Commit Graph

73122 Commits

Author SHA1 Message Date
Greg Farnum
16c39e5f8a mon: stop dereferencing pg_map for object stats in OSDMonitor
Signed-off-by: Greg Farnum <gfarnum@redhat.com>
2017-06-02 12:59:14 -04:00
Greg Farnum
44e0228ad5 mon: yet more OSDMonitor dep-pgmap-ification
Signed-off-by: Greg Farnum <gfarnum@redhat.com>
2017-06-02 12:59:14 -04:00
Greg Farnum
74c1a84a4b mon: more OSDMonitor de-pgmap-ification
Signed-off-by: Greg Farnum <gfarnum@redhat.com>
2017-06-02 12:59:13 -04:00
Greg Farnum
ff1b132f4d mon: OSDMonitor starting to abstract PGMap via PGService
We added an unfortunate get_pg_map() function but it's still easy
to find and puts off some of the logic work. We'll get to it soon.

Signed-off-by: Greg Farnum <gfarnum@redhat.com>
2017-06-02 12:59:13 -04:00
Greg Farnum
f0e74fcf8f mon: FSCommands gets pool stats via PGService
Signed-off-by: Greg Farnum <gfarnum@redhat.com>
2017-06-02 12:59:13 -04:00
Greg Farnum
2c82b9c9fb mon: FSCommands gets pool stats via PGService
Signed-off-by: Greg Farnum <gfarnum@redhat.com>
2017-06-02 12:59:13 -04:00
Greg Farnum
29acfa9349 mon: outline a PGStatService to replace raw PGMap calls
Right now, this is a quick hack that derives from PGMap. Soon, it will
be more interesting.

Signed-off-by: Greg Farnum <gfarnum@redhat.com>
2017-06-02 12:59:10 -04:00
Greg Farnum
759c7d0aee mon: mgr: send a pgmap in reports
This definitely isn't appropriate for long term but is a quick route to testing

Signed-off-by: Greg Farnum <gfarnum@redhat.com>
2017-06-02 12:58:53 -04:00
Greg Farnum
39bd83d63b mon: route manager reports down to the MgrMon
Signed-off-by: Greg Farnum <gfarnum@redhat.com>
2017-06-02 12:58:15 -04:00
Greg Farnum
ff1258e7ef mgr: introduce tick() calls through the system
Still empty for the moment

Signed-off-by: Greg Farnum <gfarnum@redhat.com>
2017-06-02 12:58:15 -04:00
Greg Farnum
efbe4667d9 mgr: create a skeleton MMonMgrReport message type
We'll use this for sending data to the monitor at tick intervals.

Signed-off-by: Greg Farnum <gfarnum@redhat.com>
2017-06-02 12:58:15 -04:00
Greg Farnum
35fd964ec0 mgr: switch from send_beacon() to tick()
We are about to want similar time-based stuff in our other
classes and it's simplest to coalesce them into a single wakeup.

Signed-off-by: Greg Farnum <gfarnum@redhat.com>
2017-06-02 12:58:15 -04:00
Greg Farnum
8d803e0483 MonClient: support "passthrough" of MonMap messages so other Dispatchers can see them
Use this in the Manager so it can notify of new monitor maps.

Signed-off-by: Greg Farnum <gfarnum@redhat.com>
2017-06-02 12:58:14 -04:00
Sage Weil
3bbcdb84e6 mon/MgrMap: prefix summary with epoch; add comma
This matches the others now:

    cluster 0d553373-5de6-4d90-bc60-cfd6f6554631
     health HEALTH_OK
     monmap e2: 1 mons, quorum a
      fsmap e2: 0/0/1 up
        mgr e3: active: x(starting), standbys: y
     osdmap e5: 1 osds: 0 up, 0 in
      pgmap v6: 24 pgs, 3 pools, 0 bytes data, 0 objects
            0 kB used, 0 kB / 0 kB avail
            100.000% pgs inactive
                  24 creating


Signed-off-by: Sage Weil <sage@redhat.com>
2017-06-02 12:58:14 -04:00
Sage Weil
ffced8cfa4 Merge pull request #15427 from tchaikov/wip-vstart-osd-id
vstart: print "start osd.$id" instead of "start osd$id"

Reviewed-by: Sage Weil <sage@redhat.com>
2017-06-02 10:34:45 -05:00
Sage Weil
b98ad7cadb Merge pull request #15217 from badone/wip-async-sleep-timer-fix
osd: Move scrub sleep timer to osdservice

Reviewed-by: Josh Durgin <jdurgin@redhat.com>
2017-06-02 10:32:15 -05:00
Sage Weil
6e406bab6a Merge pull request #15396 from jcsp/wip-ceph-status
mon: revise "ceph status" output

Reviewed-by: Sage Weil <sage@redhat.com>
2017-06-02 10:30:55 -05:00
Sage Weil
b735c5a939 Merge pull request #15011 from xiexingguo/wip-crush-tool
tools/crushtool: replicated-rule API support

Reviewed-by: Sage Weil <sage@redhat.com>
2017-06-02 10:20:44 -05:00
Sage Weil
cd4acd8d01 Merge pull request #15176 from SUSE/wip-mgr-fix-help
mgr,osd:  ceph-mgr --help, unify usage text of other daemons

Reviewed-by: Kefu Chai <kchai@redhat.com>
2017-06-02 10:20:27 -05:00
Sage Weil
6bedb4c35a Merge pull request #15395 from liewegas/wip-log-entity
common/LogEntry: include EntityName in log entries

Reviewed-by: Kefu Chai <kchai@redhat.com>
2017-06-02 10:19:58 -05:00
Sage Weil
ecef6fd839 Merge pull request #15398 from liewegas/wip-bluestore-leak
os/bluestore: fix buffers pinned by indefinitely deferred writes

Reviewed-by: Igor Fedotov <ifedotov@mirantis.com>
2017-06-02 10:19:26 -05:00
Sage Weil
773a0ac2d3 Merge pull request #15402 from liewegas/wip-bluestore-mempools
os/bluestore: improve mempool usage

Reviewed-by: Igor Fedotov <ifedotov@mirantis.com>
2017-06-02 10:19:03 -05:00
Sage Weil
a35cd1cb35 Merge pull request #15403 from liewegas/wip-mempool-dump
mempool: improve dump; fix buffer accounting bugs

Reviewed-by: Kefu Chai <kchai@redhat.com>
2017-06-02 10:18:38 -05:00
Casey Bodley
a8a9969571 Merge pull request #15431 from rzarzynski/wip-rgw-vim-modeline-for-orhpantool
rgw: add the Vim's modeline into rgw_orphan.cc.

Reviewed-by: Casey Bodley <cbodley@redhat.com>
2017-06-02 10:05:17 -04:00
Kefu Chai
d5f178f0e5 Merge pull request #15425 from yaozongyou/fix-build-error
cmake: link against fcgi only if enabled

Reviewed-by: Casey Bodley <cbodley@redhat.com>
Reviewed-by: Kefu Chai <kchai@redhat.com>
2017-06-02 21:56:03 +08:00
Radoslaw Zarzynski
e342c39a57 rgw: add the Vim's modeline into rgw_orphan.cc.
Signed-off-by: Radoslaw Zarzynski <rzarzynski@mirantis.com>
2017-06-02 14:05:39 +02:00
Kefu Chai
fda6e16cf1 vstart: print "start osd.$id" instead of "start osd$id"
"osd.9" is more consistent with other places where an osd is referenced.

Signed-off-by: Kefu Chai <kchai@redhat.com>
2017-06-02 15:10:45 +08:00
Jos Collin
371f227f10 Merge pull request #15408 from yonghengdexin735/wip-zzz-add_const
include/lru.h: add const to member functions

Reviewed-by: Jos Collin <jcollin@redhat.com>
2017-06-02 11:11:58 +05:30
Yao Zongyou
3b7ecf37e9 Fix cmake build error.
Signed-off-by: Yao Zongyou <yaozongyou@vip.qq.com>
2017-06-02 12:11:59 +08:00
Sage Weil
17751bac69 Merge pull request #15420 from yuriw/wip_remove_default_idle_timeout
qa: Removed all 'default_idle_timeout' due to chnage in rwg task
2017-06-01 22:08:31 -05:00
Sage Weil
1244c0a501 Merge pull request #15397 from dzafman/wip-20089
osd: Object level shard errors are tracked and used if no auth available

Reviewed-by: Josh Durgin <jdurgin@redhat.com>
2017-06-01 21:40:29 -05:00
Sage Weil
ea5b054e65 qa/workunits/cephtool/test.sh: adjust watch test
Signed-off-by: Sage Weil <sage@redhat.com>
2017-06-01 22:22:24 -04:00
Sage Weil
6912c4f9ea Merge pull request #15417 from smithfarm/wip-cond-epoch-2
build/ops: rpm: move _epoch_prefix below Epoch definition
2017-06-01 21:14:13 -05:00
yonghengdexin735
75c5c2b782 include/lru: add const to member function
Signed-off-by: yonghengdexin735 <zhang.zezhu@zte.com.cn>
2017-06-02 07:55:19 +08:00
Yuri Weinstein
02242ea48e Removed all 'default_idle_timeout' due to chnage in rwg task
8c74c8a639 (diff-995b04809fcabacc3e3ecfaea903a41aL539)

Signed-off-by: Yuri Weinstein <yweinste@redhat.com>
2017-06-01 14:01:30 -07:00
Nathan Cutler
3e30232607 build/ops: rpm: move _epoch_prefix below Epoch definition
Signed-off-by: Nathan Cutler <ncutler@suse.com>
2017-06-01 21:37:09 +02:00
Sage Weil
b5da257b2a Merge pull request #15400 from yehudasa/wip-20128
qa/rgw_snaps: move default_idle_timeout config under the client
2017-06-01 14:22:39 -05:00
David Zafman
1cacbea763 osd: Object level shard errors are tracked and used if no auth available
Shards with object mismatch are tracked to mark them inconsistent
Fix test because storing omap_digest in object_info not behaving as before

Fixes: http://tracker.ceph.com/issues/20089

Signed-off-by: David Zafman <dzafman@redhat.com>
2017-06-01 10:46:48 -07:00
David Zafman
955deac56f osd-scrub-repair: Fix broken test and correctly handle ec overwrite testing
Caused by: 3ca750d41d

Signed-off-by: David Zafman <dzafman@redhat.com>
2017-06-01 10:09:29 -07:00
Sage Weil
8bcf3e5781 Merge pull request #15212 from neha-ojha/wip-fix-recovery-sleep-configurable
osd: Implement asynchronous recovery sleep

Reviewed-by: Josh Durgin <jdurgin@redhat.com>
2017-06-01 11:54:58 -05:00
Sage Weil
703055b0c0 Merge pull request #15262 from Adirl/revert_cm
msg/async: Revert RDMA-CM

Reviewed-by: Haomai Wang <haomai@xsky.com>
2017-06-01 11:54:33 -05:00
Sage Weil
703125eb34 Merge pull request #15307 from ovh/bp-64bit-mem-is-zero
inline_memory: optimized mem_is_zero for non-x64

Reviewed-by: Sage Weil <sage@redhat.com>
2017-06-01 11:54:19 -05:00
Sage Weil
268d5cdeba Merge pull request #15340 from wjwithagen/wip-wjw-libcommon
cmake: do not link libcommon against some libs

Reviewed-by: Kefu Chai <kchai@redhat.com>
2017-06-01 11:53:54 -05:00
Sage Weil
f6220d1504 Merge pull request #15342 from chardan/jfw-wip-add_VLA_warning
build: Adds C++ warning flag for C Variable-Length Arrays.

Reviewed-by: Sage Weil <sage@redhat.com>
2017-06-01 11:53:32 -05:00
Sage Weil
aee60f2d92 Merge pull request #15364 from joscollin/wip-cleanup-putto-operator-const
src: put-to operator function - const input cleanup

Reviewed-by: Greg Farnum <gfarnum@redhat.com>
2017-06-01 11:52:57 -05:00
Sage Weil
d3887bfeda Merge pull request #15372 from ganeshmaharaj/reuse-isal
erasure-code: Remove duplicate of isa-l files

Reviewed-by: Loic Dachary <ldachary@redhat.com>
2017-06-01 11:52:40 -05:00
Sage Weil
4e160f20e4 os/bluestore: release aios and pinned buffers on io complete
Once we're done with our IO, clear the aio list so that the pinned buffers
are unpinned.  This ensures we release memory quickly, even if the
TransContext sticks around for a while (e.g., in the osr q).

Signed-off-by: Sage Weil <sage@redhat.com>
2017-06-01 12:28:44 -04:00
Sage Weil
d82af301c8 os/bluestore: submit deferred if txc cleanup is blocked
If we have a single deferred write, and then a uniform workload with *no*
deferred writes, we will never actually submit it.  Meanwhile, the txc is
stuck on the osr q and nothing ever gets retired.

Simple fix is to submit any deferred ops if the osr queue is blocked by
a queued deferred write and the osr queue length is above some
threshold.  This prevents memory from being pinned indefinitely.

Signed-off-by: Sage Weil <sage@redhat.com>
2017-06-01 12:28:43 -04:00
Sage Weil
aff9e37e83 Merge pull request #15209 from liewegas/wip-sortbitwise-notes
doc/release-notes: update which jewel version does sortbitwise warning
2017-06-01 11:26:06 -05:00
Sage Weil
ff2d625763 Merge pull request #15274 from liewegas/wip-auto-ccache
do_cmake.sh: enable ccache if installed

Reviewed-by: Kefu Chai <kchai@redhat.com>
2017-06-01 11:25:36 -05:00