IF YOU WOULD LIKE TO GET AN ACCOUNT, please write an
email to Administrator. User accounts are meant only to access repo
and report issues and/or generate pull requests.
This is a purpose-specific Git hosting for
BaseALT
projects. Thank you for your understanding!
Только зарегистрированные пользователи имеют доступ к сервису!
Для получения аккаунта, обратитесь к администратору.
make the build more silent
Change-Id: Ied4c87c40af4b1f1034af4a13c1153d82883751b
BUG: 3255
Reviewed-on: http://review.gluster.com/171
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Anand Avati <avati@gluster.com>
Mounting from localhost in fstab fails at boot on ubuntu when glusterd is started by debian initscript.
A new directory, extras/Ubuntu, contains the new upstart job (glusterd.conf) and a brief README.Ubuntu file.
This makes the upstart job available to package maintainers like the SysV initscripts are in extras/init.d.
This has been tested on Ubuntu 10.10 Maverick.
Change-Id: I6ccd06054e920cd6dc45dba4eb8ab89aa30f8b47
BUG: 3282
Reviewed-on: http://review.gluster.com/126
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Anand Avati <avati@gluster.com>
also print a warning message if dictionary is NULL, while sending a
mkdir request in distribute self-heal.
Change-Id: Ib9cac6ed1635203802f089986f8acb1ce416265d
BUG: 3215
Reviewed-on: http://review.gluster.com/136
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Shishir Gowda <shishirng@gluster.com>
due to the race, there was a possibility of having two (or more) threads doing
glusterd_defrag_start(), which would cause different thread to access same
pointer, and at some point making it NULL, causing other threads to crash with
SEGV.
Change-Id: Id05b99dd6f33329027b8a07f1c8da5a65fd6dae9
BUG: 3295
Reviewed-on: http://review.gluster.com/138
Reviewed-by: Anand Avati <avati@gluster.com>
Tested-by: Gluster Build System <jenkins@build.gluster.com>
This is done, so that there is no gfid mismatch. Unlink the older
linkfile if it exists, and recreate it with the correct gfid.
Also removed unused rename related code.
Change-Id: Ice0cdff51db786dafb1a019f7642c7d000e8c5f3
BUG: 2522
Reviewed-on: http://review.gluster.com/142
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Anand Avati <avati@gluster.com>
By using variable iobufs, xfer data size is no more limited to
128K (default). This helps in scaling.
Change-Id: Iab453db9223d887306d150cd6fe0b1eae9c422cc
BUG: 2472
Reviewed-on: http://review.gluster.com/13
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Amar Tumballi <amar@gluster.com>
Reviewed-by: Anand Avati <avati@gluster.com>
- exceptions raised by us will be logged as single-line error messages
(full stack strace is shown only at DEBUG loglevel)
- common/well understood exceptions are mapped to "user-parsable" error logs
Change-Id: I75f1fb848483372364b2093878d9cfed576c9739
BUG: 2778
Reviewed-on: http://review.gluster.com/125
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Anand Avati <avati@gluster.com>
Removed "volume top successful" message.
If top fails,"volume top unsuccessful" is displayed.
If errstr is set, it is displayed.
"No entries in list" messages deleted.
Change-Id: I8ed35ea7db365b7af6b10833672fcab0e4e8751d
BUG: 2722
Reviewed-on: http://review.gluster.com/121
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Shishir Gowda <shishirng@gluster.com>
- there was a possibility of double unwind in case of errors.
- use a new frame to do open in fd-based fops.
In case of errors, qr_resume_pending_ops will be called to resume all the
fops waiting on open. Hence if we use frame corresponding to fop (without
creating a new one), there is a possibility of frame being freed by the
time open would've returned to quick-read.
Change-Id: Ie4cc19907f9d6362860bdb984779c8f4cf822332
BUG: 3168
Signed-off-by: Raghavendra G <raghavendra@gluster.com>
Reviewed-on: http://review.gluster.com/34
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Anand Avati <avati@gluster.com>
New set option performance.client-io-threads added to enable/disable it.
Default behavior is io-threads is off on client side
Loaded only on fuse volume below io-stats.
Change-Id: I13aea9a564d5270901c3c356fa6bede836c6a66d
BUG: 3122
Reviewed-on: http://review.gluster.com/105
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Amar Tumballi <amar@gluster.com>
Reviewed-by: Vijay Bellur <vijay@gluster.com>
this is required because if 'CONNECT' event comes before the clnt_start()
function ends, we may loose the event if it is not initialized.
Thanks to Jeff Darcy <jdarcy@redhat.com> for pointing it out
Change-Id: Iebbd651b7563265800f41922367ca133fbbf4fc3
BUG: 3207
Reviewed-on: http://review.gluster.com/77
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Anand Avati <avati@gluster.com>
Scenario - The race window exists when before we wind to a stat call
in readdirp_cbk, whereas stats have returned to cbk. We would free up
local, which leads to a crash.
Fix - We assume all entries to be regular files, and set the wind_count
as a multiple of entries * subvol (stat calls). When we encounter a non
regular file, we decrement the windcount accordingly (subvol times). If
the last entry happens to be a non-regular file, and all stat calls have
wound back, we unwind from here itself.
Change-Id: Ia8e225ef557cbe1fedb9b72de9bd9fa61f371a24
BUG: 3257
Reviewed-on: http://review.gluster.com/115
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Amar Tumballi <amar@gluster.com>
Make sure the inode_ctx is deleted
nfs_release is not needed, as __nfs3_fdcache_remove_entry cleans up
the ctx
Change-Id: I690efa1eb1d4cd39cee258bb29692cc58ab9c380
BUG: 3250
Reviewed-on: http://review.gluster.com/103
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Shehjar Tikoo <shehjart@gluster.com>
without the proper error code, rpc won't know if transport is successfully
loaded or not.
Change-Id: Idc3ad56d2f47cf1e81ba39f90a7f7b47305d7d9a
BUG: 3260
Reviewed-on: http://review.gluster.com/112
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Anand Avati <avati@gluster.com>
This change ensures that glusterd retains 'state' information
of an ongoing replace brick operation even if it went down midway.
Change-Id: I697c2f27a96b3200c7865bd1d3a1de9692ce0da1
BUG: 3252
Reviewed-on: http://review.gluster.com/6
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Vijay Bellur <vijay@gluster.com>
in case of layout 'creation', layout->err == ENOSPC should be ignored
where as, while layout 'fixing' we should consider what was already
present in the layout.
Change-Id: Ifb613b41065813c9f1202e65e94b4b0282766d11
BUG: 2258
Signed-off-by: Amar Tumballi <amar@gluster.com>
Reviewed-on: http://review.gluster.com/15
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Shishir Gowda <shishirng@gluster.com>
An option, transport.socket.read-fail-log was added in glusterd.
This can also be added to any translator which uses socket.c.
A gf_boolean_t flag(read_fail_log) is added in socket_private_t.
Using this, logging of socket read failures can be controlled.
The options is set to 'off' in glusterd.vol by default.
Change-Id: I85cf4afc1f534f5f51018449d5d84baef18fce23
BUG: 3156
Reviewed-on: http://review.gluster.com/22
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Anand Avati <avati@gluster.com>
In the pre-GFID era, the linkfile of the destination file could be reused
the linkfile for the renamed file when dst_cached == src_cached.
This patch handles this situation and reverts the previous (wrong) fix.
Change-Id: Iba57b5eb91cf8b1fb40e74f6399cdf99b8b00410
BUG: 2464
Reviewed-on: http://review.gluster.com/90
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Amar Tumballi <amar@gluster.com>
This is required if the number of connections from the glusterd
increases.
Change-Id: Iff78773f593a8bfae13ec9eec52c5a67ea936ef2
BUG: 3142
Signed-off-by: shishir gowda <shishirng@gluster.com>
Reviewed-on: http://review.gluster.com/11
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Pavan T C <tcp@gluster.com>
Replace-brick didn't support transport types other than tcp.
Test case: replace-brick should work with tcp and rdma.
Change-Id: If33f108b70247d81791d25a1a05247f49847b4d7
BUG: 3092
Reviewed-on: http://review.gluster.com/72
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Krishnan Parthasarathi <kp@gluster.com>
With DRY_RUN=1, rfc.sh will not perform the git push just display
the push command. This can be useful if one wants to give a final
review for the submitted change list (check for typos, misplaced bug
id-s) or if one is offline.
Change-Id: I90e2016ba78e2a0ecf37044b3db0277a25914f77
BUG: 3234
Reviewed-on: http://review.gluster.com/87
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Anand Avati <avati@gluster.com>
In rebase_changes():
- Using "--all" for git fetch is bogus: the scope of this quantification is not
over the tracked refs of the upsteam repo (which we do want, but that kind
of universal quantification is already implied by the notion of "tracking"),
but over the set of all the registered remotes, which is an annoying side effect.
- $EDITOR is weaker than the git configuration settings. For the purpose of the
script, we need to override any kind of setting, and that can be done
via $GIT_EDITOR.
Cf. git-fetch(1), git-commit(1).
Change-Id: Ia91ee5c4d42dce9248f26f7697528e3c97355871
BUG: 3234
Reviewed-on: http://review.gluster.com/86
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Anand Avati <avati@gluster.com>
create_fuse_mount() now checks if it is a client process before performing
mount. If not a client process, it returns an error.
Added changes suggested
Change-Id: If8548ad60d345bb8cd09686ceed0d83522799761
BUG: 2469
Reviewed-on: http://review.gluster.com/5
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Anand Avati <avati@gluster.com>
Profile - If all bricks are down, report errror.
Top - If all bricks or the specified brick is down, report error.
Change-Id: I39d4acd631b481233ab2e9f66c8f4160caef5762
BUG: 2600
Reviewed-on: http://review.gluster.com/63
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Anand Avati <avati@gluster.com>
so the symlinks pointing to directories are not considered
as directories themself and the control flows outside
gluster's scope
Change-Id: Iae910ce6c68886d34ae6e5efe46062481b40cd25
BUG: 3191
Reviewed-on: http://review.gluster.com/30
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Anand Avati <avati@gluster.com>
Units are now shown for the latencies and duration
Changed some text to be more meaningful
Change-Id: I1891d6ac541421606dda30ed10f9b40b21e2fa52
BUG: 3028
Reviewed-on: http://review.gluster.com/18
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Vikas Gorur <vikas.gluster@gmail.com>
While peforming replace-brick, delete-brick, remove-brick... and related
operations resolve extra slashes as single path separator. i.e treat /foo//bar//
as /foo/bar.
Signed-off-by: Sachidananda Urs <sac@gluster.com>
Signed-off-by: Anand Avati <avati@gluster.com>
BUG: 3183 (When creating volumes brick paths are not handled properly.)
URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=3183
