I was wondering if there was a best practice for Node Status Polling and a time frame for collecting statistics..
I currently have our systems set at the defaults of 120secs and 10mins..
Just wondering if anyone was doing anything different?
thanks
I change mine on a node-by-node basis.
One server, maybe I don't care if it goes down - doesn't affect any users short-term.
Another device, I get calls the instant it stops working. I would prefer to hear it from NPM than a user. So I set the polling rate higher.
Another instance would be watching for specific historical data/activity/whatever. Then you might want to record it at a faster rate.
Just my $0.03/worth.
I personally think its whatever your DB can handle and the kind of response time you need to have. I have polling set to every 30 seconds and data is collected every 60 seconds. This way we know almost immediately when something is not going right. We can also look back at graphs within the last 90 days and see stats down to the minute, which can help diagnose error in code for memory leaks and things of that nature. But due to this, we have a large database, which at one point was around 70GB.
Mi recommendation is, follow the solarwinds recommendations, if you change the numbers will have database issues in the future.
documentation.solarwinds.com/.../core-optimization-polling-engines.htm
