RepoMirrors/ceph
1
0
Fork 0
mirror of https://github.com/ceph/ceph synced 2025-03-11 02:39:05 +00:00
Code Issues Packages Projects Releases Wiki Activity
58,735 Commits 219 Branches 396 Tags 1.4 GiB
700e9c406a
Commit Graph

58735 Commits

This Branch
This Branch
All Branches
Author SHA1 Message Date
Yehuda Sadeh
c16929f6bc Merge pull request #11325 from yehudasa/wip-17499 
rgw: need to 'open_object_section' before dump stats in 'RGWGetUsage_…

Reviewed-by: Yehuda Sadeh <yehuda@redhat.com>
 2016-10-04 17:02:51 -07:00
weiqiaomiao
a8d33c081b rgw: need to 'open_object_section' before dump stats in 'RGWGetUsage_ObjStore_S3::send_response()' function 
Fixes: http://tracker.ceph.com/issues/17499

Signed-off-by: weiqiaomiao <wei.qiaomiao@zte.com.cn>
 2016-10-04 17:06:36 -07:00
Yehuda Sadeh
92c2c9bc87 Merge pull request #10820 from yanghonggang/master 
rgw: fix get object instance  returned NoSuchKey error

Reviewed-by: Yehuda Sadeh <yehuda@redhat.com>
 2016-10-04 16:20:51 -07:00
Yehuda Sadeh
91a70c3e51 Merge pull request #10726 from weiqiaomiao/wqm-wip-remove-ret 
rgw: remove the field ret from class RGWPutLC

Reviewed-by: Yehuda Sadeh <yehuda@redhat.com>
 2016-10-04 15:48:05 -07:00
Sage Weil
02f457177a os/bluestore: make BufferSpace smaller 
Saves 24 bytes!

BufferSpace  104 -> 80 bytes
SharedBlob   216 -> 192 bytes

Signed-off-by: Sage Weil <sage@redhat.com>
 2016-10-04 18:14:16 -04:00
Sage Weil
7c5b77a6f8 os/bluestore: pack SharedBlob 
224 -> 216 bytes

Signed-off-by: Sage Weil <sage@redhat.com>
 2016-10-04 18:04:49 -04:00
Sage Weil
2094168a53 os/bluestore: pack Blob members 
272 -> 264 bytes

Signed-off-by: Sage Weil <sage@redhat.com>
 2016-10-04 18:03:06 -04:00
Sage Weil
cd9d61456f unittest_bluestore_types: a few more types for sizeof 
Signed-off-by: Sage Weil <sage@redhat.com>
 2016-10-04 17:38:38 -04:00
Yehuda Sadeh
a8fe5554b9 Merge pull request #6501 from dwj192/check-index-multipart 
rgw:bucket check remove _multipart_ prefix

Reviewed-by: Yehuda Sadeh <yehuda@redhat.com>
 2016-10-04 13:49:05 -07:00
Yehuda Sadeh
69b11838fe Merge pull request #10558 from cbodley/wip-rgw-mdlog-history 
rgw: store oldest mdlog period in rados

Reviewed-by: Yehuda Sadeh <yehuda@redhat.com>
 2016-10-04 13:10:29 -07:00
Mark Nelson
ed4591c595 os/bluestore: optimize intrusive sets for size. 
Signed-off-by: Mark Nelson <mnelson@redhat.com>
 2016-10-04 15:04:36 -05:00
Sage Weil
f688302df7 os/bluestore: fix has_aios -> has_pending_aios 
This is only called from safe callers (with exclusive access),
so we don't need a lock.

It also should only return true for *pending* aios--not aios
we've already submitted!

Signed-off-by: Sage Weil <sage@redhat.com>
 2016-10-04 15:28:49 -04:00
Sage Weil
3ed80d975f os/bluestore/KernelDevice: fix aio_submit race 
Behave if pending is empty.

Signed-off-by: Sage Weil <sage@redhat.com>
 2016-10-04 15:28:49 -04:00
Sage Weil
16e6e349ef Merge pull request #11312 from liewegas/wip-bluestore-assign-blob 
os/bluestore: fix crash in decode_some()
 2016-10-04 13:10:10 -05:00
Ilya Dryomov
e857b78965 rbd: recognize lock_on_read option 
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
 2016-10-04 19:50:32 +02:00
Yehuda Sadeh
b974dac04e Merge pull request #10372 from cbodley/wip-rgw-data-log-trim 
rgw multisite: trim data logs as peer zones catch up

Reviewed-by: Yehuda Sadeh <yehuda@redhat.com>
 2016-10-04 10:20:31 -07:00
Sage Weil
6b21e69a42 os/bluestore: fix crash in decode_some() 
shared_blob must be non-null when we call Extent::assign_blob().

Signed-off-by: Sage Weil <sage@redhat.com>
 2016-10-04 12:52:27 -04:00
Sage Weil
96707194c7 Merge pull request #11295 from liewegas/wip-bluestore-cache 
os/bluestore: make cache settings process-wide

Reviewed-by: Igor Fedotov <ifedotov@mirantis.com>
 2016-10-04 10:39:50 -05:00
Jason Dillaman
3e4a59d81b Merge pull request #11291 from trociny/wip-qa-mirror-asok 
qa/workunits/rbd: check status also in pool dir after asok commands

Reviewed-by: Jason Dillaman <dillaman@redhat.com>
 2016-10-04 11:22:53 -04:00
Casey Bodley
8c2dc6eacf Merge pull request #11278 from gaurav36/wip-obsolete-rgw-admin-period-prepare 
rgw: obsolete 'radosgw-admin period prepare' command

Reviewed-by: Casey Bodley <cbodley@redhat.com>
 2016-10-04 11:14:50 -04:00
Sage Weil
b16bd7781a unittest_bluestore_types: show sizeof for key types 
Signed-off-by: Sage Weil <sage@redhat.com>
 2016-10-04 11:14:08 -04:00
John Spray
fcf61342c6 Merge pull request #11307 from jcsp/wip-mgr-compile 
mgr: squash compiler warnings

Reviewed-by: Willem Jan Withagen <wjw@digiware.nl>
 2016-10-04 16:03:16 +01:00
Sage Weil
304a3b1bf8 os/bluestore: use Extent::assign_blob 
More explicit, less error-prone.

Signed-off-by: Sage Weil <sage@redhat.com>
 2016-10-04 10:59:15 -04:00
John Spray
7a93aa874e Merge pull request #11299 from liewegas/wip-mgr-fix 
mgr: implement con reset handling

Reviewed-by: John Spray <john.spray@redhat.com>
 2016-10-04 15:44:39 +01:00
John Spray
d5c6a961d3 Merge pull request #11308 from liewegas/wip-mgr-valgrind 
mgr/MgrMap: initialize all fields

Reviewed-by: John Spray <john.spray@redhat.com>
 2016-10-04 15:05:31 +01:00
Sage Weil
f48690ebbe os/bluestore: only split blobs on min_alloc_size boundaries 
Otherwise we can throw off the allocator.

Signed-off-by: Sage Weil <sage@redhat.com>
 2016-10-04 10:01:07 -04:00
Gaurav Kumar Garg
1508c70d36 rgw multisite: obsolete 'radosgw-admin period prepare' command 
All that the 'radosgw-admin period prepare' command does is create an
empty period and write it to rados. When we added the staging period
(see 'period get --staging'), uncommitted changes were written to that
instead - so the 'period prepare' command became obsolete. It should be
removed!

Fixes: http://tracker.ceph.com/issues/17387

Reported-by: Casey Bodley <cbodley@redhat.com>
Signed-off-by: Gaurav Kumar Garg <garg.gaurav52@gmail.com>
 2016-10-04 15:58:09 +02:00
Sage Weil
ed450d681b os/bluestore: prune deallocated blob tails 
This saves us space: no invalid extent, and no unused csum_data.

Signed-off-by: Sage Weil <sage@redhat.com>
 2016-10-04 09:37:00 -04:00
Sage Weil
1a39872b9e os/bluestore: try to split blobs instead of spanning them 
If a blob crosses a shard boundary, try to split it into pieces
instead of making it spanning.  This is only possible under
certain conditions (uncompressed, no in-flight writes, no unused
space, shard boundary aligned with chunk boundary).

Generally speaking this will reduce the number of blobs that end
up spanning and reduce the onode key size.

Signed-off-by: Sage Weil <sage@redhat.com>
 2016-10-04 09:36:59 -04:00
Sage Weil
5b3e71621f mgr/MgrMap: initialize all fields 
Fixes: http://tracker.ceph.com/issues/17492
Signed-off-by: Sage Weil <sage@redhat.com>
 2016-10-04 09:23:27 -04:00
John Spray
a22bb9cb75 mgr: squash compiler warnings 
Signed-off-by: John Spray <john.spray@redhat.com>
 2016-10-04 14:12:59 +01:00
John Spray
0aab9e2737 Merge pull request #11254 from xiexingguo/xxg-wip-fs-0929 
mds/MDCache: kill a comipler warning

Reviewed-by: Greg Farnum <gfarnum@redhat.com>
 2016-10-04 14:12:41 +01:00
Jason Dillaman
baa2202ea2 Merge pull request #11306 from trociny/fixup-notify_image_updated 
librbd: memory leak in MirroringWatcher::notify_image_updated

Reviewed-by: Jason Dillaman <dillaman@redhat.com>
 2016-10-04 09:05:28 -04:00
Jason Dillaman
3b021d662e Merge pull request #11303 from ceph/wip-rbd-clarify-size 
doc: clarify rbd size units

Reviewed-by: Jason Dillaman <dillaman@redhat.com>
 2016-10-04 09:00:22 -04:00
Mykola Golub
91d7de44d9 librbd: memory leak in MirroringWatcher::notify_image_updated 
Signed-off-by: Mykola Golub <mgolub@mirantis.com>
 2016-10-04 15:54:09 +03:00
Jason Dillaman
476820bf18 Merge pull request #11257 from trociny/wip-17355-fixup 
journal: ensure in-flight ops are complete destroying journaler

Reviewed-by: Jason Dillaman <dillaman@redhat.com>
 2016-10-04 08:48:05 -04:00
Boris Ranto
2d1f164613 Merge pull request #11259 from b-ranto/wip-selinux-tmp-files 
selinux: Allow ceph to manage tmp files

Reviewed-by: Ken Dreyer <kdreyer@redhat.com>
 2016-10-04 13:30:18 +02:00
Mykola Golub
6dda2b9384 rbd-mirror HA: move librbd::image_watcher::Notifier to librbd::object_watcher 
Fixes: http://tracker.ceph.com/issues/17017
Signed-off-by: Mykola Golub <mgolub@mirantis.com>
 2016-10-04 13:51:46 +03:00
John Spray
f1fb667dfd messages: fix MMDSBeacon constructor 
This was leaving garbage in some fields during decode
when decoding messages with an older version.

Fixes: http://tracker.ceph.com/issues/17466
Signed-off-by: John Spray <john.spray@redhat.com>
 2016-10-04 11:57:05 +02:00
Ilya Dryomov
d38dc29ed5 doc: clarify rbd size units 
It wasn't clear from the man page that --size defaults to M.

Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
 2016-10-04 10:21:36 +02:00
Alexey Sheplyakov
5c291855ac doc: fix librados example programs 
basically s/rados_wait_for_(safe|complete)/rados_aio_wait_for_\1/g

Signed-off-by: Alexey Sheplyakov <asheplyakov@mirantis.com>
 2016-10-04 09:39:51 +03:00
Gregory Farnum
110238608a Merge pull request #11218 from ceph/wip-getuid 
Client: pass "UserPerm" struct everywhere for security checks

Reviewed-by: Jeff Layton <jlayton@redhat.com>
 2016-10-03 22:30:55 -07:00
David Zafman
6d410e9723 osd_types,PG: force the map used in _scrub to be sorted correctly 
Previously, ScrubMap::objects was always sorted bitwise (nibblewise
before the comparator change was made.  It didn't matter because we
always scrubbed whole hash values.  Now, we need the objects in the
objectstore ordering because we may be missing objects at the tail of
the scanned range and need them to show up at the tail of the
ScrubMap::objects mapping.  We don't need to do anything else to handle
the upgrade process since the actual objects *in* the map were
determined by the objectstore ordering.

Signed-off-by: Samuel Just <sjust@redhat.com>
 2016-10-03 21:05:37 -07:00
Yang Honggang
d43b69e529 rgw: fix versioned object IO error 
When accessing a copied destination object, its source object's instance ID
    information is needed, however it's missing now in the destination object's
    manifest.

    In order to fix this problem, we can record source object's version_id/instance
    into dest object's manifest(a new filed 'tail_instance' is added). When creating
    a new object(not copy), 'tail_instance' should be equal to its instance value.
    When copy/get a object, 'tail_instance' should always be used to get the right
    tail objects.

    Fixes: http://tracker.ceph.com/issues/17111
    Signed-off-by: Yang Honggang <joseph.yang@xtaotech.com>
 2016-10-04 09:18:09 +08:00
Sage Weil
0118293a4e os/bluestore: turn bluestore_onode_cache_size 16k -> 4k 
Signed-off-by: Sage Weil <sage@redhat.com>
 2016-10-03 18:42:03 -04:00
Sage Weil
05bb253c94 os/bluestore: drop unnecessary assert 
Signed-off-by: Sage Weil <sage@redhat.com>
 2016-10-03 18:42:03 -04:00
Sage Weil
a706908340 os/bluestore: instrument cache blob and extent counts 
Signed-off-by: Sage Weil <sage@redhat.com>
 2016-10-03 18:42:03 -04:00
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