Commit Graph

69186 Commits

Author SHA1 Message Date
Jianpeng Ma
01a0805b2b common/EventTrace: fix compiler warning.
/mnt/ceph/src/osd/OSD.cc: In member function ‘void
OSD::dequeue_op(PGRef, OpRequestRef, ThreadPool::TPHandle&)’:
/mnt/ceph/src/osd/OSD.cc:9112:42: error: invalid conversion from ‘const
Message*’ to ‘Message*’ [-fpermissive]
   OID_EVENT_TRACE_WITH_MSG(op->get_req(), "DEQUEUE_OP_BEGIN", false);
                                          ^
In file included from /mnt/ceph/src/osd/OSD.h:57:0,
                 from /mnt/ceph/src/osd/OSD.cc:38:
/mnt/ceph/src/common/EventTrace.h:72:15: note:   initializing argument 1
of ‘static void EventTrace::trace_oid_event(Message*, const char*, const
char*, const char*, int, bool)’
   static void trace_oid_event(Message *m, const char *event, const char
*file,
               ^
/mnt/ceph/src/osd/OSD.cc:9158:42: error: invalid conversion from ‘const
Message*’ to ‘Message*’ [-fpermissive]
   OID_EVENT_TRACE_WITH_MSG(op->get_req(), "DEQUEUE_OP_END", true);
                                          ^
In file included from /mnt/ceph/src/osd/OSD.h:57:0,
                 from /mnt/ceph/src/osd/OSD.cc:38:
/mnt/ceph/src/common/EventTrace.h:72:15: note:   initializing argument 1
of ‘static void EventTrace::trace_oid_event(Message*, const char*, const
char*, const char*, int, bool)’
   static void trace_oid_event(Message *m, const char *event, const char
*file,
               ^
src/osd/CMakeFiles/osd.dir/build.make:62: recipe for target
'src/osd/CMakeFiles/osd.dir/OSD.cc.o' failed

Signed-off-by: Jianpeng Ma <jianpeng.ma@intel.com>
2017-02-28 00:30:58 +08:00
Sage Weil
d3c638fc5a doc/release-notes: choose terms carefully
Production-ready is in the eye of the beholder.

Signed-off-by: Sage Weil <sage@redhat.com>
2017-02-27 11:29:46 -05:00
Zack Cerza
99d942145f qa/suites/rest: Openstack volumes
Signed-off-by: Zack Cerza <zack@redhat.com>
2017-02-27 09:14:41 -07:00
Zack Cerza
964b983bdb qa/suites/ceph-ansible: Openstack volumes
Signed-off-by: Zack Cerza <zack@redhat.com>
2017-02-27 09:06:26 -07:00
Sage Weil
ef6c66287d os/filestore: debug which omap keys are set
This may help us find http://tracker.ceph.com/issues/19067

Signed-off-by: Sage Weil <sage@redhat.com>
2017-02-27 10:52:07 -05:00
Sage Weil
df3bc64532 test/CMakeLists.txt: include ceph-mgr
Signed-off-by: Sage Weil <sage@redhat.com>
2017-02-27 09:56:55 -05:00
Sage Weil
f6966ebdfa vstart.sh: fix init-ceph stop on vstart
Do this before the old ceph.conf is deleted!

Signed-off-by: Sage Weil <sage@redhat.com>
2017-02-27 09:56:55 -05:00
Sage Weil
6fff3e7478 vstart.sh: fix init-ceph path in cmake environment
Signed-off-by: Sage Weil <sage@redhat.com>
2017-02-27 09:56:55 -05:00
Sage Weil
b2ec3377ae ceph_common.sh: add mgr support to init-ceph
This is mainly for the benefit of vstart.

Signed-off-by: Sage Weil <sage@redhat.com>
2017-02-27 09:56:55 -05:00
Sage Weil
fdde99e15f stop.sh: add 'mgr' arg
Signed-off-by: Sage Weil <sage@redhat.com>
2017-02-27 09:56:55 -05:00
Sage Weil
f557edb189 vstart.sh: drop -r arg for radosgw, and use RGW=n instead
Simpler!  The start_* vars are all stupid.

Signed-off-by: Sage Weil <sage@redhat.com>
2017-02-27 09:56:55 -05:00
Sage Weil
55cd4e8b18 vstart.sh: respect CEPH_NUM_RGW
Signed-off-by: Sage Weil <sage@redhat.com>
2017-02-27 09:56:54 -05:00
Sage Weil
a559b01430 vstart.sh: start 1 ceph-mgr by default
Signed-off-by: Sage Weil <sage@redhat.com>
2017-02-27 09:56:54 -05:00
Sage Weil
cb61a6075f vstart.sh: default to -n ; add -N | --not-new for restart case
I never don't do -n; I'm guessing others are in the same boat.

Signed-off-by: Sage Weil <sage@redhat.com>
2017-02-27 09:56:54 -05:00
Sage Weil
924121b384 Merge pull request #13665 from xiexingguo/wip-fix-ec-op-out
osd/ECBackend: drop duplicated pending_commit field from << operator

Reviewed-by: Sage Weil <sage@redhat.com>
2017-02-27 08:49:53 -06:00
Sage Weil
9f03a14388 Merge pull request #13661 from dachary/wip-crush-doc
doc: crush typo in algorithm description

Reviewed-by: Kefu Chai <kchai@redhat.com>
2017-02-27 08:47:14 -06:00
songbaisen
e53f572135 osd: the osd should not share map with others when it is in stopping state.
the osd may not share its map with others when the osd is in stopping state.
     because it may from unhealthy to stopping.

Signed-off-by: song baisen <song.baisen@zte.com.cn>
2017-02-27 19:38:37 +08:00
John Spray
5124e4d881 mgr: add a get_version to the python interface
We had the metadata for all the remote services,
but no quick way for the python module to
ask "what version is the process I'm running in?"

Signed-off-by: John Spray <john.spray@redhat.com>
2017-02-27 11:29:43 +00:00
Igor Fedotov
c2c8513ff5 test/store_test: add deferred test case setup to support explicit min_alloc_size specification.
Fixes: #18857

Signed-off-by: Igor Fedotov <ifedotov@mirantis.com>
2017-02-27 11:16:36 +00:00
John Spray
6ea42d79ae mgr: use unique_ptr for MgrStandby::active_mgr
It should always have been one of these.

Signed-off-by: John Spray <john.spray@redhat.com>
2017-02-27 11:11:28 +00:00
Haomai Wang
7891484664 msg/async/rdma: remove lock protection for pending_sent_conns
Signed-off-by: Haomai Wang <haomai@xsky.com>
2017-02-27 18:24:56 +08:00
Adir Lev
a2bce6ef8b rename chunk_size
Change-Id: Icfcc80fa836b0076b839b59eac4271bdda9e1a9e
Signed-off-by: Adir Lev <adirl@mellanox.com>
2017-02-27 12:22:38 +02:00
Haomai Wang
06eadd2b9b msg/async/rdma: remove unused variable
Signed-off-by: Haomai Wang <haomai@xsky.com>
2017-02-27 18:20:30 +08:00
Haomai Wang
ab7de95dc7 msg/async/rdma: change notify to notify_worker
Signed-off-by: Haomai Wang <haomai@xsky.com>
2017-02-27 17:57:49 +08:00
Haomai Wang
c0b1e6ef10 Merge pull request #13509 from yuyuyu101/wip-tx-zerocopy
msg/async/rdma: cleanup

Reviewed-by: Adir lev <adirl@mellanox.com>
2017-02-27 17:56:52 +08:00
Kefu Chai
af5e1e8d39 Merge pull request #13531 from LiumxNL/fix-osd-utilization
mon/OSDMonitor: fix dividing by zero in OSDUtilizationDumper

Reviewed-by: Kefu Chai <kchai@redhat.com>
2017-02-27 17:51:03 +08:00
Kefu Chai
a8122af62c Merge pull request #13526 from ganeshmaharaj/tp_library_fixes
debian: Add missing tp files in deb packaging

Reviewed-by: Kefu Chai <kchai@redhat.com>
2017-02-27 17:50:40 +08:00
Kefu Chai
e094e88354 Merge pull request #13553 from yonghengdexin735/wip-zzz-dbev
os/bluestore/BlueStore.cc:remove unuse code in _open_bdev()

Reviewed-by: Igor Fedotov <ifedotov@mirantis.com>
2017-02-27 17:49:15 +08:00
Kefu Chai
b7adf4aa73 Merge pull request #13554 from songbaisen/s19
mon:  crush straw_calc_version value is 0 or 1 not 0 to 2.

Reviewed-by: Kefu Chai <kchai@redhat.com>
2017-02-27 17:48:26 +08:00
Kefu Chai
2db2f05535 Merge pull request #13653 from rzarzynski/wip-bs-bitmap-fast-exhausted-check
os/bluestore: the exhausted check in BitMapZone can be lock-less.

Reviewed-by: Sage Weil <sage@redhat.com>
2017-02-27 17:44:18 +08:00
Kefu Chai
66256aa5f7 Merge pull request #13656 from tchaikov/wip-19015
mon/MonClient: discard stray messages from non-acitve conns

Reviewed-by: Sage Weil <sage@redhat.com>
2017-02-27 17:42:20 +08:00
xie xingguo
a64addbcc1 osd/ECBackend: drop duplicated pending_commit field from << operator
Signed-off-by: xie xingguo <xie.xingguo@zte.com.cn>
2017-02-27 17:41:16 +08:00
Kefu Chai
edceabbd47 qa/tasks/workunit: use ceph.git as an alternative of ceph-ci.git for workunit repo
if we run upgrade test, where, for example, "jewel" is not in
ceph-ci.git repo, we should check ceph.git to clone the workunits.

Signed-off-by: Kefu Chai <kchai@redhat.com>
2017-02-27 17:36:05 +08:00
Haomai Wang
a6108dc4b5 msg/async/rdma: accelerate tx/rx buffer ownership lookup
Signed-off-by: Haomai Wang <haomai@xsky.com>
2017-02-27 16:33:39 +08:00
Haomai Wang
a9b13b21e8 msg/async/rdma: cleanup
Signed-off-by: Haomai Wang <haomai@xsky.com>
2017-02-27 16:22:32 +08:00
Loic Dachary
31bf2e9acf doc: crush typo in algorithm description
Signed-off-by: Loic Dachary <ldachary@redhat.com>
2017-02-27 09:18:04 +01:00
Brad Hubbard
bc458d3963 rgw: set dumpable flag after setuid post ff0e521
ff0e521 resolved the issue for the other daemons but not for rgw since
it calls setuid (via civetweb) after the new code sets PR_SET_DUMPABLE.
Add another prctl call before wait_shutdown.

Fixes: http://tracker.ceph.com/issues/19089

Signed-off-by: Brad Hubbard <bhubbard@redhat.com>
2017-02-27 16:08:41 +10:00
Haomai Wang
bd527486f6 msg/async/rdma: cleanup Infiniband, remove unused functions
Signed-off-by: Haomai Wang <haomai@xsky.com>
2017-02-27 13:55:37 +08:00
Haomai Wang
ef19e90f51 Merge pull request #13603 from Adirl/fix_compile
msg/async/rdma: Fix broken compilation

Reviewed-by: Haomai Wang <haomai@xsky.com>
2017-02-27 13:55:13 +08:00
Sage Weil
da06227c7f Merge pull request #13652 from liupan1111/wip-cleanup-rbd-nbd
rbd-nbd: fix a typo "moudle"

Reviewed-by: Mykola Golub <mgolub@mirantis.com>
2017-02-26 15:07:19 -06:00
Kefu Chai
87e0759bf2 auth/cephx: print challenges in hex
this matches the CephxServiceHandler::handle_request(), hence would be
easier for debugging.

Signed-off-by: Kefu Chai <kchai@redhat.com>
2017-02-27 00:23:07 +08:00
Kefu Chai
c0f1600666 mon/MonClient: discard stray messages from non-acitve conns
Fixes: http://tracker.ceph.com/issues/19015
Signed-off-by: Kefu Chai <kchai@redhat.com>
2017-02-27 00:23:07 +08:00
John Spray
477ddea8a8 Merge pull request #13227 from ukernel/wip-multimds-misc
mds: misc multimds fixes

Reviewed-by: John Spray <john.spray@redhat.com>
2017-02-26 11:16:40 +00:00
Kefu Chai
259eac52aa Merge pull request #13648 from tchaikov/wip-doc-osd-agent-options
doc: document `osd-agent-{max,low}-ops` options

Reviewed-by: Zhiqiang Wang <zhiqiang@xsky.com>
Reviewed-by: Brad Hubbard <bhubbard@redhat.com>
2017-02-26 17:35:54 +08:00
Pan Liu
5cc7771447 rbd-nbd: fix a typo "moudle"
Signed-off-by: Pan Liu <liupan1111@gmail.com>
2017-02-26 12:14:15 +08:00
Kefu Chai
0d858b9789 doc: document osd-agent-{max,low}-ops options
Signed-off-by: Kefu Chai <kchai@redhat.com>
2017-02-26 11:12:22 +08:00
Radoslaw Zarzynski
19ca16b856 bluestore: remove CephContext* from BmapEntry.
Each BmapEntry instance stores a pointer to the same CephContext.
As we expect to have thousands of instances the overhead might
be too high. For instance, serving 1 TiB SSD disk on x86-64,
while using the default settings, results in 32 MiB of extra
memory consumption:

  # assuming sizeof(unsigned long) * CHAR_BIT == 64
  >>> 1024 * 1024 * 1024 * 1024 / 4096 / 64
  4194304
  >>> 4194304 * 8 / 1024
  32768

Although memory is cheap, CPU's caches are not.

Signed-off-by: Radoslaw Zarzynski <rzarzynski@mirantis.com>
2017-02-25 22:21:57 +01:00
Radoslaw Zarzynski
d76f222462 bluestore: the exhausted check in BitMapZone can be lock-less.
Before the patch BitMapZone::is_exhausted() required from its
callers to acquire appropriate lock. However, fulfilling this
condition is not really necessary to use the method correctly
while it can significantly hurt performance.

The change allows BitMapAreaLeaf::child_check_n_lock() to not
acquire the lock while examining zones for being exhausted.

Signed-off-by: Radoslaw Zarzynski <rzarzynski@mirantis.com>
2017-02-25 20:47:24 +01:00
Sage Weil
af5dab0613 Merge pull request #13649 from liewegas/wip-ceph-scrub-debug
qa/tasks/ceph.py: debug which pgs aren't scrubbing

Reviewed-by: Brad Hubbard <bhubbard@redhat.com>
2017-02-25 13:15:06 -06:00
Mykola Golub
44e8c873e2 msg: end parameter in entity_addr_t::parse is optional
Signed-off-by: Mykola Golub <mgolub@mirantis.com>
2017-02-25 17:42:42 +01:00