[Bugs] [Bug 1385226] New: arbiter volume write performance is bad with sharding

[bugzilla at redhat.com]

https://bugzilla.redhat.com/show_bug.cgi?id=1385226

            Bug ID: 1385226
           Summary: arbiter volume write performance is bad with sharding
           Product: GlusterFS
           Version: 3.7.15
         Component: arbiter
          Keywords: Triaged
          Assignee: bugs at gluster.org
          Reporter: ravishankar at redhat.com
                CC: bugs at gluster.org
        Depends On: 1384906
            Blocks: 1375125, 1385224



+++ This bug was initially created as a clone of Bug #1384906 +++

Description of problem:
Max Raba reported  that arbiter volume write performance is bad with sharding.
See BZ 1375125

--- Additional comment from Worker Ant on 2016-10-14 06:58:54 EDT ---

REVIEW: http://review.gluster.org/15641 (afr: Take full locks in arbiter only
for data transactions) posted (#1) for review on master by Ravishankar N
(ravishankar at redhat.com)

--- Additional comment from Worker Ant on 2016-10-14 13:18:40 EDT ---

COMMIT: http://review.gluster.org/15641 committed in master by Pranith Kumar
Karampuri (pkarampu at redhat.com) 
------
commit 3a97486d7f9d0db51abcb13dcd3bc9db935e3a60
Author: Ravishankar N <ravishankar at redhat.com>
Date:   Fri Oct 14 16:09:08 2016 +0530

    afr: Take full locks in arbiter only for data transactions

    Problem:
    Sharding exposed a bug in arbiter config. where `dd` throughput was
    extremely slow. Shard xlator was sending a fxattrop to update the file
    size immediately after a writev. Arbiter was incorrectly over-riding the
    LLONGMAX-1 start offset (for metadata domain locks) for this fxattrop,
    causing the inodelk to be taken on the data domain. And since the
    preceeding writev hadn't released the lock (afr does a 'lazy'
    unlock if write succeeds on all bricks), this degraded to a blocking
    lock causing extra lock/unlock calls and delays.

    Fix:
    Modify flock.l_len and flock.l_start to take full locks only for data
    transactions.

    Change-Id: I906895da2f2d16813607e6c906cb4defb21d7c3b
    BUG: 1384906
    Signed-off-by: Ravishankar N <ravishankar at redhat.com>
    Reported-by: Max Raba <max.raba at comsysto.com>
    Reviewed-on: http://review.gluster.org/15641
    Smoke: Gluster Build System <jenkins at build.gluster.org>
    NetBSD-regression: NetBSD Build System <jenkins at build.gluster.org>
    CentOS-regression: Gluster Build System <jenkins at build.gluster.org>
    Reviewed-by: Pranith Kumar Karampuri <pkarampu at redhat.com>


Referenced Bugs:

https://bugzilla.redhat.com/show_bug.cgi?id=1375125
[Bug 1375125] arbiter volume write performance is bad.
https://bugzilla.redhat.com/show_bug.cgi?id=1384906
[Bug 1384906] arbiter volume write performance is bad with sharding
https://bugzilla.redhat.com/show_bug.cgi?id=1385224
[Bug 1385224] arbiter volume write performance is bad with sharding
-- 
You are receiving this mail because:
You are on the CC list for the bug.
You are the assignee for the bug.