We will now explore some of the observability that Container Insights gives us. This is a very quick look at CloudWatch….it can do a lot!
In the AWS Console make sure you are in CloudWatch (which you will probably be from the last section).
On the left navigation pane click Resources under Container Insights.
Click any of the items from the list of Resources and you will see a performance dashboard open giving you an insight into the performance:
There is a specific resource for Prometheus AppMesh from the left most dropdown box on the Performance Monitoring pane. This shows a dashboard specifically for App Mesh. We will return to this later.
On the left navigation pane click Metrics. Then click ContainerInsights under Custom Namespaces. You will then be shown a list of groups at different granularities. Click ClusterName, Namespace.
This will show a list of the metrics that you can add to the graph. Check the box next to these 2 metrics:
The graph will then be updated to display the metrics.
Now click on Log Groups under Logs. You will see a number of log groups created by Container Insights for our cluster, these are prefixed with /aws/containerinsights/gitopsworkshop.
Click /aws/containerinsights/gitopsworkshop/application. You will then see a list of the log streams for your pods. Select any log stream from the list and you will be taken to the logs:
This has been a brief look at what Container Insights and CloudWatch gives us for observability of our EKS cluster. CloudWatch enables you to do a lot more including creating alarms (a.k.a alerts), dashboards, end-to-end tracing and a lot more!