Description of problem: ============================= "replace-brick commit force" should work even when rebalance process is running. commit with "force" option is to override all the checks. If we have a dead node and we want to replace the dead node with the new node , currently we have to wait for the rebalance process if any to complete and then do a replace-brick commit force and then trigger self-heal. Version-Release number of selected component (if applicable): ============================================================= glusterfs 3.4.0.57rhs built on Jan 13 2014 06:59:05 How reproducible: ================== Often Steps to Reproduce: ======================= 1.Create a 2 x 3 distribute-replicate volume. Create fuse mount . Create files /directories from mount point. 2.Add 3 more bricks to volume to change the volume type to 3 x 3. Start rebalance. 3.While rebalance is in progress, bring down 2 nodes. 4. Try to replace the offline bricks with new bricks using "replace-brick" with "commit force" option. Actual results: =============== root@domU-12-31-39-0A-99-B2 [Jan-21-2014-12:05:17] >gluster volume replace-brick exporter ip-10-114-195-155.ec2.internal:/rhs/bricks/exporter ip-10-83-5-197.ec2.internal:/rhs/bricks/exporter commit force ; volume replace-brick: failed: Volume name exporter rebalance is in progress. Please retry after completion Expected results: ================== replace-brick commit force should be successful even when rebalance process is running.
Cloning this to 3.1. To be fixed in future.