| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
|
|
|
|
|
| |
Additional information for source and sinks are added.
Change-Id: I1704956ff86ac3ae36744efe7499c1d1c43faeaf
BUG: 968301
Signed-off-by: Venkatesh Somyajulu <vsomyaju@redhat.com>
Reviewed-on: http://review.gluster.org/5638
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Anand Avati <avati@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Problem:
internal_lock->lk_attempted_count keeps track of the number of blocking
locks attempted. lk_expected_count keeps track of the number locks expected.
Here are the sequence of steps that happen which lead to the illution that
a full file lock is achieved, even without attempting any lock.
2 mounts are doing dd on same file. Both of them witness a brick going
down and coming back up again. Both of the mounts issue self-heal
1) Both mount-1, mount-2 attempt full file locks in self-heal domain.
lets say mount-1 got the lock, mount-2 attempts blocking lock.
2) mount-1 attempts full file lock in data domain. It goes into blocking
mode because some other writes are in progress. Eventually it gets the lock.
But this results in lk_attempted_count to be still as 2 and will not be reset.
It completes syncing the data.
3) mount-1 before unlocking final small range lock attempts full file lock in
data domain to figure out the source/sink. This will be put into blocked mode
again because some other writes are in progress. But this time seeing the
stale value of lk_attempted_count being equal to lk_expected_count, blocking_lock
phase thinks it completed locking without acquiring a single lock :-O.
4) mount-1 reads xattrs without any lock but since it does not modify the xattrs,
no harm is done by this phase. It tries to do unlocks and the unlocks will fail
because the locks are never taken in data domain. mount-1 also unlocks
self-heal domain locks.
Our beloved mount-2 now gets the chance to cause horror :-(.
5) mount-2 gets the full range blocking lock in self-heal domain.
Please note that this sets lk_attempted_count to 2.
6) mount-2 attempts full range lock in data domain, since there are still
writes on going, it switches to blocking mode. But since lk_attempted_count is 2
which is same as lk_expected_count, blocking phase locks thinks it actually got
the full range locks even though not a single lock request went out the wire.
7) mount-2 reads the change-log xattrs, which would give the number of operations
in progress (lets call this 'X'). It does the syncing and at the end of the sync
decrements the changelog by 'X'. But since that 'X' was introduced by 'X' number
of transactions that are in progress, they also decrement the changelog by 'X'.
Effectively for 'X' operations 'X' number of pre-ops are done but 2 times 'X'
number of post-ops are done resulting in -ve changelog numbers.
Fix:
Reset the lk_attempted_count and inode locks array that is used to remember locks
that are granted.
Change-Id: Ic0a79cd16f32392ea7c790511343c73592bbe6bd
BUG: 1002698
Signed-off-by: Pranith Kumar K <pkarampu@redhat.com>
Reviewed-on: http://review.gluster.org/5736
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Anand Avati <avati@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
After an upgrade the NFS3 filehandle size changed (became smaller),
but when doing a live ugprade the client would send the old handle
(expect ESTALE and do fresh lookup). But when reading the old
handle we were reading it into a structure which was limited to the
size of the new handle, while we should have been reading into a
buffer which is as big as the NFS3 spec permits the handle size to
be. The actor functions declare the structure on the stack. So the
overflow is resulting in a stack corruption.
Change-Id: Ie930875ac9db46b43d1cb8ad1e6d89cdaeded7ca
BUG: 1002385
Signed-off-by: Anand Avati <avati@redhat.com>
Reviewed-on: http://review.gluster.org/5730
Reviewed-by: Rajesh Joseph <rjoseph@redhat.com>
Reviewed-by: Niels de Vos <ndevos@redhat.com>
Tested-by: Gluster Build System <jenkins@build.gluster.com>
|
|
|
|
|
|
|
|
|
|
|
| |
Else this results in a missing frame causing a hang
Change-Id: Ib5f3dc6a3999449faa2853cee2944af2fb065a20
BUG: 1002399
Signed-off-by: Anand Avati <avati@redhat.com>
Reviewed-on: http://review.gluster.org/5731
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Vijay Bellur <vbellur@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
| |
It has been available for a while now and is probably the sane
default due to the more efficient layout and performance benefit.
BUG: 1001207
Change-Id: I6275f9741866c0afd6e685f8dc5867a86485fd20
Signed-off-by: Brian Foster <bfoster@redhat.com>
Reviewed-on: http://review.gluster.org/5624
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Anand Avati <avati@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Idea is to not leave the file in FOOL-FOOL scenario in case on
all the bricks data transaction failed with EDQUOT to avoid
increasing un-necessary load of self-heals in the system.
For directory transactions don't leave pending changelog in case
the failures are seen on all the subvolumes.
Change-Id: I38a5561d1d581a78347a76a4a509514e4a0c3fb7
BUG: 969461
Signed-off-by: Pranith Kumar K <pkarampu@redhat.com>
Reviewed-on: http://review.gluster.org/5709
Reviewed-by: Anand Avati <avati@redhat.com>
Tested-by: Gluster Build System <jenkins@build.gluster.com>
|
|
|
|
|
|
|
|
|
| |
Change-Id: I9c27b1edab111031ca8eea9cc49480ea01e39089
BUG: 1002207
Signed-off-by: Kaleb S. KEITHLEY <kkeithle@redhat.com>
Reviewed-on: http://review.gluster.org/5716
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Anand Avati <avati@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
- Enhance syncenv_new() to accept scaling parameters of syncproc.
Previously the scaling parameters were hardcoded and decided at
compile time.
- New API synctask_create() which returns the created synctask. This
is similar to synctask_new which only returned the status of whether
a synctask could be created or not.
The meaning of NULL cbk in synctask_create() means the task is
"joinable". Until synctask_join() is called on such a synctask,
the task is not reaped and resources are not destroyed. The
task would be in a zombie state after synctask_fn returns and
before synctask_join() is called.
Change-Id: I368ec9037de9510d2ba951f0aad86aaf18d9a6b6
BUG: 986775
Signed-off-by: Anand Avati <avati@redhat.com>
Reviewed-on: http://review.gluster.org/5365
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Brian Foster <bfoster@redhat.com>
|
|
|
|
|
|
|
|
|
|
| |
Change-Id: Ib0c3af6babc61dc3ed45252582876e2f243d6446
BUG: 958118
Signed-off-by: Pranith Kumar K <pkarampu@redhat.com>
Reviewed-on: http://review.gluster.org/5635
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Jeff Darcy <jdarcy@redhat.com>
Reviewed-by: Anand Avati <avati@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Introduce /var/lib/glusterfs/nfs/rmtab to contain a list of NFS-clients
which have a volume mounted. The volume option 'nfs.mount-rmtab' can be
set to an alternative filename. When the file is located on shared
storage, multiple gNFS servers can use the same file to present a single
NFS-server.
This cache is read when a system administrator calls 'showmount -a' and
updated when an NFS-client calls MNT or UMNT from the MOUNT protocol.
Usage:
- create a volume for storing the shared rmtab file
- mount the volume on all storage servers, at the same location
- make sure that the volume is mounted at boot (add to /etc/fstab)
- place the rmtab file on the volume:
# gluster volume set <VOLUME> nfs.mount-rmtab <MOUNTPOINT>/<FILENAME>
- any subsequent mount requests will add an entry to this file
- 'showmount -a' requests will return the NFS-clients using the cluster
Note:
The NFS-server does currently not support reconfigure(). When a
configuration option is set/changed, the NFS-server glusterfs process
gets restarted. This causes the active NFS-clients to be forgotten (the
entries are saved in the old rmtab, but we do not have a reference to
that file any more, so we can't re-add them). Therefor a re-mount done
by the NFS-clients is needed before they get listed in the rmtab again.
Change-Id: I58f47135d60ad112849d647bea4e1129683dd2b3
BUG: 904065
Signed-off-by: Niels de Vos <ndevos@redhat.com>
Reviewed-on: http://review.gluster.org/4430
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Harshavardhana <harsha@harshavardhana.net>
Tested-by: Harshavardhana <harsha@harshavardhana.net>
Reviewed-by: Rajesh Joseph <rjoseph@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Some xlators still require lookup() fop to be sent for proper working.
This patch remembers inodes which have been linked through readdiprlus
and makes the resolver send lookups on them.
Change-Id: Ibe8a04a659539d90dfc794521b51bf2bda017a0b
BUG: 979910
Signed-off-by: Anand Avati <avati@redhat.com>
Reviewed-on: http://review.gluster.org/5267
Reviewed-by: Amar Tumballi <amarts@redhat.com>
Tested-by: Gluster Build System <jenkins@build.gluster.com>
|
|
|
|
|
|
|
|
|
| |
Change-Id: I23b8cb7223b91a55af1cd4214f61bbe0e87351f6
BUG: 952029
Signed-off-by: Amar Tumballi <amarts@redhat.com>
Reviewed-on: http://review.gluster.org/5683
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Anand Avati <avati@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
In batch forgetting of inodes, nodeid should be checked and if it is for
root, then it should not be sent forget.
Change-Id: I99bd91ba70d8be4df88ddac005e38c449f4ed7d9
BUG: 990744
Signed-off-by: Raghavendra Bhat <raghavendra@redhat.com>
Reviewed-on: http://review.gluster.org/5471
Reviewed-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Amar Tumballi <amarts@redhat.com>
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Vijay Bellur <vbellur@redhat.com>
Reviewed-by: Anand Avati <avati@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The typecast of pointer to uint64_t *, followed by setting of
64bit in inode_ctx_get() results in memory corruption on 32bit
system.
Change-Id: I32fa3bf3b853ed2690a9b9a471099a59b9d7186a
BUG: 997902
Signed-off-by: Anand Avati <avati@redhat.com>
Reviewed-on: http://review.gluster.org/5682
Tested-by: Morten Johansen <morten@bzzt.no>
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Santosh Pradhan <spradhan@redhat.com>
Reviewed-by: Brian Foster <bfoster@redhat.com>
|
|
|
|
|
|
|
|
|
| |
BUG: 952029
Change-Id: I7405d473d369a4a951836eceda4faccbad19ce0e
Signed-off-by: Amar Tumballi <amarts@redhat.com>
Reviewed-on: http://review.gluster.org/5497
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Anand Avati <avati@redhat.com>
|
|
|
|
|
|
|
|
|
| |
Change-Id: I38d2fdc47e4b805deafca6805e54807976ffdb7e
Signed-off-by: Amar Tumballi <amarts@redhat.com>
BUG: 952029
Reviewed-on: http://review.gluster.org/5496
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Anand Avati <avati@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This reverts commit 4c0f4c8a89039b1fa1c9c015fb6f273268164c20.
Conflicts:
xlators/mount/fuse/src/fuse-bridge.c
For build issues added CREATE_MODE_KEY definition in:
libglusterfs/src/glusterfs.h
Change-Id: I8093c2a0b5349b01e1ee6206025edbdbee43055e
BUG: 952029
Signed-off-by: Amar Tumballi <amarts@redhat.com>
Reviewed-on: http://review.gluster.org/5495
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Anand Avati <avati@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The 'start' variant of the remove-brick command only applies at the dht
level wherein we can remove all the bricks of a sub-volume (and remove
multiple such sub-volumes) but not select bricks of it.
This patch disallows removing individual replica bricks of multiple
sub-volumes (i.e. reducing the replcia count of the volume) using
remove-brick 'start'. The preferred method for such an operation is to use
commit force.
This patch also reverts the check to prevent removal of bricks from a
replicate volume (commit 0d415f7)
BUG: 961669
Change-Id: I447ad27f73a0963b5e09fb317bf7267a7a5a6147
Signed-off-by: Ravishankar N <ravishankar@redhat.com>
Reviewed-on: http://review.gluster.org/5566
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Amar Tumballi <amarts@redhat.com>
Reviewed-by: Anand Avati <avati@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
xattr update will result in at least ctime change. So invalidate
attributes in xattr callback.
Change-Id: Ie6e8f2fd9a11c56c27e78bd58c2ff1e1d6edce6e
BUG: 953694
Signed-off-by: Anand Avati <avati@redhat.com>
Reviewed-on: http://review.gluster.org/5641
Reviewed-by: Amar Tumballi <amarts@redhat.com>
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Vijay Bellur <vbellur@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Upon graph switch, the basefd's flags were not saved in the new fd created
for the new graph upon which all the further requests for the open file
would come. Thus posix was treating the fd as a read-only fd and was
denying the write on the fds.
Change-Id: I781b62b376a85d1a938c091559270c3f242f1a2a
BUG: 998352
Signed-off-by: Raghavendra Bhat <raghavendra@redhat.com>
Reviewed-on: http://review.gluster.org/5601
Reviewed-by: Amar Tumballi <amarts@redhat.com>
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Anand Avati <avati@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Else things can deadlock in getspec v/s glusterd_do_mount()
Change-Id: Ie70b43916e495c1c8f93e4ed0836c2fb7b0e1f1d
BUG: 997576
Signed-off-by: Anand Avati <avati@redhat.com>
Reviewed-on: http://review.gluster.org/5636
Tested-by: Joe Julian <joe@julianfamily.org>
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Amar Tumballi <amarts@redhat.com>
Reviewed-by: Krishnan Parthasarathi <kparthas@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
A volume would fail to start if any one of the bricks fails staging or
fails to start, even with the 'force' option. With this patch, when the
'force' option is given for a volume start, glusterd will continue and
start other bricks even if one fails staging or starting.
Also did a small fix in changelog, to prevent it crashing when it fails
to init.
Change-Id: I7efbd9ab13d12d69b0335ae54143fa17586f8f98
BUG: 994375
Signed-off-by: Kaushal M <kaushal@redhat.com>
Reviewed-on: http://review.gluster.org/5510
Reviewed-by: Venky Shankar <vshankar@redhat.com>
Reviewed-by: Amar Tumballi <amarts@redhat.com>
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Vijay Bellur <vbellur@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Add server uuid as an attribute to the existing brick details in the
volume info cli xml output.
Currently, when a node has more than one ip, the oVirt-engine fails
to map the corresponding server using the ip alone.
If we get the host uuid along with brick details in volume info
command it will be easy for ovirt-engine to find out the
server and thereby we can avoid confusion in finding the server.
Change-Id: I3c9c9acea80e10e0b2977477759d9af045e48959
BUG: 955588
Signed-off-by: Timothy Asir <tjeyasin@redhat.com>
Reviewed-on: http://review.gluster.org/4875
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Vijay Bellur <vbellur@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
| |
Confusing "Error" messages in logs can cause user panic
and false positives - avoid them as necessary in future.
Change-Id: I906c64eea879b19a8db099c89d1d7f874e5530db
BUG: 995784
Signed-off-by: Harshavardhana <harsha@harshavardhana.net>
Reviewed-on: http://review.gluster.org/5555
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Vijay Bellur <vbellur@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Currently, we sent GF_READDIR_SKIP_DIRS for all subvolumes if
first_subvol != first_up_subvolume.
Also first_up_subvolume can change with-in the life of a call and
cbk. Saving the first_up_subvol in dht_local for checks.
Change-Id: I6e369e63f29c9761993f2a66ed768c424bb44d27
BUG: 996474
Signed-off-by: shishir gowda <sgowda@redhat.com>
Reviewed-on: http://review.gluster.org/5577
Reviewed-by: Pranith Kumar Karampuri <pkarampu@redhat.com>
Reviewed-by: Vijay Bellur <vbellur@redhat.com>
Tested-by: Gluster Build System <jenkins@build.gluster.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
For Write Once Read Many times type of work-load choosing largest
file to be the source will always resolve fool-fool
scenarios correctly. In other cases we fsync() the files and
will have a reliable 'wise man'.
Change-Id: Ic4dbea8d06db6d578fbcb866fb65ee2d066ac7ba
BUG: 958118
Signed-off-by: Anand Avati <avati@redhat.com>
Reviewed-on: http://review.gluster.org/5519
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Pranith Kumar Karampuri <pkarampu@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
we find fd marked bad while trying to fulfill lies.
* flush was queued behind some unfulfilled write.
* A previously wound write returned an error and hence fd was marked
bad with corresponding error.
* wb_fulfill_head (invocation probably rooted in wb_flush), before
winding checks for failures of previous writes and since there was a
failure, calls wb_head_done without even winding one request in head.
* wb_head_done unrefs all the requests in list "head".
* since flush was last operation on fd (and most likely last operation
on inode itself), no one invokes wb_process_queue and flush is stuck
in request queue for eternity.
Change-Id: I3b5b114a1c401d477dd7ff64fb6119b43fda2d18
BUG: 988642
Signed-off-by: Raghavendra G <rgowdapp@redhat.com>
Reviewed-on: http://review.gluster.org/5398
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Anand Avati <avati@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Durability of appending writes is implicit in the file size. Therefore
performing an explicit fsync() is unnecessary in such cases as self-heal
can check for the size of file when pending changelog is not unambiguous.
Change-Id: I05446180a91d20e0dbee5de5a7085b87d57f178a
BUG: 927146
Signed-off-by: Anand Avati <avati@redhat.com>
Reviewed-on: http://review.gluster.org/5501
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Pranith Kumar Karampuri <pkarampu@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
| |
Also fixes for failing testcase `./tests/bugs/bug-888174.t`,
which has been failing sporadically for many patches.
Change-Id: Ic7d2c95da5d3126623cec403207afadd449bf950
BUG: 927146
Signed-off-by: Harshavardhana <harsha@harshavardhana.net>
Reviewed-on: http://review.gluster.org/5620
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Anand Avati <avati@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Currently, remove-brick supports removal of only one distributed
stripe/ replica pair at a time. Fix it to support removal of multiple
pairs. This is consistent with add-brick behaviour which supports adding
multiple stripe/replica pairs simultaneously.
Removal is successful irrespective of the order of the bricks given at
the CLI, as long as the bricks are from the same subvolume(s).
Change-Id: I7c11c1235ce07b124155978b9d48d0ea65396103
BUG: 974007
Signed-off-by: Ravishankar N <ravishankar@redhat.com>
Reviewed-on: http://review.gluster.org/5210
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Krishnan Parthasarathi <kparthas@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
at syslog side, log message is identified by its properties like
programname, pid, etc. brick/mount processes need to be identified
uniquely as they are different process of gluterfsd/glusterfs. At
rsyslog side, log separated by programname/app-name with pid works but
bit hard to identify them in long run which process is for what
brick/mount.
This patch fixes by setting identity string at openlog() which sets
programname/app-name as similar to old style log file prefixed by
gluster, glusterd, glusterfs or glusterfsd
Change-Id: Ia05068943fa67ae1663aaded1444cf84ea648db8
BUG: 928648
Signed-off-by: Bala.FA <barumuga@redhat.com>
Reviewed-on: http://review.gluster.org/5541
Reviewed-by: Vijay Bellur <vbellur@redhat.com>
Tested-by: Gluster Build System <jenkins@build.gluster.com>
|
|
|
|
|
|
|
|
|
|
|
|
| |
When user has explicitly configured to disable entry self-heal in the
client, it is wrong to do the healing in opendir. So skip it. This
is especially useful to reduce opendir() times after graph switches.
Change-Id: Ic6eb9ff2334a5b8417f2f35410a366a536bad5df
Signed-off-by: Anand Avati <avati@redhat.com>
Reviewed-on: http://review.gluster.org/5528
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Vijay Bellur <vbellur@redhat.com>
|
|
|
|
|
|
|
|
|
| |
Change-Id: I5701bf115e0aa1adb4fb52f5418534910a2268d4
BUG: 994959
Signed-off-by: Pranith Kumar K <pkarampu@redhat.com>
Reviewed-on: http://review.gluster.org/5531
Reviewed-by: Vijay Bellur <vbellur@redhat.com>
Tested-by: Gluster Build System <jenkins@build.gluster.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
When one of the bricks of a 1x2 replicate volume is down,
writes to the volume is causing a race between afr_flush_wrapper() and
afr_flush_cbk(). The latter frees up the call_frame's local variables
in the unwind, while the former accesses them in the for loop and
sending a stack wind the second time. This causes the FUSE mount process
(glusterfs) toa receive a SIGSEGV when the corresponding unwind is hit.
This patch adds the call_count check which was removed when
afr_flush_wrapper() was introduced in commit 29619b4e
Change-Id: I87d12ef39ea61cc4c8244c7f895b7492b90a7042
BUG: 988182
Signed-off-by: Ravishankar N <ravishankar@redhat.com>
Reviewed-on: http://review.gluster.org/5393
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Pranith Kumar Karampuri <pkarampu@redhat.com>
Reviewed-by: Anand Avati <avati@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
| |
Bad condition check, fix it!
Change-Id: I6e047de70f77d7b98b2ca771a467f14a76fd62fe
BUG: 994392
Signed-off-by: Anand Avati <avati@redhat.com>
Reviewed-on: http://review.gluster.org/5513
Reviewed-by: Amar Tumballi <amarts@redhat.com>
Tested-by: Amar Tumballi <amarts@redhat.com>
Tested-by: Gluster Build System <jenkins@build.gluster.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Problem:
In 3.3, inode locks of both metadata and data are competing in same
domain called data domain (old style). This coupled with eager-lock,
delayed post-ops introduce delays for metadata operations like chmod,
chown etc. To avoid this problem, inode locks for metadata ops are
moved to different domain called metadata domain in 3.4 (new style).
But when both 3.3 clients and 3.4 clients are present, 3.4 clients
for metadata operations still need to take locks in "old style" so
that proper synchronization happens across 3.3 and 3.4 clients. Only
when all clients are >= 3.4 locks will be taken in "new style" for
metadata locks. Because of this behavior as long as at least one 3.3
client is present, delays will be perceived for doing metadata
operations on all 3.4 clients while data operations are in
progress (Ex: Untar will untar one file per sec).
Fix:
Make locks xlators translate old-style metadata locks to new-style
metadata locks. Since upgrade process suggests upgrading servers
first and then clients, this approach gives good results.
Tests:
1) Tested that old style metadata locks are converted to new style by
locks xlator using gdb
2) Tested that disconnects purge locks in meta-data domain as well
using gdb and statedumps.
3) Tested that untar performance is not hampered by meta-data and
data operations.
4) Had two mounts one with orthogonal-meta-data on and other with
orthogonal-meta-data off ran chmod 777 <file> on one mount and
chmod 555 <file> on the other mount in while loops when I took
statedumps I saw that both the transports are taking lock on
same domain with same range.
18:49:30 :) ⚡ sudo grep -B1 "ACTIVE" /usr/local/var/run/gluster/home-gfs-r2_0.324.dump.*
home-gfs-r2_0.324.dump.1375794971-lock-dump.domain.domain=r2-replicate-0:metadata
home-gfs-r2_0.324.dump.1375794971:inodelk.inodelk[0](ACTIVE)=type=WRITE, whence=0, start=9223372036854775806, len=0, pid = 7525, owner=78f9e652497f0000, transport=0x15ac9e0, , granted at Tue Aug 6 18:46:11 2013
home-gfs-r2_0.324.dump.1375795051-lock-dump.domain.domain=r2-replicate-0:metadata
home-gfs-r2_0.324.dump.1375795051:inodelk.inodelk[0](ACTIVE)=type=WRITE, whence=0, start=9223372036854775806, len=0, pid = 8879, owner=0019cc3cad7f0000, transport=0x158f580, , granted at Tue Aug 6 18:47:31 2013
Change-Id: I268df4efd93a377a0c73fbc59b739ef12a7a8bb6
BUG: 993981
Signed-off-by: Pranith Kumar K <pkarampu@redhat.com>
Reviewed-on: http://review.gluster.org/5503
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Anand Avati <avati@redhat.com>
|
|
|
|
|
|
|
|
|
|
| |
Change-Id: I4352f513fc5616daa20e9a4ad51a63fb13a27dff
BUG: 847839
Signed-off-by: M S Vishwanath Bhat <vbhat@redhat.com>
Reviewed-on: http://review.gluster.org/5472
Reviewed-by: Amar Tumballi <amarts@redhat.com>
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Vijay Bellur <vbellur@redhat.com>
|
|
|
|
|
|
|
|
|
| |
Change-Id: Ic3c43291e0e1ead0d89c0436e8d70aa5dee2f543
BUG: 924488
Signed-off-by: Harshavardhana <harsha@harshavardhana.net>
Reviewed-on: http://review.gluster.org/5391
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Anand Avati <avati@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The fuse_getxattr() function was not freeing fuse_state_t resulting in a
memory leak. As a result, when continuous writes (run dd command in a loop)
were done from a FUSE mount point, the OOM killer killed the client
process (glusterfs).
Change-Id: I6ded1a4c25d26ceab0cb3b89ac81066cb51343ec
BUG: 988182
Signed-off-by: Ravishankar N <ravishankar@redhat.com>
Reviewed-on: http://review.gluster.org/5392
Reviewed-by: Pranith Kumar Karampuri <pkarampu@redhat.com>
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Anand Avati <avati@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Asynchronous tasks are shown in 'volume status' only for a normal volume
status request for either all volumes or a single volume.
Change-Id: I9d47101511776a179d213598782ca0bbdf32b8c2
BUG: 888752
Signed-off-by: Kaushal M <kaushal@redhat.com>
Reviewed-on: http://review.gluster.org/5308
Reviewed-by: Amar Tumballi <amarts@redhat.com>
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Anand Avati <avati@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
7.17
- Distinguishes between POSIX and BSD locking support via a
separate BSD locking support init flag. Older protocol versions
(since BSD support was added) export both types of locking
requests if FUSE_POSIX_LOCKS is specified. Gluster sets this
flag, so set FUSE_FLOCK_LOCKS as well on kernels that support
version 17 or newer.
7.18
- Adds ioctl() support for directories (and the associated
FUSE_IOCTL_DIR flag). Gluster does not support the ioctl
request, so no changes are required. Update the header.
- Adds support for the delete notification to allow a filesystem
to inform the kernel of a deleted inode. No gluster changes
required.
7.19
- Adds support for the fallocate request. Gluster already supports
fallocate and includes the request opcode definition and data
structure. Update the header version number.
7.20
- Adds the FUSE_AUTO_INVAL_DATA init flag to enable attribute
updates on reads and automatic cache invalidation on mtime
changes. Behavior does not change unless the init flag is
specified, no gluster changes required. Update header.
7.21
- Adds readdirplus support and updates the poll request to include
events. Gluster already supports readdirplus and includes the
relevant data structures. Poll is not supported, so no changes
are required. Update the header with some missing
READDIRPLUS_AUTO bits.
7.22
- Adds real asynchronous direct I/O support. Gluster already
supports/enables the associated bit (FUSE_ASYNC_DIO), no further
changes are required. Update the header.
BUG: 990744
Change-Id: Idf6fd75bbd48189587e548f7624626f9a75309e8
Signed-off-by: Brian Foster <bfoster@redhat.com>
Reviewed-on: http://review.gluster.org/5489
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Anand Avati <avati@redhat.com>
|
|
|
|
|
|
|
|
|
| |
Change-Id: Ia8d4bed7ccd316a83c397b53b9c1b1806024f83e
BUG: 991622
Signed-off-by: Pranith Kumar K <pkarampu@redhat.com>
Reviewed-on: http://review.gluster.org/5493
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Anand Avati <avati@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Instead of using the cluster op-version, volume op-version is used to
enable open-behind during volgen. For doing this, the volume op-versions
are updated before regenerating the volfiles.
Change-Id: I675bb549bf7c7c0279030dca698fb530781addc6
BUG: 990830
Signed-off-by: Kaushal M <kaushal@redhat.com>
Reviewed-on: http://review.gluster.org/5385
Reviewed-by: Krutika Dhananjay <kdhananj@redhat.com>
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Vijay Bellur <vbellur@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Lets say mount1 has eager-lock(full-lock) and after the eager-lock
is taken mount2 opened the same file, it won't be able to
perform any data operations until mount1 releases eager-lock.
To avoid such scenario do not enable eager-lock for transaction
if open-fd-count is > 1. Delaying of changelog piggybacking is
avoided in this situation.
Change-Id: I51b45d6a7c216a78860aff0265a0b8dabc6423a5
BUG: 910217
Signed-off-by: Pranith Kumar K <pkarampu@redhat.com>
Reviewed-on: http://review.gluster.org/5432
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: venkatesh somyajulu <vsomyaju@redhat.com>
Reviewed-by: Vijay Bellur <vbellur@redhat.com>
|
|
|
|
|
|
|
|
|
|
| |
Change-Id: I86bdf865730416150c10617dcbad5c037579acde
BUG: 910217
Signed-off-by: Pranith Kumar K <pkarampu@redhat.com>
Reviewed-on: http://review.gluster.org/5433
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Raghavendra Bhat <raghavendra@redhat.com>
Reviewed-by: Vijay Bellur <vbellur@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
do_iot_schedule() enqueues the stub and kicks the worker thread.
The stub is eventually destroyed after it has been resumed and thus
unsafe to access after being enqueued.
Though likely difficult to reproduce in a real deployment, a crash
is reproducible by running a smallfile benchmark on a replica 2
volume on a single vm. Reorder the debug log message prior to the
do_iot_schedule() call to avoid the crash.
BUG: 989579
Change-Id: Ifc6502c02ae455c959a90ff1ca62a690e31ceafb
Signed-off-by: Brian Foster <bfoster@redhat.com>
Reviewed-on: http://review.gluster.org/5418
Reviewed-by: Santosh Pradhan <spradhan@redhat.com>
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Anand Avati <avati@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
7.14
- Splice write support to fuse device node. No gluster changes
required besides header update.
7.15
- Store/retrieve notification support. No gluster changes
required besides header update.
7.16
- BATCH_FORGET request support. Implement a handler for
BATCH_FORGET requests and update the header.
- Updated ioctl() ABI. No gluster changes required besides
header update.
BUG: 990744
Change-Id: If3061a720ba566ee6731ad8b77cdc665d8fbf781
Signed-off-by: Brian Foster <bfoster@redhat.com>
Reviewed-on: http://review.gluster.org/5449
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Anand Avati <avati@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Bug 990887: During a blocking NLM call NFS server is crashing.
Cause: When nlm4_establish_callback function is called from nlm4svc_send_granted the cs->req->trans
pointer is NULL. Thus using this pointer will result in a crash. Whereas cs->trans points to a
valid transport object. NLM should use cs->trans instead of cs->req->trans.
Fix: Replaced cs->req->trans with cs->trans.
Change-Id: I425e48e0aafc9a6c130912edf2e801d8c4c9472d
BUG: 990887
Signed-off-by: Rajesh Joseph <rjoseph@redhat.com>
Reviewed-on: http://review.gluster.org/5452
Reviewed-by: Amar Tumballi <amarts@redhat.com>
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Vijay Bellur <vbellur@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Because of posix default_acl on parent directory, the mode
of linkfile can get masked with the mode in the default acl.
This breaks DHT integrity. So let the mode get explicitly reset
after mknod().
Change-Id: Ia7328e1ee7b4430bda308f9da293dba78405e081
BUG: 990410
Signed-off-by: Anand Avati <avati@redhat.com>
Reviewed-on: http://review.gluster.org/5440
Reviewed-by: Amar Tumballi <amarts@redhat.com>
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Raghavendra Talur <rtalur@redhat.com>
|
|
|
|
|
|
|
|
|
| |
Change-Id: I42d08b3a6a7a3839f5e9953e1f83959222c080f8
Signed-off-by: shishir gowda <sgowda@redhat.com>
BUG: 989846
Reviewed-on: http://review.gluster.org/5446
Reviewed-by: Vijay Bellur <vbellur@redhat.com>
Tested-by: Gluster Build System <jenkins@build.gluster.com>
|