[Samba] CTDB and locking issues reloaded

[Giuseppe Lo Presti]

On 15/04/2020 16:57, Ralph Boehme via samba wrote:
> Am 4/15/20 um 10:59 AM schrieb Giuseppe Lo Presti via samba:
>> So the question is do you acknowledge the correct behaviour should be to
>> prevent the second client from getting the lock?
> 
> yes. My bet: your cluster filesystem device nummer differs across nodes. Cf:

Bingo!

Indeed inodes are (obviously) identical, but device numbers differ. For 
the record, the solution has been to include as suggested:

   vfs objects = fileid
   fileid:algorithm = fsname

in the global configuration. With this, Microsoft Office - the real use 
case - works correctly by detecting files opened by other users, no 
matter which gateway users hit.

So thanks a lot, case solved. And what would be your advice concerning 
the other locking settings? In particular:

   posix locking = no
   strict locking = no
   oplocks = no
   level2 oplocks = no
   kernel oplocks = no

Thanks again,
Cheers,
Giuseppe