Hello,

I wrote a small daemon which we use currently for making and feeding custom
metrics into ganglia in clusters built with OSCAR-Pro. The daemon is called
cmetricd and is quite configurable in what it does:
- it runs arbitrary programs at arbitrary time intervals, kills them if they
don't return within a timeout interval,
- extracts from the output of the programs one or more metrics, the metrics
can be specified as regexp matches in a config file. This is especially useful
for dealing with sensor metrics where typically one command lists all sensor
states.
- it can also extract from the output of the invoked programs metrics for
multiple hosts and publish them as spoofed metrics (with ganglia >= 3.0.4),
this can be useful when dealing with other daemons which are aware of the
cluster, e.g. torque.
- publishes the metrics through the local gmetric command,
- it avoids invoking the external commands more frequently than necessary.

cmetricd is available for download at http://code.google.com/p/cmetricd
Some sample configuration files are there, too.

You'll need perl-Events installed for running cmetricd, a source RPM for that
is downloadable from the cmetricd site, too.

Best regards,
Erich



-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/
_______________________________________________
Ganglia-developers mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/ganglia-developers

Reply via email to