osd: switch filestore to default to rocksdb

Signed-off-by: Neha Ojha <nojha@redhat.com>
(cherry picked from commit ba96197e6d4614377c80cf16a1b29b6e6ac70f90)

Conflicts:
src/common/config_opts.h (trivial)

rocksdb: remove experimental status

We've been testing rocksdb on mons in QA for months now
with no problems.  We've also observed no problems with
its use in bluestore.

Signed-off-by: Sage Weil <sage@redhat.com>
(cherry picked from commit 68138893648e052a6549fecf23b0b58af921dfc7)

include/assert: define ceph_abort_msg(cct, msg)

Abort with an error string.

Signed-off-by: Sage Weil <sage@redhat.com>
(cherry picked from commit 23d6438f1253b593dd9a0e909e2ccbd045e168b8)

kv/RocksDBStore: abort if rocksdb EIO, don't return incorrect result

If the underlying disk is missing, the current logic won't check the actual
reason why Get/Set failed, it will result to client get a empty key/value
pair which is not expected. The correct logic should be abort
right now. Otherwise, it will leads to undefined behavior.

Signed-off-by: Haomai Wang <haomai@xsky.com>
(cherry picked from commit 8c60b55d7d49bc6fa1d8d076c025fc83283c0776)

Conflicts:
src/kv/RocksDBStore.cc (trivial)

os/kv: beautify kvdb statistics

Improves the presentation of rocksdb dump statistics

Signed-off-by: Varada Kari <varada.kari@sandisk.com>
(cherry picked from commit 8e1d4ac92e2f9cd5f7438a99459bc063adf2542b)

src/kv/RocksDBStore.cc:fix typos

Signed-off-by: xianxiaxiao <xianxia.xiao@kylin-cloud.com>
(cherry picked from commit b26555a0422423a9fef0034aa9313b3a4c1308ef)

kv/RocksDBStore: dump extended stats

Signed-off-by: Varada Kari <varada.kari@sandisk.com>
Signed-off-by: Jianpeng Ma <jianpeng.ma@intel.com>
(cherry picked from commit 9fda9ce0b0fbf795469b50f6e9ebcb9bd0bc32a4)

Conflicts:
src/common/config_opts.h
src/kv/RocksDBStore.cc
src/kv/RocksDBStore.h
(trivial resolution)

kv/RocksDBStore: no reason to make write_batch using new/delete

discards new/delete for bat

Signed-off-by: Haomai Wang <haomai@xsky.com>
(cherry picked from commit 79a6611c0aa2c5b16435fad53268dd29a8541188)

Conflicts:
src/kv/RocksDBStore.cc (trivial)

Logging rocksdb transaction in ceph log

The rocksdb transaction is now been logged into ceph log based
on debug level. If transaction is failed, the error code,
error string and the entire transaction is dumped in the log.

Signed-off-by: Somnath Roy <somnath.roy@sandisk.com>
(cherry picked from commit 46f1f9e5ca2f4e34393dfa1952fa4283a859d869)

RocksDBStore: use Get for multiget

Signed-off-by: Haomai Wang <haomai@xsky.com>
(cherry picked from commit a44f4f7238585996f83f7e35e01587671fe675d4)

kv/RocksDBStore: convert get() to use rocksdb Get() instead of iterator

For RocksDB, iterator will bypass row cache.
We don't want this to happen for get(), after row cache is enabled.

Signed-off-by: Jianjian Huo <jianjian.huo@ssi.samsung.com>
(cherry picked from commit f6ac5579190549291f6385775070c305dc8fba8b)

Conflicts:
src/kv/RocksDBStore.cc (trivial)

tests: unittest_rocksdb_option: remove deprecated option

* disable_data_sync was removed from rocksdb, see
  https://github.com/facebook/rocksdb/blob/02799ad77a16332ee5bfb570f62ab6162a788a5a/HISTORY.md#530-03082017.
  so remove it.
* and check options.compression, which was commented out.

Signed-off-by: Kefu Chai <kchai@redhat.com>
(cherry picked from commit 94f4b48cd4be1b09ae5ee74b61c069e08424c3a5)

rocksdb update

This is the version in 12.2.0, that we have been running for ~6 month or so.

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

kv/Makefile: fix rocksdb build options (USE_RTTI=1)

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

kv/RocksDBStore: Add table options for filter and index tuning.

This is a modified cherry-pick of commit 9c15338 to work with the old
config_opts settings.

Signed-off-by: Mark Nelson <mnelson@redhat.com>
(cherry picked from commit 9c15338ad3049f5171cdecafb9370d6bf5dbe913)

Merge pull request #21234 from tchaikov/wip-jewel-rocksdb-without-jemalloc

jewel: rocksdb: disable jemalloc explicitly

Reviewed-by: Nathan Cutler <ncutler@suse.com>

rocksdb: disable jemalloc explicitly

rocksdb's makefile tries to detect jemalloc automatically and compiles
with it enabled if it's found, but if ceph disable jemalloc, it won't
link against jemalloc. this causes link failure if we compile ceph
without jemalloc enabled in a host with jemalloc library installed.

Signed-off-by: Kefu Chai <kchai@redhat.com>
Conflicts:
src/kv/Makefile.am: this change is not cherry-picked from
master. as master does not use autotools anymore. and master is
using rocksdb's cmake script for building, the latter does not
jemalloc by default.

Merge pull request #21178 from tchaikov/wip-jewel-pr-21159

jewel: build/ops: drop deprecated --use-wheel option from pip invocations

Reviewed-by: Nathan Cutler <ncutler@suse.com>

src: s/--use-wheel//

--use-wheel was deprecated in favor of --only-binary in pip v7.0.0. and
--use-wheel was removed in a recent release of pip. but some packages
are source packages, so we cannot simply replace use-wheel with
only-binary. so a simpler approach is to drop --use-wheel option, as pip
respects --find-links, and will find the required package from the
wheelhouse.

Signed-off-by: Kefu Chai <kchai@redhat.com>
(cherry picked from commit a80cce4ba6afee921a7c6adce802c91e30f8c0a9)

Conflicts:
src/ceph-detect-init/CMakeLists.txt
src/ceph-disk/CMakeLists.txt: trivial resolution
src/pybind/mgr/dashboard/CMakeLists.txt: dashboard2 is not
in luminous, so drop this change.
src/tools/setup-virtualenv.sh:
src/ceph-detect-init/Makefile.am:
src/ceph-disk/Makefile.am: these files were dropped in
master, so update them also.

Merge pull request #20418 from pdvian/wip-22965-jewel

jewel: rbd-mirror: fix potential infinite loop when formatting status message

Reviewed-by: Mykola Golub <mgolub@suse.com>

Merge pull request #20524 from pdvian/wip-23065-jewel

jewel: test/librbd: utilize unique pool for cache tier testing

Reviewed-by: Mykola Golub <mgolub@suse.com>

Merge pull request #20285 from smithfarm/wip-22175-jewel

jewel: rbd: possible deadlock in various maintenance operations

Reviewed-by: Mykola Golub <mgolub@suse.com>

Merge pull request #20518 from smithfarm/wip-23076-jewel

jewel: osd: objecter sends out of sync with pg epochs for proxied ops

Reviewed-by: Sage Weil <sage@redhat.com>
Reviewed-by: Josh Durgin <jdurgin@redhat.com>
Reviewed-by: David Zafman <dzafman@redhat.com>

Merge pull request #20508 from smithfarm/wip-23066-jewel

jewel: ceph.restart + ceph_manager.wait_for_clean is racy

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

Merge pull request #18743 from shinobu-x/wip-21923-jewel

jewel: core: Objecter::C_ObjectOperation_sparse_read throws/catches exceptions on -ENOENT

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

Merge pull request #20435 from pdvian/wip-22968-jewel

jewel: cephfs: osdc/Journaler: make sure flush() writes enough data

Reviewed-by: Patrick Donnelly <pdonnell@redhat.com>

Merge pull request #20335 from pdvian/wip-22865-jewel

jewel: mds: fix scrub crash

Reviewed-by: Patrick Donnelly <pdonnell@redhat.com>

Merge pull request #20333 from pdvian/wip-22863-jewel

jewel: cephfs-journal-tool: move shutdown to the deconstructor of MDSUtility

Reviewed-by: Patrick Donnelly <pdonnell@redhat.com>

Merge pull request #20312 from pdvian/wip-22861-jewel

jewel: cephfs: osdc: "FAILED assert(bh->last_write_tid > tid)" in powercycle-wip-yuri-master-1.19.18-distro-basic-smithi

Reviewed-by: Patrick Donnelly <pdonnell@redhat.com>

Merge pull request #19993 from pdvian/wip-22700-jewel

jewel: cephfs: fuse client: ::rmdir() uses a deleted memory structure of dentry leads …

Reviewed-by: Patrick Donnelly <pdonnell@redhat.com>

Merge pull request #19961 from smithfarm/wip-22695-jewel

jewel: mds: fix dump last_sent

Reviewed-by: Patrick Donnelly <pdonnell@redhat.com>

Merge pull request #19141 from shinobu-x/wip-22241-jewel

jewel: cephfs: Processes stuck waiting for write with ceph-fuse

Reviewed-by: Patrick Donnelly <pdonnell@redhat.com>

Merge pull request #20271 from smithfarm/wip-20823-jewel

jewel: cephfs: client::mkdirs not handle well when two clients send mkdir request for a same dir

Reviewed-by: Patrick Donnelly <pdonnell@redhat.com>

Merge pull request #20982 from dzafman/wip-compact

jewel: Add compact operation to ceph-osdomap-tool

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

Merge pull request #21014 from dzafman/wip-remove-bc-jewel

jewel: test: Replace bc command with printf command

Reviewed-by: Nathan Cutler <ncutler@suse.com>

test: Replace bc command with printf command

Signed-off-by: David Zafman <dzafman@redhat.com>
(cherry picked from commit 293ac9895f86b1258434d0dc539a7d4b7046cb6d)

Merge pull request #20479 from pdvian/wip-23021-jewel

jewel: rgw: fix the max-uploads parameter not work

Reviewed-by: Casey Bodley <cbodley@redhat.com>

Merge pull request #20561 from smithfarm/wip-22894-jewel

jewel: rgw: ECANCELED in rgw_get_system_obj() leads to infinite loop

Reviewed-by: Casey Bodley <cbodley@redhat.com>

Merge pull request #20496 from pdvian/wip-23023-jewel

jewel: rgw: upldate the max-buckets when the quota is uploaded

Reviewed-by: Casey Bodley <cbodley@redhat.com>

Merge pull request #20749 from theanalyst/wip-jewel-curl-ssl

jewel: rgw: libcurl & ssl fixes

Reviewed-by: Casey Bodley <cbodley@redhat.com>

Merge pull request #19635 from linuxbox2/wip-rgw-jewel-20101

jewel: rgw: S3 POST policy should not require Content-Type

Reviewed-by: Casey Bodley <cbodley@redhat.com>

Merge pull request #19488 from dplyakin/jewel-get-website-error-code

jewel: rgw: fix GET website response error code

Reviewed-by: Casey Bodley <cbodley@redhat.com>

Merge pull request #20293 from smithfarm/wip-22572-jewel

jewel: rgw: stale bucket index entry remains after object deletion

Reviewed-by: Casey Bodley <cbodley@redhat.com>

Merge pull request #20292 from smithfarm/wip-22018-jewel

jewel: rgw: segmentation fault when starting radosgw after reverting .rgw.root

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

Merge pull request #20291 from smithfarm/wip-21954-jewel

jewel: rgw: list bucket which enable versioning get wrong result when user marker

Reviewed-by: Casey Bodley <cbodley@redhat.com>

Merge pull request #20269 from smithfarm/wip-21951-jewel

jewel: rgw: multisite: data sync status advances despite failure in RGWListBucketIndexesCR

Reviewed-by: Casey Bodley <cbodley@redhat.com>

Merge pull request #20262 from smithfarm/wip-21950-jewel

jewel: rgw: null instance mtime incorrect when enable versioning

Reviewed-by: Casey Bodley <cbodley@redhat.com>

Merge pull request #20179 from shinobu-x/wip-21655-jewel

jewel: rgw: add ability to sync user stats from admin api

Reviewed-by: Matt Benjamin <mbenjamin@redhat.com>

Merge pull request #19057 from shinobu-x/wip-22188-jewel

jewel: rgw: add cors header rule check in cors option request

Reviewed-by: Casey Bodley <cbodley@redhat.com>

Merge pull request #18772 from smithfarm/wip-22028-jewel

jewel: rgw: boto3 v4 SignatureDoesNotMatch failure due to sorting of sse-kms headers

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

ceph-osdomap-tool: Add compact comand option

Signed-off-by: David Zafman <dzafman@redhat.com>
(cherry picked from commit b1f397f4fc3313ee0325722156759bfe32286c6b)

rgw: rgw_admin use raii for curl init/teardown

Since a lot of rgw-admin commands can exit early

Signed-off-by: Abhishek Lekshmanan <abhishek@suse.com>
(cherry picked from commit 4667937e4cece63751d5406d75c140418910a823)

rgw: call setup and release curl handles in setup_curl

Since rgw admin can also use it which will fail otherwise.

Fixes: http://tracker.ceph.com/issues/23203
Signed-off-by: Abhishek Lekshmanan <abhishek@suse.com>
(cherry picked from commit 08c4c05f5d939976c9e1607b9338f0a9b287708d)

rgw: rgw_admin also inits curl

since we use http manager which in turn uses curl and uses curl multi
interfaces. While curl is initialized at the first call of curl_easy_init() this
method isn't guaranteed to be safe when multiple threads may call the function
since curl_global_init isn't reentrant. Calling curl_global_init via
rgw::curl::setup_curl which additionally sets up ssl interfaces etc. when
openssl is used as curl's ssl backend. Similarly moving rgw target link to
accomodate this change.

Signed-off-by: Abhishek Lekshmanan <abhishek@suse.com>
(cherry picked from commit 112ba0b4dae5972b520aec39fe99275a6dfa899d)

 Conflicts:
  src/rgw/CMakeLists.txt
  beast submodule related changes which are not a part of this patchset
  src/rgw/rgw_admin.cc
  trivial whitespace conflicts #

rgw: make frontend map optional for ssl init

Since rgw admin also needs to init curl globally where we don't have a frontend
map

Signed-off-by: Abhishek Lekshmanan <abhishek@suse.com>
(cherry picked from commit 55dde3df50a9735b81c6b31f5683ebda1dabd5e2)

rgw: setup locks for libopenssl

openssl <= 1.02 requires explicit callbacks for locking which libcurl doesn't
set. This causes random segmentation faults when openssl uses its global
structures across multiple threads. Providing a simple mutex lock/unlock
functions as a callback. We determine whether openssl is used for libcurl via
curl-config utility which should be installed as a part of our curl development
headers package. We also additionally check that the openssl version is < 1.1.0
which alleviates the need for these callbacks. In this patchset we have done the
following:

- move all curl related global init functionality under rgw::curl namespace
  since libcurl may need to set up various ssl libraries etc during its init
- introduce WITH_CURL_OPENSSL in cmake
  this checks the backend curl is deployed with using curl-config. Since curl
  devel is expected to be installed anyway, this binary should be available and
  can help identify the ssl backend curl was compiled with.
- we only setup the locks if beast/civetweb aren't terminated with ssl, since
  these libraries setup the locks anyway and we want to prevent double
  initialization of openssl. Also we pass in ~CURL_GLOBAL_SSL making curl not
  initialize openssl if civetwb/beast is initializing them. Unfortunately this
  flag is a noop from curl >= 7.57 wherein both the libraries will end up
  initializing openssl anyway, which might override certain settings like error
  strings if using openssl < 1.1

https://curl.haxx.se/libcurl/c/threadsafe.html
https://www.openssl.org/docs/man1.0.2/crypto/threads.html#DESCRIPTION

Fixes: http://tracker.ceph.com/issues/22951
Signed-off-by: Abhishek Lekshmanan <abhishek@suse.com>
Signed-off-by: Jesse Williamson <jwilliamson@suse.com>
(cherry picked from commit 4e30277e4b38694e6f8d61925736cd7c2cff515f)

 Conflicts:
  - CMakeLists.txt
  libcryptopp drop in master conflict
  - src/rgw/rgw_main.cc
  conflict as we removed the ac related headers in Luminous

Conflicts:
  CMake isn't supported in jewel, while the CMake changes were cherried,
  CMakelists for rgw does not exist in jewel and this commit doesn't fix that.
  So though CMake changes are backported, cmake build is untested and expected
  to be broken
src/include/config-h.in.cmake
  src/rgw/rgw_main.cc

jewel: build/rgw: configure if curl is linked against openssl

Since the Luminous/master change was cmake only, making an explicit autoconf
test.

Signed-off-by: Abhishek Lekshmanan <abhishek@suse.com>

rgw: reuse CURL* for keystone.

When using keystone with https (the recommended setting),
it is inefficient to start up a new SSL connection for
each and every operation.  Keeping a CURL* structure around
should reduce the cost of doing this.  This logic tries
to do so, but it also tries to free them fairly aggressively
(5-10 seconds).  This should still greatly reduce load on
keystone at peak times while not tying up excess resources.

Fixes: https://tracker.ceph.com/issues/23162
Signed-off-by: Marcus Watts <mwatts@redhat.com>
(cherry picked from commit 0c5cee1977eac232045571f4850e64be93271aec)

 Conflicts:
  src/rgw/rgw_common.h
  trivial conflict as lowercase_dash_http_attr isnt a part of L

build/link against curl w/ openssl not nss or gnutls. (debian)

curl + gnutls has some big significant performance hits at least
when doing keystone validation.  nss has long-term memory growth
issues.

Signed-off-by: Marcus Watts <mwatts@redhat.com>
(cherry picked from commit cb12e1afd5beedf1b8c0b85e24e412e34c2fc427)

Merge pull request #20061 from cbodley/wip-jewel-rgw-log-trim

jewel: rgw: automated trimming of datalog and mdlog

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

rgw: fix ECANCELED handling in rgw_get_system_obj()

Fixes: http://tracker.ceph.com/issues/17996
Signed-off-by: Yehuda Sadeh <yehuda@redhat.com>
(cherry picked from commit e67a3b5b5c63ee3a34f06bdbe91d3256c6660a89)

Conflicts:
    src/rgw/rgw_rados.cc: used ctx.invalidate(obj) instead of ctx.raw.invalidate(obj)
        in RGWRados::SystemObject::invalidate_state()

Merge pull request #20557 from smithfarm/wip-really-disable-pidfile-test

test/CMakeLists: disable test_pidfile.sh

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

Client: mkdirs bugs when two clients send mkdir request for a same dir

Fixes: http://tracker.ceph.com/issues/20592
Signed-off-by: dongdong tao <tdd21151186@gmail.com>
(cherry picked from commit 80bdd3fb4afd262ee3073b4e8377da9238bdaef2)

Conflicts:
        src/client/Client.cc (jewel uses "uid, gid" instead of "perms")

test/CMakeLists: disable test_pidfile.sh

Too flaky, see http://tracker.ceph.com/issues/20975

Signed-off-by: Sage Weil <sage@redhat.com>
(cherry picked from commit ff3de2304497544033837bb8d0c809a9e54a3e6e)

Conflicts:
    src/test/Makefile.am: applied change manually because jewel uses autotools

Merge pull request #20344 from pdvian/wip-22866-jewel

jewel: mon/OSDMonitor: fix dividing by zero in OSDUtilizationDumper

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

Merge pull request #20463 from dzafman/wip-23006

test: Adjust for Jewel quirk caused of differences with master

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

Merge pull request #20446 from smithfarm/wip-23010-jewel

jewel: Filestore rocksdb compaction readahead option not set by default

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

Merge pull request #20289 from smithfarm/wip-21239-jewel

jewel: test_health_warnings.sh can fail

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

Merge pull request #20284 from dzafman/wip-22912

jewel: ceph-objectstore-tool: "$OBJ get-omaphdr" and "$OBJ list-omap" scan all pgs instead of using specific pg

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

Merge pull request #19611 from ukernel/jewel-22494

jewel: include/fs_types: fix unsigned integer overflow

Reviewed-by: Patrick Donnelly <pdonnell@redhat.com>
Reviewed-by: Amit Kumar <amitkuma@redhat.com>

Merge pull request #20520 from smithfarm/wip-disable-test-pidfile

test/CMakeLists: disable test-pidfile.sh

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

test/librbd: utilize unique pool for cache tier testing

Fixes: http://tracker.ceph.com/issues/11502
Signed-off-by: Jason Dillaman <dillaman@redhat.com>
(cherry picked from commit c42c54d0ec643b272aff225134dd0eceae946af0)

Conflicts:
src/test/librbd/test_librbd.cc: Resolved in TEST_F

test/CMakeLists: disable test-pidfile.sh

Signed-off-by: Sage Weil <sage@redhat.com>
(cherry picked from commit c4936f4493d34341aab8f01488b45dc981d4ef3f)

osd/PrimaryLogPG: cancel all objecter ops atomically

We want to avoid a situation like this:

- start proxy op A (epoch E)
- start proxy op B (epoch E)
- start proxy op C (epoch E)
- objecter sends none of these because target is down in epoch E
- osdmap update to E+1
- pg cancels requeues A, B
- objecter updates to E+1
- objecter sends C
- pg cancels/requeues C

Note that the key thing is that operations on each object are canceled
atomically.  On the interval change we do it all at once.  In the other
cases, we cancel everything on the given object together.

Fixes: http://tracker.ceph.com/issues/22123
Signed-off-by: Sage Weil <sage@redhat.com>
(cherry picked from commit 93fd56ed039363c4f169259a0a560b968d1a0333)

Conflicts:
src/osd/PrimaryLogPG.cc (manually backported changes from src/osd/ReplicatedPG.cc)
src/osd/PrimaryLogPG.h (manually backported changes from src/osd/ReplicatedPG.cc)

osdc/Objecter: allow multiple ops to be canceled atomically

This allows multiple cancellations without worrying about a race with an
(objecter) osdmap update.

Signed-off-by: Sage Weil <sage@redhat.com>
(cherry picked from commit 7459d4a4794eb092cdc3ee06af12b782c892193d)

qa/tasks/ceph.py: fix timing of wait-for-* and osd markdown

Mark down osds, *then* wait for them to come up or for the cluster to be
healthy!

Signed-off-by: Sage Weil <sage@redhat.com>
(cherry picked from commit 86c0d07e32205e2b6aa417a0e4ae03f0084a1888)

tasks/ceph.restart osd fix

ceph.restart should mark restarted osds down in order to avoid a
race condition with ceph_manager.wait_for_clean

Fixes: http://tracker.ceph.com/issues/15778
Signed-off-by: Warren Usui <wusui@redhat.com>
(manual cherry pick of 1b7552c9cb331978cb0bfd4d7dc4dcde4186c176)

Conflicts:
    qa/tasks/ceph.py (original commit was in ceph/ceph-qa-suite.git)

Merge pull request #17707 from liewegas/wip-17445

jewel: osd: also check the exsistence of clone obc for "CEPH_SNAPDIR" requests

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

Merge pull request #17883 from andyli029/wip-jewel-signal-handler-typo

jewel: core: global/signal_handler.cc: fix typo

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

Merge pull request #19906 from jdurgin/jewel-filestore-split

jewel: HashIndex: randomize split threshold by a configurable amount

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

Merge pull request #20331 from dzafman/wip-21036

jewel: snapset xattr corruption propagated from primary to other shards

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

Merge pull request #18294 from smithfarm/wip-21796-jewel

jewel: Ubuntu amd64 client can not discover the ubuntu arm64 ceph cluster

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

Merge pull request #18850 from vumrao/vumrao-jewel-rocksdb-cache-size

jewel: common/config: set rocksdb_cache_size to OPT_U64

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

Merge pull request #19330 from ovh/bp-21862-jewel

jewel: tools/ceph-conf: dump parsed config in plain text or as json

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

Merge pull request #19927 from jdurgin/wip-jewel-omap-recovery

jewel: config: lower default omap entries recovered at once

Reviewed-by: Nathan Cutler <ncutler@suse.com>

Merge pull request #19978 from pdvian/wip-22693-jewel

jewel: common: compute SimpleLRU's size with contents.size() instead of lru

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

Merge pull request #20108 from tchaikov/wip-18004-jewel

jewel: osd: update heartbeat peers when a new OSD is added

Reviewed-by: Sage Weil <sage@redhat.com>
Reviewed-by: Josh Durgin <jdurgin@redhat.com>
Reviewed-by: Pan Liu <wanjun.lp@alibaba-inc.com>

Merge pull request #20146 from smithfarm/wip-22818-jewel

jewel: repair_test fails due to race with osd start

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

Merge pull request #20306 from tchaikov/wip-pr-16661-jewel

jewel: throttle: Minimal destructor fix for Luminous

Reviewed-by: Josh Durgin <jdurgin@redhat.com>
Reviewed-by: Adam C. Emerson <aemerson@redhat.com>

rgw: upldate the max-buckets when the quota is uploaded

Fixes:https://tracker.ceph.com/issues/22745

Signed-off-by: zhaokun <develop.hikdata.com>
(cherry picked from commit 0f3dd4841ee98351fe94dfc1d4dbb37ffcbb4fbb)

rgw: fix the max-uploads parameter not work

fix http://tracker.ceph.com/issues/22825

Signed-off-by: Xin Liao <liaoxinbit@gmail.com>
(cherry picked from commit 2ff9ea1568c4f5bd2bccbaa63eee3b58ebb9ca52)

test: Adjust for Jewel quirk caused of differences with master

Fixes: http://tracker.ceph.com/issues/23006
Signed-off-by: David Zafman <dzafman@redhat.com>

Merge pull request #18780 from shinobu-x/wip-21053-jewel

jewel: RHEL 7.3 Selinux denials at OSD start

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

Merge pull request #20143 from smithfarm/wip-21872-jewel

jewel: ObjectStore/StoreTest.FiemapHoles/3 fails with kstore

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

common/options.cc: Set Filestore rocksdb compaction readahead option.

fixes: http://tracker.ceph.com/issues/21505

Signed-off-by: Mark Nelson <mnelson@redhat.com>
(cherry picked from commit 66567e573836c570040c8d8148c34b0cf9dc9ce2)

Conflicts:
src/common/options.cc (master commit made the change in
           src/common/options.cc, which doesn't exist in jewel)

osdc/Journaler: make sure flush() writes enough data

If _do_flush() fails to write all data because of prezere, _do_flush()
should be called again when prezero finishes. Otherwise flush waiters
may sleep forever.

Fixes: http://tracker.ceph.com/issues/22824
Signed-off-by: "Yan, Zheng" <zyan@redhat.com>
(cherry picked from commit 94a2076b9de9e8ffb73f3d5c873bdb87a0429e65)

rbd-mirror: fix potential infinite loop when formatting status message

The improvements include:

- tag_tid values should always be increasing, so loop only if
  master.tag_tid > mirror_tag_tid in calculate_behind_master_or_send_update;
- in send_update_tag_cache don't refetch a tag if it is already in the
  cache;
- make fake tags with tag_data.predecessor.tag_tid set to zero;
- make sure the new tag is inserted to the cache if an old entry
  with this id happens to exist.

Fixes: http://tracker.ceph.com/issues/22932
Signed-off-by: Mykola Golub <mgolub@suse.com>
(cherry picked from commit 2fb99b5ed39c36879d1b6180ac47629bc5a3b315)

mon/OSDMonitor: fix deviding by zero in OSDUtilizationDumper

Signed-off-by: Mingxin Liu <mingxin@xsky.com>
(cherry picked from commit 9fe103233eaee965849c00c66c1b352f31932e01)

Merge pull request #19907 from pdvian/wip-22590-jewel

jewel: cephfs: ceph.in: pass RADOS inst to LibCephFS

Reviewed-by: Patrick Donnelly <pdonnell@redhat.com>

Merge pull request #20067 from smithfarm/wip-22762-jewel

jewel: mds: unbalanced auth_pin/auth_unpin in RecoveryQueue code

Reviewed-by: Yan, Zheng <zyan@redhat.com>