we have a WebLogic setup that has been giving us some problems.
We have a monitoring tool called HP BAC that helps us visualize the state of the servers, and we tie this tool with the Java Thread Dump utility such that when there is a low thread count situation, a thread dump will be captured.
There are situations where the server goes into a low thread count state, and we are uncertain why because all our efforts to find out - through thread dumps especially seems futile.
Given the "state of art", our thread dump capture is always triggered late / after, and always never capture thread dump useful for our investigation.
I would like to check how the rest of you guys have put in place such monitoring efforts?
I'd suggest WLST
Write a WLST script which hooks into the running server and gets the active thread count at regular intervals. (say 30 seconds)
If your active thread count goes below your pre-configured threshold, fire the kill - 3
and get the thread dump.
Some examples
http://middlewaremagic.com/weblogic/?p=5423
http://wlstbyexamples.blogspot.com/2009/04/wlst-thread-count.html
If you love us? You can donate to us via Paypal or buy me a coffee so we can maintain and grow! Thank you!
Donate Us With