Chapter 7. Metrics, Log Collection, and Monitoring
That's it. This chapter could well have ended here but I shall carry on for the benefit of those amongst us who would prefer things in more detail.
A great deal of the DevOps practice(s) evolve around the idea of being able to review and react to the state of your infrastructure at any given time – should you need to.
That is not to say, setup e-mail notifications for every time the date changes on your host, but a stream of sensible, usable amount of event data which would allow an operator to make a reasonably informed decision under stress and/or uncertainy.
If you have been paying attention in life so far, you would have noticed many a wise man talking about balance, the golden middle.
You should aim to configure your monitoring system in a way that you are notified of events of potential interest and in a timely manner. The notifications should arrive in a format that is hard to overlook, and should provide...
