How to prevent GKE nodes from running out of space

[jabalsad]

Hello,

I am mounting a directory from the underlying node into pods as a cache of sorts. What, if anything, prevents the underlying nodes in a node pool from running out of disk space? Do the pods have to take care of not letting that directory grow unbounded? I was just curious if GKE has any mechanisms in place that does this automatically. Or maybe it is the responsibility of the cluster admin to deploy a daemonset that periodically cleans up the cache?

[yrhsk]

There isn't an automatic mechanism to prevent the underlying nodes from running out of disk space when you're mounting a directory as a cache. You can deploy a DaemonSet that periodically cleans up the cache to prevent it from growing unbounded. Or, as an alternative, you can set up monitoring and alerts to keep an eye on disk usage and take action as needed.