Whamcloud - gitweb
Andreas Dilger [Thu, 30 Mar 2017 23:12:29 +0000 (17:12 -0600)]
LU-8998 llapi: rename llapi_layout_comp_move -> *use
Rename llapi_layout_comp_move() and llapi_layout_comp_move_at()
to llapi_layout_comp_use() and llapi_layout_comp_use_id(),
respectively. This avoids confusion about what "move" and "at" in
the function name implies. The component itself is not actually
being moved, just a different layout component is being selected
for access or modification. Using "_id" instead of "_at" also
makes it more clear what the difference is between these functions.
Rename LLAPI_LAYOUT_COMP_POS_{FIRST,NEXT,LAST} correspondingly to
LLAPI_LAYOUT_COMP_USE_{FIRST,NEXT,LAST} to match.
Split llapi_layout_comp_use_id.3 from llapi_layout_comp_use.3 since
they are mostly independent anyway.
Test-Parameters: trivial testlist=sanity-pfl
Signed-off-by: Andreas Dilger <andreas.dilger@intel.com>
Change-Id: I85926d4ec9774745bc49b0d178ed9b23ec2cab07
Reviewed-on: https://review.whamcloud.com/26483
Tested-by: Jenkins
Reviewed-by: Niu Yawei <yawei.niu@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Jinshan Xiong <jinshan.xiong@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
John L. Hammond [Fri, 28 Apr 2017 15:12:01 +0000 (10:12 -0500)]
LU-9306 kuc: initialize kkuc_groups at module init time
Some kkuc functions use kkuc_groups[group].next == NULL to test for an
empty group list. This is incorrect if the group was previously added
to but not empty. Remove all next == NULL tests and use module load
time initialization of the kkuc_groups array.
Signed-off-by: John L. Hammond <john.hammond@intel.com>
Change-Id: I0d6af7b8584f18d1dc03873993d6aac55b1677a9
Reviewed-on: https://review.whamcloud.com/26883
Tested-by: Jenkins
Reviewed-by: Frank Zago <fzago@cray.com>
Reviewed-by: Faccini Bruno <bruno.faccini@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Henri Doreau <henri.doreau@cea.fr>
Reviewed-by: Quentin Bouget <quentin.bouget@cea.fr>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Jinshan Xiong [Mon, 17 Apr 2017 19:29:51 +0000 (12:29 -0700)]
LU-9340 lov: readahead shouldn't exceed component boundary
Otherwise, it will extend the readahead RPC to the next component
while the actual lock of that component is not checked.
Signed-off-by: Jinshan Xiong <jinshan.xiong@intel.com>
Change-Id: Ice743d45f9df5e6fdc83b07aa6af1b182b660c9a
Reviewed-on: https://review.whamcloud.com/26677
Tested-by: Jenkins
Reviewed-by: Bobi Jam <bobijam@hotmail.com>
(cherry picked from commit
e31e234c06ac798501cdb7ec92269af83157cb21)
Reviewed-on: https://review.whamcloud.com/26861
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Niu Yawei <yawei.niu@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Wang Shilong [Sun, 9 Apr 2017 02:22:24 +0000 (10:22 +0800)]
LU-4017 quota: add project id support to lfs find
Project ID support is added into 'lfs find', which is
needed for sanity-quota.sh etc to collect project id information
if test failed
Change-Id: I60cfc4f81bb8779db0a33a5c9bae7255e9d0100c
Signed-off-by: Wang Shilong <wshilong@ddn.com>
Reviewed-on: https://review.whamcloud.com/26464
Reviewed-by: Li Xi <lixi@ddn.com>
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Niu Yawei <yawei.niu@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Wang Shilong [Fri, 7 Apr 2017 01:09:18 +0000 (09:09 +0800)]
LU-4017 quota: add project inherit attributes
Add @LUSTRE_PROJINHERIT_FL inode flag which means
creating new objects parents projid, it is disabled
in default, unless setting explictly.
It is kept same inteface as Ext4/XFS, you could use
following ioctl directly:
chattr +P <directory>
Agent inode for DNE should ignore project ID, patch also
remove some unnecessary lustre_set_wire_obdo() call for
attr fetching/updates, which mixed la flags and obdo flags.
Change-Id: I573b71c5bd7b0089172025c30c6824562444d57d
Signed-off-by: Wang Shilong <wshilong@ddn.com>
Reviewed-on: https://review.whamcloud.com/26463
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Niu Yawei <yawei.niu@intel.com>
Reviewed-by: Li Xi <lixi@ddn.com>
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Andreas Dilger [Sun, 19 Mar 2017 06:45:55 +0000 (00:45 -0600)]
LU-3308 mdc: allow setting readdir RPC size parameter
Allow the mdc.*.max_pages_per_rpc tunable to set the MDS bulk
readdir RPC size, rather than always using the default 1MB RPC
size. The tunable is set in the MDC, as it should be, rather
than in the llite superblock, which requires extra code just to
get it up from the MDC's connect_data only to send it back down.
The RPC size could be tuned independently if different types of
MDSes are used (e.g. local vs. remote).
Remove the md_op_data.op_max_pages and ll_sb_info.ll_md_brw_pages
fields that previously were used to pass the readdir size from
llite to mdc_read_page(). Reorder some 32-bit fields in md_op_data
to avoid struct holes.
Remove lprocfs_obd_rd_max_pages_per_rpc() as it is no longer used.
Move osc_obd_max_pages_per_rpc_seq_write() to obdclass along with
lprocfs_obd_max_pages_per_rpc_seq_show().
Remove debug messages from fid_flatten*() since this clutters up
the debug logs, and is redundant with other debug messages.
Fix test_24v's calculation for the number of RPCs being sent, so
that it will be correct even if the readdir RPC size is modified.
Register cleanup trap to avoid leaving lots of files behind.
Merge "slow" sanity test_24D into test_24v since they both need to
have many files in a single directory, which avoids duplication.
Change calc_osc_stats() and calc_llite_stats() into calc_stats()
to be used in test_24v and other tests. Similarly, clear_osc_stats()
and clear_llite_stats() are consolidated into clear_stats().
Signed-off-by: Andreas Dilger <andreas.dilger@intel.com>
Change-Id: Ibd814cea3b788129be8aca2866d1bb139b3ebbe5
Reviewed-on: https://review.whamcloud.com/26088
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Niu Yawei <yawei.niu@intel.com>
Reviewed-by: Lai Siyao <lai.siyao@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Andreas Dilger [Mon, 16 Jan 2017 21:24:10 +0000 (14:24 -0700)]
LU-9069 tests: improve output of sanity test_255a
Improve output of sanity.sh test_255a to contain more information.
Clean up the performance measurements and calculations to make the
test easier to read. The random_read_iops() helper routine might
be useful for other tests in the future as well.
The test does not (yet) work for ZFS, so it will skip the checks on
ZFS OSTs until "ladvise -a dontneed" is implemented for osd-zfs.
Test-Parameters: trivial
Signed-off-by: Andreas Dilger <andreas.dilger@intel.com>
Change-Id: Ia177d87e41266b058a6863bbf36108ad71ef9a00
Reviewed-on: https://review.whamcloud.com/24907
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Jian Yu <jian.yu@intel.com>
Reviewed-by: Saurabh Tandan <saurabh.tandan@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Doug Oucharek [Fri, 28 Apr 2017 21:44:26 +0000 (14:44 -0700)]
LU-9420 lnd: Remove a bad check which slipped in
When the patch for LU-5710 landed, a check for message size was landed
that should not have been. This check was part of a patch in LU-7650
which was later pulled because it broke things. LU-5718 picked up this
code via its many rebases (it too forever to land LU-5718 which is the
core problem here).
This patch removes that messaage size check.
Test-Parameters: trivial
Signed-off-by: Doug Oucharek <doug.s.oucharek@intel.com>
Change-Id: I3d114ec16cfbfd994efd9aee55e28a09159597be
Reviewed-on: https://review.whamcloud.com/26891
Tested-by: Jenkins
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Sonia Sharma <sonia.sharma@intel.com>
Reviewed-by: James Simmons <uja.ornl@yahoo.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Amir Shehata <amir.shehata@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Oleg Drokin [Wed, 26 Apr 2017 06:04:45 +0000 (02:04 -0400)]
LU-8587 osp: Remove unused osp_max_pages_per_rpc
It was never added to be seen from proc and in fact it is
completely unused anyway, so just remove it and the
osp_max_pages_per_rpc_seq_show function.
Change-Id: I4d0c582e4d48daff29795e95a7c8fa1f24340766
Signed-off-by: Oleg Drokin <oleg.drokin@intel.com>
Reviewed-on: https://review.whamcloud.com/24572
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: James Simmons <uja.ornl@yahoo.com>
Reviewed-by: Dmitry Eremin <dmitry.eremin@intel.com>
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Chris Hanna [Wed, 26 Apr 2017 13:43:55 +0000 (09:43 -0400)]
LU-9336 utils: prevent key clobber and clarify lgss_sk usage
Prevent lgss_sk from overwriting key value when modifying attributes.
Altered usage text to match, and clarified that the input source is
the key value, not a source of randomization for key generation.
Signed-off-by: Chris Hanna <hannac@iu.edu>
Change-Id: I87b9d59b65f3172b0425115441eaa1456489daeb
Reviewed-on: https://review.whamcloud.com/26838
Tested-by: Jenkins
Reviewed-by: Kit Westneat <kit.westneat@gmail.com>
Reviewed-by: Nathan Lavender <nblavend@iu.edu>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Saurabh Tandan [Tue, 25 Apr 2017 23:18:55 +0000 (16:18 -0700)]
LU-8103 tests: Skip sanity test_404 for interop mode
Skip sanity test_404 if the server version is older than
2.8.53 for interop mode.
This needs to be done due to the patch
http://review.whamcloud.com/19637/ which was introduced in
ticket LU-6601.
Test-Parameters: trivial
Signed-off-by: Saurabh Tandan <saurabh.tandan@intel.com>
Change-Id: I96271380cb7c3c09877de13c31c98367f74aad22
Reviewed-on: https://review.whamcloud.com/26827
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: James Nunez <james.a.nunez@intel.com>
Reviewed-by: Wei Liu <wei3.liu@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Amir Shehata [Mon, 17 Apr 2017 23:30:32 +0000 (16:30 -0700)]
LU-9119 lnet: Normalize ioctl interface
To avoid backwards compatibility issues between base MR
and Dynamic Discovery standardize the ioctl interface by
bringing in changes to the interface required by
Dynamic Discovery now.
Test-Parameters: trivial
Signed-off-by: Amir Shehata <amir.shehata@intel.com>
Change-Id: I014d74943b893ec24e3d42e1eb6824d755460c2b
Reviewed-on: https://review.whamcloud.com/26689
Tested-by: Jenkins
Reviewed-by: Olaf Weber <olaf@sgi.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Doug Oucharek <doug.s.oucharek@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Olaf Weber [Fri, 27 Jan 2017 15:15:50 +0000 (16:15 +0100)]
LU-9119 lnet: remove debug ioctl
Remove a debug ioctl that was added to allow for debug
messages from user space. However, the code is currently
not being used.
Test-Parameters: trivial
Signed-off-by: Amir Shehata <amir.shehata@intel.com>
Signed-off-by: Olaf Weber <olaf@sgi.com>
Change-Id: Ifd2bee73ef507bd07296af76dac1caf08ded9e64
Reviewed-on: https://review.whamcloud.com/26688
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Doug Oucharek <doug.s.oucharek@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Amir Shehata [Fri, 3 Feb 2017 02:05:20 +0000 (18:05 -0800)]
LU-9057 lnet: fix static analysis issues
Fixed a set of issues found while running static analysis.
Test-Parameters: trivial
Signed-off-by: Amir Shehata <amir.shehata@intel.com>
Change-Id: I22ddfdda86c979c7a300ab9df777efbdd5973ac5
Reviewed-on: https://review.whamcloud.com/26687
Tested-by: Jenkins
Reviewed-by: Olaf Weber <olaf@sgi.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Doug Oucharek <doug.s.oucharek@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Bob Glossman [Tue, 11 Apr 2017 13:19:20 +0000 (06:19 -0700)]
LU-9316 kernel: kernel update RHEL6.9 [2.6.32-696.1.1.el6]
Update RHEL6.9 kernel to 2.6.32-696.1.1.el6
Test-Parameters: trivial clientdistro=el6.9 mdsdistro=el6.9 \
ossdistro=el6.9 mdtfilesystemtype=ldiskfs \
ostfilesystemtype=ldiskfs testgroup=review-ldiskfs
Signed-off-by: Bob Glossman <bob.glossman@intel.com>
Change-Id: I5955da3a8953175d36e68877de52f9f9f2fd659b
Reviewed-on: https://review.whamcloud.com/26587
Tested-by: Jenkins
Reviewed-by: Yang Sheng <yang.sheng@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Minh Diep <minh.diep@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
John L. Hammond [Mon, 10 Apr 2017 14:17:20 +0000 (09:17 -0500)]
LU-9296 ptlrpc: set rq_sent when send fails due to -ENOMEM
In ptl_send_rpc() set rq_sent when we fail to send the RPC due
to insufficient memory, since this is what the upper layers expect.
Signed-off-by: John L. Hammond <john.hammond@intel.com>
Change-Id: Ia19c5d44e2999a9b347ec1088a7d448e1b548136
Reviewed-on: https://review.whamcloud.com/26470
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Jinshan Xiong <jinshan.xiong@intel.com>
Reviewed-by: Mike Pershin <mike.pershin@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Fan Yong [Sat, 3 Dec 2016 00:08:25 +0000 (08:08 +0800)]
LU-9294 mgs: ignore stale barrier locks
Currently, when MDT umount/crash, it may not notify the MGS
to cleanup related barrier lock and bitmap. Then subsequent
barrier operation may find some stale barrier locks. It is
unnecessary to return failure by mgs_barrier_glimpse_lock()
for such case, the caller will handle that properly.
Signed-off-by: Fan Yong <fan.yong@intel.com>
Change-Id: I341fd648da2eebbaab729145d1f06c420fce6455
Reviewed-on: https://review.whamcloud.com/26455
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Niu Yawei <yawei.niu@intel.com>
Reviewed-by: Lai Siyao <lai.siyao@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Wang Shilong [Wed, 19 Apr 2017 03:02:57 +0000 (11:02 +0800)]
LU-9355 obd: remove obsolete OBD_FL_LOCAL_MASK
From Andreas Dilger:
The OBD_FL_LOCAL_MASK support has not been used since
commit
e62f0a3c5b9 b=21980 cache ll_obdo_cache: Can't free all
objects which predates Jira. With the landing of patch
https://review.whamcloud.com/26463
LU-4017 quota: add project inherit attributes the handling of the "local"
flags is broken and since it is unused it is better to be removed entirely.
Change-Id: Iebd3de73f78f72851c5a664e72fa3d145729e1d6
Signed-off-by: Wang Shilong <wshilong@ddn.com>
Reviewed-on: https://review.whamcloud.com/26728
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Niu Yawei <yawei.niu@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Dmitry Eremin [Fri, 3 Mar 2017 17:11:40 +0000 (20:11 +0300)]
LU-9183 llite: handle removal the offset argument of direct_IO
In commit
c8b8e32d700fe943a935e435ae251364d016c497 the offset
argument for ->direct_IO() was removed.
Change-Id: I9f1cd5862dfcb40ad6b43b9da3072a852c550c49
Signed-off-by: Dmitry Eremin <dmitry.eremin@intel.com>
Reviewed-on: https://review.whamcloud.com/25822
Reviewed-by: James Simmons <uja.ornl@yahoo.com>
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
James Simmons [Sat, 8 Apr 2017 21:38:42 +0000 (17:38 -0400)]
LU-6401 uapi: fix up lustre_ostid.h and lustre_fid.h
Several inline functions in the header lustre_ostid.h
are using debug macros instead of returning proper errors.
Remove the debug macros and properly handle the returned
error codes. Place both UAPI headers lustre_fid.h and
lustre_ostid.h into the uapi directory.
Change-Id: Ic32afd05850b5bf02fb8de655cb1971eeb52a321
Signed-off-by: James Simmons <uja.ornl@yahoo.com>
Reviewed-on: https://review.whamcloud.com/24569
Reviewed-by: Fan Yong <fan.yong@intel.com>
Reviewed-by: Mike Pershin <mike.pershin@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Tested-by: Oleg Drokin <oleg.drokin@intel.com>
James Simmons [Sat, 8 Apr 2017 19:44:32 +0000 (15:44 -0400)]
LU-6401 uapi: turn lustre_ioctl.h into a proper UAPI header
Remove all the complex inline functions. Move all the user land
specific functions into the userland library. Unwind the kernel
specific functions and move obd_ioctl_is_valid() into the kernel
source file linux-module.c.
Change-Id: I91e69a21231f3effd23b191b6df9b5515a2ccc64
Signed-off-by: James Simmons <uja.ornl@yahoo.com>
Reviewed-on: https://review.whamcloud.com/24568
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Ben Evans <bevans@cray.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Kit Westneat [Fri, 18 Nov 2016 14:50:02 +0000 (09:50 -0500)]
LU-8851 nodemap: add uid/gid only flags to control mapping
This patch adds two flags to nodemaps which control whether or not
the nodemap should map UIDs, GIDs, or both. These flags can be
controlled via lctl as a new nodemap parameter map_mode, with values
both, uid_only, or gid_only.
Signed-off-by: Kit Westneat <kit.westneat@gmail.com>
Change-Id: I3efe6ff348d909c196a89273a0c9c046c56dbf1d
Reviewed-on: https://review.whamcloud.com/23853
Reviewed-by: Chris Hanna <hannac@iu.edu>
Reviewed-by: Sebastien Buisson <sbuisson@ddn.com>
Reviewed-by: Stephan Thiell <sthiell@stanford.edu>
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Andreas Dilger [Fri, 13 May 2016 01:53:54 +0000 (19:53 -0600)]
LU-5834 build: obsolete lustre-client if installing server
When installing the "lustre" (client+server) package, obsolete
older lustre-client package so that it can install without error.
Remove ancient Provides: and Obsoletes: lines from 1.2 days.
Fix remaining Provides: and Obsoletes: and package Group: lines
so they contain version numbers and names to make rpmlint happy.
Test-Parameters: trivial
Signed-off-by: Andreas Dilger <andreas.dilger@intel.com>
Change-Id: Ifdd1d7567ab03d0d1dfa599b592b6f28e09cab07
Reviewed-on: https://review.whamcloud.com/20162
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Minh Diep <minh.diep@intel.com>
Reviewed-by: Dmitry Eremin <dmitry.eremin@intel.com>
Emoly Liu [Mon, 24 Apr 2017 07:16:28 +0000 (15:16 +0800)]
LU-9311 tests: add sanity-pfl.sh test_13
This patch adds sanity-pfl.sh test_13 to verify the following fix:
https://review.whamcloud.com/#/c/26474/
test_13 uses 8 OSTs to make the LOVEA buffer bigger than the request
reply buffer, and writes data to the composite file to verify if the
layout write intent RPC resent by the client is reprocessed by
mdt_layout_change().
Test-parameters: trivial ostcount=8 testlist=sanity-pfl
Signed-off-by: Emoly Liu <emoly.liu@intel.com>
Change-Id: I7df6ff472c9cd879267912851d7288258db61d15
Reviewed-on: https://review.whamcloud.com/26700
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Niu Yawei <yawei.niu@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Doug Oucharek [Thu, 26 May 2016 18:26:17 +0000 (11:26 -0700)]
LU-7214 lnet: Allow min stats to be reset in peers and nis
Allow writes to the peers and nis LNet procfs files to
reset the mininum stat columns.
Signed-off-by: Doug Oucharek <doug.s.oucharek@intel.com>
Change-Id: Ic7aa647a154a424b35be1598266887d27efb4ab3
Reviewed-on: https://review.whamcloud.com/20470
Tested-by: Jenkins
Reviewed-by: Olaf Weber <olaf@sgi.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Amir Shehata <amir.shehata@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Wei Liu [Wed, 19 Apr 2017 16:49:05 +0000 (09:49 -0700)]
LU-9234 test: Skip test_70f if OSS version is older than 2.9.53
Add version check to replay-single test_70f. Skip it if OSS is older
than 2.9.53
Test-Parameters: trivial testlist=replay-single
Change-Id: I31424129e28c35eb0ac2bab9d8f321c13c978672
Signed-off-by: Wei Liu <wei3.liu@intel.com>
Reviewed-on: https://review.whamcloud.com/26739
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: James Nunez <james.a.nunez@intel.com>
Bobi Jam [Tue, 18 Apr 2017 13:58:59 +0000 (21:58 +0800)]
LU-9359 pfl: instantiate enough component at mdd_create_data
mknod creates file without layout, then truncate trigger the MDT
create OST objects, while current implementation only instantiate the
1st component, and when the truncate size locates in the other
components, the lvb size info is lost.
This patch makes MDT creates enought OST objects to cover the file's
size.
This patch fixes the misunderstanding of ost_pool::op->size, it
indicates the buffer size allocated instead of the array count.
Another issue fixed is that in lod_alloc_qos(), only fill in the ost
inused array when the lod_qos_declare_object_on() succeeds.
Signed-off-by: Bobi Jam <bobijam.xu@intel.com>
Change-Id: Ie66950e9b3d8cc009cca58f63936b275759211f1
Reviewed-on: https://review.whamcloud.com/26706
Tested-by: Jenkins
Reviewed-by: Niu Yawei <yawei.niu@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Bobi Jam [Fri, 14 Apr 2017 17:41:37 +0000 (01:41 +0800)]
LU-9346 lod: replay of PFL file open failure
During replay of PFL file open, lod_qos_parse_config()->
lod_use_defined_striping() initialed stripe LU-objects, but it keeps
the component's LCME_FL_INIT flag; and later in lod_striping_create()
these component will be skipped create OST objects, that fails the
replay, it should replay creating its OST objects.
Signed-off-by: Bobi Jam <bobijam.xu@intel.com>
Change-Id: Ic84374941df7a14b53e463f6117d5fbb9995c33d
Reviewed-on: https://review.whamcloud.com/26630
Reviewed-by: Niu Yawei <yawei.niu@intel.com>
Tested-by: Jenkins
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Fan Yong [Fri, 2 Dec 2016 20:44:27 +0000 (04:44 +0800)]
LU-9334 lfsck: reset trace file for upgrade case
With PFL introduced, the layout LFSCK on-disk trace file
will be changed, that is incompatibile with non-PFL case.
Use new layout LFSCK trace file magic for FPL case, that
will reset the layout LFSCK on-disk trace files for when
upgrade from non-PFL case.
Signed-off-by: Fan Yong <fan.yong@intel.com>
Change-Id: Id3b3fa4a89e5a408700653d8b759b76626c78912
Reviewed-on: https://review.whamcloud.com/26716
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Jinshan Xiong <jinshan.xiong@intel.com>
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Qian Yingjin [Mon, 17 Apr 2017 08:05:30 +0000 (16:05 +0800)]
LU-9356 osd-ldiskfs: add blk_plug when do bio
During 16MB bulk RPC I/O evaluation on rhel7, due to kernel
BIO_MAX_PAGES (256) limit, the 16MB IO is divided into 16 1MB
I/O submitting to underly block device one by one. And we found
that the SFA disk driver got lots of 1MB IOs.
To optimize the performance, this patch introduces blk_plug into
osd-ldiskfs when do bio, before submit IOs, it calls blk_start_plug,
after submit all 16MB IOs, calls blk_finish_plug, so that the 16MB
bulk IO will have more change to merge in the block evelvator
scheduler layer.
Signed-off-by: Qian Yingjin <qian@ddn.com>
Change-Id: If26db9f85baf97bc441cc4ad19d5c9f97bd3d7e5
Reviewed-on: https://review.whamcloud.com/26697
Tested-by: Jenkins
Reviewed-by: Alex Zhuravlev <alexey.zhuravlev@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Li Xi <lixi@ddn.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Bobi Jam [Fri, 14 Apr 2017 08:06:34 +0000 (16:06 +0800)]
LU-9342 pfl: handle uninited component in user defined LOVEA
It is possible that user defined LOVEA contains uninstantiated flag
like replay of create or layout extend or lfs swap, the partial LOVEA
will passed into
lod_declare_xattr_set()->lod_declare_striped_object()->
lod_prepare_create()->lod_qos_parse_config()->
lod_use_defined_striping(),
so that lod_use_defined_striping() also need to take care of
uninstantiated component entry.
Signed-off-by: Bobi Jam <bobijam.xu@intel.com>
Change-Id: I45d46b118bcde66f4604b80e2da51a808f381219
Reviewed-on: https://review.whamcloud.com/26619
Reviewed-by: Niu Yawei <yawei.niu@intel.com>
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Bobi Jam [Sat, 15 Apr 2017 03:57:18 +0000 (11:57 +0800)]
LU-9344 test: hung with sendfile_grouplock test12()
This is a makeshift fix.
When we hold a group lock of a file, there should no data written to
the file, since during the write IO, the file's layout could possibly
change, and the write IO will try to update its layout, which could
be blocked by itself.
Signed-off-by: Bobi Jam <bobijam.xu@intel.com>
Change-Id: Ia24310d509e8a93c3c1d849c54eacf789f40bfb8
Reviewed-on: https://review.whamcloud.com/26646
Tested-by: Jenkins
Reviewed-by: Jinshan Xiong <jinshan.xiong@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Wang Shilong [Fri, 21 Apr 2017 02:31:22 +0000 (22:31 -0400)]
LU-9339 ldiskfs: fix regression of macro check
include/linux/quota.h undefine @PRJQUOTA and make it
enum, ldiskfs macro check here will be negative, fix
it to use @HAVE_PROJECT_QUOTA, also wrap ext4-export-*.patch
to ext4-projid-feature-support.patch.
Change-Id: I415e366423408413dfb5b99dee98f7a30513f83f
Signed-off-by: Wang Shilong <wshilong@ddn.com>
Reviewed-on: https://review.whamcloud.com/26769
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Yang Sheng <yang.sheng@intel.com>
Reviewed-by: Li Xi <lixi@ddn.com>
Tested-by: Jenkins
Reviewed-by: James Simmons <uja.ornl@yahoo.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Bob Glossman [Thu, 30 Mar 2017 16:27:47 +0000 (09:27 -0700)]
LU-9276 kernel: kernel update [SLES12 SP1 3.12.69-60.64.35]
Update target and kernel_config files for new version
Test-Parameters: clientdistro=sles12 testgroup=review-ldiskfs \
mdsdistro=sles12 ossdistro=sles12 mdtfilesystemtype=ldiskfs \
ostfilesystemtype=ldiskfs
Signed-off-by: Bob Glossman <bob.glossman@intel.com>
Change-Id: Ied31b0a3601ac6e1db06da3e0167b875d1f2200a
Reviewed-on: https://review.whamcloud.com/26286
Reviewed-by: Minh Diep <minh.diep@intel.com>
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Yang Sheng <yang.sheng@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Dmitry Eremin [Thu, 2 Mar 2017 19:48:44 +0000 (22:48 +0300)]
LU-9183 llite: handle removal of parent argument of ->d_compare()
In commit
6fa67e707559303e086303aeecc9e8b91ef497d5 the parent
parameter for ->d_compare() was removed.
Change-Id: Ia241619c3ade13036154973a19f44a2083a9bc53
Signed-off-by: Dmitry Eremin <dmitry.eremin@intel.com>
Reviewed-on: https://review.whamcloud.com/25818
Reviewed-by: James Simmons <uja.ornl@yahoo.com>
Reviewed-by: Jinshan Xiong <jinshan.xiong@intel.com>
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Dmitry Eremin [Thu, 2 Mar 2017 19:39:37 +0000 (22:39 +0300)]
LU-9183 libcfs: handle get_user_pages() with gup_flags
In commit
c8fe4609827aedc9c4b45de80e7cdc8ccfa8541b the arguments
write/force for get_user_pages() were replaced with gup_flags.
Change-Id: I64b3e69ee736d1f3be48f2c4c8fe1af1dc14e857
Signed-off-by: Dmitry Eremin <dmitry.eremin@intel.com>
Reviewed-on: https://review.whamcloud.com/25817
Reviewed-by: James Simmons <uja.ornl@yahoo.com>
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Steve Guminski [Mon, 6 Mar 2017 21:05:22 +0000 (16:05 -0500)]
LU-9152 mgc: Remove unnecessary checks for config_log_put()
Because config_log_put() now checks if its parameter is NULL, it
is unnecessary to perform the check prior to calling it. This patch
removes the redundant checks.
Signed-off-by: Steve Guminski <stephenx.guminski@intel.com>
Change-Id: Id6b1fccebd5bc53a29bc364b9a3c47956649920a
Reviewed-on: https://review.whamcloud.com/25854
Tested-by: Jenkins
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: James Simmons <uja.ornl@yahoo.com>
Fan Yong [Sat, 3 Dec 2016 02:18:35 +0000 (10:18 +0800)]
LU-9306 tests: more debug info for hsm test_24d
More debug information for sanity-hsm test_24d.
Drop unnecessary condition check for mdt_hsm_cdt_stop().
Test-Parameters: testlist=sanity-hsm,sanity-hsm,sanity-hsm,sanity-hsm
Signed-off-by: Fan Yong <fan.yong@intel.com>
Change-Id: Ib4a43693d364e208468937d15e3c48a1b1afb17a
Reviewed-on: https://review.whamcloud.com/26770
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Niu Yawei <yawei.niu@intel.com>
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Ned Bass [Sat, 18 Mar 2017 00:55:46 +0000 (17:55 -0700)]
LU-8424 osd-zfs: ZFS macro DN_MAX_BONUSLEN is deprecated
The ZFS macro DN_MAX_BONUSLEN was deprecated as of ZFS 0.7.0. Lustre
should instead use the compatibility wrappers such as
osd_dmu_object_alloc() and osd_zap_create_flags(). The reason for the
API change is that ZFS 0.7.0 adds support for variable length dnodes,
so the maximum bonus length should not be treated as a fixed
constant. The maximum bonus length may vary by dnode and by dataset,
and it should be derived accordingly.
This change:
- Adds an additional compatibility function osd_obj_bonuslen(obj) to
obtain a maximum bonus length given an osd_object.
- Updates code that uses the deprecated macro to instead use
appropriate compatibility interfaces.
- Removes the definition of DN_MAX_BONUSLEN that was added in commit
49fc02fb738e9420ab10c5a7d41534c7a55b8ea0 to ensure that future
builds using the deprecated macro will fail.
- Adds DN_MAX_BONUSLEN and DN_OLD_MAX_BONUSLEN to the list of
deprecated interfaces in checkpatch.pl.
Signed-off-by: Ned Bass <bass6@llnl.gov>
Change-Id: I1fcc84e55b39ca49a88acb909b5e3294f3b46723
Reviewed-on: https://review.whamcloud.com/26078
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Alex Zhuravlev <alexey.zhuravlev@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Nathaniel Clark <nathaniel.l.clark@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Bob Glossman [Wed, 12 Apr 2017 14:49:38 +0000 (07:49 -0700)]
LU-9323 kernel: kernel update RHEL7.3 [3.10.0-514.16.1.el7]
update RHEL 7.3 kernel to 3.10.0-514.16.1.el7
Signed-off-by: Bob Glossman <bob.glossman@intel.com>
Change-Id: Ia53f001b7b277a20349b23ffdf54e0b0dbc1d065
Reviewed-on: https://review.whamcloud.com/26590
Tested-by: Jenkins
Reviewed-by: Minh Diep <minh.diep@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Yang Sheng <yang.sheng@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
James Simmons [Thu, 13 Apr 2017 19:54:23 +0000 (15:54 -0400)]
LU-9327 test: handle nodes lacking numbers in names for hostlist_expand.
Currently hostlist_expand() in test-framework assumes all node names will
contain numbers to sort them by. It is possible to have nodes that instead
are named alphabetical in nature (nodea, nodeb, ...). Handle this additional
case.
Change-Id: I014d7934a869192b3bc1908b4b0d559724d443c9
Signed-off-by: James Simmons <uja.ornl@yahoo.com>
Reviewed-on: https://review.whamcloud.com/26560
Tested-by: Jenkins
Reviewed-by: Jian Yu <jian.yu@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Minh Diep <minh.diep@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Jinshan Xiong [Tue, 11 Apr 2017 11:08:16 +0000 (19:08 +0800)]
LU-9008 pfl: Read should not trigger layout write intent
In lov_io_rw_iter_init(), only write not read operation should
trigger layout write intent.
For append write, it has to make sure all uninited components
are instantiated.
Page mkwrite should also trigger write intent.
Fixed sanity-pfl test cases as well - in order to instantiate
a component by truncate, the truncated size must be within the
extent of component.
Signed-off-by: Jinshan Xiong <jinshan.xiong@intel.com>
Signed-off-by: Bobi Jam <bobijam.xu@intel.com>
Change-Id: I3c5257c5799ae0df35a572b9ee981f583260e8ec
Reviewed-on: https://review.whamcloud.com/26499
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Niu Yawei <yawei.niu@intel.com>
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Jadhav Vikram [Tue, 4 Apr 2017 04:08:11 +0000 (09:38 +0530)]
LU-9292 utils: handle partitioned MD arrays correctly
mount.lustre doesn't handle partitioned MD arrays correctly.
The function set_blockdev_tunables doesn't correctly parse
the newer names of partitioned MD devices (e.g /dev/md0p2).
It is written to parse the older partitioned MD devices
(e.g /dev/md_d2p3). This means that MD partions like /dev/md0p2
do not receive MD-level tunings that non-partitioned MD devices
get (specifically stripe_cache_size), leading to a large
performance hit. so fix is to handle newer and older MD device
names.
Seagate-bug-id: MRP-2608
Signed-off-by: Jadhav Vikram <jadhav.vikram@seagate.com>
Reviewed-by: Ashish Purkar <ashish.purkar@seagate.com>
Reviewed-by: Christopher Walker <chris.walker@seagate.com>
Tested-by: Elena V. Gryaznova <elena.gryaznova@seagate.com>
Reviewed-by: Andriy Skulysh <andriy.skulysh@seagate.com>
Change-Id: Ia0abe63e725e3a70d2561960faa1bc48981f2fd0
Reviewed-on: https://review.whamcloud.com/26399
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Nathaniel Clark <nathaniel.l.clark@intel.com>
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Kit Westneat [Mon, 27 Mar 2017 16:27:27 +0000 (12:27 -0400)]
LU-9258 nodemap: group quota ID not properly mapped
This patch fixes a typo in which group quota IDs were mapped as user
IDs for quota commands.
Test-Parameters: trivial
Signed-off-by: Kit Westneat <kit.westneat@gmail.com>
Change-Id: I9458b4af04a3638fb8ccf3bd2a96b30305021514
Reviewed-on: https://review.whamcloud.com/26209
Tested-by: Jenkins
Reviewed-by: Stephan Thiell <sthiell@stanford.edu>
Reviewed-by: Emoly Liu <emoly.liu@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Qian Yingjin [Mon, 20 Mar 2017 02:33:54 +0000 (10:33 +0800)]
LU-9227 nrs: Rate change of a TBF rule loses control
In some test cases, i.e.
start dd_0 {dd.0} 1000
dd if=/dev/zero of=/mnt/lustre/test bs=1M count=100
start dd_1000 {dd.1000} 1000
After ran above commands, changing rate of dd_0 take no effect.
The reason is that starting rule dd_1000 increases the sequence,
but the sequence number of the class with id "dd.0" does not
change accordingly, resulting in failure of rate change.
This patch fixes this problem.
Signed-off-by: Qian Yingjin <qian@ddn.com>
Change-Id: I9cdd469fd57dea692b86285cc26040a117b120ad
Reviewed-on: https://review.whamcloud.com/26085
Reviewed-by: Emoly Liu <emoly.liu@intel.com>
Reviewed-by: Li Xi <lixi@ddn.com>
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Andreas Dilger [Mon, 27 Feb 2017 00:32:40 +0000 (17:32 -0700)]
LU-9153 llog: update llog print format to use FIDs
Print llog identifiers using FIDs instead of OSTID format, which has
been deprecated since 2.3.
Signed-off-by: Andreas Dilger <andreas.dilger@intel.com>
Change-Id: Id012c059e8151e2a78086579150f04b1b05cab07
Reviewed-on: https://review.whamcloud.com/25640
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Henri Doreau <henri.doreau@cea.fr>
Reviewed-by: Faccini Bruno <bruno.faccini@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Bob Glossman [Wed, 22 Feb 2017 17:32:41 +0000 (09:32 -0800)]
LU-9148 kernel: kernel update RHEL6.8 [2.6.32-642.15.1.el6]
Update RHEL6.8 kernel to 2.6.32-642.15.1.el6
Test-Parameters: trivial clientdistro=el6.8 mdsdistro=el6.8 \
ossdistro=el6.8 mdtfilesystemtype=ldiskfs \
ostfilesystemtype=ldiskfs testgroup=review-ldiskfs
Signed-off-by: Bob Glossman <bob.glossman@intel.com>
Change-Id: Iaac7b99246b9e9d3204c3545aaf0b9d2fd8d3185
Reviewed-on: https://review.whamcloud.com/25633
Tested-by: Jenkins
Reviewed-by: Minh Diep <minh.diep@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: James Nunez <james.a.nunez@intel.com>
Reviewed-by: Yang Sheng <yang.sheng@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Nathaniel Clark [Tue, 7 Feb 2017 13:55:06 +0000 (08:55 -0500)]
LU-8457 pacemaker: Update healthLNET to 0.99.4
Fixed minor issue with lnet connectivity
Fix License header.
Test-Parameters: trivial
Signed-off-by: Nathaniel Clark <nathaniel.l.clark@intel.com>
Change-Id: I074b93e2e3ea29e608a6f1b46600556a1b255438
Reviewed-on: https://review.whamcloud.com/25297
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Malcolm Cowe <malcolm.j.cowe@intel.com>
Vladimir Saveliev [Wed, 1 Feb 2017 15:29:56 +0000 (23:29 +0800)]
LU-8397 mgc: add comma-separated nids for primary MGS
In lustre_start_mgc(), if the primary MGS has multiple nids separated by
commas, then only the first nid is added, and the other nids are not added.
This patch fixes the above issue by adding the nids until the end of the
primary MGS nid list or hitting the first failover MGS nid.
Test-Parameters: combinedmdsmgs=false envdefinitions=ONLY=77 testlist=conf-sanity
Test-Parameters: envdefinitions=ONLY=77 testlist=conf-sanity
Signed-off-by: Vladimir Saveliev <vladimir.saveliev@seagate.com>
Signed-off-by: Jian Yu <jian.yu@intel.com>
Seagate-bug-id: MRP-2930
Change-Id: I3e84786bfc08767c75a133affb4f86325d789d6e
Reviewed-on: https://review.whamcloud.com/21308
Tested-by: Jenkins
Reviewed-by: Bobi Jam <bobijam@hotmail.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Emoly Liu <emoly.liu@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Andreas Dilger [Wed, 19 Apr 2017 02:53:00 +0000 (20:53 -0600)]
LU-9354 utils: fix lfs.c compile for PPC
Fix the lfs_setstripe_args parameter to use "unsigned long long"
for lsa_comp_end so that when passed to llapi_parse_size() it is
the correct type. PPC annoyingly uses "unsigned long" for __u64
while most other architectures use "unsigned long long".
Test-Parameters: trivial
Signed-off-by: Andreas Dilger <andreas.dilger@intel.com>
Change-Id: I27b724d5eab5f4669bde1f5fb44c75a1051ba7bd
Reviewed-on: https://review.whamcloud.com/26727
Tested-by: Jenkins
Reviewed-by: Minh Diep <minh.diep@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Reviewed-by: James Simmons <uja.ornl@yahoo.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Bobi Jam [Mon, 10 Apr 2017 17:50:04 +0000 (01:50 +0800)]
LU-9311 pfl: shouldn't reprocess done/no-op resent request
When the LOVEA buffer is bigger than the request reply buffer, the
client will resend the layout write intent RPC, and
mdt_layout_change() should not reprocess it since the 2nd process
will try to cancel the 1st granted CR lock, while client has not
get it granted yet because of the reply buffer shortage resend the
RPC.
There is another layout change resent case: the client's job has been
done by another client, referring lod_declare_layout_change -EALREADY
case, and it became a operation w/o transaction, so we should not do
the layout change, otherwise mdt_layout_change() will try to cancel
the granted server CR lock whose remote counterpart is still in hold
on the client, and a deadlock ensues.
This patch also adjusts some debug messages, makes dump_lsm() dump
uninstantiated component stripe info.
Signed-off-by: Bobi Jam <bobijam.xu@intel.com>
Change-Id: I9b063ee54d57c233eca3250502a2707997892898
Reviewed-on: https://review.whamcloud.com/26474
Reviewed-by: Niu Yawei <yawei.niu@intel.com>
Tested-by: Jenkins
Reviewed-by: Emoly Liu <emoly.liu@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Bobi Jam [Thu, 13 Apr 2017 03:36:58 +0000 (11:36 +0800)]
LU-9324 test: get layout parameters
Add test framework functions get_layout_param()/parse_layout_param
to get plain/PFL file/dir layout parameters, so that the layout
parameters can be used later.
Test-Parameters: trivial testlist=sanity-pfl
Signed-off-by: Bobi Jam <bobijam.xu@intel.com>
Change-Id: I4f43a5b4d23e353660114b7ff299a97afc405a07
Reviewed-on: https://review.whamcloud.com/26578
Tested-by: Jenkins
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: James Nunez <james.a.nunez@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Nathaniel Clark [Mon, 27 Mar 2017 16:55:31 +0000 (12:55 -0400)]
LU-9168 spec: New lustre-resource-agents rpm
This rpm bundles up all resource agents for Lustre into a single RPM.
Signed-off-by: Nathaniel Clark <nathaniel.l.clark@intel.com>
Change-Id: Ifc4422f11e9d6b4c1e03ef85fc78e35d6fcb0316
Reviewed-on: https://review.whamcloud.com/26229
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Zhiqi Tao <zhiqi.tao@intel.com>
Reviewed-by: Gabriele Paciucci <gabriele.paciucci@intel.com>
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Fan Yong [Sat, 8 Oct 2016 06:00:24 +0000 (14:00 +0800)]
LU-8840 osp: handle EA cache properly (2)
For success case, dt_xattr_get() should return the EA size
instead of zero. If such EA does not exist, return -ENODATA.
More code cleanup for OSP EA cache to avoid potential reference
leak, buffer overflow, and so on.
Signed-off-by: Fan Yong <fan.yong@intel.com>
Change-Id: I85d7c7a2cafd50334f2ea0634f5e2b21c0b3908e
Reviewed-on: https://review.whamcloud.com/25207
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Lai Siyao <lai.siyao@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Nathaniel Clark [Wed, 19 Apr 2017 14:54:51 +0000 (10:54 -0400)]
LU-2155 utils: Enable ldiskfs mmp on tunefs failover add
Enable Multi-Mount Protection in ldiskfs when adding a failover peer
via tunefs. MMP is enabled in mkfs if failover is configured
initially, but it wasn't if failover was added later.
Signed-off-by: Nathaniel Clark <nathaniel.l.clark@intel.com>
Change-Id: Ie21ef5324de240afe0fd760cb4a9b3e1b4165064
Reviewed-on: https://review.whamcloud.com/26758
Tested-by: Jenkins
Reviewed-by: Zhiqi Tao <zhiqi.tao@intel.com>
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Andreas Dilger [Tue, 18 Apr 2017 07:35:54 +0000 (01:35 -0600)]
LU-8998 utils: allow "--component-*" to be abbreviated
Allow "--component-foo" arguments to be abbreviated to "--comp-foo"
to simplify userspace, especially for options like "--comp-flags"
that do not have a short option.
Test-Parameters: trivial
Signed-off-by: Andreas Dilger <andreas.dilger@intel.com>
Change-Id: I8c434dd68c042efefc2df93d0f2d7b9bc33ebbe5
Reviewed-on: https://review.whamcloud.com/26699
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Niu Yawei <yawei.niu@intel.com>
Reviewed-by: Jinshan Xiong <jinshan.xiong@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Niu Yawei [Mon, 17 Apr 2017 12:58:32 +0000 (08:58 -0400)]
LU-9351 pfl: setstripe to existing file
It's ok to setstripe to an existing file when the file has no
LOVEA created yet, so we should not use exlusive open for the
'lfs setstripe'.
Signed-off-by: Niu Yawei <yawei.niu@intel.com>
Change-Id: I2090ba2940b391f7853cc06ccdb9cd842ad6f984
Reviewed-on: https://review.whamcloud.com/26655
Tested-by: Jenkins
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Jinshan Xiong <jinshan.xiong@intel.com>
Bobi Jam [Thu, 13 Apr 2017 16:37:25 +0000 (00:37 +0800)]
LU-9335 pfl: calculate PFL file LOVEA correctly
PFL file could contain uninstantiated component, so it could still
keeps the specified -1 stripe count,
lov_mds_md_size()/lov_user_md_size() should heed this case,
otherwise its LOVEA size could be errneous big.
Signed-off-by: Bobi Jam <bobijam.xu@intel.com>
Change-Id: Ied4bf4531f0b0ac9fdefc9efef3c97ae5ae449f4
Reviewed-on: https://review.whamcloud.com/26597
Tested-by: Jenkins
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Niu Yawei <yawei.niu@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Di Wang [Mon, 9 Jan 2017 15:07:36 +0000 (10:07 -0500)]
LU-8922 lod: check master stripes properly
When creating striped directory, it should
first check if the stripe has been created
on the current MDT, otherwise it might create
duplicate stripes on the master MDT, especially
when one MDT is deactived, and specified stripes
is larger than the active MDTs.
Test-Parameters: envdefinitions=ONLY=50 testlist=conf-sanity,conf-sanity,conf-sanity
Test-Parameters: envdefinitions=ONLY=50 testlist=conf-sanity,conf-sanity,conf-sanity
Signed-off-by: Di Wang <di.wang@intel.com>
Change-Id: Id3bba2817c4c7c9584f9129e32555f0f676b3364
Reviewed-on: https://review.whamcloud.com/24776
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Fan Yong <fan.yong@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Liang Zhen [Tue, 28 Oct 2014 03:34:26 +0000 (11:34 +0800)]
LU-5718 o2iblnd: multiple sges for work request
In current protocol, lnet router cannot align buffer for rdma,
o2iblnd may run into "too fragmented RDMA" issue while routing
non-page-aligned IO larger than 512K, because each page will
be splited into two fragments by kiblnd_init_rdma().
With this patch, o2iblnd can have multiple sges for each work
request, and combine multiple remote fragments of the same page
into one work request to resovle the "too fragmented RDMA" issue.
Signed-off-by: Liang Zhen <liang.zhen@intel.com>
Change-Id: Id57a74dc92801b012956ab785233aa87cac14263
Reviewed-on: https://review.whamcloud.com/12451
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Amir Shehata <amir.shehata@intel.com>
Reviewed-by: Nathaniel Clark <nathaniel.l.clark@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Fan Yong [Tue, 29 Nov 2016 13:18:08 +0000 (21:18 +0800)]
LU-9334 lfsck: object leak in lfsck_load_one_trace_file
In lfsck_load_one_trace_file(), if we successfully load or
create the object via local_index_find_or_create(), but the
subsequent do_index_try() failed, then we need to release
such object, otherwise it will be left there.
Signed-off-by: Fan Yong <fan.yong@intel.com>
Change-Id: Ic3c7db9239e0d10a5cf6fc2254a4c414f4cd007f
Reviewed-on: https://review.whamcloud.com/26703
Tested-by: Jenkins
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Lai Siyao <lai.siyao@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Andreas Dilger [Tue, 4 Apr 2017 23:22:51 +0000 (17:22 -0600)]
LU-1032 build: fix typo in lustre-dkms.spec changelog
Fix typo in the lustre-dkms.spec.in ChangeLog which has an
invalid date, and causes a build warning.
Test-Parameters: trivial
Signed-off-by: Andreas Dilger <andreas.dilger@intel.com>
Change-Id: I6029c93928f84ff59c8798d2f8115e872c4cab07
Reviewed-on: https://review.whamcloud.com/26358
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Faccini Bruno <bruno.faccini@intel.com>
Reviewed-by: Nathaniel Clark <nathaniel.l.clark@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Andreas Dilger [Sun, 23 Apr 2017 04:19:00 +0000 (22:19 -0600)]
LU-9384 tests: skip 2.9 filesystem images on 2.10
The 2.9 disk images upgraded to 2.10 are causing common test failures
due to small differences in space usage. Disable testing these images
until a proper fix is available.
Test-Parameters: trivial envdefinitions=ONLY=32 testlist=conf-sanity
Signed-off-by: Andreas Dilger <andreas.dilger@intel.com>
Change-Id: I64d14ac6b079c50ee234ff886c9a80d9213ebbe5
Reviewed-on: https://review.whamcloud.com/26789
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Chris Horn [Thu, 12 Mar 2015 23:25:14 +0000 (18:25 -0500)]
LU-6283 ptlrpc: Implement NRS Delay Policy
The NRS Delay policy seeks to perturb the timing of request processing
at the PtlRPC layer, with the goal of simulating high server load, and
finding and exposing timing related problems. When this policy is
active, upon arrival of a request the policy will calculate an offset,
within a defined, user-configurable range, from the request arrival
time, to determine a time after which the request should be handled.
The request is then stored using the cfs_binheap implementation,
which sorts the request according to the assigned start time.
Requests are removed from the binheap for handling once their start
time has been passed.
The behavior of the policy can be controlled via three proc files
which can be written to via lctl similar to other policies.
nrs_delay_min: Controls the minimum amount of time, in seconds, that a
request will be delayed by this policy. The default is 5 seconds.
nrs_delay_max: Controls the maximum amount of time, in seconds, that a
request will be delayed by this policy. The default is 300 seconds.
nrs_delay_pct: Control the percentage of requests that will be delayed
by this policy. The default is 100. Note, when a request is not
selected for handling by the delay policy due to this variable then
the request will be handled by whatever fallback policy is defined
for that service. If no other fallback policy is defined then the
request will be handled by the FIFO policy.
Some examples:
lctl set_param *.*.*.nrs_delay_min=reg_delay_min:5, to set the regular
request minimum delay on all PtlRPC services to 5 seconds.
lctl set_param *.*.*.nrs_delay_min=hp_delay_min:2, to set the
high-priority request minimum delay on all PtlRPC services to 2
seconds.
lctl set_param *.*.ost_io.nrs_delay_min=8, to set both the regular and
high-priority request minimum delay of the ost_io service to 8
seconds.
lctl set_param *.*.*.nrs_delay_max=reg_delay_max:20, to set the
regular request maximum delay on all PtlRPC services to 20 seconds.
lctl set_param *.*.*.nrs_delay_max=hp_delay_max:10, to set the
high-priority request maximum delay on all PtlRPC services to 10
seconds.
lctl set_param *.*.ost_io.nrs_delay_max=35, to set both the regular
and high-priority request maximum delay of the ost_io service to 35
seconds.
lctl set_param *.*.*.nrs_delay_pct=reg_delay_pct:5, to delay 5
percent of regular requests on all PtlRPC services.
lctl set_param *.*.*.nrs_delay_pct=hp_delay_pct:2, to delay 2 percent
of high-priority requests on all PtlRPC services.
lctl set_param *.*.ost_io.nrs_delay_pct=8, to delay 8 percent of both
regular and high-priority requests of the ost_io service.
Signed-off-by: Chris Horn <hornc@cray.com>
Change-Id: Iab50a639900adf31893c7b1fe83658932fd59db1
Reviewed-on: https://review.whamcloud.com/14701
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Henri Doreau <henri.doreau@cea.fr>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Fan Yong [Mon, 7 Mar 2016 21:34:36 +0000 (05:34 +0800)]
LU-9319 statahead: skip agl for the file in restoring
In case of restore, the MDT has the right size and has already sent
it back without granting the layout lock, inode is up-to-date. Then
AGL (async glimpse lock) is useless.
Also to glimpse we need the layout, in case of a running restore the
MDT holds the layout lock so the glimpse will block up to the end of
restore (statahead/agl will block).
Signed-off-by: Fan Yong <fan.yong@intel.com>
Change-Id: Iaaf138a28671c3eccfb05b08ce8d7364423256a1
Reviewed-by: Lai Siyao <lai.siyao@intel.com>
Reviewed-by: Jinshan Xiong <jinshan.xiong@intel.com>
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Reviewed-on: https://review.whamcloud.com/26501
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Emoly Liu [Fri, 14 Apr 2017 06:18:49 +0000 (14:18 +0800)]
LU-9332 tests: fix pool_list issue in conf-sanity test_82b
In conf-sanity.sh test_82b, if we use separated MGS and MDS nodes,
MDS needs more time to list the OSTs in pool, so use wait_update
instead of the pool_list command.
Signed-off-by: Emoly Liu <emoly.liu@intel.com>
Change-Id: I8f7281d44a12731d7ae2d1f8fa9bad163086cecc
Reviewed-on: https://review.whamcloud.com/26582
Tested-by: Jenkins
Reviewed-by: Parinay Kondekar <parinay.kondekar@seagate.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Jian Yu <jian.yu@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
James Simmons [Sun, 9 Apr 2017 20:27:02 +0000 (16:27 -0400)]
LU-9086 gss: handle specific mgc binding
Recent work has made mgc config log handle more than one MGC
existing on a client node which is the case when more than
one lustre file system is mounted. This broke sptlrpc
handling due to cfg_instance no longer being NULL. Since
this is the case we remove the cfg_instance NULL test in
class_config_llog_handler.
The second problem exposed was due to cfg_obdname changing from
the file system name to the actual mgc abstract client device.
Since target2fsname() was only checking fsname-[MDT|OST]XXXX we
could use this function with the obd_uuid field which has the
format of fsname-[MDT|OST]XXXX_UUID. For the case of obd names
with the format fsname-XXXXXXX those obd device have an obd_uuid
that is just some random long hexadecimal string. To handle the
obd_uuid case use the proper obd_uuid2fsname() function instead
of target2fsname(). Rename target2fsname() to obdname2fsname()
since this what it now does. Add handling the fsname-XXXXXXXX
obd name format to obdname2fsname().
Change-Id: I6c3235b3e2d41ef955b98d30aec76d47346c2369
Signed-off-by: James Simmons <uja.ornl@yahoo.com>
Reviewed-on: https://review.whamcloud.com/25959
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Mike Pershin <mike.pershin@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Nathaniel Clark [Tue, 28 Feb 2017 14:22:22 +0000 (09:22 -0500)]
LU-8458 pacemaker: Resource to manage Lustre Target
A new script to monitor health and mount/umount lustre target.
pcs resource create [Resource Name] ocf:lustre:Lustre
target=/dev/mapper/mpatha mountpoint=/mnt/MGS
This script should be located in /usr/lib/ocf/resource.d/lustre/
of both the Lustre servers with permission 755.
Test-Parameters: trivial
Signed-off-by: Nathaniel Clark <nathaniel.l.clark@intel.com>
Change-Id: I1cfa42604af3c777c1b8d6ed344e22ceccfceb62
Reviewed-on: https://review.whamcloud.com/25664
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Malcolm Cowe <malcolm.j.cowe@intel.com>
Reviewed-by: Zhiqi Tao <zhiqi.tao@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
CongXu [Fri, 10 Feb 2017 22:36:33 +0000 (14:36 -0800)]
LU-9083: Add scripts to collect and analyze Lustre RPC traces
A set of scripts to collect and analyze RPC traces in a Lustre file system.
These scripts turn the Lustre debug-daemon on and off, before and after
launching a job respectively. They also allow users to parse and visualize
the logged RPC traces.
Test-Parameters: trivial
Change-Id: I6ee598ff6e49b2a7406354172b10bd295e801cb0
Signed-off-by: CongXu <cong.xu@intel.com>
Reviewed-on: https://review.whamcloud.com/25395
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Omkar Kulkarni <omkar.kulkarni@intel.com>
Bruno Faccini [Fri, 3 Feb 2017 16:55:37 +0000 (17:55 +0100)]
LU-9075 mdt: avoid race causing mdt_coordinator_cb() err msgs
This patch mainly moves mdt_agent_record_update() call before
mdt_cdt_remove_request() in mdt_hsm_update_request_state(), to
avoid the frequent couple of
"(mdt_coordinator.c:1473:mdt_hsm_update_request_state()) ...
Cannot find running request for cookie ..."
and
"(mdt_coordinator.c:339:mdt_coordinator_cb()) ...
cannot cleanup timed out request ..."
error msgs, likely to concern active requests that have completed
and thus that have already been removed from memory in
mdt_hsm_update_request_state() (using mdt_cdt_remove_request() and
in the context of a MDT thread handling CT's MDS_HSM_PROGRESS
requests), but the corresponding action LLOG record update is stuck
awaiting for CDT to give-back cdt_llog_lock in
mdt_agent_record_update().
Others related but minor changes are, use of arr_req_change instead of
arr_req_create to more accuratelly determine if a request exceeds the
timeout, and change main debug msg in mdt_hsm_update_request_state()
to reflect if action LLOG record update will occur or not.
Signed-off-by: Bruno Faccini <bruno.faccini@intel.com>
Change-Id: I043813f1ff11a7e9e99c534fa8560a35e2c52543
Reviewed-on: https://review.whamcloud.com/25243
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Henri Doreau <henri.doreau@cea.fr>
Reviewed-by: Quentin Bouget <quentin.bouget@cea.fr>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Elena Gryaznova [Fri, 30 Sep 2016 18:15:57 +0000 (21:15 +0300)]
LU-8641 tests: tests: speedup metabench
With huge number of nfiles created by metabench
rm -rf called at the end of run_metabench() takes a very long
time.
Sine metabench cleanup is also very slow, metabench
execution without -k option is useless to use also.
Patch adds mbench_CLEANUP mode, default value is true.
When set to false -- metabench testdir is just renamed to avoid
files/dirs deletion.
Seagate-bug-id: MRP-3454
Signed-off-by: Elena Gryaznova <elena.gryaznova@seagate.com>
Reviewed-by: Alexander Zarochentsev <alexander.zarochentsev@seagate.com>
Reviewed-by: Vitaly Fertman <vitaly.fertman@seagate.com>
Change-Id: Iaa130785d27dc3a87f642723fcaad6dc5e748440
Reviewed-on: https://review.whamcloud.com/22852
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.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>
Andreas Dilger [Mon, 27 Feb 2017 10:54:28 +0000 (03:54 -0700)]
LU-8832 tests: fix module loading for Parallels
It isn't possible to insmod modules directly from Parallels prl_fs
mounting a host filesystem when testing directly from a build tree
(insmod from an RPM installed in the root image filesystem works).
Instead, copy the locally-built module into /tmp before loading it.
Use load_module() in sanity.sh test_180b and run-llog.sh.
Test-Parameters: trivial
Signed-off-by: Andreas Dilger <andreas.dilger@intel.com>
Change-Id: Id0123059e8151e2a78086579150f04b1b05cab07
Reviewed-on: https://review.whamcloud.com/25660
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: James Nunez <james.a.nunez@intel.com>
Reviewed-by: Jinshan Xiong <jinshan.xiong@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Niu Yawei [Fri, 14 Apr 2017 05:18:21 +0000 (01:18 -0400)]
LU-9280 osd-zfs: don't mark existing on failed creation
In osd_object_create(), existing flag shouldn't be set on failure.
Signed-off-by: Niu Yawei <yawei.niu@intel.com>
Change-Id: I77dd96f0e3c562c6bef659dc0af405a1ce488805
Reviewed-on: https://review.whamcloud.com/26617
Tested-by: Jenkins
Reviewed-by: Alex Zhuravlev <alexey.zhuravlev@intel.com>
Reviewed-by: Fan Yong <fan.yong@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Oleg Drokin [Sat, 22 Apr 2017 01:12:21 +0000 (21:12 -0400)]
New tag 2.9.56
Change-Id: I20adf1083b0c081ce818941c87b40f64631ad8e4
Signed-off-by: Oleg Drokin <oleg.drokin@intel.com>
Jim Casper [Fri, 14 Apr 2017 18:19:52 +0000 (14:19 -0400)]
LU-8333 test: Add replay-dual 21b to ALWAYS_EXCEPT
Replay-dual's subtest test_21b has been failing more than half the time
(since June 2016) with a COS check failure. Finding a fix for this
is taking longer than expected. We would like to except this subtest
until a better COS check test can be developed. This change only
affects the replay-dual.sh test set script.
Test-Parameters: trivial testlist=replay-dual
Change-Id: I575342ae5b6633efc198383b3351b4d9cfc1fbf0
Signed-off-by: Jim Casper <jamesx.casper@intel.com>
Reviewed-on: https://review.whamcloud.com/26631
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: James Nunez <james.a.nunez@intel.com>
Reviewed-by: Saurabh Tandan <saurabh.tandan@intel.com>
Reviewed-by: Wei Liu <wei3.liu@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Saurabh Tandan [Fri, 14 Apr 2017 00:55:35 +0000 (17:55 -0700)]
LU-9320 test: Skip sanity test_80[12]*
Skip sanity test_80[1-2]* if MDS or OSS version is before
2.9.55
Skipping following tests: test_801a, test_801b, test_801c,
test_802
Test-Parameters: trivial testlist=sanity,sanity
Signed-off-by: Saurabh Tandan <saurabh.tandan@intel.com>
Change-Id: I98ace61c312466381ffcc406644276d86faac782
Reviewed-on: https://review.whamcloud.com/26616
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: James Nunez <james.a.nunez@intel.com>
John L. Hammond [Fri, 21 Apr 2017 14:16:25 +0000 (09:16 -0500)]
LU-9383 test: use a subdir in sanityn test_71b()
In sanityn test_71b() use a test specific subdir to avoid changing the
default file striping on the filesystem root.
Signed-off-by: John L. Hammond <john.hammond@intel.com>
Change-Id: Ic3bfbf24a9f6d00f07392014b48af0c5062756f6
Reviewed-on: https://review.whamcloud.com/26771
Tested-by: Jenkins
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: James Nunez <james.a.nunez@intel.com>
Saurabh Tandan [Thu, 13 Apr 2017 18:08:44 +0000 (11:08 -0700)]
LU-9318 test: sanity-lsnapshot skipped for all non-zfs configs
The problem with the script was that it tried to check
for ZFS file system after the script tried to add snapshot
to the debug list.
As part of solution the script just needs to check for ZFS
file system or the version is too old at the start for of
the script.
Test-Parameters: trivial testlist=sanity-lsnapshot,sanity-lsnapshot,sanity-lsnapshot
Signed-off-by: Saurabh Tandan <saurabh.tandan@intel.com>
Change-Id: Ie59dac7543d1e35b924c84d19c34a5ec5b903a14
Reviewed-on: https://review.whamcloud.com/26600
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: James Nunez <james.a.nunez@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Fan Yong [Tue, 22 Nov 2016 17:39:05 +0000 (01:39 +0800)]
LU-9271 tests: remove unnecessary sanity-scrub test delay
In sanity-scrub test_5, we want to check the OI scrub state machine.
For that, we inject some failure stub to delay the test for checking
status. But sometimes, we want to check the status within some short
time that may be not enough because of the injected test delay. This
patch removes one unnecessary status checking since it has been done
in former check. On the other hand, the subsequent checking also can
indicate the correctness of the skipped status.
Test-parameters: trivial
Signed-off-by: Fan Yong <fan.yong@intel.com>
Change-Id: I14b5096db3d7ca255f3de1a8853dde85ead54215
Reviewed-on: https://review.whamcloud.com/26457
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Jian Yu <jian.yu@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Alex Zhuravlev [Tue, 11 Apr 2017 07:01:23 +0000 (10:01 +0300)]
LU-9314 osd: cleaup qsd once commit callbacks are done
otherwise osd_trans_commit_cb() calling qsd_op_end() may
find qsd structure released.
Change-Id: Ifd64b17d636f5e728ed6a4ad1a2d4d43998c90ec
Signed-off-by: Alex Zhuravlev <alexey.zhuravlev@intel.com>
Reviewed-on: https://review.whamcloud.com/26496
Tested-by: Jenkins
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Mike Pershin <mike.pershin@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Bob Glossman [Mon, 3 Apr 2017 23:03:36 +0000 (16:03 -0700)]
LU-9287 kernel: kernel update [SLES11 SP4 3.0.101-97]
Update SLES11 SP4 kernel to 3.0.101-97
Test-Parameters: mdsdistro=sles11sp4 ossdistro=sles11sp4 \
clientdistro=sles11sp4 mdtfilesystemtype=ldiskfs \
ostfilesystemtype=ldiskfs testgroup=review-ldiskfs
Signed-off-by: Bob Glossman <bob.glossman@intel.com>
Change-Id: I5214114b5c0bb0bfe5a68dbe8cf2a1f26705d352
Reviewed-on: https://review.whamcloud.com/26336
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Yang Sheng <yang.sheng@intel.com>
Reviewed-by: Minh Diep <minh.diep@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Bob Glossman [Thu, 30 Mar 2017 15:56:29 +0000 (08:56 -0700)]
LU-9275 kernel: kernel update [SLES12 SP2 4.4.49-92.14]
Update target and kernel_config files for new version
Test-Parameters: clientdistro=sles12sp2 testgroup=review-ldiskfs \
mdsdistro=sles12sp2 ossdistro=sles12sp2 \
mdtfilesystemtype=ldiskfs ostfilesystemtype=ldiskfs
Signed-off-by: Bob Glossman <bob.glossman@intel.com>
Change-Id: I1d9c109904a61989b8eeb82c21ab99154db464cf
Reviewed-on: https://review.whamcloud.com/26263
Tested-by: Jenkins
Reviewed-by: James Nunez <james.a.nunez@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Minh Diep <minh.diep@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Andreas Dilger [Wed, 29 Mar 2017 21:53:49 +0000 (15:53 -0600)]
LU-930 doc: update LFSCK documentation
Update the LFSCK usage document to include information about logging
LFSCK changes to the console or separate log file.
Improve the description of some options, putting more commonly-used
options first.
Update URLs to reference lustre.org instead of other websites.
Test-Parameters: trivial
Signed-off-by: Andreas Dilger <andreas.dilger@intel.com>
Change-Id: I980a43143297400c0a544281b716c32881d65172
Reviewed-on: https://review.whamcloud.com/26253
Tested-by: Jenkins
Reviewed-by: Fan Yong <fan.yong@intel.com>
Reviewed-by: Joseph Gmitter <joseph.gmitter@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Fan Yong [Sun, 13 Nov 2016 11:14:48 +0000 (19:14 +0800)]
LU-9259 tests: set fail_loc on the right MDT
In sanity test_17o, we need to set the fail_loc on the MDT on which
the target file resides. So "getstripe -M" should be done after the
file created. In further client side ldlm locks should be cancelled
before failover the MDT.
Signed-off-by: Fan Yong <fan.yong@intel.com>
Change-Id: I35acefb6b7a361f25593843e1f2a097e786e5f11
Reviewed-on: https://review.whamcloud.com/26225
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Jian Yu <jian.yu@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Steve Guminski [Tue, 21 Mar 2017 19:04:19 +0000 (15:04 -0400)]
LU-9166 mdt: Add trailing newlines to proc files
Adds trailing newlines to two proc files. A new sanity test ensures
that all non-empty readable proc files end with a newline.
Test-Parameters: trivial
Signed-off-by: Steve Guminski <stephenx.guminski@intel.com>
Change-Id: Id65dd5a43edc41be9f572d476b38975f0144fbaf
Reviewed-on: https://review.whamcloud.com/25977
Tested-by: Jenkins
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Reviewed-by: James Simmons <uja.ornl@yahoo.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Steve Guminski [Wed, 1 Mar 2017 18:46:19 +0000 (13:46 -0500)]
LU-9076 ptlrpc: Prevent possible dereference of NULL pointers
Check pointers for NULL before passing to other functions.
Signed-off-by: Steve Guminski <stephenx.guminski@intel.com>
Change-Id: I5ad1e6675e9c5e41aee2dc59045ed67e9da5ea0d
Reviewed-on: https://review.whamcloud.com/25909
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Reviewed-by: Dmitry Eremin <dmitry.eremin@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Andreas Dilger [Tue, 7 Mar 2017 03:05:42 +0000 (20:05 -0700)]
LU-9186 tests: fix sanityn test_76 and related code
Fix sanityn.sh test_76 to skip already open files when checking
whether mdt.*.exports.*.open_files accurately lists open files.
Otherwise, if some previous test leaves a file descriptor open
it will cause this test to fail spuriously.
Limit the number of open files to 256*MDSCOUNT rather than 2048,
so the test runs faster (80s vs. 215s in my VM with 2 MDTs).
Properly close the open shell FDs using "FD<&-" rather than
redirecting them to /dev/null, otherwise they stay open forever.
Print a '.' only once per 32 files opened/closed.
List open files in *.open_files in the order they were opened,
rather than reverse order, since this makes it easier to see which
files have been open longer.
Print out the pathname of the open files being skipped.
Don't explicitly mount $MOUNT2 in sanityn.sh. This is already
handled inside check_and_setup_lustre() at the start of sanityn.sh.
Improve free_fd() to accept a "last_fd" argument to avoid searching
all open FDs on each call if it is known the fds are not closing.
Test-Parameters: trivial
Signed-off-by: Andreas Dilger <andreas.dilger@intel.com>
Change-Id: I274b6db96f49a52c086f836c005558ef4b3ebbe5
Reviewed-on: https://review.whamcloud.com/25847
Tested-by: Jenkins
Reviewed-by: Jian Yu <jian.yu@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Ashish Purkar <ashish.purkar@seagate.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Mikhal Pershin [Tue, 6 Dec 2016 11:06:48 +0000 (14:06 +0300)]
LU-9184 grants: move grant code to the target
Make grants code common for MDT and OST.
- move grants field from OFD to the lu_target
- same for statfs
- make ofd_grants.c independent from OFD
- move ofd_grants.c to the tgt_grants.c
This work was done for Data-on-MDT feature but it
is needed also for better space control on MDT as
well, e.g. for lfsck, HSM, etc.
Signed-off-by: Mikhal Pershin <mike.pershin@intel.com>
Change-Id: I43930b32b271c9e850aa2783795e7cce12830cd5
Reviewed-on: https://review.whamcloud.com/25820
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Niu Yawei <yawei.niu@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Dmitry Eremin [Thu, 2 Mar 2017 19:23:00 +0000 (22:23 +0300)]
LU-9183 libcfs: handle dump_trace() and related callbacks removal
In commit
c8fe4609827aedc9c4b45de80e7cdc8ccfa8541b all previous users
of dump_trace() have been converted to use the new unwind interfaces,
so the dump_trace() and the related print_context_stack() and
print_context_stack_bp() callback functions were removed.
Test-Parameters: trivial
Change-Id: Ifa7a112d622b23f733d6daab05f9838afdf31a86
Signed-off-by: Dmitry Eremin <dmitry.eremin@intel.com>
Reviewed-on: https://review.whamcloud.com/25816
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Reviewed-by: James Simmons <uja.ornl@yahoo.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Andreas Dilger [Sat, 4 Mar 2017 05:27:32 +0000 (22:27 -0700)]
LU-3434 scripts: check for mdsfilesystemtype= option
Check if the Test-Parameters line is using the old mdsfilesystemtype=
option and ask for it to be removed. The mdtfilesystemtype= option
should be used instead.
Test-Parameters: trivial
Signed-off-by: Andreas Dilger <andreas.dilger@intel.com>
Change-Id: I52765fd94f4dec896593c86763737fa64f3ebbe5
Reviewed-on: https://review.whamcloud.com/25804
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Charlie Olmstead <charlie.olmstead@intel.com>
Reviewed-by: James Nunez <james.a.nunez@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Andriy Skulysh [Wed, 1 Mar 2017 18:34:11 +0000 (20:34 +0200)]
LU-9171 osc: GPF while doing ELC with no_wait_policy
osc_ldlm_weigh_ast() uses osc_object from ldlm_lock
without taking a reference.
It should take a reference like osc_ldlm_glimpse_ast()/
osc_dlm_blocking_ast0() does.
Change-Id: I33a817e4c08c2955f5395b13ffe8cd11f3bdf691
Seagate-bug-id: MRP-4179
Signed-off-by: Andriy Skulysh <andriy.skulysh@seagate.com>
Reviewed-on: https://review.whamcloud.com/25700
Reviewed-by: Jinshan Xiong <jinshan.xiong@intel.com>
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Bobi Jam <bobijam@hotmail.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Yang Sheng [Mon, 10 Apr 2017 07:31:26 +0000 (15:31 +0800)]
LU-7502 utils: add --mdt-count & --mdt-hash options
lfs find --mdt-count and --mdt-hash options
lfs getdirstripe --mdt-count and --mdt-hash options
Signed-off-by: Yang Sheng <yang.sheng@intel.com>
Change-Id: I55f2d7660c4e4fea5a197a57407b614c4004974c
Reviewed-on: https://review.whamcloud.com/25515
Tested-by: Jenkins
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Reviewed-by: Jian Yu <jian.yu@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Arnd Bergmann [Thu, 30 Mar 2017 13:32:59 +0000 (09:32 -0400)]
LU-4423 osc: use 64-bit timestamps in debugfs output
Some lustre debugfs files contain the current time. Make sure
we use a format here that does not overflow in 2038.
Linux-commit:
86655400c6e74751a73be82ea40d5ae6979c3225
Change-Id: I8a79e60bd15ce7c162b30025ca3b572fec2ee58f
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Oleg Drokin <green@linuxhacker.ru>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: James Simmons <uja.ornl@yahoo.com>
Reviewed-on: https://review.whamcloud.com/25407
Reviewed-by: Jinshan Xiong <jinshan.xiong@intel.com>
Reviewed-by: Dmitry Eremin <dmitry.eremin@intel.com>
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
James Nunez [Tue, 20 Dec 2016 21:12:28 +0000 (14:12 -0700)]
LU-8603 test: Separate tests in metadata_updates
The metadata-updates script is one continuous piece of code
that runs three different tests. These tests need to be broken
out into sub-tests and some of the comments echoed to the
console should be logged.
Test-Parameters: trivial testlist=metadata-updates
Signed-off-by: James Nunez <james.a.nunez@intel.com>
Change-Id: I7e420ee8e0e02d2c2a66ac644301bb457a8b2256
Reviewed-on: https://review.whamcloud.com/24450
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Jian Yu <jian.yu@intel.com>
Reviewed-by: Niu Yawei <yawei.niu@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Dmitry Eremin [Fri, 21 Oct 2016 12:10:39 +0000 (15:10 +0300)]
LU-8703 libcfs: use int type for CPT identification.
This patch bring the lustre CPT code into alignment with the Linux kernel.
* use int type for CPT identification like CPU identification.
* fix coding style.
Change-Id: I5df33f37757eb6d47ab4e824ec1b800a40730f61
Signed-off-by: Dmitry Eremin <dmitry.eremin@intel.com>
Reviewed-on: https://review.whamcloud.com/23304
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: James Simmons <uja.ornl@yahoo.com>
Reviewed-by: Doug Oucharek <doug.s.oucharek@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Ashish Purkar [Fri, 26 Feb 2016 04:55:47 +0000 (20:55 -0800)]
LU-7819 tests: fix checkfiemap for stripecount > 1
If sanityn test_71 is ran with stripecount > 1, it fails
with "error while ioctl 95" i.e. EOPNOTSUPP.
FIEMAP_FLAG_DEVICE_ORDER flag must be set while issuing
FS_IOC_FIEMAP ioctl.
New test case 71b to check fiemap for stripe count > 1
Renamed original test_71 to 71a
Test-Parameters: trivial testlist=sanityn envdefinitions=ONLY=71
Seagate-bug-id: MRP-2858
Signed-off-by: Ashish Purkar <ashish.purkar@seagate.com>
Change-Id: If62370fbcfd586607e0f0da2b348070641bed0d1
Reviewed-by: Artem Blagodarenko <artem.blagodarenko@seagate.com>
Reviewed-by: Ujjwal Lanjewar <ujjwal.lanjewar@seagate.com>
Tested-by: Parinay Vijayprakash Kondekar <parinay.kondekar@seagate.com>
Reviewed-on: https://review.whamcloud.com/18679
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Bobi Jam <bobijam@hotmail.com>
teddy [Sun, 9 Apr 2017 03:14:33 +0000 (11:14 +0800)]
LU-7529 test: fix tiny problems of tests in conf-sanity.sh
1)test_68:
Mount MGS in the end of test if MGS and MDS are seperated.
2)test_75:
Reformat all devices otherwise 76a might fail.
Change-Id: Iee8a5fa1c6acccbae86cdcbe14c8715a71b0ae2b
Signed-off-by: Li Xi <lixi@ddn.com>
Signed-off-by: teddy <jjkky@yahoo.com>
Reviewed-on: https://review.whamcloud.com/17507
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Emoly Liu <emoly.liu@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Bobi Jam [Sun, 9 Apr 2017 09:12:19 +0000 (17:12 +0800)]
LU-9307 lov: NULL pointer deref in lov_delete_composite
When lov_init_composite() cannot allocate memory and bails out,
lov_delete_composite() is then called, and it should check the
composite entry's existence before trying to delete each component.
Signed-off-by: Bobi Jam <bobijam.xu@intel.com>
Change-Id: I013c596a19a114b763f3eebbaafa1559fa09d8d9
Reviewed-on: https://review.whamcloud.com/26456
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Niu Yawei <yawei.niu@intel.com>
Reviewed-by: Jinshan Xiong <jinshan.xiong@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
James Simmons [Sat, 8 Apr 2017 16:52:15 +0000 (12:52 -0400)]
LU-9019 quota : migrate to 64 bit time
Replace cfs_time_current_sec() to avoid the overflow
issues in 2038 with ktime_get_real_seconds(). The
quota uses the libcfs 64 bit time abstraction it
developed before the linux kernel has created its
own. Replace all the libcfs 64 bit time handle with
generic time64_t handling.
Change-Id: Ibf9475c7ba8312561a57d48516ffa9b0b3ae4112
Signed-off-by: James Simmons <uja.ornl@yahoo.com>
Reviewed-on: https://review.whamcloud.com/26452
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Niu Yawei <yawei.niu@intel.com>
Reviewed-by: Fan Yong <fan.yong@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Alexander Boyko [Mon, 3 Apr 2017 11:35:53 +0000 (14:35 +0300)]
LU-9286 ptlrpc: fix wrong error handlers
If ptlrpc_request_pack finish with 0, we need to call
ptlrpc_req_finished for later errors.
Signed-off-by: Alexander Boyko <alexander.boyko@seagate.com>
Seagate-bug-id: MRP-4285
Change-Id: I97dec565c1aa95c4457b32cb5d72b4f8abdd3d1d
Reviewed-on: https://review.whamcloud.com/26319
Tested-by: Jenkins
Reviewed-by: Chris Horn <hornc@cray.com>
Reviewed-by: Mike Pershin <mike.pershin@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Niu Yawei <yawei.niu@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
git://git.whamcloud.com / fs / lustre-release.git / log