Hi Jamie,
I'm not sure if your book mentions this (alerts) and if it's even related to the topic or if it's considered a separate one. However, I though I will ask anyway - suppose i am gathering system metrics, for example CPU utilisation and memory utilisation (either via pull or push method), how to best implement the alerting on top to avoid false positives? I find it hard to configure as every server is different and for some 95% utilisation is actually good. Any insight into this subject would be great, thanks.
Best regards,