DB2 performance optimization criteria

The following guidelines should be taken into account when developing an overall performance optimization plan:

1. Keep in mind that the biggest performance benefit from the law of reduction in return is usually from initial efforts. Future changes generally only produce smaller and smaller benefits, and require more efforts.

2. Do not optimize for optimization to lift consistency constraints. If optimization resources are not the main cause of performance problems, such optimization will not affect the response time unless it is subject to major constraints, in addition, this optimization will actually make the subsequent optimization work more difficult. If there is any important potential improvement factor, it is to improve the performance of some resources, which are the main factor affecting the response time.

3. Consider that the entire system can never have only one parameter or system. Before any optimization, you must consider the impact it has on the entire system.

4. Modify only one parameter at a time. Do not change multiple performance optimization parameters at a time. Even if all changes are positive, there is no way to assess the impact of each change. If multiple parameters are modified at a time, the advantages and disadvantages of the changes cannot be effectively determined. If you optimize a parameter every time to improve one aspect, it will almost always affect at least one other aspect that is not considered.

5. Measurement and reconfiguration by level because only one parameter should be modified at a time, only one level of the system can be optimized at a time. You can use the following system-level list as a reference:

L hardware

L Operating System

L application server and requester

L database l SQL statements

L Application

6. Check for hardware and software problems. Some performance problems can be solved through hardware repair or software repair. If the problem can be solved through repair or repair, it does not need to spend too much time on monitoring and optimizing the system.

7. Identify the problem before upgrading the hardware. Even if you can immediately improve performance by adding memory or processor resources, you should take the time to understand the bottleneck of the system. It is possible that the system does not have any processor resources to process it after spending more money to increase disk memory.

8. before optimization, the rollback process takes effect on the existing system. If the optimization fails to achieve the expected effect or even bring negative effects, you need to cancel the changes, therefore, you must be prepared for this purpose.
 

Before optimizing DB2 performance, the content mentioned above must be strictly observed. Only in this way can DB2 performance optimization be improved. I am very happy to share this article with you. I hope the knowledge in this article will be helpful to you. This article will also be a good reference for you to optimize DB2 performance in the future.