Activity
From 03/10/2014 to 04/08/2014
04/08/2014
- 08:25 PM Bug #8004: LibCephFS.HardlinkNoOriginal hang
- 12:12 PM Bug #8026 (Resolved): shared pointer completely break multiple mds
- 07:04 AM Bug #8026: shared pointer completely break multiple mds
- 12:04 AM Bug #8026 (Resolved): shared pointer completely break multiple mds
- 06:59 AM Bug #3424: java: Add the correct JUnit package dependencies on supported platforms and ensure the...
- 06:51 AM Bug #8025: nfs-on-kclient: rm -r failed
04/07/2014
- 10:08 PM Bug #8025 (Resolved): nfs-on-kclient: rm -r failed
- teuthology-2014-04-06_23:01:11-knfs-master-testing-basic-plana/175859/...
- 09:26 PM Bug #7958 (Resolved): ceph-fuse+fsx umount hang on leaked inode reference
- 09:17 PM Feature #7319 (Resolved): qa: multimds, no failure
- 09:13 PM Bug #7739 (Resolved): mds: uninitialized field in message
- 04:18 PM Bug #8005 (Rejected): fuse hang
- no error in client log, looks like mds was killed by someone
- 12:57 PM Bug #8010: It's impossible to remove unused filesystem pools from a cluster
- It's happening with
ceph version 0.78 (f6c746c314d7b87b8419b6e584c94bfe4511dbd4)
Linux access.car.dot.com 3.13.0-... - 10:07 AM Bug #8010 (Resolved): It's impossible to remove unused filesystem pools from a cluster
- We've inadvertently made it impossible to remove a filesystem from a Ceph cluster. If there is not data in the FS, it...
04/06/2014
- 09:22 PM Bug #8005: fuse hang
- still looks like MDS was dead
- 01:44 PM Bug #8005 (Rejected): fuse hang
- ...
- 05:40 PM Bug #8006 (Rejected): fuse hang on flush (icache branch)
- The flush hang is because ceph-fuse was umounting (received signal). umounting can't finish becase MDS was dead at th...
- 01:49 PM Bug #8006 (Rejected): fuse hang on flush (icache branch)
- ...
- 05:29 PM Bug #8004: LibCephFS.HardlinkNoOriginal hang
- oh, and the 32-bit pointer thing is because ceph-fuse is running under valgrind.
- 05:27 PM Bug #8004: LibCephFS.HardlinkNoOriginal hang
- seems easy to reproduce, just hit this again with...
- 01:41 PM Bug #8004 (Resolved): LibCephFS.HardlinkNoOriginal hang
- ubuntu@teuthology:/var/lib/teuthworker/archive/sage-2014-04-05_15:44:13-multimds:verify-wip-ms-dump-testing-basic-pla...
- 01:22 PM Bug #7739: mds: uninitialized field in message
04/04/2014
- 07:28 PM Bug #7980: 0.78: MDS crash (segmentation fault) on client wake-up from suspend.
- Works as expected, problem solved, thank you.
- 12:18 AM Bug #7980: 0.78: MDS crash (segmentation fault) on client wake-up from suspend.
- Very nice, thank you. I'll test and confirm.
- 12:01 AM Bug #7980 (Resolved): 0.78: MDS crash (segmentation fault) on client wake-up from suspend.
- fixed by https://github.com/ceph/ceph/commit/fb72330fb3514be690dc60598242036aa560e023
04/03/2014
- 11:47 PM Bug #7980 (Resolved): 0.78: MDS crash (segmentation fault) on client wake-up from suspend.
- MDS crashes (segmentation fault) when I wake-up machine with CephFS (mounted using kernel client) from suspend:
<p... - 06:53 AM Bug #7958: ceph-fuse+fsx umount hang on leaked inode reference
- 05:32 AM Bug #7958: ceph-fuse+fsx umount hang on leaked inode reference
- I guess it's introduce by commit f1c7b4ef0 (client: pin Inode during readahead). Readahead raced with truncate. Objec...
04/02/2014
- 10:00 PM Bug #7958: ceph-fuse+fsx umount hang on leaked inode reference
- ubuntu@teuthology:/var/lib/teuthworker/archive/teuthology-2014-04-01_23:00:33-fs-firefly-distro-basic-plana/160589
- 10:23 AM Bug #7958 (Resolved): ceph-fuse+fsx umount hang on leaked inode reference
- ubuntu@teuthology:/var/lib/teuthworker/archive/teuthology-2014-03-31_23:00:36-fs-master-testing-basic-plana/157636
... - 04:12 PM Bug #7966 (Resolved): ceph-mds respawn doesn't always work
- ...
03/29/2014
03/28/2014
- 03:56 PM Bug #7867 (Resolved): client/Client.cc: 2087: FAILED assert(!unclean)
- 03:47 PM Bug #7867 (Pending Backport): client/Client.cc: 2087: FAILED assert(!unclean)
- 08:25 AM Bug #7780 (Resolved): When full flag is set, even MDS writes are blocked
- Fix was merged at c647a03fffb2e1e997dbdb0ff128eeb6efc47deb
03/27/2014
- 10:26 PM Bug #7880: multimds: directory gets rsynced twice
- 06:54 AM Bug #7880 (Resolved): multimds: directory gets rsynced twice
- probalby the mtime doesn't get set properly teh first time?...
- 10:40 AM Bug #7867: client/Client.cc: 2087: FAILED assert(!unclean)
03/26/2014
- 07:00 PM Bug #5787 (Duplicate): client/Client.cc: 2081: FAILED assert(!unclean) in put_inode
- 07:00 PM Bug #5787: client/Client.cc: 2081: FAILED assert(!unclean) in put_inode
- dup #7867
- 05:57 PM Bug #7867: client/Client.cc: 2087: FAILED assert(!unclean)
- pushed a simpler version, wip-7867-b, that just pins the Inode* for the duration of the readahead. This avoids a pos...
- 09:13 AM Bug #7867 (Resolved): client/Client.cc: 2087: FAILED assert(!unclean)
- here is the lsat bit of log, starting with the incomplete readahead that caused the crash...
03/25/2014
- 10:50 AM Feature #3863 (Resolved): implement a tool to lookup inode numbers without holding their path
- Merged into master in commit:83661c273e759837816d5cd7ba27233ded898455
- 06:46 AM Feature #7761 (In Progress): journal-tool: forwards-search through corrupt regions
- 06:45 AM Feature #7759 (In Progress): journal-tool: roll in resetter/dumper from MDS
03/21/2014
- 06:46 AM Feature #7810 (Resolved): libcephfs: add a test that freezes + unfreezes a client, and then verif...
03/19/2014
- 08:28 AM Bug #7780 (Fix Under Review): When full flag is set, even MDS writes are blocked
- 05:12 AM Bug #7780: When full flag is set, even MDS writes are blocked
- Just tried commenting out the check in objecter and the MDS writes get through, so not too much work to do here.
- 04:05 AM Bug #7780 (Resolved): When full flag is set, even MDS writes are blocked
We're meant to allow the MDS to write even when the OSDs think they're full, so that we can journal file deletions ...
03/18/2014
- 09:28 PM Bug #7708 (Resolved): mds: null deref in issue_caps
- 05:28 AM Feature #7764 (Resolved): InoTable/SessionMap/ manipulator (cephfs-table-tool)
Some bugs might create a situation where the metadata/journal may be okay but recovery fails because one of the glo...- 05:25 AM Feature #7763 (Resolved): journal-tool: import
- Given a binary journal dump, insert it back into the metadata pool.
- 05:23 AM Feature #7762 (Rejected): journal-tool: backwards-search after corrupt regions
- When a corrupt region is encountered, use the start pointers to walk from the end of the log back to the latest valid...
- 05:22 AM Feature #7352: mds: make classes encode/decode-able
- related, on journal-tool branch have got to a point where dencoder can handle getting the inodes out of .inode object...
- 05:20 AM Feature #7761 (Resolved): journal-tool: forwards-search through corrupt regions
When corruption is encountered, search forward byte by byte for the next sentinel word.- 05:18 AM Feature #7760 (Resolved): journal-tool: implement splice
Simple implementation based on new ENoOp event type that allows us to selectively blank out regions in existing log.- 05:17 AM Feature #7759 (Resolved): journal-tool: roll in resetter/dumper from MDS
To provide the capabilities in one convenient place and remove a few LOC from the MDS.- 05:15 AM Feature #7758 (Resolved): journal-tool: complete filtering
- Already done:
--path
--range
Remaining to do:
--by-inode <inodeno>
--by-dirfrag <frag id>:[dentry]
--by-...
03/17/2014
- 01:04 PM Bug #7750 (Can't reproduce): Attempting to mount a kNFS export of a sub-directory of a CephFS fil...
- Hi,
I have run into the following issue when trying to export a subset of a CephFS filesystem via NFS:
Machine ...
03/15/2014
- 07:39 PM Bug #7684 (Resolved): failed cfuse_workunit_kernel_untar_build.yaml test
- 11:15 AM Bug #7739 (Resolved): mds: uninitialized field in message
- ubuntu@teuthology:/a/teuthology-2014-03-14_23:01:36-multimds-master-testing-basic-plana/131675
hard to map this to...
03/14/2014
03/13/2014
- 09:01 PM Bug #7708 (Resolved): mds: null deref in issue_caps
- ...
- 08:28 PM Bug #7684: failed cfuse_workunit_kernel_untar_build.yaml test
- ...
- 11:01 AM Bug #7684: failed cfuse_workunit_kernel_untar_build.yaml test
- And again:
http://qa-proxy.ceph.com/teuthology/teuthology-2014-03-12_23:00:26-fs-firefly-testing-basic-plana/127977/... - 03:27 PM Bug #7565 (Resolved): Failed assert in check_rstats
- 06:53 AM Bug #7565: Failed assert in check_rstats
- ubuntu@teuthology:/a/teuthology-2014-03-12_04:55:11-multimds-master-testing-basic-plana/127691
failed on trivial_s...
03/12/2014
- 01:13 AM Bug #1318: directories disappear across multiple rsyncs
- please check if there was file recovery involved (MDCache::do_file_recover )
03/11/2014
- 01:57 PM Bug #7685 (Can't reproduce): hung/failed teuthology test: cfuse_workunit_misc
- http://qa-proxy.ceph.com/teuthology/teuthology-2014-03-07_23:00:50-fs-firefly-testing-basic-plana/122094
http://qa-p... - 01:46 PM Bug #7684 (Resolved): failed cfuse_workunit_kernel_untar_build.yaml test
- http://qa-proxy.ceph.com/teuthology/teuthology-2014-03-09_23:00:25-fs-firefly-testing-basic-plana/124157/
The teut... - 07:18 AM Bug #7474 (Won't Fix): Kernel oops with cephfs [ceph_write_begin -> *x8 -> wait_on_page_read]
- This looks like it's the writeback deadlock when trying to flush from the client to the OSD on a single memory-constr...
- 07:17 AM Bug #6599 (Resolved): client: invalid iterator dereference in Client::trim_caps
- 07:12 AM Feature #5486: kclient: make it work with selinux
- Hmm, Sage notes that maybe it'll work now we support ACLs. Or maybe we can use a special mount option?
- 07:00 AM Bug #2187 (Can't reproduce): pjd chown/00.t failed test 97
- 06:59 AM Bug #2740 (Resolved): mds: crash in Objecter when shutting down too early
03/10/2014
- 12:58 PM Bug #1318: directories disappear across multiple rsyncs
- By “this” I meant files with different timestamps from what they were last set to, as in the first paragraph of comme...
- 12:51 PM Bug #1318: directories disappear across multiple rsyncs
- I'm afraid this still occurs quite often with ceph 0.77 and ceph.ko 3.13.6-gnu. I have a slightly better understandi...
Also available in: Atom