Activity
From 04/01/2021 to 04/30/2021
04/30/2021
- 12:35 PM Backport #50391 (Rejected): pacific: MDS doesn't set fscrypt flag on new inodes with crypto conte...
- Yeah. This is some basic infrastructure that we merged for fscrypt support, but it's going to require more changes be...
- 05:28 AM Feature #50581 (Fix Under Review): cephfs-mirror: allow mirror daemon to connect to local/primary...
04/29/2021
- 06:15 PM Documentation #50008 (In Progress): mgr/nfs: Add troubleshooting section
- 04:50 PM Backport #50255: nautilus: mds: standby-replay only trims cache when it reaches the end of the re...
- Backport Bot wrote:
> https://github.com/ceph/ceph/pull/40744
merged - 01:43 PM Feature #50581 (Resolved): cephfs-mirror: allow mirror daemon to connect to local/primary cluster...
- This enables rook to easily deploy and manage cephfs-mirror daemons and mimics how other daemon in ceph work.
- 01:28 PM Bug #50523 (Fix Under Review): Mirroring path "remove" don't not seem to work
- 03:20 AM Feature #47264 (In Progress): "fs authorize" subcommand should work for multiple FSs too
04/28/2021
- 09:36 PM Bug #46883: kclient: ghost kernel mount
- Xiubo Li wrote:
> Ramana Raja wrote:
> > I saw the following failure multiple times in Yuri's nautilus runs,
> > h... - 01:47 AM Bug #46883: kclient: ghost kernel mount
- Ramana Raja wrote:
> I saw the following failure multiple times in Yuri's nautilus runs,
> https://pulpito.ceph.com... - 01:29 AM Bug #46883: kclient: ghost kernel mount
- I saw the following failure multiple times in Yuri's nautilus runs,
https://pulpito.ceph.com/yuriw-2021-04-21_16:19:... - 09:25 PM Backport #50179: nautilus: client: only check pool permissions for regular files
- Backport Bot wrote:
> https://github.com/ceph/ceph/pull/40730
merged - 09:17 PM Bug #50570 (Won't Fix - EOL): nautilus: qa: tasks/trim-i22073 cluster [WRN] Health check failed: ...
- https://pulpito.ceph.com/yuriw-2021-04-20_21:38:51-fs-wip-yuri8-testing-2021-04-20-0734-nautilus-distro-basic-smithi/...
- 08:53 PM Bug #50569: nautilus: qa: tasks/cfuse_workunit_suites_fsstress validater/valgrind failures
- Maybe this is same as https://tracker.ceph.com/issues/36685 [paramiko timeout not working for hung process] ?
- 08:48 PM Bug #50569 (Won't Fix): nautilus: qa: tasks/cfuse_workunit_suites_fsstress validater/valgrind fai...
- Description: fs/verify/{begin centos_latest clusters/fixed-2-ucephfs conf/{client mds mon osd} mount/fuse objectstore...
- 07:09 PM Backport #50026: nautilus: client: items pinned in cache preventing unmount
- Backport Bot wrote:
> https://github.com/ceph/ceph/pull/40722
merged - 07:08 PM Backport #49853: nautilus: mds: race of fetching large dirfrag
- Backport Bot wrote:
> https://github.com/ceph/ceph/pull/40720
merged - 05:03 PM Feature #50235: allow cephfs-shell to mount named filesystems
- I would like to work on this ticket. If that's ok, can some one please assign this ticket to me?
- 03:09 PM Bug #50561 (Fix Under Review): cephfs-mirror: incrementally transfer snapshots whenever possible
- 03:08 PM Bug #50561 (Resolved): cephfs-mirror: incrementally transfer snapshots whenever possible
- Currently, each snapshot is synchronized by purging data on remote filesystem under the directory followed by bulk co...
- 12:31 PM Bug #50559 (Fix Under Review): session dump includes completed_requests twice, once as an integer...
- 12:20 PM Bug #50559 (Resolved): session dump includes completed_requests twice, once as an integer and onc...
- ...
- 06:34 AM Bug #15783 (New): client: enable acls by default
- 06:32 AM Bug #49644 (New): vstart_runner: run_ceph_w() doesn't work with shell=True
- 05:33 AM Bug #50010 (Fix Under Review): qa/cephfs: get_key_from_keyfile() return None when key is not foun...
- 03:47 AM Backport #50445 (In Progress): pacific: client: Inode.cc: 405: FAILED ceph_assert(_ref >= 0)
- 03:47 AM Backport #50392 (In Progress): pacific: cephfs-top: exception: addwstr() returned ERR
- 01:14 AM Bug #50389 (In Progress): mds: "cluster [ERR] Error recovering journal 0x203: (2) No such file ...
04/27/2021
- 11:25 PM Bug #50546 (Won't Fix): nautilus: qa: 'The following counters failed to be set on mds daemons: {'...
- tasks/cephfs_test_exports failed in multi mds suite in nautilus testing as follows,
https://pulpito.ceph.com/yuriw-2... - 02:45 PM Backport #50541 (Resolved): pacific: libcephfs: support file descriptor based *at() APIs
- https://github.com/ceph/ceph/pull/41475
- 02:41 PM Bug #50298 (Pending Backport): libcephfs: support file descriptor based *at() APIs
- 01:25 PM Backport #50539 (Rejected): octopus: mgr/pybind/snap_schedule: do not fail when no fs snapshots a...
- 01:25 PM Backport #50538 (Resolved): pacific: mgr/pybind/snap_schedule: do not fail when no fs snapshots a...
- https://github.com/ceph/ceph/pull/41044
- 01:25 PM Backport #50537 (Resolved): pacific: "ceph fs snapshot mirror daemon status" should not use json ...
- https://github.com/ceph/ceph/pull/41475
- 01:22 PM Bug #50266 (Pending Backport): "ceph fs snapshot mirror daemon status" should not use json keys a...
- 01:22 PM Bug #49837 (Pending Backport): mgr/pybind/snap_schedule: do not fail when no fs snapshots are ava...
- 04:46 AM Bug #50523: Mirroring path "remove" don't not seem to work
- I can reproduce this when mirror daemon is not running (guessing that your case too):...
- 03:54 AM Bug #50523 (In Progress): Mirroring path "remove" don't not seem to work
- 02:18 AM Backport #50253 (Resolved): pacific: mds: "cluster [WRN] Scrub error on inode 0x1000000039d (/cli...
- 02:08 AM Bug #50532 (Resolved): mgr/volumes: hang when removing subvolume when pools are full
- 01:58 AM Backport #50225 (Resolved): pacific: mds: failed to decode message of type 29 v1: void CapInfoPay...
- 12:15 AM Feature #50531 (New): cephfs-top : show `delayed_ranks` in cephfs-top output
- cephfs-top : show `delayed_ranks` -set of active MDS ranks that are reporting stale metrics in cephfs-top output
04/26/2021
- 11:55 PM Backport #49562: nautilus: qa: file system deletion not complete because starter fs already destr...
- Backport Bot wrote:
> https://github.com/ceph/ceph/pull/40709
merged - 11:54 PM Backport #49516: nautilus: pybind/cephfs: DT_REG and DT_LNK values are wrong
- Backport Bot wrote:
> https://github.com/ceph/ceph/pull/40704
merged - 11:54 PM Backport #49473: nautilus: nautilus: qa: "Assertion `cb_done' failed."
- Backport Bot wrote:
> https://github.com/ceph/ceph/pull/40701
merged - 10:32 PM Backport #49613: nautilus: qa: racy session evicted check
- Backport Bot wrote:
> https://github.com/ceph/ceph/pull/40714
merged - 05:18 PM Backport #50285: pacific: qa: test standby_replay in workloads
- Backport Bot wrote:
> https://github.com/ceph/ceph/pull/40853
merged - 05:18 PM Backport #50287: pacific: qa: "log [ERR] : error reading sessionmap 'mds2_sessionmap'"
- Backport Bot wrote:
> https://github.com/ceph/ceph/pull/40852
merged - 05:17 PM Backport #50253: pacific: mds: "cluster [WRN] Scrub error on inode 0x1000000039d (/client.0/tmp/b...
- Backport Bot wrote:
> https://github.com/ceph/ceph/pull/40825
merged - 04:58 PM Bug #50530 (Resolved): pacific: client: abort after MDS blocklist
- ...
- 04:36 PM Bug #50528 (New): pacific: qa: fs:thrash: pjd suite not ok 20
- ...
- 04:32 PM Bug #50527 (New): pacific: qa: untar_snap_rm timeout during osd thrashing (ceph-fuse)
- ...
- 03:13 PM Bug #40284: kclient: evaluate/fix/add lazio support in the kernel
- Discussing for Quincy: Jeff is against keeping this in the kernel -- we are not really using it and it's probably inc...
- 02:40 PM Feature #16745: mon: prevent allocating snapids allocated for CephFS
- Milind, this one looks related to what you're investigating.
- 02:31 PM Tasks #38386 (Closed): qa: write kernel fscache tests
- Dup of #6373
- 01:40 PM Bug #50523 (Triaged): Mirroring path "remove" don't not seem to work
- 01:05 PM Bug #50523 (Resolved): Mirroring path "remove" don't not seem to work
- Consider the following:...
- 10:30 AM Feature #47277: implement new mount "device" syntax for kcephfs
- I did some work on this a while back but never go to posting a patch (or changing the tracker status). I'm reviving t...
- 10:28 AM Feature #47277 (In Progress): implement new mount "device" syntax for kcephfs
- 06:38 AM Bug #45434: qa: test_full_fsync (tasks.cephfs.test_full.TestClusterFull) failed
- osdstat of osd.6:...
04/24/2021
- 01:48 AM Bug #45434: qa: test_full_fsync (tasks.cephfs.test_full.TestClusterFull) failed
- Ramana Raja wrote:
> see this in octopus testing, https://pulpito.ceph.com/yuriw-2021-02-09_00:31:50-kcephfs-wip-yur...
04/23/2021
- 05:43 PM Bug #50224 (In Progress): qa: test_mirroring_init_failure_with_recovery failure
- Patrick,
PR #41000 not a fix for this tracker. I've started to look into this today.
PR #41000 is an additional... - 03:31 PM Bug #50224 (Fix Under Review): qa: test_mirroring_init_failure_with_recovery failure
- 12:30 PM Bug #50224 (In Progress): qa: test_mirroring_init_failure_with_recovery failure
- 02:49 PM Backport #50086: pacific: tasks.cephfs.test_volumes.TestSubvolumeGroups: RuntimeError: rank all f...
- Backport Bot wrote:
> https://github.com/ceph/ceph/pull/40688
merged - 02:48 PM Backport #50180: pacific: client: only check pool permissions for regular files
- Backport Bot wrote:
> https://github.com/ceph/ceph/pull/40686
merged - 02:48 PM Backport #50185: pacific: qa: "RADOS object not found (Failed to operate read op for oid cephfs_m...
- Backport Bot wrote:
> https://github.com/ceph/ceph/pull/40684
merged - 02:47 PM Backport #50190: pacific: qa: "Assertion `cb_done' failed."
- Backport Bot wrote:
> https://github.com/ceph/ceph/pull/40683
merged - 02:46 PM Backport #50225: pacific: mds: failed to decode message of type 29 v1: void CapInfoPayload::decode
- Backport Bot wrote:
> https://github.com/ceph/ceph/pull/40682
merged - 02:43 PM Backport #50127: pacific: pybind/mgr/volumes: deadlock on async job hangs finisher thread
- Backport Bot wrote:
> https://github.com/ceph/ceph/pull/40630
merged - 02:42 PM Backport #50187: pacific: ceph-dokan improvements for additional mounts
- Backport Bot wrote:
> https://github.com/ceph/ceph/pull/40627
merged - 05:27 AM Bug #50447 (Fix Under Review): cephfs-mirror: disallow adding a active peered file system back to...
- 03:26 AM Bug #50258 (Fix Under Review): pacific: qa: "run() got an unexpected keyword argument 'stdin_data'"
- 03:02 AM Bug #50495 (New): libcephfs: shutdown race fails with status 141
- Back from the dead: #43039.
/ceph/teuthology-archive/yuriw-2021-04-21_19:25:00-fs-wip-yuri3-testing-2021-04-21-093... - 02:48 AM Bug #45145 (Duplicate): qa/test_full: failed to open 'large_file_a': No space left on device
04/22/2021
- 04:25 PM Backport #50488 (Resolved): pacific: mgr/nfs: move nfs code out of volumes plugin
- https://github.com/ceph/ceph/pull/41389
- 04:21 PM Cleanup #50080 (Pending Backport): mgr/nfs: move nfs code out of volumes plugin
- 01:28 PM Backport #50445: pacific: client: Inode.cc: 405: FAILED ceph_assert(_ref >= 0)
- Patrick Donnelly wrote:
> Xiubo, please try backporting this.
Sure. - 01:28 PM Backport #50392: pacific: cephfs-top: exception: addwstr() returned ERR
- Patrick Donnelly wrote:
> Backport this one too please.
Sure.
04/21/2021
- 06:30 PM Feature #50470 (Resolved): cephfs-top: multiple file system support
- Currently only shows one file system (or combines, not sure).
- 01:15 PM Bug #50246 (Fix Under Review): mds: failure replaying journal (EMetaBlob)
- In standby_replay, if some dentries just added/linked but not get a
chance to replay the EOpen journals followed, if... - 07:13 AM Bug #50246: mds: failure replaying journal (EMetaBlob)
- ...
- 05:04 AM Bug #50246: mds: failure replaying journal (EMetaBlob)
- The [dir 0x10000000a35 /client.0/tmp/t/linux-5.4/Documentation/devicetree/bindings/hwmon/] was trimmed from the subtr...
- 12:47 AM Bug #50246: mds: failure replaying journal (EMetaBlob)
- Patrick Donnelly wrote:
> Xiubo Li wrote:
> > Before replaying the journal, the inode 0x10000000a35 was removed fro... - 12:03 PM Bug #50442 (Fix Under Review): cephfs-mirror: ignore snapshots on parent directories when synchro...
- 09:06 AM Bug #50442 (In Progress): cephfs-mirror: ignore snapshots on parent directories when synchronizin...
- 02:25 AM Bug #50442: cephfs-mirror: ignore snapshots on parent directories when synchronizing snapshots
- Patrick Donnelly wrote:
> Venky Shankar wrote:
> > Patrick Donnelly wrote:
> > > Venky Shankar wrote:
> > > > Ign... - 02:20 AM Bug #50442: cephfs-mirror: ignore snapshots on parent directories when synchronizing snapshots
- Venky Shankar wrote:
> Patrick Donnelly wrote:
> > Venky Shankar wrote:
> > > Ignore syncing snapshots starting wi... - 02:09 AM Bug #50442: cephfs-mirror: ignore snapshots on parent directories when synchronizing snapshots
- Patrick Donnelly wrote:
> Venky Shankar wrote:
> > Ignore syncing snapshots starting with an underscore since they ... - 02:05 AM Bug #50442: cephfs-mirror: ignore snapshots on parent directories when synchronizing snapshots
- Venky Shankar wrote:
> Ignore syncing snapshots starting with an underscore since they are internally used by CephFS... - 09:10 AM Bug #46985 (Resolved): common: validate type CephBool cause 'invalid command json'
- While running with --resolve-parent, the script "backport-create-issue" noticed that all backports of this issue are ...
- 09:10 AM Bug #46988 (Resolved): mds: 'forward loop' when forward_all_requests_to_auth is set
- While running with --resolve-parent, the script "backport-create-issue" noticed that all backports of this issue are ...
- 09:09 AM Documentation #48010 (Resolved): doc: document MDS recall configurations
- While running with --resolve-parent, the script "backport-create-issue" noticed that all backports of this issue are ...
- 09:08 AM Bug #49511 (Resolved): qa: "AttributeError: 'NoneType' object has no attribute 'mon_manager'"
- While running with --resolve-parent, the script "backport-create-issue" noticed that all backports of this issue are ...
- 09:07 AM Bug #49833 (Resolved): MDS should return -ENODATA when asked to remove xattr that doesn't exist
- While running with --resolve-parent, the script "backport-create-issue" noticed that all backports of this issue are ...
- 09:05 AM Backport #49903 (Resolved): nautilus: mgr/volumes: setuid and setgid file bits are not retained a...
- This update was made using the script "backport-resolve-issue".
backport PR https://github.com/ceph/ceph/pull/40270
m... - 07:09 AM Cleanup #50450 (New): mgr/nfs: Simplify the parsing of Ganesha Conf using existing pseudo-parsers
- More discussion here https://github.com/ceph/ceph/pull/40526#discussion_r607171126
- 06:58 AM Feature #50449 (Fix Under Review): mgr/nfs: Add unit tests for conf parser and others
- 06:26 AM Backport #50173: pacific: mgr/nfs: validation error on creating custom export
- This update was made using the script "backport-resolve-issue".
backport PR https://github.com/ceph/ceph/pull/40687
m... - 12:03 AM Backport #50173 (Resolved): pacific: mgr/nfs: validation error on creating custom export
- 06:21 AM Backport #50015 (Resolved): pacific: qa: "AttributeError: 'NoneType' object has no attribute 'mon...
- This update was made using the script "backport-resolve-issue".
backport PR https://github.com/ceph/ceph/pull/40645
m... - 05:49 AM Feature #50448 (New): cephfs-mirror: easy repeering
- Right now re-adding a file system as a peer to another file system requires that the cluster admin/operator cleanup s...
- 04:18 AM Bug #50447 (Resolved): cephfs-mirror: disallow adding a active peered file system back to its source
- cephfs-mirror is unidirectional atm. So, If (cluster2, fs) is an active peer for (cluster1, fs), disallow adding (clu...
- 12:42 AM Backport #50392 (Need More Info): pacific: cephfs-top: exception: addwstr() returned ERR
- Backport this one too please.
- 12:42 AM Backport #50445 (Need More Info): pacific: client: Inode.cc: 405: FAILED ceph_assert(_ref >= 0)
- Xiubo, please try backporting this.
04/20/2021
- 09:06 PM Bug #50433 (Fix Under Review): mds: Error ENOSYS: mds.a started profiler
- 05:53 AM Bug #50433 (In Progress): mds: Error ENOSYS: mds.a started profiler
- 05:53 AM Bug #50433: mds: Error ENOSYS: mds.a started profiler
- The "./bin/ceph tell mds.a heap XXX" commands succeeded, but still returning -ENOSYS errno.
- 05:51 AM Bug #50433 (Resolved): mds: Error ENOSYS: mds.a started profiler
- ...
- 09:05 PM Backport #50445 (Resolved): pacific: client: Inode.cc: 405: FAILED ceph_assert(_ref >= 0)
- https://github.com/ceph/ceph/pull/41052
- 09:02 PM Bug #49536 (Pending Backport): client: Inode.cc: 405: FAILED ceph_assert(_ref >= 0)
- We can try backporting this to pacific but I don't think it's worth the trouble farther than that.
- 09:00 PM Bug #50387: client: fs/snaps failure
- /ceph/teuthology-archive/pdonnell-2021-04-20_02:26:05-fs:thrash-master-distro-basic-smithi/6060039/teuthology.log
... - 06:12 PM Backport #49903: nautilus: mgr/volumes: setuid and setgid file bits are not retained after a subv...
- Backport Bot wrote:
> https://github.com/ceph/ceph/pull/40270
merged - 05:56 PM Bug #44384 (Can't reproduce): qa: FAIL: test_evicted_caps (tasks.cephfs.test_client_recovery.Test...
- Jeff Layton wrote:
> Lowering priority to Normal. Patrick have there been any more occurrences of this?
I have no... - 12:59 PM Bug #44384: qa: FAIL: test_evicted_caps (tasks.cephfs.test_client_recovery.TestClientRecovery)
- Lowering priority to Normal. Patrick have there been any more occurrences of this?
- 05:43 PM Bug #50442 (Resolved): cephfs-mirror: ignore snapshots on parent directories when synchronizing s...
- Ignore syncing snapshots starting with an underscore since they are internally used by CephFS.
- 04:54 PM Bug #50246: mds: failure replaying journal (EMetaBlob)
- Xiubo Li wrote:
> Before replaying the journal, the inode 0x10000000a35 was removed from the inode_map in upkeep thr... - 08:15 AM Bug #50246: mds: failure replaying journal (EMetaBlob)
Before replaying the journal, the inode 0x10000000a35 was removed from the inode_map in upkeep thread:...- 02:06 PM Bug #50266 (Fix Under Review): "ceph fs snapshot mirror daemon status" should not use json keys a...
- 07:17 AM Bug #50266: "ceph fs snapshot mirror daemon status" should not use json keys as value
- BTW, this is how the json would look for multiple active mirror daemon instances (yeh, we only support running one in...
- 07:41 AM Bug #50216: qa: "ls: cannot access 'lost+found': No such file or directory"
- Jeff Layton wrote:
> How do you reproduce this on the kclient? The MDS would have to give this inode number as the r... - 05:25 AM Bug #47276 (In Progress): MDSMonitor: add command to rename file systems
04/19/2021
- 10:44 PM Bug #50281 (Resolved): qa: untar_snap_rm timeout
- Considering this resolved -- thanks Jeff.
- 08:55 PM Bug #50281: qa: untar_snap_rm timeout
- test run: https://pulpito.ceph.com/pdonnell-2021-04-19_20:54:09-fs:snaps-master-distro-basic-smithi/
- 07:44 PM Bug #50281: qa: untar_snap_rm timeout
- Ok, I think I see the issue. We had a patch that had done a conversion from atomic_t to refcount_t for some objects, ...
- 05:25 PM Bug #50281: qa: untar_snap_rm timeout
- Same problem there:...
- 05:20 PM Bug #50281: qa: untar_snap_rm timeout
- /ceph/teuthology-archive/pdonnell-2021-04-15_01:35:57-fs-wip-pdonnell-testing-20210414.230315-distro-basic-smithi/604...
- 04:38 PM Bug #50281: qa: untar_snap_rm timeout
- Looks like the client kernel hit a GPF. The address appears to be a poisoned list value, so this may be a use-after-f...
- 01:47 PM Bug #50281 (Triaged): qa: untar_snap_rm timeout
- 04:25 PM Backport #50391 (Need More Info): pacific: MDS doesn't set fscrypt flag on new inodes with crypto...
- No longer certain we want to backport this -- Jeff is considering another approach.
- 01:53 PM Bug #50407 (Need More Info): mds_session state is stale after restart all mds
- Hi Wei, can you add more to the description please? Are you working on a PR for this? I couldn't find it on Github.
- 01:51 PM Bug #50389: mds: "cluster [ERR] Error recovering journal 0x203: (2) No such file or directory" ...
- I will take it. Thanks.
- 01:45 PM Bug #50387 (Triaged): client: fs/snaps failure
- 11:53 AM Bug #50216: qa: "ls: cannot access 'lost+found': No such file or directory"
- How do you reproduce this on the kclient? The MDS would have to give this inode number as the result of a lookup, I'd...
- 02:40 AM Bug #50216: qa: "ls: cannot access 'lost+found': No such file or directory"
- Fixed it in the kernel client too: https://patchwork.kernel.org/project/ceph-devel/list/?series=469331
Maybe Jeff ... - 02:38 AM Bug #50216 (Fix Under Review): qa: "ls: cannot access 'lost+found': No such file or directory"
- 02:13 AM Bug #50216: qa: "ls: cannot access 'lost+found': No such file or directory"
- For our privious Private-Inodes fixes, we have missed the "lost+found" dir, whose ino is 0x40.head....
- 10:15 AM Bug #50035 (Fix Under Review): cephfs-mirror: use sensible mount/shutdown timeouts
- 04:54 AM Bug #50246 (In Progress): mds: failure replaying journal (EMetaBlob)
04/17/2021
- 01:24 AM Bug #50408 (New): mds_session state is stale after restart all mds daemon
- After restart all mds daemon, client-mds session is turn into stale state and the nfs service is not recover although...
- 01:14 AM Bug #50407 (Need More Info): mds_session state is stale after restart all mds
04/16/2021
- 01:38 PM Bug #50237: cephfs-journal-tool/cephfs-data-scan: Stuck in infinite loop with "NetHandler create_...
- we ran the followings from our mon.0 (10.223.14.4):
"cephfs-data-scan pg_files / 23.4" at 2021-04-16T12:16:58.280... - 09:31 AM Bug #50266 (In Progress): "ceph fs snapshot mirror daemon status" should not use json keys as value
- 05:54 AM Backport #50186 (New): pacific: qa: daemonwatchdog fails if mounts not defined
- 04:23 AM Bug #50060 (Fix Under Review): client: access(path, X_OK) on non-executable file as root always s...
- 04:10 AM Backport #50392 (Resolved): pacific: cephfs-top: exception: addwstr() returned ERR
- https://github.com/ceph/ceph/pull/41053
- 04:10 AM Backport #50391 (Rejected): pacific: MDS doesn't set fscrypt flag on new inodes with crypto conte...
- 04:07 AM Bug #50305 (Pending Backport): MDS doesn't set fscrypt flag on new inodes with crypto context in ...
- 04:06 AM Bug #50091 (Pending Backport): cephfs-top: exception: addwstr() returned ERR
- 04:04 AM Fix #48683 (Resolved): mds/MDSMap: print each flag value in MDSMap::dump
- 04:04 AM Feature #48682 (Resolved): MDSMonitor: add command to print fs flags
- 03:57 AM Bug #50390 (Fix Under Review): mds: monclient: wait_auth_rotating timed out after 30
- 03:54 AM Bug #50390 (Resolved): mds: monclient: wait_auth_rotating timed out after 30
- Symptom:...
- 03:36 AM Bug #50389 (Resolved): mds: "cluster [ERR] Error recovering journal 0x203: (2) No such file or ...
- Symptom:...
- 03:03 AM Bug #50387 (Duplicate): client: fs/snaps failure
- ...
- 02:53 AM Bug #50220: qa: dbench workload timeout
- /ceph/teuthology-archive/pdonnell-2021-04-15_01:35:57-fs-wip-pdonnell-testing-20210414.230315-distro-basic-smithi/604...
- 02:52 AM Bug #50281: qa: untar_snap_rm timeout
- /ceph/teuthology-archive/pdonnell-2021-04-15_01:35:57-fs-wip-pdonnell-testing-20210414.230315-distro-basic-smithi/604...
04/15/2021
- 09:13 AM Bug #50216: qa: "ls: cannot access 'lost+found': No such file or directory"
- After https://github.com/ceph/ceph/pull/40868, I can reproduce it locally:...
- 08:14 AM Bug #50373 (Fix Under Review): qa: AttributeError: 'LocalCephManager' object has no attribute 'ctx'
- 04:45 AM Bug #50373 (Fix Under Review): qa: AttributeError: 'LocalCephManager' object has no attribute 'ctx'
- When running the test by using the vstart_runner.py:...
- 07:35 AM Backport #49519 (In Progress): nautilus: client: wake up the front pos waiter
- 02:09 AM Backport #49519: nautilus: client: wake up the front pos waiter
- Backport Bot wrote:
> https://github.com/ceph/ceph/pull/40707
The new one https://github.com/ceph/ceph/pull/40865 - 01:25 AM Backport #49519: nautilus: client: wake up the front pos waiter
- In nautilus we should switch the notify_one() to SignalOne() instead. More detail please see src/common/Cond.h.
- 06:04 AM Bug #50035 (In Progress): cephfs-mirror: use sensible mount/shutdown timeouts
- 04:34 AM Feature #49942: cephfs-mirror: enable running in HA
- FYI, this is HA active/active.
- 04:33 AM Feature #50372 (Resolved): test: Implement cephfs-mirror trasher test for HA active/active
04/14/2021
- 08:41 PM Backport #47084 (Rejected): nautilus: mds: 'forward loop' when forward_all_requests_to_auth is set
- can be avoided by not setting the config, which is off by default
- 08:40 PM Backport #47086 (Rejected): nautilus: common: validate type CephBool cause 'invalid command json'
- not that important
- 08:34 PM Backport #50182 (Rejected): nautilus: client: openned inodes counter is inconsistent
- Nathan Cutler wrote:
> This ticket is for tracking the nautilus backport of a follow-on fix for #46865 which was bac... - 08:34 PM Backport #50023 (Rejected): nautilus: ceph-fuse: src/include/buffer.h: 1187: FAILED ceph_assert(_...
- Nathan Cutler wrote:
> AFAICT the code being fixed does not exist in nautilus
Indeed! - 08:32 PM Backport #50251 (Rejected): nautilus: mds: "cluster [WRN] Scrub error on inode 0x1000000039d (/cl...
- This resolves warnings that typically show up with multimds scrub which isn't supported in nautilus.
- 08:32 PM Backport #50189 (Rejected): nautilus: qa: "Assertion `cb_done' failed."
- This one is not worth the effort.
- 08:31 PM Backport #48112 (Rejected): nautilus: doc: document MDS recall configurations
- 08:31 PM Backport #49519: nautilus: client: wake up the front pos waiter
- Xiubo, please take this one.
- 07:28 PM Backport #49933 (Rejected): nautilus: MDS should return -ENODATA when asked to remove xattr that ...
- https://tracker.ceph.com/issues/49833#note-12
- 07:28 PM Backport #49931 (Rejected): octopus: MDS should return -ENODATA when asked to remove xattr that d...
- https://tracker.ceph.com/issues/49833#note-12
- 05:40 PM Documentation #41725 (New): Document on-disk format of inodes
- 05:39 PM Bug #44097 (Can't reproduce): nautilus: "cluster [WRN] Health check failed: 1 clients failing to ...
- 05:35 PM Feature #10679 (In Progress): Add support for the chattr +i command (immutable file)
- 03:26 PM Backport #50282 (In Progress): pacific: MDS slow request lookupino #0x100 on rank 1 block forever...
- 03:24 PM Backport #50254 (In Progress): pacific: mds: standby-replay only trims cache when it reaches the ...
- 03:23 PM Backport #50289 (In Progress): pacific: MDS stuck at stopping when reducing max_mds
- 03:21 PM Backport #50285 (In Progress): pacific: qa: test standby_replay in workloads
- *
- 03:19 PM Backport #50287 (In Progress): pacific: qa: "log [ERR] : error reading sessionmap 'mds2_sessionmap'"
- 02:15 PM Bug #50083: CephFS file access issues using kernel driver: file overwritten with null bytes
- Hi sorry for the delay.
We were fairly panicked about this so have moved our kit onto a different kernel that does... - 01:10 PM Backport #50354 (Rejected): octopus: mgr/nfs: validation error on creating custom export
- 01:07 PM Documentation #50161: mgr/nfs: validation error on creating custom export
- Backport to Octopus too. Because this related PR is backported https://github.com/ceph/ceph/pull/40766
- 10:53 AM Bug #49833: MDS should return -ENODATA when asked to remove xattr that doesn't exist
- I don't think this should be backported to octopus and nautilus since the change that introduced the incorrect handli...
- 04:25 AM Bug #50216 (In Progress): qa: "ls: cannot access 'lost+found': No such file or directory"
- 03:53 AM Bug #48365 (Fix Under Review): qa: ffsb build failure on CentOS 8.2
- 03:37 AM Bug #48365: qa: ffsb build failure on CentOS 8.2
- From https://www.gnu.org/software/automake/manual/automake.html, it seems the ffsb is using the old form of `AM_INIT_...
- 01:43 AM Bug #48365: qa: ffsb build failure on CentOS 8.2
- ...
- 01:31 AM Bug #48365: qa: ffsb build failure on CentOS 8.2
- Patrick Donnelly wrote:
> Back from the dead: /ceph/teuthology-archive/pdonnell-2021-04-08_22:42:24-fs-wip-pdonnell-...
04/13/2021
- 02:24 PM Bug #50035: cephfs-mirror: use sensible mount/shutdown timeouts
- Patrick Donnelly wrote:
> Venky Shankar wrote:
> > Mostly setting this as a config option (in ceph.conf) would suff... - 02:04 PM Bug #50035: cephfs-mirror: use sensible mount/shutdown timeouts
- Venky Shankar wrote:
> Mostly setting this as a config option (in ceph.conf) would suffice. However, cephfs-mirror c... - 01:07 PM Bug #50035: cephfs-mirror: use sensible mount/shutdown timeouts
- Mostly setting this as a config option (in ceph.conf) would suffice. However, cephfs-mirror can connect to the remote...
- 05:52 AM Bug #49939 (Fix Under Review): cephfs-mirror: be resilient to recreated snapshot during synchroni...
04/12/2021
- 09:19 PM Bug #50083: CephFS file access issues using kernel driver: file overwritten with null bytes
- David, ping? Are you able to answer the questions above?
- 07:20 PM Bug #49873 (Duplicate): ceph_lremovexattr does not return error on file in ceph pacific
- Thanks for checking that Sidharth.
- 03:15 PM Bug #49873: ceph_lremovexattr does not return error on file in ceph pacific
- That sounds sensible to me, thanks! I did attempt to build ceph a copule of weeks ago, but unfortunately I was unabl...
- 02:54 PM Bug #49873: ceph_lremovexattr does not return error on file in ceph pacific
- I tested this incorrectly last week using libcephfs due to an incorrect reproduction of the steps. Upon retesting cor...
- 07:19 PM Bug #50305 (Resolved): MDS doesn't set fscrypt flag on new inodes with crypto context in xattr bu...
- The new fscrypt context handling code will set the "fscrypt" flag when the encryption.ctx xattr is set explicitly, bu...
- 06:03 PM Backport #50253 (In Progress): pacific: mds: "cluster [WRN] Scrub error on inode 0x1000000039d (/...
- 04:46 PM Bug #50271 (Won't Fix): vmware esxi NFS client cannot create thin provisioned vmdk files
- Closing this as WONTFIX since we can't really do it without harming performance for important workloads.
- 03:48 PM Bug #50271: vmware esxi NFS client cannot create thin provisioned vmdk files
- I responded in the original email thread. The problem here is that ceph doesn't report sparse file usage correctly. W...
- 01:44 PM Bug #50271 (Triaged): vmware esxi NFS client cannot create thin provisioned vmdk files
- 02:26 PM Bug #50237: cephfs-journal-tool/cephfs-data-scan: Stuck in infinite loop with "NetHandler create_...
- thanks for the reply!
i'll collect relevant mon/mds debug logs, and the strace outputs this week
if there is anythi... - 01:49 PM Bug #50237 (Need More Info): cephfs-journal-tool/cephfs-data-scan: Stuck in infinite loop with "N...
- 01:46 PM Bug #50258 (Triaged): pacific: qa: "run() got an unexpected keyword argument 'stdin_data'"
- 01:44 PM Bug #50279 (Triaged): qa: "Replacing daemon mds.b as rank 0 with standby daemon mds.c"
- 11:16 AM Bug #50298 (Fix Under Review): libcephfs: support file descriptor based *at() APIs
- 11:14 AM Bug #50298 (Resolved): libcephfs: support file descriptor based *at() APIs
04/11/2021
- 10:00 AM Backport #50252 (Need More Info): octopus: mds: "cluster [WRN] Scrub error on inode 0x1000000039d...
- cherry-pick applies cleanly, but the resulting code does not compile
see https://github.com/ceph/ceph/pull/40781 f...
04/10/2021
- 09:44 PM Backport #50288 (In Progress): octopus: MDS stuck at stopping when reducing max_mds
- 03:10 AM Backport #50288 (Resolved): octopus: MDS stuck at stopping when reducing max_mds
- https://github.com/ceph/ceph/pull/40768
- 09:43 PM Backport #50290 (In Progress): nautilus: MDS stuck at stopping when reducing max_mds
- 05:12 PM Backport #50290 (Need More Info): nautilus: MDS stuck at stopping when reducing max_mds
- appears to depend on #48991 which is a large, non-trivial changeset
- 03:10 AM Backport #50290 (Resolved): nautilus: MDS stuck at stopping when reducing max_mds
- https://github.com/ceph/ceph/pull/40769
- 09:34 PM Backport #50286 (In Progress): octopus: qa: "log [ERR] : error reading sessionmap 'mds2_sessionmap'"
- 03:10 AM Backport #50286 (Resolved): octopus: qa: "log [ERR] : error reading sessionmap 'mds2_sessionmap'"
- https://github.com/ceph/ceph/pull/40783
- 09:33 PM Backport #50283 (In Progress): octopus: MDS slow request lookupino #0x100 on rank 1 block forever...
- 03:00 AM Backport #50283 (Resolved): octopus: MDS slow request lookupino #0x100 on rank 1 block forever on...
- https://github.com/ceph/ceph/pull/40782
- 09:30 PM Backport #50252 (In Progress): octopus: mds: "cluster [WRN] Scrub error on inode 0x1000000039d (/...
- 09:30 PM Backport #50188 (Need More Info): octopus: qa: "Assertion `cb_done' failed."
- not clear how new yaml file fits into octopus directory structure
- 09:28 PM Backport #50188 (In Progress): octopus: qa: "Assertion `cb_done' failed."
- 09:27 PM Backport #50181 (In Progress): octopus: client: only check pool permissions for regular files
- 09:26 PM Backport #50027 (In Progress): octopus: client: items pinned in cache preventing unmount
- 09:25 PM Backport #49950 (In Progress): octopus: mgr/nfs: Update about cephadm single nfs-ganesha daemon p...
- 09:23 PM Backport #49934 (In Progress): octopus: libcephfs: test termination "what(): Too many open files"
- 09:23 PM Backport #49752 (In Progress): octopus: snap-schedule doc
- 09:20 PM Backport #49851 (In Progress): octopus: mds: race of fetching large dirfrag
- 09:16 PM Backport #49611 (In Progress): octopus: qa: racy session evicted check
- 09:14 PM Backport #49560 (In Progress): octopus: qa: file system deletion not complete because starter fs ...
- 09:13 PM Backport #49518 (In Progress): octopus: client: wake up the front pos waiter
- 09:12 PM Backport #49515 (In Progress): octopus: pybind/cephfs: DT_REG and DT_LNK values are wrong
- 09:11 PM Backport #49514 (In Progress): nautilus: client: allow looking up snapped inodes by inode number+...
- 09:10 PM Backport #49513 (In Progress): octopus: client: allow looking up snapped inodes by inode number+s...
- 09:01 PM Backport #49472 (In Progress): octopus: qa: ffsb workload: PG_AVAILABILITY|PG_DEGRADED warnings
- 08:58 PM Backport #49413 (In Progress): octopus: mgr/nfs: Update about user config
- 08:57 PM Backport #49347 (In Progress): octopus: qa: Test failure: test_damaged_dentry (tasks.cephfs.test_...
- 08:56 PM Backport #48878 (In Progress): octopus: mds: fix recall defaults based on feedback from productio...
- 08:55 PM Backport #48836 (In Progress): octopus: have mount helper pick appropriate mon sockets for ms_mod...
- 08:54 PM Backport #45853 (In Progress): octopus: cephfs-journal-tool: NetHandler create_socket couldn't cr...
- 04:00 PM Bug #50271: vmware esxi NFS client cannot create thin provisioned vmdk files
- I have upgraded now to Octopus 15.2.10 and the issue is still present
Attached is the packet capture from the esxi... - 01:03 PM Backport #50284 (In Progress): nautilus: MDS slow request lookupino #0x100 on rank 1 block foreve...
- 03:00 AM Backport #50284 (Rejected): nautilus: MDS slow request lookupino #0x100 on rank 1 block forever o...
- 12:59 PM Backport #50251 (Need More Info): nautilus: mds: "cluster [WRN] Scrub error on inode 0x1000000039...
- non-trivial backport. The cherry-pick from master applies cleanly, but the resulting code does not compile -- presuma...
- 11:53 AM Backport #50251 (In Progress): nautilus: mds: "cluster [WRN] Scrub error on inode 0x1000000039d (...
- 12:44 PM Backport #50255 (In Progress): nautilus: mds: standby-replay only trims cache when it reaches the...
- 11:58 AM Backport #50255 (Need More Info): nautilus: mds: standby-replay only trims cache when it reaches ...
- non-trivial backport
- 12:15 PM Backport #48813 (In Progress): octopus: mds: spurious wakeups in cache upkeep
- 12:11 PM Backport #50256 (In Progress): octopus: mds: standby-replay only trims cache when it reaches the ...
- 11:52 AM Backport #50189 (Need More Info): nautilus: qa: "Assertion `cb_done' failed."
- this one is non-trivial to backport because the yaml structure under qa/ has changed substantially since nautilus
- 10:58 AM Backport #50184 (Need More Info): octopus: client: openned inodes counter is inconsistent
- This ticket is for tracking the octopus backport of a follow-on fix for #46865 which was backported to pacific only?
- 10:57 AM Backport #50182 (Need More Info): nautilus: client: openned inodes counter is inconsistent
- This ticket is for tracking the nautilus backport of a follow-on fix for #46865 which was backported to pacific only?
- 03:10 AM Backport #50289 (Resolved): pacific: MDS stuck at stopping when reducing max_mds
- https://github.com/ceph/ceph/pull/40856
- 03:10 AM Backport #50287 (Resolved): pacific: qa: "log [ERR] : error reading sessionmap 'mds2_sessionmap'"
- https://github.com/ceph/ceph/pull/40852
- 03:09 AM Bug #50215 (Pending Backport): qa: "log [ERR] : error reading sessionmap 'mds2_sessionmap'"
- 03:08 AM Bug #50112 (Pending Backport): MDS stuck at stopping when reducing max_mds
- 03:05 AM Backport #50285 (Resolved): pacific: qa: test standby_replay in workloads
- https://github.com/ceph/ceph/pull/40853
- 03:03 AM Fix #50045 (Pending Backport): qa: test standby_replay in workloads
- 03:02 AM Bug #49466 (Resolved): qa: "Command failed on gibba030 with status 1: 'set -ex\nsudo dd of=/tmp/t...
- 03:00 AM Backport #50282 (Resolved): pacific: MDS slow request lookupino #0x100 on rank 1 block forever on...
- https://github.com/ceph/ceph/pull/40856
- 02:58 AM Bug #49922 (Pending Backport): MDS slow request lookupino #0x100 on rank 1 block forever on dispa...
- 02:55 AM Bug #50281 (Resolved): qa: untar_snap_rm timeout
- ...
- 02:45 AM Bug #48365 (New): qa: ffsb build failure on CentOS 8.2
- Back from the dead: /ceph/teuthology-archive/pdonnell-2021-04-08_22:42:24-fs-wip-pdonnell-testing-20210408.192301-dis...
- 02:39 AM Bug #50279 (Need More Info): qa: "Replacing daemon mds.b as rank 0 with standby daemon mds.c"
- ...
04/09/2021
- 10:24 PM Bug #47689: rados/upgrade/nautilus-x-singleton fails due to cluster [WRN] evicting unresponsive c...
- looks similar: https://tracker.ceph.com/issues/50275
- 05:20 PM Bug #50271 (Won't Fix): vmware esxi NFS client cannot create thin provisioned vmdk files
- I have a 3 node Ceph octopus 15.2.7 cluster running on fully up to date Centos 7 with nfs-ganesha 3.5.
After follo... - 03:41 PM Backport #50015: pacific: qa: "AttributeError: 'NoneType' object has no attribute 'mon_manager'"
- Backport Bot wrote:
> https://github.com/ceph/ceph/pull/40645
merged - 03:41 PM Backport #50179 (In Progress): nautilus: client: only check pool permissions for regular files
- 03:40 PM Backport #50128 (Need More Info): nautilus: pybind/mgr/volumes: deadlock on async job hangs finis...
- nautilus backport is non-trivial
- 03:39 PM Bug #50266: "ceph fs snapshot mirror daemon status" should not use json keys as value
- After discussing with Venky, it seems that a daemon can mirror multiple ffs so we need another list for the fs_id, so...
- 01:33 PM Bug #50266 (Resolved): "ceph fs snapshot mirror daemon status" should not use json keys as value
- Currently the command outputs:...
- 10:19 AM Bug #49684 (Resolved): qa: fs:cephadm mount does not wait for mds to be created
- While running with --resolve-parent, the script "backport-create-issue" noticed that all backports of this issue are ...
- 10:15 AM Backport #50025 (Resolved): pacific: client: items pinned in cache preventing unmount
- This update was made using the script "backport-resolve-issue".
backport PR https://github.com/ceph/ceph/pull/40629
m... - 10:15 AM Backport #50022 (Resolved): pacific: ceph-fuse: src/include/buffer.h: 1187: FAILED ceph_assert(_n...
- This update was made using the script "backport-resolve-issue".
backport PR https://github.com/ceph/ceph/pull/40628
m... - 10:15 AM Backport #50030 (Resolved): pacific: qa: fs:cephadm mount does not wait for mds to be created
- This update was made using the script "backport-resolve-issue".
backport PR https://github.com/ceph/ceph/pull/40528
m... - 09:01 AM Backport #50026 (In Progress): nautilus: client: items pinned in cache preventing unmount
- 09:01 AM Bug #49936: ceph-fuse: src/include/buffer.h: 1187: FAILED ceph_assert(_num <= 1024)
- @Patrick, you asked for this bugfix to be backported all the way back to nautilus, but AFAICT the code containing the...
- 09:00 AM Backport #50024 (Need More Info): octopus: ceph-fuse: src/include/buffer.h: 1187: FAILED ceph_ass...
- AFAICT the code being fixed does not exist in Octopus
- 08:58 AM Backport #50023 (Need More Info): nautilus: ceph-fuse: src/include/buffer.h: 1187: FAILED ceph_as...
- AFAICT the code being fixed does not exist in nautilus
- 08:56 AM Backport #49931 (Need More Info): octopus: MDS should return -ENODATA when asked to remove xattr ...
- AFAICT the line that introduced the bug is not present in octopus.
- 08:56 AM Bug #49833: MDS should return -ENODATA when asked to remove xattr that doesn't exist
- @Patrick, you requested that this bugfix be backported to octopus, but AFAICT the line that introduced the bug is not...
- 08:53 AM Bug #49833: MDS should return -ENODATA when asked to remove xattr that doesn't exist
- @Patrick, you requested that this bugfix be backported to nautilus, but AFAICT the line that introduced the bug is no...
- 08:54 AM Backport #49933 (Need More Info): nautilus: MDS should return -ENODATA when asked to remove xattr...
- AFAICT the line that introduced the bug is not present in nautilus.
- 07:42 AM Backport #49853 (In Progress): nautilus: mds: race of fetching large dirfrag
- 06:21 AM Bug #49912: client: dir->dentries inconsistent, both newname and oldname points to same inode, m...
- I have figured one case could reprodce it in theory:
1, I have check the `mv` source code, before doing the `renam... - 03:10 AM Feature #48791 (Rejected): mds: support file block size
- Discussions over email indicate we'll be going in another direction.
- 03:06 AM Feature #41566: mds: support rolling upgrades
- Jos's current work: https://github.com/ceph/ceph/pull/36821
04/08/2021
- 10:22 PM Backport #50025: pacific: client: items pinned in cache preventing unmount
- Backport Bot wrote:
> https://github.com/ceph/ceph/pull/40629
merged - 10:21 PM Backport #50022: pacific: ceph-fuse: src/include/buffer.h: 1187: FAILED ceph_assert(_num <= 1024)
- Backport Bot wrote:
> https://github.com/ceph/ceph/pull/40628
merged - 10:21 PM Backport #50030: pacific: qa: fs:cephadm mount does not wait for mds to be created
- Backport Bot wrote:
> https://github.com/ceph/ceph/pull/40528
merged - 09:17 PM Bug #50237: cephfs-journal-tool/cephfs-data-scan: Stuck in infinite loop with "NetHandler create_...
- The warning is probably unrelated. Please collect logs.
- 01:29 PM Bug #50237 (Need More Info): cephfs-journal-tool/cephfs-data-scan: Stuck in infinite loop with "N...
- Both tools are getting stuck in an infinite loop and only outputting this message.
Env:
version: 15.2.4
docker... - 08:40 PM Backport #49613 (In Progress): nautilus: qa: racy session evicted check
- 08:37 PM Bug #50260 (New): pacific: qa: "rmdir: failed to remove '/home/ubuntu/cephtest': Directory not em...
- ...
- 08:29 PM Bug #50258 (Resolved): pacific: qa: "run() got an unexpected keyword argument 'stdin_data'"
- ...
- 08:22 PM Backport #49471 (In Progress): nautilus: qa: ffsb workload: PG_AVAILABILITY|PG_DEGRADED warnings
- 07:54 PM Backport #49519 (Need More Info): nautilus: client: wake up the front pos waiter
- commit applies cleanly in nautilus, but does not compile. The compiler error can be seen in a comment I appended to t...
- 05:53 PM Backport #49519 (In Progress): nautilus: client: wake up the front pos waiter
- 06:40 PM Backport #50256 (Resolved): octopus: mds: standby-replay only trims cache when it reaches the end...
- https://github.com/ceph/ceph/pull/40743
- 06:40 PM Backport #50255 (Resolved): nautilus: mds: standby-replay only trims cache when it reaches the en...
- https://github.com/ceph/ceph/pull/40744
- 06:40 PM Backport #50254 (Resolved): pacific: mds: standby-replay only trims cache when it reaches the end...
- https://github.com/ceph/ceph/pull/40855
- 06:35 PM Backport #50253 (Resolved): pacific: mds: "cluster [WRN] Scrub error on inode 0x1000000039d (/cli...
- https://github.com/ceph/ceph/pull/40825
- 06:35 PM Backport #50252 (Rejected): octopus: mds: "cluster [WRN] Scrub error on inode 0x1000000039d (/cli...
- 06:35 PM Backport #50251 (Rejected): nautilus: mds: "cluster [WRN] Scrub error on inode 0x1000000039d (/cl...
- 06:35 PM Bug #50048 (Pending Backport): mds: standby-replay only trims cache when it reaches the end of th...
- 06:34 PM Bug #48805 (Pending Backport): mds: "cluster [WRN] Scrub error on inode 0x1000000039d (/client.0/...
- 06:30 PM Bug #50250 (New): mds: "log [WRN] : Scrub error on inode 0x10000004506 (/client.0/tmp/clients/cli...
- ...
- 05:58 PM Backport #49562 (In Progress): nautilus: qa: file system deletion not complete because starter fs...
- 05:54 PM Backport #49475 (In Progress): octopus: nautilus: qa: "Assertion `cb_done' failed."
- 05:11 PM Bug #50246 (Resolved): mds: failure replaying journal (EMetaBlob)
- ...
- 05:07 PM Backport #49516 (In Progress): nautilus: pybind/cephfs: DT_REG and DT_LNK values are wrong
- 05:05 PM Backport #49514 (Need More Info): nautilus: client: allow looking up snapped inodes by inode numb...
- large, non-trivial changeset
- 04:57 PM Backport #49473 (In Progress): nautilus: nautilus: qa: "Assertion `cb_done' failed."
- 03:51 PM Backport #50086 (In Progress): pacific: tasks.cephfs.test_volumes.TestSubvolumeGroups: RuntimeErr...
- 03:50 PM Backport #50173 (In Progress): pacific: mgr/nfs: validation error on creating custom export
- 03:50 PM Backport #50180 (In Progress): pacific: client: only check pool permissions for regular files
- 03:49 PM Backport #50183 (In Progress): pacific: client: openned inodes counter is inconsistent
- 03:49 PM Backport #50185 (In Progress): pacific: qa: "RADOS object not found (Failed to operate read op fo...
- 03:49 PM Feature #50150: qa: begin grepping kernel logs for kclient warnings/failures to fail a test
- "BUG:"
"Oops:"
"INFO: task .+ blocked for more than .+ seconds."
"KASAN"
I'm not sure if we build with KASAN as... - 03:48 PM Backport #50190 (In Progress): pacific: qa: "Assertion `cb_done' failed."
- 03:48 PM Backport #50225 (In Progress): pacific: mds: failed to decode message of type 29 v1: void CapInfo...
- 02:40 AM Backport #50225 (Resolved): pacific: mds: failed to decode message of type 29 v1: void CapInfoPay...
- https://github.com/ceph/ceph/pull/40682
- 03:47 PM Backport #50241 (In Progress): pacific: cephfs-mirror: update docs with `fs snapshot mirror daemo...
- 02:30 PM Backport #50241 (Resolved): pacific: cephfs-mirror: update docs with `fs snapshot mirror daemon s...
- https://github.com/ceph/ceph/pull/41475
- 03:10 PM Bug #45553: mds: rstats on snapshot are updated by changes to HEAD
- rstats propagation is working by now
but the propagation goofs up snapshot rstats
see this: https://tracker.ceph.co... - 02:29 PM Bug #49912: client: dir->dentries inconsistent, both newname and oldname points to same inode, m...
- Since the `ls` command output was correct, the ORDERED flag should have been cleared as expected, or it should show '...
- 07:58 AM Bug #49912: client: dir->dentries inconsistent, both newname and oldname points to same inode, m...
- Xiaoxi Chen wrote:
> Jeff Layton wrote:
> > I think that after the mv, the directory should no longer be considered... - 02:28 PM Documentation #50229 (Pending Backport): cephfs-mirror: update docs with `fs snapshot mirror daem...
- 09:42 AM Documentation #50229 (Resolved): cephfs-mirror: update docs with `fs snapshot mirror daemon statu...
- This was missing in the docs.
- 01:59 PM Bug #50238: mds: ceph.dir.rctime for older snaps is erroneously updated
- To fix this issue:
# the correct sequence of snap IDs must be chosen to be updated on a mksnap request (snapshot cre... - 01:54 PM Bug #50238 (New): mds: ceph.dir.rctime for older snaps is erroneously updated
- vxattr ceph.dir.rctime shows new values for older snaps when new snap is created or when live data is updated.
- 12:11 PM Feature #50235 (Resolved): allow cephfs-shell to mount named filesystems
- Currently, cephfs-shell can only mount the "default" filesystem. There may be some way to use the config file to dire...
- 02:37 AM Bug #49972 (Pending Backport): mds: failed to decode message of type 29 v1: void CapInfoPayload::...
- 02:36 AM Bug #50021 (Resolved): qa: snaptest-git-ceph failure during mon thrashing
- 02:30 AM Bug #50224 (Resolved): qa: test_mirroring_init_failure_with_recovery failure
- ...
- 02:20 AM Bug #50221 (New): qa: snaptest-git-ceph failure in git diff
- ...
- 02:12 AM Bug #50220 (New): qa: dbench workload timeout
- ...
04/07/2021
- 07:46 PM Bug #50216 (Resolved): qa: "ls: cannot access 'lost+found': No such file or directory"
- ...
- 07:28 PM Bug #50215 (Fix Under Review): qa: "log [ERR] : error reading sessionmap 'mds2_sessionmap'"
- 07:26 PM Bug #50215 (Resolved): qa: "log [ERR] : error reading sessionmap 'mds2_sessionmap'"
- ...
- 01:43 PM Backport #50015 (In Progress): pacific: qa: "AttributeError: 'NoneType' object has no attribute '...
- 06:51 AM Bug #50178: qa: "TypeError: run() got an unexpected keyword argument 'shell'"
- Fix - https://github.com/ceph/teuthology/pull/1639
- 06:50 AM Bug #50178 (Fix Under Review): qa: "TypeError: run() got an unexpected keyword argument 'shell'"
- 04:58 AM Bug #50112: MDS stuck at stopping when reducing max_mds
- 玮文 胡 wrote:
> Xiubo Li wrote:
> > For the mds private inode/dirs, it shouldn't be normal to handle the client caps,... - 04:44 AM Bug #50112: MDS stuck at stopping when reducing max_mds
- Xiubo Li wrote:
> For the mds private inode/dirs, it shouldn't be normal to handle the client caps, right ?
I thi... - 02:27 AM Bug #50112: MDS stuck at stopping when reducing max_mds
- Patrick Donnelly wrote:
> I wonder if this is related to
>
> https://tracker.ceph.com/issues/49922
IMO you are... - 01:18 AM Bug #50112: MDS stuck at stopping when reducing max_mds
- 玮文 胡 wrote:
> Xiubo Li wrote:
> > It seems the "mdsdir_in->get_num_ref()" is not zero in mds.0
>
> Yes, do you h... - 04:13 AM Backport #50186 (In Progress): pacific: qa: daemonwatchdog fails if mounts not defined
04/06/2021
- 07:25 PM Backport #50127 (In Progress): pacific: pybind/mgr/volumes: deadlock on async job hangs finisher ...
- 07:22 PM Backport #50025 (In Progress): pacific: client: items pinned in cache preventing unmount
- 07:18 PM Backport #50022 (In Progress): pacific: ceph-fuse: src/include/buffer.h: 1187: FAILED ceph_assert...
- 06:10 PM Backport #50187 (In Progress): pacific: ceph-dokan improvements for additional mounts
- 05:50 PM Backport #50187 (Resolved): pacific: ceph-dokan improvements for additional mounts
- https://github.com/ceph/ceph/pull/40627
- 05:50 PM Backport #50190 (Resolved): pacific: qa: "Assertion `cb_done' failed."
- https://github.com/ceph/ceph/pull/40683
- 05:50 PM Backport #50189 (Rejected): nautilus: qa: "Assertion `cb_done' failed."
- 05:50 PM Backport #50188 (Rejected): octopus: qa: "Assertion `cb_done' failed."
- 05:46 PM Bug #49662 (Pending Backport): ceph-dokan improvements for additional mounts
- 05:46 PM Backport #50186 (Resolved): pacific: qa: daemonwatchdog fails if mounts not defined
- https://github.com/ceph/ceph/pull/40634
- 05:45 PM Backport #50185 (Resolved): pacific: qa: "RADOS object not found (Failed to operate read op for o...
- https://github.com/ceph/ceph/pull/40684
- 05:45 PM Backport #50184 (Rejected): octopus: client: openned inodes counter is inconsistent
- 05:45 PM Backport #50183 (Resolved): pacific: client: openned inodes counter is inconsistent
- https://github.com/ceph/ceph/pull/40685
- 05:45 PM Backport #50182 (Rejected): nautilus: client: openned inodes counter is inconsistent
- 05:45 PM Bug #49500 (Pending Backport): qa: "Assertion `cb_done' failed."
- 05:45 PM Backport #50181 (Resolved): octopus: client: only check pool permissions for regular files
- https://github.com/ceph/ceph/pull/40779
- 05:45 PM Backport #50180 (Resolved): pacific: client: only check pool permissions for regular files
- https://github.com/ceph/ceph/pull/40686
- 05:45 PM Backport #50179 (Resolved): nautilus: client: only check pool permissions for regular files
- https://github.com/ceph/ceph/pull/40730
- 05:44 PM Bug #50090 (Pending Backport): client: only check pool permissions for regular files
- 05:43 PM Bug #50020 (Pending Backport): qa: "RADOS object not found (Failed to operate read op for oid cep...
- 05:41 PM Bug #50057 (Pending Backport): client: openned inodes counter is inconsistent
- 05:30 PM Bug #50178 (Rejected): qa: "TypeError: run() got an unexpected keyword argument 'shell'"
- ...
- 03:25 PM Backport #50173 (Resolved): pacific: mgr/nfs: validation error on creating custom export
- https://github.com/ceph/ceph/pull/40687
- 03:24 PM Documentation #50161 (Pending Backport): mgr/nfs: validation error on creating custom export
- 10:24 AM Documentation #50161 (In Progress): mgr/nfs: validation error on creating custom export
- 10:18 AM Documentation #50161 (Resolved): mgr/nfs: validation error on creating custom export
- ...
- 12:33 PM Documentation #49406: Exceeding osd nearfull ratio causes write throttle.
- It's unfortunate that it caught you by surprise. Would you care to draft a patch to update the documentation? Where w...
- 11:59 AM Bug #50112: MDS stuck at stopping when reducing max_mds
- Xiubo Li wrote:
> BTW, do you have more logs for the mds.0 ? I need to check the logs to get why.
Only less than ... - 11:14 AM Bug #50112: MDS stuck at stopping when reducing max_mds
- 玮文 胡 wrote:
> Xiubo Li wrote:
> > It seems the "mdsdir_in->get_num_ref()" is not zero in mds.0
>
> Yes, do you h... - 11:00 AM Bug #50112: MDS stuck at stopping when reducing max_mds
- Xiubo Li wrote:
> It seems the "mdsdir_in->get_num_ref()" is not zero in mds.0
Yes, do you have any idea why? As ... - 10:46 AM Bug #50112: MDS stuck at stopping when reducing max_mds
- It seems the "mdsdir_in->get_num_ref()" is not zero in mds.0:...
- 09:03 AM Bug #50112: MDS stuck at stopping when reducing max_mds
- 玮文 胡 wrote:
> Some more progress. mds.0 seems actually sending an empty cache_expire message to mds.1, despite sayin... - 09:22 AM Bug #50021 (Fix Under Review): qa: snaptest-git-ceph failure during mon thrashing
- 05:54 AM Bug #50021: qa: snaptest-git-ceph failure during mon thrashing
- Patrick Donnelly wrote:
> Xiubo Li wrote:
>
[...]
> > > >
> > > > What repo do you want to mirror?
> > >
... - 03:21 AM Bug #50021: qa: snaptest-git-ceph failure during mon thrashing
- Xiubo Li wrote:
> Patrick Donnelly wrote:
> > David Galloway wrote:
> > > Xiubo Li wrote:
> > > > Patrick Donnell... - 02:15 AM Bug #50021: qa: snaptest-git-ceph failure during mon thrashing
- Patrick Donnelly wrote:
> David Galloway wrote:
> > Xiubo Li wrote:
> > > Patrick Donnelly wrote:
> > > > Xiubo L... - 01:50 AM Feature #50150 (Pending Backport): qa: begin grepping kernel logs for kclient warnings/failures t...
- Right now, TMK, we are not confirming there are no warnings/errors/lockups in the kclient before passing a test. We d...
- 01:46 AM Cleanup #50149 (Resolved): client: always register callbacks before mount()
- Make Client::ll_register_callbacks() return 'int' type value instead
and return negative errno if not successful.
...
04/05/2021
- 05:56 PM Bug #48439: fsstress failure with mds thrashing: "mds.0.6 Evicting (and blocklisting) client sess...
- Deepika Upadhyay wrote:
> @Patrick I am seeing this issue on Ubuntu 18.04; doesn't seem to be related to testing pr.... - 05:02 PM Bug #48439: fsstress failure with mds thrashing: "mds.0.6 Evicting (and blocklisting) client sess...
- @Patrick I am seeing this issue on Ubuntu 18.04; doesn't seem to be related to testing pr.
Not sure if it's related ... - 05:12 PM Bug #50112: MDS stuck at stopping when reducing max_mds
- I wonder if this is related to
https://tracker.ceph.com/issues/49922
The code would indicate it's normal for md... - 04:48 PM Bug #50021: qa: snaptest-git-ceph failure during mon thrashing
- David Galloway wrote:
> Xiubo Li wrote:
> > Patrick Donnelly wrote:
> > > Xiubo Li wrote:
> > >
> > > [...]
> >... - 01:35 PM Bug #50035 (Triaged): cephfs-mirror: use sensible mount/shutdown timeouts
- 11:47 AM Feature #49304 (Resolved): nfs-ganesha: plumb xattr support into FSAL_CEPH
- Patches are merged into the ganesha next branch and should make next release.
04/04/2021
- 02:06 PM Bug #50112: MDS stuck at stopping when reducing max_mds
- Some more progress. mds.0 seems actually sending an empty cache_expire message to mds.1, despite saying "successfully...
- 09:09 AM Bug #50112: MDS stuck at stopping when reducing max_mds
- I spent some more time investigating this. Hope this is helpful.
When I issue "dump cache" to the stopping mds.1, ...
04/03/2021
- 04:35 PM Bug #49912: client: dir->dentries inconsistent, both newname and oldname points to same inode, m...
- Xiubo Li wrote:
> I am doubting that if there has two tasks are doing the rename:
>
> For task1, if it just do _l... - 04:21 PM Bug #49912: client: dir->dentries inconsistent, both newname and oldname points to same inode, m...
- Jeff Layton wrote:
> I think that after the mv, the directory should no longer be considered ORDERED. We probably _c... - 02:25 PM Backport #50128 (Resolved): nautilus: pybind/mgr/volumes: deadlock on async job hangs finisher th...
- https://github.com/ceph/ceph/pull/41394
- 02:25 PM Backport #50127 (Resolved): pacific: pybind/mgr/volumes: deadlock on async job hangs finisher thread
- https://github.com/ceph/ceph/pull/40630
- 02:25 PM Backport #50126 (Rejected): octopus: pybind/mgr/volumes: deadlock on async job hangs finisher thread
- https://github.com/ceph/ceph/pull/43269
- 02:24 PM Bug #49605 (Pending Backport): pybind/mgr/volumes: deadlock on async job hangs finisher thread
04/02/2021
- 05:44 PM Bug #50112 (Triaged): MDS stuck at stopping when reducing max_mds
- 01:42 PM Bug #50112: MDS stuck at stopping when reducing max_mds
- I've figured out "7 mds.1.cache still have replicated objects" may be the reason that this MDS cannot complete its sh...
- 12:13 PM Bug #50112 (Resolved): MDS stuck at stopping when reducing max_mds
- We are trying to upgrade to v16 today. Cephadm is trying to reduce max_mds to 1 automatically. However, MDS.1 is stuc...
- 07:10 AM Feature #46865 (Resolved): client: add metric for number of pinned capabilities
- While running with --resolve-parent, the script "backport-create-issue" noticed that all backports of this issue are ...
- 07:09 AM Bug #48559 (Resolved): qa: ERROR: test_damaged_dentry, KeyError: 'passed_validation'
- While running with --resolve-parent, the script "backport-create-issue" noticed that all backports of this issue are ...
- 07:09 AM Fix #48802 (Resolved): mds: define CephFS errors that replace standard errno values
- While running with --resolve-parent, the script "backport-create-issue" noticed that all backports of this issue are ...
- 07:08 AM Bug #48912 (Resolved): ls -l in cephfs-shell tries to chase symlinks when stat'ing and errors out...
- While running with --resolve-parent, the script "backport-create-issue" noticed that all backports of this issue are ...
- 07:08 AM Bug #49074 (Resolved): mds: don't start purging inodes in the middle of recovery
- While running with --resolve-parent, the script "backport-create-issue" noticed that all backports of this issue are ...
- 07:08 AM Bug #49121 (Resolved): vstart: volumes/nfs interface complaints cluster does not exists
- While running with --resolve-parent, the script "backport-create-issue" noticed that all backports of this issue are ...
- 07:07 AM Bug #49391 (Resolved): qa: run fs:verify with tcmalloc
- While running with --resolve-parent, the script "backport-create-issue" noticed that all backports of this issue are ...
- 07:07 AM Bug #49464 (Resolved): qa: rank_freeze prevents failover on some tests
- While running with --resolve-parent, the script "backport-create-issue" noticed that all backports of this issue are ...
- 07:07 AM Bug #49507 (Resolved): qa: mds removed because trimming for too long with valgrind
- While running with --resolve-parent, the script "backport-create-issue" noticed that all backports of this issue are ...
- 07:07 AM Bug #49607 (Resolved): qa: slow metadata ops during scrubbing
- While running with --resolve-parent, the script "backport-create-issue" noticed that all backports of this issue are ...
- 07:07 AM Feature #49619 (Resolved): cephfs-mirror: add mirror peers via bootstrapping
- While running with --resolve-parent, the script "backport-create-issue" noticed that all backports of this issue are ...
- 07:07 AM Feature #49623 (Resolved): Windows CephFS support - ceph-dokan
- While running with --resolve-parent, the script "backport-create-issue" noticed that all backports of this issue are ...
- 07:06 AM Bug #49711 (Resolved): cephfs-mirror: symbolic links do not get synchronized at times
- While running with --resolve-parent, the script "backport-create-issue" noticed that all backports of this issue are ...
- 07:06 AM Bug #49719 (Resolved): mon/MDSMonitor: standby-replay daemons should be removed when the flag is ...
- While running with --resolve-parent, the script "backport-create-issue" noticed that all backports of this issue are ...
- 07:06 AM Bug #49725 (Resolved): client: crashed in cct->_conf.get_val() in Client::start_tick_thread()
- While running with --resolve-parent, the script "backport-create-issue" noticed that all backports of this issue are ...
- 07:06 AM Bug #49822 (Resolved): test: test_mirroring_command_idempotency (tasks.cephfs.test_admin.TestMirr...
- While running with --resolve-parent, the script "backport-create-issue" noticed that all backports of this issue are ...
- 03:55 AM Bug #50108 (New): access to a file with the wrong permission when changing the parent directory's...
- Recently, we tried to manage the permission of files and directories in ceph with ACL.
Basically, we planed to set...
04/01/2021
- 06:17 PM Backport #49935 (Resolved): pacific: libcephfs: test termination "what(): Too many open files"
- 06:17 PM Backport #49932 (Resolved): pacific: MDS should return -ENODATA when asked to remove xattr that d...
- 06:17 PM Backport #49929 (Resolved): pacific: test: test_mirroring_command_idempotency (tasks.cephfs.test_...
- 06:16 PM Backport #49905 (Resolved): pacific: mgr/volumes: setuid and setgid file bits are not retained af...
- 06:16 PM Backport #49854 (Resolved): pacific: client: crashed in cct->_conf.get_val() in Client::start_tic...
- 06:16 PM Backport #49852 (Resolved): pacific: mds: race of fetching large dirfrag
- 06:16 PM Backport #49765 (Resolved): pacific: cephfs-mirror: symbolic links do not get synchronized at times
- 06:16 PM Backport #49753 (Resolved): pacific: cephfs-mirror: add mirror peers via bootstrapping
- 06:15 PM Backport #49751 (Resolved): pacific: snap-schedule doc
- 06:14 PM Backport #49713 (Resolved): pacific: mgr/nfs: Add interface to update export
- 06:14 PM Backport #49687 (Resolved): pacific: client: add metric for number of pinned capabilities
- 06:14 PM Backport #49685 (Resolved): pacific: ls -l in cephfs-shell tries to chase symlinks when stat'ing ...
- 06:14 PM Backport #49634 (Resolved): pacific: Windows CephFS support - ceph-dokan
- 06:13 PM Backport #49631 (Resolved): pacific: mds: don't start purging inodes in the middle of recovery
- 06:13 PM Backport #49630 (Resolved): pacific: qa: slow metadata ops during scrubbing
- 06:13 PM Backport #49612 (Resolved): pacific: qa: racy session evicted check
- 06:13 PM Backport #49610 (Resolved): pacific: qa: mds removed because trimming for too long with valgrind
- 06:12 PM Backport #49609 (Resolved): pacific: qa: ERROR: test_damaged_dentry, KeyError: 'passed_validation'
- 06:12 PM Backport #49608 (Resolved): pacific: mds: define CephFS errors that replace standard errno values
- 06:12 PM Backport #49569 (Resolved): pacific: qa: rank_freeze prevents failover on some tests
- 06:12 PM Backport #49563 (Resolved): pacific: qa: run fs:verify with tcmalloc
- 06:11 PM Backport #49561 (Resolved): pacific: qa: file system deletion not complete because starter fs alr...
- 06:11 PM Backport #49520 (Resolved): pacific: client: wake up the front pos waiter
- 06:11 PM Backport #49517 (Resolved): pacific: pybind/cephfs: DT_REG and DT_LNK values are wrong
- 06:11 PM Backport #49512 (Resolved): pacific: client: allow looking up snapped inodes by inode number+snap...
- 06:10 PM Backport #49474 (Resolved): pacific: nautilus: qa: "Assertion `cb_done' failed."
- 06:10 PM Backport #49470 (Resolved): pacific: qa: ffsb workload: PG_AVAILABILITY|PG_DEGRADED warnings
- 06:10 PM Backport #49346 (Resolved): pacific: vstart: volumes/nfs interface complaints cluster does not ex...
- 03:45 PM Bug #49662 (Fix Under Review): ceph-dokan improvements for additional mounts
- 02:58 PM Bug #50021: qa: snaptest-git-ceph failure during mon thrashing
- Xiubo Li wrote:
> Patrick Donnelly wrote:
> > Xiubo Li wrote:
> >
> > [...]
> >
> > > @Patrick,
> > >
> > > ... - 12:27 PM Bug #50083: CephFS file access issues using kernel driver: file overwritten with null bytes
- More questions:
1) how large are these files (generally)?
2) at what point does the corruption start?
3) How far... - 12:05 PM Bug #50083: CephFS file access issues using kernel driver: file overwritten with null bytes
- Ok, so it is reproducible in your environment. The problem is that I'm unclear on what sort of I/O is being done here...
- 08:32 AM Bug #50083: CephFS file access issues using kernel driver: file overwritten with null bytes
- We do have a reliable repro currently yes. A file that we write to several times an hour shows clear corruption corre...
- 09:18 AM Bug #50091 (Fix Under Review): cephfs-top: exception: addwstr() returned ERR
- 04:24 AM Bug #50091 (Resolved): cephfs-top: exception: addwstr() returned ERR
- When the terminal is not wide enough, this can be seen 100%.
- 01:32 AM Bug #50090 (Resolved): client: only check pool permissions for regular files
- There is no need to do a check_pool_perm() on anything that isn't
a regular file, as the MDS is what handles talking...
Also available in: Atom