We introduced ElasticKube last month to ease the management of containers on Kubernetes, and the feedback from the community has been phenomenal. ElasticKube Diagnostics adds a simple and intuitive module for troubleshooting deployments. It displays the various components and requirements of your containerized applications and their status, giving you the ability to check the health of the service and easily identify what needs to be fixed. This latest update is automatically available within the admin console and accessed at /diagnostics.
Diagnostics are extremely valuable during the installation of ElasticKube. If, for instance, ElasticKube is not properly deployed, you will find information on what might be missing or what is configured incorrectly. For common errors, users will be redirected there automatically to begin troubleshooting in real time.
The complete list of services that are constantly being checked includes:
- Kubernetes Connection: Check if ElasticKube has access to the Kubernetes API. It’s a common problem that prevents ElasticKube from working
- Mongo Connection: Check if a MongoDB Replica Controller is running
- Websocket Service and Chart Service: Check if the ElasticKube Server Replica Controller is running. If any of these services are not identified as “Ok”, ElasticKube will fail to start or may not be deployed
- Internet Connection: Internet connection is needed and most importantly, the ability to fetch the Charts from your target repository
- Heapster Connection: This is a key component for pulling data for the various metrics
- Kubernetes DNS: Checks if the required DNS is up and running
When deploying ElasticKube to manage your containers on top of a Kubernetes cluster, check out the diagnostics panel and make sure you’re aware of how your services are performing. This will help reduce downtime and provide guidance on which components need immediate attention or will be at risk in the near future.