Skip to content

Quick Perfservmon Checks

Jump to bottom
Dimos Varounis edited this page Jan 31, 2017 · 6 revisions

This is a guide for quickly testing the perfservlet and the perfservmon plugin. For more information on how to set it in Nagios check README.md

First run the following check to collect the perfservlet data from WAS Cell WAS_Cell_name: Nagios_Path/libexec/perfservmon.py -C WAS_Cell_name retrieve -N Perfservlet_Host -P Perfservlet_Port

Optionally in case you use https(optionally with --ignorecert to ignore the server certificate) and/or basic auth credentials to connect to Perfservlet Application run the following check:

Nagios_Path/libexec/perfservmon.py -C WAS_Cell_name retrieve -N Perfservlet_Host -P Perfservlet_Port -H https -u userid -p passwd [--ignorecert]

You may want to perform the check above with a custom interval(e.g. 3 minutes) to collect fresh data from Cell.

Then add Monitoring Checks like the following:

  • Check the Total live http sessions of Servlets of the Server ServerName, get critical alert when 10000, warning when 5000. Nagios_Path/libexec/perfservmon.py -C WAS_Cell_name show -n NodeName -s ServerName -M LiveSessions -c 10000 -w 5000

  • Check the usage of WebContainer Thread Pool of Server ServerName, get critical alert when over 90%, warning when over 70%. Nagios_Path/libexec/perfservmon.py -C WAS_Cell_name show -n NodeName -s ServerName -M WebContainer -c 90 -w 70

  • Check the JVM Heap usage of Server ServerName, get critical alert when over 90%, warning when over 70%.Nagios_Path/libexec/perfservmon.py -C WAS_Cell_name show -n NodeName -s ServerName -M Heap -c 90 -w 70

  • Check the usage of all the Connection Pools available for Server ServerName, get critical alert when over 90%, warning when over 70%. Nagios_Path/libexec/perfservmon.py -C WAS_Cell_name show -n NodeName -s ServerName -M DBConnectionPoolPercentUsed -c 90 -w 70

Clone this wiki locally