I have also similar experiences with Zenoss 2.2.3 process monitoring.
OS Process monitoring is behaving really non-deterministically. Either I can't 
figure out how to use it right, or there is a bug in it.

Working example: I define a process for crond (regex=^crond$, parameters 
ignored) and activate it on a device. Everything works as expected. When the 
process goes down, an event is generated. 

Not-working example: I define a process for syslogd (regex= ^syslogd$, 
parameters ignored) and activate it on a device. 
Things that go wrong: 1) when the process is activated, it shows up green, even 
when the syslog process is down, 2) when I remodel the device (while syslog 
up), the activated process disappears even when processes locked.

The ouput of ps and snmpwalk grepping either syslogd or crond is very similar 
and both processes should work the same way. Yet they don't.


The second issue that also bothers me a little is this: when I define a regex 
to cover several processes of the same service ( for example the 
(unlinkd),(squid) and squid processes of the squid daemon ) I get THREE 
different processes in the view. For httpd daemon all processes discovered by 
the regex (which happen to look exactly the same) are rolled to one.
In my opinion the summing and grouping and counting of statistics should be 
done by the regex (showing the process label not the process name itself), not 
by the exact process names returned.




-------------------- m2f --------------------

Read this topic online here:
http://community.zenoss.com/forums/viewtopic.php?p=25037#25037

-------------------- m2f --------------------



_______________________________________________
zenoss-users mailing list
[email protected]
http://lists.zenoss.org/mailman/listinfo/zenoss-users

Reply via email to