# v17.2.7



* Backport #55301: quincy: Hybrid allocator might return duplicate extents when performing online free dump.
* Backport #55614: quincy: GetBucketTagging returns wrong NoSuchTagSetError instead of NoSuchTagSet
* Backport #56677: quincy: cls_rgw: nonexists object shoud not be accounted when check_index
* Backport #57117: quincy: mon: race condition between `mgr fail` and MgrMonitor::prepare_beacon()
* Backport #57158: quincy: doc: update snap-schedule notes regarding 'start' time
* Backport #57388: quincy: [test] iscsi rest_api_create.t and rest_api_delete.t need formatting adjustment
* Backport #57458: quincy: bluefs fsync doesn't respect file truncate
* Backport #57592: quincy: STORE==USED in ceph df 
* Backport #57604: quincy: Log message is little confusing
* Backport #57886: quincy: mgr/prometheus: avoid duplicates and deleted entries for rbd_stats_pool
* Backport #58180: quincy: KernelDevice::read doesn't translate error codes correctly
* Backport #58259: quincy: rados: fix extra tabs on warning for pool copy
* Backport #58479: quincy: RGW service crashes regularly with floating point exception
* Backport #58967: quincy: do not use wholespace iterator for default cf lookups 
* Backport #58998: quincy: store names of modules that register RADOS clients in the MgrMap
* Backport #59062: quincy: [rbd-mirror] Enabling mirroring on image in a namespace falsely fails saying cannot enable mirroring in current pool mirroring mode
* Backport #59097: quincy: CLOCK_REALTIME_COARSE overridden on Windows
* Backport #59140: quincy: RocksDBTransactionImpl::rm_range_keys doesn't use bound iterator
* Backport #59213: quincy: wrong attribute for rbd_quiesce_complete api
* Backport #59336: quincy: cannot enable level_compaction_dynamic_level_bytes
* Backport #59369: quincy: unlink_peer gets stuck during blocklist
* Backport #59416: quincy: pybind/mgr/volumes: investigate moving calls which may block on libcephfs into another thread
* Backport #59439: quincy: [test] qemu-utils is not upgraded when qemu-block-extra is installed
* Backport #59456: quincy: Monitors do not permit OSD to join after upgrading to Quincy
* Backport #59479: quincy: win32_deps_build.sh attempts to install packages using pip
* Backport #59507: quincy: [rbd-mirror] don't leave an incomplete primary snapshot if the peer who is handling snapshot creation dies
* Backport #59537: quincy: osd/scrub: verify SnapMapper consistency not backported
* Backport #59565: quincy: mgr/telemetry: log exceptions as "exception" instead of "error"
* Backport #59709: quincy: rbd-mirror cannot grab lock in time with latency
* Backport #59711: quincy: [rbd_support] recover from RADOS instance blocklisting
* Backport #59726: quincy: mds: allow entries to be removed from lost+found directory
* Backport #61194: quincy: [rbd_support] recover from "double blocklisting" (being blocklisted while recovering from blocklisting)
* Backport #61196: quincy: managed_lock/GetLockerRequest: fix empty address check
* Backport #61231: quincy: Tests failing with slow scrubs with new mClock default profile
* Backport #61308: quincy: BlueFS does not prealloc space
* Backport #61341: quincy: mgr/prometheus: fix pool_objects_repaired and daemon_health_metrics format
* Backport #61423: quincy: [ceph-dencoder] fix incorrect type define for trash_watcher
* Backport #61462: quincy: rbd-wnbd: improve image map error message
* Backport #61487: quincy: ceph: osd blocklist does not accept v2/v1: prefix for addr
* Backport #61527: quincy: SimpleSchedulerObjectDispatch decrements iterator before first element
* Backport #61530: quincy: "profile rbd-read-only" OSD cap should allow calling rbd.metadata_list method on the pool
* Bug #61559: RGW crash upon parsing a TPCDS query (Trino push-down)
* Bug #61561: Trino does not stop the statement processing upon a failure of the s3select-engine 
* Bug #61636: missing SQL syntax construct in s3select
* Bug #61637: s3select: a backtick(`) on a constant date is converted into timestamp - on parsing time
* Bug #61638: RGW crashed on getObject::range-request for some of the TPCDS queries
* Backport #61705: quincy: [rbd_support] display useful error message when rbd_support module is not ready
* Backport #61724: quincy: BlueStore::_collection_list latency perf counter error
* Backport #61752: quincy: [librbd] volume data corruption when using rbd-mirror w/failover
* Backport #61757: quincy: avoid ballooning client_mount_timeout by 10x
* Backport #61766: quincy: RGW D3n cache don't work for object with slash
* Bug #61769: Bulk upload feature not working
* Backport #61870: quincy: rgw: add support http_date if http_x_amz_date is missing for sigv4
* Bug #61876: Error bootstraping first cluster node 
* Bug #61882: rgw: nothing provides libthrift-0.14.0.so
* Backport #61894: quincy: [rbd-mirror] snapshot replayer is shut down with "local image linked to unknown peer" error on force promote
* Backport #61898: quincy: pybind/cephfs: holds GIL during rmdir
* Bug #61955: S3 metadata with dot . in the key AccessDenied
* Backport #62030: quincy: ceph config set using osd/host mask not working
* Backport #62060: quincy: ceph-volume lvm new-db requires 'bluestore-block-db-size' parameter
* Backport #62092: quincy: active mgr crashes with segfault when running 'ceph osd purge'
* Backport #62102: quincy: Quincy and pacific fail to compile after Boost upgrade of main
* Backport #62112: quincy: rbd-mirror: non-primary images not deleted when the primary images are deleted
* Backport #62118: quincy: qa: use parallel gzip for compressing logs
* Backport #62302: quincy: retry metadata cache notifications with INVALIDATE_OBJ
* Backport #62307: quincy: rgw/syncpolicy: sync status doesn't reflect syncpolicy set
* Backport #62314: quincy: stale info for s3 Unsupported Header Fields
* Backport #62321: quincy: File Corruption in Multisite Replication with Encryption
* Bug #62402: TestClsRbd.mirror_snapshot failure in octopus-x-quincy
* Backport #62420: quincy: mds: adjust cap acquistion throttle defaults
* Bug #62528: Runtime Error: cephadmin version
* Backport #62554: quincy: libcephsqlite: short reads fill 0s at beginning of buffer
* Backport #62571: quincy: ceph_fs.h: add separate owner_{u,g}id fields
* Backport #62607: quincy: mgr: DaemonServer::ms_handle_authentication acquires daemon locks
* Backport #62685: quincy: hang due to exclusive lock acquisition (STATE_WAITING_FOR_LOCK) racing with blocklisting
* Backport #62690: quincy: [rbd-mirror] demote snapshot does not get removed
* Backport #62753: quincy: Object with null version when using versioning and transition
* Bug #62773: TestClsRbd.mirror_snapshot failure in quincy p2p
* Backport #62941: quincy: [test] bogus POOL_APP_NOT_ENABLED health alert throught the suite
* Backport #62943: quincy: New radosgw-admin commands to cleanup leftover OLH index entries and unlinked instance objects
* Backport #63041: quincy: CVE-2023-43040 - Improperly verified POST keys.
* Backport #63050: quincy: RGW s3website API prefetches data for range requests
* Backport #63053: quincy: SignatureDoesNotMatch when extra headers start with 'x-amzn'
* Backport #63056: quincy: high virtual memory consumption when dealing with Chunked Upload
* Backport #63059: quincy: crash: RGWSI_Notify::unwatch(RGWSI_RADOS::Obj&, unsigned long)
* Backport #63063: quincy: [test] reproducer for a deadlock which can occur when a watch error is hit while krbd is recovering from a previous watch error
* Backport #63126: quincy: osd: Is it necessary to unconditionally increase osd_bandwidth_cost_per_io in mClockScheduler::calc_scaled_cost?
* Backport #63155: quincy: another hang due to exclusive lock acquisition (STATE_WAITING_FOR_LOCK) racing with blocklisting
* Bug #63239: Corrupted rocksdb WAL log after power cut
* Bug #63357: quincy: mgr/dashboard: disable dashboard v3 in quincy
* Bug #63391: OSDs fail to be created on PVs or LVs in v17.2.7 due to failure in ceph-volume raw list