Commit Graph

44764 Commits

Author SHA1 Message Date
Sage Weil
139b5d6235 os: require Sequencer arg for apply_transaction()
Signed-off-by: Sage Weil <sage@redhat.com>
2015-09-14 20:09:50 -04:00
Sage Weil
7d89225acc Merge pull request #5923 from ceph/wip-limits
set limits in systemd unit files
2015-09-14 16:40:22 -04:00
Sage Weil
df921125b5 ceph_objectstore_test: fix warnings
In file included from test/objectstore/store_test.cc:34:0:
../src/gmock/gtest/include/gtest/gtest.h: In instantiation of ‘testing::AssertionResult testing::internal::CmpHelperEQ(const char*, const char*, const T1&, const T2&) [with T1 = unsigned int; T2 = int]’:
../src/gmock/gtest/include/gtest/gtest.h:1484:23:   required from ‘static testing::AssertionResult testing::internal::EqHelper<lhs_is_null_literal>::Compare(const char*, const char*, const T1&, const T2&) [with T1 = unsigned int; T2 = int; bool lhs_is_null_literal = false]’
test/objectstore/store_test.cc:411:5:   required from here
../src/gmock/gtest/include/gtest/gtest.h:1448:16: warning: comparison between signed and unsigned integer expressions [-Wsign-compare]
   if (expected == actual) {
                ^
../src/gmock/gtest/include/gtest/gtest.h: In instantiation of ‘testing::AssertionResult testing::internal::CmpHelperEQ(const char*, const char*, const T1&, const T2&) [with T1 = int; T2 = unsigned int]’:
../src/gmock/gtest/include/gtest/gtest.h:1484:23:   required from ‘static testing::AssertionResult testing::internal::EqHelper<lhs_is_null_literal>::Compare(const char*, const char*, const T1&, const T2&) [with T1 = int; T2 = unsigned int; bool lhs_is_null_literal = false]’
test/objectstore/store_test.cc:2003:7:   required from here
../src/gmock/gtest/include/gtest/gtest.h:1448:16: warning: comparison between signed and unsigned integer expressions [-Wsign-compare]
../src/gmock/gtest/include/gtest/gtest.h: In instantiation of ‘testing::AssertionResult testing::internal::CmpHelperEQ(const char*, const char*, const T1&, const T2&) [with T1 = int; T2 = long unsigned int]’:
../src/gmock/gtest/include/gtest/gtest.h:1484:23:   required from ‘static testing::AssertionResult testing::internal::EqHelper<lhs_is_null_literal>::Compare(const char*, const char*, const T1&, const T2&) [with T1 = int; T2 = long unsigned int; bool lhs_is_null_literal = false]’
test/objectstore/store_test.cc:2010:7:   required from here
../src/gmock/gtest/include/gtest/gtest.h:1448:16: warning: comparison between signed and unsigned integer expressions [-Wsign-compare]
../src/gmock/gtest/include/gtest/gtest.h: In instantiation of ‘testing::AssertionResult testing::internal::CmpHelperEQ(const char*, const char*, const T1&, const T2&) [with T1 = long unsigned int; T2 = int]’:
../src/gmock/gtest/include/gtest/gtest.h:1484:23:   required from ‘static testing::AssertionResult testing::internal::EqHelper<lhs_is_null_literal>::Compare(const char*, const char*, const T1&, const T2&) [with T1 = long unsigned int; T2 = int; bool lhs_is_null_literal = false]’
test/objectstore/store_test.cc:2080:5:   required from here
../src/gmock/gtest/include/gtest/gtest.h:1448:16: warning: comparison between signed and unsigned integer expressions [-Wsign-compare]

Signed-off-by: Sage Weil <sage@redhat.com>
2015-09-14 16:33:39 -04:00
Sage Weil
1002201a09 os/KeyValueStore: better osr debug
Signed-off-by: Sage Weil <sage@redhat.com>
2015-09-14 16:33:39 -04:00
Sage Weil
83c83e4440 os/KeyValueStore: kill default_osr
Signed-off-by: Sage Weil <sage@redhat.com>
2015-09-14 16:33:39 -04:00
Sage Weil
e7b57cd0da os/newstore: better sequencer debug
Signed-off-by: Sage Weil <sage@redhat.com>
2015-09-14 16:33:39 -04:00
Sage Weil
16c672c777 os/newstore: kill default_osr
Signed-off-by: Sage Weil <sage@redhat.com>
2015-09-14 16:33:39 -04:00
Sage Weil
47e7953c2b os/FileStore: improve debug output for sequencers
Signed-off-by: Sage Weil <sage@redhat.com>
2015-09-14 16:33:38 -04:00
Sage Weil
b46cdc02fa os/ObjectStore: fix leak ref on Sequencer_impl
This should start at 0 refs, not 1.

Signed-off-by: Sage Weil <sage@redhat.com>
2015-09-14 16:33:38 -04:00
Sage Weil
10b00f09d6 os/FileStore: assert Sequencer is passed in; kill default_osr
Signed-off-by: Sage Weil <sage@redhat.com>
2015-09-14 16:33:38 -04:00
Sage Weil
fedf360a3e osd: queue all transactions under an osr
This was missed in e7bbafa3bfbd5e936a8be026a30b83a89f6121c3.

Signed-off-by: Sage Weil <sage@redhat.com>
2015-09-14 16:33:28 -04:00
Yehuda Sadeh
d5f8c8e412 rgw: unit test for testing rgw_obj encoding correctness
Signed-off-by: Yehuda Sadeh <yehuda@redhat.com>
2015-09-14 13:19:36 -07:00
Yehuda Sadeh
583a2959fb rgw: json dump rgw_obj::orig_obj
useful for debugging

Signed-off-by: Yehuda Sadeh <yehuda@redhat.com>
2015-09-14 13:19:36 -07:00
Yehuda Sadeh
988979c3ef rgw: transform from cls_rgw_obj_key to rgw_obj correctly
Fixes: #12819
cls_rgw_obj_key keeps object names encoded, so need to transform from
it to rgw_obj correctly.

Signed-off-by: Yehuda Sadeh <yehuda@redhat.com>
2015-09-14 13:19:35 -07:00
Yehuda Sadeh
470afb9cc7 rgw: rgw_obj encoding fixes
Fix a few issues:
 - conversion to and from rgw_obj_key
 - strip_namespace_from_object() in the case where object starts
   with underscore
 - encoding if object has either namespace or instance and starts
   with underscore

Signed-off-by: Yehuda Sadeh <yehuda@redhat.com>
2015-09-14 13:19:35 -07:00
Sage Weil
8e13d89f0f systemd: eliminate ceph-rgw tmpfiles.d file
This is for storing the rgw socket files for fastcgi, which we do not
want to enable by default.

Signed-off-by: Sage Weil <sage@redhat.com>
2015-09-14 14:00:26 -04:00
Sage Weil
367c794cb1 systemd: no need to preprocess ceph-osd@service
This used to be necessary but now is not.

Signed-off-by: Sage Weil <sage@redhat.com>
2015-09-14 14:00:26 -04:00
Sage Weil
8453a89cb2 systemd: set nofile limit in unit files
Make it big so hopefully nobody has to change it.

Signed-off-by: Sage Weil <sage@redhat.com>
2015-09-14 14:00:26 -04:00
Sage Weil
1ff51a299d osd: drop default map cache size from 500 -> 200
OSDs don't need this from each other anymore since they send
a pg_interval_map_t with pg notify.  This cache is only used
during startup when loading things up (and 200 is still > the
max advance of 150), and when clients appear with super old maps
(pretty unusual unless they are idle for long periods).

It should mean a pretty significant cut in memory...

Signed-off-by: Sage Weil <sage@redhat.com>
2015-09-14 12:31:42 -04:00
Sage Weil
8678a8acec kill /etc/security/limits.d/ceph
This only affects login shells; not useful.

Signed-off-by: Sage Weil <sage@redhat.com>
2015-09-14 12:04:50 -04:00
Jason Dillaman
adb8478b61 include/inline_memory: out-of-bounds read on unaligned memory
When checking if a >=16 byte unaligned buffer is zeroed, the 32bit
check will read outside the buffer memory range.

Fixes: #13082
Signed-off-by: Jason Dillaman <dillaman@redhat.com>
2015-09-14 11:27:45 -04:00
Bo Cai
0b03b32d8b tools:remove the local file when get map failed.
Signed-off-by: Bo Cai <cai.bo@h3c.com>
2015-09-14 23:15:09 +08:00
Sage Weil
c2a9764fdd mon: do not return ref to MonOpRequest::get_session() caller
Most callers don't put the ref, and since we hold the op ref it's
not necessary.  This will also  make a somewhat cleaner transition
to MonSessionRef later.

Signed-off-by: Sage Weil <sage@redhat.com>
2015-09-14 10:24:22 -04:00
Sage Weil
d99e6899ca mon: fix MonSession leak when waitlisting op
Signed-off-by: Sage Weil <sage@redhat.com>
2015-09-14 09:56:57 -04:00
Orit Wasserman
2f663d9099 rgw: make radosgw-admin user rm idempotent
Fixes: #5378
Signed-off-by: Orit Wasserman <owasserm@redhat.com>
2015-09-14 10:50:16 +02:00
Sage Weil
40c3c85a39 mon: debug MonSession refs
Signed-off-by: Sage Weil <sage@redhat.com>
2015-09-13 22:04:29 -04:00
Sage Weil
640ebbfa3a mon/PGMonitor: fix use-after-free in stats ack
The MonOpRequestRef owns one ref; we need to get() another so we
can pass it to send_reply.

Signed-off-by: Sage Weil <sage@redhat.com>
2015-09-13 21:48:47 -04:00
Sage Weil
68ecc55d79 mon: do not leak ref creating MonOpRequest
Signed-off-by: Sage Weil <sage@redhat.com>
2015-09-13 21:44:51 -04:00
Sage Weil
3a7d91d250 msg/simple: debug refs on sent messages
Signed-off-by: Sage Weil <sage@redhat.com>
2015-09-13 21:42:27 -04:00
Sage Weil
7d112c6483 mon/MDSMonitor: drop incorrect m->put()
Signed-off-by: Sage Weil <sage@redhat.com>
2015-09-13 21:36:29 -04:00
Sage Weil
27d1e4d41e Merge pull request #5901 from yuyuyu101/wip-13001
AsyncMessenger: Kepp file_lock hold when accessing its event field
2015-09-13 09:12:39 -04:00
Sage Weil
786f23c5bd Merge pull request #5626 from majianpeng/objecter
Objecter: fix bus about objecter.

Reviewed-by: Sage Weil <sage@redhat.com>
2015-09-13 09:12:25 -04:00
Sage Weil
c625671c58 Merge pull request #5840 from liewegas/wip-12991
qa/workunits/cephtool/test.sh: make mds epoch check more tolerant

Reviewed-by: Loic Dachary <ldachary@redhat.com>
2015-09-13 09:08:57 -04:00
Sage Weil
770a573b0b Merge remote-tracking branch 'gh/infernalis' 2015-09-13 07:45:30 -04:00
Sage Weil
e73ac661a4 Merge pull request #5686 from liewegas/wip-hammer-feature
require osds are hammer v0.94.4 or later before upgrading past hammer

Reviewed-by: Loic Dachary <ldachary@redhat.com>
2015-09-13 07:41:56 -04:00
Sage Weil
cac601d631 Merge pull request #5905 from ceph/wip-rpm-notcmalloc
ceph.spec: respect CEPH_EXTRA_CONFIGURE_ARGS
2015-09-12 20:44:05 -04:00
Sage Weil
f8c66b4116 Merge pull request #5904 from ceph/wip-submodules
fix a few submodule issues

Reviewed-by: Loic Dachary <ldachary@redhat.com>
2015-09-12 20:43:48 -04:00
Sage Weil
af39f98b53 .gitignore: ignore src/ceph.tmpe
It makes the gitbuilders unhappy.  Not sure where it comes from...

Signed-off-by: Sage Weil <sage@redhat.com>
2015-09-12 17:43:52 -04:00
Sage Weil
98302adf99 gmock: ignore *.pyc
Signed-off-by: Sage Weil <sage@redhat.com>
2015-09-12 14:21:32 -04:00
Sage Weil
c57e8688fe rocksdb: ignore m4
Signed-off-by: Sage Weil <sage@redhat.com>
2015-09-12 14:17:01 -04:00
Sage Weil
51abff1168 ceph.spec: respect CEPH_EXTRA_CONFIGURE_ARGS
Signed-off-by: Sage Weil <sage@redhat.com>
2015-09-12 14:12:20 -04:00
Sage Weil
4a5a5b3705 qa/workunits/cephtool/test.sh: make mds epoch check more tolerant
This can race with an actual mdsmap epoch update for some other
reason.  We just need to make sure the epoch *increased*, not that
it is exactly old + 1.

Fixes: #12991
Signed-off-by: Sage Weil <sage@redhat.com>
2015-09-12 13:51:13 -04:00
Sage Weil
a511c24547 Merge remote-tracking branch 'gh/wip-10617-again' 2015-09-12 13:45:09 -04:00
Kefu Chai
9429b7735f Merge pull request #5334 from majianpeng/read-len-zero-for-ec
make read ec-object(offset=0, len=0) can work like read no-ec object.

Reviewed-by: Kefu Chai <kchai@redhat.com>
2015-09-13 00:05:27 +08:00
Sage Weil
d33fea5d38 sd/PG: tolerate missing pgmeta object
Bug 10617 left stray PG dirs around in firefly.  Hammer correctly
ignores these, assuming they are leftover cruft.  We broke this
when we dropped compat support in cd4e676e6d

See also #13060.

Signed-off-by: Sage Weil <sage@redhat.com>
2015-09-12 09:20:02 -04:00
Sage Weil
f15d9585ed osd: allow peek_map_epoch to return an error
Allow PG::peek_map_epoch to return an error indicating the PG
should be skipped.

Signed-off-by: Sage Weil <sage@redhat.com>
2015-09-12 09:20:02 -04:00
Sage Weil
2e44373eac Merge pull request #5776 from wonzhq/tmap-update
osd: force promote for ops which ec base pool can't handle

Reviewed-by: Josh Durgin <jdurgin@redhat.com>
Reviewed-by: Loic Dachary <ldachary@redhat.com>
2015-09-12 09:19:20 -04:00
Sage Weil
ff9600a6ce osd/ReplicatedPG: remove stray debug line
This snuck in

Signed-off-by: Sage Weil <sage@redhat.com>
2015-09-12 08:33:44 -04:00
Haomai Wang
6e85433832 AsyncMessenger: Kepp file_lock hold when accessing its event field
When process_event get a file event, other thread may delete this event later
but before we do event

Fix #13001
Signed-off-by: Haomai Wang <haomai@xsky.io>
2015-09-12 17:34:39 +08:00
Sage Weil
78de6fd61d Merge pull request #5896 from liewegas/wip-fix-spec
ceph.spec: include /etc/sysconfig/ceph

Reviewed-by: Ken Dreyer <kdreyer@redhat.com>
2015-09-11 13:02:29 -04:00