Viewer
Overview
The Performance Advisor in Tivoli Performance Viewer (TPV) provides advice to help tune systems for optimal performance and gives recommendations on inefficient settings by using collected Performance Monitoring Infrastructure (PMI) data. Advice is obtained by selecting the Performance Advisor icon in TPV. The Performance Advisor in TPV provides more extensive advice than the Runtime Performance Advisor. For example, TPV provides advice on setting the dynamic cache size, setting the JVM heap size and using the DB2 Performance Configuration Wizard.
1. Enable PMI services in the appserver
To monitor performance data through the PMI interfaces, you must first enable the performance monitoring service through the administrative console before restarting the server. If running in Network Deployment, you need to enable PMI services on both the server and on the node agent before restarting the server and the node agent.
2. Enable PMI services in Node Agent.
If running Network Deployment, you must enable PMI service on both the server and on the node agent, and restart the server and node agent.
3. Enable data collection.
The monitoring levels that determine which data counters are enabled can be set dynamically, without restarting the server. These monitoring levels and the data selected determine the type of advice you obtain. The Performance Advisor in TPV uses the standard monitoring level; however, the Performance Advisor in TPV can use a few of the more expensive counters (to provide additional advice) and provide advice on which counters can be enabled. This action can be completed in one of the following ways:
a. Setting performance monitoring levels .
b. Enabling performance monitoring services using the command line.
4. Start the Tivoli Performance Viewer.
5. Simulate a production level load.
Simulate a realistic production load for your application, if you are using the Performance Advisor in a test environment, or doing any other performance tuning. The application should run this load without errors. This simulation includes numbers of concurrent users typical of peak periods, and drives system resources such as CPU and memory to the levels expected in production. The Performance Advisor only provides advice when CPU utilization exceeds a sufficiently high level. For a list of IBM business partners providing tools to drive this type of load, see the article, Performance: Resources for learning in the sub-section of Monitoring performance with third party tools.
6. (Optional) Store data to a log file.
7. (Optional) Replay a performance data log file.
8. Refresh data.
Clicking refresh with server selected under the viewer icon causes TPV to:
o Query the server for new PMI and product configuration information.
Clicking refresh with server selected under the advisor icon causes TPV to:
o Refresh advice that is provided in a single instant in time.
o Not query the server for new PMI and product configuration information.
9. Tuning advice appears when the Advisor icon is chosen in the TPV Performance Advisor. Double-click an individual message for details. Since PMI data is taken over an interval of time and averaged to provide advice, details within the advice message appear as averages.
If the Refresh Rate is adjusted, the Buffer Size should also be adjusted to allow sufficient data to be collected for performing average calculations. Currently 2 minutes of data is required. Read more about adjusting the Refresh Rate or Buffer Size at:
o Change the display buffer size.
o Change the display buffer size.
10. Update the product configuration for improved performance, based on advice. Since Tivoli Performance Viewer refreshes advice at a single instant in time, take the advice from the peak load time.
Although the performance advisors attempt to distinguish between loaded and idle conditions, misleading advice might be issued if the advisor is enabled while the system is ramping up or down. This result is especially likely when running short tests. Although the advice helps in most configurations, there might be situations where the advice hinders performance. Due to these conditions, advice is not guaranteed. Therefore, test the environment with the updated configuration to ensure it functions and performs well.
Over a period of time the advisor may issue differing advice. This is due to load fluctuations and runtime state. When differing advice is received the user should look at all advice and the time period over which it was issued. Advice should be taken during the time that most closely represents peak production load.
Performance tuning is an iterative process. After applying advice, simulate a production load, update the configuration based on the advice, and retest for improved performance. This procedure should be continued until optimal performance is achieved.
11. Clearing values from tables and charts.
12. Resetting counters to zero.
Subscribe to:
Post Comments (Atom)
No comments:
Post a Comment