vPostgres Monitoring

[jbo5112]

I'm trying to set up monitoring of the database(s) on a vPostgres machine.  The "postgresql.program Absolute path to control command" setting defaults to "/etc/init.d/postgres", which does not exist, nor does "/etc/init.d/vpostgres". I cannot find a similar program, and I cannot configure the monitoring without it.  It appears that the database is started by "/opt/vmware/etc/isv/subsequentboot".  I saw an old bug on this filed in a JIRA system (HHQ-5667), but it was closed as Invalid without any explanation that I can see.

Am I not supposed to monitor the database that holds my Hyperic data?  Should I just push the "/etc/init.d/vpostgres" file from another server?  Is there something I'm misunderstanding?

[admin · Accepted solution]

Run - ps -ef | grep postgres

From output you should see where postgres is running

[jbo5112]

Thanks for the help.  The control file was "/opt/vmware/vpostgres/9.1/bin/postgres".  It appears to be working, despite still giving me an error about it being unconfigured as some metric data is now starting to appear.  For people in the future with this problem, I found the path much easier to see with "ps aux | grep postgres", but I'm aware that may not always contain the path to the executable.  For some reason, the postgres command didn't always turn up either.

I had found it before, but it's a 64-bit ELF executable instead of a text file.  I assumed it definitely wasn't it, when it gave a huge error about my settings being invalid.  It turns out that Hyperic had come up with a very strange postgresql.host name that included things like log_rotation_age.