Created attachment 585851 [details]
engine.log

* Replace a brick in an existing volume.
* Immediately try to replace another brick in the same volume.
* 2nd replace fails because:
[glusterd-replace-brick.c:304:glusterd_op_stage_replace_brick] 0-: Replace brick is already started for volume
* From command line, the failure looks like this:
[root@localhost vdsm]# /usr/sbin/gluster --mode=script volume replace-brick vol3 10.35.97.158:/export node3:/bbbbb start
replace-brick failed to start
* Instead of failing nicely we are throwing an exception.