Challenge 11 - Operations and Monitoring
Introduction
Running a cluster without knowing what is going on inside of it is a show-stopper for any serious production deployment. It is imperative that we are familiar with operationalizing our Kubernetes clusters and having a full view into day to day running and error state alerts.
Description
In this challenge you will learn how to view application logs and trouble-shoot errors. View performance metrics and identity bottlenecks.
- Find the logs for your application’s containers, using:
kubectl- Using the Kubernetes Dashboard
- Notice how you can check the logs of any of your pods individually.
- Start a bash shell into one of the containers running on a pod and check the list of running processes
- Find out if your pods had any errors.
- Figure out how to get details on a running pod to see reasons for failures.
- Azure Monitor:
- Enable “Azure Monitor for Containers” on the AKS cluster
- Show a screenshot of CPU and memory utilization of all nodes
- Show a screenshot displaying logs from the frontend and backend containers
- Kibana:
- Install Fluentd and Kibana resources on the Kubernetes cluster to use an external ElasticSearch cluster
- Create a Kibana dashboard that shows a summary of logs from the front-end app only
- Create a Kibana dashboard that shows a summary of logs from the back-end app only
- Create a Kibana dashboard that gives a count of all log events from the kubernetes cluster for the last 30 minutes only.
Success Criteria
- Show logs for the containers running in your cluster.
- Demonstrate that you can login to a running container and issue bash commands.
- Show Azure Monitor working.
- Show Kibana dashboards working.
Learning Resources
- Azure Monitoring for Containers
- ELK stack:
- Fluentd: