Use this Knowledge Script to monitor the amount of time a node spends processing a request. In a multi-node deployment, requests typically enter the system and create processes that fan out across several nodes within the distributed system. Use this script to gather and correlate data collected at each node of the system, in order to provide a breakdown of how much time is spent in different components of the request-processing pipeline.
Before using this Knowledge Script, you must run the SetRMFilters script to instruct WebSphere Application Server to collect metrics for requests that match a given set of patterns, such as client IP address, URL, or EJB method names. Once request metrics are enabled, WebSphere Application Server will write an entry to the trace log each time a matching HTTP request is received or a remote EJB method call is made. This script can then be used to gather and correlate these trace log entries.
In order for this script to gather request metrics, it must have a list of the hosts from which to collect data. Normally, you should specify all hosts in the WebSphere Application Server domain, but you can restrict the data to a particular subset of nodes by specifying only those computers. You do not need to specify individual application servers on the selected computers—this Knowledge Script analyzes the log files for every application server discovered on each selected computer.
This script measures the request-processing time for only those requests that entered the system at the node on which this script is run. Any request that enters the WebSphere Application Server network at an upstream node is ignored. Hence, run this script only on the node containing the entry point for request processing. This script can be set to raise an event if the average request-processing response time exceeds a specified threshold. The detail message associated with the event shows the breakdown of how much time was spent in each component. You can also specify that a data stream containing the average response time be written to the database.
WebSphere Application Server
The default interval for this script is Every 15 Minutes.
Set the following parameters as needed:
|
Description |
How to Set It |
|---|---|
|
Event? (y/n) |
Set to y to raise events. The default is y. |
|
Collect data? (y/n) |
Set to y to collect data for reports and graphs. The default is n. |
|
Semicolon-separated list of hosts from which to collect data |
Specify the hosts from which request metrics data should be gathered. Normally, you should specify all hosts in the WebSphere Application Server domain. |
|
Request processing time threshold |
Specify a threshold value greater than or equal to -1. Use -1 to ignore this threshold. If the average time to process a request (in seconds) exceeds this threshold, an event is raised. The default is -1. |
|
Event severity when request processing time exceeds threshold |
Set the event severity level, from 1 to 40, to indicate the importance of event. The default is 25. |