+

Search Tips   |   Advanced Search

Video: How do I collect performance information for WAS?

The following transcript is for the "How do I collect performance information for WebSphere Application Server?" Video, which describes how to collect performance information for the WAS. This transcript is the video storyboard. Audio describes narration and captions. Onscreen Action describes the content that is shown in the video.

How do I collect performance information for WAS?

Scene Audio Onscreen Action
1 Question: How do I collect performance information for WAS? Show title, How do I collect performance information for WAS?.
2

Access the WebSphere Integrated Solutions Console for the servers to monitor.

The Performance Monitoring Infrastructure (PMI) is used to identify and track Key Performance Indicators for applications.

Show the user how to log in to the WebSphere Integrated Solutions Console for the servers we want to monitor.

The Performance Monitoring Infrastructure (PMI) is used to identify and track Key Performance Indicators for applications.

3

Click Monitoring and Tuning.

Click Performance Monitoring Infrastructure (PMI).

Click a server name.

Click Monitoring and Tuning from the left navigation menu.

Click Performance Monitoring Infrastructure (PMI).

To collect performance information for a WebSphere Application server click the server name from your list of servers.

4

Select the right performance indicators is an iterative process. Selecting the Basic monitoring level is a good starting point.

Enable PMI here and select the statistics in the boxes below.

Select the appropriate monitoring level.

Basic is a good starting point for determining performance indicators.

The page also shows the information that is tracked.

Select the Basic monitoring option.

Click Apply or OK.

Select the right performance indicators is an iterative process. Selecting the Basic monitoring level is a good starting point.

Enable PMI by selecting the Box.

Select the appropriate monitoring level from the list of level options.

Basic is a good starting point for determining performance indicators.

The page also shows the information that is tracked. Show user how to expand the twistie to view the list.

Select the Basic monitoring option.

Click Apply or OK.

5

Save the changes.

The selected level of monitoring will be in place when the server is started or restarted.

The video will also show the other monitoring levels that can be selected.

Save the changes by clicking save.

The selected level of monitoring is in place when the server is started or restarted.

6

In cases where basic monitoring is not sufficient, an extended monitoring level might provide the needed level of information.

Expand the Extended section to see the extended set of statistics.

In cases where basic monitoring is not sufficient, an extended monitoring level might provide the needed level of information.

Show the user how to expand the Extended section to see the extended set of statistics.

7

Enable all statistics is not recommended but might be reviewed to see the full range of information available.

The Custom selection provides a convenient tree of selections.

Click here to see the selection menus.

The custom category provides the granularity and the depth to monitor the key indicators.

Enable all statistics is not recommended but might be reviewed to see the full range of information available. The Custom selection provides a convenient tree of selections.

Show the user where to click to see the selection menus.

The custom category provides the granularity and the depth to monitor the key indicators.

8

Click each category to look at the statistics that can be enabled.

Enterprise beans often determine the overall response.

JDBC statistics can be used as the indicators of the health of the application.

JVM Runtime shows the overall health of the server.

Once we make the selections and restart the server, we can use the performance viewer or enable logging.

Logging is recommended as it consumes fewer resources and allows performance information to be reviewed multiple times.

Click each category to look at the statistics that can be enabled. Show the user different categories and explain what they can do.

Enterprise beans often determine the overall response.

JDBC statistics can be used as the indicators of the health of the application.

JVM Runtime shows the overall health of the server.

When we have made the selections and restarted the server, we can use the performance viewer or enable logging.

Logging is recommended as it consumes fewer resources and allows performance information to be reviewed multiple times.

9

This screen shows the recommended starting selection.

Show a screen example of the recommended starting selections.

10

We can look at the current activity, but it is not recommended because live monitoring and network communications for the information might affect performance. Logging the information does not impose network load, and is less disruptive to performance.

Show the user how to look at the current activity on the left navigation menu.

It is not recommended because live monitoring and network communications for the information might affect performance. Logging the information does not impose network load, and is less disruptive to performance.

11

Select this server to show the current activity and to allow some updates to the data collection properties.

Show the user how to select a server to show the current activity and allow some updates to the data collection properties.

12

This application is not accessed yet, so statistics are not available. We can specify the location and the size of the log file in the settings menu.

Show users where to access the settings menu.

13

The default values are fine for many applications, but they should be increased if the application is complex or if there is an increase in the time period. Otherwise, many log files need to be accessed to review the information.

The name of the log file automatically includes the server name and the date stamp.

A Binary Log Output Format is recommended. If XML is used as the Log Output Format, the XML files will be compressed at the end of the collection period.

Show the user the Logging Output page with the default settings enabled.

14

Click here to start logging.

Logging is in progress and other actions can be taken in the administrative console. This concludes the exercise of enabling the Performance Monitoring Infrastructure.

Show the user where to click Start Logging to begin logging. Logging is started and the user can take other actions within the administrative console.

15

At the conclusion of the monitoring period, IBM recommends that you reset the collected statistics to none, or disable PMI.

At the conclusion of the monitoring period, IBM recommends that you reset the collected statistics to none, or disable PMI.

16

Select the application with PMI enabled.

Show the user how to select the application with PMI enabled.

17

Select None as the Currently monitored statistic set, or clear the check Box on Enable Performance Monitoring Infrastructure.

Click Apply or OK.

Show the user where to select none under the Currently monitored statistic set.

The user can also clear the check Box for Enable Performance Monitoring Infrastructure.

Click Apply or OK.

18

Save the changes.

Save the changes by clicking save.

19

Use the administrative console to restart the server that was updated.

Show the user how to restart the server that was updated using the administrative console.

20

Restart the server for the changes to take effect.

The changes are saved when the server is restarted.

Restart the server for the changes to take effect.

The changes are saved when the server is restarted.