In the last chapter we covered how good intentions can result in a well-meaning train wreck. I certainly don’t expect you to have solved all of those problems in your environment by the time you read this chapter, and that’s totally OK. Since you can now be mindful of the anti-patterns and work on solving them, you’re going to need new solutions for what to do in their place.
This chapter answers that question by presenting four design patterns that, if taken seriously and implemented, will lead you to monitoring nirvana. Let’s dig in.
Composable monitoring is the first pattern of modern monitoring design. The principle is simple: use multiple specialized tools and couple them loosely together, forming a monitoring “platform.” This pattern is directly in opposition to the monolithic tools many of you are familiar with, chief among them, Nagios. Composable monitoring can be thought of as the UNIX philosophy in action:
Write programs that do one thing and do it well. Write programs to work together.
Back in 2011, conversations about why monitoring was so bad grew around the #monitoringsucks hashtag on Twitter. This grew into #monitoringlove and the founding of the Monitorama conference in Boston. Many, many conversations were had about what could be done to improve things. One of the biggest points raised was that we needed new and better tools. More specialized tools. Composable monitoring as ...