glusterfs.git/geo-replication, branch v3.12.5

geo-rep: Fix status transition

2017-10-12T18:46:28+00:00

The status transition is as below which is
wrong.

Created->Initializing->Active->Active/Passive->Stopped

As soon as the monitor spawns the worker, the state
is changed from 'Initializing' to 'Active' and then to
'Active/Passive' based on whether worker gets the lock
or not. This is wrong and it should directly tranistion
as below.

Created->Initializing->Active/Passive->Stopped

> Change-Id: Ibf5ca5c4fdf168c403c6da01db60b93f0604aae7
> BUG: 1500284
> Signed-off-by: Kotresh HR 
(cherry picked from commit 3edf926a1bda43879c09694cf3904c214c94c9dc)

Change-Id: Ibf5ca5c4fdf168c403c6da01db60b93f0604aae7
BUG: 1500835
Signed-off-by: Kotresh HR

geo-rep: Add EINTR to retry list while doing readlink

2017-10-12T18:38:41+00:00

Worker occasionally crashed with EINTR on readlink.
This is not persistent and is transient. Worker restart
invovles re-processing of few entries in changenlogs.
So adding EINTR to retry list to avoid worker restart.

> Change-Id: Iefe641437b5d5be583f079fc2a7a8443bcd19f9d
> BUG: 1499393
> Signed-off-by: Kotresh HR 
(cherry picked from commit 34d52445a9058310d7512c9bcc8c01e709aac1ef)

Change-Id: Iefe641437b5d5be583f079fc2a7a8443bcd19f9d
BUG: 1500845
Signed-off-by: Kotresh HR

geo-rep: Add ENODATA to retry list on gfid getxattr

2017-10-12T18:38:06+00:00

During xsync crawl, worker occasionally crashed
with ENODATA on getting gfid from backend. This
is not persistent and is transient. Worker restart
invovles re-processing of few entries in changenlogs.
So adding ENODATA to retry list to avoid worker
restart.

> Change-Id: Ib78d1e925c0a83c78746f28f7c79792a327dfd3e
> BUG: 1499391
> Signed-off-by: Kotresh HR 
(cherry picked from commit b56bdb34dafd1a87c5bbb2c9a75d1a088d82b1f4)

Change-Id: Ib78d1e925c0a83c78746f28f7c79792a327dfd3e
BUG: 1500841
Signed-off-by: Kotresh HR

geo-rep: Fix passive brick's last sync time

2017-10-12T18:36:17+00:00

Passive brick's stime was not updated to the
status file immediately after updating the brick
root. As a result the last sync time was showing
'0' until it finishes first crawl if passive
worker becomes active after restart. Fix is to
update the status file immediately after upgrading
the brick root.

> Change-Id: I248339497303bad20b7f5a1d42ab44a1fe6bca99
> BUG: 1500346
> Signed-off-by: Kotresh HR 
(cherry picked from commit f18a47ee7e6e06c9a9a8893aef7957f23a18de53)

Change-Id: I248339497303bad20b7f5a1d42ab44a1fe6bca99
BUG: 1500853
Signed-off-by: Kotresh HR

geo-rep: Add ENOTSUP error to retry list

2017-10-12T18:27:13+00:00

os.listdir gives ENOTSUP on gfid path occasionally
which is not persistant. Adding it to retry list
to avoid worker to crash if it's transient error.

> Change-Id: Ic795dd1f02a27c9e5d901e20722ee32451838feb
> BUG: 1499180
> Signed-off-by: Kotresh HR 
(cherry picked from commit 3afbe4201d864d2e886509f5633468aa344fa444)

Change-Id: Ic795dd1f02a27c9e5d901e20722ee32451838feb
BUG: 1500396
Signed-off-by: Kotresh HR

geo-rep: Fix syncing of hardlink of symlink

2017-08-29T12:42:31+00:00

Problem:
If there is a hardlink to a symlink on master
and if the symlink file is deleted on master,
geo-rep fails to sync the hardlink.

Typical Usecase:
It's easily hit with rsnapshot use case where
it uses hardlinks.

Example Reproducer:
Setup geo-replication between master and slave
volume and in master mount point, do the following.

 1. mkdir /tmp/symlinkbug
 2. ln -f -s /does/not/exist /tmp/symlinkbug/a_symlink
 3. rsync -a /tmp/symlinkbug ./
 4. cp -al symlinkbug symlinkbug.0
 5. ln -f -s /does/not/exist2 /tmp/symlinkbug/a_symlink
 6. rsync -a /tmp/symlinkbug ./
 7. cp -al symlinkbug symlinkbug.1

Cause:
If the source was not present while syncing hardlink,
it was always packing the blob as regular file.

Fix:
If the source was not present while syncing hardlink,
pack the blob based on the mode.

> Change-Id: Iaa12d6f99de47b18e0650e7c4eb455f23f8390f2
> BUG: 1432046
> Signed-off-by: Kotresh HR 
> Reported-by: Christian Lohmaier 
> Reviewed-on: https://review.gluster.org/18011
> Smoke: Gluster Build System 
> CentOS-regression: Gluster Build System 
> Reviewed-by: Aravinda VK 
(cherry picked from commit e893962deaabab8e934813f8a0443a8f94e009f2)

Change-Id: Iaa12d6f99de47b18e0650e7c4eb455f23f8390f2
BUG: 1486120
Signed-off-by: Kotresh HR 
Reviewed-on: https://review.gluster.org/18127
Smoke: Gluster Build System 
CentOS-regression: Gluster Build System 
Reviewed-by: Shyamsundar Ranganathan

geo-rep: Convert gfid mismatch logs to structured logging

2017-08-03T14:11:51+00:00

Convert the logs related to entry failures fix
due to gfid mismatch logs into structured logging
format

Change-Id: I9bce950c5339b48d3ec8b84bddee38b0473b7634
> Updates: #246
> Signed-off-by: Kotresh HR 
> Reviewed-on: https://review.gluster.org/17896
> Smoke: Gluster Build System 
> CentOS-regression: Gluster Build System 
> Reviewed-by: Aravinda VK 
(cherry picked from commit de69fb7ac819c7e92d132a5ead12ea317ef1ec20)

Change-Id: I9bce950c5339b48d3ec8b84bddee38b0473b7634
Updates: #246
Signed-off-by: Kotresh HR 
Reviewed-on: https://review.gluster.org/17920
Reviewed-by: Aravinda VK 
CentOS-regression: Gluster Build System 
Smoke: Gluster Build System

geo-rep: Fix syncing of self healed hardlinks

2017-07-31T16:29:32+00:00

Problem:
In a distribute replicate volume, if the hardlinks
are created when a subvolume is down, it gets
healed from other subvolume when it comes up.
If this subvolume becomes ACTIVE in geo-rep
there are chances that those hardlinks won't
be synced to slave.

Cause:
AFR can't detect hardlinks during self heal.
It just create those files using mknod and
the same is recorded in changelog. Geo-rep
processes these mknod and ignores it as
it finds gfid already on slave.

Solution:
Geo-rep should process the mknod as link
if the gfid already exists on slave.

> Change-Id: I2f721b462b38a74c60e1df261662db4b99b32057
> BUG: 1475308
> Signed-off-by: Kotresh HR 
> Reviewed-on: https://review.gluster.org/17880
> Smoke: Gluster Build System 
> CentOS-regression: Gluster Build System 
> Reviewed-by: Aravinda VK 
(cherry picked from commit d685e4238fafba8f58bf01174c79cb5ca35203e5)

Change-Id: I2f721b462b38a74c60e1df261662db4b99b32057
Signed-off-by: Kotresh HR 
BUG: 1476208
Reviewed-on: https://review.gluster.org/17905
Smoke: Gluster Build System 
CentOS-regression: Gluster Build System 
Reviewed-by: Aravinda VK

geo-rep: Handle possible entry failures gracefully

2017-07-21T10:38:12+00:00

Updates: #246

Change-Id: If0ce83fe8dd3068bfb671f398b2e82ac831288d0
Signed-off-by: Kotresh HR 
Reviewed-on: https://review.gluster.org/17577
Smoke: Gluster Build System 
CentOS-regression: Gluster Build System 
Reviewed-by: Aravinda VK

geo-rep: Fix changelog encoding to encode only space and newline

2017-07-21T04:44:30+00:00

libgfchangelog was encoding path using spec rfc3986, but encoding only
required for SPACE and NEWLINE chars since the NEWLINE char is used as
record separator and SPACE as field separator in the parsed changelogs
output.

Changed the encoding function to encode only SPACE and NEWLINE.

BUG: 1451724
Change-Id: I1936efad31788a9e636f912c832ed7d7efea4fe2
Signed-off-by: Aravinda VK 
Reviewed-on: https://review.gluster.org/17787
Reviewed-by: Prashanth Pai 
Reviewed-by: Kotresh HR 
CentOS-regression: Gluster Build System 
Smoke: Gluster Build System