osd: avoid unnecessary calculation in agent_choose_mode()

Signed-off-by: Li Wang <liwang@ubuntukylin.com>
Reviewed-by: Yunchuan Wen <yunchuanwen@ubuntukylin.com>

Merge pull request #5445 from ceph/wip-12432

Wip 12432

rgw: set http status in civetweb

Need to set the http status in civetweb so that we report it correctly.
Fixes: #12432
Signed-off-by: Yehuda Sadeh <yehuda@redhat.com>

civetweb: update submodule to support setting of http status

Signed-off-by: Yehuda Sadeh <yehuda@redhat.com>

Merge pull request #5258 from zhouyuan/rgw_skip_first_chunk

rgw: skip prefetch first chunk if range get falls to shadow objects

Reviewed-by: Yehuda Sadeh <yehuda@redhat.com>

doc: krbd supports clones since 3.10

Signed-off-by: Ilya Dryomov <idryomov@gmail.com>

Merge pull request #5438 from ceph/fix-doc-build

doc: fix the build of placement-groups.rst

Reviewed-by: Kefu Chai <kchai@redhat.com>

doc: fix the build of placement-groups.rst

add a reference target of "mapping pgs to osds"

Signed-off-by: Kefu Chai <kchai@redhat.com>

Merge pull request #5432 from ceph/wip-fix-rbd-build

make: fix the build of "rbd"

Reviewed-by: Loic Dachary <ldachary@redhat.com>

Merge pull request #5434 from tobegit3hub/remove_execute_mode

Remove execute mode of source files

Reviewed-by: Loic Dachary <ldachary@redhat.com>

Remove execute mode of source files

Merge pull request #5398 from mingyuez/master

rbd:If the image has any protected snaps,'rbd snap purge image' will return failure without removing any snaps.

Reviewed-by: Josh During <jdurgin@redhat.com>
Reviewed-by: Haomai Wang <haomaiwang@gmail.com>

make: fix the build of "rbd"

`CEPH_GLOBAL` was removed in fa78739, but the linked libraries does not
offer all needed symbols for "rbd". so add them back.

Signed-off-by: Kefu Chai <kchai@redhat.com>

rbd:'rbd purge image' will return failure without removing any snaps if the image has a protected snap

Signed-off-by: mingyuez <zhao.mingyue@h3c.com>

Merge pull request #5395 from tchaikov/wip-12512

common/syncfs: fall back to sync(2) if syncfs(2) not available

Reviewed-by: Haomai Wang <haomaiwang@gmail.com>
Reviewed-by: Sage Weil <sage@redhat.com>
Reviewed-by: Loic Dachary <ldachary@redhat.com>

Merge pull request #5402 from ceph/wip-claim-sparse-read-buffer

osd/ReplicatedPG: claim the read buffer instead of copying it

Reviewed-by: David Zafman <dzafman@redhat.com>

Merge pull request #5428 from ceph/wip-kill-rest-bench

kill rest-bench

Reviewed-by: Ken Dreyer <kdreyer@redhat.com>

rgw: skip prefetch first chunk if range get falls to shadow objects

Currently the head object will be prefetched in each GET:
 a) This is unnecessary if the Range GET falls to shadow objects.
 b) The GET request would be quite slow if we have a big head object
This patch adds some check on the Range. If it's not in the head
object(>=rgw_max_chunk_size) then skip the prefetch.

Fixes: #12539
Signed-off-by: Yuan Zhou <yuan.zhou@intel.com>

Merge remote-tracking branch 'origin/next'

Merge pull request #5425 from ceph/wip-12479-next

librbd: test crash during client upgrade suite

Reviewed-by: Josh Durgin <jdurgin@redhat.com>

Merge pull request #5423 from oddomatik/patch-1

doc: No CRUSH involvement in deciding PG target

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

Merge pull request #5426 from ceph/wip-12383-next

rbd: remove dependency on non-ABI controlled CephContext

Reviewed-by: Josh Durgin <jdurgin@redhat.com>

Merge pull request #5399 from tchaikov/wip-12458

rbd: add "--keyring" option to help message

Reviewed-by: Josh Durgin <jdurgin@redhat.com>

Merge pull request #5414 from cxwshawn/master

doc: add bucket object version description.

Reviewed-by: Yehuda Sadeh <yehuda@redhat.com>

remove libs3 submodule

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

remove rest-bench

This is a weak tool; users should look to cosbench or others instead.

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

Merge pull request #5419 from ceph/wip-kfsx-invalidate

test_librbd_fsx: invalidate before discard in krbd mode

Reviewed-by: Josh Durgin <jdurgin@redhat.com>

No CRUSH involvement in deciding PG target

Quoting from elsewhere "[we calculate] a placement group using the object name, a hash code, the number of PGs in the pool and the pool name"

lockdep: allow lockdep to be dynamically enabled/disabled

librbd test cases attempt to enable lockdep coverage via the librados
API.  Use a configuration observer to register/unregister lockdep
support.

Signed-off-by: Jason Dillaman <dillaman@redhat.com>

tests: librbd API test cannot use private md_config_t struct

Remove all depencencies on md_config_t and instead use librados API
methods to get/set configuration values.

Fixes: #12479
Backport: hammer
Signed-off-by: Jason Dillaman <dillaman@redhat.com>

librados_test_stub: implement conf get/set API methods

Signed-off-by: Jason Dillaman <dillaman@redhat.com>

rbd: add "--keyring" option to help message

Refs: #12458
Signed-off-by: Kefu Chai <kchai@redhat.com>

Merge pull request #5401 from ceph/wip-12450

tools/rados: change the default max_ops to 16

Reviewed-by: David Zafman <dzafman@redhat.com>

doc: add bucket object version description.
bucket object version has been supported, but do not have description in the docs,
so add this part.

Signed-off-by: shawn chen <cxwshawn@gmail.com>

test_librbd_fsx: invalidate before discard in krbd mode

Commit bd050240ceef ("test_librbd_fsx: flush before discard in krbd
mode") added an fsync() before BLKDISCARD.  Don't know what I was
thinking at the time, but I missed the invalidate part, for which we
need to use the BLKFLSBUF ioctl.

Signed-off-by: Ilya Dryomov <idryomov@gmail.com>

Log::reopen_log_file: take m_flush_mutex

Otherwise, _flush() might continue to write to m_fd after it's closed.
This might cause log data to go to a data object if the filestore then
reuses the fd during that time.

Fixes: #12465
Backport: firefly, hammer
Signed-off-by: Samuel Just <sjust@redhat.com>
(cherry picked from commit 8778ab3a1ced7fab07662248af0c773df759653d)

Merge pull request #5348 from athanatos/wip-12465

log: take mutex when reopening log fd

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

tools/rados: change the first op id to 0

the first op id was 16 by default, which is okay, but a non-zero
magic number could lead to questions. max_op was mixed up with
max_ops, and changed to 16 in 51e402e3 by mistake.

Signed-off-by: Kefu Chai <kchai@redhat.com>

tools/rados: change the default max_ops to 16

otherwise the load-gen will not emit any ops by default.

Fixes: #12450
Signed-off-by: Kefu Chai <kchai@redhat.com>

common/syncfs: fall back to sync(2) if syncfs(2) not available

Fixes: #12512
Signed-off-by: Kefu Chai <kchai@redhat.com>

rbd: remove dependency on non-ABI controlled CephContext

The rbd CLI tool no longer attempts to initialize a CephContext
and pass said context to librados since it's possible that the
structure will not be ABI compatible between rbd and librados.

Signed-off-by: Jason Dillaman <dillaman@redhat.com>

Merge pull request #5139 from jmunhoz/wip-7817-ms

rgw: add minimum support for copy multipart part

Reviewed-by: Yehuda Sadeh <yehuda@redhat.com>

Merge pull request #5313 from guce/rgw_access_invalid_argument

rgw: check subuser illegal access parameter.

Reviewed-by: Yehuda Sadeh <yehuda@redhat.com>

crypto: use NSS_InitContext/NSS_ShutdownContex to avoid memory leak

Switched to context-aware NSS init/shutdown functions to avoid conflicts
with parent application.  Use a reference counter to properly shutdown the
NSS crypto library when the last CephContext is destroyed.  This avoids
memory leaks with the NSS library from users of librados.

Signed-off-by: Jason Dillaman <dillaman@redhat.com>

Merge pull request #5364 from xdonghai/master

doc:modified librados-intro.rst an error of document representation

Reviewed-by: Abhishek Lekshmanan <abhishek.lekshmanan@ril.com>
Reviewed-by: Josh Durgin <jdurgin@redhat.com>

Merge pull request #5191 from theanalyst/v9.0.2/mailmap

doc: mailmap for v9.0.2 release

Reviewed-by: Loic Dachary <ldachary@redhat.com>

Merge pull request #5130 from tchaikov/wip-12167

ceph-detect-init: do not require "argparse" on py2.6

Reviewed-by: Loic Dachary <ldachary@redhat.com>

Merge pull request #5338 from hjwsm1989/wip-12459

common/WorkQueue: fix the FTBFS with GCC 4.4.7

* add nested-name-specifier ThreadPool before WorkQueueVal

Reviewed-by: Kefu Chai <kchai@redhat.com>

Merge pull request #5350 from ceph/wip-pgls-filter-enc

osdc: fix pgls_filter op encoding

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

Merge pull request #5355 from ceph/wip-anotate-s-files

Anotate all the .s files

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

Merge pull request #5116 from yuyuyu101/wip-compressor

Wip compressor

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

modified librados-intro.rst

Signed-off-by: x11507 <xu.donghai@h3c.com>

Merge pull request #5254 from vuhuong/wip-xio

@vuhuong sorry for the latency, could you please rebase & repush , to poke the built bot again?

once the build completes, will get it merged.

thanks,

osd/ReplicatedPG: claim the read buffer instead of copying it

Signed-off-by: Kefu Chai <kchai@redhat.com>

Merge pull request #5349 from yuyuyu101/wip-async-fix-21

AsyncConnection: Exit process loop if entering fault

Merge remote-tracking branch 'origin/next'

Merge pull request #5352 from ceph/wip-rbd-help-specs

rbd: cli help and arg parsing fixes

Reviewed-by: Jason Dillaman <dillaman@redhat.com>
Reviewed-by: Josh Durgin <jdurgin@redhat.com>

Merge pull request #5336 from ceph/wip-migrate-event-seq

mds: fix val used in inode->last_journaled

Reviewed-by: Greg Farnum <gfarnum@redhat.com>

Anotate all the .s files

Recent update to erasure code library in 59aa6700 caused a regression
where .s files are no longer properly anotated and yasm sets the exec
stack for them. This patch brings back the anotations as was done before
by Dan Mick, see Bug #10114.

Signed-off-by: Boris Ranto <branto@redhat.com>

rbd: rename --object-extents option to --whole-object

--object-extents is a bit confusing - extent is generally something of
a varying length and here the meaning is "diff whole objects".  Rename
it to --whole-object (the name of diff_iterate() parameter).

Signed-off-by: Ilya Dryomov <idryomov@gmail.com>

rbd: du command should take spec as an argument

Change du to take <image-spec> | <snap-spec> as an argument instead of
going through --image option.  The new synopsis is

    (du | disk-usage) [<image-spec> | <snap-spec>]

This is to make it look more like the rest of the commands: the only
other command that takes pool as an argument is ls and it can't really
serve as a prototype for du, because the latter has to work on images
and snapshots as well.

Examples:

    # stats for pool rbd
    $ rbd du
    $ rbd -p rbd du

    # stats for pool foo
    $ rbd -p foo du

    # stats for snapshot mysnap of image baz in pool rbd
    $ rbd du baz@mysnap

    # stats for image bar in pool foo
    $ rbd du foo/bar

Signed-off-by: Ilya Dryomov <idryomov@gmail.com>

rbd: assorted cli arg parsing fixes

- import-diff needs image-spec, yet OPT_IMPORT_DIFF was in the list of
  commands that don't need imgname

- check for lock id after imgname, so that if imgname isn't specified
  we fail with imgname rather than lock id error message

- error out if locker isn't specified

- move feature-name validity check after imgname checks so that if
  imgname is malformed we fail with imgname rather than feature error
  message

- error out if image-meta key or value isn't specified

Signed-off-by: Ilya Dryomov <idryomov@gmail.com>

rbd: error out if dest_snapname is specified

No command uses it as of now, but only clone command fails; cp, mv and
import simply ignore it.  Check if it's set and exit with a generic
error message.

Signed-off-by: Ilya Dryomov <idryomov@gmail.com>

rbd: import doesn't require image-spec arg, ditto for export and path

Mark those as such in help and clarify what image-spec defaults to.

Related, all command args in our man page are enclosed into brackets.
I suppose the reason is that they are optional in the sense that you
can have commands like

    $ rbd clone --pool a --image b --snap -c --dest-pool d --dest e

with no args.  Given that we are trying to push people towards

    $ rbd clone a/b@c d/e

undo that so that real optional arguments can be marked optional.
While at it, add synopsis for each command and use backticks for
denoting commands more consistently.

Signed-off-by: Ilya Dryomov <idryomov@gmail.com>

doc: use spec syntax in rbd docs

Use spec syntax instead of --pool, --image and --snap to hopefully push
people towards the former.

Signed-off-by: Ilya Dryomov <idryomov@gmail.com>

rbd: use image-spec and snap-spec in rbd help

This patch changes image-name instances to image-spec and snap-name
instances to snap-spec to try to clarify usage for some commands and
disambiguate the term {image,snap}-name, which has been used to denote
both simple names and compound names (specs).

<image-spec> is [<pool-name>]/<image-name>
<snap-spec> is [<pool-name>]/<image-name>@<snap-name>

This patch also removes duplicate checks for image-name and snap-name.

Signed-off-by: Vikhyat Umrao <vumrao@redhat.com>
[idryomov@gmail.com: some commands take either image-spec or snap-spec,
other fixes, formatting, changelog]
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>

osdc: fix pgls_filter op encoding

This was broken with respect to the OSD's expected order
of decoding.

Signed-off-by: John Spray <john.spray@redhat.com>

Compressor: Remove thread affinity options

Signed-off-by: Haomai Wang <haomaiwang@gmail.com>

AsyncConnection: Exit process loop if entering fault

"fault" will transmit state to exception state or STATE_CONNECTING, all
transmitted states don't need to check right now.

Signed-off-by: Haomai Wang <haomaiwang@gmail.com>

Merge pull request #5342 from tchaikov/wip-fix-perf-local-on-arm

test/perf_local: disable tests on unsupported archs

Reviewed-by: Haomai Wang <haomaiwang@gmail.com>

test/perf_local: disable tests on unsupported archs

* maybe we can have div32 tests on aarch64, but i don't
  think "udiv|sdiv" supports 64bits numerator. probably
  we can use float divde for the benchmark...
* disable cpuid test on non-intel archs.

Fixes: #12453
Reported-by: Tom Deneau <tom.deneau@amd.com>
Signed-off-by: Kefu Chai <kchai@redhat.com>

Log::reopen_log_file: take m_flush_mutex

Otherwise, _flush() might continue to write to m_fd after it's closed.
This might cause log data to go to a data object if the filestore then
reuses the fd during that time.

Fixes: #12465
Backport: firefly, hammer
Signed-off-by: Samuel Just <sjust@redhat.com>

Merge pull request #5316 from shylesh/bugfix

mon: ceph osd map shows NONE when an osd is missing

Reviewed-by: Kefu Chai <kchai@redhat.com>
Reviewed-by: Loic Dachary <ldachary@redhat.com>

  common: add nested-name-specifier ThreadPool before WorkQueueVal
Fixes: #12459
Signed-off-by: huangjun <hjwsm1989@gmail.com>

Merge pull request #5322 from majianpeng/mds-log

mmds/MDLog: No make sense set offset for LogEvent in _start_entry.

Reviewed-by: John Spray <john.spray@redhat.com>

mds: fix val used in inode->last_journaled

This was getting assigned with LogEvent::get_start_offset
on an uncommitted LogEvent, which is junk.  During replay
last_journaled is compared with the metablob's event_seq,
so that's what should be used here.

This change just from code inspection -- haven't seen this
manifest as an actual misbehaviour.

Signed-off-by: John Spray <john.spray@redhat.com>

Merge pull request #5330 from liewegas/wip-cephtool-test

qa/workunits/cephtool/test.sh: escape osd.* and mon.*

Reviewed-by: Loic Dachary <ldachary@redhat.com>

qa/workunits/cephtool/test.sh: escape osd.* and mon.*

Otherwise we get filenames if the glob happens to match something in the
source dir.

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

Merge branch 'next'

Merge pull request #5327 from athanatos/wip-12223

WBThrottle::clear_object: signal if we cleared an object

Reviewed-by: David Zafman <dzafman@redhat.com>

Merge pull request #5328 from athanatos/wip-12263

ReplicatedPG::finish_promote: do not prefill new_clones

Reviewed-by: David Zafman <dzafman@redhat.com>

WBThrottle::clear_object: signal if we cleared an object

Fixes: #12223
Signed-off-by: Samuel Just <sjust@redhat.com>

Merge pull request #5189 from athanatos/wip-recovery-defaults

config_opts: turn down default recovery tunables

Reviewed-by: David Zafman <dzafman@redhat.com>

Merge pull request #5329 from athanatos/wip-8193

test/librados/tier.cc: we can exceed the hitset limit while backfilling

Reviewed-by: David Zafman <dzafman@redhat.com>

Merge pull request #5323 from ceph/wip-12426

mon: fix checks on mds add_data_pool

Reviewed-by: Greg Farnum <gfarnum@redhat.com>

Merge pull request #4982 from branch-predictor/bp-lighter-rwlocks

common/RWLock: allow disabling read/write lock counts

Reviewed-by: Samuel Just <sjust@redhat.com>

Merge pull request #5152 from branch-predictor/bp-rados-aio-overrun

tools: fix race condition in seq/rand bench

Reviewed-by: Samuel Just <sjust@redhat.com>

Merge pull request #5093 from guangyy/wip-12162

osd - pg_interval_t::check_new_interval should not rely on pool.min_size to determine if the PG was active

Reviewed-by: Samuel Just <sjust@redhat.com>

ceph-helpers.sh: don't test the osd_max_backfills value

This isn't a very meaningful assert since the default could
easily change.

Signed-off-by: Samuel Just <sjust@redhat.com>

Merge pull request #5325 from vumrao/wip-vumrao-10927

rgw: fix radosgw start-up script.

Reviewed-by: Ken Dreyer <kdreyer@redhat.com>

Merge pull request #5308 from oritwas/fix-compilation-warnings

Fix "was hidden" compilation warnings

Reviewed-by: Yehuda Sadeh <yehuda@redhat.com>

Fix "was hidden" compilation warnings

Signed-off-by: Orit Wasserman <owasserm@redhat.com>

rgw: fix radosgw start-up script.

radosgw init script is unable to start radosgw daemon.
as it is relies on requiretty being disabled.
once init script start this daemon with sudo it fails
to start the daemon.

changing 'sudo' to 'su', it will fix this issue and
will also help running radosgw daemon with our new
"ceph" UID project.

Fixes: #10927
Signed-off-by: Vikhyat Umrao <vumrao@redhat.com>

Merge pull request #5326 from tweenk/docs-improvements

Documentation for a few objects

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

Merge pull request #5261 from SUSE/wip-drop-py-requires

ceph.spec.in: drop SUSE-specific %py_requires macro

Reviewed-by: Ken Dreyer <kdreyer@redhat.com>

Merge pull request #5263 from SUSE/wip-apache-not-required

ceph.spec.in: remove SUSE-specific apache2-mod_fcgid dependency

Reviewed-by: Ken Dreyer <kdreyer@redhat.com>

docs: Document md_config_obs_t.

Signed-off-by: Krzysztof Kosiński <krzysztof.kosinski@intel.com>

docs: Document the ThreadPool and WorkQueue classes.

Signed-off-by: Krzysztof Kosiński <krzysztof.kosinski@intel.com>

docs: Document the Finisher class.

Signed-off-by: Krzysztof Kosiński <krzysztof.kosinski@intel.com>

mon: fix checks on mds add_data_pool

Use same _check_pool helper as during fs creation.

Fixes: #12426
Signed-off-by: John Spray <john.spray@redhat.com>