]>
git.apps.os.sepia.ceph.com Git - ceph.git/log
Jason Dillaman [Tue, 7 Jun 2016 05:01:09 +0000 (01:01 -0400)]
librbd: flush journal commit positions before starting op
Ensure all IO has been properly flushed and committed to the
journal before starting an op that could affect the IO path.
Signed-off-by: Jason Dillaman <dillaman@redhat.com>
Jason Dillaman [Tue, 7 Jun 2016 04:11:34 +0000 (00:11 -0400)]
librbd: leaked AioCompletion if IO issued after image shut down
Signed-off-by: Jason Dillaman <dillaman@redhat.com>
Jason Dillaman [Mon, 6 Jun 2016 23:00:47 +0000 (19:00 -0400)]
librbd: resize and snap create can hang on journal replay failure
Signed-off-by: Jason Dillaman <dillaman@redhat.com>
Jason Dillaman [Mon, 6 Jun 2016 18:54:23 +0000 (14:54 -0400)]
librbd: missing journal state for flushing/restarting replay
Fixes: http://tracker.ceph.com/issues/16077
Signed-off-by: Jason Dillaman <dillaman@redhat.com>
Jason Dillaman [Fri, 13 May 2016 01:22:56 +0000 (21:22 -0400)]
librbd: track IO in the order issued
Fixes: http://tracker.ceph.com/issues/15034
Signed-off-by: Jason Dillaman <dillaman@redhat.com>
Jason Dillaman [Thu, 12 May 2016 19:15:04 +0000 (15:15 -0400)]
librbd: AioCompletion should retrieve CephContext from ImageCtx
Signed-off-by: Jason Dillaman <dillaman@redhat.com>
Jason Dillaman [Mon, 9 May 2016 22:17:49 +0000 (18:17 -0400)]
librbd: Journal no longer requires AioCompletion for IO events
Signed-off-by: Jason Dillaman <dillaman@redhat.com>
Jason Dillaman [Mon, 9 May 2016 21:59:09 +0000 (17:59 -0400)]
librbd: bubble journal error up via existing AioObjectRequests
The journal cannot directly fail the AioCompletion because it
might have one-or-more outstanding requests associated with
it.
Fixes: http://tracker.ceph.com/issues/15791
Signed-off-by: Jason Dillaman <dillaman@redhat.com>
Kefu Chai [Tue, 7 Jun 2016 04:05:26 +0000 (12:05 +0800)]
Merge pull request #9296 from tchaikov/wip-test-chain_xattr
test: Fix the order of ASSERT_EQ() in chain_xattr tests
Reviewed-by: Brad Hubbard <bhubbard@redhat.com>
Kefu Chai [Tue, 7 Jun 2016 04:03:19 +0000 (12:03 +0800)]
Merge pull request #9361 from stiopaa1/osd_osd_override
osd/OSD.h: add override to virtual functions
Reviewed-by: Kefu Chai <kchai@redhat.com>
Casey Bodley [Mon, 6 Jun 2016 21:55:04 +0000 (17:55 -0400)]
Merge pull request #9527 from linuxbox2/rgw-ldap
rgw ldap: fix ldap bindpw parsing
Reviewed-by: Casey Bodley <cbodley@redhat.com>
Matt Benjamin [Mon, 6 Jun 2016 20:19:17 +0000 (16:19 -0400)]
rgw ldap: fix ldap bindpw parsing
Also add additional LDAP debugging output at 0, 10, and 15 to make
troubleshooting easier.
Fixes DN search issue using QE configuration of MS AD.
Signed-off-by: Matt Benjamin <mbenjamin@redhat.com>
Samuel Just [Mon, 6 Jun 2016 20:53:06 +0000 (13:53 -0700)]
Merge pull request #9525 from liewegas/wip-wq
unittest_workqueue: fix wq test for 0 threads
Reviewed-by: Samuel Just <sjust@redhat.com>
Josh Durgin [Mon, 6 Jun 2016 19:57:37 +0000 (12:57 -0700)]
Merge pull request #9517 from idryomov/wip-rbd-concurrent
krbd: don't segfault if images are unmapped concurrently
Reviewed-by: Josh Durgin <jdurgin@redhat.com>
Samuel Just [Mon, 6 Jun 2016 19:23:46 +0000 (12:23 -0700)]
Merge pull request #9432 from ceph/wip-16113
Clean up hobject_t::max() values in decode, avoid comparing directly to get_max()
Reviewed-by: Sage Weil <sage@redhat.com>
Sage Weil [Mon, 6 Jun 2016 18:20:30 +0000 (14:20 -0400)]
unittest_workqueue: fix wq test for 0 threads
In
81517aea0696e4e62c2eb0830d7b0bd378c57230 we added
support for a wq with 0 worker threads. Fix unit test
accordingly.
Signed-off-by: Sage Weil <sage@redhat.com>
Sage Weil [Mon, 6 Jun 2016 18:18:22 +0000 (14:18 -0400)]
Merge pull request #9522 from ceph/wip-rocksdb
Makefile-rocksdb: updates
Samuel Just [Fri, 3 Jun 2016 00:38:05 +0000 (17:38 -0700)]
hobject: enforce max canonical value
Signed-off-by: Samuel Just <sjust@redhat.com>
Samuel Just [Fri, 3 Jun 2016 00:13:09 +0000 (17:13 -0700)]
src/: remove all direct comparisons to get_max()
get_max() now returns a special singleton type from which hobject_t's
can be assigned and constructed, but which cannot be directly compared.
This patch also cleans up all such uses to use is_max() instead.
This should prevent some issues like 16113 by preventing us from
checking for max-ness by comparing against a sentinel value. The more
complete fix will be to make all fields of hobject_t private and enforce
a canonical max() representation that way. That patch will be hard to
backport, however, so we'll settle for this for now.
Fixes: http://tracker.ceph.com/issues/16113
Signed-off-by: Samuel Just <sjust@redhat.com>
Sage Weil [Mon, 6 Jun 2016 16:11:34 +0000 (12:11 -0400)]
Merge pull request #9518 from cbodley/wip-mds-override
mds: fix MDRequestImpl::print override
Reviewed-by: John Spray <jspray@redhat.com>
Sage Weil [Mon, 6 Jun 2016 15:16:52 +0000 (11:16 -0400)]
Makefile-rocksdb: updates
Signed-off-by: Sage Weil <sage@redhat.com>
Casey Bodley [Mon, 6 Jun 2016 14:06:41 +0000 (10:06 -0400)]
Merge pull request #9506 from tchaikov/wip-cmake
cmake: s/BUILD_SHARED_LIBS/ENABLE_SHARED/
Reviewed-by: Casey Bodley <cbodley@redhat.com>
Casey Bodley [Mon, 6 Jun 2016 13:45:49 +0000 (09:45 -0400)]
mds: fix MDRequestImpl::print override
Fixes compiler warnings:
In file included from /home/cbodley/ceph/src/mds/Locker.cc:22:0:
/home/cbodley/ceph/src/mds/Mutation.h:161:16: warning: ‘virtual void
MutationImpl::print(std::ostream&) const’ was hidden
[-Woverloaded-virtual]
virtual void print(ostream &out) const {
^
/home/cbodley/ceph/src/mds/Mutation.h:340:8: warning: by ‘void
MDRequestImpl::print(std::ostream&)’ [-Woverloaded-virtual]
void print(ostream &out);
Signed-off-by: Casey Bodley <cbodley@redhat.com>
Casey Bodley [Mon, 6 Jun 2016 13:37:07 +0000 (09:37 -0400)]
Merge pull request #9512 from Yan-waller/yj-wip-rgwdocs
doc/rgw: correct radosgw-admin command name
Reviewed-by: Casey Bodley <cbodley@redhat.com>
Sage Weil [Mon, 6 Jun 2016 13:28:50 +0000 (09:28 -0400)]
Merge pull request #9175 from majianpeng/bluestore-sync-wal-apply
os/bluestore: If using sync_wal_apply, no need start wal_tp thread.
Reviewed-by: Sage Weil <sage@redhat.com>
Sage Weil [Mon, 6 Jun 2016 13:27:40 +0000 (09:27 -0400)]
Merge pull request #9470 from liewegas/wip-bluestore
os/bluestore: shard cache to match osd op wq sharding
Sage Weil [Mon, 6 Jun 2016 13:26:56 +0000 (09:26 -0400)]
Merge pull request #9516 from liewegas/wip-bluestore-fixes
os/bluestore: misc fixes
Reviewed-by: Sage Weil <sage@redhat.com>
xie xingguo [Thu, 2 Jun 2016 11:29:01 +0000 (19:29 +0800)]
bluestore: add "0x" prefix for tracing messages
Signed-off-by: xie xingguo <xie.xingguo@zte.com.cn>
xie xingguo [Sat, 4 Jun 2016 09:43:58 +0000 (17:43 +0800)]
bluestore: simplify flag translation logic
Signed-off-by: xie xingguo <xie.xingguo@zte.com.cn>
xie xingguo [Sat, 4 Jun 2016 10:53:58 +0000 (18:53 +0800)]
bluestore: fix is_allocated() method of bluestore_blob_t
Signed-off-by: xie xingguo <xie.xingguo@zte.com.cn>
Sage Weil [Mon, 6 Jun 2016 12:52:49 +0000 (08:52 -0400)]
Merge pull request #9254 from branch-predictor/bp-lockdep-on-diet
lockdep: Lockdep on diet
Reviewed-by: Kefu Chai <kchai@redhat.com>
Ilya Dryomov [Fri, 3 Jun 2016 15:24:48 +0000 (17:24 +0200)]
krbd: don't segfault if images are unmapped concurrently
"rbd map c" can die from a NULL dereference on any of this_pool,
this_image or this_snap in wait_for_udev_add():
<image a is mapped>
rbd map c
rbd map b
rbd unmap a
rbd unmap b
However unlikely, this segfault is triggered by the rbd/concurrent.sh
workunit on a regular basis.
Similarly, "rbd showmapped" can die if an image to be listed is
unmapped.
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
Kefu Chai [Mon, 6 Jun 2016 12:48:30 +0000 (20:48 +0800)]
Merge pull request #9384 from stiopaa1/osd_replicatedpg_override
osd/ReplicatedPG.h: add override to virtual funct
Reviewed-by: Casey Bodley <cbodley@redhat.com>
Reviewed-by: Kefu Chai <kchai@redhat.com>
Sage Weil [Mon, 6 Jun 2016 12:46:25 +0000 (08:46 -0400)]
Merge pull request #9497 from Yan-waller/yj-wip-fixosdmaptools
tools/osdmaptool: fix some errors
Reviewed-by: xie xingguo <xie.xingguo@zte.com.cn>
Kefu Chai [Mon, 6 Jun 2016 12:46:00 +0000 (20:46 +0800)]
Merge pull request #9490 from stiopaa1/osd_replicatedpg_addConstOutputOp
osd/ReplicatedPG.h: add const to operator<< param
Reviewed-by: Kefu Chai <kchai@redhat.com>
Sage Weil [Mon, 6 Jun 2016 12:42:45 +0000 (08:42 -0400)]
Merge pull request #9499 from Yan-waller/yj-wip-fixrgwwarning
rgw/op: kill compiling warning
Reviewed-by: Sage Weil <sage@redhat.com>
Sage Weil [Mon, 6 Jun 2016 12:42:03 +0000 (08:42 -0400)]
Merge pull request #9501 from stiopaa1/all_operatorOutput
all: add const to operator<< param
Reviewed-by: Sage Weil <sage@redhat.com>
Sage Weil [Mon, 6 Jun 2016 12:24:02 +0000 (08:24 -0400)]
rocksdb: pull up to master
This includes a few patches from Willem Jan Withagen
<wjw@digiware.nl> that fix the build on FreeBSD.
Signed-off-by: Sage Weil <sage@redhat.com>
Yan Jun [Mon, 6 Jun 2016 11:45:16 +0000 (19:45 +0800)]
doc/rgw: correct radosgw-admin command name
Signed-off-by: Yan Jun <yan.jun8@zte.com.cn>
John Spray [Mon, 6 Jun 2016 09:47:29 +0000 (10:47 +0100)]
Merge pull request #8598 from renhwsky/wip-renhw-ceph_mds-help
mds: fix usage() display
Reviewed-by: John Spray <john.spray@redhat.com>
John Spray [Mon, 6 Jun 2016 09:46:59 +0000 (10:46 +0100)]
Merge pull request #8641 from renhwztetecs/renhw-wip-mds-cleanup
mds: MDSDaemon::init bug fix
Reviewed-by: John Spray <john.spray@redhat.com>
John Spray [Mon, 6 Jun 2016 09:45:29 +0000 (10:45 +0100)]
Merge pull request #8753 from renhwztetecs/renhw-wip-fuse-fuse_file_info
client: fuse_file_info.fh type should be uint64_t
Reviewed-by: John Spray <john.spray@redhat.com>
John Spray [Mon, 6 Jun 2016 09:43:37 +0000 (10:43 +0100)]
Merge pull request #8777 from ukernel/wip-allow-dirfrags
mds: fix 'allows_dirfrags' check for fragment_dir command
Reviewed-by: John Spray <john.spray@redhat.com>
John Spray [Mon, 6 Jun 2016 09:42:59 +0000 (10:42 +0100)]
Merge pull request #8799 from renhwztetecs/renhw-wip-fusemain
client/fuse: object release order error
Reviewed-by: Yan, Zheng <zyan@redhat.com>
Mykola Golub [Mon, 6 Jun 2016 07:20:54 +0000 (10:20 +0300)]
Merge pull request #9459 from dillaman/wip-16123
test/rbd: fsx needs to shut down the journal
Reviewed-by: Mykola Golub <mgolub@mirantis.com>
Kefu Chai [Mon, 6 Jun 2016 04:42:02 +0000 (12:42 +0800)]
cmake: s/BUILD_SHARED_LIBS/ENABLE_SHARED/
Signed-off-by: Kefu Chai <kchai@redhat.com>
Kefu Chai [Mon, 6 Jun 2016 02:57:23 +0000 (10:57 +0800)]
Merge pull request #9503 from stiopaa1/common_cephjson_duplicateDeclaration
common/ceph_json.h: remove duplicate declaration
Reviewed-by: Kefu Chai <kchai@redhat.com>
Michal Jarzabek [Sun, 5 Jun 2016 21:15:48 +0000 (22:15 +0100)]
common/ceph_json.h: remove duplicate declaration
Signed-off-by: Michal Jarzabek <stiopa@gmail.com>
Michal Jarzabek [Sat, 4 Jun 2016 22:24:06 +0000 (23:24 +0100)]
all: add const to operator<< param
Signed-off-by: Michal Jarzabek <stiopa@gmail.com>
Kefu Chai [Sun, 5 Jun 2016 06:15:02 +0000 (14:15 +0800)]
Merge pull request #9174 from majianpeng/osd-ms-dispatch
OSD: cleanup code for ms_dispatch
Reviewed-by: Sage Weil <sage@redhat.com>
Reviewed-by: Samuel Just <sjust@redhat.com>
Reviewed-by: Kefu Chai <kchai@redhat.com>
Kefu Chai [Sun, 5 Jun 2016 06:01:23 +0000 (14:01 +0800)]
Merge pull request #9178 from wjwithagen/wip-wjw-freebsd-compile-2
build: freebsd changes (part 2)
Reviewed-by: Kefu Chai <kchai@redhat.com>
Ali Maredia [Sat, 4 Jun 2016 22:28:40 +0000 (18:28 -0400)]
Merge pull request #9446 from ceph/wip-cmake
cmake changes
Reviewed-by: Ali Maredia <amaredia@redhat.com>
Samuel Just [Fri, 3 Jun 2016 00:39:09 +0000 (17:39 -0700)]
PG::replica_scrub: don't adjust pool on max object
Signed-off-by: Samuel Just <sjust@redhat.com>
Samuel Just [Fri, 3 Jun 2016 00:36:21 +0000 (17:36 -0700)]
hobject: compensate for non-canonical hobject_t::get_max() encodings
This closes a loop-hole that could allow a non-canonical in memory
hobject_t::get_max() object which would return true for is_max(), but
false for *this == hobject_t::get_max().
Fixes: http://tracker.ceph.com/issues/16113
Signed-off-by: Samuel Just <sjust@redhat.com>
Yan Jun [Sat, 4 Jun 2016 12:08:11 +0000 (20:08 +0800)]
rgw/op: kill compiling warning
rgw/rgw_op.cc: In member function 'virtual int RGWCreateBucket::verify_permission()':
rgw/rgw_op.cc:1810:37: warning: comparison between signed and unsigned integer expressions [-Wsign-compare]
if (buckets.count() >= s->user->max_buckets) {
Signed-off-by: Yan Jun <yan.jun8@zte.com.cn>
Yan Jun [Sat, 4 Jun 2016 09:08:24 +0000 (17:08 +0800)]
tools/osdmaptool: fix error message if we are unable to parse pgid
Signed-off-by: Yan Jun <yan.jun8@zte.com.cn>
Yan Jun [Sat, 4 Jun 2016 09:06:40 +0000 (17:06 +0800)]
tools/osdmaptool: exit if we are unable to resolve pg_num
Signed-off-by: Yan Jun <yan.jun8@zte.com.cn>
Piotr Dałek [Sat, 23 Apr 2016 12:06:42 +0000 (14:06 +0200)]
lockdep: Convert free_ids from list to bitmap
Convert "free_ids" list to bitmap, reducing its memory usage from
16 + MAX_LOCKS * 20 (~20KB) to MAX_LOCKS/8 (512b) and save some time
(re)populating the free id array.
Signed-off-by: Piotr Dałek <git@predictor.org.pl>
Kefu Chai [Sat, 4 Jun 2016 01:50:00 +0000 (09:50 +0800)]
cmake: use ENABLE_SHARED instead of BUILD_SHARED_LIBS
we decided to drop the static libraries from the *-dev debian packages.
so no more tricky changes for supporting creating both libraries.
* partially revert
1dbfb26 : s/BUILD_SHARED_LIBS/ENABLE_SHARED/
- add_library() command use BUILD_SHARED_LIBS as the default library
type, and all intermediate libraries should static ones. it would
cause unnecessary confusion if some developer uses add_library() without
specifying the library type, and ends up with a .so. so we use our
own setting variables for specifying the type of user facing libraries.
* revert
ac47440
- the BUILD_SHARED_LIBS setting is also populated to the gmock libraries,
which should be compiled as static ones. otherwise the unit tests will
crash on exit.
Signed-off-by: Kefu Chai <kchai@redhat.com>
Kefu Chai [Fri, 3 Jun 2016 05:39:04 +0000 (13:39 +0800)]
cmake: install binaries used by ceph-qa-suite
they are:
- ceph_test_rados: used by
tasks/rados.py
- ceph_test_rados_delete_pools_parallel: used by
suites/rados/monthrash/workloads/pool-create-delete.yaml
- ceph_test_filejournal
suites/rados/objectstore/filejournal.yaml
- ceph_test_objectstore: used by
suites/rados/objectstore/objectstore.yaml
- ceph_test_{async_driver,msgr}: used by
suites/rados/singleton-nomsgr/all/msgr.yaml
Signed-off-by: Kefu Chai <kchai@redhat.com>
Kefu Chai [Fri, 3 Jun 2016 05:32:44 +0000 (13:32 +0800)]
cmake: compile more radosstriper tests
Signed-off-by: Kefu Chai <kchai@redhat.com>
Kefu Chai [Thu, 2 Jun 2016 03:51:36 +0000 (11:51 +0800)]
cmake: do not link ceph_test_rgw_*, ceph-dencoder against cls_kvs
it is not used, and causes link errors.
Signed-off-by: Kefu Chai <kchai@redhat.com>
Kefu Chai [Thu, 2 Jun 2016 03:37:18 +0000 (11:37 +0800)]
cmake: link ceph-dencoder against os lib
os is a top level target which is linked into executables or user facing
libraries. so it is not included by any libraries linked by
ceph-dencoder so far.
Signed-off-by: Kefu Chai <kchai@redhat.com>
Kefu Chai [Thu, 2 Jun 2016 03:25:26 +0000 (11:25 +0800)]
cmake: link rbd cli against krbd_objs directly
rbd cli => Kernel.cc => krbd_create_from_context() <= krbd.cc
Signed-off-by: Kefu Chai <kchai@redhat.com>
Kefu Chai [Thu, 2 Jun 2016 03:15:26 +0000 (11:15 +0800)]
cmake: fix linkage of ceph_test_cls_${name}
they should link against cls_${name}_client, instead of against
cls_${name} directly.
Signed-off-by: Kefu Chai <kchai@redhat.com>
Kefu Chai [Wed, 1 Jun 2016 18:08:45 +0000 (02:08 +0800)]
makefile: install mount.fuse.ceph,mount.ceph into /usr/sbin
Signed-off-by: Kefu Chai <kchai@redhat.com>
Kefu Chai [Wed, 1 Jun 2016 17:50:21 +0000 (01:50 +0800)]
cmake: install ceph-disk to /usr/sbin
* add keyword "INSTALL_SCRIPT" to distutils_install_module(), so we can
override the install path of ceph-disk script.
* refactor the Distutils.cmake module a little bit, the ${option}
variable out lives the CODE snipplet. so we need to reset it
at the beginning.
Signed-off-by: Kefu Chai <kchai@redhat.com>
Michal Jarzabek [Fri, 3 Jun 2016 21:24:38 +0000 (22:24 +0100)]
osd/ReplicatedPG.h: add const to operator<< param
Signed-off-by: Michal Jarzabek <stiopa@gmail.com>
Sage Weil [Fri, 3 Jun 2016 20:42:35 +0000 (16:42 -0400)]
os/bluestore: paranoid checks on onode_map vs lru
Signed-off-by: Sage Weil <sage@redhat.com>
Sage Weil [Fri, 3 Jun 2016 20:41:37 +0000 (16:41 -0400)]
os/bluestore: replace broken get_next with map_any
The get_next implementation assumed the lru was local to the OnodeSpace,
but it includes the whole cache now.
Signed-off-by: Sage Weil <sage@redhat.com>
Sage Weil [Thu, 2 Jun 2016 20:44:03 +0000 (16:44 -0400)]
os/bluestore: better audit_lru debugging
Signed-off-by: Sage Weil <sage@redhat.com>
Sage Weil [Thu, 2 Jun 2016 20:33:01 +0000 (16:33 -0400)]
os/bluestore: shard the cache
Signed-off-by: Sage Weil <sage@redhat.com>
Sage Weil [Thu, 2 Jun 2016 20:19:04 +0000 (16:19 -0400)]
osd: communicate shard count to ObjectStore
Signed-off-by: Sage Weil <sage@redhat.com>
Sage Weil [Thu, 2 Jun 2016 20:03:00 +0000 (16:03 -0400)]
osd: use spg_t method to hash pg to (wq) shard
Signed-off-by: Sage Weil <sage@redhat.com>
Sage Weil [Thu, 2 Jun 2016 19:53:32 +0000 (15:53 -0400)]
os/bluestore: unify lrus into single Cache class
This will be sharded soon, but for now there's just one instance.
Signed-off-by: Sage Weil <sage@redhat.com>
Sage Weil [Fri, 3 Jun 2016 17:54:08 +0000 (13:54 -0400)]
Merge pull request #9483 from ceph/wip-rocksdb
rocksdb: fix build
Casey Bodley [Fri, 3 Jun 2016 17:46:11 +0000 (13:46 -0400)]
Merge pull request #9477 from tchaikov/wip-fix-cmake-unittests
cmake: do not create gtest/gmock shared libraries
Reviewed-by: Casey Bodley <cbodley@redhat.com>
Kefu Chai [Fri, 3 Jun 2016 03:53:10 +0000 (11:53 +0800)]
cmake: do not create gtest/gmock shared libraries
otherwise unit tests crash at exit.
Signed-off-by: Kefu Chai <kchai@redhat.com>
Jianpeng Ma [Mon, 23 May 2016 12:17:53 +0000 (20:17 +0800)]
os/bluestore/BlueStore: optimize WALWQ::flush()
In fact ThreadPool::drain will do
a)wait WorkQueue::empty() is true
b)wait processing thread to complete.
So it can remove the same work which wait WorkQueue::empty().
Signed-off-by: Jianpeng Ma <jianpeng.ma@intel.com>
Jianpeng Ma [Fri, 3 Jun 2016 16:32:22 +0000 (00:32 +0800)]
os/bluestore/BlueStore: Don't support dynamic set wal_apply mode.
Signed-off-by: Jianpeng Ma <jianpeng.ma@intel.com>
Jianpeng Ma [Fri, 20 May 2016 13:24:31 +0000 (21:24 +0800)]
common/WorkQueue: Make decrease thread num to zero.
Signed-off-by: Jianpeng Ma <jianpeng.ma@intel.com>
Jianpeng Ma [Fri, 20 May 2016 11:27:41 +0000 (19:27 +0800)]
common/config: make md_config_t::apply_changes recursive.
This make in md_config_obs_t::handle_conf_change to call apply_changes.
Signed-off-by: Jianpeng Ma <jianpeng.ma@intel.com>
Jianpeng Ma [Wed, 18 May 2016 16:27:15 +0000 (00:27 +0800)]
os/bluestore: If using sync_wal_apply, no need start wal_tp thread.
Signed-off-by: Jianpeng Ma <jianpeng.ma@intel.com>
Casey Bodley [Fri, 3 Jun 2016 15:34:07 +0000 (11:34 -0400)]
Merge pull request #9381 from weiqiaomiao/wqm-wip-rgw-align
rgw:fix some indentation errors of "radosgw-admin --help"
Reviewed-by: Casey Bodley <cbodley@redhat.com>
Sage Weil [Fri, 3 Jun 2016 15:22:45 +0000 (11:22 -0400)]
Makefile-rocksdb: remove .d files
Not really sure why autotools is removing these... but
whatever.
Signed-off-by: Sage Weil <sage@redhat.com>
Kefu Chai [Fri, 3 Jun 2016 15:09:49 +0000 (23:09 +0800)]
Merge pull request #9480 from ErwanAliasr1/evelu-cephtool
cephtool/test.sh: Adding iterations numbers in output log
Reviewed-by: Kefu Chai <kchai@redhat.com>
Sage Weil [Thu, 2 Jun 2016 19:04:03 +0000 (15:04 -0400)]
os/bluestore: unique type names or onode and buffer lrus
Signed-off-by: Sage Weil <sage@redhat.com>
Sage Weil [Thu, 2 Jun 2016 18:39:15 +0000 (14:39 -0400)]
os/bluestore: take collection lock when trimming collection cache
Signed-off-by: Sage Weil <sage@redhat.com>
Sage Weil [Thu, 2 Jun 2016 17:56:46 +0000 (13:56 -0400)]
os/bluestore: audit LRU at various points
This is purely for debugging.
Signed-off-by: Sage Weil <sage@redhat.com>
Sage Weil [Thu, 2 Jun 2016 19:52:12 +0000 (15:52 -0400)]
os/bluestore: don't print bnode key
Signed-off-by: Sage Weil <sage@redhat.com>
Sage Weil [Thu, 2 Jun 2016 17:55:04 +0000 (13:55 -0400)]
os/bluestore: fix buffercache discard bug
Signed-off-by: Sage Weil <sage@redhat.com>
Sage Weil [Fri, 3 Jun 2016 14:46:26 +0000 (10:46 -0400)]
Merge pull request #9465 from liewegas/wip-bluestore-zero
os/bluestore: remove aio_zero from bdev interface
Sage Weil [Fri, 3 Jun 2016 14:45:44 +0000 (10:45 -0400)]
Merge pull request #9417 from chhabaramesh/master
os/bluestore: Tree based bitmapalloc
Sage Weil [Fri, 3 Jun 2016 14:41:35 +0000 (10:41 -0400)]
Merge pull request #9466 from liewegas/wip-rocksdb
rocksdb: pull up to latest master
Sage Weil [Fri, 3 Jun 2016 14:01:37 +0000 (10:01 -0400)]
Makefile-rocksdb.am: update
Signed-off-by: Sage Weil <sage@redhat.com>
Orit Wasserman [Fri, 3 Jun 2016 13:18:15 +0000 (15:18 +0200)]
Merge pull request #9457 from yehudasa/wip-14534
rgw: can set negative max_buckets on RGWUserInfo
Reviewed-by: Orit Wasserman <owasserm@redhat.com>
Sage Weil [Fri, 3 Jun 2016 12:54:06 +0000 (08:54 -0400)]
unittest_bluefs: fix more
- wait for aio to finish
- shut down bdev, which stops the aio completion thread
- *then* deallocate the IOContexts
Signed-off-by: Sage Weil <sage@redhat.com>
Erwan Velu [Fri, 3 Jun 2016 12:53:00 +0000 (14:53 +0200)]
cephtool/test.sh: Adding iterations numbers in output log
When this test is failing and reach the limits, reading the log doesn't make
obvious that we reach them.
This simple patch adds the iterations numbers inside the output log.
Signed-off-by: Erwan Velu <erwan@redhat.com>
Kefu Chai [Fri, 3 Jun 2016 12:36:35 +0000 (20:36 +0800)]
Merge pull request #9399 from mikulely/update-ceph-deploy-link
doc: update broken link for deploy ceph by Chef
Reviewed-by: Kefu Chai <kchai@redhat.com>
Jason Dillaman [Fri, 3 Jun 2016 12:08:19 +0000 (08:08 -0400)]
Merge pull request #9441 from xinxinsh/wip-16706
librbd: remove invalid object map flag when object map is disabled
Reviewed-by: Jason Dillaman <dillaman@redhat.com>