+++ This bug was initially created as a clone of Bug #1176824 +++ This issue is same as : Dist-geo-rep : volume won't be able to stop untill the geo-rep session is deleted only difference is now it is happening for replace-brick and may be for remove brick Description of problem: Even if the geo-rep session is stopped, replace brick is not allowed [root@fedora1 glusterfs]# gluster vol replace-brick master fedora1:/bricks/brick0/b0/ fedora1:/bricks/brick2/b2 commit force volume replace-brick: failed: geo-replication sessions are active for the volume master. Stop geo-replication sessions involved in this volume. Use 'volume geo-replication status' command for more info. Even after goe-rep stop, same error msg is getting thrown. How reproducible:Happens everytime Steps to Reproduce: 1.create and start master and slave volume 2.create and start geo-rep relationship between master and slave. 3.stop the geo-rep session master and slave 4.try to replace one of the master volume brick. Actual results:It won't allow to replace a brick in volume, until geo-rep session is deleted Expected results: It should allow replace brick in volume, if the geo-rep is stopped. It shouldn't expect to delete the geo-rep session only.
REVIEW: http://review.gluster.org/9402 (glusterd/geo-rep: Allow replace/remove brick if geo-rep is stopped.) posted (#1) for review on master by Kotresh HR (khiremat)
Documentation changes are expected during remove brick and replace brick if geo-rep is configured. REMOVE BRICK: 1. Start remove brick 2. Ensure all data in the brick to be removed is synced to slave. Use geo-rep config checkpoint if necessary. 3. Ensure remove brick status and checkpoint status is completed. 4. Stop geo-replication session 5. commit remove brick. REPLACE BRICK: Geo-rep needs to stopped before commit force. Rest of the steps remains intact. This is to make sure there are no stale gsyncd processes of the removed brick. Start geo-rep once replace brick is completed.
REVIEW: http://review.gluster.org/9402 (glusterd/geo-rep: Allow replace/remove brick if geo-rep is stopped.) posted (#2) for review on master by Kotresh HR (khiremat)
REVIEW: http://review.gluster.org/9402 (glusterd/geo-rep: Allow replace/remove brick if geo-rep is stopped.) posted (#3) for review on master by Kotresh HR (khiremat)
COMMIT: http://review.gluster.org/9402 committed in master by Krishnan Parthasarathi (kparthas) ------ commit 8618abaaf07a96c0384db9bd1e7dbbe663f4f24c Author: Kotresh HR <khiremat> Date: Tue Jan 6 20:26:39 2015 +0530 glusterd/geo-rep: Allow replace/remove brick if geo-rep is stopped. Replace brick: If geo-replication was configured on a volume, replace brick used to fail. This patch allows replace brick to go through if all geo-rep sessions corresponding to that volume is stopped. Remove brick: There was no check for geo-replication for remove brick. Enforce 'remove brick commit' to fail if geo-rep session corresponding to volume is running. Allow 'remove brick commit' only if all of the geo-rep sessions corresponding to that volume is stopped. Code is re-organized for better readability. Change-Id: I02282c2764d8b81e319489c977847e6e437511a4 BUG: 1179638 Signed-off-by: Kotresh HR <khiremat> Reviewed-on: http://review.gluster.org/9402 Tested-by: Gluster Build System <jenkins.com> Reviewed-by: Aravinda VK <avishwan> Reviewed-by: ajeet jha <ajha> Reviewed-by: Avra Sengupta <asengupt> Reviewed-by: Krishnan Parthasarathi <kparthas> Tested-by: Krishnan Parthasarathi <kparthas>
This bug is getting closed because a release has been made available that should address the reported issue. In case the problem is still not fixed with glusterfs-3.7.0, please open a new bug report. glusterfs-3.7.0 has been announced on the Gluster mailinglists [1], packages for several distributions should become available in the near future. Keep an eye on the Gluster Users mailinglist [2] and the update infrastructure for your distribution. [1] http://thread.gmane.org/gmane.comp.file-systems.gluster.devel/10939 [2] http://thread.gmane.org/gmane.comp.file-systems.gluster.user