Learning to CloudWatch

Lately I’m fooling around with AWS CloudWatch service, a monitoring solution native in the best cloud platform (my opinion!)

Just want to document my learnings and share with you, hopefully you will find it useful as well!

Look at how rude Microsoft is towards aws XD

Overall Look and Feel

The first time in, you may already want to get your hand dirty, and the easiest is to look at individual metrics (by first clicking on ‘Metrics’ on the left, followed by anything in the bottom right of the screen)

Many Widgets = Dashboard

If we are trying to monitoring something frequently, for example CPU utilisation of my EC2 instances, it can be easier to go with dashboards.

Dashboards are made up of widgets, of many different types. These are the main categories:

Here below, I created 2 line widgets to monitor one of my instance:

There’s also the Metrics Explorer widget, which look like this:

Metrics explorer widgets include graphs of multiple resources that have the same tag, or share the same resource property such as an instance type.

It is a pre-build set number of metric graphs, which can save you time to build them yourselves.

A use case can be when you want to monitor your fleet of EC2 instances by assigning tags by their environments, such as production or test. You can then use these tags to filter and aggregate the operational metrics such as CPUUtilization to understand the health and performance of the EC2 instances that are associated with each tag.

Example Dashboard

As I am trying to do rightsizing of instances, here is the dashboard and metrics that I look at:

To rightsize, AWS’ recommendation is to identify instances with a maximum CPU usage and memory usage of less than 40% over a four-week period. I will do another post just on rightsizing at a later time.

Hope this helps! Tusen takk for reading!

Typically seen messing around with AWS