Redpaper DB2 9 for z/OS: Backup and Recovery I/O Related Performance Considerations
Hi DB2 User,
IBM just published a new Redpaper, "DB2 9 for z/OS: Backup and Recovery I/O Related Performance Considerations, REDP-4452-00":
This IBM® Redpaper provides best practices and I/O-related performance considerations for backup and recovery using DB2® 9 for z/OS®. It describes some guidelines for DB2 for z/OS backup and recovery, then it discusses performance considerations that may be used to predict recovery time, which can be used, in turn, to adjust backup frequencies or methods.
Table of Contents
This Redpaper has the following sections:
“ Planning the backup strategy” on page 2
This section deals with the series of decisions which a backup strategy involves, beginning on the day that disk and tape hardware are purchased. Configuring the DASD hardware is also the subject of Part 1. At this time some fundamental decisions are made that affects both the probability of a hardware failure and the amount of time that it takes you to recover from that failure. The decisions made at this time also affect the performance of online transaction workloads. Part 1 also explains and contrasts the two major types of backups, image copies and system level backups.
“ Recovery strategies” on page 16
This section presents some recovery strategies for recovering the entire DB2 system to the current point in time after a hardware failure. This part explains the factors that affect the recovery time and how to predict the recovery time. Armed with the knowledge of how long recovery will take, it is easier to develop an effective backup strategy that enables you to achieve the recovery time objectives.
“ Backup strategies” on page 22
This section analyzes hypothetical workload and failure scenarios: we examine a scenario of hardware failure and two scenarios involving programming errors. We examine related recovery to the current point in time and to a prior point in time.
This paper references and builds upon the IBM Redbooks® publication Disaster Recovery with DB2 UDB for z/OS, SG24-6370, which focused on disaster recovery, but also introduced some concepts about FlashCopy® and system level backups. That book was written prior to DB2 9 for z/OS, which added some new features that use FlashCopy and system level backups, and prior to the availability of the DS8000™ . This paper contains a discussion of DB2 9 features and DS 8000 functions.
More information about Disaster Recovering
- Redbook Disaster Recovery with DB2 UDB for z/OS, SG24-6370-00
DB2 for z/OS is the database of choice for critical data for many enterprises. It is becoming more and more important to protect this data in case of disaster and to be able to restart with a consistent copy of the DB2 data as quick as possible and with minimal losses.
A broad range of functions can be used for the disaster recovery of DB2 subsystems. The traditional DB2 based solution consists of safe keeping and restoring image copies and logs. More general functions, applicable not only to DB2 data, but to the whole system, are hardware related, such as tape vaulting or disk volumes mirroring. Other functions are specific to DB2 such as the Tracker Site. There are also products providing replication capabilities which can be used for specific propagation requirements.
DB2 UDB for z/OS Version 8 has introduced two new subsystem wide utilities, BACKUP and RESTORE, which, by interfacing the copy pools functions of DFSMS 1.5, are able to provide Point-In-Time recovery capabilities.
The disaster recovery solution consists of the combination of coherent options that best fit in with the requirements, the current environment, and the investment.
In this IBM Redbook we first introduce the main concepts, and the primary components for possible solutions. We then describe the most common solutions, and implement several recovery scenarios. All our tests were implemented with DB2 UDB for z/OS Version 8. We also include criteria for choosing a solution, and recommendations based on recovery best practices.
We focus on requirements and functions available for a disaster recovery strategy for data stored and managed by DB2 for z/OS. It is worth remembering that the non-DB2 data, logically or physically related to the DB2 applications, should be treated with equivalent and congruent solutions.
Table of Contents
Part 1. The whole picture
Chapter 1. Business continuity
Chapter 2. DB2 disaster recovery
Part 2. Disaster recovery major components
Chapter 3. Traditional recovery
Chapter 4. DB2 Tracker
Chapter 5. ESS FlashCopy
Chapter 6. SMS copy pools and DB2 point in time recovery
Chapter 7. Peer-to-Peer Remote Copy
Chapter 8. eXtended Remote Copy
Part 3. General solutions for disaster recovery
Chapter 9. Split Mirror
Chapter 10. FlashCopy Consistency Group
Chapter 11. Global Copy PPRC-XD
Chapter 12. Global Mirror PPRC
Chapter 13. XRC: Global Mirror for z/OS
Chapter 14. Geographically Dispersed Parallel Sysplex
Part 4. Implementing disaster recovery scenarios
Chapter 15. Set Log Suspend - FlashCopy - More Log - Restore System Log Only
Chapter 16. FlashCopy Consistency Group and restart
Chapter 17. PPRC - FlashCopy from secondary
Chapter 18. XRC and restart
Chapter 19. Local recovery: System PITR
Chapter 20. Restart using tape dump of copy pools
Part 5. Additional considerations
Chapter 21. Data sharing
Chapter 22. Validation and performance
Appendix A. REXX procedures
Appendix B. PITR definitions
Appendix C. Additional material