]>
git.apps.os.sepia.ceph.com Git - ceph.git/log
Samuel Just [Wed, 11 Feb 2015 18:30:51 +0000 (10:30 -0800)]
ReplicatedPG::on_change: requeue in_progress_async_reads close to last
Fixes: 10768
Signed-off-by: Samuel Just <sjust@redhat.com>
Zhiqiang Wang [Wed, 4 Feb 2015 00:22:41 +0000 (08:22 +0800)]
include/encoding: fix an compile warning
./include/encoding.h:317:7: warning: 't' may be used uninitialized in
this function [-Wmaybe-uninitialized]
Signed-off-by: Zhiqiang Wang <zhiqiang.wang@intel.com>
Reviewed-by: Sage Weil <sage@redhat.com>
Sage Weil [Wed, 4 Feb 2015 20:55:59 +0000 (12:55 -0800)]
Merge pull request #3619 from ceph/wip-8903
librbd: removed lingering TODO message from ImageWatcher
Reviewed-by: Sage Weil <sage@redhat.com>
Mykola Golub [Fri, 23 Jan 2015 09:47:45 +0000 (11:47 +0200)]
msg: fixup for
2ffacbe (crc configuration in messenger)
* Update the remaining Message::encode() calls, which now expect crc
flags to be passed as a flags parameter instead of a bool (this e.g.
fixes the issue with routing messages forwarded from an older client).
* In Message::encode() data crc is calculated when MSG_CRC_DATA is
set, but in decode_message(), Pipe::read/write_message() data crc is
calculated when MSG_CRC_HEADER is set. Fix this.
Note,
2ffacbe changed the behavior of Pipe::read_message/write_message():
previously the methods always calculated crc, now they calculate it
only if crc is enabled in the config. This means crc can not be disabled
if there are monitors of older version in the cluster.
Signed-off-by: Mykola Golub <mgolub@mirantis.com>
Reviewed-by: Sage Weil <sage@redhat.com>
Gregory Farnum [Wed, 4 Feb 2015 17:30:29 +0000 (09:30 -0800)]
Merge pull request #3613 from rzarzynski/wip-4920
client: add support for O_NOFOLLOW and O_PATH
Reviewed-by: Greg Farnum <gfarnum@redhat.com>
Radoslaw Zarzynski [Tue, 27 Jan 2015 22:25:33 +0000 (23:25 +0100)]
client: implement support for O_PATH on Linux.
Signed-off-by: Radoslaw Zarzynski <rzarzynski@mirantis.com>
Radoslaw Zarzynski [Mon, 26 Jan 2015 14:40:14 +0000 (15:40 +0100)]
client: add support for O_NOFOLLOW in Client::open().
Fixes: #4920
Signed-off-by: Radoslaw Zarzynski <rzarzynski@mirantis.com>
Loic Dachary [Wed, 4 Feb 2015 07:40:57 +0000 (08:40 +0100)]
Merge pull request #3603 from dachary/wip-10729-debian-install-deps
install-deps.sh: get lsb_release if needed
Reviewed-by: Yann Dupont <yann@objoo.org>
Dan Mick [Wed, 4 Feb 2015 04:57:47 +0000 (20:57 -0800)]
Merge pull request #3565 from rzarzynski/wip-7264
pybind/rados.py: fix error hiding and inconsistency on librados load.
Reviewed-by: Dan Mick <dan.mick@redhat.com>
John Wilkins [Tue, 3 Feb 2015 22:50:25 +0000 (14:50 -0800)]
doc: Removed references to Inktank suppport.
Signed-off-by: John Wilkins <jowilkin@redhat.com>
John Wilkins [Tue, 3 Feb 2015 22:14:22 +0000 (14:14 -0800)]
doc: Removed Inktank reference.
Signed-off-by: John Wilkins <jowilkin@redhat.com>
Gregory Farnum [Tue, 3 Feb 2015 18:43:42 +0000 (10:43 -0800)]
Merge pull request #3598 from ceph/wip-10720
mds: fix leaked MDCache::logger
Reviewed-by: Sage Weil <sage@redhat.com>
Reviewed-by: Greg Farnum <gfarnum@redhat.com>
John Wilkins [Tue, 3 Feb 2015 17:21:56 +0000 (09:21 -0800)]
Merge pull request #3262 from dachary/wip-jerasure
erasure-code: update links to jerasure upstream
Reviewed-by: John Wilkins <jowilkin@redhat.com>
Loic Dachary [Tue, 3 Feb 2015 16:40:41 +0000 (17:40 +0100)]
install-deps.sh: babeltrace is not on wheezy
The babeltrace package is not found on wheezy, not just squeeze.
Fix the regular expression that incorrectly replaced
libbabeltrace-ctf-dev with -ctf-dev leading to a strange error
E: Opening configuration file tf-dev - ifstream::ifstream (2: No such file or directory)
when running apt-get install.
Signed-off-by: Loic Dachary <ldachary@redhat.com>
Loic Dachary [Tue, 3 Feb 2015 16:18:09 +0000 (17:18 +0100)]
install-deps.sh: get lsb_release if needed
Debian does not have lsb_release installed by default.
http://tracker.ceph.com/issues/10729 Fixes: #10729
Signed-off-by: Loic Dachary <ldachary@redhat.com>
Loic Dachary [Tue, 3 Feb 2015 15:39:53 +0000 (16:39 +0100)]
Merge pull request #3601 from dachary/wip-10729-fedora-install-deps
install-deps.sh: get lsb_release if needed
Reviewed-by: Haïkel Guémar <hguemar@redhat.com>
Sage Weil [Tue, 3 Feb 2015 15:37:05 +0000 (07:37 -0800)]
doc/release-notes: v0.92
Signed-off-by: Sage Weil <sage@redhat.com>
Loic Dachary [Tue, 3 Feb 2015 15:26:17 +0000 (16:26 +0100)]
install-deps.sh: get lsb_release if needed
Fedora does not have lsb_release installed by default.
http://tracker.ceph.com/issues/10729 Fixes: #10729
Signed-off-by: Loic Dachary <ldachary@redhat.com>
Sage Weil [Tue, 3 Feb 2015 14:52:07 +0000 (06:52 -0800)]
Merge remote-tracking branch 'gh/next'
Sage Weil [Tue, 3 Feb 2015 14:50:22 +0000 (06:50 -0800)]
Jason Dillaman [Tue, 3 Feb 2015 13:57:15 +0000 (08:57 -0500)]
librbd: removed lingering TODO message from ImageWatcher
The TODO was a reminder to integrate with the watch/notify
error handler. This integration was already completed under
previous commits.
Fixes: #8903
Signed-off-by: Jason Dillaman <dillaman@redhat.com>
John Spray [Tue, 3 Feb 2015 11:19:59 +0000 (11:19 +0000)]
mds: fix leaked MDCache::logger
Fixes: #10720
Signed-off-by: John Spray <john.spray@redhat.com>
Jason Dillaman [Mon, 2 Feb 2015 20:08:42 +0000 (15:08 -0500)]
librbd: fixed object map issues discovered via fsx
The object map wasn't being properly refreshed after setting
the snapshot context on the parent image. Additionally fixed
a potential deadlock that could have occurred if no object
map update was required when trimming an image.
Fixes: #10706
Signed-off-by: Jason Dillaman <dillaman@redhat.com>
Reviewed-by: Josh Durgin <jdurgin@redhat.com>
Sage Weil [Tue, 3 Feb 2015 11:23:28 +0000 (03:23 -0800)]
vstart.sh: fix .ceph_port typo
Signed-off-by: Sage Weil <sage@redhat.com>
Loic Dachary [Tue, 3 Feb 2015 10:52:31 +0000 (11:52 +0100)]
Merge pull request #3597 from karelstriegel/master
rdb: update init-rbdmap to fix duplicate mount point
Reviewed-by: Loic Dachary <ldachary@redhat.com>
Karel Striegel [Tue, 3 Feb 2015 10:42:06 +0000 (11:42 +0100)]
rdb: update init-rbdmap to fix duplicate mount point
According to the manual 'mount -f' adds an entry in /etc/mtab.
Therefore the 'mount -v' command needs the flag -n to avoid
duplicate entries in /etc/mtab.
Signed-off-by: Karel Striegel <karel@striegel.be>
Josh Durgin [Tue, 3 Feb 2015 09:28:50 +0000 (10:28 +0100)]
Merge remote-tracking branch 'origin/hammer'
Jason Dillaman [Mon, 2 Feb 2015 16:20:26 +0000 (11:20 -0500)]
librbd: removing objects can lead to infinite loop
The AioRemove state machine is not properly advanced from
_PRE to _FLAT. This will result in an infinite state machine
loop.
Signed-off-by: Jason Dillaman <dillaman@redhat.com>
Reviewed-by: Josh Durgin <jdurgin@redhat.com>
Jason Dillaman [Mon, 2 Feb 2015 15:50:09 +0000 (10:50 -0500)]
librbd: pending AIO operations are now flushed asynchronously
If exclusive locking was enabled, the librbd aio_flush command
would block waiting for queued AIO operations to proceed once
the exclusive lock was obtained. Now librbd will no longer
block when aio_flush is invoked and AIO operations are waiting
on the exclusive lock.
Fixes: #10714
Signed-off-by: Jason Dillaman <dillaman@redhat.com>
Reviewed-by: Josh Durgin <jdurgin@redhat.com>
Sage Weil [Tue, 3 Feb 2015 08:22:52 +0000 (00:22 -0800)]
Merge pull request #3594 from ceph/wip-10707
test/cls_rgw: fix test
Reviewed-by: Sage Weil <sage@redhat.com>
Sage Weil [Tue, 3 Feb 2015 08:12:23 +0000 (00:12 -0800)]
Merge remote-tracking branch 'gh/hammer'
Sage Weil [Tue, 3 Feb 2015 08:12:17 +0000 (00:12 -0800)]
Merge remote-tracking branch 'gh/next'
Sage Weil [Tue, 3 Feb 2015 08:09:48 +0000 (00:09 -0800)]
Merge pull request #3590 from ceph/wip-10722
rgw: move perf cleanup before context cleanup
Reviewed-by: Sage Weil <sage@redhat.com>
Billy Olsen [Mon, 2 Feb 2015 23:24:59 +0000 (16:24 -0700)]
Fix memory leak in python rados bindings
A circular reference was inadvertently created when using the
CFUNCTYPE binding for callbacks for the asynchronous i/o callbacks.
This commit refactors the usage of the callbacks such that the
Ioctx object does not have a class reference to the callbacks.
Fixes: #10723
Backport: giant, firefly, dumpling
Signed-off-by: Billy Olsen <billy.olsen@gmail.com>
Reviewed-by: Dan Mick <dmick@redhat.com>
Reviewed-by: Josh Durgin <jdurgin@redhat.com>
Yehuda Sadeh [Tue, 3 Feb 2015 01:10:19 +0000 (17:10 -0800)]
test/cls_rgw: fix test
Test needed to be updated, now using meta.accounted_size for object
size.
Fixes: #10707
Signed-off-by: Yehuda Sadeh <yehuda@redhat.com>
Loic Dachary [Tue, 3 Feb 2015 00:42:34 +0000 (01:42 +0100)]
Merge pull request #3593 from ceph/wip-autogen
Fix do_autogen.sh so that -L is allowed
Reviewed-by: Loic Dachary <ldachary@redhat.com>
David Zafman [Tue, 3 Feb 2015 00:24:07 +0000 (16:24 -0800)]
Fix do_autogen.sh so that -L is allowed
Signed-off-by: David Zafman <dzafman@redhat.com>
Yehuda Sadeh [Mon, 2 Feb 2015 23:25:22 +0000 (15:25 -0800)]
rgw: move perf cleanup before context cleanup
Fixes: #10722
Fixes: #10572
Backport: giant, firefly
This fixes a regression introduced in commit
de2e5fa048639de6c9ee004a93ab295625fa3b94 .
Signed-off-by: Yehuda Sadeh <yehuda@redhat.com>
Radoslaw Zarzynski [Wed, 28 Jan 2015 16:09:20 +0000 (17:09 +0100)]
pybind: fix error hiding and inconsistency on librados load.
When loading of librados failed, the real cause was hidden in some
circumstances due to exception rewriting. Additionally, the error
reporting was inconsistent -- two paths with different exceptions
and messages existed.
Fixes: #7264
Signed-off-by: Radoslaw Zarzynski <rzarzynski@mirantis.com>
Samuel Just [Mon, 2 Feb 2015 22:18:52 +0000 (14:18 -0800)]
Merge remote-tracking branch 'upstream/next'
Jenkins [Mon, 2 Feb 2015 18:35:28 +0000 (10:35 -0800)]
0.92
Sage Weil [Mon, 2 Feb 2015 17:21:26 +0000 (09:21 -0800)]
Merge pull request #3588 from athanatos/wip-10693
Wip 10693
Reviewed-by: Sage Weil <sage@redhat.com>
Sage Weil [Mon, 2 Feb 2015 17:19:42 +0000 (09:19 -0800)]
Merge pull request #3587 from athanatos/wip-10633
ReplicatedPG::C_ProxyRead: fix dropped lock
Reviewed-by: Sage Weil <sage@redhat.com>
Samuel Just [Mon, 2 Feb 2015 17:18:58 +0000 (09:18 -0800)]
Merge pull request #3538 from athanatos/wip-7862
osd: allow recovery below min size
Reviewed-by: Sage Weil <sage@redhat.com>
Samuel Just [Mon, 2 Feb 2015 17:18:22 +0000 (09:18 -0800)]
Merge pull request #3539 from athanatos/wip-7861
osd: allow writes on degraded objects
Reviewed-by: Sage Weil <sage@redhat.com>
Yehuda Sadeh [Mon, 2 Feb 2015 17:33:03 +0000 (09:33 -0800)]
Merge remote-tracking branch 'origin/next'
Conflicts:
src/rgw/rgw_op.cc
src/rgw/rgw_rados.cc
src/rgw/rgw_rados.h
Josh Durgin [Mon, 2 Feb 2015 16:16:35 +0000 (17:16 +0100)]
Merge pull request #3578 from ceph/wip-10709
rbd: watch command should unwatch before exiting
Reviewed-by: Josh Durgin <jdurgin@redhat.com>
Sage Weil [Mon, 2 Feb 2015 15:54:53 +0000 (07:54 -0800)]
Merge pull request #3583 from jdurgin/wip-qemu-iotests
qa: use correct binary path on rpm-based systems
Reviewed-by: Sage Weil <sage@redhat.com>
Sage Weil [Mon, 2 Feb 2015 15:52:13 +0000 (07:52 -0800)]
Merge pull request #3584 from ceph/wip-optional-encoding
encoding: ignore uninitialized instantiation in boost::optional decode
Reviewed-by: Sage Weil <sage@redhat.com>
Josh Durgin [Mon, 2 Feb 2015 15:43:35 +0000 (16:43 +0100)]
qa: use correct binary path on rpm-based systems
Fixes: #10715
Signed-off-by: Josh Durgin <jdurgin@redhat.com>
Jason Dillaman [Mon, 2 Feb 2015 14:21:13 +0000 (09:21 -0500)]
rbd: watch command should unwatch before exiting
Switched the rbd CLI tool to use the new watch/unwatch API
and properly unwatch the image before exiting. Additionally,
the watch command opened the image twice -- once R/W and another
R/O. This resulted in two watches being created for the image.
Cleaned up old/new format detection by using the existing API
methods.
Fixes: #10709
Signed-off-by: Jason Dillaman <dillaman@redhat.com>
Josh Durgin [Mon, 2 Feb 2015 11:14:44 +0000 (12:14 +0100)]
encoding: ignore uninitialized instantiation in boost::optional decode
POD types won't be initialized until the decode writes to them.
Signed-off-by: Josh Durgin <jdurgin@redhat.com>
Loic Dachary [Mon, 2 Feb 2015 11:12:15 +0000 (12:12 +0100)]
Merge pull request #3562 from liewegas/wip-lttng
do_autogen.sh: default to --with-lttng, -L to build without
Reviewed-by: Loic Dachary <ldachary@redhat.com>
Sage Weil [Mon, 2 Feb 2015 11:07:36 +0000 (03:07 -0800)]
Merge pull request #3485 from jdurgin/wip-merge-diff-testing
Improve merge-diff test
Sage Weil [Mon, 2 Feb 2015 10:47:46 +0000 (02:47 -0800)]
do_autogen.sh: default to --with-lttng, -L to build without
Get developers building with LTTNG by default.
Signed-off-by: Sage Weil <sage@redhat.com>
Sage Weil [Mon, 2 Feb 2015 10:55:39 +0000 (02:55 -0800)]
Merge pull request #3549 from Ved-vampir/master
Ceph: Fix
Reviewed-by: Sage Weil <sage@redhat.com>
Ved-vampir [Fri, 30 Jan 2015 11:33:21 +0000 (13:33 +0200)]
osd: change pg_stat plain to display CRUSH_ITEM_NONE in pgmap output section.
Fixes: #10519
Signed-off-by: Kiseleva Alyona <akiselyova@mirantis.com>
Haomai Wang [Mon, 2 Feb 2015 10:26:08 +0000 (18:26 +0800)]
Merge pull request #3501 from yuyuyu101/wip-fix-async
AsyncMessenger: Fix mix auth assert failed
Josh Durgin [Mon, 2 Feb 2015 09:33:32 +0000 (10:33 +0100)]
Merge branch 'wip-rbd-listwatchers'
Reviewed-by: Sage Weil <sage@redhat.com>
Reviewed-by: Josh Durgin <jdurgin@redhat.com>
Yunchuan Wen [Wed, 21 Jan 2015 00:43:07 +0000 (00:43 +0000)]
man: add rbd status to doc/man/8/rbd.rst
Signed-off-by: Yunchuan Wen <yunchuanwen@ubuntukylin.com>
Reviewed-by: Li Wang <liwang@ubuntukylin.com>
John Spray [Mon, 2 Feb 2015 09:04:32 +0000 (10:04 +0100)]
Merge pull request #3533 from ceph/wip-disaster-doc
doc: add cephfs disaster recovery guidance
Reviewed-by: Greg Farnum <gfarnum@redhat.com
John Spray [Mon, 5 Jan 2015 15:46:38 +0000 (15:46 +0000)]
doc: add cephfs disaster recovery guidance
This is a place to put some useful notes about
the new offline recovery tooling.
Signed-off-by: John Spray <john.spray@redhat.com>
Yunchuan Wen [Wed, 21 Jan 2015 00:42:55 +0000 (00:42 +0000)]
test: add rbd status to src/test/cli/rbd/help.t
Signed-off-by: Yunchuan Wen <yunchuanwen@ubuntukylin.com>
Reviewed-by: Li Wang <liwang@ubuntukylin.com>
Yunchuan Wen [Wed, 21 Jan 2015 00:42:41 +0000 (00:42 +0000)]
rbd: add status command to list the watchers of an image
Signed-off-by: Yunchuan Wen <yunchuanwen@ubuntukylin.com>
Reviewed-by: Li Wang <liwang@ubuntukylin.com>
Josh Durgin [Sun, 1 Feb 2015 20:17:34 +0000 (21:17 +0100)]
Merge branch 'librbd-cache-iohint' of git://github.com/majianpeng/ceph
Reviewed-by: Sage Weil <sage@redhat.com>
Reviewed-by: Josh Durgin <jdurgin@redhat.com>
Conflicts:
src/librbd/ImageCtx.cc
src/librbd/ImageCtx.h
src/osdc/ObjectCacher.cc
src/osdc/ObjectCacher.h
Josh Durgin [Sun, 1 Feb 2015 12:54:52 +0000 (13:54 +0100)]
Merge pull request #3556 from trociny/image_watcher-fixup
librbd: RemoteAsyncRequest: fix comparison operator
Reviewed-by: Josh Durgin <jdurgin@redhat.com>
Loic Dachary [Sun, 1 Feb 2015 12:16:34 +0000 (13:16 +0100)]
Merge pull request #3537 from ceph/wip-10688-boost-157
#10688: support Boost 1.57.0
Reviewed-by: Loic Dachary <ldachary@redhat.com>
Mykola Golub [Sun, 1 Feb 2015 11:58:25 +0000 (13:58 +0200)]
librbd: RemoteAsyncRequest: fix comparison operator
Signed-off-by: Mykola Golub <mgolub@mirantis.com>
Jason Dillaman [Wed, 21 Jan 2015 20:23:31 +0000 (15:23 -0500)]
rbd: use IO hints for read/write operations
Import, export, and benchmark operations now use the librados
IO hints to indicate the read/write pattern and whether caching
is required.
Fixes: #10462
Signed-off-by: Jason Dillaman <dillaman@redhat.com>
Reviewed-by: Josh Durgin <jdurgin@redhat.com>
Josh Durgin [Sun, 1 Feb 2015 06:36:05 +0000 (07:36 +0100)]
Merge remote-tracking branch 'origin/wip-4087'
Reviewed-by: Josh Durgin <jdurgin@redhat.com>
Conflicts:
src/librbd/LibrbdWriteback.cc
Sage Weil [Sat, 31 Jan 2015 22:08:10 +0000 (14:08 -0800)]
Merge pull request #3545 from thesues/fix-cmp-xattr
Fix ReplicatedPG do_xattr_cmp_u64
Backport: giant, firefly
Reviewed-by: Sage Weil <sage@redhat.com>
Sage Weil [Sat, 31 Jan 2015 22:01:03 +0000 (14:01 -0800)]
Merge pull request #3555 from ceph/wip-10106
rgw: flush xml header on get acl request
Reviewed-by: Sage Weil <sage@redhat.com>
Sage Weil [Sat, 31 Jan 2015 21:16:15 +0000 (13:16 -0800)]
Merge pull request #3527 from wonzhq/op-stat
osd: don't log op stats for proxy read in the cache tier
Reviewed-by: Sage Weil <sage@redhat.com>
Sage Weil [Sat, 31 Jan 2015 21:15:47 +0000 (13:15 -0800)]
Merge pull request #3503 from ceph/wip-pg-reqids
osd: index PGLog with optional extra_reqids in each log entry
Reviewed-by: Samuel Just <sjust@redhat.com>
Reviewed-by: Wang, Zhiqiang <zhiqiang.wang@intel.com>
Josh Durgin [Sat, 31 Jan 2015 11:09:55 +0000 (12:09 +0100)]
Merge pull request #3554 from ceph/wip-10572
rgw: finalize perfcounters after shutting down storage
Reviewed-by: Josh Durgin <jdurgin@redhat.com>
Josh Durgin [Sat, 31 Jan 2015 11:06:20 +0000 (03:06 -0800)]
Merge pull request #3323 from ceph/wip-10445
rgw: use gc for multipart abort
Reviewed-by: Josh Durgin <jdurgin@redhat.com>
Sage Weil [Sat, 31 Jan 2015 10:15:53 +0000 (02:15 -0800)]
ceph_test_rados: fix typo
Signed-off-by: Sage Weil <sage@redhat.com>
Josh Durgin [Sat, 31 Jan 2015 09:03:48 +0000 (01:03 -0800)]
Merge pull request #3452 from majianpeng/iohint
Iohint in librbd cache.
Reviewed-by: Josh Durgin <jdurgin@redhat.com>
Yehuda Sadeh [Sat, 31 Jan 2015 02:42:40 +0000 (18:42 -0800)]
rgw: flush xml header on get acl request
Fixes: #10106
Backport: firefly, giant
dump_start() updates the formatter with the appropriate prefix, however,
we never flushed the formatter.
Signed-off-by: Yehuda Sadeh <yehuda@redhat.com>
Yehuda Sadeh [Fri, 30 Jan 2015 22:34:32 +0000 (14:34 -0800)]
rgw: finalize perfcounters after shutting down storage
Fixes: #10572
Backport: giant, firefly
First disable the storage subsystem, then disable perfcounters.
Signed-off-by: Yehuda Sadeh <yehuda@redhat.com>
Jianpeng Ma [Mon, 12 Jan 2015 06:35:28 +0000 (14:35 +0800)]
librbd: Don't do readahead for random read.
Signed-off-by: Jianpeng Ma <jianpeng.ma@intel.com>
Reviewed-by: Sage Weil <sage@redhat.com>
Sage Weil [Fri, 30 Jan 2015 21:22:48 +0000 (13:22 -0800)]
Merge pull request #3516 from ceph/wip-mdsmon-args
mon/MDSMonitor: fix gid/rank/state parsing
Reviewed-by: Huang Jun <hjwsm1989@gmail.com>
Reviewed-by: Dan Mick <dmick@redhat.com>
Yehuda Sadeh [Fri, 30 Jan 2015 21:14:38 +0000 (13:14 -0800)]
Merge pull request #3551 from ceph/wip-10698
rgw: fail s3 POST auth if keystone not configured
Reviewed-by: Greg Farnum <gfarnum@redhat.com>
Samuel Just [Fri, 30 Jan 2015 20:54:47 +0000 (12:54 -0800)]
ReplicatedPG::C_ProxyRead: fix dropped lock
Fixes: 10633
Signed-off-by: Samuel Just <sjust@redhat.com>
Samuel Just [Mon, 1 Dec 2014 22:52:11 +0000 (14:52 -0800)]
ReplicatedPG: only populate_obc_watchers if active
Signed-off-by: Samuel Just <sjust@redhat.com>
Samuel Just [Wed, 26 Nov 2014 22:00:34 +0000 (14:00 -0800)]
OSD: allow recovery in peered
Signed-off-by: Samuel Just <sjust@redhat.com>
Samuel Just [Fri, 7 Nov 2014 23:20:02 +0000 (15:20 -0800)]
osd/: go "peered" instead of "active" when < min_size
In the case of a replicated pool, the pg will transition to "peered"
rather than "active", where it can perform backfill bringing itself up
to min_size peers. To that end, most operations now require
is_peered rather than is_active (OSDOps being the primary exception).
Also, rather than using the query_epoch on the activation message as the
activation epoch (for last_epoch_started) on the replica, we instead
use the last_epoch_started in the info sent by the primary. This
allows the primary to not advance last_epoch_started past the last
known actual activation. This will prevent later peering epochs from
requiring the last_update from a peered epoch to go active (might be
divergent).
Fixes: #7862
Signed-off-by: Samuel Just <sam.just@inktank.com>
Samuel Just [Thu, 20 Nov 2014 23:50:50 +0000 (15:50 -0800)]
PG: move most of all_activated_and_commited into the AllReplicasActivated handler
It's confusing to have that logic split up so much.
Signed-off-by: Samuel Just <sjust@redhat.com>
Samuel Just [Tue, 18 Nov 2014 00:22:59 +0000 (16:22 -0800)]
PG::proc_master_log: assume les and history.les from log source
Otherwise, we might later go peered (not active) and not distribute
a non-0 last_epoch_started. This should be safe as the log recipient
will have a last_update reflecting that interval as well.
Signed-off-by: Samuel Just <sjust@redhat.com>
Samuel Just [Thu, 30 Oct 2014 20:21:01 +0000 (13:21 -0700)]
osd_types: add last_peered, last_became_peered
Signed-off-by: Samuel Just <sam.just@inktank.com>
Samuel Just [Wed, 29 Oct 2014 21:10:13 +0000 (14:10 -0700)]
PG: break waiting_for_peered out of waiting_for_active
waiting_for_peered now holds ops until peering completes (activation,
not necessarily state active). waiting_for_active now holds
specifically MOSDOp blocked on:
- scrub
- replay
- state active
Signed-off-by: Samuel Just <sam.just@inktank.com>
Samuel Just [Wed, 29 Oct 2014 21:07:09 +0000 (14:07 -0700)]
PG: remove very obsolete comment about backfill_pos
Signed-off-by: Samuel Just <sam.just@inktank.com>
Samuel Just [Wed, 29 Oct 2014 20:57:01 +0000 (13:57 -0700)]
osd/: add state PG_STATE_PEERED
We need a state to represent the pg state where we are below
min_size, but trying to recover.
Signed-off-by: Samuel Just <sam.just@inktank.com>
Samuel Just [Wed, 29 Oct 2014 21:03:40 +0000 (14:03 -0700)]
doc: add documentation for undersized
Signed-off-by: Samuel Just <sam.just@inktank.com>
Gregory Farnum [Fri, 30 Jan 2015 19:37:42 +0000 (11:37 -0800)]
Merge pull request #3553 from ceph/wip-10701
rgw: use strict_strtoll() for content length
Reviewed-by: Greg Farnum <gfarnum@redhat.com>
Samuel Just [Tue, 27 Jan 2015 22:08:43 +0000 (14:08 -0800)]
ReplicatedPG: block writes on degraded objects for ec pools
Signed-off-by: Samuel Just <sjust@redhat.com>
Samuel Just [Tue, 27 Jan 2015 02:31:26 +0000 (18:31 -0800)]
append_log: use remove_snap_mapped_object
Backport: giant, firefly
Signed-off-by: Samuel Just <sjust@redhat.com>
Samuel Just [Fri, 23 Jan 2015 18:01:39 +0000 (10:01 -0800)]
ReplicatedPG::agent_maybe_evict: take write lock, skip on failure
Signed-off-by: Samuel Just <sjust@redhat.com>
Samuel Just [Wed, 14 Jan 2015 22:27:02 +0000 (14:27 -0800)]
ReplicatedPG: if we got ENOENT and an obc, we need to check the locks
Otherwise, we might return ENOENT out of order.
Signed-off-by: Samuel Just <sjust@redhat.com>
Samuel Just [Tue, 13 Jan 2015 22:34:23 +0000 (14:34 -0800)]
ReplicatedPG::get_rw_locks: use excl lock for read & write_ordered
Signed-off-by: Samuel Just <sjust@redhat.com>