DM monitors sudden disappearance

[andykoh]

Hi,

We are currently monitoring 35 NT servers and 6 AIX servers. Each server has about 40 monitors deployed. Strange thing we noted is that sometimes certain monitor will stop sending back data to TEC for 2 to 5 hours and later start again.

Monitoring interval is set to 15 minutes for each monitor and threshold is set to "NORMAL". Couldn't find a reason why this happened. there were no maintenance or scheduled maintenance tasks done to the servers at the time the "disappearance" occured. Seems to happen at randomly at different time slots. Anyone encountered something like this before?

Anyway to check whether if a monitor is running other then observe from TEC?

 

[SimonERM]

There is often a log file created on the machine itself containing the last X number of values. Can't remember the exact location of it, but you'll find an info file in the same directory.
Its something like $DBDIR/.sntcfg/<large number> but do a bit of searching around as this may not be exactly right.
So, what to do with this file. See if it is getting the values inserted after each time the monitor fires.
Also, you can run the wlseng (-z for endpoints). This gives a lot of info back, but part of the returned stuff shows when it last fired and the value it received. If you get E.EXEC then this means there was something wrong with the monitor firing, and this would explain why nothing was seen at the TEC>
Hope this helps.
Simon.