Description of problem: When a file is marked as corrupted by bitrot recovering the file from the replicated volume, does not work. Version-Release number of selected component (if applicable): glusterfs-3.7.5-9.el7rhgs.x86_64 How reproducible: Incosistent Steps to Reproduce: 1. Create tiered volume with both hot and cold bricks as dist-rep and enable bitrot on the volume. 2. Fuse mount the volume and create some data into it. 3. Now edit one of the file from backend so that scrubber will mark the file as bad file in one of the subvolume. 3. Do I/O from the mount point so that all the I/O will go the other file. 4. Now try to recover the file by deleting the file and its gfid from backend and launching self heal command. Actual results: only meta data heal completes and data heal does not. Heal info always shows that there are some entries to heal. Expected results: Heal should complete sucessfully and file should be recovered. Additional info:
sos reports can be found in the link http://rhsqe-repo.lab.eng.blr.redhat.com/sosreports/1289047/