Multiple CPU alerts at the same time

I have an alert setup in Orion for high CPU load setup across multiple servers. Most of these machines have one processor. However, I always get four alerts all at the same exact time. Four for the high CPU utilization, and four for the return to normal. What I just discovered however is that each one of the four messages references a completely different NetObjectID that appears to be in now way related to the actual ID of the ID that I am monitoring. When I attempt to view that NetObjectID, I get an error telling me that ID does not exist. Anyone have any ideas on where I can go from here? I'd like to clear this up to reduce the number of alert emails I receive.

Find more posts tagged with

Alert

network_monitoring

netobject

orion

CPU

network_management

Accepted answers

bluefunelemental

Looks like it but to go fishing lets open database manager - write out the first SQL statement copy paste the 2nd half and execute.

Best way to test alert logic for now except copying whole trigger query right from Alertdefinitions table.

you might not have any high CPU but lower the threshold till it snags a river monster or boot.

All comments

adatole

Can you post your alert logic or a screengrab of the alert trigger tab?

bluefunelemental

Can you screenshot your alarm and action

even better use the drop down to change to custom SQL, screenshot, then cancel so it doesn't stay that way

jasongreb

I don't see where I can view the custom SQL, but that would certainly be nice!

bluefunelemental

Under trigger query- near top of page- change drop down from node to custom SQL and screenshot- then cancel.

jasongreb

I just noticed that the query is on Volume. Could this have anything to do with it? And thanks for the tip on the SQL!

bluefunelemental

Looks like it but to go fishing lets open database manager - write out the first SQL statement copy paste the 2nd half and execute.

Best way to test alert logic for now except copying whole trigger query right from Alertdefinitions table.

you might not have any high CPU but lower the threshold till it snags a river monster or boot.

jasongreb

I believe that was it! It was pulling in the volume names in the query. I've adjusted it to Node as the trigger query. Thanks for all the SQL tips. Once I discovered you could do custom SQL reports in Orion, I've been going crazy with the SQL.