Description of problem: I am in a 3 hosts/3 bricks cluster configuration, with a single distributed volume. Because I would like to change the FS of my bricks from xfs to ext4 (xfs is to slow with small files) I tried to remove a brick from my volume with the command : gluster volume remove-brick projets projet3:/glusterfs/projets-brick3/projet start After one day of pending status, the task was successful but, when I perform a "gluster volume status", the brick is still there : http://pastebin.com/7KZWJeet I tried to launch the command again but I have this result : > volume remove-brick start: failed: Incorrect brick projet3:/glusterfs/projets-brick3/projet for volume projets I tried to launch a removal of the server which was hosting this brick : > peer detach: failed: Brick(s) with the peer projet3 exist in cluster Other space thing, on all mount of the volume "projets", all the folder were duplicated, whith the same content. To solve this, I remove the "projets" folder on the brick3. I'm totaly lost. I just want to remove this brick and this host from the gluster configuration. Thank you for your help.
Thilam, Until you commit remove-brick transaction, brick doesn't go away, remobe-brick commit should follow remove-brick start once the status is completed. Please try the mentioned steps and close this bug accordingly. Thanks, Atin
Hi, It is now solved. Commit did not work. I had to stop all glusterfs daemon on all nodes + manually kill remaining processes. Then start again all daemons and commit the removal. Regards, T-