Home » IBM Manuals » Computer Software » IBM E16RMLL-I » Manual Viewer

IBM E16RMLL-I Implementation Guide - Page 535

THROUGHPUT thresholds, COLLOCATION

Add to My Manuals
Save this manual to your list of manuals

Page 535 highlights

Tip: If SELFTUNEbufpoolsize is set to YES, expire inventory processing will also trigger the buffer pool performance monitoring algorithm to check whether a performance improvement could be gained (based on previous server execution and available real memory). THROUGHPUT thresholds Use the THROUGHPUTTIMETHRESHOLD and THROUGHPUTDATATHRESHOLD to control how Tivoli Storage Manager handles slow performing client sessions. Slow performing sessions may pin your recovery log and result in slower overall server performance. These options allow you to specify thresholds that must be met by client sessions in order to be considered productive and avoid cancellation. The THROUGHPUTDATATHRESHOLD is specified in kilobytes per second, and specifies the throughput that a client session must reach to prevent being cancelled after the time threshold is reached. The THROUGHPUTTIMETHRESHOLD is specified in minutes, and is added to the media wait time, to determine the total client session time. It specifies the time threshold for a session after which it may be cancelled for low throughput. If the server detects that the client data transfer rate is below the specified threshold, the session is considered to be eligible for termination after the time threshold is reached. COLLOCATION The speed at which you can restore the files, among other parameters, also depends on how many tape drives are available on the server, and whether you are using collocation to keep file spaces assigned to as few volumes as possible. With tape capacities becoming larger and larger, collocation by groups of nodes is now supported. Groups of nodes can be defined, and the server can then collocate data based on these groups. Collocation by group can yield the following benefits: Reduce unused tape capacity by allowing more collocated data on individual tapes. Minimize mounts of target volumes. Minimize database scanning and reduce tape passes for sequential-to-sequential transfer. For newly defined storage pools, the default storage pool collocation setting is now group. Chapter 15. Performance considerations 505

Section	Page
Front cover	1
Contents	5
Figures	19
Notices	25
Trademarks	26
Preface	27
The team that wrote this redbook	27
Become a published author	30
Comments welcome	30
Part 1 Introduction	31
Chapter 1. Implementation checklists	33
1.1 The big picture	34
1.1.1 Our support material	35
1.2 Planning checklist	37
1.2.1 Server implementation checklist	38
1.2.2 Client implementation checklist	41
1.2.3 Operations checklist	42
1.3 Summary	43
Chapter 2. Implementation planning	45
2.1 Planning	46
2.2 Client environment data	46
2.2.1 Client name	49
2.2.2 Contact information	49
2.2.3 Operating system	49
2.2.4 Total storage available	50
2.2.5 Total storage used	50
2.2.6 GB changed per backup	50
2.2.7 Number of files backed up	51
2.2.8 Number of backup versions to be kept	51
2.2.9 Data compression	51
2.2.10 Backup window times	53
2.2.11 Backup number of hours	53
2.2.12 Required recovery time frame	53
2.2.13 Tivoli Storage Manager restore time frame	53
2.2.14 GB copied per archive	54
2.2.15 Number of files archived	54
2.2.16 Number of archives kept	54
2.2.17 Archive frequency	54
2.2.18 Archive window times	54
2.2.19 Archive number of hours	54
2.2.20 Number of image backups	55
2.2.21 Image backup frequency	55
2.2.22 Number of backup sets	55
2.2.23 Backup set frequency	55
2.2.24 Policy domain	55
2.2.25 Client option set	55
2.3 Data retention requirements	56
2.3.1 Group name	56
2.3.2 Number of backup versions	57
2.3.3 Backup file retention period	57
2.3.4 Number of deleted versions	57
2.3.5 Last deleted file version retention period	57
2.3.6 Archive retention period	58
2.3.7 Off-site copies	58
2.3.8 Onsite collocation	58
2.3.9 Off-site collocation	59
2.3.10 Image backup retention	59
2.3.11 Backup set retention	60
2.4 Server architecture considerations	60
2.4.1 Server platform	60
2.4.2 Installed user base	61
2.4.3 Cost	61
2.4.4 Capacity	62
2.4.5 Platform installation	62
2.4.6 Operation	62
2.4.7 Supported devices	63
2.5 System size	63
2.6 Multiple Tivoli Storage Manager servers	64
2.6.1 Reasons to consider multiple servers	64
2.6.2 Disadvantages of multiple servers	65
2.7 Network	65
2.7.1 Network considerations	65
2.7.2 Communication protocol	68
2.7.3 Network name resolution	69
2.8 Disk considerations and sizing	69
2.8.1 The disk subsystem	69
2.8.2 Tivoli Storage Manager database	69
2.8.3 Database performance	70
2.8.4 Database size	71
2.8.5 Archive sizing	72
2.8.6 Database volume identification	73
2.8.7 Tivoli Storage Manager recovery log	73
2.8.8 Primary disk storage pool	74
2.8.9 Disk storage pool size	76
2.8.10 Archive disk sizing	76
2.8.11 Image disk sizing	77
2.8.12 Device configuration table and volume history file	78
2.8.13 Total disk	79
2.9 Tape drives and sizing	80
2.9.1 Tape devices	82
2.9.2 Tape libraries	82
2.9.3 Number of tape drives	82
2.10 Tape volumes	83
2.10.1 On-site volumes	83
2.10.2 Off-site volumes	86
2.10.3 Database tape volumes	86
2.10.4 z/OS tape management	88
2.11 Administrator IDs	89
2.11.1 License considerations	90
2.11.2 Licensed features registration	90
2.12 Other considerations	92
2.13 Summary	93
Part 2 Installation	95
Chapter 3. Server installation	97
3.1 Software installation	98
3.2 Latest software updates	98
3.3 AIX server installation	100
3.4 Linux server installations	102
3.4.1 Installation packages	102
3.4.2 Installation commands	103
3.4.3 Post-installation steps	106
3.4.4 Uninstallation of the server	108
3.5 Windows 2000/2003 server installation	108
3.5.1 Tivoli Storage Manager server package installation	108
3.5.2 Tivoli Storage Manager License package installation	110
3.5.3 Tivoli Storage Manager Device Driver package installation	111
3.5.4 Windows 2000/2003 configuration wizards	112
3.6 Integrated Solution Console and Administration Center	113
3.6.1 ISC and Administration Center installation	115
3.7 Customization	121
3.7.1 Server options file	121
3.7.2 Settings	123
3.8 Summary	125
Chapter 4. Backup-archive client installation	127
4.1 Backup-archive client code	128
4.1.1 Backup-archive client	128
4.1.2 Web client	128
4.1.3 API client	131
4.2 Code installation	131
4.2.1 Backup-archive client code	132
4.2.2 AIX backup-archive client	132
4.2.3 Linux backup-archive client	136
4.2.4 Windows backup-archive client	138
4.3 Customization	143
4.3.1 Backup-archive client customization	143
4.3.2 Web backup-archive client	150
4.3.3 Using the client configuration wizard	150
4.4 UNIX/Linux configuration wizard	151
4.5 Use of Windows configuration wizards	159
4.5.1 Backup-archive client configuration	159
4.5.2 Web client configuration	169
4.5.3 Client scheduler configuration	179
4.5.4 Journal engine configuration	187
4.5.5 Online image support configuration	195
4.5.6 Open File Support configuration	197
4.6 Client interfaces	198
4.6.1 Command line	200
4.6.2 GUI	202
4.7 Client scheduler	203
4.7.1 Starting the client scheduler	204
4.7.2 Stopping the scheduler	210
4.8 Web client usage	211
4.8.1 Remote client GUI	211
4.8.2 Client functions	211
4.8.3 Access authorization	212
4.8.4 Starting the Web client	213
4.8.5 Stopping the remote agent services	214
4.9 Summary	216
Chapter 5. Database and recovery log	217
5.1 Database	218
5.1.1 Database design considerations	218
5.1.2 Defining database volumes	220
5.2 Recovery log	222
5.2.1 Recovery log design considerations	223
5.2.2 Defining recovery log volumes	224
5.3 Setting the log mode	226
5.4 Defining the database backup trigger	228
5.5 Setting the expansion trigger	229
5.5.1 Database space trigger	230
5.5.2 Recovery log space trigger	231
5.6 Mirroring	232
5.6.1 Database mirroring	235
5.6.2 Recovery log mirroring	236
5.7 Removing default volumes	237
5.7.1 Removing the default database volume	237
5.7.2 Removing the default recovery log volume	239
5.8 Database backup	240
5.9 Additional commands	242
5.9.1 DELETE VOLHIST TYPE=DBBACKUP	242
5.9.2 ESTIMATE DBREORGSTATS	243
5.10 Summary	244
Chapter 6. Data storage	245
6.1 Example environment	246
6.1.1 Primary storage pools	247
6.1.2 Copy storage pools	247
6.2 Devices	248
6.2.1 Defining a library	252
6.2.2 Defining a path to a library	253
6.2.3 Defining a drive in a library	255
6.2.4 Defining a path to a drive in a library	256
6.2.5 Defining a device class for a library	258
6.2.6 Defining a sequential file device class	260
6.3 Storage pools	261
6.3.1 Defining a sequential access tape storage pool	262
6.3.2 Defining a random access disk storage pool	263
6.3.3 Defining a sequential access file storage pool	264
6.3.4 Defining a copy storage pool	267
6.3.5 Deleting the default storage pools	268
6.4 Storage pool volumes	269
6.4.1 Defining random access disk volumes	269
6.4.2 Tape volumes	271
6.5 Additional commands	275
6.5.1 Auditing library contents	275
6.5.2 Auditing volume contents	276
6.5.3 Back up a storage pool	278
6.5.4 Check a tape in to a library	280
6.5.5 Check out library volumes	283
6.5.6 Deleting storage-related objects	285
6.5.7 Mounted volumes	286
6.5.8 Migration	287
6.5.9 Moving data	288
6.5.10 Querying volume contents	290
6.5.11 Querying occupancy	290
6.5.12 Rename a storage pool	291
6.5.13 Reclamation	292
6.5.14 SQL commands	294
6.6 Summary	297
Chapter 7. Data storage policies	299
7.1 Recommended setup	300
7.1.1 Defining policy domains	302
7.1.2 Defining policy sets	302
7.1.3 Defining management classes	302
7.1.4 Defining backup copy groups	304
7.1.5 Defining the archive copy group	306
7.2 Verifying policy definitions	307
7.2.1 Backup copy groups	307
7.2.2 Archive copy groups	308
7.3 Validating and activating a policy set	308
7.3.1 Validating the recommended policy sets	308
7.3.2 Activating the recommended policy sets	309
7.3.3 Deleting the STANDARD policy domain	309
7.4 Enforcing your policy	312
7.5 Summary	312
Chapter 8. Managing Tivoli Storage Manager	313
8.1 Management	314
8.1.1 Administrators	314
8.1.2 Considerations	315
8.1.3 Default environment	315
8.1.4 Recommended administrators	316
8.1.5 Working with administrators	318
8.2 Client nodes	321
8.2.1 Considerations	321
8.2.2 Default environment	322
8.2.3 Working with client nodes	322
8.3 Client option sets	328
8.3.1 Considerations	328
8.3.2 Default environment	329
8.3.3 Recommended client option sets	329
8.3.4 Working with client option sets	330
8.3.5 Associating a client node with a client option set	331
8.4 Summary	332
Chapter 9. Licensing	333
9.1 Licensed features	334
9.2 Registering licensed features	335
9.3 Saving your licenses	336
9.4 License compliance	336
9.5 Monitoring licenses	337
9.5.1 Displaying license information	337
9.5.2 Auditing licenses	338
9.5.3 Scheduling automatic license audits	338
9.6 Tivoli Storage Manager V5.2 license features	339
Chapter 10. Administrative client	341
10.1 Administration Center	342
10.1.1 Installation	342
10.1.2 Administration Center interface	343
10.2 Administrative client	346
10.2.1 Code installation	347
10.2.2 Administrative command-line client customization	348
10.2.3 Command-line interface	350
10.3 Summary	353
Part 3 Operational details	355
Chapter 11. Client operations	357
11.1 Running backup operations	358
11.1.1 Exclude rules preventing some files from being backed up	361
11.1.2 UNIX command-line examples and output	363
11.1.3 Windows GUI backup examples	372
11.1.4 Additional backup options for Windows	374
11.2 Running restore operations	375
11.2.1 UNIX command-line examples and output	378
11.2.2 Windows GUI restore examples	384
11.3 Running archive operations	389
11.3.1 UNIX command-line examples and output	391
11.3.2 Windows GUI archive examples	393
11.4 Running retrieve operations	395
11.4.1 UNIX command-line examples and output	396
11.4.2 Windows GUI retrieve examples	397
11.5 Summary	400
Chapter 12. Scheduling	401
12.1 The wheel of life	403
12.2 Administrative schedules	404
12.2.1 Defining an off-site backup schedule	407
12.2.2 Defining the volume history schedules	410
12.2.3 Defining a migration schedule	412
12.2.4 Defining an expiration schedule	412
12.2.5 Defining a reclamation schedule	413
12.2.6 Defining a licensing audit schedule	415
12.2.7 Querying administrative events	416
12.3 Client schedules	416
12.3.1 Defining a client backup schedule	417
12.3.2 Defining an enhanced client schedule	418
12.3.3 Associating a client with a schedule	421
12.3.4 Verifying the client schedules	422
Chapter 13. Routine tasks	425
13.1 Operations staff	426
13.2 Server procedures	427
13.2.1 Starting the Tivoli Storage Manager server	427
13.2.2 Stopping the Tivoli Storage Manager server	431
13.3 Event monitoring	434
13.3.1 Event receivers	435
13.4 Health monitoring	437
13.4.1 Enabling the ADMIN_CENTER account	438
13.4.2 Using Health Monitoring	439
13.5 Operational Reporting	442
13.5.1 Functions	442
13.6 Daily sanity checks	444
13.6.1 Data storage	447
13.6.2 Client-server activity	451
13.7 Storage media management	456
13.7.1 Tape use overview	456
13.7.2 Label and check in tapes	458
13.7.3 On-site and off-site tape management	460
13.7.4 Moving data from on-site to off-site	463
13.7.5 Off-site tape management to on-site	469
13.7.6 Checking volumes into a library	470
13.7.7 Reclaiming off-site tapes	471
13.7.8 Database backup management	472
13.8 Error conditions	475
13.8.1 Tivoli Storage Manager errors	476
13.8.2 Machine errors	477
13.9 Summary	481
Chapter 14. Advanced operations	483
14.1 Exporting server to server	484
14.2 Exporting data directly to another server	484
14.2.1 Preparing to export to another server for immediate import	485
14.2.2 Exporting administrator information to another server	485
14.2.3 Exporting client node information to another server	487
14.2.4 Exporting policy information to another server	488
14.3 Exporting and importing server to server	490
14.3.1 Moving a complete node’s data and meta data	490
14.3.2 Moving a node’s metadata	496
14.4 Moving a node back to an originating server	496
14.4.1 Merging file spaces	497
14.4.2 Suggestions for leveraging the export feature	501
14.5 Server groups and remote command routing	501
14.6 Reorganizing the database	503
14.7 Tivoli Storage Manager and TEC integration	508
14.8 Summary	509
Chapter 15. Performance considerations	511
15.1 How to measure performance	512
15.1.1 Network benchmarking	513
15.1.2 Tivoli Storage Manager client performance tracing	518
15.1.3 Tivoli Storage Manager server performance tracing	522
15.2 Architecture-based performance tuning	525
15.2.1 Database and recovery log	526
15.2.2 Storage pools	526
15.2.3 Versioning	527
15.2.4 Client configuration	528
15.2.5 LAN-free backup/restore	528
15.3 Tivoli Storage Manager server	528
15.3.1 Server database and recovery log	528
15.3.2 Transactions	531
15.3.3 Communication	532
15.3.4 General parameters	534
15.4 Client node	536
15.4.1 Transactions	536
15.4.2 Communication	538
15.4.3 General parameters	539
15.5 System design for performance	541
15.5.1 PCI busses	541
15.5.2 Tape busses (SCSI, Fibre)	541
15.5.3 Disk topology	542
15.5.4 System memory	542
15.5.5 Network	542
15.5.6 Tape devices	543
15.6 Special performance tips	543
15.6.1 LAN-free tuning	543
15.6.2 LTO/DLT tape tuning	543
15.6.3 File system volumes versus raw logical volumes	544
15.6.4 AIX virtual memory system tuning	544
15.6.5 Use NTFS partitions for the server	546
15.6.6 Journal-based incremental backup	547
15.7 Summary	550
Part 4 Advanced topics	553
Chapter 16. Leveraging SAN environments	555
16.1 LAN-free prerequisites	556
16.2 Server setup for LAN-free	558
16.2.1 Defining library, drives, and associated paths on the server	560
16.2.2 Defining the device class	561
16.2.3 Creating primary sequential storage pool for LAN-free	561
16.2.4 Defining policy	562
16.2.5 LAN-free validation	563
16.3 Storage Agent setup	564
16.3.1 Device drivers for shared devices	564
16.3.2 Storage Agent software installation	567
16.3.3 Configuring the Storage Agent	570
16.3.4 Device configuration on the Storage Agent	578
16.4 SAN device discovery support	582
16.4.1 Recovering from offline paths	583
16.5 Client LAN-free customization	586
16.6 Performing LAN-free operations	588
16.6.1 Determining whether the data movement is LAN-free	590
16.6.2 LAN-free data transfer considerations	592
16.7 Summary	592
Chapter 17. Server-free data movement	595
17.1 Server-free: What it is and why to use it	596
17.2 Server-free setup requirements	598
17.3 Configuration steps: Overview	600
17.4 Setting up server-free data movement	600
17.4.1 SAN configuration	600
17.4.2 SAN zoning	602
17.4.3 Setting up the SAN Data Gateway (datamover)	604
17.4.4 Configuring Tivoli Storage Manager server	605
17.4.5 Configuring the Tivoli Storage Manager Client	609
17.5 Running server-free backup and restore	609
17.6 Summary	611
Chapter 18. Network Data Management Protocol (NDMP)	613
18.1 NDMP terminology	614
18.1.1 Tivoli Storage Manager and NDMP	614
18.1.2 NDMP backup for NAS	614
18.1.3 NDMP support	615
18.1.4 Multiple NAS appliances	617
Chapter 19. Disaster Recovery Manager	619
19.1 Example of a DRM implementation	620
19.2 DRM setup	621
19.2.1 Register DRM license	621
19.2.2 Create a copy storage pool	622
19.2.3 DRM settings	624
19.2.4 Verifying the settings	628
19.3 Daily operations	628
19.3.1 Back up primary storage pools to copy storage pool	630
19.3.2 Backup of Tivoli Storage Manager database	633
19.3.3 Querying DR media	634
19.3.4 Send disaster recovery media off-site	635
19.3.5 Generate the recovery plan	639
19.3.6 Returning expired volumes	640
19.4 Server restore setup	641
19.4.1 Obtain the latest DR plan	642
19.4.2 Find a replacement server and storage	643
19.4.3 Install the operating system and the server	643
19.5 Break out the disaster recovery plan	643
19.5.1 Obtain the recovery volumes	646
19.5.2 Review the Tivoli Storage Manager macros	646
19.5.3 Review the device configuration	648
19.5.4 Start the restore Tivoli Storage Manager server scripts	650
19.5.5 Restore primary storage pools	655
19.5.6 Summary of example disaster recovery plan	657
19.6 Summary	657
Chapter 20. Bare Machine Recovery	659
20.1 Windows Bare Machine Recovery	660
20.1.1 Collect client machine information for disaster recovery	660
20.1.2 Collect partition and logical volume information with diskmap	663
20.1.3 Store system information for DRM access	663
20.1.4 Insert client machine information into DRM	664
20.1.5 Use machchar.vbs to insert machine reports into DRM	665
20.1.6 Windows systems additional information	668
20.2 Using SysBack for Bare Machine Recovery	669
20.2.1 An introduction to SysBack	671
20.2.2 System installation options	677
20.2.3 Network boot	678
20.2.4 Multivolume backup and tape device support	680
20.2.5 Partition backup, recovery, and cloning	681
20.2.6 Partition backup and reinstallation (system recovery)	681
20.2.7 Cloning backup images between partitions	687
20.2.8 Cloning from a stand-alone system to a partition	695
20.2.9 License information	696
20.2.10 Documentation	697
20.3 Integrating SysBack with Tivoli Storage Manager	697
20.3.1 Prerequisites, limitations, and exclusions	697
20.3.2 Basic setup and configuration tasks	700
20.3.3 Creating a Tivoli Storage Manager virtual device	702
20.3.4 Configuring network boot options for BMR	705
20.3.5 Recovery and system reinstallation from a server	707
Chapter 21. Data Protection configuration on the server	711
21.1 Basic assumptions	712
21.2 Policy creation requirements	712
21.2.1 Defining a management class in an existing policy domain	712
21.2.2 Defining a new policy domain	713
21.2.3 Backing up and archiving copygroup considerations	714
21.3 Register node	717
21.4 Server configuration considerations	718
Chapter 22. Tivoli Storage Manager upgrade considerations	721
22.1 General upgrade considerations	722
22.1.1 Server	722
22.1.2 Clients and Data Protection modules	724
22.1.3 Storage Agent	727
22.2 Server upgrade best practices	727
22.2.1 Server quiesce	728
22.2.2 Backing up important components	728
22.2.3 Upgrading the server	729
22.2.4 Testing new updates	729
22.2.5 Enabling production mode	730
22.3 Performing server upgrades	730
22.3.1 Migration on Windows	731
22.3.2 Migration on AIX 5L	737
22.4 Performing client upgrades	745
22.4.1 Windows	745
22.4.2 AIX 5L client upgrade	746
22.5 Storage Agent upgrade	748
22.5.1 AIX Storage Agent migration	748
22.5.2 Windows 2003 Storage Agent migration	749

Match case Limit results 1 per page

Chapter 15. Performance considerations

505

THROUGHPUT thresholds

Use the THROUGHPUTTIMETHRESHOLD and

THROUGHPUTDATATHRESHOLD to control how Tivoli Storage Manager

handles slow performing client sessions. Slow performing sessions may pin your

recovery log and result in slower overall server performance. These options

allow you to specify thresholds that must be met by client sessions in order to be

considered productive and avoid cancellation.

The THROUGHPUTDATATHRESHOLD is specified in kilobytes per second, and

specifies the throughput that a client session must reach to prevent being

cancelled after the time threshold is reached.

The THROUGHPUTTIMETHRESHOLD is specified in minutes, and is added to

the media wait time, to determine the total client session time. It specifies the

time threshold for a session after which it may be cancelled for low throughput.

If the server detects that the client data transfer rate is below the specified

threshold, the session is considered to be eligible for termination after the time

threshold is reached.

COLLOCATION

The speed at which you can restore the files, among other parameters, also

depends on how many tape drives are available on the server, and whether you

are using collocation to keep file spaces assigned to as few volumes as possible.

With tape capacities becoming larger and larger, collocation by groups of nodes

is now supported. Groups of nodes can be defined, and the server can then

collocate data based on these groups. Collocation by group can yield the

following benefits:

Reduce unused tape capacity by allowing more collocated data on individual

tapes.

Minimize mounts of target volumes.

Minimize database scanning and reduce tape passes for

sequential-to-sequential transfer.

For newly defined storage pools, the default storage pool collocation setting is

now

group

Tip:

If SELFTUNE

bufpoolsize

is set to YES,

expire inventory

processing will

also trigger the buffer pool performance monitoring algorithm to check

whether a performance improvement could be gained (based on previous

server execution and available real memory).