glusterfs.git/xlators/storage, branch v3.12.5

Disable gfid2path by default on NetBSD

2017-12-08T14:38:57+00:00

NetBSD storage of extended attributes for UFS1 badly scales when
the list of extended attributes names rises. gfid2path can add as
many extended attributes names as we have files, hence we keep it
disabled for performance sake.

> Change-Id: Id77b5f5ceb4d5eba1b3362b4b9fc693450ffbc2b
> Signed-off-by: Emmanuel Dreyfus 
> BUG: 1129939

Change-Id: Id77b5f5ceb4d5eba1b3362b4b9fc693450ffbc2b
Signed-off-by: Emmanuel Dreyfus 
BUG: 1513258

geo-rep: Filter out volume-mark xattr

2017-10-23T09:15:58+00:00

The volume-mark xattr, maintained at brick root
of slave volume is specific to geo-replication
and should be filtered out for all other clients.
It should also be filtered out from list getxattr
from all mounts including geo-rep mount as it
might cause rsync to read and set.

> Change-Id: If9eb5a3af18051083c853e70d93b2819e8eea222
> BUG: 1500433
> Signed-off-by: Kotresh HR 
(cherry picked from commit c64fd0d4b0ef313bb44aae68a376ec0c9ee8657a)


Change-Id: If9eb5a3af18051083c853e70d93b2819e8eea222
BUG: 1502104
Signed-off-by: Kotresh HR

Posix: fix additional inode ref

2017-10-06T06:33:13+00:00

        backport of https://review.gluster.org/#/c/18406/

Problem: In the iteration, the inode is being ref-ed twice and
unref-ed once. this leads to ref leak.

Fix: assign the parent to the inode instead of referencing it.

>BUG: 1496379

Change-Id: Ib154b12d38ad68220f8f5288bbc50081beccc2b9
BUG: 1497084
Signed-off-by: hari gowtham

feature/posix: Enabled gfid2path by default

2017-08-29T20:09:32+00:00

Enable gfid2path feature by default. The basic
performance tests are carried out and it doesn't
show significant depreciation. The results are
updated in issue.

Updates: #139
Change-Id: I5f1949a608d0827018ef9d548d5d69f3bb7744fd
> Signed-off-by: Kotresh HR 
> Reviewed-on: https://review.gluster.org/17950
> Smoke: Gluster Build System 
> CentOS-regression: Gluster Build System 
> Reviewed-by: Aravinda VK 
> Reviewed-by: Amar Tumballi 
(cherry picked from commit 3ec63650bb7fd874a5013e7be4a2def3b519c9b2)
Reviewed-on: https://review.gluster.org/18133
Reviewed-by: Amar Tumballi 
Smoke: Gluster Build System 
Tested-by: Shyamsundar Ranganathan 
CentOS-regression: Gluster Build System 
Reviewed-by: Shyamsundar Ranganathan

storage/posix: Use the ret value of posix_gfid_heal()

2017-08-12T13:37:58+00:00

... to make the change in commit acf8cfdf truly useful.

Without this, a race between entry creation fops and lookup
at posix layer can cause lookups to fail with ENODATA, as
opposed to ENOENT.

Backport of:
> Change-Id: I44a226872283a25f1f4812f03f68921c5eb335bb
> Reviewed-on: https://review.gluster.org/17821
> BUG: 1472758
> cherry-picked from 669868d23eaeba42809fca7be134137c607d64ed

Change-Id: I44a226872283a25f1f4812f03f68921c5eb335bb
BUG: 1479717
Signed-off-by: Krutika Dhananjay 
Reviewed-on: https://review.gluster.org/18010
CentOS-regression: Gluster Build System 
Reviewed-by: Raghavendra Bhat 
Smoke: Gluster Build System

posix: add null gfid checks

2017-08-12T13:29:47+00:00

(Backport of https://review.gluster.org/17975)

...in file/dir creation and lookup codepaths. The check is relaxed for
fops coming from trash xlator at the moment until trash has client side
logic to send the create fops with gfid-req.

Also fixed the missing trash pid assignment in creates sent by trash
xlator. Without this, truncated files won't be moved to .trashcan.

Change-Id: Ieddd7f0634850e7c7010e4fbb4ad1eead35888c8
BUG: 1479474
Signed-off-by: Ravishankar N 
Reviewed-on: https://review.gluster.org/17996
Smoke: Gluster Build System 
CentOS-regression: Gluster Build System 
Reviewed-by: Shyamsundar Ranganathan

cluster/dht: Check for open fd only on EBADF

2017-08-09T11:29:50+00:00

DHT fd based fops used to check if the fd was open
on the cached subvol before winding the call. However,
this introduced a performance regression of about
30% for reads.

This check was introduced to handle cases where files
were migrated while IOs were happening. As this is not
the common case, dht will now check if the fd is
open on the cached subvol only if the call fails
with EBADF.

This will prevent a performance hit where a rebalance
is not running.

> BUG: 1476665
> Signed-off-by: N Balachandran 
> Reviewed-on: https://review.gluster.org/17976
> Smoke: Gluster Build System 
> CentOS-regression: Gluster Build System 
> Reviewed-by: Amar Tumballi 
> Reviewed-by: Susant Palai 
> Reviewed-by: Raghavendra G 

(cherry picked from commit cdca1cb26a0aba390c6d8485c0d6d95e22ffc8bd)
Change-Id: I2035a858d63c3fcd22bb634055bbb0ad01686808
BUG: 1479303
Signed-off-by: N Balachandran 
Reviewed-on: https://review.gluster.org/17995
Smoke: Gluster Build System 
CentOS-regression: Gluster Build System 
Reviewed-by: Shyamsundar Ranganathan

posix: add sanity checks for removing the gfid symlink for directories

2017-08-08T12:24:02+00:00

...during mkdir and rmdir. Otherwise, during entry self-heal, the
directory could be left out without a .glusterfs symlink causing fops like
opendir, readdir to fail.  The only chance the missing symlink will be
created is when a fresh lookup comes on it.

> Reviewed-on: https://review.gluster.org/17945
> Smoke: Gluster Build System 
> CentOS-regression: Gluster Build System 
> Reviewed-by: Raghavendra Bhat 
(cherry picked from commit 42c057a6d1a03dd2825a278393acb15d52220c8d)

Change-Id: I2e1cf1bce8962ea80187edd8f6d73e0a09cf9f8e
BUG: 1479118
Signed-off-by: Ravishankar N 
Reviewed-on: https://review.gluster.org/17991
Smoke: Gluster Build System 
CentOS-regression: Gluster Build System 
Reviewed-by: Shyamsundar Ranganathan

storage/posix: Add virtual xattr to fetch path from gfid

2017-07-31T15:45:02+00:00

The gfid2path infra stores the "pargfid/bname" as on xattr
value for each non directory entry. Hardlinks would have a
separate xattr. This xattr key is internal and is not
exposed to applications.  A virtual xattr is exposed for
the applications to fetch the path from gfid.

Internal xattr:
trusted.gfid2path.

Virtual xattr:
glusterfs.gfidtopath

getfattr -h -n glusterfs.gfidtopath //.gfid/

If there are hardlinks, it returns all the paths separated
by ':'.

A volume set option is introduced to change the delimiter
to required string of max length 7.

gluster vol set gfid2path-separator ":::"

> Updates: #139
> Change-Id: Ie3b0c3fd8bd5333c4a27410011e608333918c02a
> Signed-off-by: Kotresh HR 
> Reviewed-on: https://review.gluster.org/17785
> Smoke: Gluster Build System 
> CentOS-regression: Gluster Build System 
> Reviewed-by: Krutika Dhananjay 

Updates: #139
Change-Id: Ie3b0c3fd8bd5333c4a27410011e608333918c02a
Signed-off-by: Kotresh HR 
Reviewed-on: https://review.gluster.org/17921
CentOS-regression: Gluster Build System 
Smoke: Gluster Build System 
Reviewed-by: Shyamsundar Ranganathan

posix: option to handle the shared bricks for statvfs()

2017-07-31T15:34:58+00:00

Currently 'storage/posix' xlator has an option called option
`export-statfs-size no`, which exports zero as values for few
fields in `struct statvfs`. In a case of backend brick shared
between multiple brick processes, the values of these variables
should be `field_value / number-of-bricks-at-node`. This way,
even the issue of 'min-free-disk' etc at different layers would
also be handled properly when the statfs() sys call is made.

Fixes #241

> Reviewed-on: https://review.gluster.org/17618
> Reviewed-by: Jeff Darcy 
> Reviewed-by: Atin Mukherjee 
> (cherry picked from commit febf5ed4848ad705a34413353559482417c61467)

Change-Id: I2e320e1fdcc819ab9173277ef3498201432c275f
Signed-off-by: Amar Tumballi 
Reviewed-on: https://review.gluster.org/17903
Smoke: Gluster Build System 
CentOS-regression: Gluster Build System 
Reviewed-by: Shyamsundar Ranganathan