ODF (Rook) does not implement a health checking mechanism for the NFS-Ganesha server. If the server remains running but is "stuck" and not responding to clients, ODF admins must restart the Ganesha server pod manually to attempt to recover from the failure. There is a workaround, as outlined above, but this is certainly not ideal for system admins. Currently Shachar has proposed upstream Rook changes to introduce a liveness probe for the Ganesha server pod. Based on initial feedback from NFS-Ganesha engineers, we believe this to be a good solution. However, we are still discussing the technical complexities and finalizing the implementation. In ODF/Rook, we have found that implementing liveness probes for Ceph daemons without detailed planning can result in behavior where the state of the ceph/daemon cluster can be negatively affected (e.g., RGW and MDS). https://github.com/rook/rook/pull/12845
Not a blocker for 4.14, let's target for 4.15. It also requires a new binary in the ceph image.
Rook's PR: https://github.com/rook/rook/pull/12845
Since the problem described in this bug report should be resolved in a recent advisory, it has been closed with a resolution of ERRATA. For information on the advisory (Important: Red Hat OpenShift Data Foundation 4.15.0 security, enhancement, & bug fix update), and where to find the updated files, follow the link below. If the solution does not work for you, open a new bug report. https://access.redhat.com/errata/RHSA-2024:1383