# v10.2.3



* Backport #15608: jewel: Ownership of run dir, pid file and asok file not set when deferring permissions drop
* Backport #15700: jewel: rados/test.sh workunit timesout on OpenStack
* Backport #15767: jewel: list-inconsistent-obj and list-inconsistent-snapset improvements
* Backport #15768: jewel: FileStore: umount hang because sync thread doesn't exit
* Backport #15806: jewel: New pools have bogus "stuck inactive/unclean" HEALTH_ERR messages until they are first active and clean
* Backport #15841: jewel: s3website: x-amz-website-redirect-location header returns malformed HTTP response
* Backport #15898: jewel: Confusing MDS log message when shut down with stalled journaler reads
* Backport #15954: jewel: rgw: initial slashes are not properly handled in Swift's BulkDelete
* Backport #15960: jewel: rgw: custom metadata aren't camelcased in Swift's responses
* Backport #15964: jewel: realm pull fails when using apache frontend
* Backport #15965: jewel: No Last-Modified, Content-Size and X-Object-Manifest headers if no segments in DLO manifest
* Backport #15967: jewel: rgw: account/container metadata not actually present in a request are deleted during POST through Swift API
* Backport #15968: jewel: ceph status mds output ignores active MDS when there is a standby replay
* Backport #15978: jewel: rgw: leak
* Backport #15998: jewel: most admin operations see a failed to create default zone error message
* Backport #15999: jewel: CephFSVolumeClient: read-only authorization for volumes
* Backport #16009: radosgw-admin: failure for user create after upgrade from hammer to jewel
* Backport #16037: jewel: MDSMonitor::check_subs() is very buggy
* Backport #16039: jewel: rgw: updating custom metadata on account/container is broken
* Backport #16040: jewel: rgw: updating CORS/ACLs might not work in some circumstances
* Backport #16041: jewel: mds/StrayManager.cc: 520: FAILED assert(dnl->is_primary())
* Backport #16071: jewel: remove radosgw-admin temp command as it got deprecated 
* Backport #16080: jewel: osd:sched_time not actually randomized
* Backport #16081: jewel: mon: "mon metadata" fails when only one monitor exists
* Backport #16083: jewel: mds: wrongly treat symlink inode as normal file/dir when symlink inode is stale on kcephfs
* Backport #16085: jewel:  A query on a static large object fails with 404 error
* Backport #16086: jewel: 'radosgw-admin zone modify' clears master_zone if no --master is given
* Backport #16099: jewel: ceph-base requires parted
* Backport #16109: jewel: retry on bucket sync failures
* Backport #16112: jewel: CORS: Access-Control-Allow-Origin should return * when set that way
* Backport #16117: jewel: rgw: aws4 parsing issue
* Backport #16135: jewel: MDS: fix getattr starve setattr
* Backport #16136: jewel: MDSMonitor fixes
* Backport #16148: jewel: Scrub error: 0/1 pinned
* Backport #16150: jewel: crash adding snap to purged_snaps in ReplicatedPG::WaitingOnReplicas
* Backport #16152: jewel: fs: client: fstat cap release
* Backport #16153: jewel: Missing export for rados_aio_get_version in src/include/rados/librados.h
* Backport #16163: jewel: rgw: cannot limit users to not create buckets
* Backport #16182: jewel: backport static sites fixes master->jewel
* Backport #16193: jewel: Fixes for list-inconsistent-*
* Backport #16194: jewel: Earlier version of jq don't have -S option
* Backport #16215: jewel: client: crash in unmount when fuse_use_invalidate_cb is enabled
* Backport #16232: jewel: Improve rbd-mirror test case coverage
* Backport #16249: jewel: sparse_read on ec pool should return extends with correct offset
* Backport #16272: jewel: rgw ldap: fix ldap bindpw parsing
* Backport #16299: jewel: mds: fix SnapRealm::have_past_parents_open()
* Backport #16311: jewel: Add STREAMING-AWS4-HMAC-SHA256-PAYLOAD support
* Backport #16312: jewel: selinux denials in RGW
* Backport #16315: jewel: When journaling is enabled, a flush request shouldn't flush the cache
* Backport #16319: jewel: radosgw-admin: inconsistency in uid/email handling
* Backport #16320: jewel: fs: fuse mounted file systems fails SAMBA CTDB ping_pong rw test with v9.0.2
* Backport #16324: jewel: rbd: create error: (38) Function not implemented in upgrade:client
* Backport #16338: jewel : rados bench : add cleanup message with time it has taken to delete the objects when cleanup start for written objects 
* Backport #16339: jewel : rgw : support size suffixes for --max-size in radosgw-admin command 
* Tasks #16344: jewel v10.2.3
* Backport #16371: jewel: rbd-mirror: ensure replay status formatter has completed before stopping replay
* Backport #16372: jewel: Unable to disable journaling feature if in unexpected mirror state
* Backport #16373: jewel: rbd-mirror: gracefully handle missing sync point snapshots
* Backport #16374: jewel: AsyncConnection::lockmsg/async lockdep cycle: AsyncMessenger::lock, MDSDaemon::mds_lock, AsyncConnection::lock
* Backport #16380: jewel: msg/async: connection race hang
* Backport #16381: jewel: comparing return code to ERR_NOT_MODIFIED in rgw_rest_s3.cc (needs minus sign)
* Backport #16392: jewel: rgw: master: build failures with boost > 1.58
* Backport #16393: jewel: rgw: Swift's object versioning doesn't support restoring a previous version on DELETE
* Backport #16422: jewel: Sporadic test failures when image update notification missed
* Backport #16423: jewel: Journal duplicate op detection can cause lockdep error
* Backport #16424: jewel: Journal needs to handle duplicate maintenance op tids
* Backport #16425: jewel: rbd-mirror: potential race condition accessing local image journal 
* Backport #16426: jewel: Possible race condition during journal transition from replay to ready
* Backport #16427: jewel: prepare_pgtemp needs to only update up_thru if newer than the existing one
* Backport #16429: jewel: OSDMonitor: drop pg temps from not the current primary
* Backport #16431: jewel: librados,osd: bad flags can crash the osd
* Backport #16437: jewel: async messenger mon crash
* Backport #16438: jewel: Delete image when a resync is requested
* Backport #16459: jewel: rbd-mirror should disable proxied maintenance ops for non-primary image
* Backport #16460: jewel: Crash when utilizing advisory locking API functions
* Backport #16461: jewel: ceph Resource Agent does not work with systemd
* Backport #16482: jewel: Timeout sending mirroring notification shouldn't result in failure
* Backport #16483: jewel: Close journal and object map before flagging exclusive lock as released
* Backport #16484: jewel: ExclusiveLock object leaked when switching to snapshot
* Backport #16485: jewel: Whitelist EBUSY error from "snap unprotect" for journal replay
* Backport #16486: jewel: Object map/fast-diff invalidated if journal replays the same snap remove event
* Backport #16487: jewel: async_messenger: implement ms_inject_delay*
* Backport #16507: jewel: "ceph_test_librbd_api: symbol lookup error: ceph_test_librbd_api: undefined symbol" in upgrade:client-upgrade-jewel-distro-basic-smithi
* Backport #16511: jewel: rbd-mirror: 'wait_for_scheduled_deletion' callback might deadlock
* Backport #16512: jewel: rbd-mirror: potential race condition when restarting image replayer
* Backport #16513: jewel: rbd-mirror: image-replayer: segfault when removing resync listener
* Backport #16514: jewel: Image removal doesn't necessarily clean up all rbd_mirroring entries
* Backport #16515: jewel: Session::check_access() is buggy
* Backport #16518: jewel: TaskFinisher: cancel all tasks wait until finisher done
* Backport #16520: jewel: librbd: potential use after free on refresh error
* Backport #16547: jewel:ObjectCacher: doesn't correctly handle read replies on split BufferHeads
* Backport #16549: jewel: Monitor die if moncommand without "prefix" item
* Backport #16560: jewel: mds: enforce a dirfrag limit on entries
* Backport #16565: jewel: rgw: data sync stops after getting error in all data log sync shards
* Backport #16576: jewel: rbd-mirror: FAILED assert(m_local_image_ctx->object_map != nullptr)"
* Backport #16577: jewel : 60-ceph-partuuid-workaround-rules still needed by debian jessie (udev 215-17)
* Backport #16586: jewel: partprobe intermittent issues during ceph-disk prepare
* Backport #16589: jewel: multisite sync races with deletes
* Backport #16593: jewel: FAILED assert(object_no < m_object_map.size())
* Backport #16599: jewel: rgw: Swift API returns double space usage and objects of account metadata
* Backport #16601: jewel: librbd: fix missing return statement if failed to get mirror image state
* Backport #16620: jewel: Fix shutting down mds timed-out due to deadlock
* Backport #16621: jewel: mds: `session evict` tell command blocks forever with async messenger (TestVolumeClient.test_evict_client failure)
* Backport #16625: jewel: Failing file operations on kernel based cephfs mount point leaves unaccessible file behind on hammer 0.94.7
* Backport #16636: jewel: rgw: document multi tenancy
* Backport #16637: jewel: add socket backlog setting for via ceph.conf
* Backport #16658: jewel: rbd-mirror: gracefully handle being blacklisted
* Backport #16659: jewel: ReplicatedBackend doesn't increment stats on pull, only push
* Backport #16696: jewel: segfault in RGWBucketShardIncrementalSyncCR
* Backport #16697: jewel: ceph-fuse is not linked to libtcmalloc
* Backport #16699: jewel: multidelete query parameter not correctly parsed
* Backport #16700: jewel: rgw: segmentation fault on error_repo in data sync
* Backport #16701: jewel: rbd-mirror: image sync throttle needs to use pool id + image id to form unique key
* Backport #16702: jewel: multisite: bucket sync failures with tenant users
* Backport #16731: jewel: failed to create bucket after upgrade from hammer to jewel
* Backport #16732: jewel: Bucket index shards orphaned after bucket delete
* Backport #16735: jewel: rbd-nbd does not properly handle resize notifications
* Backport #16747: jewel: rbd-mirror: snap rename does not correctly replicate
* Backport #16748: jewel: mount.ceph: move from ceph-base to ceph-common and add symlink in /sbin for SUSE
* Backport #16750: jewel: ceph-osd-prestart.sh contains Upstart-specific code
* Backport #16778: multisite: 400-error with certain complete multipart upload requests
* Backport #16795: jewel: libatomic-ops-devel was renamed in May 2012 - fix in ceph.spec
* Backport #16796: jewel: Renaming old format image results in "Transport endpoint is not connected" error
* Backport #16797: jewel: MDS Deadlock on shutdown active rank while busy with metadata IO
* Backport #16798: jewel: ceph command line tool chokes on ceph –w (the dash is unicode 'en dash' &ndash, copy-paste to reproduce)
* Backport #16830: jewel: CephFSVolumeClient: List authorized IDs by share
* Backport #16831: jewel: Add versioning to CephFSVolumeClient interface
* Backport #16862: jewel: "default" zone and zonegroup cannot be added to a realm
* Backport #16863: jewel: use zone endpoints instead of zonegroup endpoints
* Backport #16864: jewel: multisite segfault on ~RGWRealmWatcher if realm was deleted
* Backport #16865: jewel: saw valgrind issues in ReplicatedPG::new_repop
* Backport #16867: jewel: mkfs.xfs slow performance with discards and object map
* Backport #16869: jewel: Discard hangs when 'rbd_skip_partial_discard' is enabled
* Backport #16901: jewel: segfault in RGWOp_MDLog_Notify
* Backport #16903: jewel: Non-primary image is recording journal events during image sync 
* Backport #16904: jewel: journal should prefetch small chunks of the object during replay
* Backport #16915: Jewel: OSD crash with Hammer to Jewel Upgrade: void FileStore::init_temp_collections()
* Backport #16934: jewel: Add zone rename to radosgw_admin
* Backport #16945: jewel: RGW/civetweb no longer listens on IPv6: invalid port spec
* Backport #16950: jewel: librbd/ExclusiveLock.cc: 197: FAILED assert(m_watch_handle != 0)
* Backport #16958: jewel: Bug when using port 443s in rgw.
* Backport #16959: jewel: rpm: OBS needs ExclusiveArch
* Backport #16969: jewel: src/script/subman fails with KeyError: '\n"band'
* Backport #16978: jewel: rbd-mirror: FAILED assert(m_on_update_status_finish == nullptr)
* Backport #17004: jewel: rbd-mirror: FAILED assert(m_state == STATE_STOPPING)
* Backport #17005: jewel: ImageReplayer::is_replaying does not include flush state 
* Backport #17006: jewel: Increase log level for messages occuring while running rgw admin command
* Backport #17026: jewel: doc: format 2 now is the default image format
* Backport #17032: jewel: multisite: RGWPeriodPuller tries to pull from itself
* Backport #17033: jewel: rgw: there may be some objects not delete in some circumstances
* Backport #17034: jewel: rgw: object expirer's hints might be trimmed without processing in some circumstances
* Backport #17061: jewel: bashism in src/rbdmap 
* Backport #17063: jewel: Throttle in-flight image syncs to only a X concurrent
* Backport #17066: jewel: rbd-mirror: remove ceph_test_rbd_mirror_image_replay test case
* Backport #17080: jewel: the option 'rbd_cache_writethrough_until_flush=true' dosn't work 
* Backport #17088: jewel: Sporadic failure in TestImageReplayer.StartReplayAndWrite
* Backport #17089: jewel: OSD failed to subscribe skipped osdmaps after "ceph osd pause"
* Backport #17092: jewel: build/ops: need rocksdb commit 7ca731b12ce for ppc64le build
* Backport #17126: mds: fix double-unlock on shutdown
* Backport #17147: jewel: rgw multisite "ERROR" messages during "normal operation"
* Backport #17485: jewel: Periodically update the sync point object number during sync