Ceph » CephFS

We cannot test this with accuracy.

Because:

`ceph fs volume ls` would list the already existing volumes and the default volume(s). So we don't know what is already there and what to compare with. There is no other way to know the entire list of volumes, so that we could compare the results.

To test `ceph fs volume create`, there is no getpath function/command to verify if the new volume is actually created or not. We could catch the exception here https://github.com/ceph/ceph/blob/master/qa/tasks/cephfs/test_volumes.py#L53, but it seems the subvolume tests that uses self.volname indirectly confirms that a volume could be created.

So closing this ticket.

Jos Collin wrote:

We cannot test this with accuracy.

Because:

`ceph fs volume ls` would list the already existing volumes and the default volume(s). So we don't know what is already there and what to compare with. There is no other way to know the entire list of volumes, so that we could compare the results.

Write a method to delete all volumes and then do tests for create. Likewise for ls.

To test `ceph fs volume create`, there is no getpath function/command to verify if the new volume is actually created or not.

Write one! Then do a client mount.

We could catch the exception here https://github.com/ceph/ceph/blob/master/qa/tasks/cephfs/test_volumes.py#L53, but it seems the subvolume tests that uses self.volname indirectly confirms that a volume could be created.

The volume was not created using `fs volume`. It's just the usual file system creation as part of QA tests. (Note: `fs volume` and `fs new` are basically equivalent from the perspective of the volumes plugin.)

So closing this ticket.

Reopening!