Sage Weil
084db49ed3
remove mkcephfs
...
Fixes : #7552
Signed-off-by: Sage Weil <sage@inktank.com>
2014-04-08 14:34:15 -07:00
Samuel Just
4bb0628acf
Merge pull request #1625 from ceph/wip-8019
...
osd: fix journal umount/mount weirdness
Reviewed-by: Samuel Just <sam.just@inktank.com>
2014-04-08 12:45:28 -07:00
Sage Weil
4de49e8676
os/FileStore: reset journal state on umount
...
We observed a sequence like:
- replay journal
- sets JournalingObjectStore applied_op_seq
- umount
- mount
- initiate commit with prevous applied_op_seq
- replay journal
- commit finishes
- on replay commit, we fail assert op > committed_seq
Although strictly speaking the assert failure is harmless here, in general
we should not let state leak through from a previous mount into this
mount or else assertions are in general more difficult to reason about.
Fixes : #8019
Signed-off-by: Sage Weil <sage@inktank.com>
2014-04-08 11:02:40 -07:00
Sage Weil
1cdb7381c7
vstart.sh: make crush location match up with what init-ceph does
...
This makes is to that ./init-ceph restart osd.0 won't modify the CRUSH
tree. And in any case, the localhost/localrack thing we were doing before
was pretty useless.
Signed-off-by: Sage Weil <sage@inktank.com>
2014-04-08 11:02:40 -07:00
Gregory Farnum
ddafcc3710
Merge pull request #1623 from ceph/wip-8026
...
mds: fix shared_ptr MDRequest bugs
Reviewed-by: Greg Farnum <greg@inktank.com>
2014-04-08 10:43:14 -07:00
Sage Weil
667137cc91
Merge pull request #1621 from dachary/wip-7914
...
erasure-code: thread-safe initialization of gf-complete
This looks like a good interim solution until gf-complete exposes a simpler init function
that hides this.
Reviewed-by: Sage Weil <sage@inktank.com>
2014-04-08 10:14:46 -07:00
Yan, Zheng
55cfb142fa
mds: fix shared_ptr MDRequest bugs
...
The main change is use shared_ptr instead of weak_ptr to define
active request map. The reason is that slave request needs to be
preserved until master explicitly finishes it.
Fixes : #8026
Signed-off-by: Yan, Zheng <zheng.z.yan@intel.com>
2014-04-08 16:29:43 +08:00
Loic Dachary
2ba00bd2e8
erasure-code: thread-safe initialization of gf-complete
...
Instead of relying on an implicit initialization happening during
encoding/decoding with galois.c:galois_init_default_field, call
gf.c:gf_init_easy for each w values when the plugin is loaded.
Loading the plugin is protected against race conditions by a lock.
It does not cover all possible uses of gf-complete but it is enough for
the ceph jerasure plugin.
http://tracker.ceph.com/issues/7914 fixes #7914
Signed-off-by: Loic Dachary <loic@dachary.org>
2014-04-08 09:47:33 +02:00
Sage Weil
2a6d962f47
Merge pull request #1610 from ceph/wip-4354-shared_ptr
...
Use shared pointers for Mutations/OpRequests in the MDS
Reviewed-by: Sage Weil <sage@inktank.com>
2014-04-07 21:27:50 -07:00
Sage Weil
9ea31f152d
Merge pull request #1594 from ceph/wip-7958
...
wip 7958
Passed sage-2014-04-07_07:04:02-fs-wip-7958-testing-basic-plana.
Reviewed-by: Sage Weil <sage@inktank.com>
2014-04-07 21:27:04 -07:00
Samuel Just
010dff12c3
osd_types: fix pg_stat_t::encode, object_stat_sum_t::decode version
...
Introduced in a130a4452e
Signed-off-by: Samuel Just <sam.just@inktank.com>
Reviewed-by: Sage Weil <sage@inktank.com>
2014-04-07 17:04:15 -07:00
Greg Farnum
22894c3a12
SimpleLock: Switch MutationRef& for MutationRef in get_xlock()
...
Signed-off-by: Greg Farnum <greg@inktank.com>
2014-04-07 17:00:19 -07:00
Greg Farnum
5f2ccab68f
MDCache: use raw MutationImpl* instead of MutationRef in a few places
...
Avoid the atomic ops necessary when copying a shared_ptr.
Signed-off-by: Greg Farnum <greg@inktank.com>
2014-04-07 17:00:19 -07:00
Greg Farnum
6c3fc3eeea
Locker: use raw MutationImpl* instead of MutationRef in several places
...
Sadly, you can't implicitly convert non-const references to shared pointers, so avoid the atomic ops necessary when copying a shared_ptr.
Signed-off-by: Greg Farnum <greg@inktank.com>
2014-04-07 17:00:19 -07:00
Greg Farnum
920fd6cac1
Locker: use a null_ref instead of NULL
...
Signed-off-by: Greg Farnum <greg@inktank.com>
2014-04-07 17:00:18 -07:00
Greg Farnum
099d1d6f46
Locker: Use MutationRef instead of raw pointers
...
Signed-off-by: Greg Farnum <greg@inktank.com>
2014-04-07 17:00:18 -07:00
Greg Farnum
e800bfa918
Locker: remove Mutation param from xlock_import
...
It's not used.
Signed-off-by: Greg Farnum <greg@inktank.com>
2014-04-07 17:00:18 -07:00
Greg Farnum
1294e55e2a
MDCache: fix users of active_requests for use of shared_ptr
...
Signed-off-by: Greg Farnum <greg@inktank.com>
2014-04-07 17:00:18 -07:00
Greg Farnum
a5c7b27d72
MDCache: use a null_ref instead of NULL in a few places
...
Signed-off-by: Greg Farnum <greg@inktank.com>
2014-04-07 17:00:18 -07:00
Greg Farnum
6b76998d1f
MDCache: use MutationRef instead of raw pointers
...
Signed-off-by: Greg Farnum <greg@inktank.com>
2014-04-07 17:00:18 -07:00
Greg Farnum
a6a0800005
Server: use MutationRef instead of raw pointer
...
Signed-off-by: Greg Farnum <greg@inktank.com>
2014-04-07 17:00:17 -07:00
Greg Farnum
4dedab64c2
MDS: switch cache object classes to use MutationRef instead of raw pointers
...
Signed-off-by: Greg Farnum <greg@inktank.com>
2014-04-07 17:00:17 -07:00
Sage Weil
9a4a429ed2
ceph_test_rados_api_misc: print osd_max_attr_size
...
Very confusing results from this test in bug #8009 .
Signed-off-by: Sage Weil <sage@inktank.com>
2014-04-07 16:31:16 -07:00
Sage Weil
91a9d57429
Merge pull request #1612 from ceph/wip-7919
...
mon: MonCommands: have all 'auth' commands require 'execute' caps
Reviewed-by: Sage Weil <sage@inktank.com>
2014-04-07 16:11:51 -07:00
Sage Weil
aa301b0e49
Merge pull request #1620 from ceph/wip-8003
...
Wip 8003
Reviewed-by: Sage Weil <sage@inktank.com>
2014-04-07 16:09:40 -07:00
Sage Weil
aff20f5eb2
Merge pull request #1611 from ceph/wip-7975
...
osd: disable agent when stats_invalid (post-split)
Reviewed-by: Samuel Just <sam.just@inktank.com>
2014-04-07 15:59:37 -07:00
John Wilkins
4b66868e6f
doc: Removed --stable arg and replaced with --release arg for ceph-deploy.
...
Signed-off-by: John Wilkins <john.wilkins@inktank.com>
2014-04-07 15:49:09 -07:00
Sage Weil
7273d9ea29
osd/ReplicatedPG: warn if invalid stats prevent us from activating agent
...
Signed-off-by: Sage Weil <sage@inktank.com>
2014-04-07 15:40:47 -07:00
Sage Weil
02d7e84d30
osd/ReplicatedPG: dump agent state on pg query
...
Signed-off-by: Sage Weil <sage@inktank.com>
2014-04-07 15:40:47 -07:00
Sage Weil
fa46aec723
osd/ReplicatedPG: kickstart the agent if scrub stats become valid
...
Signed-off-by: Sage Weil <sage@inktank.com>
2014-04-07 15:40:47 -07:00
Sage Weil
51839a33c9
v0.79
...
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.11 (GNU/Linux)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=MSwq
-----END PGP SIGNATURE-----
Merge tag 'v0.79' into firefly
v0.79
2014-04-07 15:04:18 -07:00
Samuel Just
9870b75b62
Merge pull request #1619 from ceph/wip-7659
...
Wip 7659
Reviewed-by: Samuel Just <sam.just@inktank.com>
Reviewed-by: David Zafman <david.zafman@inktank.com>
2014-04-07 14:47:40 -07:00
Samuel Just
07099ecd4c
ReplicatedPG: do not evict head while clone is being promoted
...
Fixes : #8003
Signed-off-by: Samuel Just <sam.just@inktank.com>
2014-04-07 14:34:01 -07:00
Samuel Just
67b2342ab8
ReplicatedPG::trim_object: account evicted prev clone for stats
...
If the previous clone is evicted, we shouldn't adjust the stats to
account for its new clone_overlap value.
Fixes : #7964
Signed-off-by: Samuel Just <sam.just@inktank.com>
2014-04-07 14:34:01 -07:00
Samuel Just
6ff6c19d65
ReplicatedPG::make_writeable: check for evicted clone before adjusting for clone_overlap
...
Fixes : #7964
Signed-off-by: Samuel Just <sam.just@inktank.com>
2014-04-07 14:33:56 -07:00
Sage Weil
afff0ef400
Merge pull request #1617 from ceph/wip-7904
...
Wip 7904
Reviewed-by: Sage Weil <sage@inktank.com>
2014-04-07 14:02:58 -07:00
Sage Weil
008ce85d19
Merge pull request #1614 from ceph/wip-7964
...
Wip 7964
Reviewed-by: Sage Weil <sage@inktank.com>
2014-04-07 14:01:58 -07:00
Sage Weil
71fc7ae306
Merge pull request #1616 from ceph/wip-7916
...
ReplicatedPG: improve get_object_context debugging
Reviewed-by: Sage Weil <sage@inktank.com>
2014-04-07 13:59:22 -07:00
Samuel Just
a8a49a034a
ReplicatedPG: use get_clone_bytes on evict/promote
...
Fixes : #7964
Signed-off-by: Samuel Just <sam.just@inktank.com>
2014-04-07 13:51:18 -07:00
Samuel Just
32862838a1
ReplicatedPG::_scrub: account for clone_overlap on each clone
...
Otherwise, we end up subtracting off clone_overlap for evicted clones
whose sizes we did not add in.
Fixes : #7964
Signed-off-by: Samuel Just <sam.just@inktank.com>
2014-04-07 13:51:18 -07:00
Samuel Just
4d5db0616e
ReplicatedPG::find_object_context: check obs.exists on clone obc before checking snaps
...
Fixes : #7858
Signed-off-by: Samuel Just <sam.just@inktank.com>
2014-04-07 13:51:17 -07:00
Samuel Just
d0e2c98665
ReplicatedPG::finish_promote: add debugging assert for clone_size
...
Signed-off-by: Samuel Just <sam.just@inktank.com>
2014-04-07 13:51:17 -07:00
Sage Weil
1a9952c605
Merge pull request #1613 from ceph/wip-7994
...
OSD: _share_map_outgoing whenever sending a message to a peer
Reviewed-by: Sage Weil <sage@inktank.com>
2014-04-07 10:57:33 -07:00
Joao Eduardo Luis
e213ee1573
qa: workunits: mon: auth_caps.sh: test 'auth' caps requirements
...
Signed-off-by: Joao Eduardo Luis <joao.luis@inktank.com>
2014-04-07 18:30:56 +01:00
Joao Eduardo Luis
85a1cf31e6
mon: MonCommands: have all 'auth' commands require 'execute' caps
...
Earlier patch already have the entity requiring 'execute' caps for
read-only commands. This patch introduces the same requirement for *all*
auth commands, read-only and read-write alike.
While the rationale behind the earlier patch for leaving read-write
operations out of this requirement still holds, we now enforce this to
match compatibility with what was happening back on Dumpling with regard
to the 'execute' cap being required for auth commands. However, it should
be noted that back on Dumpling we were only requiring the 'execute' cap
for auth commands, regardless of read-only or read-write, and no other
caps were required.
Fixes: 7919
Signed-off-by: Joao Eduardo Luis <joao.luis@inktank.com>
2014-04-07 18:17:54 +01:00
Jenkins
4c2d73a509
0.79
2014-04-07 16:48:36 +00:00
Sage Weil
51da3bb07a
mds: fix uninit val in MMDSSlaveRequest
...
Signed-off-by: Sage Weil <sage@inktank.com>
2014-04-07 08:22:32 -07:00
Sage Weil
db281bf51e
Merge pull request #1607 from ceph/wip-7997
...
mon: wait for quorum for MMonGetVersion
Reviewed-by: Joao Eduardo Luis <joao.luis@inktank.com>
2014-04-07 08:11:00 -07:00
Sage Weil
68e27116d0
Merge pull request #1609 from ceph/wip-7739
...
mds: fix some uninitialized message fields
Reviewed-by: Zheng Yan <zheng.z.yan@intel.com>
2014-04-06 17:56:05 -07:00
Sage Weil
76cbd5dd82
mds: fix uninit MMDSSlaveRequest lock_type
...
Signed-off-by: Sage Weil <sage@inktank.com>
2014-04-06 17:36:52 -07:00