dockerdata-nfs full disk causes ONAP crash #so #dublin #mariadb #nfs #nfs-settings


Ferrero Marco Angelo
 

Hi everyone,
we experienced ONAP crash due to shared disk dockerdata-nfs full. That condition affected also some k8s-nodes that became not responsive because the kernel thread related to nfs client (i.e. [<ip_of_nfs_server>-man]) required 100% of CPU time.

Is there a nfs configuration to avoid to get nfs client unresponsiveness?

Current used nfs flags are: rw, sync, no_root_squash, no_subtree_check (file /etc/exportfs in nfs server)


Thank you
Marco Ferrero

PS:  Some more details can be found in: https://unix.stackexchange.com/questions/63569/what-are-processes-123-45-78-901-ma-on-linux-where-the-number-is-an-nfs-ser