Commit Graph

84313 Commits

Author SHA1 Message Date
Neha Ojha
bec9aaf744 ECBackend: send correct stats to async_recovery_targets
Signed-off-by: Neha Ojha <nojha@redhat.com>
2018-03-15 11:13:34 -07:00
Neha Ojha
7f6f4f90fe qa: modify TEST_recovery_sizeup() to handle async recovery
Signed-off-by: Neha Ojha <nojha@redhat.com>
2018-03-15 11:13:34 -07:00
Neha Ojha
6ea566d3e0 PG: want set size should not be below min_size
Signed-off-by: Neha Ojha <nojha@redhat.com>
2018-03-15 11:13:34 -07:00
Neha Ojha
d25ad0cfa4 PG: clear missing_loc only during async_recovery
Signed-off-by: Neha Ojha <nojha@redhat.com>
2018-03-15 11:13:34 -07:00
Neha Ojha
8cf08f3ab8 PG: add location to missing_loc only during async_recovery
Signed-off-by: Neha Ojha <nojha@redhat.com>
2018-03-15 11:13:34 -07:00
Neha Ojha
b37d895a3a PG: clear out old missing_loc in issue_repop
Signed-off-by: Neha Ojha <nojha@redhat.com>
2018-03-15 11:13:34 -07:00
Neha Ojha
20f90c5608 PG: handle degraded object in _rollback_to()
Signed-off-by: Neha Ojha <nojha@redhat.com>
2018-03-15 11:13:34 -07:00
Neha Ojha
ae6e001c5f ECBackend: update pg_log missing for async_recovery_targets
Signed-off-by: Neha Ojha <nojha@redhat.com>
2018-03-15 11:13:33 -07:00
Neha Ojha
0a1e0cbee6 PG: skip primary in peer_missing in rebuild()
Signed-off-by: Neha Ojha <nojha@redhat.com>
2018-03-15 11:13:33 -07:00
Sage Weil
52db09afe0 Merge PR #20880 into wip-sage-testing-20180315.163409
* refs/pull/20880/head:
	"Correct" license from LGPL2 to LGPL2.1

Reviewed-by: John Spray <john.spray@redhat.com>
Reviewed-by: Gregory Farnum <gfarnum@redhat.com>
Reviewed-by: Abhishek Lekshmanan <abhishek.lekshmanan@gmail.com>
Reviewed-by: Patrick Donnelly <pdonnell@redhat.com>
Reviewed-by: Greg Farnum <gfarnum@redhat.com>
2018-03-15 11:34:18 -05:00
Andrew Schoen
e27f9fb281
Merge pull request #20910 from ceph/fix_callback_plugins_dir
ceph-volume: fix action plugins path in tests

Reviewed-by: Andrew Schoen <aschoen@redhat.com>
2018-03-15 16:14:36 +01:00
Kefu Chai
363e94895b
Merge pull request #20897 from ifed01/wip-ifed-discard-on-mkfs
os/bluestore: do SSD discard on mkfs

Reviewed-by: Sage Weil <sage@redhat.com>
2018-03-15 23:09:29 +08:00
Kefu Chai
872df82aa8
Merge pull request #20895 from tchaikov/wip-drop-unused-options
common/options: drop unused options

Reviewed-by: John Spray <john.spray@redhat.com>
2018-03-15 23:08:24 +08:00
Kefu Chai
c9a201a5f9
Merge pull request #20445 from aclamk/wip-assert-smaller
common: Make code to invoke assert() smaller.

Reviewed-by: Kefu Chai <kchai@redhat.com>
2018-03-15 23:07:35 +08:00
Kefu Chai
0b7673dd93
Merge pull request #20887 from liewegas/wip-kill-backlog
osd/osd_types: remove backlog type for pg_log_entry_t

Reviewed-by: xie xingguo <xie.xingguo@zte.com.cn>
2018-03-15 23:07:02 +08:00
Ilya Dryomov
e28e0c617a osd/OSDMap: fix HAVE_FEATURE logic in encode()
Currently clients that don't have SERVER_LUMINOUS end up being fed
v6 (i.e. luminous) encoding because they also don't have SERVER_MIMIC.
This was introduced in commit 553048fbf9 ("osd/OSDMap: track newly
removed and purged snaps in each epoch").

Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2018-03-15 15:37:29 +01:00
Sage Weil
69765a788e qa/standalone/osd/repro_long_log: no-mon-config for cot
Signed-off-by: Sage Weil <sage@redhat.com>
2018-03-15 08:42:14 -05:00
Sage Weil
ef7eaa48be qa/suites/rados: fewer msgr failures
500 is a bit much... e.g., enough to hit timeouts forming mon quorum.

Signed-off-by: Sage Weil <sage@redhat.com>
2018-03-15 08:39:28 -05:00
Sage Weil
85f980223d osd: debug dispatch_context cases where queries not sent
Signed-off-by: Sage Weil <sage@redhat.com>
2018-03-15 08:32:14 -05:00
Tiago Melo
a2033e0845 mgr/dashboard_v2: fix test_perf_counters_mgr_get
Now that we can obtain perf counters for the mgr, we need to assert that we get
more than 0 counters and that they have the desired structure.

Fixes: https://tracker.ceph.com/issues/23378

Signed-off-by: Tiago Melo <tmelo@suse.com>
2018-03-15 12:50:58 +00:00
Shen-Ta Hsieh(BestSteve)
dbc240fff2
memstore: fix wrong use of lock_guard
std::lock_guard need a name to make RAII works properly
Signed-off-by: Shen-Ta Hsieh <ibmibmibm.tw@gmail.com>
2018-03-15 20:36:48 +08:00
Guillaume Abrioux
9fdc56717e ceph-volume: fix action plugins path in tox
because of this commit [1] in ceph-ansible tests in ceph-volume needs to
be modified accordingly.

[1] 60d4b75f51

Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
2018-03-15 11:18:55 +01:00
Gu Zhongyan
984dd9242f mon/ConfigMonitor: clean up prepare_command()
no more paxos->get_pending_transaction()

Signed-off-by: Gu Zhongyan <guzhongyan@360.cn>
2018-03-15 17:35:02 +08:00
Kefu Chai
2f05d2d399
Merge pull request #20836 from xiexingguo/wip-balancer-05
pybind/mgr/balancer: add sanity check against empty adjusted_map

Reviewed-by: Sage Weil <sage@redhat.com>
2018-03-15 14:28:46 +08:00
Kefu Chai
11531fe81c
Merge pull request #20857 from zealoussnow/wip-leozhang
osd/bluestore: delete unused variable in KernelDevice

Reviewed-by: xie xingguo <xie.xingguo@zte.com.cn>
2018-03-15 14:27:36 +08:00
Kefu Chai
b1abdd48e6
Merge pull request #20690 from myoungwon/wip-manifest-num-counter
osd: add num_object_manifest

Reviewed-by: Sage Weil <sage@redhat.com>
Reviewed-by: Kefu Chai <kchai@redhat.com>
2018-03-15 14:26:28 +08:00
Kefu Chai
2996068738
Merge pull request #20652 from shangfufei/wip-common
src/common: update some ms_* options to be more consistent

Reviewed-by: Haomai Wang <haomai@xsky.com>
2018-03-15 14:24:22 +08:00
Kefu Chai
5002840900
Merge pull request #20886 from liewegas/wip-kill-min-max
include: kill MIN and MAX macros

Reviewed-by: Greg Farnum <gfarnum@redhat.com>
2018-03-15 14:22:52 +08:00
Kefu Chai
e078ec6004
Merge pull request #20321 from jcsp/wip-22918
mgr: initialize PyModuleRegistry sooner

Reviewed-by: Kefu Chai <kchai@redhat.com>
2018-03-15 14:18:45 +08:00
Kefu Chai
b17ff3c54b mgr/MgrClient: guard send_pgstats() with lock
ms_handle_reset() races with send_pgstats(), and the latter could
reference a session with a null connection. so we need to guard
send_pgstats() with a lock.

Fixes: http://tracker.ceph.com/issues/23370
Signed-off-by: Kefu Chai <kchai@redhat.com>
2018-03-15 13:33:56 +08:00
Brad Hubbard
0fe0c1173b log: Fix AddressSanitizer: new-delete-type-mismatch
If you directly call the operator new function, you must also directly
 call the operator delete function, and must manually call the
 destructor as well.

Fixes: http://tracker.ceph.com/issues/23324

Signed-off-by: Brad Hubbard <bhubbard@redhat.com>
2018-03-15 12:00:31 +10:00
Neha Ojha
939c9d03d6 PG: bug fix in is_degraded_or_backfilling_object()
Signed-off-by: Neha Ojha <nojha@redhat.com>
2018-03-14 15:56:05 -07:00
Neha Ojha
fa01affde1 PG: update pg_log missing for async_recovery_targets
Signed-off-by: Neha Ojha <nojha@redhat.com>
2018-03-14 15:56:05 -07:00
Neha Ojha
853655a559 PG: reset async_recovery_targets after recovery
Signed-off-by: Neha Ojha <nojha@redhat.com>
2018-03-14 15:56:05 -07:00
Neha Ojha
13d7093ec0 PG: call choose_acting when there are async_recovery_targets
Signed-off-by: Neha Ojha <nojha@redhat.com>
2018-03-14 15:56:05 -07:00
Neha Ojha
90b9f6d825 PG: mark recovery started when fail to get recovery_read_lock
Signed-off-by: Neha Ojha <nojha@redhat.com>
2018-03-14 15:56:05 -07:00
Neha Ojha
bb0ad5d8f8 PG: set delete flag in needs_recovery_map for deletes
Signed-off-by: Neha Ojha <nojha@redhat.com>
2018-03-14 15:56:05 -07:00
Neha Ojha
d16f00c9b0 PG: update missing_loc for async_recovery_targets
Signed-off-by: Neha Ojha <nojha@redhat.com>
2018-03-14 15:56:05 -07:00
Neha Ojha
1e2e58c730 PG: do not block writes on async_recovery_targets
Signed-off-by: Neha Ojha <nojha@redhat.com>
2018-03-14 15:56:05 -07:00
Neha Ojha
3b5323d149 PG: modify should_send_op to handle async recovery
Signed-off-by: Neha Ojha <nojha@redhat.com>
2018-03-14 15:56:05 -07:00
Neha Ojha
af83cf9f7b osd: rename ECSubWrite parameter backfill to backfill_or_async_recovery
Signed-off-by: Neha Ojha <nojha@redhat.com>
2018-03-14 15:56:05 -07:00
Neha Ojha
4e78d8055b PG: cleanup and fixes for async recovery
Signed-off-by: Neha Ojha <nojha@redhat.com>
2018-03-14 15:56:04 -07:00
Neha Ojha
322a00b0e6 osd: introduce acting_recovery_backfill shards
Signed-off-by: Neha Ojha <nojha@redhat.com>
2018-03-14 15:56:04 -07:00
Neha Ojha
c6dfeb3253 PG: initial changes required for async recovery
Signed-off-by: Neha Ojha <nojha@redhat.com>
2018-03-14 15:56:04 -07:00
Josh Durgin
360bc69bcf
Merge pull request #20901 from neha-ojha/wip-fix-mgr
qa/tasks/ceph_manager: use set_config on revived osd

Reviewed-by: Josh Durgin <jdurgin@redhat.com>
2018-03-14 15:27:39 -07:00
Neha Ojha
e3899dc901 qa/tasks/ceph_manager: use set_config on revived osd
Signed-off-by: Neha Ojha <nojha@redhat.com>
2018-03-14 12:37:56 -07:00
Andrew Schoen
a8cd8b2625
Merge pull request #20824 from ceph/wip-rm23259
ceph-volume process the abspath of the executable first

Reviewed-by: Andrew Schoen <aschoen@redhat.com>
2018-03-14 20:21:37 +01:00
David Zafman
8a7e6c2349
Merge pull request #20220 from dzafman/wip-calc-stats3
osd: Improve recovery stat handling by using peer_missing and missing_loc info

Reviewed-by: Sage Weil <sage@redhat.com>
2018-03-14 11:07:44 -07:00
David Zafman
af85f3cc48 test: osd-backfill-stats.sh parallel osd-recovery-stats.sh check() changes
Signed-off-by: David Zafman <dzafman@redhat.com>
2018-03-14 10:07:11 -07:00
David Zafman
acc1f80684 test: Use "(est)" in log message when an osd doesn't have peer_missing
Consolidate check() code and common script code
TEST_recovery_multi() wasn't reliable due to delayed peer_missing

Signed-off-by: David Zafman <dzafman@redhat.com>
2018-03-14 10:07:11 -07:00