Hyperic HQ Community Edition

[ccliver]

Is there a limit to the number of hosts the community edition of Hyperic HQ can monitor? I'm running the community edition monitoring 100 hosts, and just about every Monday morning I have to restart the Hyperic service because the web page is down. I get a "page cannot be displayed" message when trying to access it. The server also doesn't appear to be writing anything (except normal scheduled tasks) to the server.log or server.out.

The process also doesn't appear to be doing anything when I try to access the web page:

root@hyperic-colo:/home/hyperic/server-4.1.2/logs# ps aux |grep ^hyperic
hyperic 32245 5.2 23.7 1805536 919120 ? Sl Sep08 449:48 /home/hyperic/server-4.1.2/jre/bin/java -server -XX:MaxPermSize=192m -Xmx512m -Xms512m -Dprogram.name=hq-server -Dserver.home=/home/hyperic/server-4.1.2 -Dengine.home=/home/hyperic/server-4.1.2/hq-engine -Djava.awt.headless=true -Djava.net.preferIPv4Stack=true -classpath /home/hyperic/server-4.1.2/hq-engine/bin/run.jar org.jboss.Main -b 0.0.0.0

root@hyperic-colo:/home/hyperic/server-4.1.2/logs# strace -p 32245
Process 32245 attached - interrupt to quit
futex(0x402228d4, 0x80 /* FUTEX_??? */, 1

[bclark_hyperic]

You might be running out of memory (although I would think that you would see that in the logs somewhere). Try looking through the Hyperic Performance Tuning page to see if any of their suggestions make sense in your environment. 100 agents is slightly above the default tuning parameters that Hyperic has setup.

http://support.hyperic.com/display/DOC/Configuring+HQ+for+Large+Environments+and+Improved+Performance

Also, if you are using PostgreSQL, see the section in this article called "Tips for large HQ Deployments".

http://support.hyperic.com/display/DOC/Set+Up+PostgreSQL

Good luck.

[ccliver]

Thanks, I will try those out.