Project

General

Profile

Activity

From 09/07/2021 to 10/06/2021

10/06/2021

01:07 PM Bug #52829: cephfs/mirroring: peer_bootstrap import failed with EPERM
Attaching mgr logs split into two files Kotresh Hiremath Ravishankar
12:55 PM Bug #52829 (New): cephfs/mirroring: peer_bootstrap import failed with EPERM
Tried setting up mirroring on vstart cluster as below but peer_bootstrap import failed with EPERM.
#Create vstart ...
Kotresh Hiremath Ravishankar
11:22 AM Bug #52821: qa/xfstest-dev.py: update to include centos stream
rishabh-2021-10-05_09:25:11-fs-wip-rishabh-vr-run-multiple-cmds-distro-basic-smithi/6423166/... Rishabh Dave
05:46 AM Fix #52824 (Closed): qa: skip internal metadata directory when scanning ceph debugfs directory
kclient patchset:
https://patchwork.kernel.org/project/ceph-devel/list/?series=556049
introduces meta direc...
Venky Shankar
12:17 AM Cleanup #51390 (Resolved): mgr/volumes/fs/operations/access.py: fix various flake8 issues
Patrick Donnelly
12:17 AM Cleanup #51407 (Resolved): mgr/volumes/fs/operations/versions/subvolume_attrs.py: fix various fla...
Patrick Donnelly
12:16 AM Cleanup #51381 (Resolved): mgr/volumes/fs/async_job.py: fix various flake8 issues
Patrick Donnelly
12:16 AM Cleanup #51403 (Resolved): mgr/volumes/fs/operations/versions/auth_metadata.py: fix various flake...
Patrick Donnelly
12:15 AM Cleanup #51398 (Resolved): mgr/volumes/fs/operations/subvolume.py: fix various flake8 issues
Patrick Donnelly
12:15 AM Cleanup #51384 (Resolved): mgr/volumes/fs/vol_spec.py: fix various flake8 issues
Patrick Donnelly
12:15 AM Backport #52823 (Resolved): pacific: mgr/nfs: add more log messages
Backport included here:
https://github.com/ceph/ceph/pull/43682
Backport Bot
12:14 AM Cleanup #51396 (Resolved): mgr/volumes/fs/operations/clone_index.py: fix various flake8 issues
Patrick Donnelly
12:14 AM Cleanup #51380 (Resolved): mgr/volumes/module.py: fix various flake8 issues
Patrick Donnelly
12:13 AM Cleanup #51392 (Resolved): mgr/volumes/fs/operations/snapshot_util.py: add extra blank line
Patrick Donnelly
12:13 AM Bug #52274 (Pending Backport): mgr/nfs: add more log messages
Patrick Donnelly
12:02 AM Bug #52822 (Fix Under Review): qa: failed pacific install on fs:upgrade
Patrick Donnelly

10/05/2021

11:59 PM Bug #52822 (Resolved): qa: failed pacific install on fs:upgrade
https://pulpito.ceph.com/pdonnell-2021-10-04_23:25:27-fs-wip-pdonnell-testing-20211002.163337-distro-basic-smithi/642... Patrick Donnelly
06:56 PM Bug #52821 (Pending Backport): qa/xfstest-dev.py: update to include centos stream
Rishabh Dave
05:05 PM Bug #52820 (Resolved): Ceph monitor crash after upgrade from ceph 15.2.14 to 16.2.6
i tried to upgrade my ceph cluster from 15.2.14 to 16.2.6 on my proxmox 7.0 severs.
after updating the packages i ...
Daniel Keller

10/02/2021

12:18 AM Cleanup #51384 (Fix Under Review): mgr/volumes/fs/vol_spec.py: fix various flake8 issues
Varsha Rao

10/01/2021

11:23 PM Cleanup #51402 (Fix Under Review): mgr/volumes/fs/operations/versions/subvolume_base.py: fix vari...
Varsha Rao
10:52 PM Cleanup #51390 (Fix Under Review): mgr/volumes/fs/operations/access.py: fix various flake8 issues
Varsha Rao
10:50 PM Cleanup #51381 (Fix Under Review): mgr/volumes/fs/async_job.py: fix various flake8 issues
Varsha Rao
10:49 PM Cleanup #51407 (Fix Under Review): mgr/volumes/fs/operations/versions/subvolume_attrs.py: fix var...
Varsha Rao
10:48 PM Cleanup #51382 (Fix Under Review): mgr/volumes/fs/async_cloner.py: fix various flake8 issues
Varsha Rao
10:27 PM Cleanup #51403 (Fix Under Review): mgr/volumes/fs/operations/versions/auth_metadata.py: fix vario...
Varsha Rao
10:05 PM Cleanup #51398 (Fix Under Review): mgr/volumes/fs/operations/subvolume.py: fix various flake8 issues
Varsha Rao
09:03 PM Cleanup #51396 (Fix Under Review): mgr/volumes/fs/operations/clone_index.py: fix various flake8 i...
Varsha Rao
08:54 PM Cleanup #51400 (Fix Under Review): mgr/volumes/fs/operations/trash.py: fix various flake8 issues
Varsha Rao
08:25 PM Bug #51630 (Fix Under Review): mgr/snap_schedule: don't throw traceback on non-existent fs
Varsha Rao
08:10 PM Cleanup #51380 (Fix Under Review): mgr/volumes/module.py: fix various flake8 issues
Varsha Rao
09:54 AM Feature #46166: mds: store symlink target as xattr in data pool inode for disaster recovery
The link in the description is broken. Here is the working link https://docs.ceph.com/en/latest/cephfs/disaster-recov... Kotresh Hiremath Ravishankar
03:14 AM Cleanup #51392 (Fix Under Review): mgr/volumes/fs/operations/snapshot_util.py: add extra blank line
Varsha Rao

09/30/2021

08:52 PM Cleanup #51392: mgr/volumes/fs/operations/snapshot_util.py: add extra blank line
This PR fixes it: https://github.com/ceph/ceph/pull/43375 Varsha Rao
04:50 PM Cleanup #51391 (Resolved): mgr/volumes/fs/operations/resolver.py: add extra blank line
Patrick Donnelly
04:50 PM Cleanup #51387 (Resolved): mgr/volumes/fs/purge_queue.py: add extra blank line
Patrick Donnelly
01:27 PM Bug #52581 (Can't reproduce): Dangling fs snapshots on data pool after change of directory layout
Closing this issue based on previous comment. Please re-open the issue if it happens on the supported versions. Kotresh Hiremath Ravishankar
11:21 AM Bug #52581: Dangling fs snapshots on data pool after change of directory layout
The ceph version of this issue 'ceph version 13.2.10 (564bdc4ae87418a232fc901524470e1a0f76d641) mimic (stable)' is no... Kotresh Hiremath Ravishankar
08:43 AM Bug #51589: mds: crash when journaling during replay
Venky Shankar wrote:
> ...
> ...
> Patrick suggested that we could defer journaling blocklisted clients in reconne...
Venky Shankar

09/29/2021

12:56 PM Bug #51589: mds: crash when journaling during replay
I was able to reproduce this in master branch. The crash happens when a standby mds takes over as active and there ar... Venky Shankar
10:41 AM Bug #52581: Dangling fs snapshots on data pool after change of directory layout
Hi Frank,
I tried reproducing this issue on the master by changing root data pool to the new one but couldn't achi...
Kotresh Hiremath Ravishankar

09/28/2021

01:03 AM Cleanup #51391 (Fix Under Review): mgr/volumes/fs/operations/resolver.py: add extra blank line
Varsha Rao

09/27/2021

11:36 PM Backport #51199 (In Progress): octopus: msg: active_connections regression
Patrick Donnelly
07:15 AM Backport #51199: octopus: msg: active_connections regression
Backport PR: https://github.com/ceph/ceph/pull/43310
This regression issue is blocking https://tracker.ceph.com/...
gerald yang
02:48 PM Cleanup #51387 (Fix Under Review): mgr/volumes/fs/purge_queue.py: add extra blank line
Varsha Rao
01:43 PM Bug #52723 (Fix Under Review): mds: improve mds_bal_fragment_size_max config option
Patrick Donnelly
01:42 PM Feature #52725 (Fix Under Review): qa: mds_dir_max_entries workunit test case
Patrick Donnelly
01:41 PM Feature #52720 (Fix Under Review): mds: mds_bal_rank_mask config option
Patrick Donnelly

09/24/2021

10:30 AM Feature #52725 (Resolved): qa: mds_dir_max_entries workunit test case
mds_dir_max_entries workunit test case Yongseok Oh
06:04 AM Bug #52723 (Resolved): mds: improve mds_bal_fragment_size_max config option
Maintain mds_bal_fragment_size_max as a member variable in Server.cc Yongseok Oh

09/23/2021

04:55 PM Feature #51162 (Fix Under Review): mgr/volumes: `fs volume rename` command
Ramana Raja
10:27 AM Feature #52720 (Resolved): mds: mds_bal_rank_mask config option
hexadecimal bitmask of the active MDS ranks to rebalance on. MDS balancer dynamically redistributes subtrees within c... Yongseok Oh
08:24 AM Fix #52715 (New): mds: reduce memory usage during scrubbing
Breadth-first search may queue lots of inodes. Change scrub traverse to depth-First search.
The new scrub code no ...
Erqi Chen

09/22/2021

07:31 PM Backport #50126 (In Progress): octopus: pybind/mgr/volumes: deadlock on async job hangs finisher ...
Cory Snyder
06:01 AM Backport #52632 (In Progress): octopus: mds,client: add flag to MClientSession for reject reason
Kotresh Hiremath Ravishankar
05:52 AM Backport #52633 (In Progress): pacific: mds,client: add flag to MClientSession for reject reason
Kotresh Hiremath Ravishankar

09/21/2021

01:23 PM Bug #52688 (New): mds: possibly corrupted entry in journal (causes replay failure with file syste...
Failed replay after a standby took over as active. This marks the file system as damaged.
The journal entry for in...
Venky Shankar
11:55 AM Bug #52642 (Fix Under Review): snap scheduler: cephfs snapshot schedule status doesn't list the s...
Venky Shankar
12:50 AM Backport #52680 (Resolved): pacific: Add option in `fs new` command to start rank 0 in failed state
https://github.com/ceph/ceph/pull/45565 Backport Bot
12:50 AM Backport #52679 (Resolved): pacific: "cluster [WRN] 1 slow requests" in smoke pacific
https://github.com/ceph/ceph/pull/43562 Backport Bot
12:50 AM Backport #52678 (Resolved): pacific: qa: test_kill_mdstable (tasks.cephfs.test_snapshots.TestSnap...
https://github.com/ceph/ceph/pull/43702 Backport Bot
12:48 AM Bug #52625 (Pending Backport): qa: test_kill_mdstable (tasks.cephfs.test_snapshots.TestSnapshots)
Patrick Donnelly
12:47 AM Bug #52572 (Pending Backport): "cluster [WRN] 1 slow requests" in smoke pacific
Patrick Donnelly
12:46 AM Feature #51716 (Pending Backport): Add option in `fs new` command to start rank 0 in failed state
Patrick Donnelly
12:09 AM Bug #52677 (Fix Under Review): qa: test_simple failure
Patrick Donnelly
12:03 AM Bug #52677 (Resolved): qa: test_simple failure
... Patrick Donnelly

09/20/2021

04:15 PM Bug #52607 (Duplicate): qa: "mon.a (mon.0) 1022 : cluster [WRN] Health check failed: Reduced data...
Patrick Donnelly
01:48 PM Fix #52591: mds: mds_oft_prefetch_dirfrags = false is not qa tested
Patrick suggested that we have this setting in the thrasher suite. Ramana Raja
01:42 PM Bug #52641 (Triaged): snap scheduler: Traceback seen when snapshot schedule remove command is pas...
Patrick Donnelly
01:41 PM Bug #52642 (Triaged): snap scheduler: cephfs snapshot schedule status doesn't list the snapshot c...
Patrick Donnelly
01:38 PM Bug #52643 (Triaged): snap scheduler: cephfs snapshot created with schedules stopped on nfs volum...
Patrick Donnelly
11:40 AM Backport #52629 (In Progress): octopus: pybind/mgr/volumes: first subvolume permissions set perms...
Kotresh Hiremath Ravishankar
11:09 AM Backport #52628 (In Progress): pacific: pybind/mgr/volumes: first subvolume permissions set perms...
Kotresh Hiremath Ravishankar

09/17/2021

09:39 AM Feature #51518 (Resolved): client: flush the mdlog in unsafe requests' relevant and auth MDSes only
While running with --resolve-parent, the script "backport-create-issue" noticed that all backports of this issue are ... Loïc Dachary
09:34 AM Backport #51833 (Resolved): pacific: client: flush the mdlog in unsafe requests' relevant and aut...
This update was made using the script "backport-resolve-issue".
backport PR https://github.com/ceph/ceph/pull/42925
m...
Loïc Dachary
09:34 AM Backport #51937: pacific: qa: test_full_fsync (tasks.cephfs.test_full.TestClusterFull) failed
This update was made using the script "backport-resolve-issue".
backport PR https://github.com/ceph/ceph/pull/42923
m...
Loïc Dachary
09:33 AM Backport #51977 (Resolved): pacific: client: make sure only to update dir dist from auth mds
This update was made using the script "backport-resolve-issue".
backport PR https://github.com/ceph/ceph/pull/42937
m...
Loïc Dachary
06:59 AM Bug #52643: snap scheduler: cephfs snapshot created with schedules stopped on nfs volume after cr...
Moving to High Priority since python traceback causes loss of functionality.
Milind Changire
06:36 AM Bug #52643 (Closed): snap scheduler: cephfs snapshot created with schedules stopped on nfs volume...
... Milind Changire
06:33 AM Bug #52642 (Resolved): snap scheduler: cephfs snapshot schedule status doesn't list the snapshot ...
... Milind Changire
06:28 AM Bug #52641 (Resolved): snap scheduler: Traceback seen when snapshot schedule remove command is pa...
# ceph fs snap-schedule remove
Error EINVAL: Traceback (most recent call last):
File "/usr/share/ceph/mgr/mgr_mod...
Milind Changire
01:58 AM Backport #52639 (In Progress): pacific: MDSMonitor: handle damaged state from standby-replay
Patrick Donnelly
01:50 AM Backport #52639 (Resolved): pacific: MDSMonitor: handle damaged state from standby-replay
https://github.com/ceph/ceph/pull/43200 Backport Bot
01:48 AM Bug #52565 (Pending Backport): MDSMonitor: handle damaged state from standby-replay
Patrick Donnelly
01:47 AM Bug #51278: mds: "FAILED ceph_assert(!segments.empty())"
Might be related to: https://tracker.ceph.com/issues/51589 Venky Shankar
01:05 AM Backport #52627 (In Progress): pacific: cephfs-mirror: cephfs-mirror daemon status for a particul...
Venky Shankar
01:00 AM Backport #52444 (In Progress): pacific: cephfs-mirror: terminating a mirror daemon can cause a cr...
Venky Shankar

09/16/2021

08:58 PM Bug #52280: Mds crash and fails with assert on prepare_new_inode
@xiubo Li
Hi Li
Thanks again
- What are the recommended values for mds_log_segment_size and mds_log_segment_size.?...
Yael Azulay
01:31 PM Bug #52280 (Fix Under Review): Mds crash and fails with assert on prepare_new_inode
Yael Azulay wrote:
> @xiubo Li
> Thanks much, Li, for your analysis
> I didnt change mds_log_segment_size and mds_...
Xiubo Li
02:36 PM Backport #52444 (Need More Info): pacific: cephfs-mirror: terminating a mirror daemon can cause a...
Patrick Donnelly
02:32 PM Backport #52627 (Need More Info): pacific: cephfs-mirror: cephfs-mirror daemon status for a parti...
Patrick Donnelly
02:35 AM Backport #52627 (Resolved): pacific: cephfs-mirror: cephfs-mirror daemon status for a particular ...
https://github.com/ceph/ceph/pull/43199 Backport Bot
01:01 PM Fix #52591: mds: mds_oft_prefetch_dirfrags = false is not qa tested
Thanks for bringing this up Dan. We'll try to have someone work on this. Patrick Donnelly
02:42 AM Backport #52636 (Resolved): pacific: MDSMonitor: removes MDS coming out of quorum election
https://github.com/ceph/ceph/pull/43698 Backport Bot
02:40 AM Backport #52635 (Resolved): pacific: mds sends cap updates with btime zeroed out
https://github.com/ceph/ceph/pull/45163 Backport Bot
02:40 AM Backport #52634 (Resolved): octopus: mds sends cap updates with btime zeroed out
https://github.com/ceph/ceph/pull/45164 Backport Bot
02:40 AM Backport #52633 (Resolved): pacific: mds,client: add flag to MClientSession for reject reason
https://github.com/ceph/ceph/pull/43251 Backport Bot
02:40 AM Backport #52632 (Resolved): octopus: mds,client: add flag to MClientSession for reject reason
https://github.com/ceph/ceph/pull/43252 Backport Bot
02:40 AM Backport #52631 (Resolved): pacific: mds: add max_mds_entries_per_dir config option
https://github.com/ceph/ceph/pull/44512 Backport Bot
02:37 AM Feature #52491 (Pending Backport): mds: add max_mds_entries_per_dir config option
Patrick Donnelly
02:36 AM Bug #43216 (Pending Backport): MDSMonitor: removes MDS coming out of quorum election
Patrick Donnelly
02:35 AM Bug #52382 (Pending Backport): mds,client: add flag to MClientSession for reject reason
Patrick Donnelly
02:35 AM Backport #52629 (Resolved): octopus: pybind/mgr/volumes: first subvolume permissions set perms on...
https://github.com/ceph/ceph/pull/43224 Backport Bot
02:35 AM Backport #52628 (Resolved): pacific: pybind/mgr/volumes: first subvolume permissions set perms on...
https://github.com/ceph/ceph/pull/43223 Backport Bot
02:35 AM Bug #52123 (Pending Backport): mds sends cap updates with btime zeroed out
Patrick Donnelly
02:34 AM Bug #51870 (Pending Backport): pybind/mgr/volumes: first subvolume permissions set perms on /volu...
Patrick Donnelly
02:33 AM Bug #51989 (Pending Backport): cephfs-mirror: cephfs-mirror daemon status for a particular FS is ...
Patrick Donnelly
02:27 AM Bug #52626 (Triaged): mds: ScrubStack.cc: 831: FAILED ceph_assert(diri)
Patrick Donnelly
02:26 AM Bug #52626 (Triaged): mds: ScrubStack.cc: 831: FAILED ceph_assert(diri)
... Patrick Donnelly
02:21 AM Bug #52625 (Fix Under Review): qa: test_kill_mdstable (tasks.cephfs.test_snapshots.TestSnapshots)
Patrick Donnelly
02:20 AM Bug #52625 (Resolved): qa: test_kill_mdstable (tasks.cephfs.test_snapshots.TestSnapshots)
... Patrick Donnelly

09/15/2021

02:16 PM Bug #52280: Mds crash and fails with assert on prepare_new_inode
@xiubo Li
Thanks much, Li, for your analysis
I didnt change mds_log_segment_size and mds_log_segment_size.
To what...
Yael Azulay
05:34 AM Bug #52280: Mds crash and fails with assert on prepare_new_inode
Hi Yael,
BTW, have you ever changed the "mds_log_segment_size" and "mds_log_events_per_segment" options ?
If no...
Xiubo Li

09/14/2021

05:22 PM Bug #51866 (Can't reproduce): mds daemon damaged after outage
> rados_osd_op_timeout: 30
yeah that'll do it. thanks for clarifying the root cause!
Dan van der Ster
04:45 PM Bug #51866: mds daemon damaged after outage
...and I've just found that we're globally setting:
```
rados_osd_op_timeout: 30
rados_mon_op_timeout: 30
```
...
Alex Kershaw
04:11 PM Bug #51866: mds daemon damaged after outage
Hi again - another update on this one.
We deploy on both OpenStack and VMware hosted VMs, with the intention that ...
Alex Kershaw
05:07 PM Bug #52572 (Fix Under Review): "cluster [WRN] 1 slow requests" in smoke pacific
Patrick Donnelly
05:00 PM Bug #52572 (In Progress): "cluster [WRN] 1 slow requests" in smoke pacific
Patrick Donnelly
03:03 PM Bug #52607 (Duplicate): qa: "mon.a (mon.0) 1022 : cluster [WRN] Health check failed: Reduced data...
/ceph/teuthology-archive/yuriw-2021-09-13_14:09:59-fs-pacific-distro-basic-smithi/6387141/teuthology.log Patrick Donnelly
02:59 PM Bug #52606 (Fix Under Review): qa: test_dirfrag_limit
Patrick Donnelly
02:55 PM Bug #52606 (Resolved): qa: test_dirfrag_limit
... Patrick Donnelly
10:52 AM Bug #52382 (Fix Under Review): mds,client: add flag to MClientSession for reject reason
Kotresh Hiremath Ravishankar

09/13/2021

03:05 PM Fix #52591 (Resolved): mds: mds_oft_prefetch_dirfrags = false is not qa tested
We'd like to use `mds_oft_prefetch_dirfrags = false` in production; in our tests it demonstrates a massive speedup wh... Dan van der Ster
01:43 PM Bug #52581 (Triaged): Dangling fs snapshots on data pool after change of directory layout
Patrick Donnelly
09:34 AM Bug #52581: Dangling fs snapshots on data pool after change of directory layout
Related ceph-users thread, relevant part towards the end: https://lists.ceph.io/hyperkitty/list/ceph-users@ceph.io/th... Frank Schilder
09:31 AM Bug #52581 (New): Dangling fs snapshots on data pool after change of directory layout
... Frank Schilder
01:14 PM Backport #52441 (In Progress): pacific: mds: slow performance on parallel rm operations for multi...
Xiubo Li
05:44 AM Backport #52441: pacific: mds: slow performance on parallel rm operations for multiple kclients
I will work on this. Xiubo Li
09:26 AM Bug #52280: Mds crash and fails with assert on prepare_new_inode
I found some memories are not counted into the MDCache memories, such as the inode_map map in the MDCache class. Xiubo Li
09:17 AM Bug #52280: Mds crash and fails with assert on prepare_new_inode
Hi Yael,
BTW, did you see any memory killer logs in the /var/log/message* log files ? It should be OS's the memori...
Xiubo Li
08:58 AM Bug #52280: Mds crash and fails with assert on prepare_new_inode
Yael Azulay wrote:
> Xiubo Li wrote:
> > Hi Yael,
> >
> > Do you have the mds side logs ? Thanks.
>
> Hi Xiub...
Xiubo Li
06:54 AM Backport #51937 (Resolved): pacific: qa: test_full_fsync (tasks.cephfs.test_full.TestClusterFull)...
Xiubo Li

09/10/2021

06:18 PM Bug #52572: "cluster [WRN] 1 slow requests" in smoke pacific

(08:28:21 AM) neha: yuriw: looks that test has been failing since July due to slow requests on the mds https://pulp...
Yuri Weinstein
06:18 PM Bug #52572 (Resolved): "cluster [WRN] 1 slow requests" in smoke pacific
Run: https://pulpito.ceph.com/yuriw-2021-09-10_14:17:02-smoke-pacific-distro-basic-smithi/
Run: 6382881
Logs: http:...
Yuri Weinstein
10:00 AM Bug #50719: xattr returning from the dead (sic!)
Ralph Böhme wrote:
> I'm going to work on trying to reproduce this with a local simple test program in the next days...
Ralph Böhme
01:48 AM Support #52551: ceph osd pool set-quota
Patrick Donnelly wrote:
> These kinds of questions should go on ceph-users mailing list, please.
I'm sorry for that.
le le
12:37 AM Bug #52565 (Fix Under Review): MDSMonitor: handle damaged state from standby-replay
Patrick Donnelly

09/09/2021

11:37 PM Bug #52565 (Resolved): MDSMonitor: handle damaged state from standby-replay
After the addition of join_fscid, the state change validation in this code:
https://github.com/ceph/ceph/blob/ac46...
Patrick Donnelly
01:00 PM Support #52551 (Rejected): ceph osd pool set-quota
These kinds of questions should go on ceph-users mailing list, please. Patrick Donnelly
02:41 AM Support #52551 (Rejected): ceph osd pool set-quota
hi~, I ask for some question about "df -h" after "ceph osd pool set-quota"
CentOS Linux release 8.4.2105
ceph ver...
le le
05:35 AM Bug #52260: 1 MDSs are read only | pacific 16.2.5
We didn't change the mclock profile to custom, it is set to high io client.
I am not sure the op_queue is the prob...
cephuser2345 user
03:07 AM Bug #52508: nfs-ganesha crash when calls libcephfs, it triggers __ceph_assert_fail
Patrick Donnelly wrote:
> le le wrote:
> > The exception because of compiler’s optimization ?
>
> Probably there...
le le

09/08/2021

09:54 PM Backport #51833: pacific: client: flush the mdlog in unsafe requests' relevant and auth MDSes only
Backport Bot wrote:
> https://github.com/ceph/ceph/pull/42925
merged
Yuri Weinstein
09:53 PM Backport #51937: pacific: qa: test_full_fsync (tasks.cephfs.test_full.TestClusterFull) failed
Backport Bot wrote:
> https://github.com/ceph/ceph/pull/42923
merged
Yuri Weinstein
07:54 PM Bug #52280: Mds crash and fails with assert on prepare_new_inode
Xiubo Li wrote:
> Hi Yael,
>
> Do you have the mds side logs ? Thanks.
Hi Xiubo
We reinstalled the setup and...
Yael Azulay
08:56 AM Bug #52280: Mds crash and fails with assert on prepare_new_inode
Hi Yael,
Do you have the mds side logs ? Thanks.
Xiubo Li
02:49 PM Bug #51866: mds daemon damaged after outage
I should mention, I've only had this reproduce once after we added the fix, when previously it was 100% reproducible,... Alex Kershaw
02:10 PM Bug #51866: mds daemon damaged after outage
Hi, I'm working with David and am now looking at this issue. Unfortunately it seems the fix hasn't worked.
I've ju...
Alex Kershaw
02:38 PM Bug #52531 (Triaged): Quotas smaller than 4MB on subdirs do not have any effect
Patrick Donnelly
12:41 PM Bug #52508: nfs-ganesha crash when calls libcephfs, it triggers __ceph_assert_fail
le le wrote:
> The exception because of compiler’s optimization ?
Probably there is a race condition not protecte...
Patrick Donnelly
08:28 AM Bug #51756: crash: std::_Rb_tree_insert_and_rebalance(bool, std::_Rb_tree_node_base*, std::_Rb_tr...
For context, we hit this nearly every time we stop an active MDS in 14.2.20. (We see the #52207 variant of this). Dan van der Ster
06:34 AM Bug #43960: MDS: incorrectly issues Fc for new opens when there is an existing writer
Could we use the socket keepalive to detect the socket connection peer's aliveness in the caps_tick() if there has re... Xiubo Li
06:16 AM Bug #43960: MDS: incorrectly issues Fc for new opens when there is an existing writer
Whenever switching to a different lock state the MDS will try to issue the allowed caps to the clients, even some cap... Xiubo Li

09/07/2021

08:02 PM Backport #51977: pacific: client: make sure only to update dir dist from auth mds
Backport Bot wrote:
> https://github.com/ceph/ceph/pull/42937
merged
Yuri Weinstein
05:11 PM Bug #52260: 1 MDSs are read only | pacific 16.2.5
cephuser2345 user wrote:
> 1. we chaned osd_opqueue to mclock_scheduler right after upgrading to pacific. that was o...
Sridhar Seshasayee
04:38 PM Bug #52531 (Triaged): Quotas smaller than 4MB on subdirs do not have any effect
This doesn’t work:
root@mon0:~# setfattr -n ceph.quota.max_bytes -v $((4*1024*1024-1)) /mnt/cephfs/
root@mon0:~# ge...
Ruben Kerkhof
03:24 PM Bug #51757 (New): crash: /lib/x86_64-linux-gnu/libpthread.so.0(
Not sure why this tracker is marked as a duplicate. Moving back to new. Venky Shankar
02:02 PM Feature #52491 (Fix Under Review): mds: add max_mds_entries_per_dir config option
Patrick Donnelly
01:49 PM Feature #52459 (Need More Info): mds: add failed connections warning
xinyu wang wrote:
> Greg Farnum wrote:
> > What specific scenario are you trying to avoid here? Messenger-level fai...
Patrick Donnelly
01:46 PM Bug #52487 (Triaged): qa: Test failure: test_deep_split (tasks.cephfs.test_fragment.TestFragmenta...
Patrick Donnelly
01:45 PM Bug #52508 (Triaged): nfs-ganesha crash when calls libcephfs, it triggers __ceph_assert_fail
Patrick Donnelly
 

Also available in: Atom