RepoMirrors/ceph
1
0
Fork 0
mirror of https://github.com/ceph/ceph synced 2025-02-15 23:07:55 +00:00
Code Issues Packages Projects Releases Wiki Activity
58,686 Commits 216 Branches 395 Tags 1.2 GiB
09a2383673
Commit Graph

58686 Commits

This Branch
This Branch
All Branches
Author SHA1 Message Date
Sage Weil
8c4cabec0f unittest_bluestore_types: attach blobs to cache 
Signed-off-by: Sage Weil <sage@redhat.com>
 2016-10-03 18:42:03 -04:00
Sage Weil
c8a06ce1ee do_cmake.sh: set .ceph_port to random port 
Signed-off-by: Sage Weil <sage@redhat.com>
 2016-10-03 18:40:07 -04:00
Sage Weil
33e26d021d os/bluestore: add perf counters for cache size 
Onodes, buffers, buffer bytes.

Signed-off-by: Sage Weil <sage@redhat.com>
 2016-10-03 18:40:07 -04:00
Sage Weil
6098761000 os/bluestore: fix typo in write_penalty_read_ops perf counter 
Signed-off-by: Sage Weil <sage@redhat.com>
 2016-10-03 18:40:06 -04:00
Sage Weil
cbfa540c68 os/bluestore: fix cache tunables to be process-wide 
Divide by the shard count so that the user isn't totally
confused.

Signed-off-by: Sage Weil <sage@redhat.com>
 2016-10-03 18:40:06 -04:00
Sage Weil
2efe05128b Merge pull request #11300 from liewegas/wip-bluestore-onode-add 
os/bluestore: fix onode cache addition race
 2016-10-03 17:11:57 -05:00
Sage Weil
2150e9914b os/bluestore: fix onode cache addition race 
Two threads may try to add the same onode to the cache.
This is rare, but allowed (in the case of the meta
collection).  If that happens, one of them will just
back off and use the winning onode ref.

Signed-off-by: Sage Weil <sage@redhat.com>
 2016-10-03 17:37:15 -04:00
Sage Weil
f7a9a6f656 mgr/MgrClient: implement con reset handling 
Signed-off-by: Sage Weil <sage@redhat.com>
 2016-10-03 14:42:35 -04:00
Gregory Farnum
8d37c68b56 Merge pull request #11298 from liewegas/wip-mgr-fix 
mgr/MgrClient: fix ms_handle_reset
Reviewed-by: Greg Farnum <gfarnum@redhat.com>
 2016-10-03 11:13:12 -07:00
Sage Weil
848ee05932 mgr/MgrClient: fix ms_handle_reset 
Return false because we don't handle the reset.

Signed-off-by: Sage Weil <sage@redhat.com>
 2016-10-03 13:59:51 -04:00
Radoslaw Zarzynski
cfc0a16e04 Merge pull request #11292 from rzarzynski/yj-wip-swift-build_token-cleanup 
rgw/swift: remove redundant assignment operation

Reviewed-by: Radoslaw Zarzynski <rzarzynski@mirantis.com>
 2016-10-03 17:53:19 +02:00
Kefu Chai
eb62120723 Merge pull request #11019 from liewegas/wip-autotools-must-die 
autotools: remove a few other remaining traces

Reviewed-by: Kefu Chai <kchai@redhat.com>
 2016-10-03 23:52:38 +08:00
Jason Dillaman
ada727047d Merge pull request #11294 from trociny/fixup-qa-rbd_mirror-metadata_set 
qa/workunits/rbd: set image-meta on primary image and wait it is replicated

Reviewed-by: Jason Dillaman <dillaman@redhat.com>
 2016-10-03 11:50:04 -04:00
Sage Weil
54134418ef osd: drop stray debug message 
This snuck in from 10da6d23cd

Signed-off-by: Sage Weil <sage@redhat.com>
 2016-10-03 11:17:56 -04:00
Sage Weil
8242afb135 Merge pull request #11288 from stiopaa1/osd_replicatedpg_moveCOSDOnAppliedClass 
osd/ReplicatedPG:move C_OSD_OnApplied class to cc

Reviewed-by: Sage Weil <sage@redhat.com>
 2016-10-03 09:26:03 -05:00
Sage Weil
dbef9d057e Merge pull request #11289 from stiopaa1/mds_sessionmap_moveStatementUnderAssertPushPv 
mds/SessionMap.h: change statement to assertion

Reviewed-by: Patrick Donnelly <pdonnell@redhat.com>
 2016-10-03 09:25:36 -05:00
Mykola Golub
eb4097b8ef qa/workunits/rbd: set image-meta on primary image and wait it is replicated 
After recently added image metadata replication it is not possible any
more to update it on non-primary image.

Signed-off-by: Mykola Golub <mgolub@mirantis.com>
 2016-10-03 17:02:59 +03:00
Sage Weil
b424a35f3a run-make-check.sh: use cmake exclusively 
..and drop the separate cmake script

Signed-off-by: Sage Weil <sage@redhat.com>
 2016-10-03 09:46:50 -04:00
Sage Weil
5bd88a54bb make-debs.sh: use make-dist 
Signed-off-by: Sage Weil <sage@redhat.com>
 2016-10-03 09:46:50 -04:00
Yan Jun
f0ae5fe1bd rgw/swift: remove redundant assignment operation 
Signed-off-by: Yan Jun <yan.jun8@zte.com.cn>
 2016-10-03 14:11:09 +02:00
Loic Dachary
00873f26e5 Merge pull request #11086 from bassamtabbara/wip-ec-simd-runtime-detection 
erasure-code: Runtime detection of SIMD for jerasure and shec

Reviewed-by: Loic Dachary <ldachary@redhat.com>
Reviewed-by: Sage Weil <sage@redhat.com>
 2016-10-03 11:43:29 +02:00
Mykola Golub
2882f8c01a qa/workunits/rbd: check status also in pool dir after asok commands 
wait_for_image_replay_stopped returns not when the state is stopped,
but when the state is not replaying. So a race was possible when an
asok command was running when the previos stop command was still in
progress, leading to unexpected results.

Signed-off-by: Mykola Golub <mgolub@mirantis.com>
 2016-10-03 10:10:33 +03:00
John Spray
92fdcf36bd mon: don't crash on invalid standby_for_fscid 
Fixes: http://tracker.ceph.com/issues/17466
Signed-off-by: John Spray <john.spray@redhat.com>
 2016-10-03 07:42:19 +01:00
John Spray
63d711d4e9 mds: make mds_role_t member order match constructor 
Previously this was mildy confusing.

Signed-off-by: John Spray <john.spray@redhat.com>
 2016-10-03 07:42:02 +01:00
Michal Jarzabek
e3c26b94fe mds/SessionMap.cc: avoid copying and add const 
Signed-off-by: Michal Jarzabek <stiopa@gmail.com>
 2016-10-02 22:06:34 +01:00
Michal Jarzabek
677c8f6aad mds/SessionMap.h: change statement to assertion 
Signed-off-by: Michal Jarzabek <stiopa@gmail.com>
 2016-10-02 10:30:20 +01:00
Michal Jarzabek
675c1e44f2 osd/ReplicatedPG:move C_OSD_OnApplied class to cc 
Signed-off-by: Michal Jarzabek <stiopa@gmail.com>
 2016-10-01 17:49:34 +01:00
Mykola Golub
a25b1d7cbe journal: ensure in-flight ops are complete destroying journaler 
Fixes: http://tracker.ceph.com/issues/17446
Signed-off-by: Jason Dillaman <dillaman@redhat.com>
 2016-10-01 11:24:43 +03:00
Mykola Golub
55762cca2b journal: complete action only after notification completed 
Signed-off-by: Mykola Golub <mgolub@mirantis.com>
 2016-10-01 11:24:43 +03:00
runsisi
5581510bdf librbd: fix rollback if failed to disable mirroring for image 
Signed-off-by: runsisi <runsisi@zte.com.cn>
 2016-10-01 13:29:23 +08:00
Samuel Just
27bdc8ce6d src/osd: relax the requirement that we scrub a whole hash value 
Previously, we needed to scrub all objects in clones in a single
hash value mainly to ensure that _scrub had access to all clones
of a single object at the same time.  Instead, just avoid letting
head or snapdir be a boundary (see the comment in the commit
for details).

Signed-off-by: Samuel Just <sjust@redhat.com>
 2016-09-30 16:43:08 -07:00
Yehuda Sadeh
6e9a915b56 rgw: don't fail if lost race when setting acls 
Fixes: http://tracker.ceph.com/issues/16930

When concurrently setting acls on object/bucket, we could lose in a race.
Instead of retry, just return success (same effect as if we won and then
other writer overwrote us).

Signed-off-by: Yehuda Sadeh <yehuda@redhat.com>
 2016-09-30 16:13:00 -07:00
Dan Mick
62d8e60cb6 Merge pull request #11285 from liewegas/wip-mgr-rpm 
ceph.spec: fix ceph-mgr version requirement

Reviewed-by: Dan Mick <dmick@redhat.com>
 2016-09-30 15:40:35 -07:00
Greg Farnum
42b15729f6 client: in UserPerm, correctly use delete[] and don't leak gids on copy 
gids is an array. We also need to clean it up when deep_copy()ing, which
requires initializing our member fields in the copy-constructor.

Signed-off-by: Greg Farnum <gfarnum@redhat.com>
 2016-09-30 15:19:17 -07:00
Sage Weil
3447601ac3 ceph.spec: fix ceph-mgr version requirement 
Signed-off-by: Sage Weil <sage@redhat.com>
 2016-09-30 18:13:50 -04:00
Yehuda Sadeh
3e34380553 rgw: set multipart copy source object as atomic, prefetch 
Object's head should only be read once.

Signed-off-by: Yehuda Sadeh <yehuda@redhat.com>
 2016-09-30 14:40:33 -07:00
John Spray
f4e53b9132 mds: include legacy client fsid in FSMap print 
Signed-off-by: John Spray <john.spray@redhat.com>
 2016-09-30 22:32:31 +01:00
Sage Weil
ad6aeb0f01 Merge pull request #11282 from liewegas/wip-bluestore-gc-fix 
os/bluestore: fix gc when blob extends past eof
 2016-09-30 16:17:57 -05:00
Sage Weil
570615b63a os/bluestore: fix gc when blob extends past eof 
This can happen if we aren't referencing the bob.  Don't try to
read what isn't logically referenced.

Signed-off-by: Sage Weil <sage@redhat.com>
 2016-09-30 17:06:47 -04:00
Sage Weil
2dc5775449 Merge pull request #11279 from liewegas/wip-bluestore-gc 
os/bluestore: compaction fixes
 2016-09-30 13:50:08 -05:00
Sage Weil
ea0aee6073 os/bluestore: fix compaction left-side length 
Signed-off-by: Sage Weil <sage@redhat.com>
 2016-09-30 14:49:35 -04:00
Sage Weil
1164a30235 os/bluestore: more concise SharedBlob print 
Signed-off-by: Sage Weil <sage@redhat.com>
 2016-09-30 14:32:31 -04:00
Sage Weil
506e4aedd4 os/bluestore: cosmetic gc cleanups 
Signed-off-by: Sage Weil <sage@redhat.com>
 2016-09-30 14:23:32 -04:00
Sage Weil
75bf4d18a2 Merge pull request #11271 from stiopaa1/osd_repicatedpg_addOverride 
osd/ReplicatedPG.h: add override to virtual funct

Reviewed-by: Sage Weil <sage@redhat.com>
 2016-09-30 12:57:02 -05:00
Sage Weil
ce9781ae8b Merge pull request #11239 from liewegas/wip-bluestore-max-blob 
os/bluestore: add max blob size; fix compressed min blob size logic

Reviewed-by: Somnath Roy <somnath.roy@sandisk.com>
 2016-09-30 12:47:58 -05:00
Sage Weil
beed433254 os/bluestore: set min compressed blob size floor at 2x min_alloc_size 
Otherwise there's not point in compressing--we won't be able to
allocation a smaller space to store it.

Signed-off-by: Sage Weil <sage@redhat.com>
 2016-09-30 13:47:23 -04:00
Sage Weil
87f6d23b2f os/bluestore: introduce bluestore_max_blob_size 
This is enforced for compressed and non-compressed blobs alike.
We keep the separate compressoin min and max tunables since these
may be adjusted on a per-pool basis as well.

Signed-off-by: Sage Weil <sage@redhat.com>
 2016-09-30 13:47:22 -04:00
Sage Weil
f32b682f13 os/bluestore: adjust compression blob sizes down 
These were way too big.  64KB is enough to get good compression,
and 4MB is way too big.

Signed-off-by: Sage Weil <sage@redhat.com>
 2016-09-30 13:47:22 -04:00
Sage Weil
a389a29e16 os/bluestore: fix blob depth encoding 
1) Fix the flag constnat
2) Redefine the meaning to be more intuitive.

Signed-off-by: Sage Weil <sage@redhat.com>
 2016-09-30 13:45:36 -04:00
Sage Weil
c72bcd667e Merge pull request #11232 from Roushan81/wip-bluestore-blob-gc 
os/bluestore: garbage collect partially overlapped blobs

Reviewed-by: Sage Weil <sage@redhat.com>
 2016-09-30 09:43:33 -05:00