Commit Graph

70663 Commits

Author SHA1 Message Date
Radoslaw Zarzynski
06b9a84851 rgw: implement the full response generation in BulkUpload of Swift API.
Signed-off-by: Radoslaw Zarzynski <rzarzynski@mirantis.com>
2017-04-02 18:35:22 +02:00
Radoslaw Zarzynski
3770e2473b rgw: optimize metadata caching in BulkUpload of Swift API.
Signed-off-by: Radoslaw Zarzynski <rzarzynski@mirantis.com>
2017-04-02 18:35:22 +02:00
Radoslaw Zarzynski
0c4d1dfabf rgw: implement the object creation in BulkUpload of Swift API.
Signed-off-by: Radoslaw Zarzynski <rzarzynski@mirantis.com>
2017-04-02 18:35:22 +02:00
Radoslaw Zarzynski
e9afe38665 rgw: implement the container creation in BulkUpload of Swift API.
Signed-off-by: Radoslaw Zarzynski <rzarzynski@mirantis.com>
2017-04-02 18:35:21 +02:00
Amir Vadai
23fd3efaaf msg/async/rdma: Fix small memory leaks detected by valgrind
Issue: 995322
Change-Id: Iccffbd0defd09aa51c4b51b78733921444ea1c5f
Signed-off-by: Amir Vadai <amir@vadai.me>
2017-04-02 16:47:46 +03:00
Radoslaw Zarzynski
6c1bb7a10e rgw: implement the AlignedStreamGetter.
Signed-off-by: Radoslaw Zarzynski <rzarzynski@mirantis.com>
2017-04-02 15:46:13 +02:00
Radoslaw Zarzynski
9ed7edaaf7 rgw: implement the basic security check for BulkUpload of Swift API.
Signed-off-by: Radoslaw Zarzynski <rzarzynski@mirantis.com>
2017-04-02 15:46:13 +02:00
Kefu Chai
ef006681b7 Merge pull request #14170 from liewegas/wip-fewer-warnings
common/ceph_context: fewer warnings about experimental features

Reviewed-by: Kefu Chai <kchai@redhat.com>
2017-04-02 12:53:29 +08:00
Kefu Chai
47080150a1 qa/tasks/workunit.py: use "overrides" as the default settings of workunit
otherwise the settings in "workunit" tasks are always overridden by the
settings in template config. so we'd better follow the way of how
"install" task updates itself with the "overrides" settings: it uses the
"overrides" as the *defaults*.

Fixes: http://tracker.ceph.com/issues/19429
Signed-off-by: Kefu Chai <kchai@redhat.com>
2017-04-02 12:26:30 +08:00
Brad Hubbard
fa3ddead5e Merge pull request #14271 from joscollin/wip-krbd-warning
rbd: warning, ‘devno’ may be used uninitialized in this function

Reviewed-by: Jason Dillaman <dillaman@redhat.com>
2017-04-02 08:59:38 +10:00
Kefu Chai
d65ee4ba52 Merge pull request #13909 from kestrels/wip-19217
crc32c: Add crc32c function optimized for ppc architecture

Reviewed-by: Kefu Chai <kchai@redhat.com>
2017-04-02 01:44:03 +08:00
Kefu Chai
09c977869d Merge pull request #14124 from Fabian-Gruenbichler/ipv6_fix
common: fix segfault in public IPv6 addr picking

Reviewed-by: Kefu Chai <kchai@redhat.com>
2017-04-02 01:41:50 +08:00
Kefu Chai
c2432a5a6b Merge pull request #14173 from badone/wip-do_request-abort
common: Fix heap buffer overflow in do_request

Reviewed-by: Kefu Chai <kchai@redhat.com>
2017-04-02 01:40:30 +08:00
Kefu Chai
e2b7cc00d9 Merge pull request #14193 from liupan1111/wip-fix-delete-full
librados: set the flag CEPH_OSD_FLAG_FULL_TRY of Op in the right place.

Reviewed-by: Sage Weil <sage@redhat.com>
2017-04-02 01:39:46 +08:00
Kefu Chai
3f92a8522f Merge pull request #14248 from badone/wip-simple_spin_lock-delay
common: Make spinlock delay more conventional

Reviewed-by: Kefu Chai <kchai@redhat.com>
2017-04-02 01:38:13 +08:00
Kefu Chai
6ae3ce9148 Merge pull request #13690 from jcsp/wip-17503
mgr: pass through cluster log to plugins

Reviewed-by: Greg Farnum <gfarnum@redhat.com>
Reviewed-by: Kefu Chai <kchai@redhat.com>
2017-04-02 01:37:31 +08:00
Radoslaw Zarzynski
4c4cdf2183 rgw: implement the TAR extraction loop of Swift's BulkUpload.
Signed-off-by: Radoslaw Zarzynski <rzarzynski@mirantis.com>
2017-04-01 15:46:26 +02:00
Radoslaw Zarzynski
278978be2f rgw: implement the TAR format interpreter.
Signed-off-by: Radoslaw Zarzynski <rzarzynski@mirantis.com>
2017-04-01 15:46:26 +02:00
Radoslaw Zarzynski
0edc4e55b2 rgw: implement RGWBulkUploadOp::DecoratedStreamGetter.
Signed-off-by: Radoslaw Zarzynski <rzarzynski@mirantis.com>
2017-04-01 15:46:25 +02:00
Radoslaw Zarzynski
0091c83498 rgw: initial class structure for BulkUpload of Swift API.
Signed-off-by: Radoslaw Zarzynski <rzarzynski@mirantis.com>
2017-04-01 15:46:25 +02:00
tang.jin
2f78a100fe osd: fix comments about pg refs and lock
Signed-off-by: tang.jin <tang.jin@istuary.com>
2017-04-01 16:44:21 +08:00
Orit Wasserman
ea9331e484 Merge pull request #14276 from atheism/rgw-rgwop-remove-unused-iter
rgw: rgw_op: remove unused variable iter
Reviewed-by: Orit Wasserman <owasserm@redhat.com>
2017-04-01 09:48:46 +03:00
Weibing Zhang
fa1fe5fc1d test: objectstore: chain_xattr: fix wrong memset usage to fill buf
Current usage:
        memset(buf, sizeof(buf), 0x1F);
    should be:
        memset(buf, 0x1F, sizeof(buf));
    to fill the buffer with '0x1F'.

Signed-off-by: Weibing Zhang <zhangweibing@unitedstack.com>
2017-04-01 13:24:42 +08:00
Weibing Zhang
d5a525c645 rgw: rgw_op: remove unused variable iter
The variable iter is not used in  RGWInitMultipart::execute().

Signed-off-by: Weibing Zhang <zhangweibing@unitedstack.com>
2017-04-01 11:52:59 +08:00
Sage Weil
c4ba267245 Merge pull request #13811 from rzarzynski/wip-bs-bitmap-mem-cleanup
os/bluestore: memory and dereference clean-up in the BitAllocator

Reviewed-by: Sage Weil <sage@redhat.com>
2017-03-31 20:54:22 -05:00
Sage Weil
cc6a88ec2c Merge pull request #14121 from majianpeng/bluefs-get-allocated
os/bluestore/BlueFS: optimize get_allocated

Reviewed-by: Sage Weil <sage@redhat.com>
2017-03-31 20:54:07 -05:00
Sage Weil
841de704cf Merge pull request #14162 from majianpeng/optimize-bluestore-force-flush
os/bluestore: clean up flush logic

Reviewed-by: Sage Weil <sage@redhat.com>
2017-03-31 20:53:57 -05:00
Sage Weil
b19613f8ab Merge pull request #14225 from ivancich/wip_fix_throttle_config_change
os/bluestore: make live changes for BlueStore throttle config work like initial config

Reviewed-by: Sage Weil <sage@redhat.com>
2017-03-31 20:52:59 -05:00
Zhang Shaowen
8b43c97812 rgw: version id doesn't work in fetch_remote_obj
Signed-off-by: Zhang Shaowen <zhangshaowen@cmss.chinamobile.com>
2017-04-01 09:47:58 +08:00
Sage Weil
b0d7e13435 osd: only send beacon on connect if active
Otherwise the mon will just ignore it and we'll wait another
cycle.

Signed-off-by: Sage Weil <sage@redhat.com>
2017-03-31 15:23:27 -04:00
Sage Weil
4b07746381 osd: fix beacon condition
- && should have been ||
- change this to instead not send when monmap epoch==0,
  and only set last_sent_beacon if we actually send.  That
  way we'll send on the next tick after we get a real monmap.

Signed-off-by: Sage Weil <sage@redhat.com>
2017-03-31 15:22:00 -04:00
Sage Weil
52fbdc5e94 osd/ReplicatedBackend: remove MOSDSubOp cruft from repop_commit
Signed-off-by: Sage Weil <sage@redhat.com>
2017-03-31 14:42:12 -04:00
Sage Weil
137b3721b4 osd: drop unused MOSDSubOp includes
Signed-off-by: Sage Weil <sage@redhat.com>
2017-03-31 14:42:12 -04:00
Sage Weil
5ecf7b0ba1 osd/ReplicatedBackend: rename sub_op_modify* do_repop*
Signed-off-by: Sage Weil <sage@redhat.com>
2017-03-31 14:42:12 -04:00
Sage Weil
d06128b764 osd/ReplicatedBackend: no empty MOSDSubOps anymore
Signed-off-by: Sage Weil <sage@redhat.com>
2017-03-31 14:42:12 -04:00
Sage Weil
3a271eac3c osd/ReplicatedBackend: use more __func__
Signed-off-by: Sage Weil <sage@redhat.com>
2017-03-31 14:42:12 -04:00
Sage Weil
68565b5c5f osd: drop unnecessary declarations in headers
Signed-off-by: Sage Weil <sage@redhat.com>
2017-03-31 14:42:12 -04:00
Sage Weil
b1a38263ba osd: use MOSDScrubReserve instead of MOSDSubOp for scrub reservations
This is the last MOSDSubOp user.

Note that while the next step is to move to AsyncReserver internally,
this isn't quite yet possible since AsyncReserve "blocks" indefinitely
so we wouldn't generate a REJECT.  Changing how we schdule scrubs
internally will take a bit more work.

Signed-off-by: Sage Weil <sage@redhat.com>
2017-03-31 14:42:11 -04:00
Sage Weil
42be51d2ab osd: use new MOSDRepScrubMap message for ScrubMap instead of MOSDSubOp
This is one of the last remaining users of MOSDSubOp!

Signed-off-by: Sage Weil <sage@redhat.com>
2017-03-31 14:42:11 -04:00
Sage Weil
dc5fe9f226 osd: use new MOSDPGBackfillRemove instead of MOSDSubOp
One of the last remaining MOSDSubOp users!

Signed-off-by: Sage Weil <sage@redhat.com>
2017-03-31 14:41:42 -04:00
Sage Weil
2effd54e4e osd/ReplicatedBackend: remove legacy push/pull op handling
These are never sent, and thus never recieved.

Signed-off-by: Sage Weil <sage@redhat.com>
2017-03-31 14:41:42 -04:00
vasukulkarni
574049a90b Merge pull request #14229 from ceph/wip-systemd
qa: Add reboot case for systemd test
2017-03-31 09:15:53 -07:00
John Spray
992b8499d0 Merge pull request #14254 from idryomov/wip-vstart-runner-ps
qa/vstart_runner: amend ps invocation

Reviewed-by: John Spray <john.spray@redhat.com>
2017-03-31 17:15:30 +01:00
vasukulkarni
b6895fbe0e Merge pull request #14226 from ceph/preset-mon
systemd: remove ceph-create-keys from presets
2017-03-31 09:15:21 -07:00
Sage Weil
59548509ef Merge pull request #14266 from tchaikov/wip-require-mgr-for-pg-dump
tests: add MGR=1 so 'pg dump' won't be blocked

Reviewed-by: Sage Weil <sage@redhat.com>
Reviewed-by: Josh Durgin <jdurgin@redhat.com>
2017-03-31 10:58:56 -05:00
Sage Weil
e5b07011b2 Merge pull request #14259 from johnwilkins/wip-doc-multisite
doc: wip-doc-multisite ports downstream multisite document upstream.
2017-03-31 10:52:53 -05:00
Sage Weil
30cd328e8a test/rgw/test_rgw_manifest: fix warning
Signed-off-by: Sage Weil <sage@redhat.com>
2017-03-31 11:25:35 -04:00
Jos Collin
87c0fd2399 rbd: warning, ‘devno’ may be used uninitialized in this function
The following warning appears during make. Fixed in both unmap_image() functions in krbd.cc

--
krbd.cc: In function ‘int krbd_unmap_by_spec(krbd_ctx*, const char*, const char*, const char*, const char*)’:
krbd.cc:608:65: warning: ‘devno’ may be used uninitialized in this function [-Wmaybe-uninitialized]
   return do_unmap(ctx->udev, devno, build_unmap_buf(id, options));
                                                                 ^
krbd.cc:591:9: note: ‘devno’ was declared here
   dev_t devno;
--

Signed-off-by: Jos Collin <jcollin@redhat.com>
2017-03-31 20:01:43 +05:30
Sage Weil
8bc197400d ceph_test_librados_api_misc: fix stupid LibRadosMiscConnectFailure.ConnectFailure test
Sometimes the cond doesn't time out and it wakes up instead.  Just repeat
the test many times to ensure that at least once it times out (usually
it doesn't; it's pretty infrequent that it doesn't).

Fixes: http://tracker.ceph.com/issues/15368
Signed-off-by: Sage Weil <sage@redhat.com>
2017-03-31 10:06:42 -04:00
Sage Weil
973829132b mon/OSDMonitor: spinlock -> std::mutex
I think spinlock is dangerous here: we're doing semi-unbounded
work (decode).  Also seemingly innocuous code like dout macros
take mutexes.

Signed-off-by: Sage Weil <sage@redhat.com>
2017-03-31 09:33:19 -04:00