glusterfs.git/xlators/mgmt/glusterd/src/glusterd-mgmt.c, branch v3.9rc1

glusterd : Introduce reset brick

2016-08-30T02:55:53+00:00

The command basically allows replace brick with src and
dst bricks as same.

Usage:
gluster v reset-brick   start
This command kills the brick to be reset. Once this command is run,
admin can do other manual operations that they need to do,
like configuring some options for the brick. Once this is done,
resetting the brick can be continued with the following options.

gluster v reset-brick    commit {force}

Does the job of resetting the brick. 'force' option should be used
when the brick already contains volinfo id.

Problem: On doing a disk-replacement of a brick in a replicate volume
the following 2 scenarios may occur :

a) there is a chance that reads are served from this replaced-disk brick,
which leads to empty reads. b) potential data loss if next writes succeed
only on replaced brick, and heal is done to other bricks from this one.

Solution: After disk-replacement, make sure that reset-brick command is
run for that brick so that pending markers are set for the brick and it
is not chosen as source for reads and heal. But, as of now replace-brick
for the same brick-path is not allowed. In order to fix the above
mentioned problem, same brick-path replace-brick is needed.
With this patch reset-brick commit {force} will be allowed even when
source and destination  are identical as long as
1) destination brick is not alive
2) source and destination brick have the same brick uuid and path.
Also, the destination brick after replace-brick will use the same port
as the source brick.

Change-Id: I440b9e892ffb781ea4b8563688c3f85c7a7c89de
BUG: 1266876
Signed-off-by: Anuradha Talur 
Reviewed-on: http://review.gluster.org/12250
Smoke: Gluster Build System 
NetBSD-regression: NetBSD Build System 
CentOS-regression: Gluster Build System 
Reviewed-by: Ashish Pandey 
Reviewed-by: Pranith Kumar Karampuri

glusterd (mgmt): fix unused variable warnings/errors

2016-08-26T06:33:47+00:00

http://review.gluster.org/14085 fixes a/the "leak" - via the
generated rpc/xdr headers - of pragmas that mask these warnings.

However 14085 won't pass the smoke test until all the warnings are
fixed.

Change-Id: Ia98f2934b179145398895600a3f34dbff841c3bd
BUG: 1369124
Signed-off-by: Kaleb S. KEITHLEY 
Reviewed-on: http://review.gluster.org/15268
CentOS-regression: Gluster Build System 
NetBSD-regression: NetBSD Build System 
Smoke: Gluster Build System 
Reviewed-by: Atin Mukherjee

glusterd: clean dead initializations

2016-04-04T21:40:49+00:00

This patch cleans unused variable initialization as well as their declarations
which are no where used in the code

Change-Id: I784165fc6e91297758079699dd9583d5203b7793
BUG: 1253831
Signed-off-by: Prasanna Kumar Kalever 
Reviewed-on: http://review.gluster.org/11929
Tested-by: Prasanna Kumar Kalever 
Smoke: Gluster Build System 
NetBSD-regression: NetBSD Build System 
CentOS-regression: Gluster Build System 
Reviewed-by: Kaleb KEITHLEY 
Reviewed-by: Jeff Darcy

glusterd/syncop: double free of frame stack

2016-03-31T17:16:50+00:00

If rpc message from glusterd during brick op phase
fails without sending, then frame was freed from
the caller function and call back function.

Change-Id: I63cb3be30074e9a074f6895faa25b3d091f5b6a5
BUG: 1322262
Signed-off-by: Mohammed Rafi KC 
Reviewed-on: http://review.gluster.org/13854
Smoke: Gluster Build System 
NetBSD-regression: NetBSD Build System 
CentOS-regression: Gluster Build System 
Reviewed-by: Jeff Darcy

tier/glusterd: Check before starting tier daemon during volume start

2015-12-09T07:16:53+00:00

We start tier daemon when volume is started without looking into the
previous status. The problem with that if detach-tier is started
and then volume force start is actually starting tier daemon.

This is also fixes a problem where tier daemon is not starting
after detach stop.

Change-Id: I15b56a711e12f0e24f5ab123561258bd448621f7
BUG: 1286974
Signed-off-by: Mohammed Rafi KC 
Reviewed-on: http://review.gluster.org/12833
Tested-by: NetBSD Build System 
Tested-by: Gluster Build System 
Reviewed-by: Atin Mukherjee

glusterd: Change volume start into v3 framework

2015-11-25T12:10:34+00:00

As part of volume start, if the volume is of tier type
then we need to start tiering daemon also. But before
starting tier daemon all the bricks should be started.

So by changing volume start into v3 framework, we can
do tier start in post validate phase

Change-Id: If921067f4739e6b9a3239fc5717696eaf382c22a
BUG: 1284372
Signed-off-by: Mohammed Rafi KC 
Reviewed-on: http://review.gluster.org/12718
Tested-by: NetBSD Build System 
Tested-by: Gluster Build System 
Reviewed-by: Avra Sengupta 
Reviewed-by: Atin Mukherjee

glusterd/add-brick: change add-brick implementation to v3 framework

2015-10-07T09:09:36+00:00

add-brick commit first happens on local node and followed by peers.
As part of the commit of local-host glusterd will send the updated
volfiles to the clients connected to the local-host even before the
commit of peers happen. If any of the newly added brick was hosted
by any peer, that brick won't be started when client (connected to
local-host) try to send fops.

By changing to v3 framework we can send post validate ops
after commit operation that helps to send volfile fetch request only
after completing commits on all nodes.

Change-Id: Ib7312e01143326128c010c11fc2ed206f37409ad
BUG: 1263549
Signed-off-by: Mohammed Rafi KC 
Reviewed-on: http://review.gluster.org/12237
Tested-by: NetBSD Build System 
Tested-by: Gluster Build System 
Reviewed-by: Atin Mukherjee

dict: dict_set_bin() should never free the pointer on error

2015-07-24T15:09:08+00:00

dict_set_bin() is handling the pointer that it passed inconsistently.
Depending on the errors that can occur, the pointer passed to the dict
can be free'd, but there is no guarantee.

It is cleaner to have the caller free the pointer that allocated it and
dict_set_bin() returned an error. When dict_set_bin() returned success,
the given pointer will be free'd when dict_unref() calls data_destroy().

Many callers of dict_set_bin() already take care of free'ing the pointer
on error. The ones that did not, are corrected with this change too.

Change-Id: I39a4f7ebc0cae6d403baba99307d7ce408f25966
BUG: 1242280
Signed-off-by: Niels de Vos 
Reviewed-on: http://review.gluster.org/11638
Tested-by: Gluster Build System 
Reviewed-by: jiffin tony Thottan 
Reviewed-by: Raghavendra G 
Tested-by: NetBSD Build System

glusterd: Porting the left out gf_log_callingfns to new framework

2015-07-18T12:08:04+00:00

Change-Id: I1b0ad54238895475ddbacc4fffacac8dc6e887fe
BUG: 1235538
Signed-off-by: Nandaja Varma 
Reviewed-on: http://review.gluster.org/11590
Tested-by: Gluster Build System 
Tested-by: NetBSD Build System 
Reviewed-by: Atin Mukherjee

glusterd/ afr : set afr pending xattrs on replace brick

2015-06-26T05:55:52+00:00

This patch is part one change to prevent data loss
in a replicate volume on doing a replace-brick commit
force operation.

Problem: After doing replace-brick commit force, there is a
chance that self heal happens from the replaced (sink) brick
rather than the source brick leading to data loss.

Solution: During the commit phase of replace brick, after old
brick is brought down, create a temporary mount and perform
setfattr operation (on virtual xattr) indicating AFR to mark
the replaced brick as sink.

As a part of this change replace-brick command is being changed
to use mgmt_v3 framework rather than op-state-machine framework.

Many thanks to Krishnan Parthasarathi for helping me out on this.

Change-Id: If0d51b5b3cef5b34d5672d46ea12eaa9d35fd894
BUG: 1207829
Signed-off-by: Anuradha 
Reviewed-on: http://review.gluster.org/10076
Tested-by: NetBSD Build System 
Reviewed-by: Ravishankar N 
Reviewed-by: Pranith Kumar Karampuri