Whamcloud - gitweb
Yu Jian [Fri, 13 May 2011 04:50:54 +0000 (12:50 +0800)]
LU-296 Add missing code from auster/yaml port to b1_8
Code related to defining and creating the $TF_FAIL file was missed in
the port of auster and yaml to b1_8. This patch adds the missing
code. Namely:
- Define TF_FAIL in acceptance-small.sh
- add 'touch $TF_FAIL' to the test-framework.sh ERR signal trap
Additionally, the file test for $TF_FAIL variable should be quoted in
acceptance-small.sh and auster.sh. Otherwise, if this variable is not
defined, bash will spit out something like
acceptance-small.sh: line 123: [: too many arguments
Bash variables should generally be quoted whenever they are referenced.
(see http://tldp.org/LDP/abs/html/quotingvar.html).
Signed-off-by: Chris Horn <hornc@cray.com>
Signed-off-by: Yu Jian <yujian@whamcloud.com>
Change-Id: I2071c859d00b8cd8bd423bf9f626b44352456191
Reviewed-on: http://review.whamcloud.com/516
Reviewed-by: Johann Lombardi <johann@whamcloud.com>
Tested-by: Hudson
Bobi Jam [Tue, 10 May 2011 02:19:07 +0000 (10:19 +0800)]
LU-240 test_171() races with background multiop
Change-Id: Ibdcc584218708bd29724f3f71804a3ebf5b1d454
Signed-off-by: Bobi Jam <bobijam@whamcloud.com>
Reviewed-on: http://review.whamcloud.com/520
Tested-by: Hudson
Reviewed-by: Johann Lombardi <johann@whamcloud.com>
yangsheng [Thu, 12 May 2011 06:26:57 +0000 (14:26 +0800)]
LU-228 kernel update to rhel5 2.6.18-238.9.1
Clear up the kernel support information.
Signed-off-by: Yang Sheng <ys@whamcloud.com>
Change-Id: I6b4364eb8d22629a1caee202e4b77c7f197b5b01
Reviewed-on: http://review.whamcloud.com/539
Tested-by: Hudson
Reviewed-by: Johann Lombardi <johann@whamcloud.com>
Lai Siyao [Tue, 10 May 2011 08:50:12 +0000 (01:50 -0700)]
LU-231 sanity_test_170: FAIL: expected 248 bad lines, but got 249
if the length of malformed line is less than sizeof(struct ptldebug_header),
it's not counted into bad lines if this line is at the end of log file.
Signed-off-by: Lai Siyao <laisiyao@whamcloud.com>
Change-Id: Ib2cb1a81f2b8d00c855f11d232f7643e66cebeb0
Reviewed-on: http://review.whamcloud.com/526
Tested-by: Hudson
Reviewed-by: Johann Lombardi <johann@whamcloud.com>
Niu Yawei [Fri, 6 May 2011 16:11:24 +0000 (09:11 -0700)]
LU-274 Update LVB from disk when glimpse callback return error
Client ll_glimpse_callback() could fail to get inode if the inode is
already been cleared, and this glimpse callback will fail for
-ELDLM_NO_LOCK_DATA, so server should update LVB from disk
(in filter_intent_policy()) when it received such error from client.
Issue:
Signed-off-by: Niu Yawei <niu@whamcloud.com>
Change-Id: I29b2fe4b35114359b0ad62106de48f5f2f696636
Reviewed-on: http://review.whamcloud.com/507
Tested-by: Hudson
Reviewed-by: Johann Lombardi <johann@whamcloud.com>
yangsheng [Fri, 22 Apr 2011 10:35:31 +0000 (18:35 +0800)]
LU-228 kernel update to rhel5.6.
build ok on 2.6.18-238.9.1.el5.
Signed-off-by: Yang Sheng <ys@whamcloud.com>
Change-Id: Id66c5fc89b233ef59e851b78ca52f4ea2159c9d8
Reviewed-on: http://review.whamcloud.com/455
Tested-by: Hudson
Reviewed-by: Brian J. Murrell <brian@whamcloud.com>
Reviewed-by: Michael MacDonald <mjmac@whamcloud.com>
Reviewed-by: Johann Lombardi <johann@whamcloud.com>
Yu Jian [Tue, 26 Apr 2011 10:47:09 +0000 (18:47 +0800)]
LU-232 do not use bash variable to pass test setup state between sub-tests
Port the patch from bug 22842 attachment 30538 to b1_8 branch:
1) fix sanity tests 42, 101 and 102
2) fix PATH variable to search utilities from $LUSTRE/tests dir first
Signed-off-by: Yu Jian <yujian@whamcloud.com>
Change-Id: Idb58789efc1a1a0a50b54db5cb1f6bad5b5f8daa
Reviewed-on: http://review.whamcloud.com/467
Tested-by: Hudson
Reviewed-by: Chris Gearing <chris@whamcloud.com>
Reviewed-by: Brian J. Murrell <brian@whamcloud.com>
Reviewed-by: Johann Lombardi <johann@whamcloud.com>
Johann Lombardi [Thu, 28 Apr 2011 18:44:59 +0000 (20:44 +0200)]
LU-118 Workaround for race causing assertion failure in clear_inode()
This patch is just a *workaround* to avoid a race found in recent
kernel causing the following assertion in clear_inode():
BUG_ON(inode->i_data.nrpages).
The race can still occur when truncate_inode_pages() & clear_inode()
are called directly by the VFS, but the problem should be harder to
reproduce.
A more complete fix is considered for 2.x.
Change-Id: If0df4b1cba00de7cf43f1b3510d54d35525c5e14
Signed-off-by: Johann Lombardi <johann@whamcloud.com>
Reviewed-on: http://review.whamcloud.com/475
Tested-by: Hudson
Reviewed-by: Jinshan Xiong <jay@whamcloud.com>
Reviewed-by: Andrew Perepechko <andrew_perepechko@xyratex.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>
Yu Jian [Tue, 3 May 2011 12:35:49 +0000 (20:35 +0800)]
LU-262 port the changes of llverdev.c and llverfs.c from master to b1_8
The main changes are:
bug 24017 make llverfs more robust against errors during testing
bug 20492 allow llverdev/llverfs to handle media errors better
Signed-off-by: Yu Jian <yujian@whamcloud.com>
Change-Id: Ib7df0c98304d8fa56f3ad69127617147c2987b8d
Reviewed-on: http://review.whamcloud.com/487
Tested-by: Hudson
Reviewed-by: Johann Lombardi <johann@whamcloud.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Yu Jian [Thu, 28 Apr 2011 12:56:20 +0000 (20:56 +0800)]
LU-235 fix sanity test 36g to work with remote OSS
Signed-off-by: Yu Jian <yujian@whamcloud.com>
Change-Id: Ia4e0b01d002783bc8d4f38e7c1e6b1ac753144d6
Reviewed-on: http://review.whamcloud.com/473
Tested-by: Hudson
Reviewed-by: Chris Gearing <chris@whamcloud.com>
Reviewed-by: Brian J. Murrell <brian@whamcloud.com>
Reviewed-by: Johann Lombardi <johann@whamcloud.com>
Wang Di [Tue, 19 Apr 2011 20:21:37 +0000 (13:21 -0700)]
LU-15 slow IO with read-intense application
Align the readahead extent by 1M after when it is trimed by ra_max_pages.
signed-off-by: Wang Di <di.wang@whamcloud.com>
Change-Id: I0359a39cef678b05e48617910081e9f27826ed22
Reviewed-on: http://review.whamcloud.com/437
Reviewed-by: Oleg Drokin <green@whamcloud.com>
Reviewed-by: Jinshan Xiong <jay@whamcloud.com>
Tested-by: Hudson
Reviewed-by: Johann Lombardi <johann@whamcloud.com>
Niu Yawei [Fri, 22 Apr 2011 03:20:29 +0000 (20:20 -0700)]
LU-202 fix the test_23 of sanity-quota.sh
test_23 uses a constant value(120000 = ~120M) to check if the system has
enough space to finish the test, so in the large OSTCOUNT testing
environment, the test will fail for short of disk space.
Changes of this patch are:
- Use a reasonable value to check free space instead of magic 120000;
- Specify the stripe index of test file for test_9/10 and test_23;
- Check free space over all OSTs, and select the ost which has enough
space to do test. (test_9/10/23)
Issue:
Signed-off-by: Niu Yawei <niu@whamcloud.com>
Change-Id: I2d9059d23cf53fee0244a226622b0e9adb800599
Reviewed-on: http://review.whamcloud.com/454
Tested-by: Hudson
Reviewed-by: Johann Lombardi <johann@whamcloud.com>
Yu Jian [Thu, 28 Apr 2011 06:25:09 +0000 (14:25 +0800)]
LU-230 fix yml_entities() to create MDS entity
There is only one MDS on b1_8.
Signed-off-by: Yu Jian <yujian@whamcloud.com>
Change-Id: I4ab3f0ad27ab76dd77ef9cc153b932b315ea9c4d
Reviewed-on: http://review.whamcloud.com/464
Reviewed-by: Bobi Jam <bobijam@whamcloud.com>
Tested-by: Hudson
Reviewed-by: Chris Gearing <chris@whamcloud.com>
Reviewed-by: Johann Lombardi <johann@whamcloud.com>
Yu Jian [Mon, 25 Apr 2011 14:38:57 +0000 (22:38 +0800)]
LU-203 pass client mount options to zconf_mount_clients() correctly
Quote the third argument and pass it to zconf_mount_clients().
Signed-off-by: Yu Jian <yujian@whamcloud.com>
Change-Id: I86c2aea284178ca561c809f439faf9fd01d6e1c6
Reviewed-on: http://review.whamcloud.com/461
Tested-by: Hudson
Reviewed-by: Brian J. Murrell <brian@whamcloud.com>
Reviewed-by: Johann Lombardi <johann@whamcloud.com>
Reviewed-by: Chris Gearing <chris@whamcloud.com>
Wang Di [Wed, 20 Apr 2011 20:53:32 +0000 (13:53 -0700)]
LU-15 slow IO with read intense application
Shrink file_max_cache_size to alleviate the memory pressure of OST.
Change-Id: I6dc26ea88f8fc28bcbc94cf7d1e6f2bdfc866ea7
signed-off-by: Wang Di <di.wang@whamcloud.com>
Reviewed-on: http://review.whamcloud.com/445
Tested-by: Hudson
Reviewed-by: Johann Lombardi <johann@whamcloud.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
yangsheng [Wed, 20 Apr 2011 06:45:51 +0000 (14:45 +0800)]
LU-15 Strange slow IO messages and bad performance
b=24183 slow I/O on new files via mballoc.
upstream patch to avoid loading bitmaps from full groups
i=wangdi
i=johann
Change-Id: If44d996f8aa6268bf9e4afd2b1dbfdd8f4460a8c
Signed-off-by: Yang Sheng <ys@whamcloud.com>
Reviewed-on: http://review.whamcloud.com/441
Tested-by: Hudson
Reviewed-by: Johann Lombardi <johann@whamcloud.com>
root [Tue, 21 Dec 2010 14:00:06 +0000 (14:00 +0000)]
LU-123 Port yaml and auster to b1_8
Changes to add the yaml data logging from the 2.0 branch in the 1.8 branch, this
patch was created by applying the 2.0 yml patch to 1.8 and then resolving the issues.
This is to say that the larger changes have been taken to be correct because of their
acceptance in the 2.0 master.
This patch also contains the addition of the auster.sh script which allows for the logging
of test results to the maloo database. The actual logging is carried out by the
maloo_upload.sh script.
For instructions on how to use auster use auster -?, for instructions on maloo refer to
the whamcloud wiki [where I will begin to place a resource of information!]
Change-Id: I602a3534f17544d857aa0a9f9f82d2873fb73a39
Signed-off-by: Chris Gearing <chris@whamcloud.com>
Signed-off-by: Bobi Jam <bobijam@whamcloud.com>
Reviewed-on: http://review.whamcloud.com/421
Tested-by: Hudson
Reviewed-by: Yu Jian <yujian@whamcloud.com>
Reviewed-by: Johann Lombardi <johann@whamcloud.com>
Bobi Jam [Wed, 20 Apr 2011 04:14:41 +0000 (12:14 +0800)]
LU-179 Should let PF_MEMALLOC cover ptlrpc_register_bulk()
Under memory pressure, the rpc memory allocation in bulk registration on the memory
release path also needs be covered.
Change-Id: Ie4986eb48d1e5b99b1756a0ddaa11fe0beb37462
Signed-off-by: Bobi Jam <bobijam@whamcloud.com>
Reviewed-on: http://review.whamcloud.com/439
Reviewed-by: Liang Zhen <liang@whamcloud.com>
Tested-by: Hudson
Reviewed-by: Johann Lombardi <johann@whamcloud.com>
Bobi Jam [Tue, 19 Apr 2011 01:56:19 +0000 (09:56 +0800)]
LU-206 Memory corruption due to MMP (1.8.5)
Change-Id: Ibc0c3ea76eb94c7a64f7e17711e9fe4b954cc37b
Signed-off-by: Bobi Jam <bobijam@whamcloud.com>
Reviewed-on: http://review.whamcloud.com/433
Tested-by: Hudson
Reviewed-by: Oleg Drokin <green@whamcloud.com>
Niu Yawei [Tue, 22 Mar 2011 08:19:15 +0000 (01:19 -0700)]
LU-129 Skip quotacheck on administratively disabled OSTs
- skip quotacheck on administratively disabled OSTs.
- introduce OBD_NOTIFY_DEACTIVATE & OBD_NOTIFY_ACTIVATE event, thus lov is now aware of
administratively enable/disable import event, and is able to udpate the 'ltd_activate'
accordingly.
Issue: LU-129
Signed-off-by: Niu Yawei <niu@whamcloud.com>
Change-Id: Id3d2c9e1e035086e653ab500459912f7960976c8
Reviewed-on: http://review.whamcloud.com/351
Tested-by: Hudson
Reviewed-by: Lai Siyao <laisiyao@whamcloud.com>
Reviewed-by: Johann Lombardi <johann@whamcloud.com>
Brian J. Murrell [Mon, 7 Mar 2011 22:37:44 +0000 (17:37 -0500)]
LU-62 rhel6 support
So that RHEL6 client can be built.
Signed-off-by: Brian J. Murrell <brian@whamcloud.com>
Change-Id: Ie5ff1249aee8725625901d1dad242a8f8c48765a
Reviewed-on: http://review.whamcloud.com/385
Tested-by: Hudson
Reviewed-by: Johann Lombardi <johann@whamcloud.com>
Brian J. Murrell [Tue, 5 Apr 2011 14:28:14 +0000 (10:28 -0400)]
LU-177 jenkins build enhancements
From
fd0bd149e66cbc06a17c372e159f9bfba024cc97:
b=22281 refactor packaging and versioning
Only the --enable-dist configure flag.
i=wangyb
i=mjmac
From
ae4652c1ac831912e186d10214b266b994d7a7e2:
b=24090 distro and target autodetection
Automatically detect the target and distro. It's in fact a little
crazy that these are two separate options. They should be merged.
i=cliff
i=minh
From
63520d0a8089ba4929fc6767442e85bde4a1ccad:
b=24091 find_linux_rpms utility
Write a utility to find the the kernel "devel" package needed.
Restructure some code to facilitate reuse of existing code to
use in the utility.
i=cliff
i=minh
From
2941269c348cfed9c22d02c45eb00687339adaaa:
LU-120 the perfect storm
There is a corner case in the debian patching/unpatching process
where a patch converts the use of a given file to a file.in that is
to be substed. The problem is that between the configure and the
make debs, a file, which a debian patch wants to remove already
exists (it was created by the configure) and dpkg-buildpackage's
unpatch step fails becuase it wants to try to create a file which
configure already created.
Signed-off-by: Brian J. Murrell <brian@whamcloud.com>
Change-Id: I185e496b82687d21ea3d744e939536c106b702f9
Reviewed-on: http://review.whamcloud.com/308
Tested-by: Hudson
Reviewed-by: Robert Read <rread@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>
From
98bf9c9e3ecaa6736dcb066d0c76e20610fa9725
LU-177 lbuild autonomy fixes
Fixes to allow lbuild to work without uneeded help from it's caller:
- CentOS is a "rhel" distro
Direct output to the output FD.
Signed-off-by: Brian J. Murrell <brian@whamcloud.com>
Change-Id: I714b90b53b14ac4ef4cd6bfe638a029656a151b0
Reviewed-on: http://review.whamcloud.com/361
Reviewed-by: Oleg Drokin <green@whamcloud.com>
Tested-by: Hudson
Reviewed-by: Michael MacDonald <mjmac@whamcloud.com>
From
8fd9914d125989b924fc6c0e1d50dfd193a73438:
LU-73 lbuild and friends for jenkins
Add the portions of this patch needed for lbuild to work in
our matrix-building jenkins environment.
Included is a re-org of the RHEL5 lbuild code to maximize code
reuse and minimze code replication for future RHEL6 landing.
Signed-off-by: Brian J. Murrell <brian@whamcloud.com>
Reviewed-by: Michael MacDonald <mjmac@whamcloud.com>
Reviewed-by: Robert Read <rread@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>
Signed-off-by: Brian J. Murrell <brian@whamcloud.com>
Change-Id: I6b4c668d16b302f62a4d5afb9c84a051c260af06
Reviewed-on: http://review.whamcloud.com/402
Tested-by: Hudson
Reviewed-by: Michael MacDonald <mjmac@whamcloud.com>
Reviewed-by: Yang Sheng <ys@whamcloud.com>
Reviewed-by: Johann Lombardi <johann@whamcloud.com>
Lai Siyao [Mon, 28 Feb 2011 02:20:02 +0000 (10:20 +0800)]
LU-73 b=22375 RHEL6 kernel support for 1.8
rhel6 patchless client support:
* different shrinker function type
* blk_queue_max_sectors/segments renamed
* removed obsolete function bio_hw_segments
signed-off-by: Lai Siyao <laisiyao@whamcloud.com>
Change-Id: Ia96cbfafc6213a86b21254aad88d7807256fd0d9
Reviewed-on: http://review.whamcloud.com/282
Tested-by: Hudson
Reviewed-by: Johann Lombardi <johann@whamcloud.com>
Bobi Jam [Thu, 31 Mar 2011 01:20:31 +0000 (09:20 +0800)]
LU-57 Allow OSTs to be created with no primary node
Add a --servicenode parameter for mkfs.lustre to treat all service
nodes equally.
Bugzilla: 19064
Change-Id: I7851cb7781cdaf59d2cf18690d2da35121432cc1
Signed-off-by: Bobi Jam <bobijam@whamcloud.com>
Elena Gryaznova [Wed, 30 Mar 2011 18:37:56 +0000 (22:37 +0400)]
b=23935 append truncate race
i=Vladimir.Saveliev
i=Oleg.Drokin
Original patch from Vitaly Fertman
Elena Gryaznova [Wed, 30 Mar 2011 12:41:06 +0000 (16:41 +0400)]
b=21847 obdfilter-survey: Syntax error in some locales
i=Vladimir.Saveliev
Vitaly Fertman [Wed, 30 Mar 2011 10:36:27 +0000 (14:36 +0400)]
b=21501 Properly cleanup flock lock on disconnect
Properly wakeup flock waiters on eviction.
Destroyed lock for flock completion ast is not an error, return success
to avoid double lock decref.
i=rread
i=adilger
LaiSiYao [Wed, 30 Mar 2011 09:59:08 +0000 (13:59 +0400)]
b=24437 revoke open lock for executable files if needed
When a normal lustre client open write/exec a file, the open lock on that file needs to be revoked
in case an NFSD lustre client still holds it.
i=green@whamcloud.com
i=vladimir.saveliev@oracle.com
Niu Yawei [Sat, 19 Feb 2011 07:18:46 +0000 (23:18 -0800)]
b=22729 Remove LPSZ & LPSSZ
Code cleanup patch for 1.8 which removes the use of LPSZ/LPSSZ to
improve the build portability.
Issue: LU-54
Change-Id: I04d8446564bf36d0806ff3ee9abf8950a33eb170
Signed-off-by: Niu Yawei <niu@whamcloud.com>
Brian J. Murrell [Tue, 29 Mar 2011 17:32:04 +0000 (13:32 -0400)]
LU-53 run autogen if a Makefile.am is patched
Signed-off-by: Brian J. Murrell <brian@whamcloud.com>
Change-Id: I40eb3cecff64a62c2eca212ed19263128ad4b50d
Dmitry Zogin [Thu, 24 Mar 2011 10:49:37 +0000 (13:49 +0300)]
b=21137 Sles11 with 1.8 is slower than 1.6 sles10 for O_DIRECT single file IOR writes
Fix ptlrpc_main() condition to start service threads correctly.
i=adilger
i=vitaly.fertman
Elena Gryaznova [Tue, 29 Mar 2011 15:41:52 +0000 (19:41 +0400)]
b=23049 t-f do_node() VERBOSE fix
i=Vladimir.Saveliev
Terry Rutledge [Mon, 28 Mar 2011 16:53:26 +0000 (09:53 -0700)]
Cleared white space warning from file. Deleted
blank line from EOF.
Brian J. Murrell [Mon, 28 Mar 2011 16:45:51 +0000 (12:45 -0400)]
b=24479 LU-92 files and dirs missing in dist tarball
Some files and dirs are missing in the "dist" tarball.
Change-Id: Id9fea606549174cac2ea1f7239d75f2fc63c144c
Signed-off-by: Brian J. Murrell <brian@whamcloud.com>
unknown [Mon, 14 Mar 2011 17:13:47 +0000 (10:13 -0700)]
Updated version for the next incremental build.
unknown [Thu, 10 Mar 2011 19:15:34 +0000 (11:15 -0800)]
Corrected a trailing white space error.
Elena Gryaznova [Thu, 10 Mar 2011 16:46:00 +0000 (19:46 +0300)]
b=19494 "lfs find" hangs when searching for an OST index
i=Emoly.Liu
i=Johann.Lombardi <johann@whamcloud.com>
- new test_88 "lfs find identifies the missing striped file segments"
- exit_status () egrep pattern fix
hongchao.zhang [Mon, 28 Feb 2011 19:38:46 +0000 (03:38 +0800)]
b=24194 increase reseed count to mitigate inconsistence in OST allocation
in alloc_rr, "LOV_CREATE_RESEED_MULT" and "LOV_CREATE_RESEED_MIN" is increased
to mitigate the inconsistence in OST allocation.
i=johann
i=andreas.dilger
Elena Gryaznova [Wed, 2 Mar 2011 21:04:39 +0000 (00:04 +0300)]
b=24451 racer test cleanup
i=Vladimir.Saveliev
i=Andrew.Perepechko <Andrew_Perepechko@xyratex.com>
- modify racer/racer.sh to wait the process killed,
exit 1 if the process are still existing;
- remove runracer;
Elena Gryaznova [Mon, 28 Feb 2011 19:49:29 +0000 (22:49 +0300)]
b=19649 sanity test_77j fix
i=Johann.Lombardi <johann@whamcloud.com>
Elena Gryaznova [Mon, 28 Feb 2011 19:05:32 +0000 (22:05 +0300)]
b=24426 add ERRLOG suffix to not ovewrite the lustre logs
i=Vladimir.Saveliev
Vitaly Fertman [Tue, 22 Feb 2011 10:29:54 +0000 (13:29 +0300)]
b=24420 avoid an LASSERT on recovery
o=Diego.Moreno
i=tappro
i=vitaly
Vitaly Fertman [Tue, 22 Feb 2011 10:29:53 +0000 (13:29 +0300)]
b=24375 Fix a race between completion and enqueue
o=green
i=vitaly
i=vs
ldlm_enqueue_tail does not obtain proper lockng when checking lock mode to see
if the lock is granted, so there is a window where ldlm_handle_completion_ast
can update lvb with correct data, but beforeit has a chance to update the lock
mode, the ldlm_enqueue_tail will check the lock mode and since the lock is not
granted yet, it will overwrite correct lvb with stale value from enqueue time.
Jian Yu [Tue, 22 Feb 2011 11:15:56 +0000 (19:15 +0800)]
b=24050 fix lustre_start to start server targets in the order of MGS->MDT->OST(s)
i=vitaly.fertman
Vladimir Saveliev [Thu, 17 Feb 2011 19:16:26 +0000 (22:16 +0300)]
b=24426 run_one(): run error() once
there is no reason to run error() (and lctl dk thereby) more than once
second lctl dk overwrites the most important logs obtained on first lctl dk
i=grev
LaiSiYao [Fri, 18 Feb 2011 15:31:40 +0000 (23:31 +0800)]
b=23787 Modified struct lprocfs_percpu to be C99 compliant.
i=yangsheng
Elena Gryaznova [Thu, 17 Feb 2011 14:17:43 +0000 (17:17 +0300)]
b=24432 mount_lustre.c/parse_options() fix
o=Chris.Horn <hornc@cray.com>
i=Elena.Gryaznova
to differentiate between 'force*' and 'force'
Vitaly Fertman [Thu, 17 Feb 2011 13:45:57 +0000 (16:45 +0300)]
b=22168 write-append-truncate: retry write when receives EINTR.
r=jian.yu
r=vitaly.fertman
Vitaly Fertman [Wed, 16 Feb 2011 20:16:33 +0000 (23:16 +0300)]
b=22984 change all references to tune.ldiskfs in lustre to tunefs.ldiskfs
o=girish
i=adilger
i=johann
Vitaly Fertman [Wed, 16 Feb 2011 20:16:32 +0000 (23:16 +0300)]
b=21135 calculate Use% for "lfs df" the same way as standard "df"
o=andreas
i=girish
i=simmonsja
Elena Gryaznova [Mon, 14 Feb 2011 17:25:35 +0000 (20:25 +0300)]
b=19944 adjust debug size to be -gt num_possible_cpus()
a=Landen <zhiyong.tian@oracle.com>
i=Elena.Gryaznova
Elena Gryaznova [Mon, 14 Feb 2011 16:53:09 +0000 (19:53 +0300)]
b=23670 exit_status () fix
i=Vladimir.Saveliev
Jian Yu [Sat, 12 Feb 2011 03:08:49 +0000 (11:08 +0800)]
b=23430 fix sanity-quota test 14a to write file in O_DIRECT mode
o=andrew.perepechko
i=johann.lombardi
i=landen
Elena Gryaznova [Tue, 8 Feb 2011 22:10:31 +0000 (01:10 +0300)]
b=24374 lov_dump_user_lmm_header () fix
i=Johann
Terry Rutledge [Wed, 2 Feb 2011 22:26:37 +0000 (14:26 -0800)]
Updated version for 1.8.6 Build 03.
Lai Siyao [Wed, 2 Feb 2011 12:45:02 +0000 (13:45 +0100)]
b=23064 create proper macro check for bdi interface
i=johann
Johann Lombardi [Wed, 2 Feb 2011 12:39:32 +0000 (13:39 +0100)]
add missing changelog entries
Liang Zhen [Mon, 31 Jan 2011 08:18:29 +0000 (16:18 +0800)]
b=14846 dynamically grow/shrink connd threads pool
i=isaac
if multiple nodes are down, all socklnd connds could be blocked for a
long while, we can workaround this by increase default nconnds but it
always requires to have unnecessary number of threads.
This patch can support dynamically grow/shrink connd threads pool, it
can create new thread if there's pending active connecting, it will
kill some threads if there are too many idle connds.
Liang Zhen [Fri, 19 Nov 2010 03:47:07 +0000 (11:47 +0800)]
b=24218 fix contention on ksock_tx_t
i=isaac
If a connection is closed before ksocknal_transmit() returns to
ksocknal_process_transmit(), then nobody has refcount on conn::ksnc_sock
and all pending ZC requests will be finalized by
ksocknal_connsock_decref->ksocknal_finalize_zcreq, ksocknal_finalize_zcreq
will mark not-acked ZC request as error by setting tx::tx_reside = -1.
This is race because ksocknal_process_transmit() will check tx::tx_resid
right after calling ksocknal_transmit(), and it can get
tx->tx_resid != 0 and rc == 0 then hit later LASSERT(rc < 0).
Elena Gryaznova [Tue, 1 Feb 2011 11:42:05 +0000 (14:42 +0300)]
b=23983 mmp test_10 fix
i=Vladimir.Saveliev
Götz Waschk [Mon, 31 Jan 2011 13:55:04 +0000 (14:55 +0100)]
Fix rhel6 build issue
Johann Lombardi [Mon, 31 Jan 2011 13:46:45 +0000 (14:46 +0100)]
b=23499 add changelog entry
Bobi Jam [Mon, 31 Jan 2011 13:33:49 +0000 (21:33 +0800)]
b=23499 ASSERTION(atomic_read(&client_stat->nid_exp_ref_count) == 0)
i=johann
i=liang
In lprocfs_exp_setup(), we need release old stats in all cases.
Johann Lombardi [Mon, 31 Jan 2011 13:32:08 +0000 (14:32 +0100)]
b=23729 cancel_lru_locks not working cause some locks are still in cache from mmap files
Fix sanity-benchmark.sh to remove files after fsx otherwise client keeps locks acquired for mmap files in cache.
Johann Lombardi [Mon, 31 Jan 2011 08:31:01 +0000 (09:31 +0100)]
b=21581 change wrong URL
Brian J. Murrell [Fri, 28 Jan 2011 22:00:53 +0000 (17:00 -0500)]
b=21581 fix typo; add Fedora
Fix a typo.
Add Fedora for the yum cases per Andreas.
LU-47
Change-Id: Ie4e6fd9860dcd1c6280cdd907c54bd2e50aab7cc
Johann Lombardi [Fri, 28 Jan 2011 08:51:23 +0000 (09:51 +0100)]
b=24427 hopefully the last libcfs_memory_pressure_* fix for liblustre
Johann Lombardi [Fri, 28 Jan 2011 08:33:08 +0000 (09:33 +0100)]
b=24427 another userspace fix for libcfs_memory_pressure_restore()
Johann Lombardi [Fri, 28 Jan 2011 07:40:18 +0000 (08:40 +0100)]
b=24427 define libcfs_memory_pressure_get for userspace
Johann Lombardi [Fri, 28 Jan 2011 07:32:14 +0000 (08:32 +0100)]
Revert "b=21776 make sure libcfs_memory_pressure_get is declared before using it in LNetPut()"
This reverts commit
941880626e036fc48f31e13c5b71bb1118346d5c.
This patch does not help ...
Johann Lombardi [Thu, 27 Jan 2011 22:13:21 +0000 (23:13 +0100)]
b=21776 make sure libcfs_memory_pressure_get is declared before using it in LNetPut()
Brian J. Murrell [Thu, 27 Jan 2011 19:52:02 +0000 (14:52 -0500)]
b=21581 too long file / path names for old tar
Instruct automake to use tar's ustar format to prevent errors when
pathnames are longer than 99 characters.
- this requires automake >= 1.9, so adjust accordingly
- including dealing with multiple versions of automake installed
i=adilger
Issue: LU-47
Change-Id: Iaafcb3573c164c2b0e3efd60eb4223dc93bf1493
Elena Gryaznova [Thu, 27 Jan 2011 13:58:48 +0000 (14:58 +0100)]
b=24410 exit with error if NFSCLIENT is set, but no nfs export found
Johann Lombardi [Thu, 27 Jan 2011 13:56:59 +0000 (14:56 +0100)]
b=24388 remove another file inadvertently added by previous commit
Johann Lombardi [Thu, 27 Jan 2011 13:54:48 +0000 (14:54 +0100)]
b=24388 remove files inadvertently added by previous commit
Elena Gryaznova [Thu, 27 Jan 2011 13:52:47 +0000 (14:52 +0100)]
b=24388 sgpdd-survey fix: use node_var_name () for variables
i=andrew
Isaac Huang [Thu, 27 Jan 2011 13:49:27 +0000 (14:49 +0100)]
b=21776 Set PF_MEMALLOC on outgoing path to prevent deadlock on memory allocation under pressure
i=johann
i=dmitry
i=maxim
Johann Lombardi [Thu, 27 Jan 2011 13:38:37 +0000 (14:38 +0100)]
b=22980 init_logging does not exist in 1.8
Brian J. Murrell [Wed, 26 Jan 2011 19:38:18 +0000 (14:38 -0500)]
b=24417 Update Build-Depends
- remove texlive-latex-recommended as a build requirement
- add missing "| automake1.7 | automake1.8 | automake1.9" to
debian/control.main
Issue: LU-52
Change-Id: I53b54a068e2dc5cf8704235031ca198784bab6c9
Brian J. Murrell [Wed, 26 Jan 2011 19:38:17 +0000 (14:38 -0500)]
b=24416 debian packaging fixes
- don't make a patch out of anything in /debian
- exclude noise files from the debian built source tarball
- fake debian/patche{s,d} for make dist
- a few more reasons to run autogen.sh
- figure out if dist tarball needs autogen.shs and include it if so
- look for and run autogen.sh in the build subdir
- make debdiff as part of make dist
- add a debian/source/format file
- mv the orig tarball and the debdiff to the debs dir
- don't try to dist /debian for non-dpkg-using build targets
Issue: LU-51
Brian J. Murrell [Wed, 26 Jan 2011 16:05:41 +0000 (11:05 -0500)]
b=24413 fix for automake > 1.9.6
We seem to be using a Makefile variable that does not exist in more
recent versions of automake. This fixes that problem.
i=oleg
v=hudson@whamcloud
Issue: LU-48
Change-Id: I886b54583b444c4911e71ae6ba6836a8ce075dd5
Terry Rutledge [Wed, 26 Jan 2011 19:02:44 +0000 (11:02 -0800)]
Revert "b=24413 fix for automake > 1.9.6"
This reverts commit
9aaebdf27eefe38f846b89ea738dd0f4f3e16d8e.
Brian J. Murrell [Wed, 26 Jan 2011 16:05:41 +0000 (11:05 -0500)]
b=24413 fix for automake > 1.9.6
We seem to be using a Makefile variable that does not exist in more
recent versions of automake. This fixes that problem.
i=oleg
v=hudson@whamcloud
Issue: LU-48
Change-Id: I886b54583b444c4911e71ae6ba6836a8ce075dd5
Niu Yawei [Wed, 26 Jan 2011 07:28:59 +0000 (23:28 -0800)]
b=22980 Support unlocked_ioctl
Adding 'unlocked_ioctl' for performance sensitive ioctls, such as OBD_IOC_BRW_READ/WRITE
i=andreas
i=johann
Change-Id: I963aaad6bf9e2fdf9b70f15659e5ad47180e0d5a
Vitaly Fertman [Wed, 26 Jan 2011 15:22:15 +0000 (18:22 +0300)]
b=24320 do not fork a new thread in mem pressure
i=johann
i=green
we already check for PF_MEMALLOC in ldlm shrinker and pass this flag
to the blocking thread, but a new thread start was still done with
no check for this flag.
Johann Lombardi [Tue, 25 Jan 2011 17:07:20 +0000 (18:07 +0100)]
b=24245 fix SA perf test to support SA disabled by default
Bob Glossman [Mon, 24 Jan 2011 14:17:35 +0000 (15:17 +0100)]
b=17275 make lustre client less verbose at startup time for Cray
i=johann
niu [Mon, 24 Jan 2011 13:10:07 +0000 (14:10 +0100)]
b=24360 fix NULL pointer deref in mds_verify_child() when ll_lookup_one_len() fails
i=johann
Andreas Dilger [Tue, 30 Nov 2010 23:22:14 +0000 (16:22 -0700)]
b=20563 Fix fid_flatten() after 1 trillion SEQ numbers
Fix the fid_flatten() function to properly handle FID mapping to
64-bit inode numbers, after the first 1 trillion SEQ numbers have
been granted out. Even with CMD this would only happen after
1024 MDTs have each had 1B client mounts, so there is little risk
of introducing collisions as a result of this change, and at worst
this is a client-local phenomenon that is not persistent.
Andreas Dilger [Tue, 30 Nov 2010 22:58:29 +0000 (15:58 -0700)]
b=20563 Fix fid_flatten32() to not lose OID bits
The original implementation of fid_flatten32() was broken due to an
error in the shift calculation (note to self - "0x00" is 8 bits, not
16 bits). This could negatively impact 32-bit clients that were
creating more than 64k files in the same directory.
This 32-bit inode number is visible only within a single client mount,
is not used in any persistent storage, and only if a 2.x server is in
use (which is basically none today) by a 32-bit client, so there is no
issue to change it at this time.
Joseph Herring [Mon, 24 Jan 2011 12:58:03 +0000 (13:58 +0100)]
b=22660 Return kernel's locking return code to when lustre reports success
i=johann
i=andrew
Zhang HongChao [Mon, 24 Jan 2011 12:53:25 +0000 (13:53 +0100)]
b=23352 modified value of at_min is not taken into account
i=Johann
i=Eri Mei
Dmitry Zogin [Mon, 18 Oct 2010 19:20:47 +0000 (15:20 -0400)]
b=22378 Correct MDS client stats
sanity test_133b fails with "The getattr counter on mds is wrong"
message.
i=Andrew Perepechko
Johann Lombardi [Thu, 20 Jan 2011 13:15:47 +0000 (14:15 +0100)]
b=15962 disable statahead by default due to important races found in the code
Bernd Schubert [Thu, 20 Jan 2011 12:39:18 +0000 (13:39 +0100)]
b=22882 MMP might sleep negative time
i=adilger
i=girish
Christopher Morrone [Thu, 20 Jan 2011 12:24:48 +0000 (13:24 +0100)]
b=21456 Patch to support lnet v1 pings in 'lctl ping'
i=maxim
i=isaac
Johann Lombardi [Wed, 19 Jan 2011 14:14:22 +0000 (15:14 +0100)]
bump version to 1.8.5.53
Johann Lombardi [Wed, 19 Jan 2011 14:12:17 +0000 (15:12 +0100)]
b=23988 Remove sd iostats patch from sles11 patch series
Terry Rutledge [Tue, 11 Jan 2011 16:37:12 +0000 (08:37 -0800)]
Updated version for 1.8.6 build 02.
Johann Lombardi [Fri, 7 Jan 2011 08:26:20 +0000 (09:26 +0100)]
add missing changelog entries