My Alarms browser takes ages to load. 1

[enttoobad]

We use ITO and have templates to collect snmp traps directly to the ITO message browser. BUT - when I load up NNM (ovw), the alarm browser takes ages to load ("Reading state file&quot, and when it finally does load, it is full of trap messages - node ups/downs, network marginal, lots of "no format defined" messages etc.
Anyone have any tips on reducing the number of alarms raised or automatically deleting the alarms?
Also - if an snmp trap is picked up by ITO and the message is subsequently acknowledged, should this delete the alert from the NNM alarms browser also? (if not, can it be made to do so?)

Thanks in anticipation!

 

[unixrobot]

Acknowledging does not delete alarms. It will only mark the alarms as being acknowledged. There is an option to delete alarms that will delete alarms forever and it is enabled by default from JAVA CLIENT unless was specifically specified not to e preset there (see man ovalarmsrv). Usually that option is enabled by default (also there is a default number of alerts to keep in browser and delete oldest alerts - enabled by default and set to I think 1500) and I disabled it, so Operators cannot tell me later they did not see trap - I can always run report and see who is doing their job. Also, there is a way to disable delete option; there is a way to limit the number of alerts to keep in alarm browser.
No format defined means you don't have MIB loaded for that particular application/device that sends the trap. Trace where the message is coming from and load appropriate MIBs to NNM, do not blindly disable traps
There is a way to control what you see and what will not appear in alarm browser. You need to configure trapd.conf using NNM and select LOG ONLY for alerts that you don't want to see, I recommend loading appropriate MIBs before you disable valuable alarms!
Reading state file will take forever to load if you:
- do not let it finish to load
- have tonns of alerts (acknowledged/unacknowledged)
- state file is corrupt, NNM tries to repair it and you may loose some alerts due to that
- ovalarmsrv is not running/not responding
- you are out of memory/disk space
- need to apply latest patches
- etc