glusterfs.git/xlators/features/arbiter, branch release-3.7

arbiter: Fix memleak in arbiter_inode ctx

2016-08-24T21:51:47+00:00

Backport of: http://review.gluster.org/#/c/15289/

Problem:
The iattbuf ptr stored in arbiter's inode context was not freed during inode
forget.

Fix:
Change it to a statically allocated value so that we don't have to deal
with allocating/freeing it.

Change-Id: Id1b73b8aee1fb5c4174d0734bd20e168432b1abd
BUG: 1369752
Reported-by: Benjamin Edgar 
Signed-off-by: Ravishankar N 
(cherry picked from commit 4aa52061a51b97c4f865b402f977b3b43f5471a7)
Reviewed-on: http://review.gluster.org/15307
Smoke: Gluster Build System 
NetBSD-regression: NetBSD Build System 
CentOS-regression: Gluster Build System 
Reviewed-by: Pranith Kumar Karampuri

arbiter: write performance improvement

2016-04-11T12:04:04+00:00

Backport of: http://review.gluster.org/#/c/13906

Problem: The throughput for a 'dd' workload was much less for arbiter
configuration when compared to normal replica-3 volume. There were 2
issues:

i)arbiter_writev was using the request dict as response dict while
unwinding, leading to incorect GLUSTERFS_WRITE_IS_APPEND and
GLUSTERFS_OPEN_FD_COUNT values (=4), leading to immediate post-ops
because is_afr_delayed_changelog_post_op_needed() failed due to
afr_are_multiple_fds_opened() check.

ii) The arbiter code in afr was setting local->transaction.{start and len} =0
to take full file locks. What this meant was even for simultaenous but
non-overlapping writevs, afr_transaction_eager_lock_init() was not
happening because afr_locals_overlap() always stays true. Consequently
is_afr_delayed_changelog_post_op_needed() failed due to
local->delayed_post_op not being set.

Fix:
i) Send appropriate response dict values in arbiter_writev.
ii) Modify flock params instead of local->transaction.{start and len} to
take full file locks in the transaction.

Also changed _fill_writev_xdata() in posix to fill rsp_xdata for
whatever key is requested for.

Change-Id: I1c5fc5e98aba49ade540bb441a022e65b753432a
BUG: 1324809
Signed-off-by: Ravishankar N 
Reported-by: Robert Rauch 
Reported-by: Russel Purinton 
Reviewed-on: http://review.gluster.org/13925
Smoke: Gluster Build System 
Reviewed-by: Pranith Kumar Karampuri 
CentOS-regression: Gluster Build System 
NetBSD-regression: NetBSD Build System

build: remove broken and unused _la_LIBADD variable

2015-03-20T03:55:11+00:00

While building, the following warning is displayed:

  xlators/features/arbiter/src/Makefile.am:7: variable `_la_LIBADD' is defined but no program or
  xlators/features/arbiter/src/Makefile.am:7: library has `_la' as canonical name (possible typo)

The _la_LIBADD really seems like a typo, dropping it should not be harmful to
anything, except for the warning that will now be gone.

BUG: 1199985
Change-Id: I3f3ba911f59df2e51fdc6387295fff4bbcc5a12d
Signed-off-by: Niels de Vos 
Reviewed-on: http://review.gluster.org/9950
Reviewed-by: Kaleb KEITHLEY 
Tested-by: Gluster Build System 
Reviewed-by: Ravishankar N 
Reviewed-by: Vijay Bellur

afr: arbiter xlator

2015-03-19T17:45:49+00:00

This patch adds the arbiter translator into the tree. This is a server
side xlator used for replica 3 volumes. It sits above posix and will be
loaded on the 3rd (last) brick of every afr subvolume in a replica 3
configuration. It intercepts inode read/write operations: reads are
unwound with ENOTCONN, inode writes are unwound with success without
actually passing them down to posix. Metadata operations are allowed to
pass through.

The CLI for creating a 3 way replica with arbiter is also added but kept
disabled (A 'normal' 3 way replica is created instead).

This patch is a part of the arbiter logic implementation for 3 way AFR,
details of which can be found at http://review.gluster.org/#/c/9656/

Change-Id: I395b81f49d5da52c466daf5c8518f1bbad9c16fa
BUG: 1199985
Signed-off-by: Ravishankar N 
Reviewed-on: http://review.gluster.org/9840
Tested-by: Gluster Build System 
Reviewed-by: Pranith Kumar Karampuri 
Reviewed-by: Vijay Bellur