Volume/nfs interface expects a valid path. I have updated the doc with this PR[https://github.com/ceph/ceph/pull/38742].
@varsha, Can we restrict creation of a export when the non existent path is used ?
No, see my earlier comment. We expect valid path to be specified.
We can but we will have to mount cephfs. For now I would prefer not doing it.
Even with master (17.x) the `ceph nfs` command to create NFS export of non-existent CephFS paths succeeds. The mgr/nfs module does not check whether the CephFS path exists. As mentioned earlier in this BZ, the mgr/nfs module would need a CephFS client to do this. The other solution could be to rely on the NFS-Ganesha server to report back error to mgr/nfs module when reloading its exports of which one export cannot be created. For this, https://github.com/nfs-ganesha/nfs-ganesha/issues/757 would have to be solved. This would allow the mgr/nfs module to fail the `ceph nfs export create` of non-existent CephFS paths. @Yogesh Mane, can you check if you can remove/clean up the incorrect export using the `ceph nfs export rm` command?
Moving this out to 6.0 and resetting to NEW — we don't even have a viable solution to work towards at this time.
Upstream tracker - https://tracker.ceph.com/issues/58228 The doc changes should already be available downstream, Yogesh - https://github.com/ceph/ceph/pull/38742
Hi Venky, I am not seeing these changes in https://access.redhat.com/documentation/en-us/red_hat_ceph_storage/5/html-single/file_system_guide/index Could provide the link of downstream doc?
The needinfo request[s] on this closed bug have been removed as they have been unresolved for 120 days