Comment 4 for bug 1834875

Hey Tobi, could you clarify what the failure here actually is? Looking at the journal (specifically the kernel EXT4 entries), it looks like some resizing _does_ happen. (Obviously we don't want a traceback in the log so we can chase that down, but I'd like to understand the concrete effect that this is having on users in order to prioritise it. :)

(Also, if you could get us access to a failing instance, that would be great! I don't have cloud-test-framework checked out any longer, else I'd try to do it.)