RepoMirrors/ceph
1
0
Fork 0
mirror of https://github.com/ceph/ceph synced 2025-01-19 17:41:39 +00:00
Code Issues Packages Projects Releases Wiki Activity
45,744 Commits 216 Branches 0 Tags 1.8 GiB
40522828c1
Commit Graph

45744 Commits

This Branch
This Branch
All Branches
Author SHA1 Message Date
Josh Durgin
15f8ea00e4 Merge pull request #6394 from dillaman/wip-13560 
krbd: remove deprecated --quiet param from udevadm

Reviewed-by: Josh Durgin <jdurgin@redhat.com>
 2015-11-09 23:30:42 -08:00
Josh Durgin
de95d204bd Merge pull request #6508 from dillaman/wip-13726 
librbd: deadlock while attempting to flush AIO requests

Reviewed-by: Josh Durgin <jdurgin@redhat.com>
 2015-11-09 23:28:46 -08:00
Josh Durgin
220ed185f3 Merge pull request #6447 from dillaman/wip-13674 
librbd: resize should only update image size within header

Reviewed-by: Josh Durgin <jdurgin@redhat.com>
 2015-11-09 23:27:08 -08:00
xiexingguo
f7f5a08051 internal: remove unused local variables 
Remove unused variables bl and bl2 in snap_list and ictx_refresh.
Signed-off-by: xie xingguo <xie.xingguo@zte.com.cn>
 2015-11-10 07:29:54 +08:00
xiexingguo
c8fe5ae357 librados: cast oid to object explicitly before call ioctx methods 
Cast oid to object explicitly before call ioctx methods.
Signed-off-by: xie xingguo <xie.xingguo@zte.com.cn>
 2015-11-10 07:27:40 +08:00
Yehuda Sadeh
866d0a0892 Merge pull request #6513 from ceph/revert-6419-wip-openssl 
Revert 0374bb4a2f
 2015-11-09 15:26:12 -08:00
xiexingguo
e986ade817 IoCtxImpl: remove unused variable sName 
remove unused local variable sName in rollback.
Signed-off-by: xie xingguo <xie.xingguo@zte.com.cn>
 2015-11-10 07:26:03 +08:00
Yehuda Sadeh
a5651b867a Revert 0374bb4a2f 2015-11-09 15:24:29 -08:00
Yehuda Sadeh
0374bb4a2f Merge pull request #6419 from ceph/wip-openssl 
rgw: link against system openssl (instead of dlopen at runtime)

Reviewed-by: Yehuda Sadeh <yehuda@redhat.com>
 2015-11-09 13:56:29 -08:00
Jason Dillaman
d93a7951d7 Merge pull request #6453 from trociny/wip-2468 
librbd: flush and invalidate cache via admin socket

Reviewed-by: Jason Dillaman <dillaman@redhat.com>
 2015-11-09 16:36:58 -05:00
Sage Weil
62f122d4ab Merge pull request #6511 from liewegas/wip-fix-kv 
kv: fix bug in kv key optimization

Reviewed-by: Samuel Just <sjust@redhat.com>
Reviewed-by: Piotr Dałek <piotr.dalek@ts.fujitsu.com>
 2015-11-09 14:30:14 -05:00
Yehuda Sadeh
f5ac4a0c23 Merge branch 'wip-12955' 2015-11-09 11:03:40 -08:00
brian
7496741ebb rgw: fix modification to index attrs when setting acls 
Fixes: #12955

 - add ACL change after initial population
 - populate bufferlist with object attributes

Signed-off-by: Brian Felton <bjfelton@gmail.com>
 2015-11-09 11:03:02 -08:00
Sage Weil
9689fe0ae0 kv: fix string ctor usage 
When constructing a string from a char* there is only a single
length argument, no offset.  The 3 argument variant we were using
was converting to a std::string first (an prematurely terminating
on \0).

Signed-off-by: Sage Weil <sage@redhat.com>
 2015-11-09 13:40:41 -05:00
Yehuda Sadeh
754e37a952 Merge pull request #6047 from Sandy4999/wip-13140 
rgw: fix swift API returning incorrect account metadata

Reviewed-by: Yehuda Sadeh <yehuda@redhat.com>
 2015-11-09 10:38:46 -08:00
Josh Durgin
c26ec99ce5 Merge pull request #6504 from dillaman/wip-13727 
tests: fix typo in TestClsRbd.snapshots test case

Reviewed-by: Josh Durgin <jdurgin@redhat.com>
 2015-11-09 10:19:09 -08:00
Jason Dillaman
bfeb90e5fe librbd: fixed deadlock while attempting to flush AIO requests 
In-flight AIO requests might force a flush if a snapshot was created
out-of-band.  The flush completion was previously invoked asynchronously,
potentially via the same thread worker handling the AIO request. This
resulted in the flush operation deadlocking since it can't complete.

Fixes: #13726
Backport: infernalis, hammer
Signed-off-by: Jason Dillaman <dillaman@redhat.com>
 2015-11-09 11:22:24 -05:00
Jason Dillaman
a9729d9553 tests: new test case to catch deadlock on RBD image refresh 
Signed-off-by: Jason Dillaman <dillaman@redhat.com>
 2015-11-09 10:48:10 -05:00
Mykola Golub
d33842d237 tests: librbd: admin socket commands to flush and invalidate cache 
Signed-off-by: Mykola Golub <mgolub@mirantis.com>
 2015-11-09 16:14:11 +02:00
Mykola Golub
0996f9dfce librbd: flush and invalidate cache via admin socket 
Fixes: #2468
Signed-off-by: Mykola Golub <mgolub@mirantis.com>
 2015-11-09 16:14:11 +02:00
Mykola Golub
39503f5841 librbd: perf counter for cache invalidates 
Signed-off-by: Mykola Golub <mgolub@mirantis.com>
 2015-11-09 16:14:11 +02:00
Sage Weil
6ab35330e6 Merge pull request #6312 from liewegas/wip-kv 
kv: optimize and clean up internal key/value interface

Reviewed-by: Piotr Dałek <piotr.dalek@ts.fujitsu.com>
 2015-11-09 09:10:37 -05:00
Jason Dillaman
3b39226709 tests: fix typo in TestClsRbd.snapshots test case 
Fixes: #13727
Signed-off-by: Jason Dillaman <dillaman@redhat.com>
 2015-11-09 09:06:05 -05:00
Sage Weil
b5332742ad Merge pull request #6300 from XinzeChi/wip-throttle-max 
common: fix reset max in Throttle using perf reset command

Reviewed-by: Loic Dachary <ldachary@redhat.com>
 2015-11-09 09:04:25 -05:00
Sage Weil
d74de14f7c Merge pull request #6364 from dillaman/wip-13491 
osd: fix broken balance / localized read handling

Reviewed-by: Kefu Chai <kchai@redhat.com>
 2015-11-09 09:03:41 -05:00
Sage Weil
8ad594fb05 tracing: fix librados signed/unsigned warnings 
Signed-off-by: Sage Weil <sage@redhat.com>
 2015-11-09 09:03:23 -05:00
Piotr Dałek
057d39a107 os/osd: disable extra iterator validation 
In a number of loops using ObjectMapIterator, the iterator is validated
twice, first as an loop break condition, then during iter->next() call.
Suppress the validation in next() method in those cases for better
performance.

Signed-off-by: Piotr Dałek <piotr.dalek@ts.fujitsu.com>
 2015-11-09 09:03:23 -05:00
Piotr Dałek
117f40c690 os/KeyValueDB: don't call self.valid() from next() and prev() 
In most loops using ObjectMapIteratorImpl, there is a call to valid()
followed by a call to next(), which calls valid() again. Calling
valid() alone is particularly expensive, so don't do it twice if
possible.

Signed-off-by: Piotr Dałek <piotr.dalek@ts.fujitsu.com>
 2015-11-09 09:03:22 -05:00
Piotr Dałek
66b7b920cf mon/MonitorDBStore: use single-key LevelDB::get() method 
Make MonitorDBStore use single-key ::get() method for some performance
increase in heavy K/V workloads.

Signed-off-by: Piotr Dałek <piotr.dalek@ts.fujitsu.com>
 2015-11-09 09:03:22 -05:00
Piotr Dałek
a3f88918c9 os/DBObjectMap: use single-key LevelDB::get method 
Make FileStore use single-key ::get() method for some performance
increase in k/v-heavy workloads.

Signed-off-by: Piotr Dałek <piotr.dalek@ts.fujitsu.com>
 2015-11-09 09:03:22 -05:00
Piotr Dałek
76eb04aeef kv/LevelDBStore: simpler code for single-key fetches 
Often there's a need to fetch value for a specified, single key. Publish
a new method in LevelDB that takes key and prefix and puts just value at
specified address, without the need for interim keysets or resultsets.

Signed-off-by: Piotr Dałek <piotr.dalek@ts.fujitsu.com>
 2015-11-09 09:03:22 -05:00
Piotr Dałek
e184ca2339 os/LevelDBStore: faster LevelDBTransactionImpl::set 
This patch builds on Sage's idea to reduce bufferlist copying on
::set() calls. Initial patch reduced LevelDB's Slice generation
time from ~57ns to ~5ns in best-case scenario (bufferlist with
single bufferptr or contiguous bufferptrs), this patch reduces
slice generation time from ~57ns to ~11ns in worst case scenario
(under assumption that entire value length is at most 128KB),
leaving old code path for extremely-bad cases.

Signed-off-by: Piotr Dałek <piotr.dalek@ts.fujitsu.com>
 2015-11-09 09:03:22 -05:00
Sage Weil
48ceaaf0cc kv/RocksDBStore: do not Delete before Put 
A put implicitly overwrites the previous value.

Signed-off-by: Sage Weil <sage@redhat.com>
 2015-11-09 09:03:22 -05:00
Sage Weil
1e3c2fa21a kv/LevelDBStore: do not Delete before Put 
A put implicitly overwrites the previous value.

Signed-off-by: Sage Weil <sage@redhat.com>
 2015-11-09 09:03:22 -05:00
Sage Weil
338b4ed4be osd/ReplicatedPG: use bl-based setkeys/rmkeys 
FIXME: what about the tracepoint?

Signed-off-by: Sage Weil <sage@redhat.com>
 2015-11-09 09:03:21 -05:00
Kefu Chai
2d8bd53af7 Merge pull request #6502 from dachary/wip-13691-rados-test 
tests: fix test_rados_tools.sh rados lookup

Reviewed-by: Kefu Chai <kchai@redhat.com>
 2015-11-09 21:53:51 +08:00
Jason Dillaman
d167489a13 Merge pull request #6403 from xinxinsh/wip-13500 
optimize clone write path if object-map is enabled

Reviewed-by: Jason Dillaman <dillaman@redhat.com>
 2015-11-09 08:33:22 -05:00
Jason Dillaman
127905359f Merge pull request #6494 from trociny/wip-13720 
librbd: start perf counters after id is initialized

Reviewed-by: Jason Dillaman <dillaman@redhat.com>
 2015-11-09 08:20:33 -05:00
Loic Dachary
402d181ede tests: fix test_rados_tools.sh rados lookup 
http://tracker.ceph.com/issues/13691 Fixes: #13691

Signed-off-by: Loic Dachary <loic@dachary.org>
 2015-11-09 12:57:11 +01:00
Kefu Chai
cbe25efe39 Merge pull request #6464 from dachary/wip-13691-rados-test 
tests: concatenate test_rados_test_tool from src and qa

Reviewed-by: Kefu Chai <kchai@redhat.com>
 2015-11-09 19:31:00 +08:00
John Spray
8965a4943a Merge pull request #6497 from ukernel/wip-13714 
client: use null snapc to check pool permission

Reviewed-by: John Spray <john.spray@redhat.com>
 2015-11-09 10:34:26 +00:00
Yan, Zheng
fad3772fb7 client: use null snapc to check pool permission 
snap inodes' ->snaprealm can be NULL, so dereferencing it in
check_pool_perm() can cause segment fault. The pool permission
check does not write any data, so it's safe to use null snapc.

Fixes: #13714
Signed-off-by: Yan, Zheng <zyan@redhat.com>
 2015-11-09 11:51:12 +08:00
Mykola Golub
f33dd766de librbd: start perf counters after id is initialized 
So the perf counters section name has non-empty id.

Fixes: #13720
Signed-off-by: Mykola Golub <mgolub@mirantis.com>
 2015-11-07 21:57:27 +02:00
Yehuda Sadeh
a5265d51b2 Merge pull request #6228 from zhouyuan/fix_rgw_delete_exipre 
rgw: fix response of delete expired objects

Reviewed-by: Yehuda Sadeh <yehuda@redhat.com>
 2015-11-06 14:18:49 -08:00
Yehuda Sadeh
899d7754fe Merge pull request #6214 from jmunhoz/wip-bucket-request-payment 
rgw: bucket request payment support

Reviewed-by: Yehuda Sadeh <yehuda@redhat.com>
 2015-11-06 14:16:41 -08:00
Yehuda Sadeh
0507cff4eb Merge pull request #5969 from rzarzynski/wip-conf-md-add 
rgw: extend rgw_extended_http_attrs to affect Swift accounts and containers as well

Reviewed-by: Yehuda Sadeh <yehuda@redhat.com>
 2015-11-06 13:58:05 -08:00
Yehuda Sadeh
a030e56469 Merge pull request #5938 from zhuangzeqiang/rgw-cerr 
rgw:  clarify the error message when trying to create an existed user

Reviewed-by: Yehuda Sadeh <yehuda@redhat.com>
 2015-11-06 13:48:43 -08:00
Sage Weil
24380c8ea3 Merge pull request #6479 from ceph/wip-systemd-rbdmap 
rbdmap: systemd support

Reviewed-by: Kefu Chai <kchai@redhat.com>
 2015-11-06 16:24:26 -05:00
Jason Dillaman
b55a41b579 Merge pull request #6467 from fullerdj/wip-13697 
cls_rbd: change object_map_update to return 0 on success, add logging

Reviewed-by: Jason Dillaman <dillaman@redhat.com>
 2015-11-06 14:39:58 -05:00
Yehuda Sadeh
bbd5584b22 Merge pull request #6059 from dwj192/dwj-master-tip 
rgw:mdlog trim add usage prompt

Reviewed-by: Yehuda Sadeh <yehuda@redhat.com>
 2015-11-06 11:33:40 -08:00