RepoMirrors/ceph
1
0
Fork 0
mirror of https://github.com/ceph/ceph synced 2025-01-16 16:03:29 +00:00
Code Issues Packages Projects Releases Wiki Activity
78,728 Commits 216 Branches 394 Tags 1.6 GiB
032d819a26
Commit Graph

78728 Commits

This Branch
This Branch
All Branches
Author SHA1 Message Date
Kefu Chai
5c8e65a2a4 osdc/Journaler: use new style options 
also use uint64_t for these options, as they are casted to uint64_t when
reading from conf.

Signed-off-by: Kefu Chai <kchai@redhat.com>
 2017-09-22 12:28:21 +08:00
Kefu Chai
3988bf26ed Merge pull request #17830 from Yan-waller/wip-walle-0920primarypg 
osd/PrimaryLogPG: misc cleanups

Reviewed-by: Kefu Chai <kchai@redhat.com>
 2017-09-22 12:24:07 +08:00
Carl Xiong
a385b5b0c3 ceph-disk: fix '--runtime' omission for ceph-osd service 
f425a127b introduces a regression that ceph-disk omits "--runtime" when
enabling ceph-osd@$ID.service units for device-backed OSDs.

Fixes: http://tracker.ceph.com/issues/21498

Signed-off-by: Carl Xiong <cxiong@suse.com>
 2017-09-22 11:55:33 +08:00
Kefu Chai
42be200c56 qa/tasks: prolong revive_osd() timeout to 6 min 
bluestore_fsck_on_mount and bluestore_fsck_on_mount_deep are enabled by
default. and bluestore is used as the default store backend. it takes
longer to perform the deep fsck with verbose log. so prolong the
revive_osd()'s timeout from 150 sec to 360 sec.

Fixes: http://tracker.ceph.com/issues/21474
Signed-off-by: Kefu Chai <kchai@redhat.com>
 2017-09-22 10:58:41 +08:00
Kefu Chai
04697cdbdc cmake: update the error message for gperftools bug 
it's a follow-up of #17788

Signed-off-by: Kefu Chai <kchai@redhat.com>
 2017-09-22 10:40:29 +08:00
Boris Ranto
4042c8eba5 Merge pull request #17891 from b-ranto/wip-selinux-lnk-getattr 
selinux: Allow getattr on lnk sysfs files

Reviewed-by: Ken Dreyer <kdreyer@redhat.com>
 2017-09-22 04:35:26 +02:00
Sage Weil
3a6a0f4ab0 Merge pull request #17739 from liewegas/wip-filestore-configs 
common/options: document filestore and filejournal options

Reviewed-by: Kefu Chai <kchai@redhat.com>
 2017-09-21 21:25:52 -05:00
Sage Weil
787ea8aba0 Merge pull request #17740 from liewegas/wip-objecter-configs 
common/options: document objecter, filer, and journal options

Reviewed-by: Kefu Chai <kchai@redhat.com>
 2017-09-21 21:25:36 -05:00
Sage Weil
91b65236c8 Merge pull request #17579 from liewegas/wip-kill-snapdirs 
osd: kill snapdirs

Reviewed-by: Kefu Chai <kchai@redhat.com>
 2017-09-21 21:25:17 -05:00
Kefu Chai
bf6619cb22 Merge pull request #17895 from ivancich/wip-fix-mclock-coverity 
tests: fix uninitialized value found by coverity scan

Reviewed-by: Kefu Chai <kchai@redhat.com>
 2017-09-22 10:04:38 +08:00
Vasu Kulkarni
56d1511bdf use ovh instead of vps 
Signed-off-by: Vasu Kulkarni <vasu@redhat.com>
 2017-09-21 16:43:24 -07:00
J. Eric Ivancich
63be40152c tests: fix uninitialized value found by coverity scan 
Unit test for mclock priority queue had a default constructor, which
didn't initialize a value.

Signed-off-by: J. Eric Ivancich <ivancich@redhat.com>
 2017-09-21 15:24:19 -04:00
Sage Weil
6169cf6442 osd: make shutdown debug conditional (and off by default) 
Signed-off-by: Sage Weil <sage@redhat.com>
 2017-09-21 15:19:47 -04:00
Sage Weil
65bec679f2 common/options: document filestore and filejournal options 
Signed-off-by: Sage Weil <sage@redhat.com>
 2017-09-21 14:37:37 -04:00
Sage Weil
c065ef48a5 osd/PrimaryLogPG: can_create false for cache ops 
Some cache ops do writes, but none of them *create* objects, so we
should not set can_create.

Signed-off-by: Sage Weil <sage@redhat.com>
 2017-09-21 14:29:35 -04:00
Sage Weil
d5efb8a7be osd/PrimaryLogPG: drop odd maintain_ssc hack 
This was added two years ago by b61f3e43f1,
but I can see why it is needed here.  The test case for that
original bug (from a3442cf078) passes
fine.

Signed-off-by: Sage Weil <sage@redhat.com>
 2017-09-21 14:29:35 -04:00
Sage Weil
e4c4942467 ceph-objectstore-tool: drop pre-snapset support for setting up snapmapper 
Signed-off-by: Sage Weil <sage@redhat.com>
 2017-09-21 14:29:35 -04:00
Sage Weil
c6275e7d4d ceph-objectstore-tool: drop head_exists corruption 
Signed-off-by: Sage Weil <sage@redhat.com>
 2017-09-21 14:29:35 -04:00
Sage Weil
76f1c14c77 ceph_test_rados_api_pool: make LibRadosPools.PoolList tolerate long pool names 
There was a baked-in assumption that the first pool was <20 chars.
No more!

Signed-off-by: Sage Weil <sage@redhat.com>
 2017-09-21 14:29:35 -04:00
Sage Weil
cc17c22b43 osd/osd_types: object_info_t: remove legacy_snaps 
Signed-off-by: Sage Weil <sage@redhat.com>
 2017-09-21 14:29:35 -04:00
Sage Weil
8c3f484813 osd/PrimaryLogPG: drop last legacy_snaps users (just debug output!) 
Signed-off-by: Sage Weil <sage@redhat.com>
 2017-09-21 14:29:34 -04:00
Sage Weil
900d23a94d osd/osd_types: SnapSet: remove head_exists 
Signed-off-by: Sage Weil <sage@redhat.com>
 2017-09-21 14:29:34 -04:00
Sage Weil
0e707ae40b osd: drop last head_exists = true calls 
No more users!

Signed-off-by: Sage Weil <sage@redhat.com>
 2017-09-21 14:29:34 -04:00
Sage Weil
5e13884acc osd/PrimaryLogPG: find_object_context: drop !head_exists path 
Signed-off-by: Sage Weil <sage@redhat.com>
 2017-09-21 14:29:34 -04:00
Sage Weil
a9145997db osd/PrimaryLogPG: do_osd_ops: head_exists is always true for LIST_SNAPS 
Signed-off-by: Sage Weil <sage@redhat.com>
 2017-09-21 14:29:34 -04:00
Sage Weil
b5ac17874d osd/osd_types: SnapSet: drop is_legacy 
Signed-off-by: Sage Weil <sage@redhat.com>
 2017-09-21 14:29:34 -04:00
Sage Weil
fca22aae5d osd/PrimaryLogPG: start_flush: drop legacy snapset handling 
Signed-off-by: Sage Weil <sage@redhat.com>
 2017-09-21 14:29:34 -04:00
Sage Weil
222a5e6f13 osd/osd_types: default head_exists = true 
This will be helpful as we start removing code that would otherwise
set this to true in some convoluted way.

Signed-off-by: Sage Weil <sage@redhat.com>
 2017-09-21 14:29:33 -04:00
Sage Weil
8924bbeb5f osd/PrimaryLogPG: finish_promote: drop legacy snapset handling 
Signed-off-by: Sage Weil <sage@redhat.com>
 2017-09-21 14:29:33 -04:00
Sage Weil
cec378d94d osd/PrimaryLogPG: do_copy_get: drop legacy snapset path 
Signed-off-by: Sage Weil <sage@redhat.com>
 2017-09-21 14:29:33 -04:00
Sage Weil
3671bef26b osd/PrimaryLogPG: _delete_oid: drop legacy snapset handling 
Signed-off-by: Sage Weil <sage@redhat.com>
 2017-09-21 14:29:33 -04:00
Sage Weil
8d55ba7e2c osd/PrimaryLogPG: do_osd_ops: drop legacy snapset handling from LIST_SNAPS 
This is so much simpler, yay!

Signed-off-by: Sage Weil <sage@redhat.com>
 2017-09-21 14:29:33 -04:00
Sage Weil
1f90d10b2a osd/PrimaryLogPG: on_local_recover: drop legacy snapset handling 
This was there to deal with jewel peers, which are no more.

Signed-off-by: Sage Weil <sage@redhat.com>
 2017-09-21 14:29:33 -04:00
Sage Weil
e3779d9e0f osd/PG: _scan_snaps: drop legacy snapset handling 
Signed-off-by: Sage Weil <sage@redhat.com>
 2017-09-21 14:29:33 -04:00
Sage Weil
89c3439191 osd/PrimaryLogPG: find_object_context: drop legacy snapset handling 
Signed-off-by: Sage Weil <sage@redhat.com>
 2017-09-21 14:29:33 -04:00
Sage Weil
bd36330ff8 osd/PrimaryLogPG: issue_repop: drop obsolete condition on head_obc 
The only user is trim_object, which sets obc = clone, head_obc =
head.

Signed-off-by: Sage Weil <sage@redhat.com>
 2017-09-21 14:29:32 -04:00
Sage Weil
b01783fb97 osd/PrimaryLogPG: snapset_obc -> head_obc 
This is only used for trim_object, now!

Signed-off-by: Sage Weil <sage@redhat.com>
 2017-09-21 14:29:32 -04:00
Sage Weil
93ae366bff osd/PrimaryLogPG: do_op: drop snapset_obc lookup on snapdir 
We'll never find it.  This also means we can drop the goofy hack
to clear it for cache tiers.

Signed-off-by: Sage Weil <sage@redhat.com>
 2017-09-21 14:29:32 -04:00
Sage Weil
0d45f26243 osd/PrimaryLogPG: trim_object: ctx->snapset_obc -> head_obc 
Signed-off-by: Sage Weil <sage@redhat.com>
 2017-09-21 14:29:32 -04:00
Sage Weil
a53e731f8f osd/PrimaryLogPG: trim_object: simplify a bit more for post-snapdir world 
Signed-off-by: Sage Weil <sage@redhat.com>
 2017-09-21 14:29:32 -04:00
Sage Weil
203cd6fe81 osd/PrimaryLogPG: trim_object: drop legacy SnapSet handling 
Signed-off-by: Sage Weil <sage@redhat.com>
 2017-09-21 14:29:32 -04:00
Sage Weil
08d2c8875b common/AsyncReserver: get a cct 
Signed-off-by: Sage Weil <sage@redhat.com>
 2017-09-21 14:22:18 -04:00
Sage Weil
31a34946f7 osd: PG_STATE_BACKFILL -> PG_STATE_BACKFILLING 
Match user-facing string

Signed-off-by: Sage Weil <sage@redhat.com>
 2017-09-21 14:21:42 -04:00
Sage Weil
6fa40e44dc osd/osd_types: make BACKFILL <-> "backfilling" for parser 
We render BACKFILL as "backfilling"; make sure parse works that
way too.

Signed-off-by: Sage Weil <sage@redhat.com>
 2017-09-21 14:21:41 -04:00
Sage Weil
5bcfaf4bd4 osd/osd_types: remove weird BACKFILL state hack 
Signed-off-by: Sage Weil <sage@redhat.com>
 2017-09-21 14:21:41 -04:00
Sage Weil
2e45497a20 osd/PG: Cancel{Recovery,Backfill} -> Defer{Recovery,Backfill} 
"Defer" is more accurate here; we aren't canceling anything, just
rescheduling the work.

Signed-off-by: Sage Weil <sage@redhat.com>
 2017-09-21 14:21:33 -04:00
Zac Medico
caf6803b13 PGPool::update: optimize removed_snaps comparison when possible 
In self/unmanaged snaps mode, optimize removed_snaps comparison
for cases where removed_snaps has not changed. This exploits the
fact that remove_unmanaged_snap adds a dummy removed snapshot
to the end of removed_snaps, allowing for inexpensive detection
of changes. In cases where removed_snaps is very large, this
optimization improves performance dramatically.

Signed-off-by: Zac Medico <zmedico@gmail.com>
 2017-09-21 10:33:23 -07:00
Kefu Chai
344f112bb3 common: use mono clock for HeartbeatMap 
to avoid problems when admin sets system clock.

see https://gcc.gnu.org/bugzilla/show_bug.cgi?id=53901 for the reason
why we cannot use atomic<time_point>.

Signed-off-by: Kefu Chai <kchai@redhat.com>
Signed-off-by: Xinze Chi <xinze@xsky.com>
 2017-09-22 01:12:13 +08:00
Boris Ranto
394c26adb9 selinux: Allow getattr on lnk sysfs files 
This showed up during downstream testing for luminous. We are doing
getattr on the sysfs lnk files and the current policy does not allow
this.

Signed-off-by: Boris Ranto <branto@redhat.com>
 2017-09-21 17:24:10 +02:00
amitkuma
c6737e1b8a rgw: Error check on return of read_line() 
** 1406089 Unused value
CID 1406089 (#1 of 1): Unused value (UNUSED_VALUE)
returned_value: Assigning value from this->read_line
(bl, chunk_size, reached_boundary, done) to r here,
but that stored value is overwritten before it can be used.

Signed-off-by: Amit Kumar <amitkuma@redhat.com>
 2017-09-21 19:46:23 +05:30