Commit Graph

38179 Commits

Author SHA1 Message Date
Zhiqiang Wang
8e145e08ed osd: tiering: proxy reads during promote
wip 9980. Do proxy read and async promotion for writeback.

Signed-off-by: Zhiqiang Wang <zhiqiang.wang@intel.com>
2015-01-10 20:35:30 -08:00
Zhiqiang Wang
4dc9990e7e osd: tiering: add cache mode READPROXY
Signed-off-by: Zhiqiang Wang <zhiqiang.wang@intel.com>
2015-01-10 20:35:30 -08:00
Zhiqiang Wang
70d3d08a0b osd: tiering: add proxy read support
wip 9979

Signed-off-by: Zhiqiang Wang <zhiqiang.wang@intel.com>
2015-01-10 20:35:30 -08:00
Sage Weil
9ed2ac780e osd/ReplicatedPG: separate promotion from the triggering op
Remove the triggering op from the internal promote machinery.

We keep the optional op arg to promote_object() only because we may
block on an object other than the original obc.

Signed-off-by: Sage Weil <sage@redhat.com>
2015-01-10 20:35:03 -08:00
Sage Weil
a03d375ed0 osd/ReplicatedPG: pass promote error to all blocked operations
This isn't the most elegant strategy, but it is the best we can do
right now.

Signed-off-by: Sage Weil <sage@redhat.com>
2015-01-10 20:35:03 -08:00
Sage Weil
a4a414944e osd/ReplicatedPG: drop unnecessary cache_mode checks
This currently enumerates all cache modes except none, and we don't
arrive in this function when caching is disabled.  And creating a whiteout
is not cache_mode dependent.  Simplify!

Signed-off-by: Sage Weil <sage@redhat.com>
2015-01-10 20:35:03 -08:00
Sage Weil
2a67b0a090 osd/ReplicatdPG: adjust braces (no semantic change)
Signed-off-by: Sage Weil <sage@redhat.com>
2015-01-10 20:35:03 -08:00
Sage Weil
5361bc043f osd/ReplicatedPG: factor out must_promote case from all cache modes
Signed-off-by: Sage Weil <sage@redhat.com>
2015-01-10 20:35:03 -08:00
Sage Weil
8645ed7664 osd/ReplicatedPG: factor out common exists case from all cache modes
Signed-off-by: Sage Weil <sage@redhat.com>
2015-01-10 20:35:03 -08:00
Sage Weil
1d89f18158 osd/ReplicatedPG: make op argument to promote_object optional
For now, we still always pass it.  In preparation, however, we modify
promote_object() so that it will work when op is null.

Signed-off-by: Sage Weil <sage@redhat.com>
2015-01-10 20:35:03 -08:00
Wido den Hollander
d5e8ef5092 doc: Fix PHP librados documentation 2015-01-10 14:21:27 +01:00
Loic Dachary
b2a3c50508 Merge pull request #3348 from ceph/wip-mon-wishlist
doc: mon janitorial list is now a wishlist

Reviewed-by: Loic Dachary <ldachary@redhat.com>
2015-01-10 13:54:56 +01:00
Joao Eduardo Luis
a16e72cf44 doc: mon janitorial list is now a wishlist
Signed-off-by: Joao Eduardo Luis <joao@redhat.com>
2015-01-10 12:09:16 +00:00
Sage Weil
585efcb7e4 Merge remote-tracking branch 'gh/next' 2015-01-09 21:43:49 -08:00
Sage Weil
58b5e02890 Merge pull request #3327 from ceph/wip-peeringqueue
osd: fix peering queue bug

Reviewed-by: Samuel Just <sjust@redhat.com>
2015-01-09 21:43:04 -08:00
Josh Durgin
18d6a08331 Merge pull request #3344 from ceph/wip-librbd-snap-unprotect
librbd: shadow variable in snap_unprotect

Reviewed-by: Josh Durgin <jdurgin@redhat.com>
2015-01-09 16:56:39 -08:00
Yehuda Sadeh
d375532436 rgw: return InvalidAccessKeyId instead of AccessDenied
Fixes: #10334

Signed-off-by: Yehuda Sadeh <yehuda@redhat.com>
(cherry picked from commit 56af795b10)
2015-01-09 15:13:51 -08:00
Yehuda Sadeh
dd57af2f0a rgw: return SignatureDoesNotMatch instead of AccessDenied
Fixes: #10329

Signed-off-by: Yehuda Sadeh <yehuda@redhat.com>
(cherry picked from commit ef75d720f2)
2015-01-09 15:13:43 -08:00
John Wilkins
c1080b22fa doc: Clean up pool usage.
Signed-off-by: John Wilkins <jowilkin@redhat.com>
2015-01-09 14:54:30 -08:00
John Wilkins
38f424d7a1 doc: Cleanup RGW pool usage.
Signed-off-by: John Wilkins <jowilkin@redhat.com>
2015-01-09 14:54:06 -08:00
Gregory Farnum
8b9c58c4e8 Merge pull request #3341 from liewegas/wip-10504
client: add ceph version to metadata

Reviewed-by: Greg Farnum <gfarnum@redhat.com>
2015-01-09 14:52:19 -08:00
Sage Weil
e6a4ab1391 client: include ceph and git version in client metadata
Fixes: #10504
Signed-off-by: Sage Weil <sage@redhat.com>
2015-01-09 14:41:34 -08:00
Josh Durgin
ac6674c107 Merge pull request #3325 from ceph/wip-nits
allow 'ops' instead of 'dump_ops_in_flight'

Reviewed-by: Josh Durgin <jdurgin@redhat.com>
2015-01-09 14:30:44 -08:00
Josh Durgin
3fb080ffa5 Merge pull request #3250 from ceph/wip-10372
osdc/Objecter: improve pool deletion detection

Reviewed-by: Josh Durgin <jdurgin@redhat.com>
2015-01-09 14:12:22 -08:00
Yehuda Sadeh
f90c48f271 Revert "rgw: switch to new watch/notify API"
This reverts commit dc67cd6960.

Conflicts:
	src/rgw/rgw_rados.cc
2015-01-09 14:18:32 -08:00
Gregory Farnum
6d7acd3710 Merge pull request #3313 from ceph/wip-asok-get-subtrees
Add MDS "get subtrees" asok command

Reviewed-by: Greg Farnum <gfarnum@redhat.com>
Reviewed-by: Yan, Zheng <zyan@redhat.com>
2015-01-09 13:45:26 -08:00
Gregory Farnum
fe56c43927 Merge pull request #3312 from ceph/wip-mdscacheobject-const
mds: support constness in MDSCacheObjects

Reviewed-by: Greg Farnum <gfarnum@redhat.com>
2015-01-09 13:29:59 -08:00
Loic Dachary
7f3497b0a1 Merge pull request #3248 from dachary/wip-table-formatter
add table to Formatter

Reviewed-by: Loic Dachary <ldachary@redhat.com>
2015-01-09 19:34:35 +01:00
John Wilkins
f88781726f doc: Added section to install priorities/preferences.
Signed-off-by: John Wilkins <jowilkin@redhat.com>
2015-01-09 10:26:51 -08:00
Josh Durgin
7d17a17dbd Merge pull request #3324 from mattrichards/rados_translate_op_flag
librados: Translate operation flags from C APIs

Reviewed-by: Josh Durgin <jdurgin@redhat.com>
2015-01-09 09:13:09 -08:00
Sage Weil
7c5faceb6e Merge pull request #3328 from xiaoxichen/memstore_size_u64
Bump memstore_device_bytes from U32 to U64

Reviewed-by: Sage Weil <sage@redhat.com>
2015-01-09 07:52:20 -08:00
Haomai Wang
ca87973f03 Merge pull request #3337 from dachary/wip-10494-disable-unittest-msgr
tests: temporarily disable unittest_msgr
2015-01-09 23:42:05 +08:00
Loic Dachary
55a5c5f292 tests: temporarily disable unittest_msgr
http://tracker.ceph.com/issues/10494 Refs: #10494

Signed-off-by: Loic Dachary <ldachary@redhat.com>
2015-01-09 16:21:19 +01:00
Haomai Wang
d4aec89e13 Merge pull request #3331 from dachary/wip-10493-async-port
msg: initialize AsyncConnection::port

Reviewed-by: Haomai Wang<haomaiwang@gmail.com>
2015-01-09 22:48:19 +08:00
John Spray
26e7d5fab6 mds: add asok command for getting subtreemap
For when we want to inspect this from a test or
during debugging.

Signed-off-by: John Spray <john.spray@redhat.com>
2015-01-09 14:27:15 +00:00
John Spray
1c1897e6ac mds: give CDir a dump() method for JSON output
Useful when listing subtrees via admin socket.

Signed-off-by: John Spray <john.spray@redhat.com>
2015-01-09 14:27:15 +00:00
John Spray
a58d104288 mds: support constness in MDSCacheObjects
So that one can have const CInode and CDir references
from time to time.

Signed-off-by: John Spray <john.spray@redhat.com>
2015-01-09 14:20:59 +00:00
Jason Dillaman
e9033bb121 librbd: shadow variable in snap_unprotect and list_children
The shadow variable prevented snap_unprotect from returning the
correct error return code.

Signed-off-by: Jason Dillaman <dillaman@redhat.com>
2015-01-09 08:28:18 -05:00
Wido den Hollander
cf432fa670 doc: Add Librados PHP documentation 2015-01-09 14:16:08 +01:00
Loic Dachary
9e356cddeb Merge pull request #3332 from dachary/wip-9570-filejournal
journal related cleanups

Reviewed-by: Jianpeng Ma <jianpeng.ma@intel.com>
2015-01-09 13:57:11 +01:00
Andreas Peters
d994e601d2 common: Formatter: cosmetic re-indent
Signed-off-by: Andreas Peters <andreas.joachim.peters@cern.ch>
2015-01-09 13:20:44 +01:00
Andreas-Joachim Peters
e797dcf6e9 common: Formatter: add TableFormatter class
For more human readable and shell parsable output.

Signed-off-by: Andreas Peters <andreas.joachim.peters@cern.ch>
2015-01-09 13:20:43 +01:00
Loic Dachary
bcdbbd560c os: fix confusing indentation in FileJournal::corrupt
Signed-off-by: Loic Dachary <loic-201408@dachary.org>
2015-01-09 13:05:09 +01:00
Loic Dachary
ecc3bcabde os: remove debug message leftover in FileJournal
The len of the buffer shows in the message above anyway.

Signed-off-by: Loic Dachary <loic-201408@dachary.org>
2015-01-09 13:05:09 +01:00
Loic Dachary
63c8dd0a3f msg: initialize AsyncConnection::port
http://tracker.ceph.com/issues/10493 Fixes: #10493

Signed-off-by: Loic Dachary <ldachary@redhat.com>
2015-01-09 11:57:58 +01:00
Xiaoxi Chen
c60f88ba8a Bump memstore_device_bytes from U32 to U64
U32 limit the max size of memstore to a few GB, which
block our test on memstore performance(as a phototype).

Bump it to U64 will suit for more widely usage

Signed-off-by: Xiaoxi Chen <xiaoxi.chen@intel.com>
2015-01-09 16:15:06 +08:00
Sage Weil
20be188d5f osd: assert there is a peering event
This became conditional way back in 12e22b3d44
for unclear reasons.  It probably predates the in_use checks.  In any case,
at this point, we should only arrive here if the PG was queued, implying
that there will always be an event to process.

Signed-off-by: Sage Weil <sage@redhat.com>
2015-01-08 16:14:56 -08:00
Sage Weil
492ccc900c osd: requeue PG when we skip handling a peering event
If we don't handle the event, we need to put the PG back into the peering
queue or else the event won't get processed until the next event is
queued, at which point we'll be processing events with a delay.

The queue_null is not necessary (and is a waste of effort) because the
event is still in pg->peering_queue and the PG is queued.

Note that this only triggers when we exceeed osd_map_max_advance, usually
when there is a lot of peering and recovery activity going on.  A
workaround is to increase that value, but if you exceed osd_map_cache_size
you expose yourself to crache thrashing by the peering work queue, which
can cause serious problems with heavily degraded clusters and bit lots of
people on dumpling.

Backport: giant, firefly
Fixes: #10431
Signed-off-by: Sage Weil <sage@redhat.com>
2015-01-08 16:14:56 -08:00
Matt Richards
49d114f1ff librados: Translate operation flags from C APIs
The operation flags in the public C API are a distinct enum
and need to be translated to Ceph OSD flags, like as happens in
the C++ API. It seems like the C enum and the C++ enum consciously
use the same values, so I reused the C++ translation function.

Signed-off-by: Matthew Richards <mattjrichards@gmail.com>
2015-01-08 13:16:17 -08:00
Sage Weil
1e444e9103 Merge pull request #3321 from cernceph/wip-nobarrier-doc
doc: don't suggest mounting xfs with nobarrier

Reviewed-by: Sage Weil <sage@redhat.com>
2015-01-08 13:06:07 -08:00