+++ This bug was initially created as a clone of Bug #1242317 +++ This is an umbrella bug to track the ongoing effort of improving client I/O latencies observed when objects are signed. --- Additional comment from Anand Avati on 2015-07-12 23:19:56 EDT --- REVIEW: http://review.gluster.org/11461 (features/bitrot: move inode state just at the last moment) posted (#2) for review on master by Venky Shankar (vshankar) --- Additional comment from Anand Avati on 2015-07-12 23:20:03 EDT --- REVIEW: http://review.gluster.org/11461 (features/bitrot: move inode state just at the last moment) posted (#3) for review on master by Venky Shankar (vshankar) --- Additional comment from Venky Shankar on 2015-07-12 23:28:19 EDT --- Not moving bug state as there would be multiple patches (first improvement posted) to improve client I/O performance. --- Additional comment from Anand Avati on 2015-07-13 05:14:21 EDT --- COMMIT: http://review.gluster.org/11461 committed in master by Raghavendra Bhat (raghavendra) ------ commit 14e80e6e5d2d5d4fa568019fbe30cffa52714f87 Author: Venky Shankar <vshankar> Date: Tue Jun 30 13:10:46 2015 +0530 features/bitrot: move inode state just at the last moment Which was done at half the set expiry time resulting in actual IOs incrementing the object version. Now this is done just at the last moment with re-notification now cut-shorting into checksum calculation without waiting in the timer-wheel. BUG: 1242317 Change-Id: If655b77d822ebf7b2a4f65e1b5583dd3609306e7 Signed-off-by: Venky Shankar <vshankar> Reviewed-on: http://review.gluster.org/11461 Tested-by: NetBSD Build System <jenkins.org> Tested-by: Gluster Build System <jenkins.com> Reviewed-by: Raghavendra Bhat <raghavendra> --- Additional comment from Venky Shankar on 2015-07-13 05:21:36 EDT --- Let's keep this ASSIGNED as there would be other enhancements too (throttle signer, etc.)
REVIEW: http://review.gluster.org/11653 (features/bitrot: move inode state just at the last moment) posted (#1) for review on release-3.7 by Venky Shankar (vshankar)
COMMIT: http://review.gluster.org/11653 committed in release-3.7 by Raghavendra Bhat (raghavendra) ------ commit 07eefaa6a51771a31343f031568e80b55f1f16c1 Author: Venky Shankar <vshankar> Date: Tue Jun 30 13:10:46 2015 +0530 features/bitrot: move inode state just at the last moment Backport of http://review.gluster.org/11461 Which was done at half the set expiry time resulting in actual IOs incrementing the object version. Now this is done just at the last moment with re-notification now cut-shorting into checksum calculation without waiting in the timer-wheel. BUG: 1242718 Change-Id: If655b77d822ebf7b2a4f65e1b5583dd3609306e7 Signed-off-by: Venky Shankar <vshankar> Reviewed-on: http://review.gluster.org/11653 Tested-by: NetBSD Build System <jenkins.org> Tested-by: Gluster Build System <jenkins.com> Reviewed-by: Raghavendra Bhat <raghavendra>
This bug is getting closed because a release has been made available that should address the reported issue. In case the problem is still not fixed with glusterfs-3.7.3, please open a new bug report. glusterfs-3.7.3 has been announced on the Gluster mailinglists [1], packages for several distributions should become available in the near future. Keep an eye on the Gluster Users mailinglist [2] and the update infrastructure for your distribution. [1] http://thread.gmane.org/gmane.comp.file-systems.gluster.devel/12078 [2] http://thread.gmane.org/gmane.comp.file-systems.gluster.user