Skip to content

REANA pods report healthy but not operational due to stale NFS #924

@volodymyrss

Description

@volodymyrss

In ETAP we are running VRE including REANA.
We are using NFS storage as one of the advised options.

At some point we noticed that it's not possible to run REANA jobs. All REANA pods were live and ready.

Looking into logs showed that many of the REANA container logs has "Stale file handle" entries.
The triggering problem was that NFS provisioner restarted. But we'd expect that REANA pods identified their problem and restarted.

What do you think? Should there be extra health checks?

Metadata

Metadata

Assignees

No one assigned

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions