Cluster Administration


You likely know what a cluster is by now—a group of pods that will work together and share a network. Within Kubernetes, you will run these clusters. But there are various things you will need to know about utilizing clusters in Kubernetes. Much of this should be automated but isn’t naturally until you set it up.

Planning a Cluster

Solutions for establishing, creating and configuring Kubernetes clusters are called distros. The creator chooses a distros of choice, but it is important to note that not all of them are actively maintained. In order to choose an active solution, pick one that has been tested in a recent Kubernetes version.

The distros best suited for your application’s needs will be determined by certain details that apply to your app. You can use the distros to build clusters on-premise and other distros may build based in-cloud. Some will want to use a Kubernetes host, while others will allow you to use your own. Since hybrid clusters aren’t supported by k8s, multiple clusters might need to be established if you go that route.

Managing a Cluster

Pods are mortal and will be replaced as their life cycle ends or if resources become tight. You will upgrade cluster masters and workers (nodes) with maintenance and upgrades. You will be able to set up the resource quota and establish rules for the creation or deletion of these clusters.

Securing a Cluster

You will need certificates generated for different tools. Within the container environment, Kubernetes uses controls that restrict the permissions for users to fit their service accounts. Utilizing authentication, users can ensure the app is secure and authorization establishes calls to the HTTP. There are a number of plugins that snag requests to Kubernetes AP and double-check the authentication and authorization of traffic. A developer can use audit logs to ensure the data is being adapted to for a better application performance.