]> git.apps.os.sepia.ceph.com Git - ceph.git/log
ceph.git
10 years agolibrbd: update in-memory object map before on-disk update completes
Jason Dillaman [Fri, 20 Mar 2015 14:40:22 +0000 (10:40 -0400)]
librbd: update in-memory object map before on-disk update completes

When multiple IO updates are occurring within the same object, this
will allow only a single object map update request to be sent to the
OSD.

Signed-off-by: Jason Dillaman <dillaman@redhat.com>
10 years agocls_rbd: treat zero-byte object maps as missing
Jason Dillaman [Fri, 20 Mar 2015 03:29:44 +0000 (23:29 -0400)]
cls_rbd: treat zero-byte object maps as missing

Acquiring the lock on a missing object map will create an
empty object. Treat the empty object as a non-existant
object map to support rebuilding corrupt object maps while
holding the exclusive lock.

Signed-off-by: Jason Dillaman <dillaman@redhat.com>
10 years agorbd: add object map rebuild command
Jason Dillaman [Thu, 12 Mar 2015 16:10:32 +0000 (12:10 -0400)]
rbd: add object map rebuild command

An invalid object map can now be rebuilt via the rbd CLI.

Signed-off-by: Jason Dillaman <dillaman@redhat.com>
10 years agolibrbd: add rebuild_object_map to public API
Jason Dillaman [Fri, 20 Mar 2015 01:19:22 +0000 (21:19 -0400)]
librbd: add rebuild_object_map to public API

Signed-off-by: Jason Dillaman <dillaman@redhat.com>
10 years agolibrados_test_stub: add another overload of aio_operate
Jason Dillaman [Thu, 19 Mar 2015 20:28:34 +0000 (16:28 -0400)]
librados_test_stub: add another overload of aio_operate

Signed-off-by: Jason Dillaman <dillaman@redhat.com>
10 years agolibrbd: connect async rebuild object map to state machine
Jason Dillaman [Thu, 19 Mar 2015 20:26:40 +0000 (16:26 -0400)]
librbd: connect async rebuild object map to state machine

Signed-off-by: Jason Dillaman <dillaman@redhat.com>
10 years agolibrbd: require callers to ObjectMap::aio_update to acquire lock
Jason Dillaman [Thu, 19 Mar 2015 19:35:59 +0000 (15:35 -0400)]
librbd: require callers to ObjectMap::aio_update to acquire lock

This is needed to allow an atomic compare and update operation
from the rebuild object map utility.

Signed-off-by: Jason Dillaman <dillaman@redhat.com>
10 years agolibrbd: added RebuildObjectMapRequest state machine
Jason Dillaman [Thu, 19 Mar 2015 19:33:07 +0000 (15:33 -0400)]
librbd: added RebuildObjectMapRequest state machine

It will verify the object map is properly sized, verify
the existence of each object within the image (snapshot),
and clear the invalid object map flag once complete.

Signed-off-by: Jason Dillaman <dillaman@redhat.com>
10 years agolibrbd: correct basic object map errors during refresh
Jason Dillaman [Fri, 20 Mar 2015 16:16:14 +0000 (12:16 -0400)]
librbd: correct basic object map errors during refresh

If the object map is corrupt on-disk or too small for the image,
correct these basic issues as soon as possible.  The object map
is still flagged as invalid, but there will be less required repair
work if future IO is able to properly update the object map.

Signed-off-by: Jason Dillaman <dillaman@redhat.com>
10 years agolibrbd: use generic helper for issuing async requests
Jason Dillaman [Wed, 18 Mar 2015 15:51:47 +0000 (11:51 -0400)]
librbd: use generic helper for issuing async requests

resize, flatten, and rebuild object map now use the same
bootstrap code for sending the request to the remote lock owner
or executing the request locally.

Signed-off-by: Jason Dillaman <dillaman@redhat.com>
10 years agolibrbd: connect ImageWatch rebuild request to async rebuild method
Jason Dillaman [Wed, 18 Mar 2015 15:08:23 +0000 (11:08 -0400)]
librbd: connect ImageWatch rebuild request to async rebuild method

Signed-off-by: Jason Dillaman <dillaman@redhat.com>
10 years agolibrbd: add preliminary methods for rebuilding object map
Jason Dillaman [Wed, 18 Mar 2015 15:07:35 +0000 (11:07 -0400)]
librbd: add preliminary methods for rebuilding object map

Signed-off-by: Jason Dillaman <dillaman@redhat.com>
10 years agolibrbd: add hooks for rebuild object map to ImageWatcher
Jason Dillaman [Wed, 18 Mar 2015 14:38:09 +0000 (10:38 -0400)]
librbd: add hooks for rebuild object map to ImageWatcher

Signed-off-by: Jason Dillaman <dillaman@redhat.com>
10 years agolibrbd: add new notify payload for rebuild object map
Jason Dillaman [Wed, 18 Mar 2015 14:21:17 +0000 (10:21 -0400)]
librbd: add new notify payload for rebuild object map

The request will be sent to the client which owns the exclusive
lock to rebuild the object map for the current image HEAD.

Signed-off-by: Jason Dillaman <dillaman@redhat.com>
10 years agoMerge pull request #4306 from dachary/wip-ec-troubleshooting
Gregory Farnum [Thu, 9 Apr 2015 22:41:05 +0000 (15:41 -0700)]
Merge pull request #4306 from dachary/wip-ec-troubleshooting

doc: an erasure code crush ruleset can be dynamically modified

Reviewed-by: Greg Farnum <gfarnum@redhat.com>
10 years agoMerge pull request #4320 from liewegas/wip-crush-straw2-div0
Loic Dachary [Thu, 9 Apr 2015 22:05:36 +0000 (00:05 +0200)]
Merge pull request #4320 from liewegas/wip-crush-straw2-div0

crush/mapper: fix divide-by-0 in straw2

Reviewed-by: Yann Dupont <yann@objoo.org>
Reviewed-by: Loic Dachary <ldachary@redhat.com>
10 years agocrush/mapper: fix divide-by-0 in straw2 4320/head 4321/head
Sage Weil [Thu, 9 Apr 2015 21:42:34 +0000 (14:42 -0700)]
crush/mapper: fix divide-by-0 in straw2

If the item weight is 0 we don't want to divide; instead draw a minimal
value.

Fixes: #11357
Reported-by: Yann Dupont <yd@objoo.org>
Tested-by: Yann Dupont <yd@objoo.org>
Signed-off-by: Sage Weil <sage@redhat.com>
10 years agoMerge pull request #4174 from trociny/wip-10976.master
Samuel Just [Thu, 9 Apr 2015 18:25:03 +0000 (11:25 -0700)]
Merge pull request #4174 from trociny/wip-10976.master

osd: fix PG::all_unfound_are_queried_or_lost for non-existent osds

Reviewed-by: Kefu Chai <tchaikov@gmail.com
Reviewed-by: Samuel Just <sjust@redhat.com>
10 years agoMerge pull request #4111 from wonzhq/hitsets-check
Samuel Just [Thu, 9 Apr 2015 18:22:55 +0000 (11:22 -0700)]
Merge pull request #4111 from wonzhq/hitsets-check

osd/ReplicatedPG: correct the checking if the promoting object is in other hit sets

Reviewed-by: Kefu Chai <tchaikov@gmail.com>
10 years agoMerge pull request #4136 from XinzeChi/wip-pg-delete-sleep
Samuel Just [Thu, 9 Apr 2015 18:21:37 +0000 (11:21 -0700)]
Merge pull request #4136 from XinzeChi/wip-pg-delete-sleep

osd: return fast if PG::deleting is true in snap_trimmer and PG::scrub

Reviewed-by: Sage Weil <sage@redhat.com>
10 years agoMerge pull request #4178 from XinzeChi/wip-tracker-message
Samuel Just [Thu, 9 Apr 2015 18:21:20 +0000 (11:21 -0700)]
Merge pull request #4178 from XinzeChi/wip-tracker-message

osd: more useful message to find out potential unhealth osd

Reviewed-by: Samuel Just <sjust@redhat.com>
Reviewed-by: Sage Weil <sage@redhat.com>
10 years agoMerge pull request #4155 from wonzhq/prdop
Samuel Just [Thu, 9 Apr 2015 18:20:24 +0000 (11:20 -0700)]
Merge pull request #4155 from wonzhq/prdop

osd/ReplicatedPG: don't check order in finish_proxy_read

Reviewed-by: Kefu Chai <tchaikov@gmail.com>
Reviewed-by: Sage Weil <sage@redhat.com>
10 years agoMerge pull request #4139 from Ved-vampir/oscnt
Samuel Just [Thu, 9 Apr 2015 18:19:23 +0000 (11:19 -0700)]
Merge pull request #4139 from Ved-vampir/oscnt

Perf counters descriptions

Reviewed-by: Greg Farnum <gfarnum@redhat.com>
Reviewed-by: Samuel Just <sjust@redhat.com>
10 years agoMerge pull request #3928 from dmitryya/rados_bench_with_iops_metric
Samuel Just [Thu, 9 Apr 2015 18:17:55 +0000 (11:17 -0700)]
Merge pull request #3928 from dmitryya/rados_bench_with_iops_metric

rados bench: Add IOPS metric to `rados bench'

Reviewed-by: Loic Dachary <ldachary@redhat.com>
10 years agoMerge pull request #4253 from majianpeng/scrub-fix2
Samuel Just [Thu, 9 Apr 2015 18:16:56 +0000 (11:16 -0700)]
Merge pull request #4253 from majianpeng/scrub-fix2

Scrub bug fix

Reviewed-by: Samuel Just <sjust@redhat.com>
10 years agoMerge pull request #4224 from rzarzynski/wip-10682_bucketlisting
Yehuda Sadeh [Thu, 9 Apr 2015 17:02:08 +0000 (10:02 -0700)]
Merge pull request #4224 from rzarzynski/wip-10682_bucketlisting

rgw: add support for end_marker parameter for GET on Swift container.

Reviewed-by: Yehuda Sadeh <yehuda@redhat.com>
10 years agoMerge pull request #4305 from Ved-vampir/rgw
Yehuda Sadeh [Thu, 9 Apr 2015 16:55:25 +0000 (09:55 -0700)]
Merge pull request #4305 from Ved-vampir/rgw

rgw: add perf counters descriptions

Reviewed-by: Yehuda Sadeh <yehuda@redhat.com>
10 years agoMerge pull request #4240 from 9seconds/post_check_quotas
Yehuda Sadeh [Thu, 9 Apr 2015 16:51:35 +0000 (09:51 -0700)]
Merge pull request #4240 from 9seconds/post_check_quotas

rgw: Quota not respected in POST object

Reviewed-by: Yehuda Sadeh <yehuda@redhat.com>
10 years agoMerge pull request #4268 from guangyy/wip-11322
Yehuda Sadeh [Thu, 9 Apr 2015 16:50:02 +0000 (09:50 -0700)]
Merge pull request #4268 from guangyy/wip-11322

rgw : Issue AIO for next chunk first before flush the (cached) data.

Reviewed-by: Yehuda Sadeh <yehuda@redhat.com>
10 years agoMerge pull request #4315 from dachary/wip-workflow
Loic Dachary [Thu, 9 Apr 2015 12:39:14 +0000 (14:39 +0200)]
Merge pull request #4315 from dachary/wip-workflow

doc: development workflows

Reviewed-by: Alistair Israel <aisrael@gmail.com>
10 years agodoc: when and why publish a point release 4315/head
Loic Dachary [Thu, 9 Apr 2015 11:53:13 +0000 (13:53 +0200)]
doc: when and why publish a point release

Explain who decides when to cut a point release and the tradeoffs
involved.

Signed-off-by: Loic Dachary <ldachary@redhat.com>
10 years agodoc: add Hammer to the release timeline
Loic Dachary [Thu, 9 Apr 2015 11:51:15 +0000 (13:51 +0200)]
doc: add Hammer to the release timeline

Signed-off-by: Loic Dachary <ldachary@redhat.com>
10 years agoMerge pull request #4233 from ceph/wip-librbd-aio-request
Josh Durgin [Wed, 8 Apr 2015 22:19:24 +0000 (15:19 -0700)]
Merge pull request #4233 from ceph/wip-librbd-aio-request

librbd: simplify AioRequest constructor parameters

Reviewed-by: Josh Durgin <jdurgin@redhat.com>
10 years agoMerge remote-tracking branch 'origin/wip-2862'
Josh Durgin [Wed, 8 Apr 2015 20:51:31 +0000 (13:51 -0700)]
Merge remote-tracking branch 'origin/wip-2862'

Reviewed-by: Josh Durgin <jdurgin@redhat.com>
Conflicts:
src/common/ceph_argparse.cc
src/rbd.cc

10 years agoMerge pull request #4307 from dzafman/wip-dz-cmd-fix
Dan Mick [Wed, 8 Apr 2015 20:28:12 +0000 (13:28 -0700)]
Merge pull request #4307 from dzafman/wip-dz-cmd-fix

osd: Fix wrong usage for "ceph tell osd.* debug dump_missing"

Reviewed-by: Dan Mick <dmick@redhat.com>
10 years agoosd: Fix wrong usage for "ceph tell osd.* debug dump_missing" 4307/head
David Zafman [Wed, 8 Apr 2015 19:45:37 +0000 (12:45 -0700)]
osd: Fix wrong usage for "ceph tell osd.* debug dump_missing"

Signed-off-by: David Zafman <dzafman@redhat.com>
10 years agoMerge pull request #4304 from Ved-vampir/mds
John Spray [Wed, 8 Apr 2015 18:12:05 +0000 (19:12 +0100)]
Merge pull request #4304 from Ved-vampir/mds

mds: add perf counters descriptions

Reviewed-by: John Spray <john.spray@redhat.com>
10 years agorgw: add perf counters descriptions 4305/head
Ved-vampir [Wed, 8 Apr 2015 17:27:16 +0000 (20:27 +0300)]
rgw: add perf counters descriptions

Sign-off-by: Alyona Kiseleva <akiselyova@mirantis.com>
10 years agodoc: an erasure code crush ruleset can be dynamically modified 4306/head
Loic Dachary [Wed, 8 Apr 2015 17:33:05 +0000 (19:33 +0200)]
doc: an erasure code crush ruleset can be dynamically modified

Signed-off-by: Loic Dachary <ldachary@redhat.com>
10 years agomds: add perf counters descriptions 4304/head
Ved-vampir [Wed, 8 Apr 2015 17:08:09 +0000 (20:08 +0300)]
mds: add perf counters descriptions

Sign-off-by: Alyona Kiseleva <akiselyova@mirantis.com>
10 years agodoc/release-notes: note about SHEC
Sage Weil [Wed, 8 Apr 2015 16:52:59 +0000 (09:52 -0700)]
doc/release-notes: note about SHEC

Signed-off-by: Sage Weil <sage@redhat.com>
10 years agoMerge pull request #4274 from liewegas/wip-osd-cleanup
Kefu Chai [Wed, 8 Apr 2015 15:56:54 +0000 (23:56 +0800)]
Merge pull request #4274 from liewegas/wip-osd-cleanup

osd: remove some dead(ish) code

Reviewed-by: Kefu Chai <kchai@redhat.com>
10 years agoMerge pull request #4301 from nilamdyuti/wip-doc-ceph-object-gateway
Ken Dreyer [Wed, 8 Apr 2015 13:32:41 +0000 (07:32 -0600)]
Merge pull request #4301 from nilamdyuti/wip-doc-ceph-object-gateway

doc: Corrects rgw.conf file path for Debian-based and RPM-based distros
in radosgw man page.

Reviewed-by: Ken Dreyer <kdreyer@redhat.com>
10 years agodoc: Corrects rgw.conf file path for Debian-based and RPM-based distros 4301/head
Nilamdyuti Goswami [Wed, 8 Apr 2015 10:18:39 +0000 (15:48 +0530)]
doc: Corrects rgw.conf file path for Debian-based and RPM-based distros
in radosgw man page.

Signed-off-by: Nilamdyuti Goswami <ngoswami@redhat.com>
10 years agoMerge pull request #4294 from ktdreyer/wip-readme-rm-old-ubuntu
Loic Dachary [Wed, 8 Apr 2015 08:55:37 +0000 (10:55 +0200)]
Merge pull request #4294 from ktdreyer/wip-readme-rm-old-ubuntu

README: rm references to old Ubuntu distros

Reviewed-by: Loic Dachary <ldachary@redhat.com>
10 years agoMerge pull request #4289 from dachary/wip-docker-user
Kefu Chai [Wed, 8 Apr 2015 08:28:51 +0000 (16:28 +0800)]
Merge pull request #4289 from dachary/wip-docker-user

tests: allow multiple users of docker-tests.sh

Reviewed-by: Kefu Chai <kchai@redhat.com>
10 years agoMerge pull request #4210 from trociny/fix-11261
Kefu Chai [Wed, 8 Apr 2015 07:19:29 +0000 (15:19 +0800)]
Merge pull request #4210 from trociny/fix-11261

ceph_argparse: return false instead exiting when argument is missed

Reviewed-by: Kefu Chai <kchai@redhat.com>
10 years agoTestCase: Change in testcase output
Rajesh Nambiar [Wed, 8 Apr 2015 06:40:58 +0000 (23:40 -0700)]
TestCase: Change in testcase output

For the local sanity test to pass this change is needed as the output message changes.

Signed-off-by: Rajesh Nambiar <rajesh.n@msystechnologies.com>
10 years ago Fix to some of the command line parsing (including rbd)
Rajesh Nambiar [Wed, 8 Apr 2015 06:39:07 +0000 (23:39 -0700)]
 Fix to some of the command line parsing (including rbd)

Fix#: 2862

Changes to some of the common files for command line parsing

Change to ceph_argparse.cc
-------------------------

Added function ceph_arg_value_type()
  Given an input it will determine
   i) If that input is an option or not
   ii) If input is numeric in nature or not.

  It will set the flag bool_option and bool_numeric appropriately.
  This function is called by ceph_argparse_witharg() to figure out if
  the input parameter to those functions are numeric in nature and not
  an option.  If the input parameter to ceph_argparse_witharg()
  happens to be an option then it implies that user didn't supply
  value to the option.

Changes to strol.cc
-------------------
Changes to strict_strtoll() and strict_strtol()

  Both these functions reponsibility is to convert the string to long or to int.
  I felt it may be not be good for it to display error message within this function,
  rather caller of this function who has better understanding of the function's purpose
  can display the error message.

  Made change in this function to just create a generic error message,Its the
  caller of this function decides what to do with this message.

Signed-off-by: Rajesh Nambiar <rajesh.n@msystechnologies.com>
10 years agorbd: create command throws inappropriate error messages
Rajesh Nambiar [Wed, 8 Apr 2015 06:31:48 +0000 (23:31 -0700)]
rbd: create command throws inappropriate error messages

Fixes: #2862
Changes related to rbd file

Changes to rbd.cc
-----------------

Change 1: line# 2744 to 2747

 If the option is --order then do the check of its value if its less
 than 12 or greaterthan 25 then throw error. Correct value of --order
 is 12 to 25.

Change 2: Removal of validation from line# 3205 to 3209

  Since the check for correct value of --order is done before hence the
  check here is not needed.

Signed-off-by: Rajesh Nambiar <rajesh.n@msystechnologies.com>
10 years agoMerge pull request #4288 from fullerdj/wip-rbd-xfstests-201504
Josh Durgin [Wed, 8 Apr 2015 00:04:43 +0000 (17:04 -0700)]
Merge pull request #4288 from fullerdj/wip-rbd-xfstests-201504

RBD: build xfsdump and xfsprogs from source; set new expunge set for xfstests

Reviewed-by: Josh Durgin <jdurgin@redhat.com>
10 years agoRBD: update expunge set for latest test, parameterize test script 4288/head
Douglas Fuller [Mon, 6 Apr 2015 21:36:18 +0000 (14:36 -0700)]
RBD: update expunge set for latest test, parameterize test script

Update expunge set for xfstests to known-good set for RHEL7 and Ubuntu 14.04.
Expunged tests without justifying comments are long-running and have not yet
been thoroughly investigated.

Add an environment variable (optionally passed from xfstests_url variable in
test YAML) to override the default xfstests repo.

Signed-off-by: Douglas Fuller <dfuller@redhat.com>
10 years agoRBD: build prerequisites for latest xfstests and update test configuration
Douglas Fuller [Mon, 6 Apr 2015 21:35:20 +0000 (14:35 -0700)]
RBD: build prerequisites for latest xfstests and update test configuration

To standardize results from xfstests, install specific versions of xfstests,
xfsprogs, and xfsdump in /tmp. Move test mountpoints to /tmp/cephtest to allow
access for fsgqa user expected by xfstests.

XXX: xfsdump required significant hacks to build and install on Ubuntu 14.04.
A developer removed the definitions of the min and max macros; this commit is
reverted by this script. The installation paths are difficult to override and
the installer breaks (trying to ln one of the binaries to itself) when an
alternate installation location is specified. This script ignores this error
(make -k install || true).

Signed-off-by: Douglas Fuller <dfuller@redhat.com>
10 years agoREADME: rm references to old Ubuntu distros 4294/head
Ken Dreyer [Tue, 7 Apr 2015 23:38:44 +0000 (17:38 -0600)]
README: rm references to old Ubuntu distros

Support for Ubuntu 11.04 officially ended on 28 October 2012.

Support for Ubuntu Oneiric Ocelot was officially ended on 9 May 2013.

Remove the references to these EOL distro versions.

Signed-off-by: Ken Dreyer <kdreyer@redhat.com>
10 years agoMerge pull request #4292 from ceph/wip-11342
Josh Durgin [Tue, 7 Apr 2015 22:55:33 +0000 (15:55 -0700)]
Merge pull request #4292 from ceph/wip-11342

librbd: moved snap_create header update notification to initiator

Reviewed-by: Josh Durgin <jdurgin@redhat.com>
10 years agoMerge pull request #4281 from wonzhq/readahead
Josh Durgin [Tue, 7 Apr 2015 21:47:44 +0000 (14:47 -0700)]
Merge pull request #4281 from wonzhq/readahead

librbd: readahead fixes

Reviewed-by: Josh Durgin <jdurgin@redhat.com>
10 years agodoc/release-notes: spelling
Sage Weil [Tue, 7 Apr 2015 21:10:33 +0000 (14:10 -0700)]
doc/release-notes: spelling

Signed-off-by: Sage Weil <sage@redhat.com>
10 years agodoc/release-notes: correction
Sage Weil [Tue, 7 Apr 2015 21:09:37 +0000 (14:09 -0700)]
doc/release-notes: correction

There isn't an upgrade f-g-x suite

Signed-off-by: Sage Weil <sage@redhat.com>
10 years agolibrbd: moved snap_create header update notification to initiator 4292/head
Jason Dillaman [Tue, 7 Apr 2015 19:39:13 +0000 (15:39 -0400)]
librbd: moved snap_create header update notification to initiator

When handling a proxied snap_create operation, the client which
invoked the snap_create should send the header update notification
to avoid a possible race condition where snap_create completes but
the client doesn't see the new snapshot (since it didn't yet receive
the notification).

Fixes: #11342
Signed-off-by: Jason Dillaman <dillaman@redhat.com>
10 years agodoc/release-notes: final hammer notes
Sage Weil [Tue, 7 Apr 2015 20:02:33 +0000 (13:02 -0700)]
doc/release-notes: final hammer notes

Signed-off-by: Sage Weil <sage@redhat.com>
10 years agoMerge remote-tracking branch 'gh/hammer'
Sage Weil [Tue, 7 Apr 2015 19:57:07 +0000 (12:57 -0700)]
Merge remote-tracking branch 'gh/hammer'

Conflicts:
src/gmock

10 years agoMerge pull request #4291 from ceph/wip-doc-fixes
Loic Dachary [Tue, 7 Apr 2015 19:11:43 +0000 (21:11 +0200)]
Merge pull request #4291 from ceph/wip-doc-fixes

Wip doc fixes

Reviewed-by: Loic Dachary <ldachary@redhat.com>
10 years agodoc/release-notes: make a note about rgw deployment; tweak dedication
Sage Weil [Tue, 7 Apr 2015 18:36:49 +0000 (11:36 -0700)]
doc/release-notes: make a note about rgw deployment; tweak dedication

Signed-off-by: Sage Weil <sage@redhat.com>
10 years agoMerge pull request #4283 from oritwas/multizone
Yehuda Sadeh [Tue, 7 Apr 2015 17:43:31 +0000 (10:43 -0700)]
Merge pull request #4283 from oritwas/multizone

rgw: remove meta file after deleting bucket

Reviewed-by: Yehuda Sadeh <yehuda@redhat.com>
10 years ago0.94 v0.94
Jenkins [Tue, 7 Apr 2015 17:05:41 +0000 (10:05 -0700)]
0.94

10 years agolibrbd: simplify AioRequest constructor parameters 4233/head
Jason Dillaman [Tue, 31 Mar 2015 16:28:11 +0000 (12:28 -0400)]
librbd: simplify AioRequest constructor parameters

Moved all parent overlap computation to within AioRequest so that
callers don't need to independently compute the overlap.  Also
removed the need to pass the snap_id for write operations since
it can only be CEPH_NOSNAP.

Signed-off-by: Jason Dillaman <dillaman@redhat.com>
10 years agoman: enable warning msgs for "all" target 4291/head
Kefu Chai [Tue, 7 Apr 2015 10:29:35 +0000 (18:29 +0800)]
man: enable warning msgs for "all" target

* and remove `local-check` target. because we have killed all warnings,
  and `make check` depends on `am-all` actually. so previously, `make check`
  basically does nothing at all for validating the man pages' syntax.
  `make all` has rendered them already.
* do not run sphinx-build if $(top_srcdir)/doc/man does not exist. this
  is exactly the case of `make install` from a dist tarball.
* remove generated man pages in `make maintainer-clean` not in `make
  clean`, because we don't ship .rst files in the dist tarball.

Signed-off-by: Kefu Chai <kchai@redhat.com>
10 years agodoc: ref 9/ceph-mon using relative path
Kefu Chai [Tue, 7 Apr 2015 10:15:01 +0000 (18:15 +0800)]
doc: ref 9/ceph-mon using relative path

sphinx complains when it failed to find the referenced file using the
absolute path whose root is the $(top_srcdir)/doc/man when rendering
man pages. but it is able to find it when rendering the html, because
build-doc specifies $(top_srcdir)/doc/ as the source directory.

Signed-off-by: Kefu Chai <kchai@redhat.com>
10 years agodoc: add ":orphan:" field to avoid sphinx warnings
Kefu Chai [Tue, 7 Apr 2015 10:11:29 +0000 (18:11 +0800)]
doc: add ":orphan:" field to avoid sphinx warnings

when building man pages from rst files, sphinx complains if any
rst file is not referenced by a doctree.

Signed-off-by: Kefu Chai <kchai@redhat.com>
10 years agoman: ignore fieldlist when getting desc for manpages
Kefu Chai [Tue, 7 Apr 2015 07:45:20 +0000 (15:45 +0800)]
man: ignore fieldlist when getting desc for manpages

this change will enable us to specify `:orphan:` in the fieldlist of
a manpage rst file, otherwise sphinx-build complains at seeing it if
it is not referenced by a toc doc.

Signed-off-by: Kefu Chai <kchai@redhat.com>
10 years agodoc: fix sphinx warnings
Kefu Chai [Tue, 7 Apr 2015 08:37:07 +0000 (16:37 +0800)]
doc: fix sphinx warnings

* fix `Inline literal start-string without end-string` warnings.
* fix `Title underline too short` warning

Signed-off-by: Kefu Chai <kchai@redhat.com>
10 years agodoc: fix sphinx warnings
Kefu Chai [Tue, 7 Apr 2015 08:20:44 +0000 (16:20 +0800)]
doc: fix sphinx warnings

fix `Inline literal start-string without end-string` warnings.

Signed-off-by: Kefu Chai <kchai@redhat.com>
10 years agoman: appease sphinx by providing a toc doc
Kefu Chai [Mon, 6 Apr 2015 02:36:22 +0000 (10:36 +0800)]
man: appease sphinx by providing a toc doc

Fixes: #11324
Signed-off-by: Kefu Chai <kchai@redhat.com>
10 years agoman: point man/conf.py and Makefile to doc/man
Kefu Chai [Tue, 7 Apr 2015 04:29:04 +0000 (12:29 +0800)]
man: point man/conf.py and Makefile to doc/man

Fixes: #11320
Signed-off-by: Kefu Chai <kchai@redhat.com>
10 years agoman: move man/8/*.rst back to doc/man/8
Kefu Chai [Tue, 7 Apr 2015 08:02:18 +0000 (16:02 +0800)]
man: move man/8/*.rst back to doc/man/8

sphinx requires that a document tree to be in a single directory,
and the man pages are references by our document. so a simple solution
would be to put all the rst file under a single directory.

Fixes: #11320
Signed-off-by: Kefu Chai <kchai@redhat.com>
10 years agodoc: fix a typo which use "::" for ":"
Kefu Chai [Tue, 7 Apr 2015 04:26:49 +0000 (12:26 +0800)]
doc: fix a typo which use "::" for ":"

Signed-off-by: Kefu Chai <kchai@redhat.com>
10 years agodoc: fix "Title underline too short" warning
Kefu Chai [Tue, 7 Apr 2015 03:01:42 +0000 (11:01 +0800)]
doc: fix "Title underline too short" warning

Signed-off-by: Kefu Chai <kchai@redhat.com>
10 years agodoc: fix sphinx warnings
Kefu Chai [Tue, 7 Apr 2015 04:41:30 +0000 (12:41 +0800)]
doc: fix sphinx warnings

* kill "Inline strong start-string without end-string" warning by
  removing superfluous '*'

Signed-off-by: Kefu Chai <kchai@redhat.com>
10 years agodoc: remove duplicated option doc
Kefu Chai [Mon, 6 Apr 2015 02:39:22 +0000 (10:39 +0800)]
doc: remove duplicated option doc

sphinx complains:
SEVERE: Duplicated ID: "cmdoption-radosgw-admin--bucket"
at seeing duplicated `--bucket` option in radosgw-admin.rst. and
there is a global `--bucket` for all commands of radosgw, so remove
the one for the `quota` command.

Signed-off-by: Kefu Chai <kchai@redhat.com>
10 years agorgw: add support for end_marker for GET on Swift container. 4224/head
Radoslaw Zarzynski [Mon, 30 Mar 2015 13:00:59 +0000 (15:00 +0200)]
rgw: add support for end_marker for GET on Swift container.

Fixes: #10682
Backport: hammer
Signed-off-by: Radoslaw Zarzynski <rzarzynski@mirantis.com>
10 years agotests: remove dead code in docker-test-helper.sh 4289/head
Loic Dachary [Tue, 7 Apr 2015 14:15:31 +0000 (16:15 +0200)]
tests: remove dead code in docker-test-helper.sh

Signed-off-by: Loic Dachary <ldachary@redhat.com>
10 years agotests: allow multiple users of docker-tests.sh
Loic Dachary [Tue, 7 Apr 2015 14:07:40 +0000 (16:07 +0200)]
tests: allow multiple users of docker-tests.sh

The docker image created by docker-tests.sh for a given operating system
is parameterized with the user name. If two users on the same machine
try to use the same image, they will compete and fail with an error
like:

... user get supplementary groups Unable to find user ...

Add the $USER to the image name to reflect the fact that they contain an
account for this user.

Signed-off-by: Loic Dachary <ldachary@redhat.com>
10 years agoMerge pull request #4273 from dachary/wip-docker-ccache
Kefu Chai [Tue, 7 Apr 2015 13:21:19 +0000 (21:21 +0800)]
Merge pull request #4273 from dachary/wip-docker-ccache

tests: docker-test must bind mount .ccache

Reviewed-by: Kefu Chai <kchai@redhat.com>
10 years agoMerge pull request #4255 from Ved-vampir/osdccnt
John Spray [Tue, 7 Apr 2015 13:10:04 +0000 (14:10 +0100)]
Merge pull request #4255 from Ved-vampir/osdccnt

Perf counters descriptions, new part

Reviewed-by: John Spray <john.spray@redhat.com>
Reviewed-by: Joao Eduardo Luis <jluis@suse.com>
10 years agoMerge pull request #4286 from tchaikov/readable-href-in-admonition
Loic Dachary [Tue, 7 Apr 2015 12:48:09 +0000 (14:48 +0200)]
Merge pull request #4286 from tchaikov/readable-href-in-admonition

doc: show href anchor in white in important block

Reviewed-by: Loic Dachary <ldachary@redhat.com>
10 years agoMerge pull request #4287 from dachary/wip-sphinx
Loic Dachary [Tue, 7 Apr 2015 12:36:00 +0000 (14:36 +0200)]
Merge pull request #4287 from dachary/wip-sphinx

spec.in: sphinx -b man needs sphinx > 1.0 (part 2)

Reviewed-by: Owen Synge <osynge@suse.com>
10 years agodoc: show href anchor in white in important block 4286/head
Kefu Chai [Tue, 7 Apr 2015 09:02:55 +0000 (17:02 +0800)]
doc: show href anchor in white in important block

Fixes: #11310
Signed-off-by: Kefu Chai <kchai@redhat.com>
10 years agomon: add perf counters description 4255/head
Ved-vampir [Fri, 3 Apr 2015 08:12:10 +0000 (11:12 +0300)]
mon: add perf counters description

Sign-off-by: Alyona Kiseleva <akiselyova@mirantis.com>
10 years agoMerge pull request #4265 from trociny/wip-11319
John Spray [Tue, 7 Apr 2015 10:58:54 +0000 (11:58 +0100)]
Merge pull request #4265 from trociny/wip-11319

ceph daemonperf: add watch interval and count parameters

Reviewed-by: John Spray <john.spray@redhat.com>
10 years agospec.in: sphinx -b man needs sphinx > 1.0 (part 2) 4287/head
Loic Dachary [Tue, 7 Apr 2015 09:47:28 +0000 (11:47 +0200)]
spec.in: sphinx -b man needs sphinx > 1.0 (part 2)

Instead of listing the operating system versions that do not require the
python-sphinx10 package, switch to listing the operating system versions that
require the python-sphinx10 package. It's easier to maintain because
there only are a few.

Signed-off-by: Loic Dachary <ldachary@rehdat.com>
10 years agoMerge pull request #4271 from nehaummareddy/master
Loic Dachary [Tue, 7 Apr 2015 09:54:19 +0000 (11:54 +0200)]
Merge pull request #4271 from nehaummareddy/master

doc: Replaced with apt word and fixed grammatical error

Reviewed-by: Loic Dachary <ldachary@redhat.com>
10 years agorgw: remove meta file after deleting bucket 4283/head
Orit Wasserman [Fri, 3 Apr 2015 11:43:31 +0000 (13:43 +0200)]
rgw: remove meta file after deleting bucket
The meta file is deleted only if the bucket meta data is not synced

Signed-off-by: Orit Wasserman <owasserm@redhat.com>
Fixes: #11149
Backport: hammer, firefly

10 years agodoc: Replaced with apt word and fixed grammatical error 4271/head
nehaummareddy [Fri, 3 Apr 2015 21:02:00 +0000 (03:02 +0600)]
doc: Replaced with apt word and fixed grammatical error

Signed-off-by: Neha Ummareddy <nehaummareddy@gmail.com>
10 years agorgw: quota not respected in POST object 4240/head
9seconds [Wed, 1 Apr 2015 06:12:06 +0000 (09:12 +0300)]
rgw: quota not respected in POST object

Signed-off-by: Sergey Arkhipov <nineseconds@yandex.ru>
Backport: hammer, firefly
Fixes: #11323
10 years agolibrbd: don't do readahead when m_readahead_pos reaching limit 4281/head
Zhiqiang Wang [Tue, 7 Apr 2015 06:23:38 +0000 (14:23 +0800)]
librbd: don't do readahead when m_readahead_pos reaching limit

When m_readahead_pos reaches the limit, there's no need to call
_compute_readahead to calculate the readahead. Just return with no
readahead.

Signed-off-by: Zhiqiang Wang <zhiqiang.wang@intel.com>
10 years agolibrbd: start readahead from m_last_pos when the size of the continuing
Zhiqiang Wang [Tue, 7 Apr 2015 05:43:36 +0000 (13:43 +0800)]
librbd: start readahead from m_last_pos when the size of the continuing
triggering request is big enough

If the size of the read triggering the continuing readahead is such big
that exceeding m_readahead_pos, should do the readahead starting from
m_last_pos.

Signed-off-by: Zhiqiang Wang <zhiqiang.wang@intel.com>
10 years agoMerge pull request #3867 from wonzhq/fiemap
Sage Weil [Mon, 6 Apr 2015 22:48:27 +0000 (15:48 -0700)]
Merge pull request #3867 from wonzhq/fiemap

FileStore: fiemap implementation using SEEK_HOLE/SEEK_DATA

Reviewed-by: Haomai Wang <haomaiwang@gmail.com>
Reviewed-by: Sage Weil <sage@redhat.com>
10 years agoMerge pull request #3170 from tchaikov/python-rados-refactor
Sage Weil [Mon, 6 Apr 2015 22:47:33 +0000 (15:47 -0700)]
Merge pull request #3170 from tchaikov/python-rados-refactor

python-rados: extract type-checking into a decorator

Reviewed-by: Josh Durgin <jdurgin@redhat.com>
Reviewed-by: Loic Dachary <ldachary@redhat.com>
10 years agoMerge pull request #3560 from XinzeChi/wip-pmon-all
Sage Weil [Mon, 6 Apr 2015 22:47:09 +0000 (15:47 -0700)]
Merge pull request #3560 from XinzeChi/wip-pmon-all

ceph.in: add ceph ping mon.* for ping all monitor

Reviewed-by: Loic Dachary <ldachary@redhat.com>
Reviewed-by: Kefu Chai <kchai@redhat.com>
10 years agoMerge pull request #3879 from liewegas/wip-cache-mode-none
Sage Weil [Mon, 6 Apr 2015 22:46:21 +0000 (15:46 -0700)]
Merge pull request #3879 from liewegas/wip-cache-mode-none

mon: warn if cache_mode is none and pool is a read or writ eoverlay tier

Reviewed-by: Sage Weil <sage@redhat.com>