Ceph - v18.2.3 open 77% 70 issues (54 closed — 16 open) Related issues Bug #64441: reef: qa: add upgrade testing from minor release of reef (v18.2.[01]) to reef HEAD Actions
Ceph - v19.1.0 open 84% 65 issues (55 closed — 10 open) Related issues Bug #55725: MDS allows a (kernel) client to exceed the xattrs key/value limits Actions Bug #56067: Cephfs data loss with root_squash enabled Actions Bug #64490: mds: some request errors come from errno.h rather than fs_types.h Actions Bug #64503: client: log message when unmount call is received Actions Bug #64748: reef: snaptest-git-ceph.sh failure Actions Tasks #63669: qa: add teuthology tests for quiescing a group of subvolumes Actions
Ceph - v20.0.0 open T release 13% 151 issues (17 closed — 134 open) Related issues Bug #40159: mds: openfiletable prefetching large amounts of inodes lead to mds start failure Actions Bug #40197: The command 'node ls' sometimes output some incorrect information about mds. Actions Bug #48562: qa: scrub - object missing on disk; some files may be lost Actions Bug #50821: qa: untar_snap_rm failure during mds thrashing Actions Bug #51197: qa: [WRN] Scrub error on inode 0x10000001520 (/client.0/tmp/t/linux-5.4/Documentation/driver-api) see mds.f log and `damage ls` output for details Actions Bug #51282: pybind/mgr/mgr_util: .mgr pool may be created too early causing spurious PG_DEGRADED warnings Actions Bug #57676: qa: error during scrub thrashing: rank damage found: {'backtrace'} Actions Bug #62123: mds: detect out-of-order locking Actions Bug #62188: AttributeError: 'RemoteProcess' object has no attribute 'read' Actions Bug #63866: mount command returning misleading error message Actions Bug #63931: qa: test_mirroring_init_failure_with_recovery failure Actions Bug #64008: mds: CInode::item_caps used in two different lists Actions Bug #64064: mds config `mds_log_max_segments` throws error for value -1 Actions Bug #64198: mds: Fcb caps issued to clients when filelock is xlocked Actions Bug #64477: pacific: rados/cephadm/mgr-nfs-upgrade: [WRN] client session with duplicated session uuid 'ganesha-nfs.foo.XXX' denied Actions Bug #64486: qa: enhance labeled perf counters test for cephfs-mirror Actions Bug #64502: pacific/quincy/v18.2.0: client: ceph-fuse fails to unmount after upgrade to main Actions Bug #64537: mds: lower the log level when rejecting a session reclaim request Actions Bug #64542: Difference in error code returned while removing system xattrs using removexattr() Actions Bug #64563: mds: enhance laggy clients detections due to laggy OSDs Actions Bug #64572: workunits/fsx.sh failure Actions Bug #64602: tools/cephfs: cephfs-journal-tool does not recover dentries with alternate_name Actions Bug #64616: selinux denials with centos9.stream Actions Bug #64641: qa: Add multifs root_squash testcase Actions Bug #64685: mds: disable defer_client_eviction_on_laggy_osds by default Actions Bug #64700: Test failure: test_mount_all_caps_absent (tasks.cephfs.test_multifs_auth.TestClientsWithoutAuth) Actions Bug #64707: suites/fsstress.sh hangs on one client - test times out Actions Bug #64711: Test failure: test_cephfs_mirror_cancel_mirroring_and_readd (tasks.cephfs.test_mirroring.TestMirroring) Actions Bug #64717: MDS stuck in replay/resolve use Actions Bug #64729: mon.a (mon.0) 1281 : cluster 3 [WRN] MDS_SLOW_METADATA_IO: 3 MDSs report slow metadata IOs" in cluster log Actions Bug #64730: fs/misc/multiple_rsync.sh workunit times out Actions Bug #64746: qa/cephfs: add MON_DOWN and `deprecated feature inline_data' to health ignorelist. Actions Bug #64747: postgresql pkg install failure Actions Bug #64751: cephfs-mirror coredumped when acquiring pthread mutex Actions Bug #64752: cephfs-mirror: valgrind report leaks Actions Bug #64761: cephfs-mirror: add throttling to mirror daemon ops Actions Bug #64912: make check: QuiesceDbTest.MultiRankRecovery Failed Actions Bug #64947: qa: fix continued use of log-whitelist Actions Bug #64985: qa: mgr logs do not include client debugging Actions Bug #64986: qa: "cluster [WRN] Health detail: HEALTH_WARN 1 filesystem is online with fewer MDS than max_mds" in cluster log " Actions Bug #64987: qa: "cluster [WRN] Health check failed: 1 pool(s) do not have an application enabled (POOL_APP_NOT_ENABLED)" in cluster log " Actions Bug #64988: qa: fs:workloads mgr client evicted indicated by "cluster [WRN] evicting unresponsive client smithi042:x (15288), after 303.306 seconds" Actions Bug #65001: mds: ceph-mds might silently ignore client_session(request_close, ...) message Actions Bug #65018: PG_DEGRADED warnings during cluster creation via cephadm: "Health check failed: Degraded data redundancy: 2/192 objects degraded (1.042%), 1 pg degraded (PG_DEGRADED)" Actions Bug #65019: qa/suites/fs/top: [WRN] Health check failed: 1 pool(s) do not have an application enabled (POOL_APP_NOT_ENABLED)" in cluster log Actions Bug #65020: qa: Scrub error on inode 0x1000000356c (/volumes/qa/sv_0/2f8f6bb4-3ea9-47a0-bd79-a0f50dc149d5/client.0/tmp/clients/client7/~dmtmp/PARADOX) see mds.b log and `damage ls` output for details" in cluster log Actions Bug #65021: qa/suites/fs/nfs: cluster [WRN] Health check failed: 1 stray daemon(s) not managed by cephadm (CEPHADM_STRAY_DAEMON)" in cluster log Actions Bug #65022: qa: test_max_items_per_obj open procs not fully cleaned up Actions Bug #65039: mds: standby-replay segmentation fault in md_log_replay Actions Bug #65043: Unable to set timestamp to value > UINT32_MAX Actions Bug #65073: pybind/mgr/stats/fs: log exceptions to cluster log Actions Bug #65094: mds STATE_STARTING won't add root ino for root rank and not correctly handle when fails at STATE_STARTING Actions Bug #65116: squid: kclient: "ld: final link failed: Resource temporarily unavailable" Actions Bug #65136: QA failure: test_fscrypt_dummy_encryption_with_quick_group Actions Bug #65157: cephfs-mirror: set layout.pool_name xattr of destination subvol correctly Actions Bug #65171: Provide metrics support for the Replication Start/End Notifications Actions Bug #65182: mds: quiesce_inode op waiting on remote auth pins is not killed correctly during quiesce timeout/expiration Actions Bug #65224: mds: fs subvolume rm fails Actions Bug #65225: ceph_assert on dn->get_projected_linkage()->is_remote Actions Bug #65246: qa/cephfs: test_multifs_single_path_rootsquash (tasks.cephfs.test_admin.TestFsAuthorize) Actions Bug #65260: mds: Reduce log level for messages when mds is stopping Actions Bug #65262: qa/cephfs: kernel_untar_build.sh failed due to build error Actions Bug #65265: qa: health warning "no active mgr (MGR_DOWN)" occurs before and after test_nfs runs Actions Bug #65271: qa: cluster [WRN] Health detail: HEALTH_WARN 1 pool(s) do not have an application enabled" in cluster log Actions Bug #65276: MDS daemon is using 50% CPU when idle Actions Bug #65301: fs:upgrade still uses centos_8* distro Actions Bug #65308: qa: fs was offline but also unexpectedly degraded Actions Bug #65309: qa: dbench.sh failed with "ERROR: handle 10318 was not found" Actions Bug #65314: valgrind error: Leak_PossiblyLost posix_memalign UnknownInlinedFun ceph::buffer::v15_2_0::list::refill_append_space(unsigned int) Actions Bug #65342: mds: quiesce_counter decay rate initialized from wrong config Actions Bug #65350: mgr/snap_schedule: restore yearly spec from uppercase Y to lowercase y Actions Bug #65372: qa: The following counters failed to be set on mds daemons: {'mds.exported', 'mds.imported'} Actions Bug #65388: The MDS_SLOW_REQUEST warning is flapping even though the slow requests don't go away Actions Bug #65389: The ceph_readdir function in libcephfs returns incorrect d_reclen value Actions Bug #65472: mds: avoid recalling Fb when quiescing file Actions Bug #65496: mds: ceph.dir.subvolume and ceph.quiesce.blocked is not properly replicated Actions Bug #65508: qa: lockup not long enough to for test_quiesce_authpin_wait Actions Bug #65518: mds: regular file inode flags are not replicated by the policylock Actions Bug #65545: Quiesce may fail randomly with EBADF due to the same root submitted to the MDCache multiple times under the same quiesce request Actions Bug #65564: Test failure: test_snap_schedule_subvol_and_group_arguments_08 (tasks.cephfs.test_snap_schedules.TestSnapSchedulesSubvolAndGroupArguments) Actions Bug #65572: Command failed (workunit test fs/snaps/untar_snap_rm.sh) on smithi155 with status 1 Actions Bug #65580: mds/client: add dummy client feature to test client eviction Actions Bug #65595: mds: missing policylock acquisition for quiesce Actions Bug #65603: mds: quiesce timeout due to a freezing directory Actions Bug #65606: workload fails due to slow ops, assert in logs mds/Locker.cc: 551 FAILED ceph_assert(!lock->is_waiter_for(SimpleLock::WAIT_WR) || lock->is_waiter_for(SimpleLock::WAIT_XLOCK)) Actions Bug #65614: client: resends request to same MDS it just received a forward from if it does not have an open session with the target Actions Bug #65616: pybind/mgr/snap_schedule: 1m scheduled snaps not reliably executed (RuntimeError: The following counters failed to be set on mds daemons: {'mds_server.req_rmsnap_latency.avgcount'}) Actions Bug #65618: qa: fsstress: cannot execute binary file: Exec format error Actions Bug #65647: Evicted kernel client may get stuck after reconnect Actions Bug #65658: mds: MetricAggregator::ms_can_fast_dispatch2 acquires locks Actions Bug #65660: mds: drop client metrics during recovery Actions Bug #65669: QuiesceDB responds with a misleading error to a quiesce-await of a terminated set. Actions Bug #65700: qa: Health detail: HEALTH_WARN Degraded data redundancy: 40/348 objects degraded (11.494%), 9 pgs degraded" in cluster log Actions Bug #65701: qa: quiesce cache/ops dump not world readable Actions Bug #65704: mds+valgrind: beacon thread blocked for 60+ seconds Actions Bug #65705: qa: snaptest-multiple-capsnaps.sh failure Actions Bug #65716: mds: quiesce_path blocks on acquiring auth_pins for dentries to root inode to be quiesced Actions Bug #65733: mds: upgrade to MDS enforcing CEPHFS_FEATURE_MDS_AUTH_CAPS_CHECK with client having root_squash in any MDS cap causes eviction for all file systems the client has caps for Actions Fix #63432: qa: run TestSnapshots.test_kill_mdstable for all mount types Actions Fix #64984: qa: probabilistically ignore PG_AVAILABILITY/PG_DEGRADED Actions Fix #65408: qa: under valgrind, restart valgrind/mds when MDS exits with 0 Actions Fix #65579: mds: use _exit for QA killpoints rather than SIGABRT Actions Fix #65617: qa: increase debugging for snap_schedule Actions Feature #57481: mds: enhance scrub to fragment/merge dirfrags Actions Feature #61334: cephfs-mirror: use snapdiff api for efficient tree traversal Actions Feature #63374: mds: add asok command to kill/respond to request Actions Feature #63663: mds,client: add crash-consistent snapshot support Actions Feature #63664: mds: add quiesce protocol for halting I/O on subvolumes Actions Feature #63665: mds: QuiesceDb to manage subvolume quiesce state Actions Feature #63666: mds: QuiesceAgent to execute quiesce operations on an MDS rank Actions Feature #63668: pybind/mgr/volumes: add quiesce protocol API Actions Feature #64506: qa: update fs:upgrade to test from reef/squid to main Actions Feature #64507: pybind/mgr/snap_schedule: support crash-consistent snapshots Actions Feature #64531: mds,mgr: identify metadata heavy workloads Actions Feature #65503: mgr/stats, cephfs-top: provide per volume/sub-volume based performance metrics to monitor / troubleshoot performance issues Actions Feature #65637: mds: continue sending heartbeats during recovery when MDS journal is large Actions Cleanup #65689: mds: move specialized cleanup for fragment_dir to MDCache::request_cleanup Actions Cleanup #65690: mds: move specialized cleanup for export_dir to MDCache::request_cleanup Actions Tasks #63707: mds: AdminSocket command to control the QuiesceDbManager Actions Tasks #63708: mds: MDS message transport for inter-rank QuiesceDbManager communications Actions