DB2 Workload Management helps prevent, detect, and resolve conflicting resource requirements on a DB2 data server based on defined business objectives. This paper details the current best practices for DB2 workload manager.
Today, for both strategic and financial reasons, many businesses are consolidating
multiple individual data servers onto a single shared data server. As each new data
server is merged, it potentially adds a very different type of workload to the mix with
different system interactions. The consolidated server must now support a variety of
different, concurrent workloads. Good workload management practices are critical to
meeting business commitments in such an environment.
When your database system encounters performance degradation due to the different,
and sometimes conflicting, resource demands from the work being executed on it, then
you need DB2 workload manager to help you prevent, detect, and resolve these conflicts.
This paper introduces you to the current best practices for DB2 workload manager on
your DB2 Version 9.5 for Linux, UNIX, and Windows data server which can be used to
help you meet your business objectives for the work being executed on DB2 data servers.
The best practices presented here are based on IBM field experience in benchmark and
proof-of-concept exercises as well as feedback from customer adoptions of DB2 Version
9.5.
There are three sections to this paper covering: - Best practices in design
- Best practices in implementation
- Best practices in monitoring
There are also appendices in this paper that provide you with more in-depth information
and guidance such as: - A specific scenario that shows you how to help meet your business priorities
with DB2 workload manager in a data warehouse environment
- An example script to use with the AIX® Workload Manager to achieve dynamic
processor allocation while using hard maximum limits on CPU use
- Guidance on how to upgrade from Query Patroller and DB2 Governor
This paper will be most useful to you if you are already familiar with some of the basic
concepts of DB2 workload management. If you find that you would like more
background information first, you can refer to “Introduction to DB2 9.5 workload
management” or to some of the further reading material suggested at the end of this
paper
- Executive summary
- Best Practices: Design
- Understand existing systems
- Assigning business priority to database work
- Take advantage of your experience
- Start simply and change slowly
- Best practices: Implementation
- DB2 workloads
- DB2 service classes
- DB2 work action sets
- DB2 thresholds
- Best practices: Monitoring
- Conclusion
- Further Reading
- Appendices
- Scenario: Best practices in a data warehousing environment
- Understand the work your data warehouse server performs
- Determine the connection attributes for use in workload definitions
- Obtain consistent response times with priority settings
- Define work action sets to distinguish between types of work
- Integrate with AIX Workload Manager
- Protect your system from being overloaded
- Tame the monster query
- Limit the number of concurrent load operations
- Monitor expensive queries
- Analyze event monitor data
- Moderating AIX Workload Manager hard maximum processor use limits
- Think DB2 workload manager when upgrading from Query Patroller and DB2
Governor
- Why is there no tool to upgrade from Query Patroller automatically?I
think
- Notices
"
Best Practices:
DB2 Workload Management
"
(October 2008)
DB2 Workload Management helps prevent, detect, and resolve conflicting resource requirements on a DB2 data server based on defined business objectives. This paper details the current best practices for DB2 workload manager. (pdf; 1.45MB; 48 pages)
Comment, edit, or add your own insights to the DB2 Workload Management best practices
on the IBM Database Wiki.
Check out all the other Best Practices papers and see how you can improve your experience with DB2 for Linux, UNIX, and Windows.
| 
