If you’ve ever had performance issues within ServiceNow, chances are you’ve gone to look at the performance dashboards to see if anything looks “funny” there. If not, you then check the stats.do page.
Recently I’ve been working on automated monitoring, spotting a situation arising before the customer notices the system is going into slowdown.
While doing this I’ve discovered a new table with a whole load of metrics to read from.
This table is cmdb_metric table. It has loads of extensions to it (25 to be precise) which monitor stores results for all sorts of stuff, semaphore queues, memory usage, cpu utilization – all of this separated by node too.
All these metrics are extremely useful for either automated monitoring or if looking into it manually, you can go back to a point in time and see how the performance was during a period of slowness.
I haven’t checked how long the metrics are kept for, but I’d expect them to be held for no longer than 30 days in a similar fashion to most of ServiceNow logging.