Whamcloud - gitweb
fs/lustre-release.git
7 years agoLU-2874 tests: mark slow sync zfs tests as EXCEPT_SLOW
Nathaniel Clark [Thu, 28 Feb 2013 06:16:18 +0000 (01:16 -0500)]
LU-2874 tests: mark slow sync zfs tests as EXCEPT_SLOW

For tests that take a very long time on ZFS but not on ldiskfs, place
them in EXCEPT_SLOW when running on zfs.

Signed-off-by: Nathaniel Clark <nathaniel.l.clark@intel.com>
Change-Id: I52193d9a1c51e1276a5ac86c46f32e5ba5dd6299
Reviewed-on: http://review.whamcloud.com/5553
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
7 years agoLU-2877 tests: sanity test_34h needs to flush cache after write
Oleg Drokin [Wed, 27 Feb 2013 03:30:24 +0000 (22:30 -0500)]
LU-2877 tests: sanity test_34h needs to flush cache after write

We need to ensure that the cache is clean after dd in the test,
otherwise subsequent multiop might block trying to flush the
dirty pages for more than 2 seconds and trigger a false failure.

Change-Id: Ifb5a0aa0f9c627b353abe0d402c42a4e14d67609
Signed-off-by: Oleg Drokin <oleg.drokin@intel.com>
Reviewed-on: http://review.whamcloud.com/5541
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Jinshan Xiong <jinshan.xiong@intel.com>
7 years agoLU-2645 ldlm: use correct lvb size to reply 1.8 lock enqueue
Fan Yong [Tue, 5 Feb 2013 16:38:04 +0000 (00:38 +0800)]
LU-2645 ldlm: use correct lvb size to reply 1.8 lock enqueue

For 1.8 client, it does not support variable-sized LVB.
The 2.4 server should correctly distinguish whether the
client support it or not, and fill the reply buffer with
suitable LVB size when processing lock enqueue.

Test-Parameters: envdefinitions=SLOW=yes,ENABLE_QUOTA=yes \
clientjob=lustre-b1_8 clientbuildno=256 testlist=runtests

Signed-off-by: Fan Yong <fan.yong@intel.com>
Change-Id: I9241efe25dc64b26e86c4e75da72ab74bb1bc750
Reviewed-on: http://review.whamcloud.com/5459
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Jinshan Xiong <jinshan.xiong@intel.com>
Reviewed-by: Jian Yu <jian.yu@intel.com>
Reviewed-by: Keith Mannthey <keith.mannthey@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
7 years agoLU-2832 ptlrpc: cleanup bulk for resend case
Hongchao Zhang [Sun, 3 Feb 2013 07:27:49 +0000 (15:27 +0800)]
LU-2832 ptlrpc: cleanup bulk for resend case

when the request with bulk(ptlrpc_bulk_desc) is resent or replayed,
the stats of the bulk should be cleaned up for it will be reused.

Change-Id: Ie340c8aa43e1a19595c50bed05134537d8d07d74
Signed-off-by: Hongchao Zhang <hongchao.zhang@intel.com>
Reviewed-on: http://review.whamcloud.com/5532
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Fan Yong <fan.yong@intel.com>
Reviewed-by: Mike Pershin <mike.pershin@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
7 years agoLU-2776 tests: sleep longer time to yield CPU
Jinshan Xiong [Mon, 11 Feb 2013 20:55:52 +0000 (12:55 -0800)]
LU-2776 tests: sleep longer time to yield CPU

It used to be 0.1 seconds and turned out too less.

Signed-off-by: Jinshan Xiong <jinshan.xiong@intel.com>
Change-Id: I2776d8387aee8a55f325459999c8c9454dd2b4fa
Reviewed-on: http://review.whamcloud.com/5321
Tested-by: Hudson
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
7 years agoLU-2793 fld: send special seq lookup request to MDT0
wangdi [Thu, 28 Nov 2013 12:03:17 +0000 (04:03 -0800)]
LU-2793 fld: send special seq lookup request to MDT0

Since almost all special sequence locates on MDT0, we should
send all seq lookup req to MDT0, especially in an evironment
where other MDTs might not be started at all.

Signed-off-by: wang di <di.wang@intel.com>
Change-Id: Ifda9cb434a217d1e54dc2ef4fcb7628fca049d9d
Reviewed-on: http://review.whamcloud.com/5319
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Fan Yong <fan.yong@intel.com>
7 years agoLU-2683 lov: release all locks in closure to release sublock
Jinshan Xiong [Wed, 30 Jan 2013 00:35:49 +0000 (16:35 -0800)]
LU-2683 lov: release all locks in closure to release sublock

We used to only release current parent lock, this may cause deadlock
if the sublock is shared. See stacktrace of LU-2683 and LU-874 for
details.

Signed-off-by: Jinshan Xiong <jinshan.xiong@intel.com>
Change-Id: Ibe5fc364ef22a279f23bb24ad1311a1ad09be369
Reviewed-on: http://review.whamcloud.com/5208
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Bobi Jam <bobijam@gmail.com>
Reviewed-by: Niu Yawei <yawei.niu@intel.com>
Reviewed-by: Prakash Surya <surya1@llnl.gov>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
7 years agoLU-2805 tests: compare the file correctly for layout swapping
Jinshan Xiong [Wed, 13 Feb 2013 21:34:53 +0000 (13:34 -0800)]
LU-2805 tests: compare the file correctly for layout swapping

The size of file ref2 may be less than copied bytes so the result of
command cmp will be false due to EOF. Check differ string instead.

Signed-off-by: Jinshan Xiong <jinshan.xiong@intel.com>
Change-Id: I344f90bf4b75f535961f1acdc5574aeca8e85126
Reviewed-on: http://review.whamcloud.com/5420
Tested-by: Hudson
Reviewed-by: Jian Yu <jian.yu@intel.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
7 years agoLU-2449 mdd: set linkea on /ROOT
Alex Zhuravlev [Wed, 13 Feb 2013 10:52:05 +0000 (14:52 +0400)]
LU-2449 mdd: set linkea on /ROOT

osd-zfs/ needs linkea set to handle lookup(..) and readdir.

Signed-off-by: Alex Zhuravlev <alexey.zhuravlev@intel.com>
Change-Id: I35d601dbbc0598da4aa73abb69a59480fa108e0a
Reviewed-on: http://review.whamcloud.com/5413
Tested-by: Hudson
Reviewed-by: Fan Yong <fan.yong@intel.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Mike Pershin <mike.pershin@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
7 years agoLU-1812 osd-ldiskfs: handle 4-arg fsync properly
Jeff Mahoney [Mon, 4 Feb 2013 22:07:07 +0000 (17:07 -0500)]
LU-1812 osd-ldiskfs: handle 4-arg fsync properly

Fix locking around fsync in osd_object_sync.

Signed-off-by: Jeff Mahoney <jeffm@suse.com>
Change-Id: Idc35c1649af7ef7d65fba316ffbcd86574fa847a
Reviewed-on: http://review.whamcloud.com/5115
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Niu Yawei <yawei.niu@intel.com>
Reviewed-by: chas williams - CONTRACTOR <chas@cmf.nrl.navy.mil>
Reviewed-by: James Simmons <uja.ornl@gmail.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
7 years agoLU-1546 mdd: return nlink as zero for dead objects
John L. Hammond [Wed, 16 Jan 2013 11:53:19 +0000 (05:53 -0600)]
LU-1546 mdd: return nlink as zero for dead objects

In mdd_attr_get_internal() check for dead objects and set nlink to
zero accordingly.  Add a sanity check to verify that this works as
expected.

Change-Id: I9dd574cce63bcbf47c9f8f3c7ae6d151c704a45a
Signed-off-by: John L. Hammond <john.hammond@intel.com>
Reviewed-on: http://review.whamcloud.com/3154
Tested-by: Hudson
Reviewed-by: Bobi Jam <bobijam@gmail.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
7 years agoLU-2467 utils: Update wirecheck for new connect flags
Li Wei [Tue, 5 Feb 2013 06:54:21 +0000 (14:54 +0800)]
LU-2467 utils: Update wirecheck for new connect flags

This patch adds OBD_CONNECT_PINGLESS to wirecheck and updates
wiretests accordingly.  Also, the connect flag readme comment is
updated to include this as a guidance for adding connect flags in the
future.

Signed-off-by: Li Wei <wei.g.li@intel.com>
Change-Id: I131291dd2dcfda07045a09cbfde4833128464966
Reviewed-on: http://review.whamcloud.com/5269
Reviewed-by: Niu Yawei <yawei.niu@intel.com>
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
7 years agoLU-1812 ldiskfs: Add kernel patch series for SLES11SP2
James Simmons [Wed, 13 Feb 2013 18:11:43 +0000 (13:11 -0500)]
LU-1812 ldiskfs: Add kernel patch series for SLES11SP2

Add a kernel patch series for SLES11SP2.

Signed-off-by: James Simmons - <uja.ornl@gmail.com>
Signed-off-by: chas williams - CONTRACTOR <chas@cmf.nrl.navy.mil>
Signed-off-by: Jeff Mahoney <jeffm@suse.com>
Change-Id: I5e3902eb90e69c9652cdfbfc0d8ea1d99d15b453
Reviewed-on: http://review.whamcloud.com/4972
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
7 years agoLU-2548 quota: proper iam index truncation
Niu Yawei [Thu, 7 Feb 2013 06:58:18 +0000 (01:58 -0500)]
LU-2548 quota: proper iam index truncation

In quota migration, the iam container needs be reinitialized
after truncation.

Signed-off-by: Niu Yawei <yawei.niu@intel.com>
Change-Id: I622cc8e4e095cc0c9dcd2d2ce0e245a6de376b3d
Reviewed-on: http://review.whamcloud.com/5292
Reviewed-by: Johann Lombardi <johann.lombardi@intel.com>
Reviewed-by: Fan Yong <fan.yong@intel.com>
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
7 years agoLU-2844 mdd: mdt prepare failure should not oops
Lai Siyao [Fri, 22 Feb 2013 14:35:29 +0000 (22:35 +0800)]
LU-2844 mdd: mdt prepare failure should not oops

If mdd prepare fails, lfsck data may not be initialized, mdd
lfsck stop should check this flag.

Add one test to fail mdt prepare, and system should not oops
upon this failure.

Signed-off-by: Lai Siyao <lai.siyao@intel.com>
Change-Id: Ie3e81a2a18aedaa47965800881292c122937506a
Reviewed-on: http://review.whamcloud.com/5519
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Ned Bass <bass6@llnl.gov>
Reviewed-by: Fan Yong <fan.yong@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
7 years agoLU-1812 ext4: ec_type has been dropped from ext4_ext_cache
chas williams - CONTRACTOR [Mon, 7 Jan 2013 19:59:53 +0000 (14:59 -0500)]
LU-1812 ext4: ec_type has been dropped from ext4_ext_cache

In short, EXT4_EXT_CACHE_EXTENT is indicated when both
ee_len != 0 and ee_start != 0.

see kernel commit b05e6ae58a13b56e3e11882c1fc71948c9b29760

Signed-off-by: chas williams - CONTRACTOR <chas@cmf.nrl.navy.mil>
Change-Id: I88bb07c0e09ea13ac17399edb47de8e16f105543
Reviewed-on: http://review.whamcloud.com/4970
Reviewed-by: James Simmons <uja.ornl@gmail.com>
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Peng Tao <bergwolf@gmail.com>
Reviewed-by: Niu Yawei <yawei.niu@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
7 years agoLU-2473 ldiskfs: Reorganize ldiskfs kernel patches
Christopher J. Morrone [Tue, 29 Jan 2013 01:21:11 +0000 (17:21 -0800)]
LU-2473 ldiskfs: Reorganize ldiskfs kernel patches

This commit makes no changes to the patches, only moving
their location.

For some time we have supported only a single kernel per major
OS release, e.g. RHEL 6.3, or RHEL 6.2, but not both.  While
it is understandable to only support a single kernel, it is
quite another to intentionally destroy the patchset for the previous
kernel every time we upgrade.

This makes the logistics of upgrading an OS release (even a minor
one like RHEL 6.2 to 6.3, or 6.3 to 6.4) quite complicated.

We should really start a new patch series when we update support
for a new kernel.  That way we can leave the patch sets for
previous kernels in place.  We do not need to "support" them, but
at least leaving them unmolested until the users of the previous
patch set have time to upgrade would be greatly appreciated.

This commit attempts to organize the patches into subdirectories
according to the kernel/os in which they were first created.  A
later commit will then add support for RHEL6.4.

Also, fix build/confirmpatches.sh to allow checking
ldiskfs/kernel_patches (a bug prevent checking anything but the
default).

Extend both confirmpatches.sh and clearpatches.sh to work with
series files that point to patches organized into subdirectories.

Change-Id: I6552fc271fff1f00657ba1430c4a1215dea5b530
Signed-off-by: Christopher J. Morrone <morrone2@llnl.gov>
Reviewed-on: http://review.whamcloud.com/4803
Reviewed-by: James Simmons <uja.ornl@gmail.com>
Tested-by: Hudson
Reviewed-by: Jeff Mahoney <jeffm@suse.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
7 years agoLU-2658 llite: too many arguments in generic_file_llseek_size
Minh Diep [Fri, 1 Feb 2013 16:30:32 +0000 (08:30 -0800)]
LU-2658 llite: too many arguments in generic_file_llseek_size

A generic_file_llseek_size starting kernel 3.5 has 5 arguments.
We need to detect and use with the correct number of arguments.

Signed-off-by: Minh Diep <minh.diep@intel.com>
Change-Id: I966de37a880a94cde54be84062dec2b91e8334f5
Reviewed-on: http://review.whamcloud.com/5244
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Yang Sheng <yang.sheng@intel.com>
Reviewed-by: Peng Tao <bergwolf@gmail.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
7 years agoLU-2828 osp: correct osp device finialize order
Bobi Jam [Tue, 26 Feb 2013 01:55:58 +0000 (09:55 +0800)]
LU-2828 osp: correct osp device finialize order

* Should stop osp precreate thread before releasing its last used
  oid/seq files.

* git commit 71bdcf99 changes the return value of
  writeconf_or_reformat, it's a cleanup function and when writeconf
  does not work it reformat device and should return 0 (success)
  in the end.

Signed-off-by: Bobi Jam <bobijam.xu@intel.com>
Change-Id: I82307dea94bd9ec62088bc931c6b4dd7b90776a0
Reviewed-on: http://review.whamcloud.com/5528
Tested-by: Hudson
Reviewed-by: Niu Yawei <yawei.niu@intel.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Mike Pershin <mike.pershin@intel.com>
Reviewed-by: Alex Zhuravlev <alexey.zhuravlev@intel.com>
7 years agoLU-2781 mdt: initialize mti_spec.sp_rm_entry
John L. Hammond [Tue, 26 Feb 2013 05:08:12 +0000 (23:08 -0600)]
LU-2781 mdt: initialize mti_spec.sp_rm_entry

Prevent previous assignments to mti_spec.sp_rm_entry from affecting
the current operation by setting it to 0 in mdt_thread_info_init().

Signed-off-by: John L. Hammond <john.hammond@intel.com>
Change-Id: Iee41cd0e48718338d83b62e0610bd5d23d80dc47
Reviewed-on: http://review.whamcloud.com/5327
Reviewed-by: Niu Yawei <yawei.niu@intel.com>
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Mike Pershin <mike.pershin@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
7 years agoLU-2469 test: Skip failing ZFS tests
Nathaniel Clark [Tue, 19 Feb 2013 23:05:05 +0000 (18:05 -0500)]
LU-2469 test: Skip failing ZFS tests

Skip tests that are currently failing for ZFS.  Associate with bugs so
that they can be unskipped when they are fixed.

Test-Parameters: mdsfilesystemtype=zfs ostfilesystemtype=zfs   mdtfilesystemtype=zfs
Signed-off-by: Nathaniel Clark <nathaniel.l.clark@intel.com>
Change-Id: I7fe3a2ad51659c4e31240594702a4cdc87a67336
Reviewed-on: http://review.whamcloud.com/5472
Tested-by: Hudson
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Jian Yu <jian.yu@intel.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>
7 years agoLU-2280 tests: wait osc import to be FULL after sanityn 34.
wangdi [Tue, 20 Nov 2012 18:09:22 +0000 (10:09 -0800)]
LU-2280 tests: wait osc import to be FULL after sanityn 34.

1. It should wait osc import to be FULL before proceeding to the
following test.

2. fixes in osc_wait_import_event for waiting multiple osc imports.

Signed-off-by: Wang Di <di.wang@intel.com>
Change-Id: Ia054264954a6e0aacadc7c8062c2f353cd87f834
Reviewed-on: http://review.whamcloud.com/4468
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Jian Yu <jian.yu@intel.com>
Reviewed-by: Jinshan Xiong <jinshan.xiong@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
7 years agoLU-2624 ptlrpc: improve stop of ptlrpcd threads
Gregoire Pichon [Wed, 16 Jan 2013 12:08:03 +0000 (13:08 +0100)]
LU-2624 ptlrpc: improve stop of ptlrpcd threads

This fix improves the stop of ptlrpcd threads so that
it occurs in parallel. On machines with a large number
of cores this significantly reduce the time of unload of
ptlrpc module.

Signed-off-by: Gregoire Pichon <gregoire.pichon@bull.net>
Change-Id: I744de5d9887bfaa4c7a742d34923680e3a536ee7
Reviewed-on: http://review.whamcloud.com/5039
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Faccini Bruno <bruno.faccini@intel.com>
Reviewed-by: Niu Yawei <yawei.niu@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
7 years agoLU-2720 clio: remove CILR_PEEK optimization
Vitaly Fertman [Thu, 31 Jan 2013 12:40:07 +0000 (16:40 +0400)]
LU-2720 clio: remove CILR_PEEK optimization

some issues with that optimization in mkwrite are found:
1) ENODATA handling wrote the page and invalidated it, but the page
could have been re-added to the cache in between these 2 steps;
2) lock could be busy, so we get ENODATA on trylock, whereas a racy
ll_fault waits for lock mutex and re-adds the page to the cache;

The patch also adds a loop over trying to take a cached lock while
it is used.

Signed-off-by: Vitaly Fertman <vitaly_fertman@xyratex.com>
Change-Id: I7799b45ff3a0a60f307e05ffd1618d23e0e77b72
Xyratex-bug-id: MRP-752
Reviewed-on: http://review.whamcloud.com/5222
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Jinshan Xiong <jinshan.xiong@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
7 years agoLU-2591 lov: race between ptlrpc_rcv and umount/umount thread
Hiroya Nozaki [Wed, 9 Jan 2013 08:33:53 +0000 (17:33 +0900)]
LU-2591 lov: race between ptlrpc_rcv and umount/umount thread

The race which is refered here happens in the following scenario

1) mount runs but fails to communicate with some OSTs. Then the
   import objects which represents the OSTs are registered to
   a pinger list.
2) pinger succeeds to communicate with, at least, ONE OST. Then
   ptlrpc_rcv calls lov_set_osc_active to activate the OST and
   holds lov_refcount.
4) For some reason ... possibly mount finally fails or umount runs,
   ll_put_super is called
5) ll_put_super tries to disconnect all OSTs with lov_disconnect
   and this func calls lov_del_target to set all OSC's
   target->ltd_reap flags in order for lov_putref to handle all
   of them.
6) ptlrpc_rcv thread puts lov_refcount and if lov_refcount becomes
   0 here, the thread has to disconnect all the OSCs whose
   ltd->reap has been set by __lov_del_obd.
7) Some OSCs' imports have still been LUSTRE_IMP_CONNECTING state
   because of (2), so ptlrpc_rcv thread has to wait for these
   import states to be changed to non-recovery states, such as FULL,
   CLOSED or DISCON at ptlrpc_disconnect_import.

Now that ptlrpc_rcv thread is waiting for the import states to be
changed to non-recovery states but ptlrpc_rcv is the one who is
supposed to change a recovery state to a non-recovery state, So
ptlrpc_rcv mush hung, And mount/umount thread which has called
ll_put_super has to also wait for changing state at
ptlrpc_disconnect_import, so umount must hung too.

Signed-off-by: Hiroya Nozaki <nozaki.hiroya@jp.fujitsu.com>
Change-Id: Idcf6831d1ee6b72332c943dfde5316fddba6c13f
Reviewed-on: http://review.whamcloud.com/4979
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Jinshan Xiong <jinshan.xiong@intel.com>
Reviewed-by: Keith Mannthey <keith.mannthey@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
7 years agoLU-2792 fld: add dot_seq into in_fldb check
wangdi [Thu, 28 Nov 2013 11:37:39 +0000 (03:37 -0800)]
LU-2792 fld: add dot_seq into in_fldb check

Add FID_SEQ_DOT_LUSTRE into fid_seq_in_fldb, because the
dot_seq is inserted into fldb cache list in fld_server_init.

Signed-off-by: wang di <di.wang@intel.com>
Change-Id: I6417bab77d48c9d9c991119b29f3d2ab04e88ef2
Reviewed-on: http://review.whamcloud.com/5318
Tested-by: Hudson
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Alex Zhuravlev <alexey.zhuravlev@intel.com>
Reviewed-by: Fan Yong <fan.yong@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
7 years agoLU-2801 mdt: checking remote object in mdt_swap_layouts
wangdi [Sat, 30 Nov 2013 09:44:28 +0000 (01:44 -0800)]
LU-2801 mdt: checking remote object in mdt_swap_layouts

It should use mdt_remote_object to check remote object in
mdt_swap_layouts, instead of mdt_object_exists < 0.

Send swap layout req to the right MDT.

Signed-off-by: wang di <di.wang@intel.com>
Change-Id: Ia24635e16cb67e5683abc9797ec6c9e61f648c62
Reviewed-on: http://review.whamcloud.com/5332
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Jinshan Xiong <jinshan.xiong@intel.com>
Reviewed-by: Alex Zhuravlev <alexey.zhuravlev@intel.com>
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
7 years agoLU-2739 mdt: Deny non-DNE client to access remote directory
wangdi [Sun, 1 Dec 2013 02:03:19 +0000 (18:03 -0800)]
LU-2739 mdt: Deny non-DNE client to access remote directory

Return EIO to old clients, if they try to access the remote
directory.

Signed-off-by: wang di <di.wang@intel.com>
Change-Id: I1cc4e8fa6e0f3a4b31b5610642e204bd100c1743
Reviewed-on: http://review.whamcloud.com/5425
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Fan Yong <fan.yong@intel.com>
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
7 years agoLU-2729 mgs: fix cleanup of nonexistant MGSSELF_NAME proc file
Jeff Mahoney [Mon, 4 Feb 2013 22:05:02 +0000 (17:05 -0500)]
LU-2729 mgs: fix cleanup of nonexistant MGSSELF_NAME proc file

The mgs teardown code gets invoked in the failure case but assumes
that the MGSSELF_NAME proc file has been created. If it hasn't been
created, the proc code issues a WARN_ON.

The setup code clears the FSDB_MGS_SELF from &fsdb->fsdb_flags when it
creates the proc file, so we can key off of that to decide whether to
remove it.

Signed-off-by: Jeff Mahoney <jeffm@suse.com>
Change-Id: Idf58d1fa572eeced5dbaa72f03e38ee46a708166
Reviewed-on: http://review.whamcloud.com/5122
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Niu Yawei <yawei.niu@intel.com>
Reviewed-by: James Simmons <uja.ornl@gmail.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
7 years agoLU-2642 osd-ldiskfs: Use __REQ_WRITE instead of BIO_RW.
Jeff Mahoney [Fri, 18 Jan 2013 08:21:55 +0000 (03:21 -0500)]
LU-2642 osd-ldiskfs: Use __REQ_WRITE instead of BIO_RW.

Commit 7b6d91da removed BIO_RW and replaced it with __REQ_WRITE.
(v2.6.35)

Signed-off-by: Jeff Mahoney <jeffm@suse.com>
Change-Id: I33f87d1e7d8ffe90efcd8b173ad56d9ecf723d7a
Reviewed-on: http://review.whamcloud.com/5121
Reviewed-by: James Simmons <uja.ornl@gmail.com>
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
7 years agoLU-2158 lvfs: move obdclass related functions to obclass
James Simmons [Tue, 19 Feb 2013 17:29:29 +0000 (12:29 -0500)]
LU-2158 lvfs: move obdclass related functions to obclass

Currently a lot of functions exist in the lvfs layer
that really belongs to different parts of the stack.
This patch moves obdclass specific code that is located
in the lvfs layer into the proper location.

Signed-off-by: James Simmons <uja.ornl@gmail.com>
Change-Id: I5be4dede1afdbbc4af0157cbcd13a5133ce7a1db
Reviewed-on: http://review.whamcloud.com/5250
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
7 years agoLU-2158 lvfs: move ldiskfs only functions to osd-ldiskfs
James Simmons [Tue, 5 Feb 2013 15:24:09 +0000 (10:24 -0500)]
LU-2158 lvfs: move ldiskfs only functions to osd-ldiskfs

Currently a lot of functions exist in the lvfs layer
that really belongs to different parts of the stack.
This patch moves osd-ldiskfs specific code that
is located in the lvfs layer into the proper
location.

Signed-off-by: James Simmons <uja.ornl@gmail.com>
Change-Id: Ia197391599cc7fffad396d487873edd9d6c0b1c4
Reviewed-on: http://review.whamcloud.com/5246
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Mike Pershin <mike.pershin@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
7 years agoLU-2789 lod: initialize objects in a detached stripe array
John L. Hammond [Thu, 14 Feb 2013 23:56:02 +0000 (17:56 -0600)]
LU-2789 lod: initialize objects in a detached stripe array

Initializing the ldo_stripe member of lod_device in place exposes some
races with attribute setting so instead we use a detached array and
assign that to ldo_stripe once it's complete.

Signed-off-by: John L. Hammond <john.hammond@intel.com>
Change-Id: I63d57a21ed7402829d4b3889c1f32d4ddec1f096
Reviewed-on: http://review.whamcloud.com/5422
Reviewed-by: Alex Zhuravlev <alexey.zhuravlev@intel.com>
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Mike Pershin <mike.pershin@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
7 years agoLU-2770 llite: specify ea_size for md_getattr_name
Fan Yong [Sun, 27 Jan 2013 16:54:20 +0000 (00:54 +0800)]
LU-2770 llite: specify ea_size for md_getattr_name

For nfs re-exported case, miss to set ea_size for md_getattr_name(),
then causes the ll_get_parent failure.

Signed-off-by: Fan Yong <fan.yong@intel.com>
Change-Id: I99c8301f9b00cafdf79a6c15541f868f18e4a109
Reviewed-on: http://review.whamcloud.com/5455
Tested-by: Hudson
Reviewed-by: Lai Siyao <lai.siyao@intel.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Bobi Jam <bobijam@gmail.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
7 years agoLU-2620 tests: Flush changes from test 5 in replay-ost-single
Bob Glossman [Wed, 16 Jan 2013 19:57:51 +0000 (11:57 -0800)]
LU-2620 tests: Flush changes from test 5 in replay-ost-single

Use wait_delete_completed_mds to flush out and wait for all the
effects on kbytesavail from test 5 to finish before test 6 starts.

Test-Parameters: testlist=replay-ost-single envdefinitions=SLOW=yes

Signed-off-by: Bob Glossman <bob.glossman@intel.com>
Change-Id: Id37cf23402dd609e6743c36a1aff8e71dd246776
Reviewed-on: http://review.whamcloud.com/5042
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Alex Zhuravlev <alexey.zhuravlev@intel.com>
Reviewed-by: Minh Diep <minh.diep@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
7 years agoLU-2675 cleanup: make bitfield (un)signedness explicit
John L. Hammond [Mon, 11 Feb 2013 18:23:33 +0000 (12:23 -0600)]
LU-2675 cleanup: make bitfield (un)signedness explicit

Fix this common sparse warning (dubious one-bit signed bitfield) in
case ones-complement machines come back or something.

Signed-off-by: John L. Hammond <john.hammond@intel.com>
Change-Id: I24ea468efedd8a1269ea92022d13e802f4838ba1
Reviewed-on: http://review.whamcloud.com/5166
Tested-by: Hudson
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Reviewed-by: Sebastien Buisson <sebastien.buisson@bull.net>
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
7 years agoLU-2786 llite: allow setting pool name via fsetxattr()
Ned Bass [Fri, 8 Feb 2013 22:09:09 +0000 (14:09 -0800)]
LU-2786 llite: allow setting pool name via fsetxattr()

Setting the pool name in the striping data of a file fails when the
fsetxattr() interface is used to store the lustre extended attributes,
while using ioctl() works as expected.  This is because ll_setxattr()
assumes the size of the lustre user metadata structure to be that of
the older v1 format which doesn't include a pool name field.  Fix this
by using the magic number to detect the format version and setting
the size accordingly when passing the user data down the stack.

Signed-off-by: Ned Bass <bass6@llnl.gov>
Change-Id: I6f6969dfaf2149f268d52272f538fee05459697b
Reviewed-on: http://review.whamcloud.com/5313
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Hudson
Reviewed-by: Keith Mannthey <keith.mannthey@intel.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>
7 years agoLU-2227 build: fix 'memory illegal access' errors
Sebastien Buisson [Thu, 25 Oct 2012 12:32:15 +0000 (14:32 +0200)]
LU-2227 build: fix 'memory illegal access' errors

Fix 'memory illegal access' defects found by Coverity version
6.0.3:
Buffer not null terminated (BUFFER_SIZE_WARNING)
Calling strncpy with a certain maximum size argument on
destination array of same size might leave the destination string
unterminated.
Out-of-bounds read (OVERRUN_STATIC)
Overrunning static array with n elements, at position n.
String not null terminated (STRING_NULL)
Function does not terminate string.

Signed-off-by: Sebastien Buisson <sebastien.buisson@bull.net>
Change-Id: I5ded09054b2e0e5296ba17d7528518fed298fc89
Reviewed-on: http://review.whamcloud.com/4391
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Reviewed-by: Keith Mannthey <keith.mannthey@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
7 years agoLU-2548 quota: don't apply migration on global index copy
Niu Yawei [Mon, 4 Feb 2013 06:56:04 +0000 (01:56 -0500)]
LU-2548 quota: don't apply migration on global index copy

Quota migration shouldn't be applied on the global index copy of
quota slave on mdt.

Signed-off-by: Niu Yawei <yawei.niu@intel.com>
Change-Id: I9f36b79e867ede20562223765dfc56358a4f3831
Reviewed-on: http://review.whamcloud.com/5259
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Fan Yong <fan.yong@intel.com>
Reviewed-by: Johann Lombardi <johann.lombardi@intel.com>
7 years agoLU-2391 build: separate lustre-osd-[ldiskfs,zfs] RPMs
Bruno Faccini [Mon, 18 Feb 2013 15:22:01 +0000 (16:22 +0100)]
LU-2391 build: separate lustre-osd-[ldiskfs,zfs] RPMs

Add support for separate lustre-osd-[ldiskfs,zfs] RPMs creation in
lustre.spec.in template. This will prevent osd_[ldiskfs,zfs].ko
modules from being shipped as part of lustre-modules RPM and thus
avoids having many unsatisfied external references error messages
reported when either lustre-ldiskfs or zfs-modules RPMs are missing.
Users willing to build Lustre FSs on top of ldiskfs/zfs back-ends
will have to install the corresponding RPM.

Signed-off-by: Bruno Faccini <bruno.faccini@intel.com>
Change-Id: Ib406443186026ed5ac3dc7b3ec30db41aedac2a2
Reviewed-on: http://review.whamcloud.com/4869
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Brian J. Murrell <brian.murrell@intel.com>
Reviewed-by: Christopher J. Morrone <chris.morrone.llnl@gmail.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
7 years agoLU-1812 ldiskfs: detect generated utsrelease.h
James Simmons [Thu, 14 Feb 2013 13:07:28 +0000 (08:07 -0500)]
LU-1812 ldiskfs: detect generated utsrelease.h

Fix lustre-ldiskfs.spec.in to use SLES11's generated header files.

Signed-off-by: James Simmons <uja.ornl@gmail.com>
Signed-off-by: chas williams - CONTRACTOR <chas@cmf.nrl.navy.mil>
Change-Id: I0eb198274015bfd38c48e9f68ceb15fe9e601d17
Reviewed-on: http://review.whamcloud.com/5188
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
7 years agoLU-2634 ldiskfs: clear extent flag for fast symlink
Emoly Liu [Mon, 28 Jan 2013 18:50:34 +0000 (02:50 +0800)]
LU-2634 ldiskfs: clear extent flag for fast symlink

Fast symlinks on MDT filesystems formatted with the "extents" feature
are created with EXT4_EXTENTS_FL. e2fsck considers this a corruption
and deletes the symlink.

The patch is to clear LDISKFS_INODE_EXTENTS flag in
osd_ldiskfs_writelink() just as ext4_symlink() does.

Signed-off-by: Liu Ying <emoly.liu@intel.com>
Change-Id: Ib7c1eee7becfaf2b80805d3f6c13658686ad5699
Reviewed-on: http://review.whamcloud.com/5154
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Alex Zhuravlev <alexey.zhuravlev@intel.com>
7 years agoLU-582 test: fix bug in get_ost_param in sanity.sh
Hongchao Zhang [Fri, 17 Aug 2012 08:25:32 +0000 (16:25 +0800)]
LU-582 test: fix bug in get_ost_param in sanity.sh

in get_ost_param(), the ouput should be gl_sum, which is
the sum of 'gl' in all OSS.

Signed-off-by: Hongchao Zhang <hongchao.zhang@whamcloud.com>
Change-Id: I51d3aa591c3ca875f8a559974cf05d250b0c032a
Reviewed-on: http://review.whamcloud.com/3693
Reviewed-by: Jian Yu <jian.yu@intel.com>
Tested-by: Jian Yu <jian.yu@intel.com>
Tested-by: Hudson
Reviewed-by: Fan Yong <fan.yong@intel.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
7 years agoLU-2775 osp: enable fid-on-OST only for DNE.
wangdi [Mon, 25 Nov 2013 13:50:05 +0000 (05:50 -0800)]
LU-2775 osp: enable fid-on-OST only for DNE.

1. MDT0 will only create IDIF fids, so old clients, who
only know IDIF, can still access files on MDT0.

2. Using fid_ostid_pack to pack the FID into ostid, otherwise
IDIF will be packed wrongly and caused the panic on the OST0.

3. Add dd into compatible tests of conf-sanity.sh(32) to
check OST fids on the upgrade FS.

Signed-off-by: wang di <di.wang@intel.com>
Change-Id: Ic9322bd1c66e5fedec68c10aa2f300d5ecb3b6db
Reviewed-on: http://review.whamcloud.com/5307
Tested-by: Hudson
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Fan Yong <fan.yong@intel.com>
Reviewed-by: Alex Zhuravlev <alexey.zhuravlev@intel.com>
7 years agoLU-2625 scripts: prevent several errors in lustre_rmmod
Gregoire Pichon [Wed, 16 Jan 2013 13:59:53 +0000 (14:59 +0100)]
LU-2625 scripts: prevent several errors in lustre_rmmod

This fix prevents lustre_rmmod scripts from returning
errors when lustre modules are partially or not loaded.

Signed-off-by: Gregoire Pichon <gregoire.pichon@bull.net>
Change-Id: Iefb4166b9b14315319858031f53ecb148c589474
Reviewed-on: http://review.whamcloud.com/5040
Tested-by: Hudson
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Faccini Bruno <bruno.faccini@intel.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>
7 years agoLU-2608 tests: check_catastrophe success on failure to connect
Nathaniel Clark [Tue, 29 Jan 2013 18:46:52 +0000 (13:46 -0500)]
LU-2608 tests: check_catastrophe success on failure to connect

Allow check_catastrophe to pass on failure to connect to host.
Otherwise LBUG assertions crop up when hosts are overloaded.

Signed-off-by: Nathaniel Clark <nathaniel.l.clark@intel.com>
Change-Id: Ie873b817911613369a226341d8efe06707b08bc4
Reviewed-on: http://review.whamcloud.com/5200
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Faccini Bruno <bruno.faccini@intel.com>
Reviewed-by: Jian Yu <jian.yu@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
7 years agoLU-2785 mdd: do not use mti_fid in mdd_find_or_create_root()
Alex Zhuravlev [Fri, 8 Feb 2013 19:41:59 +0000 (23:41 +0400)]
LU-2785 mdd: do not use mti_fid in mdd_find_or_create_root()

as in the subsequent callchain it can be used by
mdd_create_sanity_check()

Signed-off-by: Alex Zhuravlev <alexey.zhuravlev@intel.com>
Change-Id: Ia357f587192c54e3c406a35054dca47580077056
Reviewed-on: http://review.whamcloud.com/5311
Reviewed-by: wangdi <di.wang@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Mike Pershin <mike.pershin@intel.com>
7 years agoLU-2494 tests: add interop support to framework check
Bob Glossman [Mon, 14 Jan 2013 21:44:26 +0000 (13:44 -0800)]
LU-2494 tests: add interop support to framework check

Modify the test-framework wait_mds_ost_sync() function to work
with mds_sync when old_sync_processed doesn't exist.
This should cover both old and new servers.

Signed-off-by: Bob Glossman <bob.glossman@intel.com>
Change-Id: I2e5c430a96ac40a1ca87990696e7daaad5a69102
Reviewed-on: http://review.whamcloud.com/5021
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Jian Yu <jian.yu@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
7 years agoLU-2734 tests: Remove XXXDEV overrides in init_facets_vars()
Li Wei [Mon, 4 Feb 2013 02:27:29 +0000 (10:27 +0800)]
LU-2734 tests: Remove XXXDEV overrides in init_facets_vars()

init_facets_vars() assigns outputs of xxxdevname() calls to XXXDEVs.
This is incorrect because XXXDEVs are supposed to contain device names
while xxxdevname() calls may print dataset names for ZFS-based facets.
Actually, XXXDEV should not be changed at all.  This patch removes the
XXXDEV assignments.

Change-Id: Id484390cbd814c6fa919c02e390700f914561434
Signed-off-by: Li Wei <wei.g.li@intel.com>
Reviewed-on: http://review.whamcloud.com/5243
Tested-by: Hudson
Reviewed-by: Jian Yu <jian.yu@intel.com>
Reviewed-by: Nathaniel Clark <nathaniel.l.clark@intel.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
7 years agoNew tag 2.3.61 2.3.61 v2_3_61 v2_3_61_0
Oleg Drokin [Sun, 10 Feb 2013 05:25:48 +0000 (00:25 -0500)]
New tag 2.3.61

Change-Id: If171f34ebcec4d651ca091af4197c71d017ee645

7 years agoLU-1940 test: use default resend count for 118c
Hongchao Zhang [Sun, 20 Jan 2013 06:53:56 +0000 (14:53 +0800)]
LU-1940 test: use default resend count for 118c

in sanity, the 118c could fail with the small resend count
for SLOW=no case, using the default value instead.

Signed-off-by: Hongchao Zhang <hongchao.zhang@intel.com>
Change-Id: I3e5ecc6af2eba560f82a80eb75223f8ab299f397
Reviewed-on: http://review.whamcloud.com/4694
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
7 years agoLU-1812 ldiskfs: don't use sb->s_qf_inums
Jeff Mahoney [Mon, 4 Feb 2013 21:52:14 +0000 (16:52 -0500)]
LU-1812 ldiskfs: don't use sb->s_qf_inums

ext4-quota-first-class.patch was finally accepted into the mainline
Linux kernel via commit 7c319d32 (v3.6). The final version doesn't
include the cached inode numbers so they aren't available on newer
kernels that include this patch (or that are using an ldiskfs
patched using the usptream version).

Since the file system will refuse mounting when the inodes can't
be read, the lookup code isn't invoked without valid inode numbers
anyway.

Signed-off-by: Jeff Mahoney <jeffm@suse.com>
Change-Id: Ie1b824183cf734a0f242bb6483a192f1919dc4be
Reviewed-on: http://review.whamcloud.com/5264
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: James Simmons <uja.ornl@gmail.com>
Reviewed-by: Johann Lombardi <johann.lombardi@intel.com>
Reviewed-by: Niu Yawei <yawei.niu@intel.com>
7 years agoLU-2675 cleanup: remove User Mode Linux specific code
John L. Hammond [Thu, 31 Jan 2013 19:55:54 +0000 (13:55 -0600)]
LU-2675 cleanup: remove User Mode Linux specific code

The UML specific code has been uncompilable for some time and no one
has complained so it goes.

Signed-off-by: John L. Hammond <john.hammond@intel.com>
Change-Id: Ie91c6e4484fee64fec4a2e38effd8a1021e8cddb
Reviewed-on: http://review.whamcloud.com/5175
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
7 years agoLU-2673 procfs: call lprocfs_free_xxx_stats() later
John L. Hammond [Thu, 24 Jan 2013 21:41:56 +0000 (15:41 -0600)]
LU-2673 procfs: call lprocfs_free_xxx_stats() later

Defer calls to lprocfs_free_{obd,md}_stats() and
lprocfs_job_stats_fini() until after the corresponding proc entries
have been removed. Change the return of some proc fini()s to void.

Signed-off-by: John L. Hammond <john.hammond@intel.com>
Change-Id: Ic410a4cf12eeb084270f5e679222c54b12d80266
Reviewed-on: http://review.whamcloud.com/5160
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Keith Mannthey <keith.mannthey@intel.com>
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
7 years agoLU-2642 quota: update quota APIs
James Simmons [Wed, 30 Jan 2013 14:31:01 +0000 (09:31 -0500)]
LU-2642 quota: update quota APIs

Commit 871a2931 renamed vfs_dq_init to dquot_initialize. (v2.6.34)
Commit 9f754758 renamed vfs_dq_drop to dquot_drop. (v2.6.34)
Commit b43fa828 renamed vfs_dq_transfer to dquot_transfer. (v2.6.34)
Commit 0f0dd62f killed vfs_dq_off and vfs_dq_quota_on_remount and
replaced them with dquot_suspend and dquot_resume. (v2.6.34)

Signed-off-by: James Simmons <uja.ornl@gmail.com>
Signed-off-by: Jeff Mahoney <jeffm@suse.com>
Change-Id: I61e5a4d6c4a519457146b4a65ce78bfddebe2899
Reviewed-on: http://review.whamcloud.com/5119
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Niu Yawei <yawei.niu@intel.com>
Reviewed-by: Johann Lombardi <johann.lombardi@intel.com>
7 years agoLU-2642 osd-ldiskfs: use struct fs_disk_quota for quotas
James Simmons [Wed, 30 Jan 2013 13:59:20 +0000 (08:59 -0500)]
LU-2642 osd-ldiskfs: use struct fs_disk_quota for quotas

Commit b9b2dd36 modified ->get_dqblk to use struct fs_disk_quota.
(v2.6.34)

Signed-off-by: James Simmons <uja.ornl@gmail.com>
Signed-off-by: Jeff Mahoney <jeffm@suse.com>
Change-Id: Ifb8a79db37e3f88046c00eb906b8e7264005481b
Reviewed-on: http://review.whamcloud.com/5118
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Niu Yawei <yawei.niu@intel.com>
Reviewed-by: Johann Lombardi <johann.lombardi@intel.com>
7 years agoLU-1199 ldiskfs: Remove HAVE_{EXT4,JBD2}_JOURNAL_CALLBACK_{SET,ADD}
Christopher J. Morrone [Mon, 28 Jan 2013 17:28:25 +0000 (09:28 -0800)]
LU-1199 ldiskfs: Remove HAVE_{EXT4,JBD2}_JOURNAL_CALLBACK_{SET,ADD}

Lustre servers now only support kernels/ldiskfs with
{ext4,ldiskfs}_journal_callback_add().  Additionally, the
HAVE_EXT4_JOURNAL_CALLBACK_ADD define is not in the upstream
3.3 kernel where the ext4_journal_callback_add() patch landed.

Therefore we remove the HAVE_{LDISKFS,EXT4}_JOURNAL_CALLBACK_{SET,ADD}
checks and just assume that it is set.

Change-Id: I10cc79fcf07dd05a10ae83014f188c613ab1620c
Signed-off-by: Christopher J. Morrone <morrone2@llnl.gov>
Reviewed-on: http://review.whamcloud.com/5055
Reviewed-by: James Simmons <uja.ornl@gmail.com>
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
7 years agoLU-1431 ptlrpc: Support for over 1MB bulk I/O RPC
Sergii Glushchenko [Tue, 5 Feb 2013 08:17:34 +0000 (10:17 +0200)]
LU-1431 ptlrpc: Support for over 1MB bulk I/O RPC

Increase the maximum size of BRW RPCs between OSCs and OSTs to 4MB
by increasing the number of LNet Memory Descriptors assosiated with
a bulk.  Each bulk transfer is still limited to LNET_MTU_SIZE, so
that it can be passed through routers and does not break RDMA limits.

The client and server negotiate the maximum BRW RPC size at connect
time via ocd_brw_size, with the client sending the maximum size it
supports, and the server returning min(client_max, server_max) back
to the client.

For each RPC, the number of bulk MDs that are registered depends on
the actual RPC size at sending time.  The (max_brw_count - 1) is sent
to the OST in the high 16 bits of obd_ioobj.ioo_max_brw and forms a
mask for the bulk RPC XIDs.  The actual number of bulk transfers
is encoded in the BRW RPC XID.  The masked value of the request XID
determines the starting bulk transfer match bits, and the last bulk
match is the RPC XID.

For older clients this means that the starting and ending bulk match
bits are the RPC XID itself (as it always was) and the old ioo_type
(now ioo_max_brw) field being set to zero would result in a mask of
"1", which just keeps the same RPC XID as it was before.

It is important to note that the client and server do NOT share the
value of PTLRPC_BULK_OPS_MASK, or this would incorrectly mask the
RPC XID if the PTLRPC_BULK_OPS_BITS was ever increased on the server.

The actual BRW RPC size can be controlled for each client separately
by changing the client's max_pages_per_rpc.  For example, this will
set it to 4MB:

        lctl set_param osc.*.max_pages_per_rpc=1024

It is also possible to specify a units suffix, so that the RPC size
can be set independently of the client PAGE_SIZE:

        lctl set_param osc.*.max_pages_per_rpc=4M

By default, the size is 1MB for for OST BRW RPCs until more testing
is done. All other bulk I/O (e.g. MDC<->MDS) is left at 1MB.

With PTLRPC_MAX_BRW_SIZE growing larger, it doesn't make sense for
the readahead to increase the window size so dramatically at one time
Instead, limit readahead growth by the current inode i_blkbits value,
which currently defaults to 4MB.

If we want to tune the readahead growth so that it matches the actual
cl_max_pages_per_rpc value, then this can be done much more easily by
changing the per-inode i_blkbits value, since this is tied to specific
OSTs with specific RPC settings (which may be different for e.g. local
and remote OSTs), and only needs to be checked once.

Change the ras_*() functions to take the inode as an argument, and
always put it first in the argument list for consistency.

Signed-off-by: Andreas Dilger <andreas.dilger@intel.com>
Signed-off-by: Sergii Glushchenko <sergii_glushchenko@xyratex.com>
Reviewed-by: Alexey Lyashkov <alexey_lyashkov@xyratex.com>
Reviewed-by: Andrew Perepechko <andrew_perepechko@xyratex.com>
Change-Id: I757b14a04e5d4cc053576e41e47864c743c35b8b
Xyratex-bug-id: MRP-319
Reviewed-on: http://review.whamcloud.com/4993
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Reviewed-by: Jinshan Xiong <jinshan.xiong@intel.com>
7 years agoLU-2710 build: avoid function resolution mistakes by Coverity
Sebastien Buisson [Wed, 30 Jan 2013 15:32:27 +0000 (16:32 +0100)]
LU-2710 build: avoid function resolution mistakes by Coverity

Remove function resolution mistakes made by Coverity version
6.0.3.
The .hs_keycmp function is incorrectly resolved by Coverity,
making the variable passed as its first 'key' parameter to
be considered of the wrong type. Data structures being of
different sizes, it ends up with this Coverity defect:
Out-of-bounds access (OVERRUN_STATIC)
Overrunning static array by passing it as an argument to a
function which indexes it at a higher byte position.

Signed-off-by: Sebastien Buisson <sebastien.buisson@bull.net>
Change-Id: Ie062b05a73db16d63c63db89dc4d0e264f3b1f8a
Reviewed-on: http://review.whamcloud.com/5211
Tested-by: Hudson
Reviewed-by: Keith Mannthey <keith.mannthey@intel.com>
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
7 years agoLU-642 lov: make up obd_connect for inactive OSC
Bobi Jam [Mon, 28 Jan 2013 12:29:31 +0000 (20:29 +0800)]
LU-642 lov: make up obd_connect for inactive OSC

When OSC is inactivated before lov tries to connect it, lov_connect()
miss the chance to connect it to OST devices even when it is
activated later.

We need make up the connection for the initially inactive OSC when it
is activated.

Signed-off-by: Bobi Jam <bobijam.xu@intel.com>
Change-Id: I4aca442bc4a77baa2bc14adaf2b0f8f5c7bfa7df
Reviewed-on: http://review.whamcloud.com/4691
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Lai Siyao <lai.siyao@intel.com>
Reviewed-by: Jinshan Xiong <jinshan.xiong@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
7 years agoLU-2444 build: fix 'error handling' issues
Sebastien Buisson [Fri, 7 Dec 2012 15:12:38 +0000 (16:12 +0100)]
LU-2444 build: fix 'error handling' issues

Fix 'error handling issues' defects found by Coverity version
6.0.3:
Unchecked return value (CHECKED_RETURN)
Value returned from a function is not checked for errors before
being used.
Argument cannot be negative (NEGATIVE_RETURNS)
Negative value used as argument to a function expecting a
positive value.

Signed-off-by: Sebastien Buisson <sebastien.buisson@bull.net>
Change-Id: I44022a7d7227ee46345b77701b4c5199a74c386d
Reviewed-on: http://review.whamcloud.com/4772
Tested-by: Hudson
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Keith Mannthey <keith.mannthey@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
7 years agoLU-2302 scripts: make lfs_migrate compatible with 1.8
Keith Mannthey [Thu, 17 Jan 2013 17:33:28 +0000 (09:33 -0800)]
LU-2302 scripts: make lfs_migrate compatible with 1.8

Earlier versions of lfs setstripe/getstripe use option -s instead
of -S. Make sure lfs_migrate can be used with any version.

Signed-off-by: Daniel Kobras <d.kobras@science-computing.de>
Signed-off-by: Keith Mannthey <keith.mannthey@intel.com>
Change-Id: Ic3caaeaf3d6c764afbd75cd7a8a7279462d21d47
Reviewed-on: http://review.whamcloud.com/4702
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
7 years agoLU-2590 obdclass: correct swab lov_mds_md
Bobi Jam [Wed, 23 Jan 2013 03:08:10 +0000 (11:08 +0800)]
LU-2590 obdclass: correct swab lov_mds_md

For caller's convenience, the 2nd parameter of obd_free_diskmd()
could be host endianness, it needs swab to LE if necessary, while just
lov_mds_md header needs it for figuring out how much memory needs to
be freed.

Signed-off-by: Bobi Jam <bobijam.xu@intel.com>
Change-Id: I537e83b56b4be6e08a7431bfda1de71efefa33ee
Reviewed-on: http://review.whamcloud.com/5145
Tested-by: Hudson
Reviewed-by: Prakash Surya <surya1@llnl.gov>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>
7 years agoLU-2158 fsfilt: Remove unused opcodes
Prakash Surya [Mon, 28 Jan 2013 16:41:32 +0000 (08:41 -0800)]
LU-2158 fsfilt: Remove unused opcodes

This change removes the unused fsfilt opcode definitions and the
accompanying code in fsfilt_ext3_start. This is strictly cleanup.

Signed-off-by: Prakash Surya <surya1@llnl.gov>
Change-Id: Ifccb6e7c9c728bc98e4ddfa52093e660dd117b8b
Reviewed-on: http://review.whamcloud.com/5192
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: James Simmons <uja.ornl@gmail.com>
7 years agoLU-2158 fsfilt: Remove lvfs_common.c
Prakash Surya [Mon, 28 Jan 2013 17:23:51 +0000 (09:23 -0800)]
LU-2158 fsfilt: Remove lvfs_common.c

This change removes the lvfs_common.c file, and moves it's only
remaining functionality directly into the obd_lfs_fid2dentry function.
This should have no functional impact, and is merely a code cleanup
exercise.

Signed-off-by: Prakash Surya <surya1@llnl.gov>
Change-Id: I5f8588a17f6e7659a7d33c2a5f79792826976de4
Reviewed-on: http://review.whamcloud.com/5193
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Mike Pershin <mike.pershin@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
7 years agoLU-2360 ptlrpc: don't clobber rq_history_seq with CPT_ANY
John L. Hammond [Wed, 30 Jan 2013 02:23:13 +0000 (20:23 -0600)]
LU-2360 ptlrpc: don't clobber rq_history_seq with CPT_ANY

In ptlrpc_req_add_history() don't allow service_parts with cpt CPT_ANY
(-1) to clobber the timestamp and sequence bits of rq_history_seq.

Signed-off-by: John L. Hammond <john.hammond@intel.com>
Change-Id: I52161d6f36dcd70dd569b0be294e6d23cc5ac4a9
Reviewed-on: http://review.whamcloud.com/5207
Tested-by: Hudson
Reviewed-by: Liang Zhen <liang.zhen@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>
7 years agoLU-2764 tests: Remove obsolete comment in local.sh
Li Wei [Wed, 6 Feb 2013 02:24:43 +0000 (10:24 +0800)]
LU-2764 tests: Remove obsolete comment in local.sh

The comment above LOAD_MODULES_REMOTE in local.sh was written for the
special case of the 2.3 OSS stack and is no longer applicable to
current master.

Signed-off-by: Li Wei <wei.g.li@intel.com>
Change-Id: I75518e9cebfa717ccfb625e2786c8d191c6bddbc
Reviewed-on: http://review.whamcloud.com/5277
Tested-by: Hudson
Reviewed-by: Niu Yawei <yawei.niu@intel.com>
Reviewed-by: Jian Yu <jian.yu@intel.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
7 years agoLU-2749 ofd: check return value of ofd_seq_load
Nathaniel Clark [Mon, 4 Feb 2013 21:07:38 +0000 (16:07 -0500)]
LU-2749 ofd: check return value of ofd_seq_load

Return value ofd_seq_load was not always checked for both NULL and
errno.  This caused a NULL pointer dereference.

Signed-off-by: Nathaniel Clark <nathaniel.l.clark@intel.com>
Change-Id: Ib2e1658f16ba2012c9410966c08201cab7a5aa5c
Reviewed-on: http://review.whamcloud.com/5263
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: wangdi <di.wang@intel.com>
7 years agoLU-2414 tests: cleanup replay-single/0c and 44c
Nathaniel Clark [Tue, 18 Dec 2012 22:16:21 +0000 (17:16 -0500)]
LU-2414 tests: cleanup replay-single/0c and 44c

Switch from return codes to error messages in test_0c and test_44c.

Signed-off-by: Nathaniel Clark <nathaniel.l.clark@intel.com>
Change-Id: I61895442fa52e2f7bff421e0b8309f0f11c97de8
Reviewed-on: http://review.whamcloud.com/4856
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Tested-by: Hudson
Reviewed-by: Jian Yu <jian.yu@intel.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
7 years agoLU-1434 utils: extend wireshark parsing
Nathaniel Clark [Thu, 24 Jan 2013 14:24:26 +0000 (09:24 -0500)]
LU-1434 utils: extend wireshark parsing

Add support for MGS_CONFIG_READ
Add capabilities where they're defined in layout.c
Fix processing of MGS_BULK_PORTAL (not a Lustre packet)
Make use of LUSTRE_BUFFER_LEN everywhere.
Add dissecting of capabilities field
Add dissecting of fld field
Add dissecting of MGS_CONFIG_BODY
Fixed all compiler warnings.
Fixed parsing of get/set (x)attrs

Signed-off-by: Nathaniel Clark <nathaniel.l.clark@intel.com>
Change-Id: Ic75bb4814570b30a1f1dca19d2f49311294ff2fe
Reviewed-on: http://review.whamcloud.com/5155
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Doug Oucharek <doug.s.oucharek@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
7 years agoLU-2689 interop: use stripe md in ll_update_inode()
John L. Hammond [Thu, 7 Feb 2013 01:05:08 +0000 (19:05 -0600)]
LU-2689 interop: use stripe md in ll_update_inode()

When connected to servers that do not support LL_SBI_LAYOUT_LOCK, if
stripe md is available in ll_inode_update() then use it.

Signed-off-by: John L. Hammond <john.hammond@intel.com>
Change-Id: I949a0129c3fc92c5434288df0549a54ccd440f97
Reviewed-on: http://review.whamcloud.com/5289
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
7 years agoLU-2562 test: cleanup previous test result
Hongchao Zhang [Tue, 15 Jan 2013 12:54:11 +0000 (20:54 +0800)]
LU-2562 test: cleanup previous test result

in sanity.sh, test_51ba only depends on test_51b, but test_51a
also create files under $DIR/$tdir, which causes test_51ba fail
for it also takes those files into account.

Change-Id: I42f491f64c8d2d65bee656c167c9523274cd7200
Signed-off-by: Hongchao Zhang <hongchao.zhang@whamcloud.com>
Reviewed-on: http://review.whamcloud.com/5183
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Niu Yawei <yawei.niu@intel.com>
Reviewed-by: Jian Yu <jian.yu@intel.com>
Reviewed-by: Nathaniel Clark <nathaniel.l.clark@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
7 years agoLU-2449 osd: osd-zfs to handle ./.. properly
Alex Zhuravlev [Mon, 7 Jan 2013 17:30:17 +0000 (21:30 +0400)]
LU-2449 osd: osd-zfs to handle ./.. properly

ZFS does not store directory records . and .. on a disk.
Instead they are generated upon request (like in zfs_readdir()).
with the patch osd-zfs does the same for namespace directories.

Signed-off-by: Alex Zhuravlev <alexey.zhuravlev@intel.com>
Change-Id: I910f9bad807c0e5a2bee1fd481470576ceaf4444
Reviewed-on: http://review.whamcloud.com/4990
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
7 years agoLU-2240 mds: Assign special fid sequence to root.
wangdi [Tue, 19 Nov 2013 14:48:38 +0000 (06:48 -0800)]
LU-2240 mds: Assign special fid sequence to root.

In current implementation, we still use IGIF as the ROOT FID,
but ZFS does not have IGIF at all, so it will move root ROOT fid
to a new sequence.

But for compatible with clients, which are being mounted
during upgrade, it will still keep the old IGIF root FID.

Signed-off-by: Wang Di <di.wang@intel.com>
Change-Id: I2e6b81cd359436d2e5b96c8a315a56c415f8e631
Reviewed-on: http://review.whamcloud.com/5257
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Alex Zhuravlev <alexey.zhuravlev@intel.com>
Reviewed-by: Mike Pershin <mike.pershin@intel.com>
7 years agoLU-1866 lfsck: LFSCK for namespace consistency (3)
Fan Yong [Tue, 15 Jan 2013 09:57:33 +0000 (17:57 +0800)]
LU-1866 lfsck: LFSCK for namespace consistency (3)

It is the third part of LFSCK 1.5 for namespace consistency
check and repair. It mainly implements the LFSCK namespace
component functionality lfsck_operations::lfsck_double_scan:
process multiple-linked objects.

1) Find out and repair crashed linkEA entry.

2) Find out and remove invalid linkEA entry

Test-Parameters: envdefinitions=ENABLE_QUOTA=yes testlist=sanity-lfsck

Signed-off-by: Fan Yong <fan.yong@intel.com>
Change-Id: Ia38fa343ed4da10635c1af4d90b9039c5e01688c
Reviewed-on: http://review.whamcloud.com/4914
Tested-by: Hudson
Reviewed-by: Alex Zhuravlev <alexey.zhuravlev@intel.com>
Reviewed-by: Jian Yu <jian.yu@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
7 years agoLU-2726 procfs: Invalid assertion for counter_{add,sub}
Bobi Jam [Fri, 1 Feb 2013 01:47:33 +0000 (09:47 +0800)]
LU-2726 procfs: Invalid assertion for counter_{add,sub}

lprocfs_stats::lc_sum_irq is only used for obd_memory for accurate
memory usage counting, other procfs stats counters do not need it,
even though they can be used in RCU callbacks.

Signed-off-by: Bobi Jam <bobijam.xu@intel.com>
Change-Id: I4c2dec50eecc0c63631f1915ba30a54a5da2fec1
Reviewed-on: http://review.whamcloud.com/5237
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: John Hammond <johnlockwoodhammond@gmail.com>
Tested-by: Oleg Drokin <oleg.drokin@intel.com>
7 years agoLU-2017 mdc: add layout swap between 2 objects
Jinshan Xiong [Fri, 1 Feb 2013 18:33:09 +0000 (10:33 -0800)]
LU-2017 mdc: add layout swap between 2 objects

This patch adds the client and MDT code to swap layouts
between 2 files:
- a lfs command
- a llapi call, based on a new ioctl
- the mdc/mdt codes for the ioctl (with layout locking)

Signed-off-by: JC Lafoucriere <jacques-charles.lafoucriere@cea.fr>
Signed-off-by: Jinshan Xiong <jinshan.xiong@intel.com>
Change-Id: I7e710a1ab3ca38e8b26582e49d08e9943aa445cd
Reviewed-on: http://review.whamcloud.com/4507
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Johann Lombardi <johann.lombardi@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
7 years agoLU-1866 lfsck: LFSCK for namespace consistency (2)
Fan Yong [Tue, 15 Jan 2013 09:53:46 +0000 (17:53 +0800)]
LU-1866 lfsck: LFSCK for namespace consistency (2)

It is the second part of LFSCK 1.5 for namespace consistency
check and repair. It mainly implements the LFSCK namespace
component functionality lfsck_operations::lfsck_exec_dir:
for linkEA verification and repair.

The basic functionality is to find out missed linkEA entries
and add them to the object. It is also part of handling IGIF
objects upgrading.

Test-Parameters: envdefinitions=ENABLE_QUOTA=yes testlist=sanity-lfsck

Signed-off-by: Fan Yong <fan.yong@intel.com>
Change-Id: I9df2a1fddfa4b15838ecbdfb6279c240d01d0f95
Reviewed-on: http://review.whamcloud.com/4913
Tested-by: Hudson
Reviewed-by: Alex Zhuravlev <alexey.zhuravlev@intel.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
7 years agoLU-1866 lfsck: FID-in-{dirent,LMA} check and repair
Fan Yong [Tue, 15 Jan 2013 09:50:49 +0000 (17:50 +0800)]
LU-1866 lfsck: FID-in-{dirent,LMA} check and repair

As part of LFSCK 1.5, it implementes the following functionalities:

1) Verify and rebuild FID-in-dirent, mainly for MDT file-level
   backup/restore.

2) (Partly) verify and rebuild FID-in-LMA: if the FID-in-dirent
   is valid but FID-in-LMA is missed, then the later one will be
   repaired; but if both are missed, then it is OI scrub's duty
   to repair the inconsistency.

3) (Partly) handle IGIF objects upgrading: build IGIF-in-dirent
   for IGIF objects. The other two parts work for IGIF objects
   upgrading are:

3.1) Generate IGIF-in-LMA, which has been done by OI scrub in
     former patch(es).

3.2) Generate linkEA, which will be done by up layer LFSCK in
     other patch(es).

Test-Parameters: envdefinitions=ENABLE_QUOTA=yes testlist=sanity-lfsck

Signed-off-by: Fan Yong <fan.yong@intel.com>
Change-Id: Ibb317f0f3f3e9cf6db56bfa25f55d2017b0e1937
Reviewed-on: http://review.whamcloud.com/4912
Tested-by: Hudson
Reviewed-by: Alex Zhuravlev <alexey.zhuravlev@intel.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
7 years agoLU-1866 lfsck: LFSCK 1.5 user space control
Fan Yong [Tue, 15 Jan 2013 09:47:44 +0000 (17:47 +0800)]
LU-1866 lfsck: LFSCK 1.5 user space control

Include the user space command to start/stop the LFSCK for
namespace consistency check and repair. And proc interface
to query the LFSCK status, statistical informtion, and ect.

Test-Parameters: envdefinitions=ENABLE_QUOTA=yes testlist=sanity-lfsck

Signed-off-by: Fan Yong <fan.yong@intel.com>
Change-Id: Icf4046b313dc1653b5b8b1a3add7fa0329ca7ad0
Reviewed-on: http://review.whamcloud.com/4911
Tested-by: Hudson
Reviewed-by: Alex Zhuravlev <alexey.zhuravlev@intel.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Jian Yu <jian.yu@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
7 years agoLU-1866 lfsck: LFSCK for namespace consistency (1)
Fan Yong [Tue, 15 Jan 2013 09:47:16 +0000 (17:47 +0800)]
LU-1866 lfsck: LFSCK for namespace consistency (1)

It is the first part of LFSCK 1.5 for namespace consistency
check and repair. It mainly implements the LFSCK namespace
component internal framework and external APIs.

The functionality parts for the LFSCK namespace component,
such as lfsck_operations::lfsck_{exec_dir,dump,double_scan},
are not included. They will be implemented in other patches.

Signed-off-by: Fan Yong <fan.yong@intel.com>
Change-Id: I011e06c991577e632eadf364b87be74f1f5535c6
Reviewed-on: http://review.whamcloud.com/4910
Reviewed-by: Mike Pershin <mike.pershin@intel.com>
Tested-by: Hudson
Reviewed-by: Alex Zhuravlev <alexey.zhuravlev@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
7 years agoLU-1866 lfsck: LFSCK main engine
Fan Yong [Tue, 15 Jan 2013 09:45:58 +0000 (17:45 +0800)]
LU-1866 lfsck: LFSCK main engine

There are two engines for LFSCK 1.5:

1) One is low layer object table based iteration, which can
   guarantee that all the objects can be scanned at least once.

2) The other is namespace based directory traversal. All the
   namespace based components, such as FID-in-dirent and linkEA,
   can be handled during that.

Signed-off-by: Fan Yong <fan.yong@intel.com>
Change-Id: Ic5a404cde35fdb20c24f6522faa8b21117cf4254
Reviewed-on: http://review.whamcloud.com/4909
Tested-by: Hudson
Reviewed-by: Alex Zhuravlev <alexey.zhuravlev@intel.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
7 years agoLU-1866 lfsck: general framework for LFSCK 1.5
Fan Yong [Tue, 15 Jan 2013 07:07:07 +0000 (15:07 +0800)]
LU-1866 lfsck: general framework for LFSCK 1.5

LFSCK 1.5 main data structure definations, LFSCK
component APIs, setup/cleanup/start/stop interfaces.

Signed-off-by: Fan Yong <fan.yong@intel.com>
Change-Id: I4507b9432289ca8d54cbf686893b706052d0eeb3
Reviewed-on: http://review.whamcloud.com/4908
Tested-by: Hudson
Reviewed-by: Alex Zhuravlev <alexey.zhuravlev@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>
7 years agoLU-1866 lfsck: ancillary work for namespace LFSCK
Fan Yong [Tue, 15 Jan 2013 07:05:35 +0000 (15:05 +0800)]
LU-1866 lfsck: ancillary work for namespace LFSCK

1) Share related linkEA functions with LFSCK.

2) Multiple linked objects are rare, and the objects with
   a lot of linkea entries are more rare. It is unnecessary
   to worry about performance under such cases. So drop the
   linkea entries count limitation: LINKEA_MAX_COUNT.

Signed-off-by: Fan Yong <fan.yong@intel.com>
Change-Id: I75f90c9619c56f2f62063435aea26ed01337d6dd
Reviewed-on: http://review.whamcloud.com/4907
Reviewed-by: Alex Zhuravlev <alexey.zhuravlev@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
7 years agoLU-2733 osd: osd-zfs to initialize existing sequence
Alex Zhuravlev [Fri, 1 Feb 2013 19:35:04 +0000 (23:35 +0400)]
LU-2733 osd: osd-zfs to initialize existing sequence

osd_find_or_add_seq() to use osd_oi_find_or_create() so that
existing sequence is loaded if exists

Signed-off-by: Alex Zhuravlev <alexey.zhuravlev@intel.com>
Change-Id: I9f173615ef047dd4696d8ed51687c7c268b32dbf
Reviewed-on: http://review.whamcloud.com/5247
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Mike Pershin <mike.pershin@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
7 years agoLU-2728 fld: check for NULL param in fldb_seq_stop()
John L. Hammond [Fri, 1 Feb 2013 01:51:04 +0000 (19:51 -0600)]
LU-2728 fld: check for NULL param in fldb_seq_stop()

Check for NULL param in fldb_seq_stop().

Signed-off-by: John L. Hammond <john.hammond@intel.com>
Change-Id: I44ed10f9ce8bceabb203e6ef0b843b23340ced2b
Reviewed-on: http://review.whamcloud.com/5238
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
7 years agoLU-2737 osc: double unlock in osc_lru_shrink()
Niu Yawei [Mon, 4 Feb 2013 03:49:26 +0000 (22:49 -0500)]
LU-2737 osc: double unlock in osc_lru_shrink()

It could unlock twice in osc_lru_shrink() if the cl_io_init() failed.

Signed-off-by: Niu Yawei <yawei.niu@intel.com>
Change-Id: Idf3ebafced7f7d3a68c13ea20d12cf221e9d74ce
Reviewed-on: http://review.whamcloud.com/5258
Reviewed-by: Bobi Jam <bobijam@gmail.com>
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Jinshan Xiong <jinshan.xiong@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
7 years agoLU-1187 mdt: Add MDS_INODELOCK_PERM lock for remote dir
Wang Di [Thu, 26 Jul 2012 11:02:58 +0000 (04:02 -0700)]
LU-1187 mdt: Add MDS_INODELOCK_PERM lock for remote dir

Add MDS_INODELOCK_PERM for remote directory on remote MDT to
protect permission changes on remote MDT. So, Master MDT, where
the parent is, will grant LOOKUP lock to the client. Remote MDT,
where the child is, will grant UPDATE|PERM lock to the client.

And client will clear the dcache during, when either PERM or LOOKUP
lock is being revoked.

Change-Id: Ie11d14bbe7c706a76f7c4bd8f75b07807e2dac02
Signed-off-by: Wang Di <di.wang@intel.com>
Reviewed-on: http://review.whamcloud.com/4346
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Reviewed-by: Mike Pershin <mike.pershin@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
7 years agoLU-2722 clio: directIO thread races with completion thread
Vitaly Fertman [Thu, 31 Jan 2013 12:12:42 +0000 (16:12 +0400)]
LU-2722 clio: directIO thread races with completion thread

ll_direct_IO_26()) ASSERTION( obj->cob_transient_pages == 0 ) failed

The directio thread puts pages to the transfer in osc_io_submit(), and
waits for its completion. Upon completion, osc_completion() is called,
which notifies waiters IO is completed and only after that put the
page. the original directIO thread wakes up and may succeed to assert
on the last page which is not yet put by osc_completion().

Signed-off-by: Vitaly Fertman <vitaly_fertman@xyratex.com>
Change-Id: Ib912e913885a9ff2cb1e9c865ae2003a676c2b5b
Xyratex-bug-id: MRP-838
Reviewed-on: http://review.whamcloud.com/5223
Reviewed-by: Jinshan Xiong <jinshan.xiong@intel.com>
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Niu Yawei <yawei.niu@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
7 years agoLU-2467 ptlrpc: Allow OBD_PINGs to be suppressed
Li Wei [Wed, 30 Jan 2013 12:38:59 +0000 (20:38 +0800)]
LU-2467 ptlrpc: Allow OBD_PINGs to be suppressed

This patch introduces a new ptlrpc module parameter, "suppress_pings",
to provide an option for reducing excessive OBD_PINGs in large
clusters.  The parameter affects all MDTs and OSTs on a node.  It is
off (zero) by default, giving a behavior identical to current
implementation.  If it is on (non-zero), all clients of the affected
targets who understand OBD_CONNECT_PINGLESS will know, at connect
time, that pings are not required.  When suppressing pings, there must
be an external mechanism to notify the targets of client deaths, via
the targets' "evict_client" procfs entries.  In addition, a highly
available standalone MGS is also recommended when suppressing pings,
so that clients are notified (through Imperative Recovery) of target
recoveries.

The changes do two basically independent things.  One is initializing
import and export states (i.e., imp_connect_data and
exp_obd_chain_timed) according to "suppress_pings", since whether to
ping or not is a property of each import-export pair.  MGC pings can
not be suppressed, because maintaining MGS connections is dictated by
the reliance on Imperative Recovery.

The other thing is changing pinger and import routines to respect the
import property set earlier.  (The export side does not need any
change at all.)  Pings are still needed to query last committed
transactions if there are uncommitted requests on an import, so that
resources pinned for replays can be released even when applications
become idle.  An early version of this patch removes imports that do
not need to be pinged from pinger_imports and add them back when last
committed transactions are needed or recoveries must be initiated.
This version does not do that because a) the overheads of iterating
through 10,000 imports are not prohibitively large---around 10 ms, b)
adding imports back to pinger_imports requires the global pinger_mutex
to be held, and c) the imp_lock contention added on each import is
small.

Change-Id: Iabc84d395c978c3f156c52aebfad83621facb4fe
Signed-off-by: Li Wei <wei.g.li@intel.com>
Reviewed-on: http://review.whamcloud.com/5009
Reviewed-by: Lai Siyao <lai.siyao@intel.com>
Tested-by: Hudson
Reviewed-by: Emoly Liu <emoly.liu@intel.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
7 years agoLU-2467 protocol: Add OBD_CONNECT_PINGLESS
Li Wei [Thu, 31 Jan 2013 09:48:23 +0000 (17:48 +0800)]
LU-2467 protocol: Add OBD_CONNECT_PINGLESS

Reserve a bit for OBD_CONNECT_PINGLESS, which indicates a client is
capable of suppressing keep-alive OBD_PINGs.  If granted by a server,
it means the server does not require (but still allows) pings.

Signed-off-by: Li Wei <wei.g.li@intel.com>
Change-Id: I7575082244a84f2e633a1936fcf894da3c6ee7dd
Reviewed-on: http://review.whamcloud.com/5229
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
7 years agoLU-2723 clio: a client hangs on osc_extent_wait in dio
Vitaly Fertman [Thu, 31 Jan 2013 12:36:19 +0000 (16:36 +0400)]
LU-2723 clio: a client hangs on osc_extent_wait in dio

If dio overlaps with mapped region, first overlapped pages are sent
by generic_file_direct_write->filemap_write_and_wait_range->writepages
After that the dio pages are sent by write_begin/write_end, writepages
However, write_end->osc_page_cache_add does not release extent because
the IO is not considered as synchronous. as the result, the following
writepages->osc_cache_writeback_range() does not send pages as the
extent is still ACTIVE and osc_cache_wait_range() hangs forever.

The patch make directIO to be considered as sync by lustre so that
extent was released at write_end.

Signed-off-by: Vitaly Fertman <vitaly_fertman@xyratex.com>
Xyratex-bug-id: MRP-839
Change-Id: I0c0e1599ee180f467826e4b6715c3cbfd0c6eb5c
Reviewed-on: http://review.whamcloud.com/5224
Reviewed-by: Jinshan Xiong <jinshan.xiong@intel.com>
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
7 years agoLU-2194 tests: add debugging to recovery-small/19
Nathaniel Clark [Mon, 21 Jan 2013 20:15:20 +0000 (15:15 -0500)]
LU-2194 tests: add debugging to recovery-small/19

Add error messages to 19a and 19b

Signed-off-by: Nathaniel Clark <nathaniel.l.clark@intel.com>
Change-Id: Ief35c063734654c5a6f3dab0d283c4b7a32706e8
Reviewed-on: http://review.whamcloud.com/5141
Reviewed-by: Li Wei <wei.g.li@intel.com>
Reviewed-by: Jian Yu <jian.yu@intel.com>
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
7 years agoLU-1187 mdt: directory remote open fix
Wang Di [Tue, 2 Oct 2012 13:54:53 +0000 (06:54 -0700)]
LU-1187 mdt: directory remote open fix

Directory remote open continue as normal open, i.e. return the
remote fid to client, and client will send the open request to
the right MDT.

Set correct it disposition in mdt_open_by_fid.

Signed-off-by: wang di <di.wang@intel.com>
Change-Id: I32d5dee9cd1eac47787e2d85a5db1e164d2846d3
Reviewed-on: http://review.whamcloud.com/4934
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Hudson
Reviewed-by: Alex Zhuravlev <alexey.zhuravlev@intel.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Mike Pershin <mike.pershin@intel.com>
7 years agoLU-1187 mdd: a few missing stuff in MD stack for DNE.
wangdi [Tue, 1 Oct 2013 11:25:49 +0000 (04:25 -0700)]
LU-1187 mdd: a few missing stuff in MD stack for DNE.

1. Assign the index operation for the directory create
in declare phase, which is needed for creating the
object in OUT.

2. Declare dotdot insertion for remote directory creation,
so the insert update can be packed into RPC and send to
the remote MDT (out) to be executed.

3. Add hint to the object declare phase, by which OSP can
get some info during update RPC packing, like parent fid.

4. Separate lu_object_exists into lu_object_remote and
lu_object_exists, so it can check whether remote object
exists as well.

Signed-off-by: wang di <di.wang@intel.com>
Change-Id: Icb359443d9982ee8567af933c5def42dc51a3a7a
Reviewed-on: http://review.whamcloud.com/4930
Tested-by: Hudson
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Alex Zhuravlev <alexey.zhuravlev@intel.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Mike Pershin <mike.pershin@intel.com>
7 years agoLU-2654 ldlm: ldlm_resource_dump should be called with resource locked
Oleg Drokin [Sat, 2 Feb 2013 04:08:59 +0000 (23:08 -0500)]
LU-2654 ldlm: ldlm_resource_dump should be called with resource locked

Fix ldlm_resource_complain to lock resource before calling
ldlm_resource_dump

Change-Id: I9dfbe84bd63de3093c99e7379daeca30d02fc875
Signed-off-by: Oleg Drokin <green@whamcloud.com>
Reviewed-on: http://review.whamcloud.com/5254
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Alex Zhuravlev <alexey.zhuravlev@intel.com>
Reviewed-by: Liang Zhen <liang.zhen@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Tested-by: Oleg Drokin <oleg.drokin@intel.com>
7 years agoLU-1187 mdt: add sanity check for rename and link
wangdi [Sun, 21 Oct 2012 10:49:02 +0000 (03:49 -0700)]
LU-1187 mdt: add sanity check for rename and link

Add sanity check for rename/link, so remote rename/link
will return EXDEV.

Signed-off-by: wang di <di.wang@whamcloud.com>
Change-Id: I841c8ffaa74b162fcca5cd1661c383215350cccc
Reviewed-on: http://review.whamcloud.com/4348
Tested-by: Hudson
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Alex Zhuravlev <alexey.zhuravlev@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
7 years agoLU-1187 mdt: unlink remote directory
wangdi [Tue, 19 Nov 2013 14:34:21 +0000 (06:34 -0800)]
LU-1187 mdt: unlink remote directory

Send unlink req to the slave MDT, so both open/close
and unlink will send request to the slave MDT, where
the remote object resides. Then it would be able to
check orphan (unlink open file) locally.

Add lock_object api for enqueue remote object.

Change-Id: I7483b0f023e4e3de6597da58d3d9f3e96c0d53b7
Signed-off-by: Wang Di <di.wang@intel.com>
Reviewed-on: http://review.whamcloud.com/4339
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Alex Zhuravlev <alexey.zhuravlev@intel.com>
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
7 years agoLU-1187 utils: add lfs setdirstripe/getdirstripe
wangdi [Sat, 16 Nov 2013 09:19:02 +0000 (01:19 -0800)]
LU-1187 utils: add lfs setdirstripe/getdirstripe

1.Add lfs setdirstripe/mkdir/getdirstripe to create
remote directory.

2.Add lfs rm_entry to just remove the entry of remote
directory.

Change-Id: Idaecfbb9dd97e3ae9fbc70e5b902a91a7b2ed18e
Signed-off-by: Wang Di <di.wang@intel.com>
Reviewed-on: http://review.whamcloud.com/4341
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Alex Zhuravlev <alexey.zhuravlev@intel.com>
7 years agoLU-1866 lfsck: enhance otable-based iteration
Fan Yong [Sat, 12 Jan 2013 11:47:12 +0000 (19:47 +0800)]
LU-1866 lfsck: enhance otable-based iteration

1) Use ::load() method to unplug OSD layer iteration.

2) Use ::put() method to wakeup the LFSCK main engine
   if it is blocked at low layer for ::next() call.

Test-Parameters: envdefinitions=ENABLE_QUOTA=yes testlist=sanity-scrub
Signed-off-by: Fan Yong <fan.yong@intel.com>
Change-Id: I28e5689ad3a4ee96a3f17a64e8d6dca553a86d77
Reviewed-on: http://review.whamcloud.com/4906
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Tested-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Alex Zhuravlev <alexey.zhuravlev@intel.com>