Commit Graph

70790 Commits

Author SHA1 Message Date
Casey Bodley
d346b1b835 Merge pull request #12175 from dongbula/fix-slavezonegroup-cannot-enable-bucket-versioning
rgw: fix slave zonegroup cannot enable the bucket versioning

Reviewed-by: Casey Bodley <cbodley@redhat.com>
Reviewed-by: Orit Wasserman <owasserm@redhat.com>
2017-04-06 10:41:18 -04:00
Daniel Gryniewicz
6a43cddce2 Merge pull request #14342 from theanalyst/cleanup/rgw/rgw-lc-inc
rgw_lc: drop a bunch of unused headers
2017-04-06 10:22:00 -04:00
Casey Bodley
4e793bfd2b Merge pull request #13622 from zhangsw/feature-lifecycle-multipart
rgw: add support for multipart upload expiration.

Reviewed-by: Daniel Gryniewicz <dang@redhat.com>
2017-04-06 10:02:29 -04:00
Nathan Cutler
3c89ca8fb6 Merge pull request #14337 from smithfarm/wip-19492
common: support s390 and unknown architectures in spin-wait loop

Reviewed-by: Kefu Chai <kchai@redhat.com>
Reviewed-by: Brad Hubbard <bhubbard@redhat.com>
2017-04-06 15:25:53 +02:00
Matt Benjamin
69fe2f2718 Merge pull request #14339 from haklein/master
add RGW ldap auth documentation
2017-04-06 09:14:22 -04:00
shiqi
08c9dad46e mailmap: shiqi affiliation
Signed-off-by: shiqi <1454927420@qq.com>
2017-04-06 20:45:45 +08:00
Mohamad Gebai
22875e50cb cmake: build tracepoint libraries for vstart target
Avoid a segfault in TracepointProvider::verify_config() when building
for the 'vstart' target. TracepointProvider tries to load the tracepoint
providers libraries at runtime, but these aren't built as part of the
'vstart' target.

Signed-off-by: Mohamad Gebai <mgebai@suse.com>
2017-04-06 08:34:28 -04:00
Wei Jin
3f514bda1c mgr: clean nonused function declaration
Signed-off-by: Wei Jin <wjin.cn@gmail.com>
2017-04-06 19:11:17 +08:00
Kefu Chai
db9e5312e8 mgr: move "osd pool stats" to mgr
Signed-off-by: Kefu Chai <kchai@redhat.com>
2017-04-06 19:01:58 +08:00
tone-zhang
e96db21307 rgw: fix response header of Swift API
Response header of Swift API returned by radosgw does not contain
"x-openstack-request-id", but Swift returns it. Enhance the
compatibility of radosgw.

Fixes: http://tracker.ceph.com/issues/19443

Signed-off-by: tone-zhang <tone.zhang@linaro.org>
2017-04-06 17:56:05 +08:00
Nathan Cutler
2807d28c1a common: support s390 and unknown architectures in spin-wait loop
This fixes the s390 build which was broken by 34481768cb

It also reinstates the previous architecture-independent for-loop
implementation as a catch-all for "unknown" (not-explicitly-supported)
architectures.

Fixes: http://tracker.ceph.com/issues/19492
Signed-off-by: Nathan Cutler <ncutler@suse.com>
2017-04-06 10:41:37 +02:00
Mykola Golub
af80ee62d6 Merge pull request #14240 from dillaman/wip-rbd-mirror-peer-uuids
rbd-mirror: pool watcher should track mirror uuid

Reviewed-by: Mykola Golub <mgolub@mirantis.com>
2017-04-06 11:02:26 +03:00
Loic Dachary
81527f237f Merge pull request #14293 from badone/wip-ceph-disk-list-selinux
ceph-disk: Populate mount options when running "list"

Reviewed-by: Loic Dachary <ldachary@redhat.com>
2017-04-06 08:28:20 +02:00
Dan Mick
3681683df7 cmake: Add simple recursive ctags target for Ceph source only
Excludes submodules and random JavaScript.

Signed-off-by: Dan Mick <dan.mick@redhat.com>
Signed-off-by: Kefu Chai <kchai@redhat.com>
2017-04-06 13:27:12 +08:00
Jos Collin
ec04593a63 .organizationmap: Added my Name and Email ID
Signed-off-by: Jos Collin <jcollin@redhat.com>
2017-04-06 09:36:14 +05:30
liuchang0812
e9289bb590 rgw: initialize is_truncated to false explicitly for function rgw_read_user_buckets
Signed-off-by: liuchang0812 <liuchang0812@gmail.com>
2017-04-06 10:15:13 +08:00
Jason Dillaman
d3663113f8 Merge pull request #14032 from joscollin/wip-18511-returntype
librbd: changed the return type of ImageRequestWQ::discard()

Reviewed-by: Jason Dillaman <dillaman@redhat.com>
2017-04-05 20:12:42 -04:00
Jason Dillaman
184df4abcf Merge pull request #13957 from wangzhengyong/import
rbd: error out if import image format failed

Reviewed-by: Jason Dillaman <dillaman@redhat.com>
2017-04-05 20:11:33 -04:00
Jason Dillaman
d271e8c279 Merge pull request #13836 from gaurav36/bug-14535-rbd-snap-pool-name-validatation
rbd: validate pool and snap name optionals

Reviewed-by: Jason Dillaman <dillaman@redhat.com>
2017-04-05 20:10:20 -04:00
Yuri Weinstein
d302deb504 Merge pull request #14282 from wjwithagen/wip-wjw-freebsd-ceph-fuse
build: let FreeBSD build ceph-fuse

Reviewed-by: Kefu Chai <kchai@redhat.com>
2017-04-05 15:02:41 -07:00
Yuri Weinstein
618767b12a Merge pull request #14308 from badone/SyntheticClient_random_walk_warning
client/SyntheticClient.cc: Fix warning in random_walk

Reviewed-by: Sage Weil <sage@redhat.com>
2017-04-05 15:00:13 -07:00
Yuri Weinstein
04c5374d5e Merge pull request #10146 from yonghengdexin735/wip-zzz-filestore-write
os/filestore: handle error returned from write_fd()

Reviewed-by: Kefu Chai <kchai@redhat.com>
Reviewed-by: Sage Weil <sage@redhat.com>
2017-04-05 14:58:28 -07:00
Willem Jan Withagen
e8f03e205d core: Prefix /proc/ with FreeBSD emulation
Signed-off-by: Willem Jan Withagen <wjw@digiware.nl>
2017-04-05 21:36:22 +02:00
Casey Bodley
7f2871fe59 rgw: allow larger payload for period commit
testing with 3 zonegroups and 3 zones each, the period json grew larger
than 4k and caused decode failures on period commit

updated to use the new config variable rgw_max_put_param_size

Fixes: http://tracker.ceph.com/issues/19505

Signed-off-by: Casey Bodley <cbodley@redhat.com>
2017-04-05 14:32:43 -04:00
Casey Bodley
39349a4fe9 Merge pull request #13815 from theanalyst/cleanup/rgw/rest-read-input
rgw: don't read all user input for a few param requests

Reviewed-by: Casey Bodley <cbodley@redhat.com>
2017-04-05 14:30:14 -04:00
Sage Weil
2faa01bf86 mon/MgrMonitor: make debug prefix useful
Signed-off-by: Sage Weil <sage@redhat.com>
2017-04-05 11:58:57 -04:00
Radoslaw Zarzynski
796c34de91 os/bluestore: don't use virtual for accessing BitMapZone::is_exhausted.
Signed-off-by: Radoslaw Zarzynski <rzarzynski@mirantis.com>
2017-04-05 17:38:30 +02:00
Sage Weil
4652f34c5f os/bluestore: do not wake kv thread if only deferred events pending
No need to wake up if there is only deferred work; we can do
it lazily.

Signed-off-by: Sage Weil <sage@redhat.com>
2017-04-05 10:26:56 -04:00
Sage Weil
36dd549893 os/bluestore: bluestore_deferred_batch_ops = 32 (not 8)
A bit more batching.

Signed-off-by: Sage Weil <sage@redhat.com>
2017-04-05 10:26:56 -04:00
Sage Weil
b4b800f3e2 os/bluestore: unify throttling model
Implement a super simple model for the cost of a transaction
for the purposes of throttling.  This replaces two independent
throttles (one for transaction ops, one for bytes) and puts
them both under the 'bytes' throttle.  The txc model cost is
expressed in terms of bytes as this is probably the simplest
thing for users to reason about.

Signed-off-by: Sage Weil <sage@redhat.com>
2017-04-05 10:26:56 -04:00
Sage Weil
5075b27f00 os/bluestore: release throttle before commit
Signed-off-by: Sage Weil <sage@redhat.com>
2017-04-05 10:26:56 -04:00
Sage Weil
0bd142b42c ceph_test_objectstore: error via derr too
This helps establish timing relative to the log.

Signed-off-by: Sage Weil <sage@redhat.com>
2017-04-05 10:23:38 -04:00
Sage Weil
0541ee1e6c os/bluestore: increase aio queue depth
We've seen some reports of the aio EAGAIN backoff in logs; use a larger
aio ring buffer to reduce the risk of this.

Signed-off-by: Sage Weil <sage@redhat.com>
2017-04-05 10:23:38 -04:00
Sage Weil
7499a87632 os/bluestore: reduce some debug level
Signed-off-by: Sage Weil <sage@redhat.com>
2017-04-05 10:23:38 -04:00
Sage Weil
554972ded2 os/bluestore: put fsck structures in mempools
This lets us track them individually in debug mode.

Signed-off-by: Sage Weil <sage@redhat.com>
2017-04-05 10:23:38 -04:00
Sage Weil
6af489121b mempool: do not skip first 2 mempools
This broke when we sorted the pool definitions in the header.

Signed-off-by: Sage Weil <sage@redhat.com>
2017-04-05 10:23:38 -04:00
Sage Weil
066514c37c os/bluestore: flush osr even on failure in _deferred_replay
This is one less thing for fsck to worry about.

Signed-off-by: Sage Weil <sage@redhat.com>
2017-04-05 10:23:38 -04:00
Sage Weil
c67e9e2530 os/bluestore: trim cache during fsck
Signed-off-by: Sage Weil <sage@redhat.com>
2017-04-05 10:23:38 -04:00
Sage Weil
f7a51d35a9 os/bluestore: be noisy about bitmap init (it's slow!)
Signed-off-by: Sage Weil <sage@redhat.com>
2017-04-05 10:23:38 -04:00
Sage Weil
5f2db7f83b Merge pull request #14316 from liewegas/wip-bluestore-fsck-unused
os/bluestore: fsck: verify blob.unused field

Reviewed-by: Igor Fedotov <ifedotov@mirantis.com>
2017-04-05 09:22:51 -05:00
Sage Weil
703213b1f0 os/bluestore: use typedef for blob unused (uint16_t)
Signed-off-by: Sage Weil <sage@redhat.com>
2017-04-05 10:18:17 -04:00
Sage Weil
f49b1d24c6 os/bluestore: fsck: verify blob.unused field
Two checks:

- verify that no logical extents reference the portion of the
  blob marked unused, and
- verify that the csum (if present) is zero for any unused
  region.

Signed-off-by: Sage Weil <sage@redhat.com>
2017-04-05 10:18:17 -04:00
Sage Weil
4759c705e9 Merge pull request #14157 from liewegas/wip-19379
os/bluestore: fix deferred writes vs collection split race

Reviewed-by: Igor Fedotov <ifedotov@mirantis.com>
2017-04-05 08:56:16 -05:00
Sage Weil
c618d22d25 Merge pull request #14258 from liewegas/wip-bluestore-tool
ceph-bluestore-tool: rename from bluefs-tool; improve usage
2017-04-05 08:55:36 -05:00
Sage Weil
d4525121c9 Merge pull request #14294 from tchaikov/wip-check-last-osd-report-if-luminous
mon/OSDMonitor: check last_osd_report only when the whole cluster is lu…

Reviewed-by: Sage Weil <sage@redhat.com>
2017-04-05 08:55:23 -05:00
Sage Weil
6f8cd66b94 Merge pull request #14303 from liewegas/wip-more-mgr-commands
mgr: move 'osd perf' and 'osd blocked-by' to mgr

Reviewed-by: Kefu Chai <kchai@redhat.com>
2017-04-05 08:55:03 -05:00
Casey Bodley
aad7e6e2da Merge pull request #14010 from zhangsw/fix-rgw-fetchremote-bug
rgw: version id doesn't work in fetch_remote_obj

Reviewed-by: Casey Bodley <cbodley@redhat.com>
2017-04-05 09:38:41 -04:00
Yan, Zheng
424e0c6744 mds: make C_MDSInternalNoop::complete() delete 'this'
Fixes: http://tracker.ceph.com/issues/19501
Signed-off-by: "Yan, Zheng" <zyan@redhat.com>
2017-04-05 21:29:10 +08:00
liuchang0812
7a3fbbfac6 rgw: use is_truncated to judge whether get all bucket already
That use is_truncated is simple and avoid nessesary list. We have to iterate 11
times initially if there is 100 buckets and iter_max_bucket is 10. We only need
10 times now.

Signed-off-by: liuchang0812 <liuchang0812@gmail.com>
2017-04-05 20:06:11 +08:00
liuchang0812
00cdfce27c rgw/rgw_bucket: update is_truncated option in rgw_read_user_buckets function
The function called `rgw_read_user_buckets` has one option `bool *is_truncated`
which should be updated, and we could use `is_truncated` to avoid unnessesary
iteration.

Fixes: http://tracker.ceph.com/issues/19365

Signed-off-by: liuchang0812 <liuchang0812@gmail.com>
2017-04-05 19:44:47 +08:00