Kubernetes nodes crashing due to DiskPressure

Question

I don't know what DiskPressure means in kubernetes and how to avoid it. I'm using gcp to deploy the cluster. Mostly all of my new nodes crash because of high disk pressure.

Answer 1

From the ll find that DiskPressure raises when:

According to the official kubernetes docs:

Available disk space and inodes on either the node’s root filesytem or image filesystem has satisfied an eviction threshold

You should have the knowledge about how much space/inodes etc left in your system before deploying new nodes. you can do this by some basic resource monitoring(Resource Usage Monitoring). Now if you have prior knowledge of such resources you should adjust the --low-diskspace-threshold-mb, --image-gc-high-threshold and --image-gc-low-threshold parameters of you kubelet accordingly. This way there's always space for your day to day operations or you can consider adding more space for your nodes.