How to monitor Hyperic server? watchdog script? auto-restart?

[kallon]

Hi,

Does anyone monitor Hyperic server itself, to make sure it doesn't fall over?  I have had issues where the Hyperic HQ 4.6 server crashes, it gets killed by the Linux kernel oom-killer (out-of-memory).  And so I want to be able to watch for this occurance and hopefully restart the Hyperic Server automatically.

I was thinking of writing a watchdog script managed by Cron, that watches Hyperic server process and restarts it with hq-server.sh.  Does anyone do this or have a good way of achieving this?

Maybe hyperic has it's own internal restart mechanisim?

Any help would be great

Thanks

[staceyeschneide]

I think at one point there was an effort around this - although it may have been around the agent only. A watchdog script could be useful, and I have seen another user post how to automatically pair starting Hyperic up at startup. This may give you a headstart: http://vashal.hubpages.com/hub/How-To-Start-The-Hyperic-HQ-Server-Service-At-Startup-With-Linux

If you do figure this out, please update us so we can share from your experiences!

[kallon]

Hi,

i created this solution attached.  place the cron job into /etc/cron.d/.  And make sure the init.d script is part of the boot up procedures for a Hyperic server that starts upone server boot up.