glusterfs.git, branch v3.4.2qa1

tests: Change 'volume create' to 'volume create force'

2013-11-27T05:39:30+00:00

Using 'force' when creating volumes prevents errors when creating
bricks in the root partition.
This fixes test bug-823081.t for bug-962226

Change-Id: I210ffda6bdc2cced4bd54755a2b44fdfd05b813b
Original-author: John Smith 
BUG: 962226
Signed-off-by: John Smith 
Signed-off-by: Krutika Dhananjay 
Reviewed-on: http://review.gluster.org/6356
Tested-by: Gluster Build System 
Reviewed-by: Anand Avati

glusterd: changes in 'volume create' behaviour

2013-11-27T05:39:16+00:00

        Backport of http://review.gluster.org/4740

Note: This patch is needed by oVirt's Gluster Integration Project

BUG: 948729
Change-Id: I7a9d7c36b7e15269f95637cd9061abac6f8a97de
Signed-off-by: Krutika Dhananjay 
Reviewed-on: http://review.gluster.org/6355
Tested-by: Gluster Build System 
Reviewed-by: Anand Avati

cli: write 'volume rebalance' error message in xml format when

2013-11-27T00:47:03+00:00

--xml is specified

When 'volume rebalance' encounters an error the cli prints the
error message in plain text independent of whether --xml is
specified. This throws off client application that expect xml
output (as mentioned in bz1026143).

Now, if the --xml flag is supplied, the cli print 'volume
rebalance' error messages in xml format.

(cherry picked from commit
 cbb47056ab09ae09eed2550584f190fd5a42fb12)

Change-Id: Iec266d56cb2ac34be0ac05a473ae3a08988cca38
BUG: 1026143
Signed-off-by: Dawit Alemu 
Reviewed-on: http://review.gluster.org/6250
Tested-by: Gluster Build System 
Reviewed-by: Anand Avati

cluster/dht: set layout in inode ctx even if linkfile fails

2013-11-26T20:52:25+00:00

Creating linkfile could have failed, but we dont care about linkfile
for setting layout in the inode ctx (could be EEXIST etc.)

So ignore @inode in cbk and pick it up from local->loc.inode

Change-Id: I2952799d7ae0d3441b84b2ca2981afd75d7576e2
BUG: 1032859
Signed-off-by: Anand Avati 
Reviewed-on: http://review.gluster.org/6358
Tested-by: Gluster Build System

cli: add peerid to volume status xml output

2013-11-26T19:53:13+00:00

This patch adds  tag to bricks and nfs/shd like services to
volume status xml output.

BUG: 955548
Change-Id: I0e58e323534a19d485c9523466bce215bd466160
Signed-off-by: Bala.FA 
Reviewed-on: http://review.gluster.org/6267
Tested-by: Gluster Build System 
Reviewed-by: Anand Avati

cli: use proper copy to set node-name

2013-11-26T19:51:34+00:00

Previously node-name is set to point to node-uuid which could cause
memory leak.  This is fixed by having memory copy of node-uuid.

BUG: 1012296
Change-Id: I4a7123771e2d8c31c5db4f78d022a9f4fbfc2667
Signed-off-by: Bala.FA 
Reviewed-on: http://review.gluster.org/6331
Tested-by: Gluster Build System 
Reviewed-by: Anand Avati

core: fix errno for non-existent GFID

2013-11-26T19:49:05+00:00

When clients refer to a GFID which does not exist, the errno to
be returned in ESTALE (and not ENOENT). Even though ENOENT might
look "proper" most of the time, as the application eventually expects
ENOENT even if a parent directory does not exist, not returning
ESTALE results in resolvers (FUSE and GFAPI) to not retry resolution
in uncached mode. This can result in spurious ENOENTs during
concurrent path modification operations.

Change-Id: I7a06ea6d6a191739f2e9c6e333a1969615e05936
BUG: 1032894
Signed-off-by: Anand Avati 
Reviewed-on: http://review.gluster.org/6322
Tested-by: Gluster Build System

gfapi: remove unnecessary call to glfs_resolve_base()

2013-11-26T19:02:17+00:00

Calling glfs_resolve_base() on the root inode for every resolver
invocation is unnecessary and wasteful.

Here are the results from running a test program which performs
path based operations (creates and deletes 1000 files):

Without patch:
[root@blackbox ~]# sync
[root@blackbox ~]# time ./a.out 1

real    0m4.314s
user    0m1.923s
sys     0m1.144s
[root@blackbox ~]# sync
[root@blackbox ~]# time ./a.out 1

real    0m4.383s
user    0m1.940s
sys     0m1.177s
[root@blackbox ~]# sync
[root@blackbox ~]# time ./a.out 1

real    0m4.339s
user    0m1.863s
sys     0m1.129s

With patch:

[root@blackbox ~]# sync
[root@blackbox ~]# time ./a.out 1

real    0m3.005s
user    0m1.162s
sys     0m0.816s
[root@blackbox ~]# sync
[root@blackbox ~]# time ./a.out 1

real    0m3.188s
user    0m1.222s
sys     0m0.867s
[root@blackbox ~]# sync
[root@blackbox ~]# time ./a.out 1

real    0m2.999s
user    0m1.131s
sys     0m0.832s

Change-Id: Id160a24f44b4dccfcfce99a6f69ddb8938523cd5
BUG: 953694
Signed-off-by: Anand Avati 
Reviewed-on: http://review.gluster.org/6321
Tested-by: Gluster Build System

gfapi: use native STACK_WIND for read _async() calls

2013-11-26T19:02:01+00:00

There is little value in using synctask wrappers for async IO
requests, as STACK_WIND is asynchronous by nature already.

Skip going through synctask for read/write async calls.

Change-Id: Ifde331d7c97e0f33426da6ef4377c5ba70dddc06
BUG: 1009134
Signed-off-by: Anand Avati 
Reviewed-on: http://review.gluster.org/6325
Tested-by: Gluster Build System

gfapi: Fix iobuf leaks in gfapi

2013-11-26T18:39:39+00:00

In glfs_readv, if syncop_readv() fails with return value <= 0
the iobref was not being unrefd which would cause iobuf leaks.

Change-Id: I9850ae149e53cf75ba26f8b5f4c5446cce4b5991
BUG: 1018176
Signed-off-by: Poornima 
Reviewed-on: http://review.gluster.org/6324
Tested-by: Gluster Build System 
Reviewed-by: Vijay Bellur