Description of problem: ---------------------- I had lot of files on my ec volume on 3.12.2-6 build. I did an offline server upgrade to 3.12.2-7 . Also upgraded the clients to the same build(did a fresh mount post upgrade) I then tried to delete the old files, and when we access the files I am seeing below messages logged for every entry in fuse mount log [2018-04-13 12:11:00.215891] I [MSGID: 109005] [dht-selfheal.c:2454:dht_selfheal_directory] 0-zen-dht: Directory selfheal failed: Unable to form layout for directory /rhs-client19.lab.eng.blr.redhat.com/linux-4.15.13/arch/arm/mach-orion5x [2018-04-13 12:11:00.317994] I [MSGID: 109005] [dht-selfheal.c:2454:dht_selfheal_directory] 0-zen-dht: Directory selfheal failed: Unable to form layout for directory /rhs-client19.lab.eng.blr.redhat.com/linux-4.15.13/arch/arm/mach-prima2 [2018-04-13 12:11:00.360088] I [MSGID: 109005] [dht-selfheal.c:2454:dht_selfheal_directory] 0-zen-dht: Directory selfheal failed: Unable to form layout for directory /rhs-client19.lab.eng.blr.redhat.com/linux-4.15.13/arch/arm/mach-qcom Volume Name: zen Type: Distributed-Disperse Volume ID: a6470510-3f32-4f34-8004-521d9670bec9 Status: Started Snapshot Count: 0 Number of Bricks: 2 x (4 + 2) = 12 Transport-type: tcp Bricks: Brick1: dhcp35-205.lab.eng.blr.redhat.com:/gluster/brick1/zen Brick2: dhcp35-169.lab.eng.blr.redhat.com:/gluster/brick1/zen Brick3: dhcp35-145.lab.eng.blr.redhat.com:/gluster/brick1/zen Brick4: dhcp35-177.lab.eng.blr.redhat.com:/gluster/brick1/zen Brick5: dhcp35-29.lab.eng.blr.redhat.com:/gluster/brick1/zen Brick6: dhcp35-14.lab.eng.blr.redhat.com:/gluster/brick1/zen Brick7: dhcp35-205.lab.eng.blr.redhat.com:/gluster/brick2/zen Brick8: dhcp35-169.lab.eng.blr.redhat.com:/gluster/brick2/zen Brick9: dhcp35-145.lab.eng.blr.redhat.com:/gluster/brick2/zen Brick10: dhcp35-177.lab.eng.blr.redhat.com:/gluster/brick2/zen Brick11: dhcp35-29.lab.eng.blr.redhat.com:/gluster/brick2/zen Brick12: dhcp35-14.lab.eng.blr.redhat.com:/gluster/brick2/zen Options Reconfigured: nfs.disable: on [root@dhcp35-205 ~]# Version-Release number of selected component (if applicable): ------------ 3.12.2-7 Steps to Reproduce: 1.have an ec volume 2x(4+2) in 3.12.2-6 and have data on the volume 2.offline upgrade the volume to 3.12.2-7 3.remount the clients after upgrading them 4. do a rm -rf or even an ls and you will see above messages Additional Info: ================= post successful rm -rf of old files, I created new dirs and deleted them, but didnt see these messages.
sosreports http://rhsqe-repo.lab.eng.blr.redhat.com/sosreports/nchilaka/bug.1567100/
https://review.gluster.org/19727
Verified this BZ while updating the machines from glusterfs-3.12.2-9 to glusterfs-3.12.2-10. Followed the same steps as in the description, didn't see the messages mentioned in the description above. 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://access.redhat.com/errata/RHSA-2018:2607