book

SAN Volume Controller: Best Practices and Performance Guidelines

Name: SAN Volume Controller: Best Practices and Performance Guidelines
ISBN: 0738485780

by Jon Tate, Deon George, Thorsten Hoss, Ronda Hruby, Ian MacQuarrie, Barry Mellish, Peter Mescher

March 2008

Intermediate to advanced

326 pages

10h 22m

English

IBM Redbooks

Read now

Unlock full access

Notices
Page xi
Trademarks
Page xii
Preface
The team that wrote this book
Page xiiiPage xiv
Become a published author
Comments welcome
Page xvPage xvi
1. SAN fabric
Page 1
1.1. SVC SAN topology
1.1.1. Redundancy
1.1.2. Topology basics
Page 2

1.1.3. ISL oversubscription
Page 3
1.1.4. IBM 2109-M12/Brocade 12000 in an SVC environment
1.1.5. Switch port layout for large edge switches
Page 4
1.1.6. Switch port layout and hardware selection for director-class core switches
1.1.7. Single switch SVC SANs
1.1.8. Basic core-edge topology
Page 5
1.1.9. Four-SAN core-edge topology
Page 6
1.1.10. Cisco VSANs
Page 7
1.1.11. Common topology issues
Page 8Page 9
1.2. Tape and disk on your SAN
1.3. Switch interoperability
Page 10
1.4. Distance extension for mirroring
1.4.1. Optical multiplexors
1.4.2. Long-distance SFPs/XFPs
1.4.3. Fibre Channel: IP Conversion
Page 11
1.5. Zoning
1.5.1. Type of zoning
Page 12Page 13
1.5.2. Pre-zoning tips and shortcuts
1.5.3. SVC cluster zone
1.5.4. SVC: Storage zones
1.5.5. SVC: Host zones
Page 14Page 15
1.5.6. Sample standard SVC zoning configuration
Page 16Page 17Page 18Page 19
1.5.7. Zoning with multiple SVC clusters
1.5.8. Split controller configurations
1.6. Switch Domain IDs
1.7. TotalStorage Productivity Center for Fabric
Page 20
2. SAN Volume Controller cluster
Page 21
2.1. Advantages of virtualization
2.1.1. How does the SVC fit into your environment
2.2. Scalability of SVC clusters
Page 22
2.2.1. Advantage of multi cluster as opposed to single cluster
2.2.2. Performance expectations by adding an SVC
Page 23
2.2.3. Growing or splitting SVC clusters
Page 24Page 25
2.3. SVC cache improves subsystem performance
Page 26Page 27Page 28
2.3.1. Cache destage operations
Page 29
2.4. Cluster upgrade
Page 30Page 31Page 32
3. Master console
Page 33
3.1. Managing the master console
3.1.1. Managing a single master console
Page 34Page 35Page 36Page 37Page 38
3.1.2. Managing multiple master consoles
Page 39
3.1.3. Administration roles
Page 40Page 41Page 42Page 43Page 44Page 45
3.1.4. Audit logging
Page 46Page 47Page 48Page 49
3.1.5. Managing IDs and passwords
Page 50Page 51
3.1.6. Saving the SVC configuration
Page 52
3.1.7. Restoring the SVC cluster configuration
Page 53Page 54
4. I/O Groups and nodes
Page 55
4.1. Determining I/O Groups
4.2. Node shutdown and node failure
Page 56
4.2.1. Impact when running single node I/O Groups
Page 57
4.3. Adding or upgrading SVC node hardware
Page 58Page 59Page 60
5. Storage controller
Page 61
5.1. Controller affinity and preferred path
5.1.1. ADT for DS4000
Page 62
5.1.2. Ensuring path balance prior to MDisk discovery
5.2. Pathing considerations for EMC Symmetrix/DMX and HDS
5.3. LUN ID to MDisk translation
5.3.1. ESS
Page 63
5.3.2. DS6000 and DS8000
Page 64
5.4. MDisk to VDisk mapping
Page 65
5.5. Mapping physical LBAs to Extents
5.6. Media error logging
5.6.1. Host encountered media errors
Page 66
5.6.2. SVC-encountered media errors
Page 67
5.7. Selecting array and cache parameters
5.7.1. DS4000 array width
Page 68
5.7.2. Segment size
Page 69
5.7.3. DS8000
5.8. Considerations for controller configuration
5.8.1. Balancing workload across DS4000 controllers
Page 70
5.8.2. Balancing workload across DS8000 controllers
Page 71Page 72
5.8.3. DS8000 ranks/extent pools
Page 73
5.8.4. Mixing array sizes within an MDG
5.8.5. Determining the number of controller ports for ESS/DS8000
5.8.6. Determining the number of controller ports for DS4000
Page 74
5.9. LUN masking
Page 75Page 76
5.10. WWPN to physical port translation
Page 77
5.11. Using TPC to identify storage controller boundaries
Page 78
5.12. Using TPC to measure storage controller performance
Page 79
5.12.1. Approximations
Page 80
5.12.2. Establish a performance baseline
5.12.3. Performance metric guidelines
Page 81
5.12.4. Storage controller back end
Page 82Page 83Page 84
6. MDisks
Page 85
6.1. Back-end queue depth
6.2. MDisk transfer size
6.2.1. Host I/O
Page 86
6.2.2. FlashCopy I/O
6.2.3. Coalescing writes
6.3. Selecting LUN attributes for MDisks
Page 87
6.4. Tiered storage
6.5. Adding MDisks to existing MDGs
Page 88
6.5.1. Adding MDisks for capacity
6.5.2. Checking access to new MDisks
6.5.3. Persistent reserve
Page 89
6.6. Removing MDisks from existing MDGs
Page 90Page 91
6.7. Remapping managed MDisks
Page 92
6.8. Controlling extent allocation order for VDisk creation
Page 93Page 94Page 95Page 96
7. Managed disk groups
Page 97
7.1. Availability considerations for planning MDGs
Page 98
7.1.1. Performance consideration
Page 99
7.1.2. Selecting the MDisk Group
Page 100Page 101
7.2. Selecting number of LUNs per array
7.2.1. Performance comparison of one compared to two LUNs per array
Page 102Page 103Page 104
7.3. Selecting the number of arrays per MDG
Page 105Page 106Page 107Page 108Page 109Page 110
7.4. Striping compared to sequential type
Page 111
7.5. Selecting storage controllers
Page 112
8. VDisks
Page 113
8.1. Creating VDisks
Page 114Page 115
8.1.1. Selecting the MDisk Group
8.1.2. Changing the preferred node within an I/O Group
Page 116
8.1.3. Moving a VDisk to another I/O Group
Page 117Page 118
8.2. VDisk migration
8.2.1. Migrating across MDGs
8.2.2. Image type to striped type migration
8.2.3. Migrating to image type VDisk
Page 119Page 120
8.2.4. Preferred paths to a VDisk
Page 121
8.2.5. Governing of VDisks
Page 122Page 123Page 124
8.3. Cache-disabled VDisks
8.3.1. Using underlying controller remote copy with SVC cache-disabled VDisks
Page 125
8.3.2. Using underlying controller PiT copy with SVC cache-disabled VDisks
Page 126
8.3.3. Changing cache mode of VDisks
Page 127Page 128
8.4. VDisk performance
Page 129Page 130Page 131
8.4.1. VDisk performance
Page 132Page 133Page 134Page 135Page 136Page 137Page 138
8.5. The effect of load on storage controllers
Page 139Page 140Page 141Page 142
9. Copy services
Page 143
9.1. SAN Volume Controller Advanced Copy Services functions
9.1.1. SVC copy service functions
9.1.2. Using both Metro Mirror and Global Mirror between two clusters
9.1.3. Performing three-way copy service functions
Page 144
9.1.4. Using native controller Advanced Copy Services functions
Page 145
9.2. Copy service limits
Page 146
9.3. Setting up FlashCopy copy services
Page 147
9.3.1. Steps to making a FlashCopy VDisk with application data integrity
Page 148Page 149Page 150
9.3.2. Making multiple related FlashCopy VDisks with data integrity
Page 151Page 152
9.3.3. Creating multiple identical copies of a VDisk
Page 153
9.3.4. Understanding FlashCopy dependencies
Page 154Page 155
9.3.5. Using FlashCopy with your backup application
9.3.6. Using FlashCopy to help with migration
Page 156Page 157
9.3.7. Summary of FlashCopy rules
9.4. Metro Mirror and Global Mirror
9.4.1. Configuration requirements for long distance links
Page 158
9.4.2. Global mirror guidelines
Page 159Page 160Page 161
9.4.3. Migrating a Metro Mirror relationship to Global Mirror
9.4.4. Recovering from suspended Metro Mirror or Global Mirror relationships
Page 162
9.4.5. Diagnosing and fixing 1920 errors
Page 163Page 164
9.4.6. Using Metro Mirror or Global Mirror with FlashCopy
9.4.7. Saving bandwidth creating Metro Mirror and Global Mirror relationships
Page 165
9.4.8. Using TPC to monitor Global Mirror performance
Page 166
9.4.9. Summary of Metro Mirror and Global Mirror rules
Page 167Page 168
10. Hosts
Page 169
10.1. Configuration recommendations
10.1.1. The number of paths
Page 170
10.1.2. Host ports
10.1.3. Port masking
10.1.4. Host to I/O Group mapping
Page 171
10.1.5. VDisk size as opposed to quantity
10.1.6. Host VDisk mapping
Page 172Page 173Page 174Page 175
10.1.7. Server adapter layout
10.1.8. Availability as opposed to error isolation
10.2. Host pathing
Page 176
10.2.1. Preferred path algorithm
10.2.2. Path selection
Page 177
10.2.3. Path management
Page 178
10.2.4. Dynamic reconfiguration
Page 179
10.2.5. VDisk migration between I/O Groups
Page 180Page 181
10.3. I/O queues
10.3.1. Queue depths
Page 182Page 183
10.4. Multipath software
10.5. Host clustering and reserves
Page 184Page 185
10.5.1. AIX
Page 186Page 187Page 188Page 189
10.5.2. SDD compared to SDDPCM
Page 190
10.5.3. Virtual I/O server
Page 191Page 192
10.5.4. Windows
Page 193
10.5.5. Linux
10.5.6. Solaris
Page 194Page 195
10.5.7. VMWare
Page 196
10.6. Mirroring considerations
10.6.1. Host-based mirroring
10.7. Monitoring
Page 197
10.7.1. Automated path monitoring
10.7.2. Load measurement and stress tools
Page 198Page 199Page 200
11. Applications
Page 201
11.1. Application workloads
11.1.1. Transaction-based processes (IOPS)
Page 202
11.1.2. Throughput-based processes (MBps)
11.1.3. Host considerations
11.2. Application considerations
11.2.1. Transaction environments
Page 203
11.2.2. Throughput environments
11.3. Data layout overview
Page 204
11.3.1. Layers of volume abstraction
11.3.2. Storage administrator and AIX LVM administrator roles
Page 205
11.3.3. General data layout recommendations
Page 206Page 207
11.3.4. Database strip size considerations (throughput workload)
11.3.5. LVM volume groups and logical volumes
Page 208
11.4. When the application does its own balancing of I/Os
11.4.1. DB2 I/O characteristics and data structures
Page 209Page 210
11.4.2. DB2 data layout example
11.4.3. Striped VDisk recommendation
Page 211
11.5. Data layout with the AIX virtual I/O (VIO) server
11.5.1. Overview
Page 212
11.5.2. Data layout strategies
11.6. VDisk size
Page 213
11.7. Failure boundaries
Page 214
12. Monitoring
Page 215
12.1. Configuring TPC to analyze the SVC
Page 216
12.2. Using the TPC to verify fabric configuration
12.2.1. Verifying SVC node ports
Page 217Page 218
12.2.2. Ensure that all SVC ports are online
Page 219
12.2.3. Verifying SVC port zones
Page 220Page 221
12.2.4. Verifying paths to storage
Page 222Page 223
12.2.5. Verifying host paths to the SVC
Page 224Page 225Page 226
12.3. Methods for collecting data
Page 227
12.3.1. Setting up TPC to collect performance information
12.3.2. Viewing TPC-collected information
Page 228Page 229Page 230Page 231Page 232Page 233Page 234
12.3.3. Using TPC to alert on performance constraints
Page 235Page 236
13. Maintenance
Page 237
13.1. Configuration and change tracking
13.1.1. SAN
Page 238Page 239Page 240
13.1.2. SVC
Page 241
13.1.3. Storage
13.1.4. General inventory
13.1.5. Change tickets and tracking
13.1.6. Configuration archiving
Page 242
13.2. Standard operating procedures
Page 243Page 244Page 245
13.3. TotalStorage Productivity Manager
13.4. Code upgrades
13.4.1. Which code levels
13.4.2. How often
13.4.3. What order
Page 246
13.4.4. Preparing for upgrades
13.4.5. Host code upgrades
Page 247
13.5. SAN hardware changes
13.5.1. Cross-referencing the SDD adapter number with the WWPN
13.5.2. Changes that result in the modification of the destination FCID
Page 248
13.5.3. Switch replacement with a like switch
Page 249
13.5.4. Switch replacement or upgrade with a different kind of switch
13.5.5. HBA replacement
Page 250
13.6. Naming convention
13.6.1. Hosts, zones, and SVC ports
13.6.2. Controllers
13.6.3. MDisks
13.6.4. VDisks
13.6.5. MDGs
Page 251Page 252
14. Other useful information
Page 253
14.1. Cabling
14.1.1. General cabling advice
14.1.2. Long distance optical links
14.1.3. Labeling
14.1.4. Cable management
Page 254
14.1.5. Cable routing and support
14.1.6. Cable length
14.1.7. Cable installation
Page 255
14.2. Power
14.2.1. Bundled uninterruptible power supply units
14.2.2. Rack power feeds
14.3. Cooling
Page 256
14.4. SVC scripting
14.5. IBM Support Notifications Service
14.6. SVC Support Web site
14.7. SVC-related publications and classes
Page 257
14.7.1. IBM Redbooks publications
14.7.2. Courses
Page 258
15. Troubleshooting and diagnostics
Page 259
15.1. Common problems
15.1.1. Host problems
15.1.2. SVC problems
Page 260Page 261
15.1.3. SAN problems
15.1.4. Storage subsystem problems
15.2. Collecting data and isolating the problem
Page 262Page 263
15.2.1. Host data collection
Page 264
15.2.2. Multipathing driver: SDD data
Page 265Page 266
15.2.3. SVC data collection
Page 267Page 268
15.2.4. SAN data collection
15.2.5. Storage subsystem data collection
Page 269
15.3. Recovering from problems
15.3.1. Solving host problems
Page 270Page 271
15.3.2. Solving SVC problems
Page 272Page 273Page 274
15.3.3. Solving SAN problems
15.3.4. Typical SVC storage problems
Page 275Page 276
15.3.5. Solving storage subsystem problems
Page 277Page 278Page 279Page 280
15.3.6. Common error recovery steps
15.4. Livedump
Page 281Page 282
16. SVC 4.2 performance highlights
Page 283
16.1. SVC and continual performance enhancements
Page 284Page 285
16.2. SVC 4.2 code improvements
16.3. Test results
Page 286Page 287Page 288
16.3.1. Performance scaling of I/O Groups
Page 289Page 290
Related publications
IBM Redbooks publications
Other resources
Page 291
Referenced Web sites
How to get IBM Redbooks publications
Help from IBM
Page 292Page 293Page 294
Index
Page 295Page 296Page 297Page 298Page 299Page 300Page 301Page 302Page 303Page 304Page 305Page 306Page 307
Back cover
Page 308

Content preview from SAN Volume Controller: Best Practices and Performance Guidelines

Chapter 11.3.2. Storage administrator and AIX LVM administrator roles

Page 205

Chapter 11. Applications 205 layers, and avoid the performance problems and hot spots that come with poor data layout. Your goal is to balance I/ Os evenly across the physical disks in the back- end storage devices. You can treat sequential I/ O applications the same as random I/ O applications unless the sequential rate is high enough to matter. We will specifically show you how to lay out storage for DB2 ® applications as a good example of how an application might balance its I/ Os within the application. There are also different implications for the host data layout based on whether you utilize image mode or striped mode VDisks. 11.3.1 Layers of volume abstraction ...

Become an O’Reilly member and get unlimited access to this title plus top books and audiobooks from O’Reilly and nearly 200 top publishers, thousands of courses curated by job role, 150+ live events each month,
and much more.

Read now

Unlock full access

More than 5,000 organizations count on O’Reilly

O’Reilly covers everything we've got, with content to help us build a world-class technology community, upgrade the capabilities and competencies of our teams, and improve overall team performance as well as their engagement.

Julian F.

Head of Cybersecurity

I wanted to learn C and C++, but it didn't click for me until I picked up an O'Reilly book. When I went on the O’Reilly platform, I was astonished to find all the books there, plus live events and sandboxes so you could play around with the technology.

Addison B.

Field Engineer

I’ve been on the O’Reilly platform for more than eight years. I use a couple of learning platforms, but I'm on O'Reilly more than anybody else. When you're there, you start learning. I'm never disappointed.

Amir M.

Data Platform Tech Lead

I'm always learning. So when I got on to O'Reilly, I was like a kid in a candy store. There are playlists. There are answers. There's on-demand training. It's worth its weight in gold, in terms of what it allows me to do.

Mark W.

Embedded Software Engineer

IBM System Storage SAN Volume Controller Best Practices and Performance Guidelines

Mary Lovelace, Katja Gebuhr, Ivo Gomilsek, Ronda Hruby, Paulo Neto, Jon Parkes, Otavio Rocha Filho, Leandro Torolho

IBM SAN Volume Controller Best Practices and Performance Guidelines for IBM Spectrum Virtualize Version 8.4.2

Antonio Rainero, Carlton Beatty, David Green, Hartmut Lonzer, Jonathan Wilkie, Kendall Williams, Konrad Trojok, Mandy Stevens, Nezih Boyacioglu, Nils Olsson, Renato Santos, Rene Oehme, Sergey Kubin, Thales Noivo Ferreira, Uwe Schreiber, Vasfi Gucer

IBM System Storage SAN Volume Controller, IBM Storwize V7000, and IBM FlashSystem 7200 Best Practices and Performance Guidelines

Jon Tate, Tiago Moreira Candelaria Bastos, Jon Herd, Sang-hyun Kim, Sergey Kubin, Dirk Peitzmann, Antonio Rainero

Storage Area Networks For Dummies®

Christopher Poelker, Alex Nikitin

Publisher Resources

ISBN: 0738485780Purchase Link

Cloud Computing

Data Engineering

Data Science

AI & ML

Programming Languages

Software Architecture

IT/Ops

Security

Design

Business

Soft Skills