In this long chapter, you learned how to set up monitoring of your application running in Windows containers on Kubernetes. First, we took a look at available monitoring solutions and determined which fit our use case with Windows nodes—the best choice currently is using a dedicated instance of Prometheus together with Grafana. Next, you learned how to make Windows nodes observable in terms of hardware, operating system, and container runtime using WMI Exporter and the experimental Docker Engine metrics service. We have shown how you can install and configure these agents on an AKS Engine cluster using extensions.
The next step was the Deployment of Prometheus and Grafana using Helm charts. You needed to ensure that Prometheus scraping jobs are capable of discovering the new metrics endpoints on Windows nodes. After that, we focused on monitoring inside the container...