]> git.apps.os.sepia.ceph.com Git - ceph.git/log
ceph.git
9 years agocmake: put virtualenv directories under TMPDIR 8822/head
Casey Bodley [Thu, 28 Apr 2016 20:58:15 +0000 (16:58 -0400)]
cmake: put virtualenv directories under TMPDIR

this was required for cmake-check to find ceph-disk and ceph-detect-init
in the PATH

Signed-off-by: Casey Bodley <cbodley@redhat.com>
9 years agoMerge pull request #8442 from trociny/wip-rbd-mirror-status
Jason Dillaman [Fri, 29 Apr 2016 06:41:46 +0000 (02:41 -0400)]
Merge pull request #8442 from trociny/wip-rbd-mirror-status

rbd-mirror: replay status now available via the rbd api and cli

Reviewed-by: Jason Dillaman <dillaman@redhat.com>
9 years agoObjectCacher: don't loop the extents of OSDWrite in map_writeReviewed-by
Jason Dillaman [Fri, 29 Apr 2016 14:59:18 +0000 (09:59 -0500)]
ObjectCacher: don't loop the extents of OSDWrite in map_writeReviewed-by

ObjectCacher: don't loop the extents of OSDWrite in map_write

Reviewed-by: Jason Dillaman <dillaman@redhat.com>
9 years agoMerge pull request #7830 from yuyuyu101/wip-14907
Jason Dillaman [Fri, 29 Apr 2016 13:39:34 +0000 (08:39 -0500)]
Merge pull request #7830 from yuyuyu101/wip-14907

IoCtxImpl: delete C_WatchCB2 when free LingerOp

Reviewed-by: Jason Dillaman <dillaman@redhat.com>
9 years agoMerge pull request #8635 from runsisi/wip-fix-md-list
Jason Dillaman [Fri, 29 Apr 2016 13:38:46 +0000 (08:38 -0500)]
Merge pull request #8635 from runsisi/wip-fix-md-list

cls_rbd: fix loop condition when listing image metadata

Reviewed-by: Jason Dillaman <dillaman@redhat.com>
9 years agoMerge pull request #8661 from runsisi/wip-fix-snapc-validation
Jason Dillaman [Fri, 29 Apr 2016 13:38:34 +0000 (08:38 -0500)]
Merge pull request #8661 from runsisi/wip-fix-snapc-validation

librbd: put the validation of image snap context earlier

Reviewed-by: Jason Dillaman <dillaman@redhat.com>
9 years agoMerge pull request #7800 from renhwsky/renhw-wip-osdmonitor-have_pending_crush
Sage Weil [Fri, 29 Apr 2016 12:43:05 +0000 (08:43 -0400)]
Merge pull request #7800 from renhwsky/renhw-wip-osdmonitor-have_pending_crush

mon/OSDMonitor: _have_pending_crush of return value should be bool

Reviewed-by: xie xingguo <xie.xingguo@zte.com.cn>
9 years agoqa/workunits/rbd: test mirror status in pool directory 8442/head
Mykola Golub [Sun, 24 Apr 2016 16:39:40 +0000 (19:39 +0300)]
qa/workunits/rbd: test mirror status in pool directory

Signed-off-by: Mykola Golub <mgolub@mirantis.com>
9 years agorbd-mirror: in replay status store number of entries behind master
Mykola Golub [Thu, 21 Apr 2016 06:52:47 +0000 (09:52 +0300)]
rbd-mirror: in replay status store number of entries behind master

Signed-off-by: Mykola Golub <mgolub@mirantis.com>
9 years agoMerge pull request #8305 from tchaikov/wip-fix-vstart
Jason Dillaman [Fri, 29 Apr 2016 04:01:07 +0000 (23:01 -0500)]
Merge pull request #8305 from tchaikov/wip-fix-vstart

vstart: make -k with optional mon_num (part 2)

Reviewed-by: Jason Dillaman <dillaman@redhat.com>
9 years agoosd/osdmonitor: _have_pending_crush of return value should be bool 7800/head
huanwen ren [Fri, 29 Apr 2016 02:53:14 +0000 (10:53 +0800)]
osd/osdmonitor: _have_pending_crush of return value should be bool

have_pending_crush of return value should be bool, fix it

Signed-off-by: huanwen ren <ren.huanwen@zte.com.cn>
9 years agoMerge pull request #8823 from aderumier/deb-default-tcmalloc-cache-size
Sage Weil [Fri, 29 Apr 2016 01:47:07 +0000 (21:47 -0400)]
Merge pull request #8823 from aderumier/deb-default-tcmalloc-cache-size

etc/default/ceph: set 128MB tcmalloc cache size

Reviewed-by: Sage Weil <sage@redhat.com>
9 years agoset 128MB tcmalloc cache size by bytes 8823/head
Alexandre Derumier [Fri, 29 Apr 2016 01:30:05 +0000 (03:30 +0200)]
set 128MB tcmalloc cache size by bytes

9 years agoetc/default/ceph: set 128MB tcmalloc cache size
Alexandre Derumier [Thu, 28 Apr 2016 20:55:37 +0000 (22:55 +0200)]
etc/default/ceph: set 128MB tcmalloc cache size

Signed-off-by: Alexandre Derumier <aderumier@odiso.com>
9 years agoMerge pull request #8568 from ddiss/osd-librados-writesame
Josh Durgin [Thu, 28 Apr 2016 19:57:02 +0000 (14:57 -0500)]
Merge pull request #8568 from ddiss/osd-librados-writesame

OSD and librados writesame support

Reviewed-by: Josh Durgin <jdurgin@redhat.com>
Reviewed-by: Haomai Wang <haomai@xsky.com>
9 years agoMerge pull request #8791 from dillaman/wip-15574
Josh Durgin [Thu, 28 Apr 2016 19:55:20 +0000 (14:55 -0500)]
Merge pull request #8791 from dillaman/wip-15574

librbd: parent image is closed twice if error encountered while opening

Reviewed-by: Josh Durgin <jdurgin@redhat.com>
9 years agoMerge pull request #8790 from dillaman/wip-15630
Josh Durgin [Thu, 28 Apr 2016 19:54:57 +0000 (14:54 -0500)]
Merge pull request #8790 from dillaman/wip-15630

rbd-mirror: interrupting image replay startup can cause crash

Reviewed-by: Josh Durgin <jdurgin@redhat.com>
9 years agoMerge pull request #8785 from tchaikov/wip-librbd-bad-header
Josh Durgin [Thu, 28 Apr 2016 19:54:30 +0000 (14:54 -0500)]
Merge pull request #8785 from tchaikov/wip-librbd-bad-header

librbd: does not crash if image header is too short

Reviewed-by: Josh Durgin <jdurgin@redhat.com>
9 years agoMerge pull request #8813 from trociny/fix-test-librbd-aio_write
Jason Dillaman [Thu, 28 Apr 2016 18:25:28 +0000 (13:25 -0500)]
Merge pull request #8813 from trociny/fix-test-librbd-aio_write

test: librbd: aio write/discard completion should always return 0 on success

Reviewed-by: Jason Dillaman <dillaman@redhat.com>
9 years agotest: librbd: aio write/discard completion should always return 0 on success 8813/head
Mykola Golub [Thu, 28 Apr 2016 12:19:48 +0000 (15:19 +0300)]
test: librbd: aio write/discard completion should always return 0 on success

Signed-off-by: Mykola Golub <mgolub@mirantis.com>
9 years agoMerge pull request #8795 from runsisi/wip-fix-librbd-ut
Jason Dillaman [Thu, 28 Apr 2016 13:40:39 +0000 (08:40 -0500)]
Merge pull request #8795 from runsisi/wip-fix-librbd-ut

cmake: add library cls_journal for target unittest_librbd

Reviewed-by: Jason Dillaman <dillaman@redhat.com>
9 years agoMerge pull request #8796 from runsisi/wip-fix-invalidate
Jason Dillaman [Thu, 28 Apr 2016 13:38:57 +0000 (08:38 -0500)]
Merge pull request #8796 from runsisi/wip-fix-invalidate

librbd: fix potential assertion during object map invalidation

Reviewed-by: Jason Dillaman <dillaman@redhat.com>
9 years agoMerge pull request #8812 from liewegas/wip-dh-systemd
Sage Weil [Thu, 28 Apr 2016 13:14:41 +0000 (09:14 -0400)]
Merge pull request #8812 from liewegas/wip-dh-systemd

debian/control: dh_systemd_start is in the dh-systemd package

Reviewed-by: Kefu Chai <kchai@redhat.com>
9 years agodebian/control: dh_systemd_start is in the dh-systemd package 8812/head
Sage Weil [Thu, 28 Apr 2016 13:13:50 +0000 (09:13 -0400)]
debian/control: dh_systemd_start is in the dh-systemd package

Fixes: http://tracker.ceph.com/issues/15573
Signed-off-by: Sage Weil <sage@redhat.com>
9 years agoMerge pull request #8700 from tchaikov/wip-systemd-deb
Sage Weil [Thu, 28 Apr 2016 03:15:52 +0000 (23:15 -0400)]
Merge pull request #8700 from tchaikov/wip-systemd-deb

debian: package systemd targets

Reviewed-by: Javier M. Mellid <jmunhoz@igalia.com>
9 years agocmake: add library cls_journal for target unittest_librbd 8795/head
runsisi [Thu, 28 Apr 2016 02:22:07 +0000 (10:22 +0800)]
cmake: add library cls_journal for target unittest_librbd

otherwise if we want to generate target unittest_librbd we need to
(re)generate cls_journal manually

Signed-off-by: runsisi <runsisi@zte.com.cn>
9 years agodebian: install systemd target files 8700/head
Kefu Chai [Fri, 22 Apr 2016 13:07:58 +0000 (21:07 +0800)]
debian: install systemd target files

* enable it using dh_systemd_enable
* start the target using dh_systemd_start
* move the dh_installinit, dh_systemd_enable, dh_systemd_start calls
  down, so they can identify the service files if they care about them.

Fixes: http://tracker.ceph.com/issues/15573
Signed-off-by: Kefu Chai <kchai@redhat.com>
9 years agolibrbd: fix potential assertion for object map invalidation 8796/head
runsisi [Thu, 28 Apr 2016 01:05:33 +0000 (09:05 +0800)]
librbd: fix potential assertion for object map invalidation

if update_flags fails we should not go any further, or the async request
will be removed from m_image_ctx.async_requests twice

Signed-off-by: runsisi <runsisi@zte.com.cn>
9 years agoMerge pull request #8792 from cbodley/wip-defer-drop-privileges
Sage Weil [Wed, 27 Apr 2016 21:13:03 +0000 (17:13 -0400)]
Merge pull request #8792 from cbodley/wip-defer-drop-privileges

common: fix value of CINIT_FLAG_DEFER_DROP_PRIVILEGES

Reviewed-by: Sage Weil <sage@redhat.com>
9 years agocommon: fix value of CINIT_FLAG_DEFER_DROP_PRIVILEGES 8792/head
Casey Bodley [Wed, 27 Apr 2016 20:53:26 +0000 (16:53 -0400)]
common: fix value of CINIT_FLAG_DEFER_DROP_PRIVILEGES

0x16 (binary 10110) was overlapping two other flags

Signed-off-by: Casey Bodley <cbodley@redhat.com>
9 years agolibrbd: propagate any image refresh errors back to caller 8791/head
Jason Dillaman [Wed, 27 Apr 2016 19:56:11 +0000 (15:56 -0400)]
librbd: propagate any image refresh errors back to caller

Previously, the saved error code was only propagated when an image
flush was required.

Signed-off-by: Jason Dillaman <dillaman@redhat.com>
9 years agolibrbd: avoid second close attempt of invalid parent image
Jason Dillaman [Wed, 27 Apr 2016 19:45:25 +0000 (15:45 -0400)]
librbd: avoid second close attempt of invalid parent image

Fixes: http://tracker.ceph.com/issues/15574
Signed-off-by: Jason Dillaman <dillaman@redhat.com>
9 years agorbd-mirror: interrupting image replay startup can cause crash 8790/head
Jason Dillaman [Wed, 27 Apr 2016 19:22:59 +0000 (15:22 -0400)]
rbd-mirror: interrupting image replay startup can cause crash

Properly shut down the journaler if a shut down has been requested
while start up is in-progress.

Fixes: http://tracker.ceph.com/issues/15630
Signed-off-by: Jason Dillaman <dillaman@redhat.com>
9 years agoMerge pull request #8714 from ceph/wip-14941
Boris Ranto [Wed, 27 Apr 2016 18:13:38 +0000 (20:13 +0200)]
Merge pull request #8714 from ceph/wip-14941

rpm: implement scriptlets for the post-split daemon packages

Reviewed-by: Tim Serong <tserong@suse.com>
Reviewed-by: Nathan Cutler <ncutler@suse.com>
Reviewed-by: Boris Ranto <branto@redhat.com>
9 years agolibrbd: does not crash if image header is too short 8785/head
Kefu Chai [Wed, 27 Apr 2016 16:59:34 +0000 (00:59 +0800)]
librbd: does not crash if image header is too short

if something goes wrong with the object, and returns a chunk shorter
than expected, don't panic

Signed-off-by: Kefu Chai <kchai@redhat.com>
9 years agorpm: Add rpm scripts for ceph-rbd-mirror 8714/head
Boris Ranto [Wed, 27 Apr 2016 16:10:51 +0000 (18:10 +0200)]
rpm: Add rpm scripts for ceph-rbd-mirror

We are currently missing the systemd scripts for the new ceph-rbd-mirror
daemons. This patch introduces them.

Signed-off-by: Boris Ranto <branto@redhat.com>
9 years agorpm: Start all the targets in %post
Boris Ranto [Wed, 27 Apr 2016 16:00:08 +0000 (18:00 +0200)]
rpm: Start all the targets in %post

We need to have the sub-targets active for the ceph.target to be able
to propagate its calls to the services. If the sub-target is inactive,
the main target won't propagate the stop/restart calls.

Signed-off-by: Boris Ranto <branto@redhat.com>
9 years agorpm: implement scriptlets for the post-split daemon packages
Nathan Cutler [Tue, 19 Apr 2016 12:48:41 +0000 (14:48 +0200)]
rpm: implement scriptlets for the post-split daemon packages

This patch gives each of the ceph-{mds,mon,osd,radosgw} packages its own
%post, %preun, and %postun scriptlets dealing with the package's unit files.

The scriptlets of ceph-base are adapted to handle the ceph.target unit file
only.

The scriptlets of ceph-mon handle ceph-create-keys services in addition to ceph-mon.

The scriptlets of ceph-osd handle ceph-disk services in addition to ceph-osd.

Fixes: http://tracker.ceph.com/issues/14941
Signed-off-by: Nathan Cutler <ncutler@suse.com>
Signed-off-by: Boris Ranto <branto@redhat.com>
9 years agosystemd: enable all the ceph .target services by default
Nathan Cutler [Tue, 19 Apr 2016 12:51:01 +0000 (14:51 +0200)]
systemd: enable all the ceph .target services by default

Some distros, like Fedora and openSUSE, have a policy that all services are
disabled by default.

This patch changes that default for the ceph.target and
ceph-{mds,mon,osd,radosgw}.target services.

Signed-off-by: Nathan Cutler <ncutler@suse.com>
Signed-off-by: Boris Ranto <branto@redhat.com>
9 years agoSigned-off-by: Tamil Muthamizhan <tmuthami@redhat.com>
Tamil Muthamizhan [Tue, 26 Apr 2016 23:13:28 +0000 (16:13 -0700)]
Signed-off-by: Tamil Muthamizhan <tmuthami@redhat.com>
removed ceph-deploy folder, as its been already replaced with ceph-tests.

9 years agoMerge pull request #8746 from yehudasa/wip-15598
Casey Bodley [Tue, 26 Apr 2016 15:49:42 +0000 (11:49 -0400)]
Merge pull request #8746 from yehudasa/wip-15598

Fixes: http://tracker.ceph.com/issues/15598
Signed-off-by: Yehuda Sadeh <yehuda@redhat.com>
Reviewed-by: Casey Bodley <cbodley@redhat.com>
9 years agoMerge pull request #8762 from ceph/wip-15595
Yehuda Sadeh [Tue, 26 Apr 2016 15:37:42 +0000 (08:37 -0700)]
Merge pull request #8762 from ceph/wip-15595

rgw: fix bad free in RGWCreateBucket_ObjStore_S3

Reviewed-by: Yehuda Sadeh <yehuda@redhat.com>
9 years agorgw: fix bad free in RGWCreateBucket_ObjStore_S3 8762/head
Casey Bodley [Tue, 26 Apr 2016 14:36:58 +0000 (10:36 -0400)]
rgw: fix bad free in RGWCreateBucket_ObjStore_S3

get_params() will accept -ERR_LENGTH_REQUIRED from
rgw_rest_read_all_input(), in which case 'data' will not be allocated
and we attempt to free() an uninitialized pointer

Fixes: http://tracker.ceph.com/issues/15595
Signed-off-by: Casey Bodley <cbodley@redhat.com>
9 years agomsg/async: add numevents statistics for external_eventsReviewed-by: Haomai Wang ...
Haomai Wang [Tue, 26 Apr 2016 14:14:31 +0000 (22:14 +0800)]
msg/async: add numevents statistics for external_eventsReviewed-by: Haomai Wang <haomai@xsky.com>

msg/async: add numevents statistics for external_events

Reviewed-by: Haomai Wang <haomai@xsky.com>
9 years agorbd: CLI to retrieve rbd mirror state for a pool / specific image
Mykola Golub [Tue, 5 Apr 2016 11:42:09 +0000 (14:42 +0300)]
rbd: CLI to retrieve rbd mirror state for a pool / specific image

Fixes: #15144
Signed-off-by: Mykola Golub <mgolub@mirantis.com>
9 years agolibrbd: API to retrieve image mirroring status
Mykola Golub [Fri, 15 Apr 2016 05:42:17 +0000 (08:42 +0300)]
librbd: API to retrieve image mirroring status

Signed-off-by: Mykola Golub <mgolub@mirantis.com>
9 years agorbd-mirror: store replay status in mirroring object
Mykola Golub [Tue, 5 Apr 2016 11:40:50 +0000 (14:40 +0300)]
rbd-mirror: store replay status in mirroring object

Fixes: #14420
Signed-off-by: Mykola Golub <mgolub@mirantis.com>
9 years agotest: MockJournaler: add get_cached_client method
Mykola Golub [Wed, 6 Apr 2016 11:57:29 +0000 (14:57 +0300)]
test: MockJournaler: add get_cached_client method

Signed-off-by: Mykola Golub <mgolub@mirantis.com>
9 years agotest: cls_rbd: methods to store/retrieve mirroring status
Mykola Golub [Tue, 5 Apr 2016 11:25:11 +0000 (14:25 +0300)]
test: cls_rbd: methods to store/retrieve mirroring status

Signed-off-by: Mykola Golub <mgolub@mirantis.com>
9 years agocls::rbd:: methods to store/retrieve mirroring status
Mykola Golub [Tue, 5 Apr 2016 11:24:07 +0000 (14:24 +0300)]
cls::rbd:: methods to store/retrieve mirroring status

Signed-off-by: Mykola Golub <mgolub@mirantis.com>
9 years agoobjclass: add method to list watchers
Mykola Golub [Fri, 1 Apr 2016 05:27:03 +0000 (08:27 +0300)]
objclass: add method to list watchers

Signed-off-by: Mykola Golub <mgolub@mirantis.com>
9 years agoMerge pull request #8683 from Ved-vampir/zlib_fix
Haomai Wang [Tue, 26 Apr 2016 03:43:51 +0000 (11:43 +0800)]
Merge pull request #8683 from Ved-vampir/zlib_fix

compressor: fix of error code control in zlib decompress

Reviewed-by: Haomai Wang <haomai@xsky.com>
9 years agoMerge pull request #8489 from SUSE/wip-15419
Sage Weil [Tue, 26 Apr 2016 02:33:20 +0000 (22:33 -0400)]
Merge pull request #8489 from SUSE/wip-15419

systemd: make Ceph daemons dependent upon time-sync.target

Reviewed-by: Tim Serong <tserong@suse.com>
Reviewed-by: James Page <james.page@ubuntu.com>
Reviewed-by: Ken Dreyer <kdreyer@redhat.com>
9 years agoMerge pull request #8721 from ibravo/patch-2
Sage Weil [Tue, 26 Apr 2016 02:21:12 +0000 (22:21 -0400)]
Merge pull request #8721 from ibravo/patch-2

doc: update quick-ceph-deploy.rst

Reviewed-by: Sage Weil <sage@redhat.com>
9 years agoMerge pull request #8713 from HeyoQiang/master
Sage Weil [Tue, 26 Apr 2016 02:20:21 +0000 (22:20 -0400)]
Merge pull request #8713 from HeyoQiang/master

osd/PG: remove unused variable have

Reviewed-by: xie xingguo <xie.xingguo@zte.com.cn>
9 years agoPG:clear an unused variable: have 8713/head
YongQiang [Sat, 23 Apr 2016 10:30:34 +0000 (06:30 -0400)]
PG:clear an unused variable: have

Signed-off-by: Yongqiang He <he.yongqiang@h3c.com>
9 years agoMerge pull request #8723 from cooboos/wrong_spell
Sage Weil [Tue, 26 Apr 2016 02:18:51 +0000 (22:18 -0400)]
Merge pull request #8723 from cooboos/wrong_spell

doc: correct wrong spell 'mnay' to 'many'

Reviewed-by: xie xingguo <xie.xingguo@zte.com.cn>
9 years agoMerge pull request #6764 from kylinstorage/fix-check_become_tier
Sage Weil [Tue, 26 Apr 2016 02:01:19 +0000 (22:01 -0400)]
Merge pull request #6764 from kylinstorage/fix-check_become_tier

mon: fix OSDMonitor::_check_become_tier

Reviewed-by: Xinze Chi <xinze@xsky.com>
9 years agoMerge pull request #8103 from adamemerson/wip-sign-of-the-time-debug
Sage Weil [Tue, 26 Apr 2016 01:59:40 +0000 (21:59 -0400)]
Merge pull request #8103 from adamemerson/wip-sign-of-the-time-debug

common: signed coarse time

Reviewed-by: Kefu Chai <kchai@redhat.com>
9 years agoMerge pull request #8407 from dx9/wip-os-release
Sage Weil [Tue, 26 Apr 2016 01:59:06 +0000 (21:59 -0400)]
Merge pull request #8407 from dx9/wip-os-release

common/util: add support for distro info from /etc/os-release file

Reviewed-by: Sage Weil <sage@redhat.com>
9 years agoMerge pull request #8437 from jjhuo/single_delete
Sage Weil [Tue, 26 Apr 2016 01:58:55 +0000 (21:58 -0400)]
Merge pull request #8437 from jjhuo/single_delete

os/kv, os/bluestore: Enable SingleDelete with RocksDB and Bluestore

Reviewed-by: Sage Weil <sage@redhat.com>
9 years agoMerge pull request #8451 from tchaikov/wip-signal-hanlder-typo
Sage Weil [Tue, 26 Apr 2016 01:58:22 +0000 (21:58 -0400)]
Merge pull request #8451 from tchaikov/wip-signal-hanlder-typo

global/signal_handler: fix typo

Reviewed-by: Sage Weil <sage@redhat.com>
9 years agoMerge pull request #8503 from yuyuyu101/wip-spdk-improve
Sage Weil [Tue, 26 Apr 2016 01:58:03 +0000 (21:58 -0400)]
Merge pull request #8503 from yuyuyu101/wip-spdk-improve

os/bluestore/NVMEDevice: performance improvements

9 years agoMerge pull request #8702 from kmroz/wip-fix-run-dir-chown
Sage Weil [Tue, 26 Apr 2016 01:56:42 +0000 (21:56 -0400)]
Merge pull request #8702 from kmroz/wip-fix-run-dir-chown

global-init: fixup chown of the run directory along with log and asok files

Reviewed-by: Sage Weil <sage@redhat.com>
9 years agoMerge pull request #8728 from liewegas/wip-notify
Sage Weil [Tue, 26 Apr 2016 01:55:54 +0000 (21:55 -0400)]
Merge pull request #8728 from liewegas/wip-notify

osd/osd_types: render notify op properly

Reviewed-by: Kefu Chai <kchai@redhat.com>
9 years agoos/bluestore/BlueStore: cleanup overlay using single delete 8437/head
Jianjian Huo [Mon, 4 Apr 2016 22:07:01 +0000 (15:07 -0700)]
os/bluestore/BlueStore: cleanup overlay using single delete

Overlay entries are also only written once, eligible to use single delete.

Signed-off-by: Jianjian Huo <jianjian.huo@ssi.samsung.com>
9 years agoos/bluestore/BlueStore: cleanup the wal entries using single delete
Jianjian Huo [Mon, 4 Apr 2016 21:50:19 +0000 (14:50 -0700)]
os/bluestore/BlueStore: cleanup the wal entries using single delete

By use of single delete, RocksDB should be able to remove deleted wal
entries with only one compaction in theory, when wal entries land on level0.
This should reduce bluestore wal entries incurred WAF.

Signed-off-by: Jianjian Huo <jianjian.huo@ssi.samsung.com>
9 years agokv/RocksDBStore: implement rm_single_delete
Jianjian Huo [Mon, 4 Apr 2016 21:44:53 +0000 (14:44 -0700)]
kv/RocksDBStore: implement rm_single_delete

Use SingleDelete to delete keys which were never overwritten.

Signed-off-by: Jianjian Huo <jianjian.huo@ssi.samsung.com>
9 years agokv/KeyValueDB: add a new interface for single delete.
Jianjian Huo [Mon, 4 Apr 2016 21:18:50 +0000 (14:18 -0700)]
kv/KeyValueDB: add a new interface for single delete.

This is useful for log-insert-merge tree based key value store, such as
RocksDB, to avoid more LSM compactions for already deleted key value pairs.

Signed-off-by: Jianjian Huo <jianjian.huo@ssi.samsung.com>
9 years agoMerge pull request #8474 from xiaoxichen/wip-pycephfs-namespace
John Spray [Mon, 25 Apr 2016 22:49:45 +0000 (17:49 -0500)]
Merge pull request #8474 from xiaoxichen/wip-pycephfs-namespace

pycephfs: using namespace for data isolation

Reviewed-by: John Spray <john.spray@redhat.com>
9 years agorgw: fix suffix in RGWZoneParams::fix_pool_names() 8746/head
Yehuda Sadeh [Mon, 25 Apr 2016 22:09:47 +0000 (15:09 -0700)]
rgw: fix suffix in RGWZoneParams::fix_pool_names()

Fixes: http://tracker.ceph.com/issues/15598
Signed-off-by: Yehuda Sadeh <yehuda@redhat.com>
9 years agoceph.spec.in: recommend ntp-daemon on SUSE 8489/head
Nathan Cutler [Sat, 23 Apr 2016 09:42:25 +0000 (11:42 +0200)]
ceph.spec.in: recommend ntp-daemon on SUSE

SUSE has settled on "ntp-daemon" as the generic package name. The "ntp" and
"chrony" etc. packages have "Provides: ntp-daemon" in their respective spec
files.

References: http://tracker.ceph.com/issues/15419
Signed-off-by: Nathan Cutler <ncutler@suse.com>
9 years agodeb: make ceph-base and radosgw recommend time-daemon
Nathan Cutler [Thu, 7 Apr 2016 18:54:21 +0000 (20:54 +0200)]
deb: make ceph-base and radosgw recommend time-daemon

References: http://tracker.ceph.com/issues/15419

Signed-off-by: Nathan Cutler <ncutler@suse.com>
9 years agoMerge pull request #8715 from runsisi/wip-fix-cmake
Ali Maredia [Mon, 25 Apr 2016 17:30:45 +0000 (13:30 -0400)]
Merge pull request #8715 from runsisi/wip-fix-cmake

cmake: use LINK_PRIVATE instead of PRIVATE for CMake version == 2.8.11

Reviewed-by: Ali Maredia <amaredia@redhat.com>
9 years agoMerge pull request #8730 from ibravo/patch-3
Sage Weil [Mon, 25 Apr 2016 17:20:27 +0000 (13:20 -0400)]
Merge pull request #8730 from ibravo/patch-3

doc: Updated ceph-deploy osd to choose underlying file system

9 years agoMerge pull request #8745 from ErwanAliasr1/evelu-deps
Sage Weil [Mon, 25 Apr 2016 16:24:50 +0000 (12:24 -0400)]
Merge pull request #8745 from ErwanAliasr1/evelu-deps

install-deps: Adding gcc for debian based systems

Reviewed-by: Sage Weil <sage@redhat.com>
9 years agoadmin-socket: use chown instead of fchown 8702/head
Karol Mroz [Mon, 25 Apr 2016 15:24:07 +0000 (17:24 +0200)]
admin-socket: use chown instead of fchown

fchown() returns success, but does not change asok file
ownership. chown() does.

Signed-off-by: Karol Mroz <kmroz@suse.com>
9 years agoinstall-deps: Adding gcc for debian based systems 8745/head
Erwan Velu [Mon, 25 Apr 2016 15:11:45 +0000 (17:11 +0200)]
install-deps: Adding gcc for debian based systems

When running install-deps on a minimalistic system, we reach that situation :

dpkg-checkbuilddeps --admindir=/tmp/install-deps.5526 debian/control
    sh: 1: gcc: not found
    dpkg-checkbuilddeps: warning: Couldn't determine gcc system type, falling back to default (native compilation)
    dpkg-checkbuilddeps: error: cannot open /tmp/install-deps.5526/status: No such file or directory

This means that we shall install gcc before calling dpkg-checkbuilddeps.

Signed-off-by: Erwan Velu <erwan@redhat.com>
9 years agoMerge pull request #8743 from ceph/revert-8502-song20
Kefu Chai [Mon, 25 Apr 2016 14:04:02 +0000 (22:04 +0800)]
Merge pull request #8743 from ceph/revert-8502-song20

Revert "crushtool: do not allow to compile crushmap with repeat rules…

Reviewed-by: Sage Weil <sage@redhat.com>
9 years agoRevert "crushtool: do not allow to compile crushmap with repeat ruleset." 8743/head
Kefu Chai [Mon, 25 Apr 2016 14:00:47 +0000 (22:00 +0800)]
Revert "crushtool: do not allow to compile crushmap with repeat ruleset."

9 years agoTestRados: add writesame support 8568/head
David Disseldorp [Thu, 14 Apr 2016 15:52:53 +0000 (17:52 +0200)]
TestRados: add writesame support

Can be tested via "--op writesame". Requests are currently dispatched
*without* a multiplication factor, i.e. data_len == write_len.

Signed-off-by: David Disseldorp <ddiss@suse.de>
9 years agoceph_test_rados_api_c_write_ops: add writesame test
David Disseldorp [Sun, 10 Apr 2016 12:45:03 +0000 (14:45 +0200)]
ceph_test_rados_api_c_write_ops: add writesame test

Signed-off-by: David Disseldorp <ddiss@suse.de>
9 years agoceph_test_rados_api_io: ETOBIG tests for writesame
David Disseldorp [Tue, 12 Apr 2016 23:00:04 +0000 (01:00 +0200)]
ceph_test_rados_api_io: ETOBIG tests for writesame

Signed-off-by: David Disseldorp <ddiss@suse.de>
9 years agoceph_test_rados_api_aio: add writesame tests
David Disseldorp [Thu, 7 Apr 2016 13:06:45 +0000 (15:06 +0200)]
ceph_test_rados_api_aio: add writesame tests

Write a buffer a number of times using writesame, read the full range
back, and check that it matches. Do this using rados_aio_writesame(),
ioctx.aio_writesame() and ioctx.aio_operate(op.writesame()).

Signed-off-by: David Disseldorp <ddiss@suse.de>
9 years agoceph_test_rados_api_misc: add preliminary writesame tests
David Disseldorp [Wed, 6 Apr 2016 17:17:47 +0000 (19:17 +0200)]
ceph_test_rados_api_misc: add preliminary writesame tests

Test the new ioctx.writesame() and rados_writesame() API functions.

Signed-off-by: David Disseldorp <ddiss@suse.de>
9 years agolibrados: add writesame API
David Disseldorp [Wed, 6 Apr 2016 17:13:18 +0000 (19:13 +0200)]
librados: add writesame API

The writesame operation allows callers to write the same data buffer
multiple times to a given object.

Signed-off-by: David Disseldorp <ddiss@suse.de>
9 years agoceph osd: add support for new op writesame
Mike Christie [Wed, 6 Apr 2016 17:13:14 +0000 (12:13 -0500)]
ceph osd: add support for new op writesame

This adds a new ceph request writesame that writes a buffer of length
writesame.data_length bytes at writesame.offset over
writesame.length bytes.

This command maps to SCSI's WRITE SAME request, so users like LIO+rbd
can pass this to the OSD. Right now, it only saves having to transfer
writesame.length bytes over the network, but future versions will be
to fully offload it by passing it directly to the FS/devices if they
support it.

v2:
- Fix tab/spaces to matching coding style.
- Allow zero write length. Check for invalid data lengths.

Signed-off-by: Mike Christie <mchristi@redhat.com>
Reviewed-by: David Disseldorp <ddiss@suse.de>
9 years agocompressor: fix of error code control in zlib decompress 8683/head
Ved-vampir [Thu, 21 Apr 2016 15:40:23 +0000 (18:40 +0300)]
compressor: fix of error code control in zlib decompress

Signed-off-by: Alyona Kiseleva <akiselyova@mirantis.com>
9 years agoMerge pull request #8725 from liewegas/wip-make-check
Kefu Chai [Mon, 25 Apr 2016 05:58:20 +0000 (13:58 +0800)]
Merge pull request #8725 from liewegas/wip-make-check

test/encoding/readable.sh: fix make check for autotool

Reviewed-by: Kefu Chai <kchai@redhat.com>
9 years agotest/encoding/readable.sh: fix make check for autotools 8725/head
Sage Weil [Mon, 25 Apr 2016 05:12:45 +0000 (01:12 -0400)]
test/encoding/readable.sh: fix make check for autotools

Use ./ if ceph-dencoder is in the local directory.

Signed-off-by: Sage Weil <sage@redhat.com>
9 years agoMerge pull request #8719 from SUSE/wip-fix-multidump
Kefu Chai [Mon, 25 Apr 2016 03:04:47 +0000 (11:04 +0800)]
Merge pull request #8719 from SUSE/wip-fix-multidump

multi-dump.sh: fix usage function

Reviewed-by: Kefu Chai <kchai@redhat.com>
9 years agoDoc: Updated ceph-deploy osd to choose underlying file system 8730/head
ibravo [Mon, 25 Apr 2016 00:28:35 +0000 (20:28 -0400)]
Doc: Updated ceph-deploy osd to choose underlying file system

Included paragraph about --fs-type and --bluestore

Signed-off-by: Ignacio Bravo ibravo@hotmail.com
9 years agoDoc: Update quick-ceph-deploy.rst 8721/head
ibravo [Sat, 23 Apr 2016 19:14:06 +0000 (15:14 -0400)]
Doc: Update quick-ceph-deploy.rst

Signed-off-by: Ignacio Bravo ibravo@hotmail.com
9 years agoosd/osd_types: render notify OSDOp properly 8728/head
Sage Weil [Sun, 24 Apr 2016 20:03:05 +0000 (16:03 -0400)]
osd/osd_types: render notify OSDOp properly

Signed-off-by: Sage Weil <sage@redhat.com>
9 years agoMerge pull request #8640 from xiexingguo/xxg-wip-osdsideeffect
Sage Weil [Sun, 24 Apr 2016 19:51:06 +0000 (15:51 -0400)]
Merge pull request #8640 from xiexingguo/xxg-wip-osdsideeffect

osd/ReplicatedPG: allow CMPXATTR ops to trigger side effects too

Reviewed-by: Sage Weil <sage@redhat.com>
9 years agoMerge pull request #8694 from wuxingyi/fixincomplete
Sage Weil [Sun, 24 Apr 2016 19:50:00 +0000 (15:50 -0400)]
Merge pull request #8694 from wuxingyi/fixincomplete

osd/PG: PG is only likely to go active when min_size turns smaller.

Reviewed-by: Sage Weil <sage@redhat.com>
9 years agoosd: honour result code passed in; tidy-up changesReviewed-by: Sage Weil <sage@redhat...
Sage Weil [Sun, 24 Apr 2016 19:49:09 +0000 (15:49 -0400)]
osd: honour result code passed in; tidy-up changesReviewed-by: Sage Weil <sage@redhat.com>

osd: honour result code passed in; tidy-up changes

Reviewed-by: Sage Weil <sage@redhat.com>
9 years agoMerge pull request #8675 from roidayan/xio-build_deb_with_xio
Sage Weil [Sun, 24 Apr 2016 19:48:35 +0000 (15:48 -0400)]
Merge pull request #8675 from roidayan/xio-build_deb_with_xio

xio: add MNop.h to dist tarball

Reviewed-by: Sage Weil <sage@redhat.com>
9 years agoMerge pull request #8678 from jack-changtao/pgbackend
Sage Weil [Sun, 24 Apr 2016 19:48:25 +0000 (15:48 -0400)]
Merge pull request #8678 from jack-changtao/pgbackend

osd/PGBackend: if min > max, PGBackend::objects_list_partial will list all objects of that pg

Reviewed-by: Sage Weil <sage@redhat.com>
9 years agoMerge pull request #8680 from ceph/rm-distribute
Sage Weil [Sun, 24 Apr 2016 19:48:09 +0000 (15:48 -0400)]
Merge pull request #8680 from ceph/rm-distribute

tools: remove installation for distribute. It is no longer maintained

Reviewed-by: Sage Weil <sage@redhat.com>