Kubernetes monitoring: Prometheus and Grafana
It’s essential to monitor the resources and applications of your Kubernetes Cluster. Do you still have enough resources for another application? Do you have something crashing constantly? Do you have any pods using more resources than you planned? Why is that? These and other questions can be responded to with a good logging and monitoring system.
In this article, we will talk about 2 of the most famous tools in the Kubernetes world for this job: Prometheus and Grafana. You will understand what these tools do, how to install and configure them in your home lab, and never be blinded again about your cluster.
Let’s resolve the storage first.
We talked about different ways to handle storage in Kubernetes Cluster, including a method used by several clusters in production these days, using NFS to serve all our Kubernetes apps storage.
But I want to test something different here. The NFS works fine, but I want to use the nodes space directly and better visualize what’s happening rather than just looking at the NFS folders. Besides that, as I’m using proxmox to virtualize all my nodes, I’m already taking snapshots of them, and if the data is saved on the nodes, it will also be in the backup snapshots.
