]>
git.apps.os.sepia.ceph.com Git - ceph.git/log
Sage Weil [Wed, 19 Aug 2015 21:03:28 +0000 (17:03 -0400)]
os/RocksDBStore: log any options we pass to rocksdb
Signed-off-by: Sage Weil <sage@redhat.com>
Sage Weil [Wed, 19 Aug 2015 11:04:07 +0000 (07:04 -0400)]
os/FileStore: ensure sync() doesn't wait forever
Avoid
2015-08-19 01:46:37.586450
7f89a53ca980 10 filestore(b.10) start_sync
2015-08-19 01:46:37.586451
7f89a53ca980 10 filestore(b.10) sync waiting
2015-08-19 01:46:37.586528
7f899e2b1700 20 filestore(b.10) sync_entry woke after 5.398564
2015-08-19 01:46:37.586539
7f899e2b1700 20 filestore(b.10) sync_entry waiting for another 35994.601435 to reach min interval 36000.000000
by setting force_sync flag.
Signed-off-by: Sage Weil <sage@redhat.com>
Sage Weil [Tue, 18 Aug 2015 22:40:59 +0000 (18:40 -0400)]
common/hobject: rename get_bitreverse_key* -> get_bitwise_key*
Signed-off-by: Sage Weil <sage@redhat.com>
Sage Weil [Tue, 18 Aug 2015 19:16:45 +0000 (15:16 -0400)]
os/GenericObjectMap: use bitwise hash, not nibblewise
Signed-off-by: Sage Weil <sage@redhat.com>
Sage Weil [Tue, 18 Aug 2015 19:13:35 +0000 (15:13 -0400)]
common/hobject: rename get_filestore_key* -> get_nibblewise_key*
Signed-off-by: Sage Weil <sage@redhat.com>
Sage Weil [Tue, 18 Aug 2015 20:05:29 +0000 (16:05 -0400)]
rbd: fix bench-write
Make each IO get a unique offset!
Signed-off-by: Sage Weil <sage@redhat.com>
Sage Weil [Tue, 18 Aug 2015 18:39:08 +0000 (14:39 -0400)]
ceph-kvstore-tool: take a db type too
Signed-off-by: Sage Weil <sage@redhat.com>
Sage Weil [Tue, 18 Aug 2015 18:25:38 +0000 (14:25 -0400)]
common/hobject: make hobject_t::set_key avoid dups
If the key matches the oid name, clear it.
Signed-off-by: Sage Weil <sage@redhat.com>
Sage Weil [Tue, 18 Aug 2015 14:52:46 +0000 (10:52 -0400)]
test_objectstore_memstore.sh on make check
This'll give us some coverage from ceph_test_objectstore, albeit
on the memstore backend. Upside is that this takes < 10s.
Signed-off-by: Sage Weil <sage@redhat.com>
Sage Weil [Tue, 28 Apr 2015 16:30:32 +0000 (09:30 -0700)]
ceph_test_objectstore: test many small writes
This exercises the overlay code better.
Signed-off-by: Sage Weil <sage@redhat.com>
Sage Weil [Tue, 18 Aug 2015 14:11:20 +0000 (10:11 -0400)]
ceph_test_objectstore: strengthen clone test w/ new assert
Signed-off-by: Sage Weil <sage@redhat.com>
Sage Weil [Tue, 18 Aug 2015 14:10:42 +0000 (10:10 -0400)]
ceph_test_objectstore: test omap clone
Signed-off-by: Sage Weil <sage@redhat.com>
Sage Weil [Tue, 18 Aug 2015 14:10:18 +0000 (10:10 -0400)]
ceph_test_objectstore: much cleanup, new tests
These changes are necessary to make newstore (which is more picky
about unique collection names and objects that land *inside* the
correct collection) pass. There are also several new tests.
Signed-off-by: Sage Weil <sage@redhat.com>
Sage Weil [Tue, 28 Apr 2015 23:17:03 +0000 (16:17 -0700)]
os/fs: fix open_handle when name_to_handle not present
Signed-off-by: Sage Weil <sage@redhat.com>
Sage Weil [Fri, 10 Apr 2015 23:55:05 +0000 (16:55 -0700)]
os/fs: add zero / hole punch support
Signed-off-by: Sage Weil <sage@redhat.com>
Sage Weil [Tue, 7 Apr 2015 22:21:50 +0000 (15:21 -0700)]
os/fs: add simple FS abstraction layer
- open by handle
- xfs allocation hint
Signed-off-by: Sage Weil <sage@redhat.com>
Sage Weil [Mon, 17 Aug 2015 19:23:14 +0000 (15:23 -0400)]
os/ObjectStore: drop sync(Context *onsync)
No users.
Signed-off-by: Sage Weil <sage@redhat.com>
Sage Weil [Mon, 17 Aug 2015 19:22:57 +0000 (15:22 -0400)]
os/ObjectStore: drop sync()
No callers.
Signed-off-by: Sage Weil <sage@redhat.com>
Sage Weil [Mon, 17 Aug 2015 19:22:35 +0000 (15:22 -0400)]
os/ObjectStore: drop flush()
No more callers.
Signed-off-by: Sage Weil <sage@redhat.com>
Sage Weil [Mon, 17 Aug 2015 19:22:13 +0000 (15:22 -0400)]
os/ObjectStore: drop sync_and_flush()
The semantics of this call were FileStore-specific; and there are no more
callers.
Signed-off-by: Sage Weil <sage@redhat.com>
Sage Weil [Tue, 28 Jul 2015 17:22:27 +0000 (13:22 -0400)]
osd: improve osd bench
Signed-off-by: Sage Weil <sage@redhat.com>
Sage Weil [Thu, 16 Apr 2015 00:01:59 +0000 (17:01 -0700)]
buffer: add prepare_iov() method
Signed-off-by: Sage Weil <sage@redhat.com>
Sage Weil [Sat, 11 Apr 2015 00:13:09 +0000 (17:13 -0700)]
osd: drop dead suicide() code
Signed-off-by: Sage Weil <sage@redhat.com>
Sage Weil [Sat, 11 Apr 2015 00:39:31 +0000 (17:39 -0700)]
ceph-objectstore-tool: drop explicit sync_and_flush() calls
We implicitly sync on shutdown.
Signed-off-by: Sage Weil <sage@redhat.com>
Sage Weil [Mon, 17 Aug 2015 19:15:24 +0000 (15:15 -0400)]
ceph_test_objectstore: drop useless sync_and_flush
Signed-off-by: Sage Weil <sage@redhat.com>
Sage Weil [Sat, 11 Apr 2015 00:07:58 +0000 (17:07 -0700)]
osd: sync (no flush) on osd suicide
Not that we even need to sync here!
Signed-off-by: Sage Weil <sage@redhat.com>
Sage Weil [Sat, 11 Apr 2015 00:07:34 +0000 (17:07 -0700)]
osd: drop explicit sync/flush calls before umount
Do it implicity on umount.
Signed-off-by: Sage Weil <sage@redhat.com>
Sage Weil [Sat, 11 Apr 2015 00:06:26 +0000 (17:06 -0700)]
osd: fix flush_journal command
We need to sync_and_flush *and* sync() again in order to retire any items
in the FileStore journal.
Signed-off-by: Sage Weil <sage@redhat.com>
Sage Weil [Mon, 6 Apr 2015 20:55:46 +0000 (13:55 -0700)]
osd: explicitly use a 'meta' Sequencer
Use an explicit Sequencer for all transactions.
Signed-off-by: Sage Weil <sage@redhat.com>
Sage Weil [Mon, 6 Apr 2015 20:54:46 +0000 (13:54 -0700)]
osd: use osr for scan_snaps
Updates should be ordered with respect to PG metadata. Not that it matters
that much since we are doing cleanup/repair work here.
Signed-off-by: Sage Weil <sage@redhat.com>
Sage Weil [Mon, 17 Aug 2015 19:09:47 +0000 (15:09 -0400)]
os/ObjectStore: do not let Sequencer::flush_commit delete the context
Let the caller manage the lifecycle; this allows us to potentially put it
on the stack, for instance.
Signed-off-by: Sage Weil <sage@redhat.com>
Sage Weil [Mon, 17 Aug 2015 19:09:29 +0000 (15:09 -0400)]
os/ObjectStore: ref count Sequencer_impl
This allows the implementation to have a lifecycle that extends beyond
the Sequencer interface object we expose to the user.
Signed-off-by: Sage Weil <sage@redhat.com>
Xiaoxi Chen [Wed, 8 Apr 2015 13:43:29 +0000 (21:43 +0800)]
os/KeyValueDB: return -ENOENT if key doesn't exist
Previous code return 0 in all cases, caller doesn't
know if the key is exists in the db.
Signed-off-by: Xiaoxi Chen <xiaoxi.chen@intel.com>
Sage Weil [Wed, 1 Apr 2015 00:32:36 +0000 (17:32 -0700)]
os/ObjectStore: pass bits to create_collection
This informs the backend how many bits of the hash range will be covered
by the collection. Those bits are implicitly matched against the spg_t in
coll_t.
Signed-off-by: Sage Weil <sage@redhat.com>
Sage Weil [Wed, 1 Apr 2015 00:24:35 +0000 (17:24 -0700)]
osd/osd_types: pg_t: add contains() helper to check objects is in a pg
This calls the ghobject_t method that does the same, along with ps().
Signed-off-by: Sage Weil <sage@redhat.com>
Sage Weil [Wed, 1 Apr 2015 00:05:30 +0000 (17:05 -0700)]
os/ObjectStore: return const refs for oid, cid
Should be a bit faster.
Signed-off-by: Sage Weil <sage@redhat.com>
Sage Weil [Wed, 1 Apr 2015 00:04:07 +0000 (17:04 -0700)]
common/WorkQueue: add wait()
Signed-off-by: Sage Weil <sage@redhat.com>
Sage Weil [Tue, 31 Mar 2015 00:54:13 +0000 (17:54 -0700)]
vstart.sh: allow rocksdb
Signed-off-by: Sage Weil <sage@redhat.com>
Sage Weil [Sat, 15 Aug 2015 19:30:27 +0000 (15:30 -0400)]
os/RocksDbStore: init pointers in ctor
Signed-off-by: Sage Weil <sage@redhat.com>
Sage Weil [Sat, 28 Mar 2015 00:06:44 +0000 (17:06 -0700)]
osd/osd_types: make coll_t operator== less picky
For meta coll's do not compare pg.
Signed-off-by: Sage Weil <sage@redhat.com>
Sage Weil [Tue, 17 Feb 2015 19:29:35 +0000 (11:29 -0800)]
os/KeyValueDB: add sipler get() variant
Signed-off-by: Sage Weil <sage@redhat.com>
Sage Weil [Mon, 9 Feb 2015 08:56:50 +0000 (00:56 -0800)]
common/Formatter: add dump_object helper
Signed-off-by: Sage Weil <sage@redhat.com>
Sage Weil [Tue, 31 Mar 2015 22:31:02 +0000 (15:31 -0700)]
os/FileStore: clear cached fd for renamed-over object
Signed-off-by: Sage Weil <sage@redhat.com>
Sage Weil [Wed, 19 Aug 2015 15:07:14 +0000 (11:07 -0400)]
Merge pull request #5470 from ceph/wip-newstore-sort
osd: change sort order for hobject_t
Reviewed-by: Samuel Just <sjust@redhat.com>
Sage Weil [Wed, 19 Aug 2015 13:21:42 +0000 (09:21 -0400)]
Merge pull request #5424 from lebauce/fix_swift_user_command_line
doc: fix command line for swift user creation
Reviewed-by: Sage Weil <sage@redhat.com>
Sage Weil [Wed, 19 Aug 2015 13:21:15 +0000 (09:21 -0400)]
Merge pull request #5048 from SUSE/wip-objecter-comment-cleanup
Objecter.cc: fix name of _take_op_budget() function in comment
Reviewed-by: Sage Weil <sage@redhat.com>
Kefu Chai [Wed, 19 Aug 2015 11:55:53 +0000 (19:55 +0800)]
Merge pull request #5606 from liewegas/wip-scrub-errormsg
osd/ReplicatedPG: fix missing whitespace in scrub error message
Reviewed-by: Kefu Chai <kchai@redhat.com>
Kefu Chai [Wed, 19 Aug 2015 10:48:32 +0000 (18:48 +0800)]
Merge pull request #5604 from oritwas/wip-cmake-fix-rbd
cmake: add DiffIterate.cc to librbd
Reviewed-by: Kefu Chai <kchai@redhat.com>
Rohan Mars [Thu, 13 Aug 2015 04:19:31 +0000 (21:19 -0700)]
moved to use boost uuid implementation, based on commit
4fe89a7b14c97b2ed7f357132901beb2bdcec551
Signed-off-by: Rohan Mars <code@rohanmars.com>
Reviewed-by: Casey Bodley <cbodley@redhat.com>
Sage Weil [Wed, 19 Aug 2015 10:44:20 +0000 (06:44 -0400)]
Merge pull request #5433 from zqkkqz/master
common/Thread: pthread_attr_destroy(thread_attr) when done with it
Reviewed-by: Kefu Chai <kchai@redhat.com>
Sage Weil [Wed, 19 Aug 2015 10:40:17 +0000 (06:40 -0400)]
osd/ReplicatedPG: fix missing whitespace in scrub error message
Signed-off-by: Sage Weil <sage@redhat.com>
Sage Weil [Wed, 19 Aug 2015 10:29:40 +0000 (06:29 -0400)]
Merge pull request #5538 from ceph/wip-selinux-improve-update
selinux: Relabel files if and only if the policy version changed
Reviewed-by: Milan Broz <mbroz@redhat.com>
Boris Ranto [Wed, 19 Aug 2015 09:25:52 +0000 (11:25 +0200)]
ceph.spec.in: Require awk for selinux post script
Signed-off-by: Boris Ranto <branto@redhat.com>
Boris Ranto [Tue, 11 Aug 2015 07:13:01 +0000 (09:13 +0200)]
selinux: Relabel files if and only if the policy version changed
Currently, the ceph files are being relabelled every time the package is
rebuilt. Fix this by checking the policy versions and relabel the files
only if the policy actually changed (different policy version was
detected).
Signed-off-by: Boris Ranto <branto@redhat.com>
branto1 [Wed, 19 Aug 2015 09:00:40 +0000 (11:00 +0200)]
Merge pull request #5596 from mbroz/wip-selinux-systemd
Update ceph-selinux rpm spec
Reviewed-by: Boris Ranto <branto@redhat.com>
Orit Wasserman [Wed, 19 Aug 2015 08:56:27 +0000 (10:56 +0200)]
cmake: add DiffIterate.cc to librbd
Signed-off-by: Orit Wasserman <owasserm@redhat.com>
Loic Dachary [Tue, 18 Aug 2015 19:19:37 +0000 (21:19 +0200)]
Merge pull request #5600 from liewegas/wip-osd-bind
doc: ceph-osd binds to up to 4 ports, not 3
Reviewed-by: Loic Dachary <ldachary@redhat.com>
Sage Weil [Tue, 18 Aug 2015 17:57:37 +0000 (13:57 -0400)]
doc: ceph-osd binds to up to 4 ports, not 3
Signed-off-by: Sage Weil <sage@redhat.com>
Sage Weil [Tue, 18 Aug 2015 15:52:53 +0000 (11:52 -0400)]
Merge remote-tracking branch 'gh/next'
branto1 [Tue, 18 Aug 2015 13:05:02 +0000 (15:05 +0200)]
Merge pull request #5597 from dachary/wip-12718-restorecon
ceph-disk: only call restorecon when available
Reviewed-by: Boris Ranto <branto@redhat.com>
Loic Dachary [Tue, 18 Aug 2015 12:43:15 +0000 (14:43 +0200)]
ceph-disk: only call restorecon when available
9db80da12803d42bb676d67f37442c0c54d83448 added an unconditional call to
restorecon after mounting the filesystem. It fails when restorecon is
not available and must be made conditional.
http://tracker.ceph.com/issues/12718 Fixes: #12718
Signed-off-by: Loic Dachary <ldachary@redhat.com>
Milan Broz [Tue, 18 Aug 2015 12:18:26 +0000 (14:18 +0200)]
Do not use hardcoded paths in spec post section for selinux commands.
Signed-off-by: Milan Broz <mbroz@redhat.com>
Milan Broz [Mon, 17 Aug 2015 12:46:31 +0000 (14:46 +0200)]
Add dependence to selinux-policy-base to fix installation in kickstart.
Signed-off-by: Milan Broz <mbroz@redhat.com>
Milan Broz [Mon, 17 Aug 2015 10:48:17 +0000 (12:48 +0200)]
Update rpm spec to properly restart systemd ceph.target.
If we are using systemd, ceph target should be used.
Also fail to restart service should not stop policy to load.
Signed-off-by: Milan Broz <mbroz@redhat.com>
Gregory Farnum [Tue, 18 Aug 2015 12:07:09 +0000 (13:07 +0100)]
Merge pull request #5595 from ceph/wip-12710
qa/fsstress.sh: fix 'cp not writing through dangling symlink'
Reviewed-by: Greg Farnum <gfarnum@redhat.com>
Gregory Farnum [Tue, 18 Aug 2015 12:01:17 +0000 (13:01 +0100)]
Merge pull request #5594 from ceph/wip-12711
mds: properly set client incarnation
Reviewed-by: Greg Farnum <gfarnum@redhat.com>
Yan, Zheng [Tue, 18 Aug 2015 07:22:55 +0000 (15:22 +0800)]
qa/fsstress.sh: fix 'cp not writing through dangling symlink'
On some test machines, /usr/lib/ltp/testcases/bin/fsstress is
dangling symlink. 'cp -f' is impotent in this case.
Fixes: #12710
Signed-off-by: Yan, Zheng <zyan@redhat.com>
Yan, Zheng [Tue, 18 Aug 2015 06:56:24 +0000 (14:56 +0800)]
mds: properly set client incarnation
Fixes: #12711
Signed-off-by: Yan, Zheng <zyan@redhat.com>
Samuel Just [Mon, 17 Aug 2015 19:59:18 +0000 (12:59 -0700)]
Merge pull request #5593 from ceph/revert-4927-snapset-obc
Revert "osd/ReplicatedPG: snapset is not persisted"
Reviewed-by: Samuel Just <sjust@redhat.com>
Samuel Just [Mon, 17 Aug 2015 19:58:58 +0000 (12:58 -0700)]
Revert "osd/ReplicatedPG: snapset is not persisted"
Yehuda Sadeh [Mon, 17 Aug 2015 18:19:23 +0000 (11:19 -0700)]
Merge pull request #5577 from oritwas/wip-next-12363
rgw: we should not overide Swift sent content type
Reviewed-by: Yehuda Sadeh <yehuda@redhat.com>
Jason Dillaman [Mon, 17 Aug 2015 15:42:51 +0000 (11:42 -0400)]
Merge pull request #5558 from s09816/rbd-fix
rbd:improve the error handle of rbd,check the return value.
Reviewed-by: Jason Dillaman <dillaman@redhat.com>
Reviewed-by: Josh Durgin <jdurgin@redhat.com>
Sage Weil [Mon, 17 Aug 2015 14:29:48 +0000 (10:29 -0400)]
osd/OSDMap: test_flag returns bool
Otherwise comparing the result to a bool will promote the bool to
an int and fail the comparison.
Signed-off-by: Sage Weil <sage@redhat.com>
Sage Weil [Mon, 10 Aug 2015 14:14:03 +0000 (10:14 -0400)]
osdc/Objecter: restart listing this PG if sort order changes
If the cluster sort order changes mid-way through our listing, our
cursor within this pg is meaningless and we need to restart at the
beginning of the PG.
Signed-off-by: Sage Weil <sage@redhat.com>
Sage Weil [Fri, 7 Aug 2015 20:47:43 +0000 (16:47 -0400)]
osd/ReplicatedPG: fix missing set sort order on [N]PGLS
This is the only thing that cares about the missing set sort order,
so only adjust it here as needed.
Signed-off-by: Sage Weil <sage@redhat.com>
Sage Weil [Fri, 7 Aug 2015 20:47:08 +0000 (16:47 -0400)]
osd/osd_types: add pg_missing_t::resort() method
Set/change sort order as needed.
Signed-off-by: Sage Weil <sage@redhat.com>
Sage Weil [Fri, 7 Aug 2015 20:46:50 +0000 (16:46 -0400)]
osd/osd_types: make pg_missing_t sort order dynamic
Signed-off-by: Sage Weil <sage@redhat.com>
Sage Weil [Fri, 7 Aug 2015 20:27:43 +0000 (16:27 -0400)]
osd: refuse to boot if SORTBITWISE not set but backend cannot sort nibblewise
This prevenents OSDs with new backends from starting if the
cluster is still using the legacy sort order.
Signed-off-by: Sage Weil <sage@redhat.com>
Noah Watkins [Mon, 17 Aug 2015 13:00:39 +0000 (07:00 -0600)]
Merge pull request #5586 from ceph/wip-jni-loader
wip-jni-loader
Signed-off-by: Noah Watkins <noahwatkins@gmail.com>
Kefu Chai [Mon, 17 Aug 2015 05:20:34 +0000 (13:20 +0800)]
Merge pull request #5588 from zhouyuan/isal_yasm_fix
configure: Fix checking for yasm compatibility
Reviewed-by: Kefu Chai <kchai@redhat.com>
Yuan Zhou [Mon, 17 Aug 2015 01:52:19 +0000 (09:52 +0800)]
configure: Fix checking for yasm compability
Fix typo when checking yasm
Signed-off-by: Yuan Zhou <yuan.zhou@intel.com>
Noah Watkins [Sun, 9 Aug 2015 20:12:48 +0000 (14:12 -0600)]
java: add libcommon to deps
Why do indirect dependencies seem to work randomly:
undefined symbol: _ZN4ceph3log3Log12create_entryEii
Signed-off-by: Noah Watkins <noahwatkins@gmail.com>
Noah Watkins [Sun, 9 Aug 2015 19:33:22 +0000 (13:33 -0600)]
java: search for JNI bits in common dirs
Signed-off-by: Noah Watkins <noahwatkins@gmail.com>
Kefu Chai [Sat, 15 Aug 2015 03:56:22 +0000 (11:56 +0800)]
Merge pull request #5572 from xinxinsh/wip-rados-bench-error
fix print error of rados bench
Reviewed-by: Kefu Chai <kchai@redhat.com>
s09816 [Sat, 15 Aug 2015 02:31:57 +0000 (22:31 -0400)]
rbd:improve the error handle of rbd,check the return value.
Signed-off-by: s09816 <shi.lu@h3c.com>
Josh Durgin [Fri, 14 Aug 2015 22:50:53 +0000 (15:50 -0700)]
Merge pull request #5443 from ceph/wip-wrlock
cleanup: remove all traces of rados 'lock' operations
Reviewed-by: Josh Durgin <jdurgin@redhat.com>
Josh Durgin [Fri, 14 Aug 2015 21:18:12 +0000 (14:18 -0700)]
Merge remote-tracking branch 'origin/next'
Josh Durgin [Fri, 14 Aug 2015 21:03:56 +0000 (14:03 -0700)]
Merge pull request #5537 from s09816/master
rbd:modify the log of purging snaps so that it is more appropriate
Reviewed-by: Josh Durgin <jdurgin@redhat.com>
Josh Durgin [Fri, 14 Aug 2015 21:03:15 +0000 (14:03 -0700)]
Merge pull request #5560 from solesoul1127/master
rbd:Check the dest image name, if it is empty string, return error and give a message
Reviewed-by: Josh Durgin <jdurgin@redhat.com>
xinxin shu [Thu, 13 Aug 2015 03:57:58 +0000 (11:57 +0800)]
fix print error of rados bench
Total time run: 12.279167
Total writes made: 92
Write size:
4194304
Bandwidth (MB/sec): 30
Stddev Bandwidth: 23.4
Max bandwidth (MB/sec): 64
Min bandwidth (MB/sec): 2
Average IOPS: 7
Stddev IOPS: 6
Max IOPS: 32767
Min IOPS: -
1537890352
Average Latency: 2.12
Stddev Latency: 1.35
Max latency: 6.05
Min latency: 0.501
Signed-off-by: xinxin shu <xinxin.shu@intel.com>
Yan, Zheng [Fri, 14 Aug 2015 07:45:47 +0000 (15:45 +0800)]
Merge pull request #5569 from ceph/wip-unused-var
client: fix unused var warning
s09816 [Fri, 14 Aug 2015 03:24:41 +0000 (23:24 -0400)]
rbd:modify the log of purging snaps so that it is more appropriate.
Signed-off-by: s09816 <shi.lu@h3c.com>
Josh Durgin [Thu, 13 Aug 2015 20:04:19 +0000 (13:04 -0700)]
Merge pull request #4744 from ceph/wip-11625
librbd: diff_iterate should issue concurrent ops
Reviewed-by: Josh Durgin <jdurgin@redhat.com>
Loic Dachary [Thu, 13 Aug 2015 19:11:03 +0000 (21:11 +0200)]
Merge pull request #5555 from dachary/wip-mon-test-timeouts
tests: be more generous with mon tests timeouts
Reviewed-by: Kefu Chai <kchai@redhat.com>
Sage Weil [Thu, 13 Aug 2015 19:03:34 +0000 (15:03 -0400)]
Merge pull request #5576 from liewegas/wip-dencoder
simplify handling for objects w/ nondeterministic encoding
Reviewed-by: John Spray <john.spray@redhat.com>
Orit Wasserman [Thu, 13 Aug 2015 13:02:03 +0000 (15:02 +0200)]
rgw: we should not overide Swift sent content type
Fixes: #12363
backport: hammer
Signed-off-by: Orit Wasserman <owasserm@redhat.com>
Sage Weil [Thu, 13 Aug 2015 18:49:40 +0000 (14:49 -0400)]
test/encoding/check-generated: test sorted json dumps for nondeterministic objects
Nondeterministic objects dump nondeterministically (usually due
to unordered_map or _set). Compare their sorted output. This
is a weaker test but is better than nothing.
Signed-off-by: Sage Weil <sage@redhat.com>
Sage Weil [Thu, 13 Aug 2015 18:37:32 +0000 (14:37 -0400)]
Revert "osd/HitSet: make subclasses dump deterministically"
This reverts commit
34c048b3cb2b16658e692ac5110115612031510d .
Sage Weil [Thu, 13 Aug 2015 18:37:16 +0000 (14:37 -0400)]
Revert "mon/PGMap: dump osd_epochs in deterministic order"
This reverts commit
3df6438cd2c9a77a06b7a8d011bafbed00f1fc5e .
Sage Weil [Thu, 13 Aug 2015 15:35:12 +0000 (11:35 -0400)]
Merge pull request #4864 from stevecapperlinaro/for-ceph/journal-page-size-fix
FileJournal: Remove CEPH_PAGE_SIZE assumptions
Reviewed-by: Sage Weil <sage@redhat.com>