Description of problem: If the geo-replication session is active, then the volume stop with throw an warning/error message saying the geo-rep session is active and use force if you still want to stop the volume. This message should not be shown if the geo-rep session is stopped by geo-rep stop cli. Since the session is not active, the volume stop should also work just fine. Version-Release number of selected component (if applicable): glusterfs-3.4.0.14rhs-1.el6rhs.x86_64 How reproducible: Many times in my setup. Steps to Reproduce: 1. Create and start a geo-rep session between master and slave. 2. Now stop the session using geo-rep stop CLI. 3. Try to stop volume using the command volume stop. Actual results: [root@typhoon ~]# gluster v stop master --mode=script volume stop: master: failed: geo-replication sessions are active for the volume 'master'. Use 'volume geo-replication status' command for more info. Use 'force' option to ignore and stop the volume. But the geo-rep session is not really active. It's been stopped. [root@typhoon ~]# gluster v geo master status NODE MASTER SLAVE HEALTH UPTIME ------------------------------------------------------------------------------------ typhoon.blr.redhat.com master ssh://falcon::slave Stopped N/A mustang.blr.redhat.com master ssh://falcon::slave Stopped N/A harrier.blr.redhat.com master ssh://falcon::slave Stopped N/A spitfire.blr.redhat.com master ssh://falcon::slave Stopped N/A Expected results: volume stop should succeed when the geo-rep session is stopped. Additional info: logs from the glusterd ~ [2013-08-05 08:41:21.865044] W [glusterd-volume-ops.c:1037:glusterd_op_stage_stop_volume] 0-management: geo-replication sessions activefor the volume master [2013-08-05 08:41:21.865150] E [glusterd-syncop.c:872:gd_stage_op_phase] 0-management: Staging of operation 'Volume Stop' failed on localhost : geo-replication sessions are active for the volume 'master'. Use 'volume geo-replication status' command for more info. Use 'force' option to ignore and stop the volume. [2013-08-05 08:41:34.644440] I [glusterd-geo-rep.c:274:__glusterd_handle_gsync_set] 0-management: slave not found, whilehandling geo-replication options [2013-08-05 08:41:59.465206] W [glusterd-volume-ops.c:1037:glusterd_op_stage_stop_volume] 0-management: geo-replication sessions activefor the volume master [2013-08-05 08:41:59.465443] E [glusterd-syncop.c:872:gd_stage_op_phase] 0-management: Staging of operation 'Volume Stop' failed on localhost : geo-replication sessions are active for the volume 'master'. Use 'volume geo-replication status' command for more info. Use 'force' option to ignore and stop the volume. [2013-08-05 08:56:00.419505] I [glusterd-geo-rep.c:274:__glusterd_handle_gsync_set] 0-management: slave not found, whilehandling geo-replication options I also ran 'ps -aef | grep gluster' in all the master nodes to check if there is any geo-rep related processes running. But there weren't any.
This is fixed as part of Bug 1002822 and is ON_QA http://review.gluster.org/#/c/6821/ Hence marking as duplicate. *** This bug has been marked as a duplicate of bug 1002822 ***