Alerts are very versatile. You can send email (to any number of of addresses), run a script (that receives a set of related values in environment variables) and execute an action like restarting a stopped service. Alerts can be associated to anything that has metrics. So server or service availability (metadata server, object spawner, web app server), but also CPU load, memory, disk space etc. Both SAS and non-SAS components of your environment are being monitored.
Slight warning for the less-than-intuitive interface of EVM. Let's say it's an acquired taste. But is is very effective and does the job well. Also it improves with every version bump. And mind you it isn't easy to do the same as a DIY project.
The chapter on alerts and events is here.
Regards,
- Jan.
... View more