]>
git.apps.os.sepia.ceph.com Git - ceph.git/log
David Zafman [Tue, 22 Sep 2015 16:32:18 +0000 (09:32 -0700)]
Merge pull request #5922 from trociny/wip-pool-set-noscrub
osd: per pool control for scrub and deep-scrub
Reviewed-by: David Zafman <dzafman@redhat.com>
Kefu Chai [Tue, 22 Sep 2015 15:54:32 +0000 (23:54 +0800)]
Merge pull request #5846 from xiexingguo/xxg-wip-13002
msg/simple: start over after fails to bind a port in specified range
Reviewed-by: Kefu Chai <kchai@redhat.com>
Samuel Just [Tue, 22 Sep 2015 14:24:27 +0000 (07:24 -0700)]
Merge pull request #6008 from liewegas/wip-13190
osd/ReplicatedPG: tolerate promotion completion with stopped agent
Reviewed-by: Samuel Just <sjust@redhat.com>
Sage Weil [Tue, 22 Sep 2015 13:32:08 +0000 (09:32 -0400)]
Merge pull request #6011 from liewegas/wip-systemd
systemd: increase nproc ulimit
Sage Weil [Tue, 22 Sep 2015 12:29:16 +0000 (08:29 -0400)]
Merge pull request #6019 from liewegas/wip-mon-dispatch-debug
mon/PaxosService: include msg and con ptr in dispatch debug
Josh Durgin [Tue, 22 Sep 2015 07:07:43 +0000 (03:07 -0400)]
Merge pull request #5956 from xinxinsh/wip-13124
unify order limit
Reviewed-by: Josh Durgin <jdurgin@redhat.com>
Josh Durgin [Tue, 22 Sep 2015 05:19:58 +0000 (01:19 -0400)]
Merge pull request #5976 from yuyuyu101/wip-fix-metadata-conf
Librbd: Fix incorrect metadata filter behavior
Reviewed-by: Josh Durgin <jdurgin@redhat.com>
Josh Durgin [Mon, 21 Sep 2015 23:13:37 +0000 (19:13 -0400)]
Merge pull request #5750 from wonzhq/rbd-write-full
librbd: do write_full for whole object write
Reviewed-by: Jason Dillaman <dillaman@redhat.com>
Reviewed-by: Josh Durgin <jdurgin@redhat.com>
Gregory Farnum [Mon, 21 Sep 2015 23:02:33 +0000 (16:02 -0700)]
Merge pull request #5416 from zhouyuan/ceph_fuse_12204
ceph-fuse: warn and shut down when there is no MDS present
Reviewed-by: Yan, Zheng <zyan@redhat.com>
Reviewed-by: Greg Farnum <gfarnum@redhat.com>
Gregory Farnum [Mon, 21 Sep 2015 22:57:13 +0000 (15:57 -0700)]
Merge pull request #5826 from ukernel/wip-osx-port
fix compilation on OSX/FreeBSD
Reviewed-by: Greg Farnum <gfarnum@redhat.com>
Greg Farnum [Mon, 21 Sep 2015 22:55:04 +0000 (15:55 -0700)]
Merge branch 'wip-12971' of git://github.com/jcsp/ceph
Reviewed-by: Greg Farnum <gfarnum@redhat.com>
Conflicts:
src/osdc/Objecter.cc
Signed-off-by: Greg Farnum <gfarnum@redhat.com>
Gregory Farnum [Mon, 21 Sep 2015 22:52:10 +0000 (15:52 -0700)]
Merge pull request #5644 from majianpeng/mds-support-iohint
Mds journal support iohint
Reviewed-by: John Spray <john.spray@redhat.com>
Gregory Farnum [Mon, 21 Sep 2015 22:50:28 +0000 (15:50 -0700)]
Merge pull request #5966 from ukernel/wip-12506
Wip 12506
Reviewed-by: Greg Farnum <gfarnum@redhat.com>
Gregory Farnum [Mon, 21 Sep 2015 22:49:34 +0000 (15:49 -0700)]
Merge pull request #5962 from majianpeng/client-fix
Client fix mini bugs.
Reviewed-by: Greg Farnum <gfarnum@redhat.com>
Sage Weil [Mon, 21 Sep 2015 21:04:50 +0000 (17:04 -0400)]
mon/PaxosService: include msg and con ptr in dispatch debug
Signed-off-by: Sage Weil <sage@redhat.com>
Sage Weil [Mon, 21 Sep 2015 18:51:50 +0000 (14:51 -0400)]
Merge pull request #6014 from athanatos/wip-13108
Revert "osd: purge the object from the cache when proxying and not pr…
Reviewed-by: Sage Weil <sage@redhat.com>
Sage Weil [Mon, 21 Sep 2015 18:51:07 +0000 (14:51 -0400)]
Merge pull request #6015 from athanatos/wip-13143
osd/: find_object_context: return obc for head for snapdir if non null
Reviewed-by: Sage Weil <sage@redhat.com>
Sage Weil [Mon, 21 Sep 2015 18:46:33 +0000 (14:46 -0400)]
Merge pull request #6017 from athanatos/wip-13150
ReplicatedPG::get_snapset_context: set exists to false for new ssc wi…
Reviewed-by: Sage Weil <sage@redhat.com>
Sage Weil [Mon, 21 Sep 2015 18:43:55 +0000 (14:43 -0400)]
Merge pull request #6013 from athanatos/wip-12828
Wip 12828
Reviewed-by: Sage Weil <sage@redhat.com>
Sage Weil [Thu, 17 Sep 2015 22:28:38 +0000 (18:28 -0400)]
systemd: increase nproc ulimit
We were observed to be hitting the limit on centos7
(triggering pthread_create failures) on a ~2000 OSD cluster.
Increasing this resolves it!
Reported-by: Dan van der Ster <daniel.vanderster@cern.ch>
Signed-off-by: Sage Weil <sage@redhat.com>
Sage Weil [Mon, 21 Sep 2015 15:41:03 +0000 (11:41 -0400)]
osd/ReplicatedPG: tolerate promotion completion with stopped agent
We may start a promotion, then get a pool update that disables the
agent, and then complete it. We should not segfault in this case.
Fixes: #13190
Signed-off-by: Sage Weil <sage@redhat.com>
Yehuda Sadeh [Mon, 21 Sep 2015 15:35:30 +0000 (08:35 -0700)]
Merge pull request #5918 from oritwas/wip-5378
rgw: make radosgw-admin user rm idempotent
Reviewed-by: Yehuda Sadeh <yehuda@redhat.com>
Yehuda Sadeh [Mon, 21 Sep 2015 15:22:29 +0000 (08:22 -0700)]
Merge pull request #5996 from rzarzynski/wip-12780
rgw: don't read actual data on user manifest HEAD
Yehuda Sadeh [Mon, 21 Sep 2015 15:20:55 +0000 (08:20 -0700)]
Merge pull request #5974 from Sandy4999/wip-rgw-errmsg
rgw: add a new error message for user conflict when using Admin Ops API
Reviewed-by: Yehuda Sadeh <yehuda@redhat.com>
John Spray [Mon, 21 Sep 2015 09:25:39 +0000 (10:25 +0100)]
Merge pull request #6003 from ukernel/wip-13166
Wip 13166
Reviewed-by: John Spray <john.spray@redhat.com>
Yan, Zheng [Mon, 21 Sep 2015 07:20:15 +0000 (15:20 +0800)]
mds: adjust MDSRank::incarnation according to mdsmap
When a standby-replay MDS replace failed MDS, we need update its
incarnation.
Fixes: #13166
Signed-off-by: Yan, Zheng <zyan@redhat.com>
Mykola Golub [Mon, 14 Sep 2015 08:03:25 +0000 (11:03 +0300)]
osd: new pool settings: scrub intervals
ceph osd pool set $POOL scrub_min_interval N
ceph osd pool set $POOL scrub_max_interval N
ceph osd pool set $POOL deep_scrub_interval N
If N > 0, this value is used for the pool instead of
the corresponding global parameter from the config
(osd_scrub_min_interval, osd_scrub_max_interval or
osd_deep_scrub_interval).
Fixes: #13077
Signed-off-by: Mykola Golub <mgolub@mirantis.com>
Mykola Golub [Fri, 11 Sep 2015 08:05:42 +0000 (11:05 +0300)]
osd: new pool flags: noscrub, nodeep-scrub
This allows to disable periodic scrub on per pool basis.
Fixes: #13076
Signed-off-by: Mykola Golub <mgolub@mirantis.com>
Mykola Golub [Fri, 11 Sep 2015 18:34:52 +0000 (21:34 +0300)]
osd: make 'ceph osd pool get' work for all settable pool flags
Signed-off-by: Mykola Golub <mgolub@mirantis.com>
Mykola Golub [Fri, 11 Sep 2015 18:02:41 +0000 (21:02 +0300)]
osd: refactor setting write_fadvise_dontneed pool flag
Handle together with other flags.
Signed-off-by: Mykola Golub <mgolub@mirantis.com>
Sage Weil [Mon, 21 Sep 2015 00:57:13 +0000 (20:57 -0400)]
Merge pull request #6001 from dachary/wip-13182-crushtool
tests: update to match crushmap validation message
Loic Dachary [Sun, 20 Sep 2015 21:42:45 +0000 (23:42 +0200)]
tests: update to match crushmap validation message
http://tracker.ceph.com/issues/13182 Fixes: #13182
Signed-off-by: Loic Dachary <ldachary@redhat.com>
Sage Weil [Sun, 20 Sep 2015 17:54:19 +0000 (13:54 -0400)]
Merge remote-tracking branch 'gh/infernalis'
Sage Weil [Sun, 20 Sep 2015 17:50:26 +0000 (13:50 -0400)]
Merge pull request #5979 from ceph/wip-crushtest
mon: fix crush testing for new pools
Reviewed-by: Loic Dachary <ldachary@redhat.com>
Sage Weil [Sun, 20 Sep 2015 17:49:36 +0000 (13:49 -0400)]
Merge pull request #5950 from tchaikov/wip-13112-infernalis
mon/PGMap: calc min_last_epoch_clean when decode
Reviewed-by: Sage Weil <sage@redhat.com>
Loic Dachary [Sun, 20 Sep 2015 15:12:32 +0000 (17:12 +0200)]
Merge pull request #5998 from dachary/wip-13136-python3-coverage
install-deps: enable python3
Loic Dachary [Fri, 18 Sep 2015 10:59:09 +0000 (12:59 +0200)]
install-deps: enable python3
The upstream regression has been fixed, we can re-enable python3
support.
http://tracker.ceph.com/issues/13136 Fixes: #13136
Signed-off-by: Loic Dachary <ldachary@redhat.com>
Sage Weil [Sat, 19 Sep 2015 21:21:10 +0000 (17:21 -0400)]
Merge pull request #5994 from SUSE/wip-13175
doc: do not promise backports to Dumpling
Reviewed-by: Sage Weil <sage@redhat.com>
Nathan Cutler [Sat, 19 Sep 2015 20:25:31 +0000 (22:25 +0200)]
doc: do not promise backports to Dumpling
Dumpling was retired in May 2015.
http://tracker.ceph.com/issues/13175 Fixes: #13175
Signed-off-by: Nathan Cutler <ncutler@suse.cz>
John Spray [Sat, 19 Sep 2015 11:03:57 +0000 (12:03 +0100)]
Merge pull request #5965 from jcsp/wip-get-or-create-output
mon: fix auth get-or-create output
Reviewed-by: Greg Farnum <gfarnum@redhat.com>
Ken Dreyer [Fri, 18 Sep 2015 18:41:42 +0000 (12:41 -0600)]
Merge pull request #5984 from ceph/wip-packages
Fixes locations from ceph.com to download.ceph.com
(For the record, this is to fix http://tracker.ceph.com/issues/13162)
Reviewed-by: Ken Dreyer <kdreyer@redhat.com>
Alfredo Deza [Fri, 18 Sep 2015 18:20:12 +0000 (14:20 -0400)]
doc: remove mention of ceph-extra as a requirement
Signed-off-by: Alfredo Deza <adeza@redhat.com>
Alfredo Deza [Fri, 18 Sep 2015 18:13:21 +0000 (14:13 -0400)]
doc: remove ceph-extras
Signed-off-by: Alfredo Deza <adeza@redhat.com>
Alfredo Deza [Fri, 18 Sep 2015 18:13:02 +0000 (14:13 -0400)]
doc: correct links to download.ceph.com
Signed-off-by: Alfredo Deza <adeza@redhat.com>
John Spray [Thu, 17 Sep 2015 14:51:20 +0000 (15:51 +0100)]
mon: fix auth get-or-create output
Previously the caps were omitted from the
output: they should be present for this
to fulfil the 'get' part of get-or-create.
Signed-off-by: John Spray <john.spray@redhat.com>
Samuel Just [Thu, 17 Sep 2015 23:49:53 +0000 (16:49 -0700)]
ReplicatedPG::get_snapset_context: set exists to false for new ssc with can_create
Otherwise, if we do a proxy_write and don't promote it, the ssc will
stick around with exists = true and cause future reads on snaps for the
object to fail to fall through to the base tier.
Fixes: #13150
Signed-off-by: Samuel Just <sjust@redhat.com>
Yehuda Sadeh [Thu, 17 Sep 2015 16:10:39 +0000 (09:10 -0700)]
rgw: don't read actual data on user manifest HEAD
Fixes: #12780
We unconditionally read all the data, which is not needed
when doing HEAD operation on user manifest objects.
Signed-off-by: Yehuda Sadeh <yehuda@redhat.com>
Signed-off-by: Radoslaw Zarzynski <rzarzynski@mirantis.com>
Sage Weil [Fri, 18 Sep 2015 15:40:26 +0000 (11:40 -0400)]
ceph-osd-prestart.sh: no ceph-disk chown
chown -R ceph:ceph $1 is easy enough.
Signed-off-by: Sage Weil <sage@redhat.com>
Sage Weil [Fri, 18 Sep 2015 15:06:24 +0000 (11:06 -0400)]
Merge pull request #5980 from liewegas/wip-13059
ceph.spec: fix up /var/run/ceph
Reviewed-by: Sage Weil <sage@redhat.com>
Boris Ranto [Fri, 18 Sep 2015 15:00:30 +0000 (17:00 +0200)]
ceph.spec.in: Fix up (/var)/run/ceph creation
Fixes: #13059
Signed-off-by: Boris Ranto <branto@redhat.com>
Kefu Chai [Fri, 18 Sep 2015 14:32:41 +0000 (22:32 +0800)]
Merge pull request #5977 from tchaikov/wip-size_t-monstore-txn
mon: return size_t from MonitorDBStore::Transaction::size()
Reviewed-by: Joao Eduardo Luis <joao@suse.de>
Sage Weil [Fri, 18 Sep 2015 13:42:47 +0000 (09:42 -0400)]
mon/OSDMonitor: fix crush injection error message
Signed-off-by: Sage Weil <sage@redhat.com>
Sage Weil [Fri, 18 Sep 2015 13:41:25 +0000 (09:41 -0400)]
mon/OSDMonitor: only test crush ruleset for the newly created pool
Otherwise, we test *all* crush rules.. which might be a lot, and
which is a big waste of time and effort.
Signed-off-by: Sage Weil <sage@redhat.com>
Sage Weil [Fri, 18 Sep 2015 13:40:13 +0000 (09:40 -0400)]
crush/CrushTester: allow testing by ruleset
Signed-off-by: Sage Weil <sage@redhat.com>
Haomai Wang [Fri, 18 Sep 2015 04:17:53 +0000 (12:17 +0800)]
Librbd: Fix incorrect metadata filter behavior
Signed-off-by: Haomai Wang <haomai@xsky.com>
Kefu Chai [Fri, 18 Sep 2015 09:05:09 +0000 (17:05 +0800)]
mon: return size_t from MonitorDBStore::Transaction::size()
Signed-off-by: Kefu Chai <kchai@redhat.com>
Sangdi [Fri, 18 Sep 2015 03:02:07 +0000 (11:02 +0800)]
rgw: add a new error message for user conflict when using Admin Ops API
Add a new error message to indicate user conflict, which was using the error code for bucket conflict.
Signed-off-by: Sangdi Xu <xu.sangdi@h3c.com>
Sage Weil [Fri, 18 Sep 2015 02:14:56 +0000 (22:14 -0400)]
Merge pull request #5948 from flyd1005/master
doc: delete wrong description of installing RPMs
Samuel Just [Thu, 17 Sep 2015 17:54:18 +0000 (10:54 -0700)]
osd/: find_object_context: return obc for head for snapdir if non null
The head obc may have an in progress operation holding a lock, we
need to block on the lock for ordering reasons.
Fixes: #13143
Signed-off-by: Samuel Just <sjust@redhat.com>
Sage Weil [Thu, 17 Sep 2015 19:09:17 +0000 (15:09 -0400)]
Merge pull request #5926 from dillaman/wip-13082
include/inline_memory: out-of-bounds read on unaligned memory
Reviewed-by: Sage Weil <sage@redhat.com>
Yan, Zheng [Thu, 17 Sep 2015 15:56:14 +0000 (23:56 +0800)]
mds: Make sure wanted_state of first MDSBeacon is MDSMap::STATE_BOOT
Fixes: #12506
Signed-off-by: Yan, Zheng <zyan@redhat.com>
Sage Weil [Thu, 17 Sep 2015 14:47:26 +0000 (10:47 -0400)]
new release key
The previous release key,
pub 4096R/
460F3994 2015-09-15
uid Ceph.com (release key) <security@ceph.com>
may have been compromised.
The new release key is
pub 4096R/
460F3994 2015-09-15
uid Ceph.com (release key) <security@ceph.com>
Signed-off-by: Sage Weil <sage@redhat.com>
Kefu Chai [Thu, 17 Sep 2015 15:44:57 +0000 (23:44 +0800)]
Merge pull request #5947 from weiqian/master
doc: replaces 'osd host' with 'host' in ceph.conf docs.
Reviewed-by: Kefu Chai <kchai@redhat.com>
Yan, Zheng [Thu, 17 Sep 2015 12:55:34 +0000 (20:55 +0800)]
Merge pull request #5964 from jcsp/wip-13138
libcephfs: fix calling init() then mount()
John Spray [Thu, 17 Sep 2015 12:46:30 +0000 (13:46 +0100)]
libcephfs: fix calling init() then mount()
Previously only ever called these separately, but
it should be allowed for callers to use one after
the other.
Fixes: #13138
Signed-off-by: John Spray <john.spray@redhat.com>
Orit Wasserman [Thu, 17 Sep 2015 10:59:43 +0000 (12:59 +0200)]
Merge pull request #5963 from jcsp/wip-cmake-cephfs
CMake: fix libcephfs shared lib generation
Kefu Chai [Thu, 17 Sep 2015 10:54:24 +0000 (18:54 +0800)]
Merge pull request #5857 from ceph/wip-da-SCA-
20150812
SCA and compiler warning fixes
Reviewed-by: Kefu Chai <kchai@redhat.com>
Kefu Chai [Thu, 17 Sep 2015 10:52:54 +0000 (18:52 +0800)]
Merge pull request #5870 from s09816/master
osd:the fuction osd::shutdown Lock failed.
Reviewed-by: Kefu Chai <kchai@redhat.com>
Kefu Chai [Thu, 17 Sep 2015 10:51:36 +0000 (18:51 +0800)]
Merge pull request #5096 from tchaikov/wip-add-bl-const-iterator
common/buffer: add bufferlist::const_iterator
Reviewed-by: Sage Weil <sage@redhat.com>
John Spray [Thu, 17 Sep 2015 10:26:32 +0000 (11:26 +0100)]
CMake: fix libcephfs shared lib generation
Previously weren't generating versioned symlinks
etc, so python bindings didn't find it.
Signed-off-by: John Spray <john.spray@redhat.com>
Kefu Chai [Thu, 17 Sep 2015 09:09:18 +0000 (17:09 +0800)]
Merge pull request #5920 from caibo2014/fix-tool-bug
tools: ceph_monstore_tool: remove the output file when get map fails
Reviewed-by: Kefu Chai <kchai@redhat.com>
Kefu Chai [Thu, 17 Sep 2015 08:59:00 +0000 (16:59 +0800)]
Merge pull request #5960 from dachary/wip-13136-python3-coverage
install-deps.sh: disable python3
Reviewed-by: Kefu Chai <kchai@redhat.com>
Loic Dachary [Thu, 17 Sep 2015 07:35:11 +0000 (09:35 +0200)]
install-deps.sh: disable python3
Disable python3 support until
https://bugs.launchpad.net/ubuntu/+source/python-coverage/+bug/
1496715
is fixed.
Nothing in the build process depends on python3 right now, there is no
harm disabling it.
http://tracker.ceph.com/issues/13136 Fixes: #13136
Signed-off-by: Loic Dachary <ldachary@redhat.com>
Jianpeng Ma [Thu, 17 Sep 2015 03:14:31 +0000 (11:14 +0800)]
client/MetaRequest: optimize func can_forward/auth_is_best.
Signed-off-by: Jianpeng Ma <jianpeng.ma@intel.com>
xinxin shu [Thu, 17 Sep 2015 03:07:47 +0000 (11:07 +0800)]
unify order limit
Fixes: #13124
Signed-off-by: xinxin shu <xinxin.shu@intel.com>
Jianpeng Ma [Thu, 17 Sep 2015 02:05:46 +0000 (10:05 +0800)]
client/MetaRequest: open w/ O_CREAT|O_TRUNC is write.
From the function name is_write, it judge MetaRequest whether write.
For open, if it w/ O_CREAT|O_TRUNC, it's a write op and vice versa.
Signed-off-by: Jianpeng Ma <jianpeng.ma@intel.com>
Jianpeng Ma [Thu, 17 Sep 2015 01:57:04 +0000 (09:57 +0800)]
clinet/MetaRequest: Remove the useless condition.
"head.op & CEPH_MDS_OP_WRITE" include "head.op == CEPH_MDS_OP_CREATE".
So remove the conditon statement "head.op == CEPH_MDS_OP_CREATE".
Signed-off-by: Jianpeng Ma <jianpeng.ma@intel.com>
Kefu Chai [Wed, 16 Sep 2015 07:08:17 +0000 (15:08 +0800)]
mon/PGMap: calc min_last_epoch_clean when decode
Fixes: #13112
Signed-off-by: Kefu Chai <kchai@redhat.com>
Sage Weil [Wed, 16 Sep 2015 13:12:16 +0000 (09:12 -0400)]
Merge pull request #5841 from ceph/wip-12994-rbd-replay-packaging
#12994: packaging: move rbd-replay* to ceph-common
Reviewed-by: James Page <jamespage@ubuntu.com>
Reviewed-by: Nathan Cutler <ncutler@suse.com>
Sage Weil [Wed, 16 Sep 2015 13:11:33 +0000 (09:11 -0400)]
Merge pull request #5946 from liewegas/wip-rgw-logdir
radosgw: log to /var/log/ceph instead of /var/log/radosgw
Reviewed-by: Yehuda Sadeh <yehuda@redhat.com>
Li Peng [Wed, 16 Sep 2015 06:05:05 +0000 (14:05 +0800)]
doc: delete wrong description of installing RPMs
In CentOS/RHEL 6/7, gitk and git-gui are available in default repository actually.
Signed-off-by: Li Peng <lip@dtdream.com>
weiqian [Wed, 16 Sep 2015 03:04:52 +0000 (11:04 +0800)]
doc:Replaces 'osd host' with 'host'
Signed-off-by: weiqian <weiq@dtdream.com>
Sage Weil [Tue, 15 Sep 2015 22:05:59 +0000 (18:05 -0400)]
radosgw: log to /var/log/ceph instead of /var/log/radosgw
This is simpler.
Signed-off-by: Sage Weil <sage@redhat.com>
Sage Weil [Tue, 15 Sep 2015 21:42:42 +0000 (17:42 -0400)]
Merge remote-tracking branch 'gh/infernalis'
Sage Weil [Tue, 15 Sep 2015 21:36:47 +0000 (17:36 -0400)]
Merge pull request #5725 from liewegas/wip-scrub-stats
osd/ReplicatedPG: snaptimmer: adjust stats through ctx->delta_stats
Reviewed-by: Samuel Just <sjust@redhat.com>
Sage Weil [Tue, 15 Sep 2015 21:35:02 +0000 (17:35 -0400)]
Merge pull request #5925 from liewegas/wip-mon-leaks
fix various mon leaks
Reviewed-by: Joao Eduardo Luis <joao@suse.de>
Sage Weil [Tue, 15 Sep 2015 21:34:13 +0000 (17:34 -0400)]
Merge pull request #5929 from liewegas/wip-osd-leaks
misc osd leaks
Reviewed-by: David Zafman <dzafman@redhat.com>
Reviewed-by: Kefu Chai <kchai@redhat.com>
John Spray [Tue, 15 Sep 2015 21:11:20 +0000 (22:11 +0100)]
osd: permit MDS to send writes on full pools
Same behaviour as the global full flag, for the
per-pool full flag. The FS was broken by
67de12bf ,
which added the pool flag handling.
Fixes: #12971
Signed-off-by: John Spray <john.spray@redhat.com>
Samuel Just [Tue, 15 Sep 2015 20:58:32 +0000 (13:58 -0700)]
Revert "osd: purge the object from the cache when proxying and not promoting the op"
Fixes: #13108
This reverts commit
da68bb371f8b06b85a73077cd069a294193ac6c0 .
John Spray [Tue, 15 Sep 2015 20:35:48 +0000 (21:35 +0100)]
Merge pull request #5944 from tchaikov/wip-13090
mds: cast numbers for mds health to string when print
Reviewed-by: John Spray <john.spray@redhat.com>
John Spray [Tue, 15 Sep 2015 20:24:18 +0000 (21:24 +0100)]
osdc/Objecter: make unlocked 'full' helpers private
These never needed to be public, let's be safe.
Signed-off-by: John Spray <john.spray@redhat.com>
John Spray [Tue, 15 Sep 2015 13:05:13 +0000 (14:05 +0100)]
osdc: fix honor_osdmap_full
This was being ignored in new code that tested
the per-pool full flags. Refactor it so that
all the tests of full flags go through _osdmap_pool_full
and thereby experience the honor_osdmap_full policy.
Fixes: #12971
Signed-off-by: John Spray <john.spray@redhat.com>
Ken Dreyer [Tue, 8 Sep 2015 16:41:02 +0000 (10:41 -0600)]
packaging: move rbd-replay* to ceph-common
The rbd-replay* utilities are useful for Ceph users with RBD clients.
Currently the rbd-replay* utilities ship in the "ceph-test" package, and
we intend this ceph-test package for Ceph developers and contributors,
not normal users.
Move the rbd-replay* utilities to "ceph-common".
http://tracker.ceph.com/issues/12994 Fixes: #12994
Signed-off-by: Ken Dreyer <kdreyer@redhat.com>
Samuel Just [Fri, 28 Aug 2015 20:13:08 +0000 (13:13 -0700)]
src/test/osd/Object: allow a hole at the end of the object
This way, we test the case with holes at the end of the file more
often.
Signed-off-by: Samuel Just <sjust@redhat.com>
Samuel Just [Fri, 28 Aug 2015 19:49:00 +0000 (12:49 -0700)]
osd_types.cc: add size to ObjectRecoveryInfo print
Signed-off-by: Samuel Just <sjust@redhat.com>
Samuel Just [Fri, 28 Aug 2015 19:46:57 +0000 (12:46 -0700)]
ReplicatedBackend::prepare_pull: set recover_info.size for clones
Fixes: #12828
Signed-off-by: Samuel Just <sjust@redhat.com>
Sage Weil [Tue, 15 Sep 2015 16:49:36 +0000 (12:49 -0400)]
mon: debug refs on output replies
Signed-off-by: Sage Weil <sage@redhat.com>
Sage Weil [Tue, 15 Sep 2015 16:46:25 +0000 (12:46 -0400)]
msg/simple: discard local queued messages on shutdown
Otherwise these generate leak detection noise.
Signed-off-by: Sage Weil <sage@redhat.com>
Sage Weil [Tue, 15 Sep 2015 16:35:46 +0000 (12:35 -0400)]
mon/Monitor: fix MonSession Leak
dispatch() does not need to keep a ref; we hold one indirectly
via the MonOpRequest.
Signed-off-by: Sage Weil <sage@redhat.com>
Sage Weil [Tue, 15 Sep 2015 16:32:02 +0000 (12:32 -0400)]
mon: fix MonSession operator<<
Print the object, not its pointer. Make output unambiguously
terminated.
Signed-off-by: Sage Weil <sage@redhat.com>