# v16.2.6



* Backport #49377: pacific: building libcrc32
* Backport #49555: pacific: notifications: handle exceptions inside the persistent notification thread
* Backport #50094: pacific: notification: librgw is not initializing kafka and amqp libraries
* Backport #50156: pacific: notification: persistent notifications stop when ack-levl=none
* Backport #50157: pacific: notifications: setting "persistent=false" creates a persistent topic
* Backport #50209: pacific: 'list index out of range'
* Backport #50214: pacific: rgw: object lock: improve client error messages
* Backport #50240: pacific: clang build fails on rgw_log_backing.cc
* Backport #50367: pacific: rgw: during reshard lock contention, adjust logging
* Backport #50379: pacific: ceph-16.2.0 builds have started failing in Fedora 35/rawhide w/ librabbitmq-0.11.0
* Backport #50386: pacific: SimpleRADOSStriper: use debug_cephsqlite for SimpleRadosStriper dout
* Backport #50463: pacific: rgw admin hides sync enable/disable failure on non-master
* Backport #50465: pacific: per bucket notification object is never deleted
* Backport #50599: pacific: Ceph-osd refuses to bind on an IP on the local loopback lo
* Backport #50642: pacific: assumed-role: s3api head-object returns 403 Forbidden, even if role has ListBucket, for non-existent object
* Backport #50645: pacific: rgw: allow rgw-orphan-list to process multiple data pools
* Backport #50679: pacific: Reproducible crash in radosgw (nautilus and later)
* Backport #50708: pacific: rgw: fix bucket object listing when initial marker matches prefix
* Backport #50711: pacific: radosgw-admin mfa resync crashes when supplied with only one totp-pin
* Backport #50728: pacific: multisite: crash in RGWRESTStreamRWRequest::do_send_prepare() with empty url
* Backport #50731: pacific: rgw_file: RGWLibFS::read success executed, but nodata readed
* Backport #50748: pacific: max_misplaced was replaced by target_max_misplaced_ratio
* Backport #50803: pacific: slow radosgw-admin startup when large value of rgw_gc_max_objs configured
* Backport #50845: pacific: deprecate civetweb in pacific
* Backport #50862: pacific: add ceph-volume lvm [new-db|new-wal|migrate] commands
* Backport #50896: pacific: ceph-volume lvm activate will consider /dev/root mounted directories as 'unmounted' and mount tmpfs on top of them
* Backport #50900: pacific: PGs always go into active+clean+scrubbing+deep+repair in the LRC
* Backport #50907: pacific: [rbd-nbd] default pool isn't picked up
* Backport #50935: pacific: osd-bluefs-volume-ops.sh fails 
* Backport #50991: pacific: mgr/nfs: skipping conf file or passing empty file throws traceback
* Backport #51014: pacific: rgw: remove quota soft threshold
* Backport #51080: pacific: Bad error message on bucket chown
* Backport #51081: pacific: multisite: full sync broken for tenanted buckets
* Backport #51130: pacific: In poweroff conditions BlueFS can create corrupted files
* Backport #51140: pacific: radosgw-admin user create error message is confusing if user with supplied email address already exists
* Backport #51174: pacific: mgr/nfs: add nfs-ganesha config hierarchy
* Backport #51188: pacific: mgr/telemetry: pass leaderboard flag even w/o ident
* Backport #51198: pacific: msg: active_connections regression
* Backport #51200: pacific: qa: test_subvolume_clone_in_progress_snapshot_rm is racy
* Backport #51221: pacific: NFS bucket removal can crash
* Backport #51241: pacific: mgr module fails in focal, due to ceph-mgr-rook module
* Backport #51285: pacific: mds: unknown metric type is always -1
* Backport #51290: pacific: mon: stretch mode clusters do not sanely set default crush rules
* Backport #51310: pacific: Add Support customed CA certificate from vault KMS for SSE encryption
* Backport #51329: pacific: rgw: deleting a bucket results in infinite loop
* Backport #51374: pacific: mgr/dashboard: Grafana dashboards clean-up
* Backport #51411: pacific: pybind/mgr/volumes: purge queue seems to block operating on cephfs connection required by dispatch thread
* Backport #51424: pacific: mgr module INFO should be logged by default
* Backport #51467: Backport Bucket Remove Op: Pass in user 
* Backport #51471: pacific: User has assume_role permission can access to any bucket
* Backport #51498: pacific: mgr spamming with repeated set pgp_num_actual while merging
* Backport #51499: pacific: qa: test_ls_H_prints_human_readable_file_size failure
* Backport #51500: pacific: qa: FileNotFoundError: [Errno 2] No such file or directory: '/sys/kernel/debug/ceph/3fab6bea-f243-47a4-a956-8c03a62b61b5.client4721/mds_sessions'
* Backport #51514: pacific: notifications: metadata does not work for COPY events
* Backport #51542: pacific: mgr: radosgw: include realm_{id,name} in service map
* Backport #51544: pacific: mgr/volumes: use a dedicated libcephfs handle for subvolume API calls
* Backport #51547: pacific: src/pybind/mgr/mirroring/fs/snapshot_mirror.py: do not assume a cephfs-mirror daemon is always running
* Backport #51549: pacific: cephadm bootstrap on arm64 fails to start ceph/ceph-grafana service 
* Backport #51553: pacific: rebuild-mondb hangs
* Backport #51556: pacific: mon: return -EINVAL when handling unknown option in 'ceph osd pool get'
* Backport #51558: pacific: 400 error caused by multi-subtag language in Accept-Language header
* Backport #51568: pacific: pool last_epoch_clean floor is stuck after pg merging
* Backport #51570: pacific: CommandCrashedError: Command crashed: 'mkdir -p -- /home/ubuntu/cephtest/mnt.0/client.0/tmp && cd -- /home/ubuntu/ceph...
* Backport #51579: pacific: FTBFS with fmt-8.x
* Backport #51584: pacific: osd does not proactively remove leftover PGs
* Backport #51603: pacific: qa/standalone: Add missing teardowns at the end of a subset of osd and scrub tests
* Backport #51605: pacific: bufferlist::splice() may cause stack corruption in bufferlist::rebuild_aligned_size_and_memory()
* Backport #51636: pacific: monitoring: fix Physical Device Latency unit
* Backport #51649: pacific: Bluestore repair might erroneously remove SharedBlob entries.
* Backport #51657: pacific: mgr/dashboard: remove usage of 'rgw_frontend_ssl_key'
* Backport #51663: pacific: rados/test_envlibrados_for_rocksdb.sh: cmake: symbol lookup error: cmake: undefined symbol: archive_write_add_filter_zstd in centos
* Backport #51664: pacific: bluestore doesn't respect "bluestore_warn_on_spurious_read_errors" config parameter
* Backport #51675: pacific: Manual datalog trim segfaaults
* Backport #51679: pacific: Potential race condition in robust notify
* Backport #51685: pacific: mgr/dashboard: test_error force maintenance dashboard check
* Backport #51696: pacific: rgw: fail as expected when set-bucket-website attempted on a non-existent bucket
* Backport #51699: pacific: rgw: beast: lack of TLS settings
* Backport #51731: pacific: mgr/dashboard: Add configurable MOTD or wall notification
* Backport #51751: pacific: rgw_orphan_list.sh causing a crash in the OSD
* Backport #51766: pacific: scrub/osd-scrub-repair.sh: TEST_auto_repair_bluestore_tag failed
* Backport #51771: pacific: unittest_rgw_dmclock_scheduler Failed
* Backport #51779: pacific: Creating a role in another tenant seems to be possible
* Backport #51781: pacific: Return correct XML response when delete role fail.
* Backport #51785: pacific: multisite: metadata sync does not keep retrying failed entries
* Backport #51801: pacific: radosgw-admin should print error on missing --end-marker argument
* Backport #51804: pacific: http/notifications: "Content-Type" is missing in POST messages
* Backport #51819: pacific: cephfs-mirror: removing a mirrored directory path causes other sync failures for other directory paths
* Backport #51830: pacific: set a non-zero default value for osd_client_message_cap
* Backport #51832: pacific: mds: META_POP_READDIR, META_POP_FETCH, META_POP_STORE, and cache_hit_rate are not updated
* Backport #51833: pacific: client: flush the mdlog in unsafe requests' relevant and auth MDSes only
* Backport #51834: pacific: mon/MDSMonitor: allow creating a file system with a specific fscid
* Backport #51836: pacific: ceph.spec: after being eliminated, FIRST_ARG crept back in
* Backport #51840: pacific: osd: snaptrim logs to derr at every tick
* Backport #51851: pacific: ci/tests: update ansible configuration
* Backport #51852: pacific: ZeroDivisionError when specifying only LVs as --db-devices
* Backport #51932: pacific: mds: MDCache.cc:5319 FAILED ceph_assert(rejoin_ack_gather.count(mds->get_nodeid()))
* Backport #51935: pacific: mds: improve debugging for mksnap denial
* Backport #51937: pacific: qa: test_full_fsync (tasks.cephfs.test_full.TestClusterFull) failed
* Backport #51940: pacific: MDSMonitor: monitor crash after upgrade from ceph 15.2.13 to 16.2.4
* Backport #51951: pacific: insights module can generate too much data, fail to put in config-key
* Backport #51955: pacific: Change of public type in snappy
* Backport #51958: pacific: tests: destroying OSDs fails because osd not yet marked as down
* Backport #51969: pacific: rpm/luarocks: simplify conditional and support Leap 15.3
* Backport #51977: pacific: client: make sure only to update dir dist from auth mds
* Backport #51983: pacific: mon/MDSMonitor: do not pointlessly kill standbys that are incompatible with current CompatSet
* Support #51985: Help-seeking for rbd persistent cache usage problems
* Backport #51992: pacific: monitoring/grafana/cluster: use per-unit max and limit values
* Backport #51994: pacific: mgr/dashboard: cephadm-e2e script: improvements
* Backport #52006: pacific: [pybind] mirror_image_get_status() throws TypeError if remote_statuses isn't empty
* Backport #52007: pacific: rgw: avoid occuring radosgw daemon crash when access a conditionally redirected static website
* Backport #52024: pacific: kv/RocksDBStore: enrich debug message
* Backport #52036: pacific: mon/MDSMonitor.cc: fix join fscid not applied with pending fsmap at boot
* Backport #52048: pacific: mgr/dashboard: fix ssl cert validation for rgw service creation
* Backport #52049: pacific: mgr/dashboard: error on showing rgw svc. perf. counters
* Backport #52051: pacific: rgw: when an already-deleted object is removed in a versioned bucket, an unneeded delete marker is created
* Backport #52084: pacific: pybind/mgr/stats: KeyError
* Backport #52106: pacific: rbd-mirror: potential hang on shutdown
* Backport #52133: pacific: Phantom Atari partitions when using raw mode
* Backport #52135: pacific: mgr/dashboard: monitoring: replace Grafana JSON with code-based generators
* Backport #52244: pacific: Deferred writes are unexpectedly applied to large writes on spinners
* Backport #52246: pacific: cannot enable osd resharding on pacific
* Backport #52248: pacific: mgr/dashboard: tox.ini: delete useless env. 'apidocs'
* Backport #52250: pacific: mgr/dashboard: cephadm e2e start script: add "--expanded" option
* Backport #52291: pacific: mgr/dashboard: rename "inventory" to "disks"
* Backport #52293: pacific: mgr/dashboard: Creation of snapshot for subvol's folder is failing with 500 Internal error
* Backport #52294: pacific: mgr/dashboard: Request for refresh button in iscsi targets page
* Backport #52308: pacific: mgr/dashboard: unable to completely silence alerts in dashboard
* Backport #52325: pacific: mgr/dashboard: redirect to active mgr ipv6 lacks [], fails
* Backport #52330: pacific: multisite upgrade: error repo entries before upgrade can't be overwritten or removed
* Backport #52336: pacific: ceph df detail reports dirty objects without a cache tier
* Backport #52378: pacific: mgr/dashboard: stats=false not working when listing buckets
* Backport #52380: pacific: mgr/dashboard: fix grafonnet build error (FTBFS when building on a system that is not connected to the Internet)
* Bug #52415: rocksdb: build error with rocksdb-6.22.x
* Backport #52433: pacific: BlueFS might end-up with huge WAL files when upgrading OMAPs
* Backport #52449: pacific: remove remaining references to lsb_release
* Backport #52452: pacific: "rbd unmap" misbehaves if the image name contains glob metacharacters
* Backport #52461: pacific: rbd-mirrror: expose perf dump info for snapshot replayer
* Backport #52471: pacific: ceph.spec.in: drop gdbm from build deps
* Backport #52477: pacific: BlueFS superblock might contain incomplete list of physical extents
* Backport #52492: pacific: bluefs mount failed to replay log: (5) Input/output error
* Backport #52518: pacific: ceph-bluestore-tool:bluefs-bdev-migrate might result in broken OSD
* Backport #52519: pacific: pg_autoscaler causes device_health_metrics pool to use 128 pgs preventing rgw from being deployed 
* Backport #52525: pacific: pvs --no-heading fail
* Backport #52538: pacific: raw: list is failing when using logical partition on the host
* Backport #52542: pacific: documentation: add lvm new-db, new-wal and migrate
* Backport #52545: pacific: ceph-volume lvm migrate can't work with container
* Backport #52547: pacific: ceph-volume can't deactivate all
* Backport #52554: pacific: ceph-volume lvm migrate without args raises an exception
* Backport #52560: pacific: UnboundLocalError: local variable 'tags' referenced before assignment