Description of problem: ======================= When the below steps were performed remove-brick status output is showing status of layout fixing instead of remove-brick output. Version-Release number of selected component (if applicable): 3.8.4-2.el7rhgs.x86_64 How reproducible: ================= 1/1 Steps to Reproduce: =================== 1. Create a distributed replica volume and start it. 2. FUSE mount the volume. 3. Fix the layout by issuing "gluster volume rebalance <vol-name> fix-layout start". 4. After completion of step-3, start remove-brick and check the status of the remove-brick, it will show the correct output. [root@dhcp42-7 ~]# gluster volume remove-brick distrep replica 2 10.70.42.7:/bricks/brick3/b3 10.70.41.211:/bricks/brick3/b3 status Node Rebalanced-files size scanned failures skipped status run time in h:m:s --------- ----------- ----------- ----------- ----------- ----------- ------------ -------------- localhst 50 113.7KB 9969 0 0 completed 0:2:7 10.70.41.211 0 0Bytes 0 0 0 completed 0:1:17 5. After completion of remove-brick, commit it. 6. Now again Fix the layout using command gluster volume rebalance <vol-name> fix-layout start". 7. After completion of step-6, start remove-brick and check the status. It is showing status of fix-layout instead of remove-brick output. [root@dhcp42-7 ~]# gluster v remove-brick distrep replica 2 10.70.43.141:/bricks/brick2/b2 10.70.43.156:/bricks/brick2/b2 status Node Rebalanced-files size scanned failures skipped status run time in h:m:s --------- ----------- ----------- ----------- ----------- ----------- ------------ -------------- 10.70.43.141 fix-layout in progress 0:0:7 10.70.43.156 fix-layout in progress 0:0:7 ... .. . [root@dhcp42-7 ~]# gluster v remove-brick distrep replica 2 10.70.43.141:/bricks/brick2/b2 10.70.43.156:/bricks/brick2/b2 status Node Rebalanced-files size scanned failures skipped status run time in h:m:s --------- ----------- ----------- ----------- ----------- ----------- ------------ -------------- 10.70.43.141 fix-layout completed 0:5:12 10.70.43.156 fix-layout completed 0:1:41 Actual results: =============== remove-brick status output is showing status of layout fixing. Expected results: ================= remove-brick status output should original remove-brick status output.
Steps to reproduce this on 3.2.0: 1. On a 2 node cluster, create a volume with 1 brick on each node. 2. From node1, run gluster v rebalance <volname> fix-layout start 3. Once the fix-layout has completed, from node1, run gluster v remove-brick <volname> <brick on node2> start 4. On node1, run gluster v remove-brick <volname> <brick on node2> status This will print the fix-layout output. Running the command on node2 prints the output correctly. gluster v remove-brick <volname> <brick on node2> status
Upstream patch at: http://review.gluster.org/15749
upstream mainline : http://review.gluster.org/15749 upstream 3.9 : http://review.gluster.org/#/c/15870/
Verified this BZ on glusterfs version 3.8.4-8.el7rhgs.x86_64. Below are the steps, 1) Created a distributed replica volume and started it. 2) FUSE mount the volume. 3) From node1, fix the layout by issuing "gluster volume rebalance <vol-name> fix-layout start". 4) After fixing the layout, from node 1 removed the peer subvol bricks. gluster v remove-brick <volname> <brick on node3> <brick on node4>start 5) From node 1, checked the remove-brick status and it is showing the correct remove-brick status output. Moving this BZ to Verified.
Since the problem described in this bug report should be resolved in a recent advisory, it has been closed with a resolution of ERRATA. For information on the advisory, and where to find the updated files, follow the link below. If the solution does not work for you, open a new bug report. https://rhn.redhat.com/errata/RHSA-2017-0486.html