A common pain point in Kubernetes clusters is running out of disk space. When the node is provisioned, the root file system often shares space with container runtime data. The container runtime usually writes to /var , which can be located as a separate partition to prevent logs or container layers from filling the root file system.
Hard links respect inodes; editing a file in the clone will break the link (copy-on-write), but until you modify, zero extra disk is used. cri file system tools link
Use crictl stats to view real-time resource usage, including disk write metrics per container. A common pain point in Kubernetes clusters is