Description of problem: Once a brick becomes completely (100%) full, the gluster filesystem becomes inconsistent as xattr operations fail and recovery is very difficult. A rebalance cannot be used to migrate data off the full brick as xattr operations will fail. DHT uses a min-free-disk option to try to keep some reserve space on all bricks. However, this approach is not fool proof and should ideally be handled by the posix xlator. Version-Release number of selected component (if applicable): How reproducible: Steps to Reproduce: 1. 2. 3. Actual results: Expected results: Additional info:
Patch is posted on upstream https://review.gluster.org/#/c/17780/
Verified this RFE on glusterfs version: 3.12.2-5.el7rhgs.x86_64 1) Able to set/reset storage.reserve volume set option. 2) Reserved the space to 50% and created files using dd created files. Once the backend bricks reached the reserved space, file creation is failing with ENOSPC errors. I have filed below BZs while testing this RFE, https://bugzilla.redhat.com/show_bug.cgi?id=1554291 https://bugzilla.redhat.com/show_bug.cgi?id=1550991 https://bugzilla.redhat.com/show_bug.cgi?id=1550982 https://bugzilla.redhat.com/show_bug.cgi?id=1554298 Moving this BZ to Verified considering the basic functional is working and have bugs filed for the seen issues. Will file new BZs on this feature if seen during our regression testing.
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://access.redhat.com/errata/RHSA-2018:2607
The needinfo request[s] on this closed bug have been removed as they have been unresolved for 1000 days