v14.2.11 Nautilus released

TheAnalyst

This is the eleventh release in the Nautilus series. This release brings a number of bugfixes across all major components of Ceph. We recommend that all Nautilus users upgrade to this release.

Notable Changes

  • RGW: The radosgw-admin sub-commands dealing with orphans – radosgw-admin orphans find, radosgw-admin orphans finish, radosgw-admin orphans list-jobs – have been deprecated. They have not been actively maintained and they store intermediate results on the cluster, which could fill a nearly-full cluster. They have been replaced by a tool, currently considered experimental, rgw-orphan-list.

  • Now when noscrub and/or nodeep-scrub flags are set globally or per pool, scheduled scrubs of the type disabled will be aborted. All user initiated scrubs are NOT interrupted.

  • Fixed a ceph-osd crash in committed osd maps when there is a failure to encode the first incremental map. issue#46443

Changelog

  • bluestore: core: os/bluestore: fix large (>2GB) writes when bluefs_buffered_io = true (pr#35404, Igor Fedotov)

  • bluestore: os/bluestore: implement Hybrid allocator (pr#35500, Adam Kupczyk, Kefu Chai, Igor Fedotov, xie xingguo)

  • build/ops: build/ops: selinux: allow ceph_t amqp_port_t:tcp_socket (pr#36190, Kaleb S. KEITHLEY, Thomas Serlin)

  • ceph-volume: add dmcrypt support in raw mode (pr#35831, Guillaume Abrioux)

  • cephfs,pybind: pybind/cephfs: fix custom exception raised by cephfs.pyx (pr#36180, Ramana Raja)

  • cephfs: ceph_fuse: add the ‘-d’ option back for libfuse (pr#35398, Xiubo Li)

  • cephfs: client: fix directory inode can not call release callback (pr#36177, sepia-liu)

  • cephfs: client: fix setxattr for 0 size value (NULL value) (pr#36173, Sidharth Anupkrishnan)

  • cephfs: client: fix snap directory atime (pr#36169, Luis Henriques)

  • cephfs: client: introduce timeout for client shutdown (issue#44276, pr#36215, Venky Shankar)

  • cephfs: client: release the client_lock before copying data in read (pr#36294, Chencan)

  • cephfs: client: static dirent for readdir is not thread-safe (pr#36511, Patrick Donnelly)

  • cephfs: mds: add config to require forward to auth MDS (pr#35377, simon gao)

  • cephfs: mds: cleanup uncommitted fragments before mds goes to active (pr#35397, “Yan, Zheng”)

  • cephfs: mds: do not raise “client failing to respond to cap release” when client working set is reasonable (pr#36513, Patrick Donnelly)

  • cephfs: mds: do not submit omap_rm_keys if the dir is the basedir of merge (pr#36178, Chencan)

  • cephfs: mds: fix filelock state when Fc is issued (pr#35841, Xiubo Li)

  • cephfs: mds: fix hang issue when accessing a file under a lost parent directory (pr#36179, Zhi Zhang)

  • cephfs: mds: fix nullptr dereference in MDCache::finish_rollback (pr#36439, “Yan, Zheng”)

  • cephfs: mds: flag backtrace scrub failures for new files as okay (pr#35400, Milind Changire)

  • cephfs: mds: initialize MDSlaveUpdate::waiter (pr#36462, “Yan, Zheng”)

  • cephfs: mds: make threshold for MDS_TRIM configurable (pr#36175, Paul Emmerich)

  • cephfs: mds: preserve ESlaveUpdate logevent until receiving OP_FINISH (pr#35394, Varsha Rao, songxinying)

  • cephfs: mds: reset heartbeat in EMetaBlob replay (pr#36170, Yanhu Cao)

  • cephfs: mgr/fs/volumes misc fixes (pr#36167, Patrick Donnelly, Kotresh HR, Ramana Raja)

  • cephfs: mgr/volumes: Add snapshot info command (pr#35672, Kotresh HR)

  • cephfs: mgr/volumes: Deprecate protect/unprotect CLI calls for subvolume snapshots (pr#36166, Shyamsundar Ranganathan)

  • cephfs: qa: add debugging for volumes plugin use of libcephfs (pr#36512, Patrick Donnelly)

  • cephfs: qa: skip cache_size check (pr#36526, Patrick Donnelly)

  • cephfs: tools/cephfs: don’t bind to public_addr (pr#35401, “Yan, Zheng”)

  • cephfs: vstart_runner: set mounted to True at the end of mount() (pr#35396, Rishabh Dave)

  • core,mon: mon/OSDMonitor: Reset grace period if failure interval exceeds a threshold (pr#35798, Sridhar Seshasayee)

  • core: mgr/DaemonServer.cc: make ‘config show’ on fsid work (pr#36074, Neha Ojha)

  • core: mgr/alert: can’t set inventory_cache_timeout/service_cache_timeout from CLI (pr#36104, Kiefer Chang)

  • core: osd/PG: fix history.same_interval_since of merge target again (pr#36161, xie xingguo)

  • core: osd/PeeringState.h: Fix pg stuck in WaitActingChange (pr#35389, chen qiuzhang)

  • core: osd: Cancel in-progress scrubs (not user requested) (pr#36292, David Zafman)

  • core: osd: fix crash in _committed_osd_maps if incremental osdmap crc fails (pr#36339, Neha Ojha, Dan van der Ster)

  • core: osd: make “missing incremental map” a debug log message (pr#35386, Nathan Cutler)

  • core: osd: make message cap option usable again (pr#35738, Neha Ojha, Josh Durgin)

  • mgr/dashboard: Allow to edit iSCSI target with active session (pr#35998, Ricardo Marques)

  • mgr/dashboard: Prevent dashboard breakdown on bad pool selection (pr#35367, Stephan Müller)

  • mgr/dashboard: Prometheus query error in the metrics of Pools, OSDs and RBD images (pr#35884, Avan Thakkar)

  • mgr/dashboard: add popover list of Stand-by Managers & Metadata Servers (MDS) in landing page (pr#34095, Kiefer Chang, Avan Thakkar)

  • mgr/dashboard: fix Source column i18n issue in RBD configuration tables (pr#35822, Kiefer Chang)

  • mgr/k8sevents: sanitise kubernetes events (pr#35563, Paul Cuzner)

  • mgr/prometheus: improve Prometheus module cache (pr#35918, Patrick Seidensal)

  • mgr: mgr/progress: Skip pg_summary update if _events dict is empty (pr#36075, Manuel Lausch)

  • mgr: mgr/telemetry: force –license when sending while opted-out (pr#35390, Yaarit Hatuka)

  • mgr: mon/PGMap: do not consider changing pg stuck (pr#35959, Kefu Chai)

  • monitoring: fixing some issues in RBD detail dashboard (pr#35464, Kiefer Chang)

  • msgr: New msgr2 crc and secure modes (msgr2.1) (pr#35733, Jianpeng Ma, Ilya Dryomov)

  • rbd: librbd: new ‘write_zeroes’ API methods to suppliment the discard APIs (pr#36250, Jason Dillaman)

  • rbd: mgr/dashboard: work with v1 RBD images (pr#35712, Ernesto Puerta)

  • rbd: rbd: librbd: Watcher should not attempt to re-watch after detecting blacklisting (pr#35385, Jason Dillaman)

  • rgw,tests: test/rgw: update hadoop versions (pr#35778, Casey Bodley, Vasu Kulkarni)

  • rgw: Add subuser to OPA request (pr#36187, Seena Fallah)

  • rgw: Add support wildcard subuser for bucket policy (pr#36186, Seena Fallah)

  • rgw: add “rgw-orphan-list” tool and “radosgw-admin bucket radoslist …” (pr#34127, J. Eric Ivancich)

  • rgw: add check for index entry’s existing when adding bucket stats during bucket reshard (pr#36189, zhang Shaowen)

  • rgw: add quota enforcement to CopyObj (pr#36184, Casey Bodley)

  • rgw: bucket list/stats truncates for user w/ >1000 buckets (pr#36165, J. Eric Ivancich)

  • rgw: cls_bucket_list_(un)ordered should clear results collection (pr#36163, J. Eric Ivancich)

  • rgw: fix loop problem with swift stat on account (pr#36185, Marcus Watts)

  • rgw: lc: fix Segmentation Fault when the tag of the object was not found (pr#36086, yupeng chen, zhuo li)

  • rgw: ordered listing lcv not managed correctly (pr#35882, J. Eric Ivancich)

  • rgw: radoslist incomplete multipart uploads fix marker progression (pr#36191, J. Eric Ivancich)

  • rgw: rgw/iam: correcting the result of get role policy (pr#36193, Pritha Srivastava)

  • rgw: rgw/url: fix amqp urls with vhosts (pr#35384, Yuval Lifshitz)

  • rgw: stop realm reloader before store shutdown (pr#36192, Casey Bodley)

  • tools: Add statfs operation to ceph-objecstore-tool (pr#35713, David Zafman)