Whamcloud - gitweb
yury [Tue, 13 Sep 2005 18:42:37 +0000 (18:42 +0000)]
- possible fix to #49 with good comment
tappro [Tue, 13 Sep 2005 17:58:15 +0000 (17:58 +0000)]
b=9312
I need Big Book of Logic it seems.
niu [Tue, 13 Sep 2005 15:06:33 +0000 (15:06 +0000)]
- correct some return value.
- add more assertions.
tappro [Tue, 13 Sep 2005 11:12:39 +0000 (11:12 +0000)]
b=8654
- cut&paste bug in previous commit
tappro [Tue, 13 Sep 2005 11:07:14 +0000 (11:07 +0000)]
b=8654
- fix for client STAT logging
alex [Tue, 13 Sep 2005 08:53:04 +0000 (08:53 +0000)]
b=9294
- debug message to make sure we interpret LASSERT() right
yury [Tue, 13 Sep 2005 07:56:47 +0000 (07:56 +0000)]
- fixed objects creating on OST with correct uid and gid on setattr path.
- removed not needed prototype from filter_internal.h
- removed uid/gid stuff from filter_preprw_write() as it is not used there.
tappro [Tue, 13 Sep 2005 07:34:04 +0000 (07:34 +0000)]
b=8654
store setattr mask in nid for debug purposes
tappro [Tue, 13 Sep 2005 03:26:55 +0000 (03:26 +0000)]
b=9300
correct incrementing.
niu [Tue, 13 Sep 2005 02:57:12 +0000 (02:57 +0000)]
oops, fix typo.
ericm [Tue, 13 Sep 2005 02:51:25 +0000 (02:51 +0000)]
quiet some crypto messages.
niu [Tue, 13 Sep 2005 02:49:34 +0000 (02:49 +0000)]
enlarge namelen to make caller can free mem with correct size
tappro [Tue, 13 Sep 2005 00:15:32 +0000 (00:15 +0000)]
b=9300
wrong LASSERT in previous commit.
tappro [Mon, 12 Sep 2005 23:21:21 +0000 (23:21 +0000)]
b=9300
no space was allocated for final zero at the end of string.
adilger [Mon, 12 Sep 2005 22:30:05 +0000 (22:30 +0000)]
Branch HEAD
Ensure that libsysio is also checked out with the current version tag if
specified for releases 1.4.3 and beyond.
adilger [Mon, 12 Sep 2005 22:23:40 +0000 (22:23 +0000)]
Branch HEAD
Verify that the to-be-branched directory exists.
Suppress bogus error about CVS/Tag not missing if we are branching from HEAD.
alex [Mon, 12 Sep 2005 20:22:59 +0000 (20:22 +0000)]
- don't need that debug anymore, fsync() is safe enough WRT 11/17
eeb [Mon, 12 Sep 2005 19:41:46 +0000 (19:41 +0000)]
* Added temporary LNET branch b_newconfig_rdmarouting
tappro [Mon, 12 Sep 2005 19:34:49 +0000 (19:34 +0000)]
b=9300
re-enable path restoring
wangdi [Mon, 12 Sep 2005 17:31:50 +0000 (17:31 +0000)]
Branch: HEAD
some fix of mds add
1)not reset uuid when lmv_add_mdc in mds_dt_update_config
2)reset mds_max_mdsize after config update of mds
tappro [Mon, 12 Sep 2005 14:48:33 +0000 (14:48 +0000)]
b=9300
turn off pathname restoring for debug purpose
niu [Mon, 12 Sep 2005 11:42:57 +0000 (11:42 +0000)]
- correct type checking logic in filldir().
- some debug message.
niu [Mon, 12 Sep 2005 11:35:22 +0000 (11:35 +0000)]
return 128 as filetype for cross-ref entry
r: alex
alex [Mon, 12 Sep 2005 10:38:43 +0000 (10:38 +0000)]
- test multinode creation w/o separate portal for MDS_CLOSE
alex [Sun, 11 Sep 2005 17:52:33 +0000 (17:52 +0000)]
- one more CWARN() to simplify debugging
alex [Sun, 11 Sep 2005 12:08:34 +0000 (12:08 +0000)]
b=7237
- always close transaction before calling lov/osc
tappro [Sun, 11 Sep 2005 05:40:09 +0000 (05:40 +0000)]
change message level from warn to info
tappro [Sun, 11 Sep 2005 03:48:06 +0000 (03:48 +0000)]
b=8654
- fix cross-ref issue while scanning dir by inode number
tappro [Sun, 11 Sep 2005 03:21:41 +0000 (03:21 +0000)]
b=8654
- fix issue when more than needed chars in name are shown
tappro [Sun, 11 Sep 2005 01:07:48 +0000 (01:07 +0000)]
b=8654
temporary fix for pathname restoring procedure
tappro [Sun, 11 Sep 2005 00:37:46 +0000 (00:37 +0000)]
b=8654
r=alex
- get mds num from obd for just created file/directory to store
correct value in audit log
tappro [Sun, 11 Sep 2005 00:34:26 +0000 (00:34 +0000)]
b=8654
r=alex
- add audit flag on UNNAMED dir to enable audit inheritance on newly created objects
alex [Sat, 10 Sep 2005 23:44:32 +0000 (23:44 +0000)]
- better dump a backtrace, but not too often
alex [Sat, 10 Sep 2005 16:06:46 +0000 (16:06 +0000)]
b=7273
- additional debug check
alex [Sat, 10 Sep 2005 07:20:52 +0000 (07:20 +0000)]
- make obdfilter able to create object upon OST_CREATE as before,
need this for lustre-iokit.
alex [Fri, 9 Sep 2005 19:56:41 +0000 (19:56 +0000)]
b=7268
- additional debug message to simplify data gathering from script
alex [Fri, 9 Sep 2005 19:55:34 +0000 (19:55 +0000)]
- DEBUG_REQ() on server shows where request came from
lsy [Fri, 9 Sep 2005 13:44:30 +0000 (13:44 +0000)]
fix:
* mds_id might be 0, so +1 when calculating hash.
* link new capa into its hash list head, not the capa_hash.
yury [Fri, 9 Sep 2005 11:53:06 +0000 (11:53 +0000)]
- fixes in GNS stuff - bug #9283
tappro [Fri, 9 Sep 2005 09:02:12 +0000 (09:02 +0000)]
b=8654
finish previous commit
tappro [Fri, 9 Sep 2005 08:53:06 +0000 (08:53 +0000)]
b=8654
fixed bug with setting fs_audit. lustre root inode id was used to set audit. For fs audit inode = NULL should be used.
tappro [Fri, 9 Sep 2005 04:41:20 +0000 (04:41 +0000)]
b=8654
debug for audit setting
tappro [Fri, 9 Sep 2005 00:59:53 +0000 (00:59 +0000)]
b=8654
debug for audit issues on MDS
tappro [Thu, 8 Sep 2005 19:46:08 +0000 (19:46 +0000)]
b=9262
update local audit setting on client while set_audit()
adilger [Thu, 8 Sep 2005 18:16:22 +0000 (18:16 +0000)]
Branch b1_4
Description: liblustre could not open files whose last component is a symlink
Details : sysio_path_walk() would incorrectly pass the open intent to
intermediate path components.
b=6363
r=oleg, lee, devesh
yury [Thu, 8 Sep 2005 08:34:57 +0000 (08:34 +0000)]
- landed share_write_lock.patch from #7410
ericm [Wed, 7 Sep 2005 17:07:16 +0000 (17:07 +0000)]
adjust gss upcall timeout value just be secinit timeout, service gss upcall
might be very slow.
yury [Wed, 7 Sep 2005 12:24:23 +0000 (12:24 +0000)]
- fixed oops in mds_reint_unlink_remote()
wangdi [Wed, 7 Sep 2005 06:28:51 +0000 (06:28 +0000)]
Branch: HEAD
merge HP fixes about ost add/del
ericm [Wed, 7 Sep 2005 00:46:19 +0000 (00:46 +0000)]
minor fix to make gss error report more accurate.
lsy [Tue, 6 Sep 2005 14:00:26 +0000 (14:00 +0000)]
fix: for each open_client_handle, there will be only one mds_file_data,
so capability should obtain mfd handle from och->och_fh, instead of
the one from reply message body, for the latter one has been cleared.
tappro [Mon, 5 Sep 2005 15:44:26 +0000 (15:44 +0000)]
r=alex
dparent can be valid not only if resent_req == 0
alex [Mon, 5 Sep 2005 14:47:38 +0000 (14:47 +0000)]
- complain if kernel doesn't support 1MB request
alex [Mon, 5 Sep 2005 14:09:29 +0000 (14:09 +0000)]
- set SCSI_MAX_PHYS_SEGMENTS to 256 for 1MB requests
alex [Mon, 5 Sep 2005 13:39:19 +0000 (13:39 +0000)]
- disable debug
alex [Mon, 5 Sep 2005 12:56:32 +0000 (12:56 +0000)]
- debug info to investigate a source of small writes
niu [Mon, 5 Sep 2005 09:46:37 +0000 (09:46 +0000)]
rename confusing mdc_uuid to tgt_uuid.
wangdi [Mon, 5 Sep 2005 08:58:54 +0000 (08:58 +0000)]
Branch: HEAD
remove obsolete lgs.sh
lsy [Mon, 5 Sep 2005 06:05:07 +0000 (06:05 +0000)]
add some debug message for capbility key update.
a small fix for key expiry.
alex [Sun, 4 Sep 2005 23:19:20 +0000 (23:19 +0000)]
- typo fixed
alex [Sun, 4 Sep 2005 22:23:38 +0000 (22:23 +0000)]
- mds_postrecov_common() can return number of cleaned orphans as a success
alex [Sun, 4 Sep 2005 22:21:50 +0000 (22:21 +0000)]
b=7273
- additional debug info
alex [Sun, 4 Sep 2005 14:38:34 +0000 (14:38 +0000)]
- make gcc happy
wangdi [Sun, 4 Sep 2005 03:35:45 +0000 (03:35 +0000)]
Branch: HEAD
add ost add/del test script
wangdi [Sun, 4 Sep 2005 03:21:20 +0000 (03:21 +0000)]
Branch: HEAD
remove lov_tgt_pending, which is useless for ost add/del
alex [Sat, 3 Sep 2005 16:16:31 +0000 (16:16 +0000)]
- can hit during forced umount
alex [Sat, 3 Sep 2005 16:12:11 +0000 (16:12 +0000)]
- i think we don't need that now
alex [Sat, 3 Sep 2005 12:13:46 +0000 (12:13 +0000)]
- disable extents/mballoc for benchmarking purposes
alex [Sat, 3 Sep 2005 09:37:02 +0000 (09:37 +0000)]
- serialize allocations for 2.1, 2.2 and 2.3
alex [Fri, 2 Sep 2005 21:15:46 +0000 (21:15 +0000)]
- revert hacks back
alex [Fri, 2 Sep 2005 19:31:56 +0000 (19:31 +0000)]
b=7272
- more debug to catch the case
alex [Fri, 2 Sep 2005 19:29:21 +0000 (19:29 +0000)]
- one more field in struct mds_rec_create to replay i/o epoch file was open in
lsy [Fri, 2 Sep 2005 13:31:46 +0000 (13:31 +0000)]
fixes:
* also round for short expiry, in case timer might miss.
* add mds_mfd_put() because mds_handle2mfd() will add ref.
* filter_verify_capa() logic error.
yury [Fri, 2 Sep 2005 12:46:00 +0000 (12:46 +0000)]
- few fixes in comments (typos)
- removed wrong stuff in class_add_conn(). Apparently it is result of wrong merge when newcfg stuff was landed. Issues related to it were observed in #48.
alex [Thu, 1 Sep 2005 23:15:39 +0000 (23:15 +0000)]
- eeb's sd iostat bits have been added
alex [Thu, 1 Sep 2005 22:53:41 +0000 (22:53 +0000)]
- disable fo_alloc_sem
- don't wait for transaction commit
NOTE: to evaluate impact
alex [Thu, 1 Sep 2005 21:53:36 +0000 (21:53 +0000)]
- add fo_alloc_lock serialization support for extents/mballoc case
alex [Thu, 1 Sep 2005 19:59:45 +0000 (19:59 +0000)]
- bump OST_MAX_THREADS to 48 for benchmarking
alex [Thu, 1 Sep 2005 18:14:13 +0000 (18:14 +0000)]
- serialize all allocations on OSS by fo_alloc_lock. for benchmarking
wangdi [Thu, 1 Sep 2005 07:58:17 +0000 (07:58 +0000)]
Branch: HEAD
lproc for update config should return count, not 0
wangdi [Thu, 1 Sep 2005 06:01:38 +0000 (06:01 +0000)]
Branch: HEAD
lli_posix_acl and lli_key_info is not reliable on checking whether the acl and key_info
is valid on the client
pjkirner [Thu, 1 Sep 2005 04:02:09 +0000 (04:02 +0000)]
Add support for b_lnet branch of the ./build tree to cleanly
seperate the changes related to LNET reorg that is going on on the
b1_4_newconfig branch(s).
(Consulted with Jacob and Andreas on this change)
adilger [Wed, 31 Aug 2005 20:00:53 +0000 (20:00 +0000)]
Branch HEAD
Check out the tagged version of the lustre-build scripts, so we are sure
to use the same ones as when the packages were originally built.
lsy [Wed, 31 Aug 2005 16:05:47 +0000 (16:05 +0000)]
change error msg to warning.
some cleanup.
alex [Wed, 31 Aug 2005 15:35:05 +0000 (15:35 +0000)]
- as we don't set PG_writeback on pages being flushed, we need another
way to support fsync(2). the simplest one is to wait for empty list
lli_pending_write_llaps
pjkirner [Wed, 31 Aug 2005 02:20:33 +0000 (02:20 +0000)]
Fix build breakage introduced by lnet reorg.
Completely removed changes until something more compatable can be designed.
Backed out changes to build tree:
autogen.sh reverted back to rev 1.4
autoMakefile.am.toplevel reverted back to rev 1.8
buildcvs reverted back to rev 1.44
Makefile.in.toplevel reverted back to rev 1.1
autoconf/lustre-build-darwin.m4 reverted back to rev 1.8
autoconf/lustre-build-linux.m4 reverted back to rev 1.15
autoconf/lustre-build.m4 reverted back to rev 1.28
tappro [Wed, 31 Aug 2005 02:11:23 +0000 (02:11 +0000)]
b=7390, 7409
changes in auditing of failed operations when audit is set on directory
pjkirner [Tue, 30 Aug 2005 20:09:53 +0000 (20:09 +0000)]
Phase 1 of LNET Reorg
r=eeb
r=pjk
wangdi [Tue, 30 Aug 2005 14:14:40 +0000 (14:14 +0000)]
Branch: HEAD
forget lprof->lp_gkc in profile delete
alex [Tue, 30 Aug 2005 11:51:19 +0000 (11:51 +0000)]
- give concurrent allocation one more chance
ericm [Mon, 29 Aug 2005 20:59:19 +0000 (20:59 +0000)]
hack to always use root credential between client and oss.
alex [Mon, 29 Aug 2005 19:42:30 +0000 (19:42 +0000)]
b=7268
- dump time ost_brw_write() spent waiting for bulk completion
alex [Mon, 29 Aug 2005 12:28:58 +0000 (12:28 +0000)]
- replay i/o epochs along with other attributes to improve isize consistency
without involving OST
lsy [Mon, 29 Aug 2005 10:33:29 +0000 (10:33 +0000)]
minor fix:
* for partial write only search capability once.
alex [Mon, 29 Aug 2005 10:30:37 +0000 (10:30 +0000)]
- ll_writepages() waits for I/O completion if filesystem is being umounted
alex [Mon, 29 Aug 2005 10:28:16 +0000 (10:28 +0000)]
- dump log if OSC recovery takes too long
alex [Mon, 29 Aug 2005 09:49:28 +0000 (09:49 +0000)]
r=tappro
- couple unused vars have been removed
wangdi [Mon, 29 Aug 2005 07:22:19 +0000 (07:22 +0000)]
Branch: HEAD
revert back the fix for b7379, it is not right.
alex [Sun, 28 Aug 2005 17:57:15 +0000 (17:57 +0000)]
- fill io epoch in MDS_CLOSE only if file is writtable
alex [Sun, 28 Aug 2005 13:16:48 +0000 (13:16 +0000)]
- set and check io epoch only for writtable files