glusterfs.git/xlators/mgmt/glusterd/src/glusterd-utils.h, branch v3.8.3

glusterd: Fix gsyncd upgrade issue

2016-07-15T16:47:44+00:00

Backport of http://review.gluster.org/14898/

Problem:
    gluster upgrade is not generating new volfiles

Cause:
During upgrade, "glusterd --xlator-option *.upgrade=on -N"
is run to generate new volfiles. It is run post 'glusterfs'
rpm installation. The above command fails during upgrade
if geo-replication is installed. This is because on
glusterd start 'gsyncd' binary is called to configure
geo-replication related stuff. Since 'glusterfs' rpm is
installed prior to 'geo-rep' rpm, the 'gsyncd' binary
used to glusterd upgrade command is of old version and
hence it fails before generating new volfiles.

Solution:
Don't call geo-replication configure during upgrade/downgrade.
Geo-replication configuration happens during start of glusterd
after upgrade.

Change-Id: Id58ea44ead9f69982f86fb68dc5b9ee3f6cd11a1
BUG: 1356439
Signed-off-by: Kotresh HR 
Reviewed-on: http://review.gluster.org/14898
(cherry picked from commit 1b998788ece8c8b52657e8b9aae65d3279690c5b)
Reviewed-on: http://review.gluster.org/14916
Smoke: Gluster Build System 
CentOS-regression: Gluster Build System 
NetBSD-regression: NetBSD Build System 
Reviewed-by: Atin Mukherjee

glusterd: compare uuid instead of hostname address resolution

2016-07-05T14:43:24+00:00

Backport of http://review.gluster.org/14849

In glusterd_get_brickinfo () brick's hostname is address resolved. This adds an
unnecessary latency since it uses calls like getaddrinfo (). Instead given the
local brick's uuid is already known a comparison of MY_UUID and brickinfo->uuid
is much more light weight than the previous approach.

On a scale testing where cluster hosting ~400 volumes spanning across 4 nodes,
if a node goes for a reboot, few of the bricks don't come up. After few days of
analysis its found that glusterd_pmap_sigin () was taking signficant amount of
latency and further code walthrough revealed this unnecessary address
resolution. Applying this fix solves the issue and now all the brick processes
come up on a node reboot.

Backport of commit 633e6fe265bc2de42dade58dc6a15c285957da76:
> Change-Id: I299b8660ce0da6f3f739354f5c637bc356d82133
> BUG: 1352279
> Signed-off-by: Atin Mukherjee 
> Reviewed-on: http://review.gluster.org/14849
> Smoke: Gluster Build System 
> NetBSD-regression: NetBSD Build System 
> CentOS-regression: Gluster Build System 
> Reviewed-by: Prashanth Pai 
> Reviewed-by: Samikshan Bairagya 
> Reviewed-by: Kaushal M 

Change-Id: I299b8660ce0da6f3f739354f5c637bc356d82133
BUG: 1352817
Signed-off-by: Atin Mukherjee 
Reviewed-on: http://review.gluster.org/14860
NetBSD-regression: NetBSD Build System 
CentOS-regression: Gluster Build System 
Reviewed-by: Niels de Vos 
Smoke: Gluster Build System

cli/glusterd: add/remove brick fixes for arbiter volumes

2016-05-24T08:35:57+00:00

Backport of: http://review.gluster.org/14126

1.Provide a command to convert replica 2 volumes to arbiter volumes.
Existing self-heal logic will automatically heal the file hierarchy into
the arbiter brick, the progress of which can be monitored using the
heal info command.

Syntax: gluster volume add-brick  replica 3 arbiter 1


2. Add checks when removing bricks from arbiter volumes:
- When converting from arbiter to replica 2 volume, allow only arbiter
  brick to be removed.
- When converting from arbiter to plain distribute volume, allow only if
  arbiter is one of the bricks that is removed.

3. Some clean-up:
- Use GD_MSG_DICT_GET_SUCCESS instead of GD_MSG_DICT_GET_FAILED to
log messages that are not failures.
- Remove unused variable `brick_list`
- Move 'brickinfo->group' related functions to glusted-utils.

Change-Id: Ifa75d137c67ffddde7dcb8e0df0873163e713119
BUG: 1337387
Signed-off-by: Ravishankar N 
Reviewed-on: http://review.gluster.org/14502
Smoke: Gluster Build System 
NetBSD-regression: NetBSD Build System 
CentOS-regression: Gluster Build System 
Reviewed-by: Atin Mukherjee

glusterd: copy real_path from older brickinfo during brick import

2016-05-20T09:33:10+00:00

Backport of http://review.gluster.org/14306

In glusterd_import_new_brick () new_brickinfo->real_path will not be populated
for the first time and hence if the underlying file system is bad for the same
brick, import will fail resulting in inconsistent configuration data.

Fix is to populate real_path from old brickinfo object.

Also there were many cases where we were unnecessarily calling realpath() and
that may cause in failure. For eg - if a remove brick is executed with a brick
whoose underlying file system has crashed, remove-brick fails since realpath()
call fails. We'd need to call realpath() here as the value is of no use.Hence
passing construct_realpath as _gf_false in glusterd_volume_brickinfo_get_by_brick ()
is a must in such cases.

Change-Id: I7ec93871dc9e616f5d565ad5e540b2f1cacaf9dc
BUG: 1337114
Signed-off-by: Atin Mukherjee 
Reviewed-on: http://review.gluster.org/14306
Smoke: Gluster Build System 
NetBSD-regression: NetBSD Build System 
CentOS-regression: Gluster Build System 
Reviewed-by: Kaushal M 
Reviewed-on: http://review.gluster.org/14411

quota/glusterd: enhance quota enable and disable process

2016-04-29T10:11:35+00:00

Previously quota crawl was done from the single mount point,
this is very slow process if there are huge number of files exists
in the volume

This RFE will now spawn crawl process for each brick in the
volume, and  files are looked in parallel independently for each
brick. This improves the speed of crawling process for
entire files-system

This patch also fixes below problem
* Previously, mountdir was created under '/tmp'.
  If someone tries to cleanup '/tmp'/ directory
  then it is very dangerous that we loose volume data
  So create a mount point under /var/run/gluster/tmp
  instead

* Previously, file-system crawl is performed from all the nodes,
  which is a redundant operation and performance will degrade
  The problem is fixed with this patch

Change-Id: Icabedeb44182139ace9c8106793803122388cab8
BUG: 1290766
Signed-off-by: vmallika 
Reviewed-on: http://review.gluster.org/12952
Smoke: Gluster Build System 
NetBSD-regression: NetBSD Build System 
CentOS-regression: Gluster Build System 
Reviewed-by: Manikandan Selvaganesh 
Reviewed-by: Atin Mukherjee

glusterd: populate brickinfo->real_path conditionally

2016-04-11T18:39:16+00:00

glusterd_brickinfo_new_from_brick () is called from multiple places and one of
them is glusterd_brick_rpc_notify where its very well possible that an
underlying brick's file system has crashed and a disconnect event has been
received. In this case glusterd tries to build the brickinfo from the brickid in
the RPC request, however the same fails as glusterd_brickinfo_new_from_brick ()
fails from realpath.

Fix is to skip populating real_path if its a disconnect event.

Change-Id: I9d9149c64a9cf2247abb731f219c1b1eef037960
BUG: 1325841
Signed-off-by: Atin Mukherjee 
Reviewed-on: http://review.gluster.org/13965
Smoke: Gluster Build System 
CentOS-regression: Gluster Build System 
NetBSD-regression: NetBSD Build System 
Reviewed-by: Jeff Darcy

glusterd / afr : Enable auto heal when replica count increases

2016-03-21T17:51:00+00:00

In replicate volumes, when a brick is added to a replicate
group, heal to the new brick should be triggered.
Also, the new brick should not be considered as source for
healing till it is up to date.

Previously, extended attributes had to be set manually on
the bricks for this to happen. This patch is part 1 patch
to automate this process.

Change-Id: I29958448618372bfde23bf1dac5dd23dba1ad98f
BUG: 1276203
Signed-off-by: Anuradha Talur 
Reviewed-on: http://review.gluster.org/12451
Reviewed-by: Atin Mukherjee 
NetBSD-regression: NetBSD Build System 
CentOS-regression: Gluster Build System 
Reviewed-by: Ravishankar N 
Smoke: Gluster Build System

Tier: "tier start force" command implementation

2015-12-23T00:24:11+00:00

The start command doesnt restart the tier deamon if the deamon
is running at one node. hence to bring up the tierd on the nodes
where the deamon is down, the force command is implemented.
It skips the check for tierd running.

Change-Id: I0037d3e5ecfe56637d0da201a97903c435d26436
BUG: 1292112
Signed-off-by: hari gowtham 
Reviewed-on: http://review.gluster.org/12983
Reviewed-by: Atin Mukherjee 
Tested-by: Gluster Build System 
Reviewed-by: Dan Lambright 
Tested-by: Dan Lambright

glusterd: Disallow peer with existing volumes to be probed in cluster

2015-12-07T06:23:08+00:00

As of now we do allow peer to get added in the trusted storage pool even if it
has a volume configured. This is definitely not a supported configuration and
can lead to issues as we never claim to support merging clusters. A single node
running a standalone volume can be considered as a cluster.

Change-Id: Id0cf42d6e5f20d6bfdb7ee19d860eee67c7c45be
BUG: 1287992
Signed-off-by: Atin Mukherjee 
Reviewed-on: http://review.gluster.org/12864
Tested-by: NetBSD Build System 
Reviewed-by: Kaushal M 
Tested-by: Gluster Build System

glusterd/tier: Reset to reconfigured values after detach-tier

2015-11-26T19:39:17+00:00

After detach-tier commit , we have to reset the option
reconfigured for tier volume

Change-Id: Iae0210259720d6ac14ccc0cc339dc9f54a0c4571
BUG: 1285046
Signed-off-by: Mohammed Rafi KC 
Reviewed-on: http://review.gluster.org/12736
Tested-by: NetBSD Build System 
Tested-by: Gluster Build System 
Reviewed-by: Dan Lambright 
Tested-by: Dan Lambright