0.000 MB/s
3
0.000 MB/s
4
0.000 MB/s
5
0.000 MB/s
Average
0.000 MB/s
MB/s
MB/s
MB/s
MB/s
MB/s
MB/s
Total IOPS
61397
61396
61417
61407
61437
61410
The Violin 6000 Series Memory Array write bandwidth reaches a maximum of roughly 1 GB/s on x8
PCIe.
432
535-0008-00 Rev 10
SNMP Traps
APPENDIX F
This chapter focuses on SNMP use with the Violin 6000 Series Memory Array, and
covers the following topics:
Note: You must use the MIB associated with the product version. MIBs are not
guaranteed to be backwards/forwards compatible from version to version at this
time.
535-0008-00 Rev 10
433
MIB files are available for direct download from the Violin Memory Array Web UI Administration >
Information page. Each file is described in the table below.
This file...
VIOLIN-MEMORY-ROOT-MIB.txt
VIOLIN-MEMORY-ARRAY-MIB.txt
VIOLIN-MEMORY-MEDIA-MIB.txt
VIOLIN-MEMORY-TRAP-MIB.txt
TallMaple-MIB.txt
Below are example URLs for MIBs on a Violin Memory Array named HOSTNAME:
http://HOSTNAME/doc/VIOLIN-MEMORY-MIB.txt
http://HOSTNAME/doc/TallMaple-MIB.txt
Traps:
http://HOSTNAME/doc/VIOLIN-MEMORY-TRAP-MIB.txt
434
535-0008-00 Rev 10
public
yes
public
535-0008-00 Rev 10
435
Trap/Notify Setting
user-login
user-logout
process-exit
liveness-failure
cpu-util-high
cpu-util-ok
436
Class.
Logged Event
Event
Event
Event
Event
Event
Event
/mgmtd/session/events/login
/mgmtd/session/events/logout
/pm/events/failure/unexpected_exit
/pm/events/failure/liveness
/stats/events/cpu_util_indiv/rising/error
/stats/events/cpu_util_indiv/rising/clear
Event
Event
/stats/events/paging/rising/error
/stats/events/paging/rising/clear
Event
Event
Event
Event
/stats/events/fs_mnt_bytes/falling/error
/stats/events/fs_mnt_bytes/falling/clear
/vshare/events/array/free_space_trigger
/stats/events/memory_pct_used/rising/error
Event
Event
/stats/events/memory_pct_used/rising/clear
/stats/events/intf_util/rising/error
Event
Event
Event
Event
/stats/events/intf_util/rising/clear
/stats/events/disk_io/rising/error
/stats/events/disk_io/rising/clear
/cluster/events/unexpected_join
Event
/cluster/events/unexpected_leave
Event
/cluster/events/unexpected_size
Failure /smart/events/warning
Event
/crd/events/internal_fault
Failure
Failure
Event
Event
Event
/hwmon/events/failure/filesystem
/system/events/unexpected_shutdown
/net/interface/events/link_down
/net/interface/events/link_up
/hwmon/events/failure/mce
Event
/media/events/device/health_warning
Event
/media/events/device/state_change
Event
/media/events/device/state_change
535-0008-00 Rev 10
media-device-unknowntype
media-device-lifetimewarn
media-system-swapstate
media-system-swapstate
excessive-shutdowns
Excessive unexpected
shutdowns
license-state-change
A license for %s %s
conntrack-entries-high
Too many connection
tracking entries
conntrack-entries-ok
Connection tracking
entries alarm clearing
vimm-temperature-high
VIMM temperature
alarm detected
vimm-temperature-ok
VIMM temperature
alarm clearing
chassis-temperature-high Chassis temperature
alarm detected
chassis-temperature-ok
Chassis temperature
alarm clearing
lid-ajar-time-high
Excessive lid ajar time
detected
lid-ajar-time-ok
Lid ajar alarm clearing
lid-ajar
Array lid ajar state
change detected
array-pcie-link-up
Array PCIE link up
detected
array-pcie-link-down
Array PCIE link down
detected
array-data-plane-ready
Array data plane ready
state change detected
array-raid-rebuild
Array RAID rebuild state
change detected
vimm-state-change
Array VIMM state
change detected
array-psu-state
Array PSU state change
detected
array-led-change
Array LED state change
detected
array-fan-change
Array FAN state change
detected
fc-port-state-change
vSHARE FC port state
change detected
ib-port-state-change
vSHARE IB port state
change detected
vmg-fan-change
Gateway FAN state
change detected
535-0008-00 Rev 10
Event
/media/events/device/unknown_device_type
Event
/media/events/device/life_remain_warning
Event
/media/events/system/swap_state_change
Event
/media/events/system/swap_state_change
Failure /shutdown_check/events/
excessive_shutdowns
Event /license/events/key_state_change
Event /stats/events/conntrack_entries/rising/error
Event
/stats/events/conntrack_entries/rising/clear
Event
/media/events/vimm_temp/rising/error
Event
/media/events/vimm_temp/rising/clear
Event
/media/events/chassis_temp/rising/error
Event
/media/events/chassis_temp/rising/clear
Event
/media/events/lid_ajar_time/rising/error
Event
Event
/media/events/lid_ajar_time/rising/clear
/media/events/array/lid_ajar
Event
/media/events/array/pcie_link_up
Event
/media/events/array/pcie_link_down
Event
/media/events/array/data_plane_ready
Event
/media/events/array/raid_rebuild
Event
/media/events/array/vimm_state_change
Event
/media/events/array/psu_state
Event
/media/events/array/led_change
Event
/media/events/array/fan_change
Event
/vshare/events/fc/port_state_change
Event
/vshare/events/ib/port_state_change
Event
/vshare/events/vmg/fan_change
437
vmg-psu-change
vmg-boot-raid-change
Event
/vshare/events/vmg/psu_change
Event
/vshare/events/vmg/boot_raid_change
Examples
Example Failure Email Subject System failure on <system>: Unexpected shutdown
Example Event Email Subject System event on <system>: Unexpected cluster size event on <cluster
name>
Example Event Log
May 25 01:44:17 njvshare mgmtd[4148]: [mgmtd.INFO]: EVENT:
/vshare/events/fc/port_state_change
Available Traps
The following section covers the available traps. The first table lists the traps that are enabled by
default. The second table lists the traps that are disabled by default, and that you can enable as
needed.
438
535-0008-00 Rev 10
Available Traps
Traps
Description
ib-port-state-change
lid-ajar
vmg-boot-raid-change
vmg-fan-change
vmg-psu-change
Traps
Description
array-data-plane-ready:
array-led-change:
array-pcie-link-down:
array-pcie-link-up:
array-psu-state:
array-raid-rebuild:
lid-ajar-time-high:
lid-ajar-time-ok:
vimm-state-change:
vimm-temperature-high:
vimm-temperature-ok:
Traps
Description
cpu-util-high:
disk-space-low:
interface-down:
interface-up:
liveness-failure:
paging-high:
process-crash:
process-exit:
535-0008-00 Rev 10
439
snmp-authtrap:
unexpected-shutdown:
Trap
Description
array-fan-change:
disk-io-high:
fc-port-state-change:
license-state-change:
media-device-healthwarn:
media-device-lifetimewarn:
media-device-statechange:
media-device-unknowntype:
media-system-swapstate:
memusage-high:
netusage-high:
unexpected-cluster-join:
unexpected-clusterleave:
unexpected-cluster-size:
user-login:
user-logout:
Trap Name
array-data-planeready
440
Threshold
Value
Unit
State Change 0 or 1
Description
Solution
Sev.
High /
Low
535-0008-00 Rev 10
Available Traps
Trap Name
Threshold
Value
Unit
Description
Solution
Sev.
array-fan-change
Medium
array-led-change
High /
Low
array-pcie-link-down
Event
Triggered
High
array-pcie-link-up
Event
Triggered
Nothing.
Low
array-psu-state
High /
Low
535-0008-00 Rev 10
441
Trap Name
Threshold
Value
Unit
Description
Solution
Sev.
array-raid-rebuild
State Change 0 or 1
Medium
chassis-temperaturehigh
75
Celsius
Temperature inside a
Violin Array chassis
has exceeded normal
operating range
High
chassis-temperatureok
70
Celsius
Temperature inside a
Violin Array chassis
has dropped into
normal operating
range
Nothing.
Low
cpu-util-high
98
Percent
A Violin Memory
Gateway has detected
that a CPU has
exceeded utilization
above the threshold
level
Please contact
Violin customer
support if the
system is not under
full data traffic load.
Low
disk-io-high
5120
KBytes /
sec
Low
disk-space-low
Percent
Free
A Violin Memory
Gateway disk space
has crossed the
threshold of percent of
bytes free.
Please contact
Violin customer
support.
Medium
442
535-0008-00 Rev 10
Available Traps
Trap Name
Threshold
Value
Unit
Description
Solution
Sev.
fc-port-state-change
Medium
free-space-low
User
configurable
Medium
ib-port-state-change
vSHARE InfiniBand
(IB) port state
changed.
interface-down
Event
Triggered
A network interface on
a Violin Memory
Gateway has lost link.
High
interface-up
Event
Triggered
A network interface on
a Violin Memory
Gateway has detected
link up.
Nothing.
Low
535-0008-00 Rev 10
Percent
Free
443
Trap Name
Threshold
Value
Unit
Description
Solution
Sev.
license-state-change
A Violin Memory
Gateway feature
license state has
changed. The old and
new active state is
passed as well as the
specific feature
license. The active
state is a true or
false value. When
true, the feature
license is active.
If a license has
changed to active =
false, please
contact Violin
customer support to
see if your license
has expired.
Medium
lid-ajar
High
lid-ajar-time-high
60
Seconds
lid-ajar-time-ok
Seconds
Nothing.
Low
liveness-failure
Event
Triggered
An internal process on
a Violin Memory
Gateway has been
detected as hung.
Please contact
Violin customer
support
High
media-device-healthwarn
10
Percent
444
High
535-0008-00 Rev 10
Available Traps
Trap Name
Threshold
Value
Unit
Solution
Sev.
A Violin Memory
Gateway media device
(Violin Array) has
crossed the threshold
for S.M.A.R.T attr 233:
Media Wearout
Indicator. This tracks
the number of erase
cycles for flash as a
percentage of life
remaining.
Contact Violin
customer support to
determine if any
VIMMs should be
replaced.
High
media-devicelifetime-warn
media-device-statechange
A Violin Memory
Gateway media device
(Violin Array) has
changed state. The
possible states are:
unknown, online,
offline, error, removed,
disabled, or onlining.
High /
Low
media-deviceunknown-type
Event
A Violin Memory
Gateway has detected
a media device
connected of an
unknown type. The
Model, Serial #, and
size are provided.
Low
media-system-swapstate
An internal Memory
Gateway disk has
changed state for
having an active swap
partition. The active
state is true when a
swap partition has
been activated or false
for deactivation.
Please contact
Violin customer
support.
Low
memusage-high
90
A Violin Memory
Gateway has detected
that system RAM has
exceeded a
percentage in use
threshold.
Please contact
Violin customer
support.
Medium
535-0008-00 Rev 10
Percent
Description
Triggered
Percent
445
Trap Name
Threshold
Value
Unit
Description
Solution
Sev.
netusage-high
10485760
Low
paging-high
2000
Duration /
sec
Memory paging on a
Violin Memory
Gateway has
exceeded a threshold.
Please contact
Violin customer
support if the
system is not under
full data traffic load.
Medium
process-crash
Event
Triggered
An internal process on
a Violin Memory
Gateway has crashed.
A callhome event is
generated with the
details of the event.
Please contact
Violin customer
support.
High
process-exit
Event
Triggered
An internal process on
a Violin Memory
Gateway has
unexpectedly exited. A
callhome event is
generated with the
details of the event.
Please contact
Violin customer
support.
High
unexpected-clusterjoin
Event
Triggered
A Violin Memory
Gateway has
unexpectedly joined a
vCLUSTER after the
180 seconds cluster
startup time has
elapsed.
Low
unexpected-clusterleave
Event
Triggered
A Violin Memory
Gateway has
unexpectedly left a
vCLUSTER after
already being a
member.
High
446
535-0008-00 Rev 10
Available Traps
Trap Name
Threshold
Value
Unit
Description
Solution
Sev.
unexpected-clustersize
Expected
Nodes
Node
count
After a vCLUSTER
startup time of 180
seconds has elapsed,
the number of detected
nodes should match
cluster expectednodes configured. This
trap is sent when the
number of nodes has
changed and is not the
expected value.
Low
unexpectedshutdown
Event
Triggered
A Violin Memory
Gateway has
unexpectedly
shutdown. This may
happen during a
software upgrade
when the upgrade did
not cleanly apply and
the system reverted
back to the previous
version prior to
upgrade. A callhome
event is generated.
Please contact
Violin customer
support.
High
user-login
Event
Triggered
A login to a Violin
Memory Gateway via
either the CLI or Web
UI has occurred.
Low
user-logout
Event
Triggered
Low
vimm-state-change
High /
Low
535-0008-00 Rev 10
447
Trap Name
vimm-temperaturehigh
Threshold
Value
Unit
Description
Solution
Sev.
80
Celsius
High
vimm-temperature-ok 75
Celsius
Nothing.
Low
vmg-boot-raidchange
Active,
Degraded,
Resyncing,
Unknown
Degraded and
Unknown status
mean that the RAID
disk is likely down
and needs to be
replaced.
High
NA
Active is a normal
state.
Resyncing means
the RAID group is
rebuilding, which is
also normal.
Degraded means
the RAID disk is
down.
Unknown means
the RAID disk is
unavailable and
could be down.
vmg-fan-change
NA
Off,
Absent,
Low,
Medium,
High,
An absent status
could mean a failed
fan. Replace the fan
as soon as
possible.
High
vmg-psu-change
NA
On, Off,
Absent
If the status
changes to off or
absent, investigate
the issue
immediately.
High
448
535-0008-00 Rev 10
Enabling Traps
Use the snmp-server traps event command to enable traps:
violin [violin: master] (config) # snmp-server traps event cpu-util-high
*violin [violin: master] (config) # wr mem
Testing Traps
Use the snmp-server traps send-test command to test traps:
violin-acma [violin: master] (config) # snmp-server traps ?
community
Set the default community for traps sent to hosts which
do not have a custom community string set
event
Specify which events will be sent as traps
send-test
Send a test trap
violin-acma [violin: master] (config) # snmp-server traps send-test
This will send a test trap to all configured trap sinks. This will be the 'testTrap' notification from the
TMS-MIB. This trap is only ever sent on request from the user; it is not triggered automatically. This
trap is not available for enable or disable through configuration; it is always enabled, meaning it will
always be sent when requested by the user.
535-0008-00 Rev 10
449
Spare VIMMs
To query the spare VIMMs in the system:
[user@lab-2 mibs]$ snmpwalk -Ou -M +. -m +./VIOLIN-MEMORY-MIB.txt -v 2c -c public
lab-ib-srv1 violin-memory | grep spare | grep "Gauge32: 1"
enterprises.violinmemory.products.memoryGateway.appliance.media.arrayVimmTable.arrayVimmEntry.spa
re."lab-system-098"."vimm30" = Gauge32: 1
enterprises.violinmemory.products.memoryGateway.appliance.media.arrayVimmTable.arrayVimmEntry.spa
re."lab-system-098"."vimm31" = Gauge32: 1
enterprises.violinmemory.products.memoryGateway.appliance.media.arrayVimmTable.arrayVimmEntry.spa
re."lab-system-098"."vimm32" = Gauge32: 1
enterprises.violinmemory.products.memoryGateway.appliance.media.arrayVimmTable.arrayVimmEntry.spa
re."lab-system-098"."vimm52" = Gauge32: 1
Failed VIMMs
To query the failed VIMMs in the system:
snmpwalk -Ou -M +. -m +./VIOLIN-MEMORY-MIB.txt -v 2c -c public lab-ib-srv1 violinmemory | grep fail | grep "Gauge32: 1"
[user@lab-2 mibs]$
(none)
PSU States
To query the state of the power supplies in the system:
snmpwalk -Ou -M +. -m +./VIOLIN-MEMORY-MIB.txt -v 2c -c public lab-ib-srv1 violinmemory | grep -i psu
enterprises.violinmemory.products.memoryGateway.appliance.media.chassisSystemArrayTable.chassisSy
stemArrayEntry.chassisSystemPowerPSUA."lab-fender-098" = STRING: "ON"
enterprises.violinmemory.products.memoryGateway.appliance.media.chassisSystemArrayTable.chassisSy
stemArrayEntry.chassisSystemPowerPSUB."lab-fender-098" = STRING: "ON"
[user@lab-ib-cn2 mibs]$
450
535-0008-00 Rev 10
535-0008-00 Rev 10
451
Performance Stats
To query performance statistics for the system:
[test@host1 mibs]$ snmpwalk -Ou -M +. -m +./VIOLIN-MEMORY-MIB.txt -v 2c -c public
lab-vs3-n2 violin-memory | grep ata-VIOLIN
enterprises.violinmemory.products.memoryGateway.appliance.media.mediaDevTable.mediaDevEntry.media
DevIdx."ata-VIOLIN_MEMORY_ARRAY_2110CR00000304" = STRING: ataVIOLIN_MEMORY_ARRAY_2110CR00000304
enterprises.violinmemory.products.memoryGateway.appliance.media.mediaDevTable.mediaDevEntry.devId
."ata-VIOLIN_MEMORY_ARRAY_2110CR00000304" = STRING: ataVIOLIN_MEMORY_ARRAY_2110CR00000304
enterprises.violinmemory.products.memoryGateway.appliance.media.mediaDevTable.mediaDevEntry.fwVer
sion."ata-VIOLIN_MEMORY_ARRAY_2110CR00000304" = STRING: 3.7.2
enterprises.violinmemory.products.memoryGateway.appliance.media.mediaDevTable.mediaDevEntry.g6Mod
el."ata-VIOLIN_MEMORY_ARRAY_2110CR00000304" = STRING: Violin Memory Array
enterprises.violinmemory.products.memoryGateway.appliance.media.mediaBlockTable.mediaBlockEntry.d
evId."2110CR00000304" = STRING: ata-VIOLIN_MEMORY_ARRAY_2110CR00000304
enterprises.violinmemory.products.memoryGateway.appliance.media.mediaBlockTable.mediaBlockEntry.d
evPath."2110CR00000304" = STRING: /dev/disk/by-id/ataVIOLIN_MEMORY_ARRAY_2110CR00000304-part4
enterprises.violinmemory.products.memoryGateway.appliance.media.mediaStatsDevTable.mediaStatsDevE
ntry.mediaStatsDevIdx."ata-VIOLIN_MEMORY_ARRAY_2110CR00000304" = STRING: ataVIOLIN_MEMORY_ARRAY_2110CR00000304
enterprises.violinmemory.products.memoryGateway.appliance.media.mediaStatsDevTable.mediaStatsDevE
ntry.kernelOpsInProgress."ata-VIOLIN_MEMORY_ARRAY_2110CR00000304" = STRING: 0
enterprises.violinmemory.products.memoryGateway.appliance.media.mediaStatsDevTable.mediaStatsDevE
ntry.kernelOpsTime."ata-VIOLIN_MEMORY_ARRAY_2110CR00000304" = STRING: 47374
enterprises.violinmemory.products.memoryGateway.appliance.media.mediaStatsDevTable.mediaStatsDevE
ntry.kernelOpsTimeWeighted."ata-VIOLIN_MEMORY_ARRAY_2110CR00000304" = STRING:
79475
enterprises.violinmemory.products.memoryGateway.appliance.media.mediaStatsDevTable.mediaStatsDevE
ntry.kernelReadCompleted."ata-VIOLIN_MEMORY_ARRAY_2110CR00000304" = STRING:
1157474
enterprises.violinmemory.products.memoryGateway.appliance.media.mediaStatsDevTable.mediaStatsDevE
ntry.kernelReadKbytes."ata-VIOLIN_MEMORY_ARRAY_2110CR00000304" = STRING:
11750483
452
535-0008-00 Rev 10
535-0008-00 Rev 10
453
454
535-0008-00 Rev 10
535-0008-00 Rev 10
455
456
535-0008-00 Rev 10
535-0008-00 Rev 10
457
458
535-0008-00 Rev 10
535-0008-00 Rev 10
459
460
535-0008-00 Rev 10
535-0008-00 Rev 10
461
462
535-0008-00 Rev 10
535-0008-00 Rev 10
463
464
535-0008-00 Rev 10
APPENDIX G
This appendix demonstrates ways in which you can optimize the connectivity of
initiator (client-side) hosts with Violin Memory Arrays. The following topics are
covered:
http://technet.microsoft.com/en-us/library/ee619752(WS.10).aspx
2. To ensure a proper MPIO operation, complete the following steps:
a. Install SP1 on Windows Server 2008 R2.
b. Install HotFix KB2522766 (from the Microsoft Support site).
c. Install HotFix KB2460971 (from the Microsoft Support site).
d. Install HotFix KB2468345 (from the Microsoft Support site).
e. Install HotFix KB2591462 (from the Microsoft Support site).
535-0008-00 Rev 10
465
Note: You must be a registered Violin Memory customer to download this driver. If you are
not registered, you are prompted to complete the registration process.
3. Click the
Software Downloads.
4. Click the
MPIO Driver DSM D4.5.0 link within Drivers and Utilities to download the software.
If you are using the Violin driver, the load balance policy is set to 1. For higher performance, this
should be set to 2 (Round Robin).
To change the load balance policy setting, complete the following steps:
1. Open the Registry Editor and navigate to:
HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Services\v1010dsm\Parameters\Dsm
LoadBalanceSettings\<EnumeratedViolin LUN>
2. Modify DWORD: DsmLoadBalancePolicy and change the value to 2 for all discovered LUNs.
3. Restart the client machine.
466
535-0008-00 Rev 10
System Requirements
This section covers the requirements for AIX operating systems and Violin Memory Gateway
software versions.
devices.common.IBM.fc.rte
devices.common.IBM.rte
devices.fcp.disk.rte
devices.fcp.tape.rte
Note: Refer to the AIX operating system documentation for instructions if any of the required file
sets listed above are not present on the AIX initiator host.
Note: You can use smitty, the System Management Interface Tool, to install software on an AIX
system. Or, you can use the command suggested in the following procedure.
To download and install the Violin driver for AIX MPIO, do the following:
1. Go to the Violin Memory Support Site (http://www.violin-memory.com/support/).
2. Download the appropriate driver for AIX MPIO (devices.fcp.disk.vmem)based on the version
of AIX running on your systemto a temporary location on the AIX initiator host, such as
/var/tmp/violin.
535-0008-00 Rev 10
467
2. Ensure that the maximum transfer size (max_xfer_size) is set to 0x200000, using the
following commands:
# chdev -l fcs0 -a max_xfer_size=0x200000
# chdev -l fcs0 -a num_cmd_elems=2000
3. Ensure the NACA bit is set for LUNs that are exported to AIX initiator hosts. For more
information, see the Violin 6000 Series Memory Array Users Guide.
4. Configure the following settings for each fscsi driver object in the following way, to ensure non-
468
535-0008-00 Rev 10
The Violin driver for AIX MPIO automatically configures the following attributes for optimum
connectivity:
algorithm: round_robin
queue_depth: 255
reserve_policy: no_reserve
rw_timeout: 60
Stop all applications that may be accessing LUNs on Violin Memory Arrays.
off) before deleting LUNs from the Violin Memory Array. Use the following command.
# varyoffvg <vg_name>
3. Ensure that the LUN entries are removed from the device tree, using the following command.
# rmdev -dl <hdisk##>
4. Use the following command to uninstall the Violin driver.
# installp -ug devices.fcp.disk.vmem.rte
5. Use the following command to verify that the Violin driver has been successfully removed.
E lslpp -l devices.fcp.disk.vmem.rte
535-0008-00 Rev 10
469
470
535-0008-00 Rev 10
System Specifications
535-0008-00 Rev 10
471
System Specifications
Series
6600
Model
Flash Type
SLC
6TB/6.6TB
6200
Violin Memory 6212
MLC
16TB/17.6TB
12TB/13.2TB
32TB/35.2TB
60 + 4
20 + 4
60 + 4
Reliability/Resiliency
Connectivity Options
Dimensions
Height:
Width:
Depth:
Cable Management:
Weight
76lbs / 34.5kg
89lbs / 40.4kg
76lbs / 34.5kg
89lbs / 40.4kg
Power*
1110W
1679W
1344W
1709W
Cooling
3822 BTU/hr
5780 BTU/hr
4626 BTU/hr
5884 BTU/hr
Environmental
Inlet Temp:
Humidity:
Altitude:
Electrical
Input Power Voltage:
Operating Range:
Current Rating:
200-240 VAC
180-264 VAC
15 A maximum
Interfaces
Management (CLI)
472
535-0008-00 Rev 10
8/4/2Gbps auto-negotiation
200,000 initiator and target IOPS per port
2 and 3
FC-AL, FC-AL2, point-to-point, switched fabric
FCP-3-SCSI
FC-Tape (FCP-2)
PCIe Express Base Specification rev. 2.0
PCIe Express Card Electromechanical Specification rev. 2.0
PCIe Bus Power Management Interface Specification rev. 1.2
PCI Hot Plug Specification rev. 1.0
Connectivity
Ports:
HBA Specifications
Airflow:
Power consumption:
Form factor:
Temperature:
No airflow required
6.2 Watts (typical)
Low-profile PCI Express card (6.6 in. x 2.54 in.)
0 to 55C (operating)
-40 to 70C (non-operating)
10% to 90% (operating, non-condensing)
5% to 93% (non-operating, non-condensing)
RoHS 6
Relative Humidity:
RoHS compliance:
The following table is a description of the LED states of the QLogic QLE2562 Host Bus Adapter.
See the QLogic Web site (http://www.qlogic.com) for the most complete and up-to-date information.
Yellow
(8Gbps)
Green
(4Gbps)
Orange
(2Gbps)
Off
Off
Off
Power off
On
On
On
Flashing
Flashing
Flashing
Description
Off
Off
On/Flashing
Off
On/Flashing
Off
On/Flashing
Off
Off
535-0008-00 Rev 10
473
System Specifications
QoS:
Data rate:
QDR: InfiniBand
Connectivity:
Link Speeds:
40 Gbits/s
Application Performance:
Active Components:
Topologies:
Cable Connections:
Power Consumption:
7 watts
Mechanical Dimensions:
Operating Environment:
474
535-0008-00 Rev 10
Bus type:
Bus width:
Interrupt levels:
Operating temperature:
0C-55C (32F-131F)
Connectors:
Physical dimensions:
535-0008-00 Rev 10
475
System Specifications
476
535-0008-00 Rev 10
Index
Numerics
4kB alignment
See partitions
A
AC Power Failure, Shutting Down alarm 373
active configuration file 230, 238
backed up 231
admin role 159
access privileges 34, 119, 159
commands in Config mode 225
Admin user 34
admin username
password 158
AIX
configure initiator host 468
discover devices 469
file sets, required 467
hosts 466
MPIO 466
supported OS 467
uninstall Violin driver 469
Violin driver, installing 467
AIX compatibility 179
AIX compatibility, NACA 179
Alarm LED 370
alarms
Alarm LED states 370
commands 371
non-service affecting alarms 377385
service affecting alarms 372377
ALUA
attributes 257
definition 66
enable 66
535-0008-00 Rev 10
iSCSI-3 8
performance optimization 257
preferred port 257
settings, view 259
ARP configuration 319
Array Controller
about 19
location 12
Array status screen 148
asterisk before CLI prompt 230, 238
authenticated sign-on 248
authentication
bidirectional CHAP 248
CHAP 248
iSCSI CHAP 248
local 299
SMTP configuration 130
unidirectional CHAP 248
B
bidirectional authentication 248
block size, of LUNs 63, 191, 334
block storage
See vSHARE
block storage media devices
enabling or disabling 207209
initializing Memory Arrays 180181
read and write statistics 210
block zeroing, VAAI 212
Board Sensor Unreadable alarm 378
board status
viewing in Web interface 149
Board Status page 49, 50
boot images 114
Booting Suspended, Fans Missing alarm 372
477
C
cache media devices
read and write statistics 210
call home
auto support 128
fault reporting 128
notification 128
settings 128
call home settings 168
caution icons 3, 217
CHAP
authenticated sign-on 248
bidirectional 248
bidirectional, configure 250
requirements 249
secret, establish 249
secret, remove 251
set up iSCSI 248
show configurations 252, 328
types 248
unidirectional 248
unidirectional, configure 250
chassis
front view 13, 1516
mounting on equipment rack 11
overview of 11
rear view 14
serviceability of 11
temperature alarms
non-service affecting 378
service-affecting 376
top view of modules 12
CLI
ARP configuration 319
boot commands 309
command history 291
configuration commands 306
connecting to 227
create user account 299
diagnostic commands 303
display settings 291
hidden commands 290
logging in 227229
LUN statistics 216, 337
options 289
paging output 290
reboot 290
reload 290
See command line interface (CLI)
Standard mode
commands available 224
tcpdump 311
terminal type 291
user account commands 299
user inactivity 289
cluster
Cluster Status page 48
management address, change 123, 163
Cluster Status page 48, 49
478
command
array rebalance 242
command line interface (CLI)
ambiguous command 222
asterisk for unsaved changes 230, 238
command completion
finding options with ? 222
finding values with ? 224
tab completion 224
command descriptions 288289
command list
display with tabs 224
command modes 225
command prompts 226
INTERACTIVE commands 289
key to parameters 288
shorthand form 222
showing help 222
command modes
Config mode 225
Enable mode 225
prompt and response conventions 226
Standard mode 225
user privileges 159
commands
unified logging upload 219
Commit Changes button 230
Config mode 225
admin role 159
command prompt 226
entering 225
unsaved changes (asterisk) 226
configuration
ARP 319
e-mail 129
network bond 315
RADIUS 328
reverting to saved configuration 232
save settings
configuration file 230
file management 230, 238
saving settings 230, 238
showing 231
SMTP 130
TACAS+ 329
configuration file 230
active 230, 238
backed up 231
copying 232
deleting 232
file management 230
revert to saved configuration 232
save to a different file
example 230
saving to a different file 230
show configuration files
example 231
showing configuration 231
showing configuration files 231
active file 231
535-0008-00 Rev 10
D
data management 253
best practices 255
scenarios 255
terminology 254
use cases 255
Data Plane Unavailable alarm 375
diagnostic tools 303
diskpart utility 426
DNS servers
configuring 315
DNS settings 124, 164
domain name CLI parameter 288
E
ECC errors 382, 384, 410
e-mail
alert consolidation 217, 317
alerts 217
configuration 129
consolidation 217
customizing 218
E-mail Settings page 129, 162, 169
events 132
network settings 129
Enable mode
command prompt 226
disable 225
monitor role 159
error messages
ambiguous commands 222
535-0008-00 Rev 10
F
Failed to Validate Configuration of VIMMs
Raid Group alarm 382
failover protection 22
fan
location 12
Fan Missing alarm 378
Fan Running Too Slowly alarm
non-service affecting
equipment alarm 379
fans 16
Fatal Flash Error, Hex-Value alarm 372
FATAL RAID Error alarm 374
fdisk utility 426
feature licenses 138, 170
Fibre Channel
CLI commands 331
protocol 177
target ports 183
file management, configuration 306
firmware
upgrading
Memory Array 237
upgrading using CLI 233
flash
alarms 373
non-service affecting 379380
service-affecting 372
Flash CPL IRQs on VIMM alarm 379
Flash Format Failed on VIMM alarm 379
Flash Sequence Number Mismatch alarm 379
flash-based media
media health 209
read and write statistics 210
show media 203
format
changing format capacity 313, 320
default capacity 21
performance tuning 424
showing current format capacity 325
free space management 279
monitor 279
reclamation process 282
reclamation rules 283
set thresholds 283
thresholds 282
Front and Rear Panel page 54
front panel
viewing in Web interface 56
full copy, VAAI 212
479
G
global default gateway 123, 164
H
help
command completion
tab completion 224
command-line help 222
question mark in CLI 222
showing in CLI 222
hidden commands, show 290
High Temperature on Board Sensor X X C alarm 378
host
AIX initiator 466
initiator 465
Windows initiator 465
host, initiator 465
hostname
CLI parameter 288
setting or clearing 315
I
I/O Timeout From Host alarm 375
ID LED 16
remotely activating 56
igroups
See initiator groups
immediate upgrade, Memory Gateway 235
immediate upgrade, system 234
initiator groups 67
configuration of 187188
create, delete 69
delete initiator 71
nested 72, 332
initiator hosts, connectivity 465
initiators
create groups 69
delete from group 71
managing 69
Insufficient VIMMs alarm 376
INTERACTIVE commands 289
interfaces
bonding of 315
480
535-0008-00 Rev 10
M
MAC address CLI parameter 288
Main Controller Does Not Support VIMMs alarm 377
Main Power Failure, System Shut Down alarm 373
Manage 221
LUNs 59
Manage Initiators page 69
Manage Targets page 75
management interfaces 27
RJ-45 Ethernet CLI 14
serial console port 14
serial console port CLI 227
mask
netmask format 288
Master ACM 156
connecting to Violin Web interface 143
Master Gateway
See master node
master node 114
connect to Web Interface 31
statistics, viewing 48
media
enable or disable
all media 208
read and write statistics 210
media, health status 209
Memory Array
about 8
alarms
understanding 369
rebooting 239
shutting down modules 239
version information 157
Memory Arrays, initializing for block storage 180181
Memory Gateway
boot images 114
master node, Web Interface connection 31
Memory Gateways
about 19
location 12
shutting down 236237
memory plane 18
memory, show media 203
message bar
Commit Changes button 230
modules
powering off and on 240
shutting down 239
Monitor
Monitor LUNs page 110, 111
monitor role 159
access privileges 34, 119, 159
commands in Enable mode 225
Monitor user 34
monitor username
password 158
MTBF 18
535-0008-00 Rev 10
N
NACA 179
NACA bit support 179
netmask CLI parameter 288
network
bond 315
Network Interfaces
about 19
location 13
network interfaces
See interfaces
No PCIe Host Detected on Enabled Port alarm 373
No Spare Memory Modules alarm 382
nodes, cluster
cluster node id CLI parameter 288
NTP
modify settings 125
O
OpenVMS 195
over provisioning 255, 256
P
partitions
4kB alignment 425427
creating 397400
password
admin 158
change 120, 160
monitor 158
PCIe switching 10
performance optimization 257
performance testing
using Perf_Test 428432
port
CLI parameter 288
Port X Negotiated to 0 Lanes alarm 373
ports
alarms 373, 380
POSIX threads 420
Power Controllers
about 18
location 12
power supplies 16
alarms 373, 380
Power Supply Failure alarm 380
Power Supply Missing alarm 380
Power Supply Polling Failure alarm 380
481
R
RADIUS configuration 328
RAID
alarms 374, 380381
automatic balancing 244
discovering unbalanced group 243
rebalance commands 242
rebalancing 242
scheduling group rebalance 245
showing rebalance settings 245
RAID Consistency Error alarm 374
RAID Group Faulted alarm 374
RAID Group Unprotected alarm 381
RAID Rebuild in Progress alarm 381
Raid Rebuild Of Group %D Completed with X Errors Starting at 4K Address Y alarm 381
RAID Rebuild of Group Completed with Errors (No Additional Information Available) alarm 381
RAID V2P Error alarm 374
rear panel
viewing in Web interface 54
reboot 290
REST 27
REST API 29
revert to saved configuration
example 232
roles
specify for a user 119, 159
user privileges 119, 158
routing
commands 316
S
save configuration 230, 238
active configuration file 230, 238
configuration file 230, 238
configuration write 230
write memory 230
schedule
modify schedule 276
schedule management 274
create schedule 274
delete schedule 275
show list 279
482
535-0008-00 Rev 10
T
TACAS+ configuration 329
target ports
configuring 182
Fibre Channel ports 183
iSCSI ports, configuration of 184187
verifying port activity and status 201
targets
manage 75
Manage Targets page 75
Monitor Targets page 111
monitoring, Web Interface 111
TCP port
range of values 288
tcpdump, commands 311
Telnet 14
telnet command 224
thin provisioning 254, 260
thresholds, alert, reclaim 103
thresholds, free space 283
thresholds, usage 256
tools, diagnostic 303
trap sinks, SNMP 135
V
VAAI support 212
vcounts command 406409
veeprom command 409
VIMM
about 9
VIMM Booting (X% Complete) alarm 383
VIMM Configuration Out of Spec alarm 383
VIMM Extremely Hot 384
VIMM Failed
Excessive ECC Errors alarm 384
VIMM Failed alarm 384
VIMM Failed alarms 383
VIMM Flash Format In Progress alarm 377
VIMM Hot alarm 384
VIMM in Maintenance State alarm 385
VIMM Incompatible Formatted Capacity alarm 385
VIMM Isolated alarm 385
VIMM Shutdown alarm 384
VIMM status 49
VIMM Status LED 15, 370
VIMM Temperature Unreadable, VIMM Shut Down alarm
385
U
Unable to Poll VIMM Temperatures 375
Unformatted System alarm 373
unidirectional authentication 248
Unified Logging Upload 219, 319
unpriv role
commands in Standard mode 225
Unsupported VIMM Configuration alarm 382
upgrading
Memory Array firmware 237
Memory Gateway software 235
URL
CLI parameter 288
usage thresholds 256
usage thresholds, set 283
USB LEDs 15
used space thresholds 103
user
change password 120, 160
535-0008-00 Rev 10
483
W
warning icons 3
Web Admin page 126, 167
Web Interface
Administration 113, 154
Board Status page 49
Call Home Settings 128, 168
Cluster Status page 48
484
X
xinetd server, settings 311
XTree interface
accessing 173
Z
zeroing-out data 191, 334
535-0008-00 Rev 10