Commit Graph

43540 Commits

Author SHA1 Message Date
David Disseldorp
b98b104829 rbdnamer: drop unneccessary tr usage
tr is not present by default in Dracut generated initramfs images. The
same logic can anyhow be performed via the existing sed call.

Signed-off-by: David Disseldorp <ddiss@suse.de>
2015-08-07 19:41:14 +02:00
Sage Weil
144792efe0 Merge pull request #5310 from ceph/wip-pipefd-leak
Wip pipefd leak

Reviewed-by: Jianpeng Ma <jianpeng.ma@intel.com>
2015-08-07 12:59:39 -04:00
Samuel Just
180a500a7f Merge pull request #5511 from ceph/wip-cxx11
c++11 ftw

Reviewed-by: Samuel Just <sjust@redhat.com>
Reviewed-by: Sage Weil <sage@redhat.com>
2015-08-07 08:56:38 -07:00
Casey Bodley
3260adf146 c++11: xlist needs a value_type for back_inserter
Signed-off-by: Casey Bodley <cbodley@redhat.com>
2015-08-07 11:42:59 -04:00
Casey Bodley
511106e353 c++11: fix shared_ptr conversions to bool
operator bool() is marked as explicit, which prevents it from being used
directly in gtest macros

Signed-off-by: Casey Bodley <cbodley@redhat.com>
2015-08-07 11:42:58 -04:00
Casey Bodley
3a6808049a c++11: remove hash_namespace.h and default to std
Signed-off-by: Casey Bodley <cbodley@redhat.com>
2015-08-07 11:42:58 -04:00
Casey Bodley
f4d2032ce1 c++11: disambiguate std::isnan
c++11's isnan() is a constexpr function, while c's is a macro

Signed-off-by: Casey Bodley <cbodley@redhat.com>
2015-08-07 11:42:58 -04:00
Casey Bodley
11a5d29299 c++11: fixes for preprocessor string concatenation
c++11 adds user-defined literals through type suffixes. for example,
"foo"s is a literal of type std::basic_string. this means that any
non-whitespace characters following a string literal will be interpreted
as a suffix. this causes issues with string concatenation done by the
preprocessor, which can only be resolved by adding whitespace

Signed-off-by: Casey Bodley <cbodley@redhat.com>
2015-08-07 11:42:58 -04:00
Casey Bodley
f7c478c6d9 c++11: replace boost::assign with initializer list
Signed-off-by: Casey Bodley <cbodley@redhat.com>
2015-08-07 11:42:46 -04:00
Samuel Just
ec619c7d86 Merge pull request #4832 from ceph/wip-11511
osd: Even in objects_read_sync() case don't read past oi.size

Reviewed-by: Samuel Just <sjust@redhat.com>
2015-08-07 08:38:48 -07:00
Haomai Wang
9cdc3278f4 AsyncConnection: Fix wrong order of local message delivering
Signed-off-by: Haomai Wang <haomaiwang@gmail.com>
2015-08-07 23:13:30 +08:00
Sage Weil
7811c776c5 Merge pull request #5331 from liewegas/wip-collection-list
os: clean up collection_list

Reviewed-by: Samuel Just <sjust@redhat.com>
2015-08-07 10:15:10 -04:00
Sage Weil
1beab53c08 Merge pull request #5271 from wonzhq/cache-rate
mon: add cache tier IO rate in the 'osd pool stats' command

Reviewed-by: Sage Weil <sage@redhat.com>
2015-08-07 09:35:27 -04:00
Sage Weil
2a7605380c Merge pull request #5507 from athanatos/wip-12436
blkdev.cc::get_device_by_uuid: do not leak cache

Reviewed-by: Sage Weil <sage@redhat.com>
2015-08-07 08:55:57 -04:00
Sage Weil
01d3aa7ce2 Merge pull request #5506 from athanatos/wip-12338
OSD: break connection->session->waiting message->connection cycle

Reviewed-by: Sage Weil <sage@redhat.com>
2015-08-07 08:55:25 -04:00
Sage Weil
654c16ed2e os/ObjectStore: drop get_ideal_list_min()
Unused.  Fix comment on get_ideal_list_max while we are here.

Signed-off-by: Sage Weil <sage@redhat.com>
2015-08-07 08:45:34 -04:00
Sage Weil
cdb1fc011a os: drop snapid_t arg to collection_list
All this does is filter out objects with old snapids, but this can be
just as easily done in the one caller that specifies a non-zero value
(PGLS and PGNLS), which in fact already do that filtering.

Signed-off-by: Sage Weil <sage@redhat.com>
2015-08-07 08:45:34 -04:00
Sage Weil
e46a855e5f s/collection_list_impl/collection_list/
Signed-off-by: Sage Weil <sage@redhat.com>
2015-08-07 08:45:34 -04:00
Xiaoxi Chen
7ee3eeffa6 Kill collection_list in CollectionIndex
Now we use CollectionIndex.collection_list_partial.

Signed-off-by: Xiaoxi Chen <xiaoxi.chen@intel.com>
2015-08-07 08:45:34 -04:00
Xiaoxi Chen
5df688a5df Kill Flat_index.
Too old, remove it entirely.

Signed-off-by: Xiaoxi Chen <xiaoxi.chen@intel.com>
2015-08-07 08:45:33 -04:00
Xiaoxi Chen
c5dc4047e2 Kill collection_list
use collection_list_impl as well.

Signed-off-by: Xiaoxi Chen <xiaoxi.chen@intel.com>
2015-08-07 08:45:32 -04:00
Xiaoxi Chen
2d5ed30f02 Kill collection_list_partial
Use collection_list_impl directly.

Signed-off-by: Xiaoxi Chen <xiaoxi.chen@intel.com>
2015-08-07 08:45:14 -04:00
Xiaoxi Chen
716551090e Kill collection_list_range
We can directly use collection_list_impl.

Signed-off-by: Xiaoxi Chen <xiaoxi.chen@intel.com>
2015-08-07 08:45:14 -04:00
Xiaoxi Chen
7bf999cfd2 os/KeyValuestore:Refactor collection_list_range and collection_list_partial
Add collection_list_impl which abstract the common process
of collection list behavior.

Signed-off-by: Xiaoxi Chen <xiaoxi.chen@intel.com>
2015-08-07 08:45:14 -04:00
Sage Weil
26668d6f9b os/CollectionIndex: use const ref for end
Signed-off-by: Sage Weil <sage@redhat.com>
2015-08-07 08:45:09 -04:00
Samuel Just
80b6961d3a Merge pull request #5509 from ceph/wip-12536
common/hobject_t: correctly decode pre-infernalis hobject_t min

Reviewed-by: Samuel Just <sjust@redhat.com>
2015-08-07 05:44:49 -07:00
Xiaoxi Chen
921c4586f1 os/Filestore:Refactor collection_list_range and collection_list_partial
Add collection_list_impl which abstract the common process
of coleection list behavior.

Todo: Refactor Index::collection_list_partial as
well.

Signed-off-by: Xiaoxi Chen <xiaoxi.chen@intel.com>
2015-08-07 08:43:43 -04:00
Xiaoxi Chen
d171537bd4 os/Memstore:Refactor collection_list_range and collection_list_partial
Add collection_list_impl which abstract the common process
of collection list behavior.

Signed-off-by: Xiaoxi Chen <xiaoxi.chen@intel.com>
2015-08-07 08:43:43 -04:00
Kefu Chai
ddc83b6bda Merge pull request #5508 from athanatos/wip-12410
OSDMonitor::preprocess_get_osdmap: send the last map as well

Reviewed-by: Kefu Chai <kchai@redhat.com>
2015-08-07 20:17:30 +08:00
Samuel Just
8e0a4445a2 Merge pull request #5413 from guangyy/wip-12523
osd: copy the RecoveryCtx::handle for when creating RecoveryCtx instance from another one

Reviewed-by: Samuel Just <sjust@redhat.com>
Reviewed-by: Sage Weil <sage@redhat.com>
Reviewed-by: Kefu Chai <kchai@redhat.com>
2015-08-07 04:50:20 -07:00
Yan, Zheng
e35d27fc69 Merge pull request #5492 from ceph/wip-vxattr-setall
mds: fix setting whole layout in one vxattr
2015-08-07 10:04:32 +08:00
Ruifeng Yang
e3147b8ddc rgw:segmentation fault when rgw_gc_max_objs > HASH_PRIME
obj_names[] is index out of bounds when rgw_gc_max_objs > HASH_PRIME

Fixes: #12630
Signed-off-by: Ruifeng Yang <149233652@qq.com>
2015-08-07 08:27:24 +08:00
Ruifeng Yang
9420d24c21 rgw:the arguments 'domain' should not be assigned when return false
Hostnames: [B.A]
    Inputs: [X.BB.A]
    Return: [false]
    Output: [B.A] it is wrong.

Fixes: #12629
Signed-off-by: Ruifeng Yang <149233652@qq.com>
2015-08-07 08:24:03 +08:00
John Spray
b3f1b3acfa Merge pull request #5488 from oritwas/wip-cmake-fix-rbd
cmake: add global lib to rbd

Reviewed-by: John Spray <john.spray@redhat.com>
2015-08-06 17:17:25 +01:00
Vikhyat Umrao
cd4ac1cae1 rbd: support size suffixes for size-based options
This patch will provide support for size suffixes for
below given options:

--stripe-unit <size in B/K/M>
--io-size <size in B/K/M/G/T>
--io-total <size in B/K/M/G/T>

Signed-off-by: Vikhyat Umrao <vumrao@redhat.com>
2015-08-06 20:40:51 +05:30
Radoslaw Zarzynski
d1735a4e8e rgw: rework X-Trans-Id header to be conform with Swift API.
Fixes: #12108
Signed-off-by: Radoslaw Zarzynski <rzarzynski@mirantis.com>
2015-08-06 16:27:04 +02:00
John Spray
278a6ae097 qa: add fs layout case for stripe_size decrease
The case fixed by commit:
 mds: fix setting whole layout in one vxattr

Signed-off-by: John Spray <john.spray@redhat.com>
2015-08-06 14:59:36 +01:00
Sage Weil
b63f2afadb Merge pull request #5475 from cxwshawn/com-fix
rgw_user.h: modify interface comments.

Reviewed-by: Sage Weil <sage@redhat.com>
2015-08-06 09:29:02 -04:00
Sage Weil
e61976b245 Merge pull request #5483 from oddomatik/patch-1
Fix location of --access=full in SWIFT user creation

Reviewed-by: Sage Weil <sage@redhat.com>
2015-08-06 09:25:55 -04:00
Sage Weil
0ab7926c00 Merge pull request #5484 from cxwshawn/de-fix
ceph-dencoder: add RGWRegion, RGWZoneParams, RGWOLHInfo support.

Reviewed-by: Sage Weil <sage@redhat.com>
2015-08-06 09:25:17 -04:00
Sage Weil
b436a35f3e Merge pull request #5421 from ceph/wip-selinux-policy-no-user
SELinux support for ceph

Reviewed-by: Kefu Chai <kchai@redhat.com>
Tested-by: Yuri Weinstein <yweinstei@redhat.com>
2015-08-06 09:12:24 -04:00
John Spray
880ffe9de5 mds: fix setting whole layout in one vxattr
Previously we were validating the layout
after setting each field, so even when
setting multiple fields in one go (to
something valid) the intermediate states
could be invalid and cause a bogus EINVAL.

For example try setting object_size and stripe_unit
both to 2M in one go.

Signed-off-by: John Spray <john.spray@redhat.com>
2015-08-06 13:14:44 +01:00
John Spray
0de2ea65f1 Merge pull request #5487 from ceph/wip-12619
mds: initialize InodeStoreBase::damage_flags

Reviewed-by: John Spray <john.spray@redhat.com>
2015-08-06 10:10:03 +01:00
Orit Wasserman
1559d5e953 cmake: add global lib to rbd
Signed-off-by: Orit Wasserman <owasserm@redhat.com>
2015-08-06 10:20:04 +02:00
Yan, Zheng
6b29233065 mds: initialize InodeStoreBase::damage_flags
Fixes: #12619
Signed-off-by: Yan, Zheng <zyan@redhat.com>
2015-08-06 16:14:01 +08:00
shawn
5d7cb4c212 ceph-dencoder: add RGWRegion, RGWZoneParams, RGWOLHInfo support.
since ceph-dencoder does not support the up-list structures, so
    I add these three supports to analyze data more quickly.

Signed-off-by: shawn chen <cxwshawn@gmail.com>
2015-08-06 10:48:27 +08:00
Brian Andrus
e67539e370 Fix location of --access=full in SWIFT user creation
--access=full was incorrectly placed in the output of the command, while it was supposed to have been a flag run with the command.
2015-08-05 16:25:48 -07:00
Boris Ranto
8c53a5824a ceph.spec.in: Make SELinux opt-out, not opt-in
We should enable SELinux support by default on rpm-based systems that
build with SELinux support to get better QE coverage in future runs.

Signed-off-by: Boris Ranto <branto@redhat.com>
2015-08-05 17:55:53 +02:00
Loic Dachary
f4853d8831 Merge pull request #5478 from dachary/wip-test-rados-striper
tests: rados striper tests use 7116 instead of 7113

Reviewed-by: Sebastien Ponce <sebastien.ponce@cern.ch>
2015-08-05 17:06:02 +02:00
Sage Weil
1092ed95a7 Merge pull request #5459 from cxwshawn/master
radosgw-admin: use cout instead of cerr to print help message.

Reviewed-by: Sage Weil <sage@redhat.com>
2015-08-05 10:34:21 -04:00