Activity
From 07/25/2012 to 08/23/2012
08/23/2012
- 05:59 PM rgw Feature #2797: rgw: support multi-objects delete
- 05:58 PM rgw Feature #2839: rgw: garbage collection
- 05:58 PM rgw Feature #3037 (Resolved): rgw: unit test for rgw objclass
- 04:10 PM Feature #2829 (Resolved): report on cluster size/status (for service billing purposes)
- at some point we need the receiving end of this: extract the json, validate the crc, and stick it in some database or...
- 03:55 PM Feature #2477 (Resolved): rados bench cleanup
- 01:16 PM rbd Bug #2948 (Resolved): rbd: fails to close image on error
- commit:fed8aea662bf919f35a5a72e4e2a2a685af2b2ed in master
- 12:59 PM Feature #2840 (Resolved): mon: $mon_data/cluster_fsid file
- 10:12 AM Linux kernel client Bug #3031 (Resolved): btrfs: lock returned to userspace
- [19490.018682]
[19490.038366] ================================================
[19490.063495] [ BUG: lock held whe... - 07:27 AM Subtask #2745: mon: Single-Paxos: Sync: Add new message support to the Monitor class
- Currently, most timeout callbacks simply assert. This has been allowing us to successfully debug some unforeseen situ...
- 07:13 AM Subtask #2757: mon: Single-Paxos: Sync: pack chunks of the MonitorDBStore into transactions
- Must still test how it behaves when we are only interested in synchronizing part of the store.
- 07:11 AM Subtask #2744 (Resolved): mon: Single-Paxos: Sync: Create new Message type
- Only thing missing: adjusting the commit message to fully describe the message in detail.
08/22/2012
- 01:57 PM Bug #2784: osd hit suicide timeout
- This test hung in the nightlies.
Logs: ubuntu@teuthology:/a/teuthology-2012-08-22_00:00:07-regression-next-testing... - 01:51 PM Bug #3030: config/option parser: Avoid needing to list command line options in a global config list
- Another example: daemonize.
- 01:46 PM Bug #3030 (Won't Fix): config/option parser: Avoid needing to list command line options in a glob...
- Having "monmap" in config_opts, when it's only really used by ceph-osd --mkfs, is pretty confusing. This should be be...
- 01:44 PM Bug #3029 (Won't Fix): config/option parser: Avoid needing to list obscure one-use options in glo...
- num_client is only used by ceph-syn, but still needs to be listed in the config_opts list, which a horribly generic n...
- 11:40 AM rgw Documentation #2991: doc: expand/complete RGW Swift API reference
- Sorry. Previous update intended for RGW config. This is checked in. Location is: ceph/doc/radosgw/swift. Accessible v...
- 11:34 AM rgw Documentation #2991 (In Progress): doc: expand/complete RGW Swift API reference
- Yehuda needs to review the doc and sign off. Updated doc sent via email. Current location is ceph/doc/radosgw/config-...
- 10:52 AM Bug #2947 (Resolved): osd: out of order reply
- commit:1113a6c56739a56871f01fa13da881dab36a32c4
08/21/2012
- 10:49 PM CephFS Bug #1947: mds: SIGBUS during _mark_dirty
- logs: ubuntu@teuthology:/a/teuthology-2012-08-21_02:00:04-regression-testing-testing-basic/5691
- 10:57 AM CephFS Bug #1947: mds: SIGBUS during _mark_dirty
- added debugging to kernel ffsb task
- 06:34 PM Bug #2947 (In Progress): osd: out of order reply
- ooof, the saga continues: ubuntu@teuthology:/a/sage-gfoo2/5974
- 10:50 AM Bug #2947 (Resolved): osd: out of order reply
- commit:4a0704e64a733b7bb14fb4103cd1cd54e4e7da8a
- 06:03 PM Bug #2954: osd: scrub stat mismatch, got 18/19 objects, 14/15 clones, 22478527/25385282 bytes.
- another one. ms failure injection may have contributed.
ubuntu@teuthology:/a/sage-gfoo2/5925 - 05:43 PM Bug #3026 (Resolved): ref counting error argonaut
- (11:36:55 AM) Sage Weil: -1> 2012-08-21 07:00:24.285153 7ff5abba6700 1 -- 10.214.131.24:6806/20124 --> 10.214.13...
- 05:42 PM Bug #3025 (Resolved): WaitActingChange
- We should not transition to WaitActingChange from Acting due to recovery complete.
- 05:32 PM rbd Feature #2720: rbd: add children command
- First implementation from Josh has edges sanded off, sorta running. Needs better testing and manpage updates.
- 05:11 PM Feature #1515 (Duplicate): osd: pg split
- 04:17 PM rbd Feature #2560: rbd: safe parent deletion
- I *think* this is more or less implemented. The commands are "snap protect" and
"snap unprotect", but they behave a... - 03:46 PM RADOS Feature #3011 (Fix Under Review): Remove "pool" terminology from CRUSH
- 09:38 AM RADOS Feature #3011: Remove "pool" terminology from CRUSH
- agreed.
i'll stick this in the backlog! - 09:32 AM RADOS Feature #3011: Remove "pool" terminology from CRUSH
- Since it's a hierarchy of nodes, I'd vote for "root." Also, the term "bucket" is confusing, because we use that term ...
- 08:15 AM RADOS Feature #3011: Remove "pool" terminology from CRUSH
- You're talking about 'pool=default', right? I agree. What term should be use instead for the root of the tree?
'... - 02:38 PM devops Feature #3023 (Closed): juju: automated QA of OpenStack RBD integration
- 02:38 PM devops Feature #3022 (Closed): juju: automated QA of Ceph
- 02:36 PM devops Feature #3021 (Closed): juju: change glance to use rbd
- 02:36 PM devops Feature #3020 (Closed): juju: change nova to use rbd
- 02:36 PM devops Feature #3019 (Closed): juju: modernize ceph charm, mon & osd bootstrap
- 02:35 PM devops Feature #3018 (Closed): juju: test deploy of openstack
- 02:35 PM devops Feature #3017 (Closed): juju: dev env setup
- 02:13 PM CephFS Bug #2863: client: does not tolerate traceless replies from mds
- 02:13 PM CephFS Bug #2863 (Resolved): client: does not tolerate traceless replies from mds
- 02:10 PM Feature #2829 (Fix Under Review): report on cluster size/status (for service billing purposes)
- 01:46 PM Feature #2829: report on cluster size/status (for service billing purposes)
- see wip-mon-report
- 01:31 PM Cleanup #3016 (Resolved): make ceph osd crush set ${id} osd.${id} not require the ID twice
- That is lame and confusing.
- 01:01 PM CephFS Bug #1945: blogbench hang on caps
- ubuntu@teuthology:/var/lib/teuthworker/archive/teuthology-2012-08-21_02:00:04-regression-testing-testing-basic/5675
- 12:45 PM devops Cleanup #3015 (Resolved): order of arguments should not matter for init-ceph
- -c ceph.conf start works
start -c ceph.conf does not.
Boo! - 12:45 PM Bug #3014 (Resolved): ceph mds set_data_pool pool doesn't fail
- If you specify a name instead of a pool ID, it just adds pool id 0!
- 12:27 PM Bug #2762: mon crash ceph::__ceph_assert_fail (assertion=0x63d150 "begin->last_committed == last_...
- May be clearly reproduced with >500 active clients, e.g. booting vms, and one monitor.
- 11:21 AM RADOS Bug #3013 (New): doc: Document ceph-osd --mkfs --osd-uuid, --get-osd-uuid, and friends
- ceph-osd --mkfs --osd-uuid <uuid> -i 123 ...
--get-osd-fsid and --get-cluster-fsid
Go through the source and lo... - 11:02 AM rgw Bug #2961 (Resolved): rgw: bad content range
- 08:04 AM rgw Bug #2961: rgw: bad content range
- it only with >4G objects. A test like that would just take too long. Maybe it's possible to put it as an optional tes...
- 10:14 AM Feature #2668 (Resolved): Build linux-tools-common package for perf
- 09:55 AM RADOS Feature #3012 (New): come up with some way to do gossip among daemons on a host
- In discussion, it occurred to me that really OSDs on a host ought to gossip about certain kinds of information (altho...
- 09:07 AM Feature #3010 (In Progress): Make it easy to find a list of data locations from a cephfs file
- is this what they're after?...
- 08:50 AM Bug #3005 (Resolved): bootstrapped mon crashes after win_standalone_election
- logs on #3006
also reproduced w/ vstart by doing 'ceph log foo &' every .01 seconds in a loop, and then removing m... - 08:50 AM Bug #3006 (Duplicate): mon: removing a running monitor can crash ceph
- see #3005
08/20/2012
- 09:42 PM rbd Bug #2937: btrfs filesystem on rbd device kernel BUG writing large file
- This reproduces on plana. Details: two machine cluster, one monitor, two OSDs:
roles:
- [mon.0, osd.0]
- [osd.1... - 09:28 PM RADOS Feature #3011 (Resolved): Remove "pool" terminology from CRUSH
- Users get confused and conflate RADOS pools and CRUSH pools. I don't think we actually use that term in many places i...
- 08:58 PM Feature #3010 (Resolved): Make it easy to find a list of data locations from a cephfs file
- Large cluster designers would like to be able to get as much information about a CephFS file's location as possible. ...
- 07:12 PM Bug #3009 (Resolved): if you mkfs an OSD with --filestore-xattr-use-omap and then don't start the...
- Apparently we auto-detect filestore-xattr-use-omap, but we don't store it anywhere in the OSD's data directory. Which...
- 06:57 PM RADOS Cleanup #3008 (New): Consider making MLog messages not require MON_CAP_X
- Right now, the permissions for an incoming MLog are checked against PAXOS_LOG, MON_CAP_X. This means that the MDS and...
- 06:01 PM Bug #3006 (Duplicate): mon: removing a running monitor can crash ceph
- While rewriting the ceph add/remove monitor documentation (http://ceph.com/docs/master/ops/manage/grow/mon/), I added...
- 05:18 PM Bug #3005 (Resolved): bootstrapped mon crashes after win_standalone_election
- I created the mon from #3004 and got it running correctly. It crashed since it won without being rank 0....
- 05:17 PM Bug #3001 (Resolved): mkcephfs: -a fails if only "host=localhost" sections seen in ceph.conf
- ...
- 04:22 PM Bug #3001: mkcephfs: -a fails if only "host=localhost" sections seen in ceph.conf
- And for the record, for now I'm recommending this: don't use "host=localhost", put in the actual host name.
- 04:21 PM Bug #3001 (Resolved): mkcephfs: -a fails if only "host=localhost" sections seen in ceph.conf
- This was reported earlier on the list as http://thread.gmane.org/gmane.comp.file-systems.ceph.devel/8051/focus=8092 a...
- 05:13 PM Bug #3004 (Resolved): bootstrapped initial monitor can't find its own keyring with relative paths
- I ran the following sequence of commands, which I sourced from vstart (while extracting the ceph.conf):...
- 05:05 PM rbd Documentation #2992 (In Progress): doc: RBD parent/child snapshot
- 11:31 AM rbd Documentation #2992 (Need More Info): doc: RBD parent/child snapshot
- 04:54 PM Bug #3002 (Resolved): ceph-authtool: --print does not work
- this already got fixed in master, it looks like (--print-key instead of --print). don't think it's worth backporting...
- 04:42 PM Bug #3002 (Resolved): ceph-authtool: --print does not work
- ...
- 04:53 PM Bug #3003 (Resolved): mon: race/crash after removing monitors
- commit:d521dde9b565098765a20dd001d8650ad02c2bef
- 04:47 PM Bug #3003 (Resolved): mon: race/crash after removing monitors
- ...
- 03:51 PM Bug #2691 (In Progress): osd/ReplicatedPG.cc: 5888: FAILED assert(latest->is_update())
- Recent log: ubuntu@teuthology:/a/teuthology-2012-08-20_00:00:04-regression-next-testing-basic/4822...
- 03:36 PM Feature #3000 (Resolved): osd: balance recovery vs client io
- 03:35 PM Linux kernel client Bug #1347 (Can't reproduce): forced unmount kernel bug
- 03:34 PM Bug #2451 (Can't reproduce): qa: networking doesn't always start after reboot
- i havne't seen this in a long time.
- 03:26 PM rgw Bug #2961 (In Progress): rgw: bad content range
- Can we add an s3tests for this?
- 03:26 PM rgw Bug #2961 (Resolved): rgw: bad content range
- 03:16 PM Feature #2668 (In Progress): Build linux-tools-common package for perf
- 03:06 PM Bug #2999 (Resolved): osd: msgr crash in OSD::complete_notify
- Logs: ubuntu@teuthology:/a/teuthology-2012-08-17_19:00:07-regression-master-testing-gcov/3549...
- 03:05 PM Bug #2956 (Resolved): osd:FAILED assert(waiting_for_ondisk.begin()->first == repop->v)
- commit:dd4c1dc9f9dae43e4761caca049bfe7361d9ebfb
- 12:35 PM Bug #2956: osd:FAILED assert(waiting_for_ondisk.begin()->first == repop->v)
- 11:17 AM Bug #2956: osd:FAILED assert(waiting_for_ondisk.begin()->first == repop->v)
- ...
- 02:28 PM Documentation #2998 (Can't reproduce): doc: validate install docs on ubuntu server
- 02:00 PM RADOS Bug #2874: apparent CRUSH mapping failure
- I'd like to report that I was seeing what I believe to be the same issue (at least the symptoms were the same: a 3-OS...
- 01:58 PM Bug #2761: osd: failed to recover before timeout expired
- just reproduced this one with osd and msgr logs:...
- 12:56 PM Bug #2761: osd: failed to recover before timeout expired
- Logs: ubuntu@teuthology:/a/teuthology-2012-08-20_04:00:05-regression-stable-master-basic/5044
- 01:31 PM Feature #2911 (Duplicate): osd: Restrict recovery when the OSD full list is nonempty
- 01:31 PM Feature #1637 (Duplicate): OSDs running full take down other OSDs
- 01:09 PM Bug #2924 (Resolved): doc: Adjust for mon. key being in external keyring
- ceph auth get mon. > /tmp/monkey
- 01:04 PM Bug #2947: osd: out of order reply
- More Logs: ubuntu@teuthology:/a/teuthology-2012-08-19_02:00:05-regression-testing-testing-basic/4288
- 11:38 AM Bug #2947: osd: out of order reply
- ...
- 12:35 PM rbd Bug #2967 (Resolved): librbd: cls_rbd.parents unit test failure
- 11:36 AM rbd Bug #2967: librbd: cls_rbd.parents unit test failure
- 10:59 AM rbd Bug #2967: librbd: cls_rbd.parents unit test failure
- I think this is resolved by the about-to-be-merged layering code; testing is in progress
- 11:53 AM Bug #2997 (Resolved): ceph-mon --mkfs allows you to create one without an id which then crashes o...
- And that sucks, especially when it crashes in a demo and you don't know why.
- 11:41 AM rgw Documentation #2483 (Fix Under Review): doc: radosgw api diffs to swift
- 11:40 AM rgw Documentation #2483: doc: radosgw api diffs to swift
- Can you check the latest master build of docs and see if this has been updated to your satisfaction? Thanks!
- 11:30 AM Documentation #2978 (Need More Info): doc: write RADOS restore from backup procedure
- 11:30 AM Documentation #2977 (Need More Info): doc: write RADOS backup procedure
- 11:30 AM Documentation #2979 (Need More Info): doc: write doc on how to use / rollback to RADOS snapshots
- 11:30 AM devops Documentation #2975 (Need More Info): doc: update docs to match new ceph-disk-prepare syntax
- 11:29 AM Documentation #2995 (In Progress): doc: restructure documentation (its getting messy!)
- 11:23 AM Documentation #2981 (In Progress): doc: write add/remove a monitor
- 11:22 AM Documentation #2970 (In Progress): doc: expand/complete osd settings reference
- 11:22 AM Documentation #2971 (In Progress): doc: expand/complete mon settings reference
- 11:22 AM Documentation #2973 (In Progress): doc: expand/complete ceph general settings
- 10:56 AM Feature #2840: mon: $mon_data/cluster_fsid file
- wip-mon-mkfs
- 10:55 AM Feature #2840 (Fix Under Review): mon: $mon_data/cluster_fsid file
- 09:22 AM Bug #2803 (Can't reproduce): filer: probe crash
- 09:21 AM CephFS Bug #2959 (Resolved): mds: returns null dentry on getattr
- 09:20 AM Linux kernel client Bug #2936 (Resolved): Remounting cephfs with non-existing path causes kernel panic
08/19/2012
- 04:10 PM Documentation #2996 (Resolved): doc: write install Ceph with RPMs doc
- 04:09 PM Documentation #2995 (Resolved): doc: restructure documentation (its getting messy!)
- 04:07 PM Documentation #2994 (Resolved): doc: expand/complete librados API doc
- 04:04 PM rgw Documentation #2993 (Resolved): doc: write quick RGW guide (if feasible)
- 04:03 PM rbd Documentation #2992 (Resolved): doc: RBD parent/child snapshot
- 03:59 PM rgw Documentation #2991 (Resolved): doc: expand/complete RGW Swift API reference
- The reference for the [client.radosgw.gateway] sections of ceph.conf need to be completed by John Wilkins and reviewe...
- 03:58 PM rgw Documentation #2990 (Resolved): doc: expand/complete RGW S3 API reference
- 03:57 PM rgw Documentation #2989 (Resolved): doc: write RGW troubleshooting
- 03:57 PM CephFS Documentation #2988 (Resolved): doc: write MDS troubleshooting
- 03:57 PM Documentation #2987 (Rejected): doc: write MON troubleshooting
- 03:57 PM Documentation #2986 (Rejected): doc: write OSD troubleshooting
- 03:56 PM Documentation #2985 (Rejected): doc: write install troubleshooting
- 03:56 PM Documentation #2984 (Rejected): doc: write performance tuning
- 03:56 PM Documentation #2983 (Rejected): doc: write performance monitoring
- 03:56 PM CephFS Documentation #2982 (Resolved): doc: write add/remove a metadata server
- 03:52 PM Documentation #2981 (Resolved): doc: write add/remove a monitor
- 03:52 PM Documentation #2980 (Resolved): doc: write upgrading Ceph version
- 03:52 PM Documentation #2979 (Closed): doc: write doc on how to use / rollback to RADOS snapshots
- 03:51 PM Documentation #2978 (Closed): doc: write RADOS restore from backup procedure
- 03:51 PM Documentation #2977 (Closed): doc: write RADOS backup procedure
- 03:51 PM devops Documentation #2976 (Closed): doc: update chef doc to git clone with http, not ssh
- 03:50 PM devops Documentation #2975 (Rejected): doc: update docs to match new ceph-disk-prepare syntax
- 03:50 PM devops Documentation #2974 (Resolved): doc: update chef docs for mon key distribution
- 03:50 PM Documentation #2973 (Resolved): doc: expand/complete ceph general settings
- 03:49 PM rgw Documentation #2972 (Resolved): doc: expand/complete rgw settings reference
- 03:49 PM Documentation #2971 (Resolved): doc: expand/complete mon settings reference
- 03:48 PM Documentation #2970 (Resolved): doc: expand/complete osd settings reference
- 03:47 PM CephFS Documentation #2969 (Resolved): doc: expand/complete mds settings reference
- 03:46 PM Documentation #2968 (Resolved): doc: complete architecture section
- 02:23 PM rbd Feature #2850 (Duplicate): libceph: support multi-operation transactions
- 01:07 PM Bug #2784 (Can't reproduce): osd hit suicide timeout
- 12:49 PM Bug #2856 (Resolved): osd: bound size of transactions trimming old osdmaps
- 09:13 AM Linux kernel client Bug #2936: Remounting cephfs with non-existing path causes kernel panic
- There are patches to do that pending, but i haven't pushed them to the tree yet because a regression in 3.6-rc1 break...
- 09:10 AM Linux kernel client Bug #2936: Remounting cephfs with non-existing path causes kernel panic
- I see the change for #2959 is in the mds.
However, the kernel still shouldn't hang on bad data from the mds, so I ... - 08:32 AM Linux kernel client Bug #2936: Remounting cephfs with non-existing path causes kernel panic
- this is the same issue Yan hit, #2959.
- 09:09 AM rbd Bug #2532 (Resolved): rbd command allows passing in -K </path/to/secret>, but long version of (--...
- 09:05 AM rbd Bug #2967 (Resolved): librbd: cls_rbd.parents unit test failure
- ...
08/18/2012
08/17/2012
- 04:18 PM Bug #2954: osd: scrub stat mismatch, got 18/19 objects, 14/15 clones, 22478527/25385282 bytes.
- logs: ubuntu@teuthology:/a/teuthology-2012-08-17_00:00:25-regression-next-testing-basic/2877
- 04:09 PM rbd Bug #2958 (Resolved): librbd: discard can return -ENOENT
- 04:08 PM rbd Bug #2958 (Fix Under Review): librbd: discard can return -ENOENT
- 03:35 PM Bug #2960 (Resolved): ceph osd create claims you can specify '<osd-id>'; really means UUID. Could...
- just merged a fix for this
- 11:38 AM Bug #2960 (Resolved): ceph osd create claims you can specify '<osd-id>'; really means UUID. Could...
- I think we should consider a global pass making "id" clearer in context, but the
ceph osd create usage message, name... - 03:08 PM rgw Bug #2961 (Resolved): rgw: bad content range
- Partial download of large file (> 4G), the content range is bad:...
- 11:59 AM Bug #2947 (In Progress): osd: out of order reply
- 11:28 AM Bug #2761: osd: failed to recover before timeout expired
- logs: ubuntu@teuthology:/a/teuthology-2012-08-17_02:00:04-regression-testing-testing-basic/3038
- 11:27 AM Bug #2955: monitors failed to open new election
- logs: ubuntu@teuthology:/a/teuthology-2012-08-17_02:00:04-regression-testing-testing-basic/2973
- 08:50 AM CephFS Bug #2959 (Resolved): mds: returns null dentry on getattr
- the kclient open_root_dentry issues a getattr request like #1/some/path, but the mds must not return a dentry in the ...
08/16/2012
- 09:14 PM CephFS Bug #1945: blogbench hang on caps
- ...
- 05:10 PM rbd Documentation #2670 (Resolved): Docs shouldn't direct users to echo to /sys/bus/rbd for normal use
- 05:01 PM rbd Bug #2958 (Resolved): librbd: discard can return -ENOENT
- Sometimes discard tries to remove nonexistent objects, and does not translate the -ENOENT to 0 for its callers. This ...
- 04:55 PM Bug #2957 (Resolved): osd: crash in PG::gen_prefix()
- 03:30 PM Bug #2957 (Resolved): osd: crash in PG::gen_prefix()
- ...
- 04:45 PM rbd Feature #2719 (In Progress): librbd: provide functions for listing parents and their children
- 04:43 PM rbd Feature #2723 (Resolved): librbd: protect/unprotect as appropiate during cloning
- 04:43 PM rbd Subtask #2606 (Resolved): librbd layering: copyup on missing child object
- 04:43 PM rbd Feature #2722 (Resolved): cls_rbd: add class methods to get/set protected status
- 04:43 PM rbd Subtask #2605 (Resolved): librbd layering: guard writes
- 04:43 PM rbd Subtask #2604 (Resolved): librbd layering: read path
- 04:43 PM rbd Subtask #2603 (Resolved): librbd layering: open parent on open
- 04:43 PM rbd Feature #2562 (Resolved): librbd: open parent images, read path, write path
- 04:43 PM rbd Feature #2607 (Resolved): librbd: copyup helper
- 04:43 PM rbd Feature #2561 (Resolved): rbd: copyup command
- 04:42 PM rbd Feature #2559 (Resolved): cls_rbd: copyup method
- 02:15 PM Bug #2954: osd: scrub stat mismatch, got 18/19 objects, 14/15 clones, 22478527/25385282 bytes.
- several more failures in /a/sage-a3 to look at.
- 10:11 AM Bug #2954 (Resolved): osd: scrub stat mismatch, got 18/19 objects, 14/15 clones, 22478527/2538528...
- ...
- 01:46 PM rbd Bug #2948: rbd: fails to close image on error
- This affects operations that fail partway through. One example is:
rbd export <image> <existing-file>
export err... - 10:41 AM rbd Bug #2948: rbd: fails to close image on error
- 01:30 PM Bug #2946 (Resolved): osd: build fails on g++ 4.7
- 01:29 PM Bug #2823 (Duplicate): osd: out of order ACKs
- 01:21 PM Bug #2947: osd: out of order reply
- 12:04 PM Bug #2761: osd: failed to recover before timeout expired
- Recent logs: ubuntu@teuthology:/a/teuthology-2012-08-16_02:00:06-regression-testing-testing-basic/2211
- 11:32 AM Bug #2956 (Resolved): osd:FAILED assert(waiting_for_ondisk.begin()->first == repop->v)
- Logs: ubuntu@teuthology:/a/teuthology-2012-08-15_19:00:16-regression-master-testing-gcov/1878...
- 11:29 AM Bug #2873 (Resolved): Stack trace thrown when using obsync
- commit:47b24c0562bcb44964a0b8f6c4847bb0f05924e0 in stable-next
commit:5962a9dde051c95b7f39e60dcd16b339392685b8 in ne... - 11:18 AM Bug #2955 (Can't reproduce): monitors failed to open new election
- logs: ubuntu@teuthology:/a/teuthology-2012-08-16_00:00:15-regression-next-testing-basic/2077
08/15/2012
- 06:33 PM rbd Bug #2950 (Resolved): ObjectCacher: leaks memory
- commit:825f7334eef7cc69c6f439c21dd0bbb215dbf09d
it wasn't the buffers, it was some BufferHeads that had references... - 11:41 AM rbd Bug #2950 (Resolved): ObjectCacher: leaks memory
- As reported in http://permalink.gmane.org/gmane.comp.file-systems.ceph.devel/7746
- 06:06 PM Bug #2922: mkcephfs fails with error "read: arg count"
- Hmm, my testing of the modifications has a little buggy itself sorry. But after more careful analysis I can confirm t...
- 05:55 PM Bug #2922 (Resolved): mkcephfs fails with error "read: arg count"
- ah, it's just a stupid bash vs dash thing with the 'read' command. i tested on debian (bash), breaks on ubuntu. pus...
- 05:41 PM Bug #2922: mkcephfs fails with error "read: arg count"
- just fyi you can enclose things in pre tags to make redmine skip its own formatting:...
- 05:39 PM Bug #2922 (In Progress): mkcephfs fails with error "read: arg count"
- building right now to test this out... i could have sworn i tested the directory exists situation, but i guess not!
... - 05:36 PM Bug #2922: mkcephfs fails with error "read: arg count"
- Sorry that last one does *not* work properly either.
- 05:29 PM Bug #2922: mkcephfs fails with error "read: arg count"
- This might be cleaner (I'll avoid a diff as they seem to get mangled):
Replacing:
if test -d $mon_data && ! f... - 03:19 PM Bug #2922: mkcephfs fails with error "read: arg count"
- Hmm - I don't think so:
The amended code works ok if the directory does not exist, but fails if it exists and is e... - 02:13 PM Feature #2953 (Resolved): append() in librados is not exposed to python API
- the append to an object is not available at the pyton API level and needs to be implemented.
- 11:51 AM rbd Feature #2952 (Resolved): librbd: use generic rados locking class
- Replace calls to cls_rbd's locking methods with calls to the generic lock class.
- 11:49 AM rbd Feature #2951 (Resolved): cls_rbd: remove locking methods
- Remove the unused cls_rbd locking methods, and merge the tests with the cls_lock tests.
- 10:27 AM rbd Bug #2948 (Resolved): rbd: fails to close image on error
- calling exit() doesn't run the Image destructor, which leads to the watch on the header sticking around. After that, ...
- 10:10 AM rbd Feature #2723 (Fix Under Review): librbd: protect/unprotect as appropiate during cloning
- 10:09 AM rbd Feature #2722 (Fix Under Review): cls_rbd: add class methods to get/set protected status
- 10:09 AM rbd Feature #2718 (Fix Under Review): librbd: map parent -> child in a per-pool rbd_children object w...
- 10:09 AM rbd Feature #2717 (Fix Under Review): cls_rbd: add methods for maintaining mapping from parent to chi...
- 10:09 AM rbd Feature #2562 (Fix Under Review): librbd: open parent images, read path, write path
- 10:09 AM rbd Feature #2562 (Need More Info): librbd: open parent images, read path, write path
- 10:08 AM rbd Subtask #2605 (Fix Under Review): librbd layering: guard writes
- 10:08 AM rbd Subtask #2604 (Fix Under Review): librbd layering: read path
08/14/2012
- 05:41 PM Bug #2947 (Resolved): osd: out of order reply
- triggered by thrashing by this job:...
- 04:45 PM Bug #2922 (Resolved): mkcephfs fails with error "read: arg count"
- commit:24a26c627400d191bbb07cdd3ecfa644c9e313eb
- 04:28 PM Bug #2946 (Resolved): osd: build fails on g++ 4.7
- ...
- 04:06 PM Feature #2918 (Resolved): OSD ID numbers determine OSD count and thus default pg_cnt
- 02:14 PM Feature #2918 (Fix Under Review): OSD ID numbers determine OSD count and thus default pg_cnt
- 02:58 PM Feature #2942 (Resolved): mon: throttle client, server connections
- 02:34 PM Feature #2619 (Resolved): filejournal: instrument with perfcounters
- commit:9fc79584728f87938d13757d5176c5d19d3ca2cb
- 02:07 PM Feature #2940 (Resolved): daemons do not print out version to log on startup
- 12:18 PM Feature #2940: daemons do not print out version to log on startup
- 01:58 PM rbd Bug #2777 (Resolved): qemu: report discard support
- 01:18 PM Bug #2945 (Won't Fix): package upgrade from v0.46 to v0.48argonaut fails
- I saw this once but assumed I had broken dependencies with my version mangling, but then it came up during a third pa...
- 01:13 PM RADOS Subtask #2793 (Resolved): osd: require tunable feature if current osdmap uses non-default tunables
- 01:13 PM RADOS Subtask #2792 (Resolved): mon: require tunable feature bit if current osdmap uses non-default tun...
- 01:13 PM RADOS Feature #2705 (Resolved): crush: graceful transition to new default tunables
- 12:18 PM RADOS Feature #2705 (In Progress): crush: graceful transition to new default tunables
- 12:19 PM Feature #2320 (Duplicate): mon: detect and throttle osd flapping
- 12:18 PM Feature #2742 (In Progress): qa: ms socket inject failures in regression suite
- 12:14 PM Feature #1754 (Resolved): qa: run other suites nightly as well
- 12:13 PM Feature #1514 (Duplicate): filestore: api to repartition a collection
- 12:12 PM Feature #2440: osd: understand btrfs performance
- 12:12 PM Feature #2440 (Won't Fix): osd: understand btrfs performance
- 12:12 PM Feature #2564 (Resolved): teuthology: install kernels from local dir
- 11:45 AM Feature #2944 (Duplicate): mon: dynamically adjust heartbeat grace
- Basically:
1) Keep track of when an OSD boots if it reports itself as fresh or as
wrongly-marked-down. Maintain the... - 11:44 AM Feature #2943 (Resolved): mon: norecovery and/or nobackfill
- 11:42 AM Cleanup #2763 (Resolved): move rbd locking infrastructure to a separate objclass
- 11:42 AM Feature #2768 (Resolved): teuthology: make workunit task work on different branch/sha1 etc
- 11:41 AM Feature #2857 (Resolved): compile non-production builds with -fno-omit-frame-pointer
- 09:37 AM Bug #2761: osd: failed to recover before timeout expired
- Recent logs: ubuntu@teuthology:/a/teuthology-2012-08-13_19:00:07-regression-master-testing-gcov/108
08/13/2012
- 09:48 PM Bug #2922: mkcephfs fails with error "read: arg count"
- FWIW - this seems to happen even if the mon directory does not exist - there should probably be a check of the form:
... - 07:54 PM Bug #2938 (Resolved): ceph-osd --mkfs failure to create journal is logged with dout(0), probably ...
- commit:294c25bb37aa39caacee51cc405a1f2deebb6331
- 11:09 AM Feature #2942 (Resolved): mon: throttle client, server connections
- 10:57 AM rgw Feature #2941 (Resolved): rgw: improve streaming read performance
- 10:51 AM Bug #2823: osd: out of order ACKs
- Recent logs: ubuntu@teuthology:/a/teuthology-2012-08-11_00:00:08-regression-next-testing-basic/6401
- 10:46 AM Feature #2940 (Resolved): daemons do not print out version to log on startup
- I imagine this applies to the other daemons too, but maybe not. Make it print out the version so we can be sure it's ...
- 09:28 AM devops Feature #2939 (Rejected): chef: Write up how cluster shrinking should work
- Expanding the cluster is pretty trivial, and practically identical with initial install, but shrinking needs a little...
08/12/2012
- 10:30 AM CephFS Bug #2444: null pointer deference in ceph_d_prune inside kvm
- problem doesent seem to be reproductible after upgrading to 3.5.0-9-generic (Ubuntu Quantal)
- 03:38 AM rbd Bug #2937: btrfs filesystem on rbd device kernel BUG writing large file
- I activated some extra debugging.
This appears just before the BUG:...
08/11/2012
- 06:33 PM Bug #2887: pjd open/08.t failed test 2
- this is an upstream fuse regression in the 3.6-rc1 kernel. reported to miklos and the fuse list.
- 06:28 PM Linux kernel client Bug #2868 (Resolved): kclient: crash in __kick_osd_requests -> __reset_osd -> __remove_osd
08/10/2012
- 08:31 PM Bug #2919 (Fix Under Review): ceph kernel module looks for :/ in path, but / stripped by precise ...
- 08:29 PM Bug #2938: ceph-osd --mkfs failure to create journal is logged with dout(0), probably should be derr
- yeah, just change it to derr
- 06:34 PM Bug #2938 (Resolved): ceph-osd --mkfs failure to create journal is logged with dout(0), probably ...
- A customer mistakenly named a directory as his osd journal location; the failure printed to his terminal with no hint...
- 08:25 PM Linux kernel client Bug #2801 (Resolved): msgr crash in ceph_msg_new
- 08:24 PM Linux kernel client Bug #2392 (Resolved): First read of symlink after ceph filesystem mounted gives error
- 04:26 PM Bug #2887: pjd open/08.t failed test 2
- ubuntu@teuthology:/a/teuthology-2012-08-09_00:00:04-regression-next-testing-basic/5752
- 03:59 PM Bug #2887: pjd open/08.t failed test 2
- ubuntu@teuthology:/a/teuthology-2012-08-09_02:00:13-regression-testing-testing-basic/5857
- 01:59 PM rbd Bug #2937 (Duplicate): btrfs filesystem on rbd device kernel BUG writing large file
- Writing a large file with dd on btrfs filesystem mounted from rbd device causes kernel bug
Stock kernel 3.5.1, con... - 01:48 PM Linux kernel client Bug #2936 (Resolved): Remounting cephfs with non-existing path causes kernel panic
- Steps to reproduce:
First mount the root somewhere... - 10:38 AM Bug #2913 (Resolved): monclient: asserts when no monitor addresses found due to dns failure
- Fortunately I was wrong about the string splitting - that was just a confusing message from the parsing stage.
The... - 10:09 AM rgw Feature #771: rgw: POST
- Support the S3 POST object operation referenced in
http://docs.amazonwebservices.com/AmazonS3/latest/API/RESTObje... - 09:40 AM rgw Bug #2935 (Resolved): rgw: radosgw-admin bucket link clobbers index
- radosgw-admin bucket unlink, then radosgw-admin bucket link overrides the bucket index, so objects cannot be listed a...
08/09/2012
- 04:06 PM Feature #2934: crush: create a visualizer for crush maps
- 'ceph osd tree' provides a good start on the command line, but it'd be nice to have that in the crushtool as well if ...
- 04:04 PM Feature #2934 (New): crush: create a visualizer for crush maps
- The language used in crush maps is very well defined and
hierarchical. I don't know how to do this sort of thing,
... - 03:55 PM rbd Bug #2933 (Resolved): rbd: bio_pair leak in bio_chain_clone()
- Guangliang Zhao <gzhao@suse.com> pointed out this problem on the
mailing list. Here's the latest edition of his pro... - 02:18 PM devops Feature #2932 (Rejected): chef: logstash integration
- 02:18 PM devops Feature #2931 (Rejected): chef: StatsD integration
- 01:54 PM rgw Feature #2499 (Resolved): rgw: ability to delete users without first emptying and deleting all bu...
- done, commit:45f7f0602c90073af27041f92166724ca9472197.
- 01:53 PM rgw Feature #2786 (Resolved): radosgw-admin: ability to remove objects/buckets
- object removal done, commit:cc8eac2427c745e154ad40eeb84ef28dbed99d36
bucket removal done, commit:45f7f0602c90073af27... - 01:32 PM rgw Bug #2504 (Resolved): rgw: use multiple notifications objects
- Done, commit:b28db08ea8b84ec9f1d2df88ac4edd6aea0ba7d4
- 12:29 PM Bug #2924 (Resolved): doc: Adjust for mon. key being in external keyring
- This doc is outdated
http://ceph.com/docs/master/ops/manage/grow/mon/#adding-a-monitor
as per
http://thread.gmane.... - 11:13 AM CephFS Bug #2444: null pointer deference in ceph_d_prune inside kvm
- same bug here with Ceph 0.49 on Ubuntu 12.04 LTS (GNU/Linux 3.2.0-27-generic x86_64)
- 10:58 AM rgw Feature #2923 (Resolved): rgw: non hard-coded pool names
- Don't have pool names hard coded, make them configurable.
- 10:44 AM rgw Bug #2665 (Resolved): rest-bench hangs periodically
- This was fixed a while ago.
08/08/2012
- 04:58 PM Bug #2922 (Resolved): mkcephfs fails with error "read: arg count"
- Branch: wip-auth
ceph version 0.49-306-gfc3681f (commit:fc3681f59c4f49298f5a7a5172c30be63068c330)
tamil@tamil-Vir... - 04:08 PM rgw Bug #2841 (Resolved): rgw: fix usage trim
- Fixed, commit:6bc1067fc878cbfb6761146cb154c2985c9d9bd7 and commit:04a0eacd92b0c923cb9d1efc7d751a05d544dc85
- 03:35 PM rgw Feature #2869 (Resolved): rgw: expand date format support
- Fixed, commit:074c3c0fe0c005e54f4776c60463a16305dbab10
- 03:34 PM rgw Bug #2879 (Resolved): rgw: xml parser doesn't work correctly with escape sequences
- Fixed, commit:03b787e0ee1d94e054cfb17059e5e108a7162d7b
- 03:34 PM rgw Bug #2878 (Resolved): rgw: chunked encoding for POST requests (e.g., complete multipart uploads)
- Fixed, commit:d39ea1d4b51afdbbd51254ff41c8285e8f5697df.
- 03:33 PM rgw Bug #2877 (Resolved): rgw: ETag parsing in complete multipart upload should xml decode ETag
- Fixed, commit:3809e34448e47d7baa02d7a0f9240494aba0e337.
- 02:06 PM Bug #2845 (Resolved): mkcephfs hasn't learned about new default keyring locations in argonaut
- fixed, commit:96b1a496cdfda34a5efdb6686becf0d2e7e3a1c0
- 12:48 PM Bug #2875 (Resolved): osd: pg stuck in GetLog
- 12:48 PM Bug #2834 (Resolved): osd/ReplicatedPG.cc: 3577: FAILED assert(waiting_for_ack.begin()->first == ...
- hasn't come up recently
- 11:11 AM Bug #2887: pjd open/08.t failed test 2
- Logs: ubuntu@teuthology:/a/teuthology-2012-08-06_00:00:02-regression-next-testing-basic/5012
- 10:03 AM Bug #2887: pjd open/08.t failed test 2
- Recent logs: ubuntu@teuthology:/a/teuthology-2012-08-08_00:00:07-regression-next-testing-basic/5542
- 10:48 AM Bug #2761: osd: failed to recover before timeout expired
- Recent logs: ubuntu@teuthology:/a/teuthology-2012-08-08_00:00:07-regression-next-testing-basic/5616...
- 10:33 AM rgw Bug #2915: rgw: copy of large object times out
- workaround: bump up fastcgi timeout
- 10:11 AM Feature #2921 (Rejected): doc: Provide epub docs
- Sphinx supports it. Current output seems to only include the top-level file and indexes, skipping most of the content...
- 10:04 AM Feature #2920 (Rejected): doc: Provide PDF docs
- Sphinx supports it, but we'd need to fix other parts of our toolchain.
To see where we are:
1. add this patch:
...
08/07/2012
- 05:56 PM Bug #2919 (Resolved): ceph kernel module looks for :/ in path, but / stripped by precise mountall
- I think this is really a bug in mountall (see https://bugs.launchpad.net/ubuntu/+source/mountall/+bug/809221), but it...
- 05:42 PM Feature #2918 (Resolved): OSD ID numbers determine OSD count and thus default pg_cnt
- An IRC user (maelfius) had a problem with a 1-monitor, 3-OSD cluster; the monitor chewed up all memory before it star...
- 12:50 PM rgw Bug #2916: radosgw does not check command line options for correctness
- that's a generic ceph command lines parsing issue
- 11:52 AM rgw Bug #2916 (Resolved): radosgw does not check command line options for correctness
- It is possible to pass any command line option to radosgw without error. For example
./radosgw -c /home/caleb/cep... - 12:49 PM rgw Bug #2915: rgw: copy of large object times out
- The problem is that apache is timing out. We should return an early 200 and encode any error in the response code, as...
- 10:34 AM rgw Bug #2915 (Resolved): rgw: copy of large object times out
- 09:26 AM devops Feature #2808 (Rejected): crowbar: upgrade to fred3 (get bind bug fix)
- Looks like upstream Crowbar is still buggy with regard to DNS.
- 06:37 AM Bug #2913: monclient: asserts when no monitor addresses found due to dns failure
- I am using 0.48argonaut-1precise.
08/06/2012
- 04:22 PM Bug #2914 (Resolved): librados set_complete_callback, set_safe_callback clobber each other's argu...
- 02:46 PM Bug #2913: monclient: asserts when no monitor addresses found due to dns failure
- hmm, looking closer that's a second bug - it's not splitting 'thinkmate3:6789;thinkmate4:6789' into separate addresse...
- 02:28 PM Bug #2913: monclient: asserts when no monitor addresses found due to dns failure
- I'm not so sure this is a DNS issue. Here is how name service is set up on my ceph/kvm test cluster.
On each node,... - 01:58 PM Bug #2913 (Resolved): monclient: asserts when no monitor addresses found due to dns failure
- This should be an error returned up to the user, not an assert.
From https://www.redhat.com/archives/libvirt-users... - 02:16 PM Bug #2887: pjd open/08.t failed test 2
- recent logs: ubuntu@teuthology:/a/teuthology-2012-08-06_02:00:02-regression-testing-testing-basic/5117
- 09:53 AM Feature #2911 (Duplicate): osd: Restrict recovery when the OSD full list is nonempty
- See the conversation at http://www.spinics.net/lists/ceph-devel/msg08010.html
It would be nice if we could somehow...
08/03/2012
- 06:31 PM Bug #2909 (Resolved): the rados_trunc function did not get implemented in rados.py Python
- Thanks! Added in commit:43291951fad241a6d3f8b8daa37d3665c9d842d6, with a simple test and spacing normalized to the re...
- 03:58 PM Bug #2909: the rados_trunc function did not get implemented in rados.py Python
- Yes you may.
- 03:35 PM Bug #2909: the rados_trunc function did not get implemented in rados.py Python
- That looks good to me. Can I add your signed-off-by to the patch?
- 01:53 PM Bug #2909 (Resolved): the rados_trunc function did not get implemented in rados.py Python
- This code seems to work in the Ioctx class:
def trunc(self,key,size):
self.require_ioctx_open()
... - 02:23 PM devops Feature #2910: crowbar: Use JBOD mode for ceph-osd
- There's a map in the deployer object from role name to BIOS and RAID configuration to set on the node. We can add cep...
- 02:22 PM devops Feature #2910 (Closed): crowbar: Use JBOD mode for ceph-osd
- 11:38 AM Bug #2908 (Resolved): ceph osd crush remove <name>
- (11:34:50 AM) Kyle Bader: so it looks like ceph -h is missing crush rm
(11:34:54 AM) Kyle Bader: could we add
(11:... - 07:56 AM Subtask #2738 (Rejected): mon: Single-Paxos: Sync: Add snapshot support to the monitor store
- This task was superseded by task #2756, which provides a much more broad implementation using directly the available ...
- 07:45 AM Subtask #2737: mon: Single-Paxos: Sync: Force trimming to be proposed through Paxos
- 07:44 AM Subtask #2805 (Resolved): mon: Single-Paxos: Sync: Create a test unit to verify the correctness o...
- 07:44 AM Subtask #2758 (Resolved): mon: Single-Paxos: Sync: Extend the in-memory mock-up of KeyValueDB to ...
- 07:43 AM Subtask #2756 (Resolved): mon: Single-Paxos: LevelDBStore: Make iterator thread-safe
08/02/2012
- 04:51 PM Bug #2907: rados benchmarking tool which does not always do creates
- Why not record the raw data and let other tools produce percentiles and other statistics?
- 04:44 PM Bug #2907 (Resolved): rados benchmarking tool which does not always do creates
- Features:
Pluggable distribution for choosing objects (zipifan?, random?, sequential?)
configurable numbe... - 04:09 PM Bug #2904 (Resolved): ceph-authtool: Adds keys on typos, expected error message
- ...
- 03:50 PM CephFS Feature #2903 (Resolved): ceph-fuse: Support -o noallow_other
- Currently, ceph-fuse hardcodes the -o allow_other option to FUSE_ARGS_INIT.
https://github.com/ceph/ceph/blob/5db3... - 01:35 PM rgw Bug #2841 (Fix Under Review): rgw: fix usage trim
- 01:31 PM rgw Bug #1855 (Resolved): Creation of a subuser that appears to own an s3 key is possible, and removi...
- Commit 5db3a9e71c6b757660d0702efada40af6be63eb8 pushed. We disallow creating s3 key when subuser is created in order ...
- 01:27 PM devops Feature #2398: chef: external osd journal support
- Shuffling old notes here:
see if "osd journal" was overridden in $cluster.conf; if yes, do not attempt discovery
... - 12:59 PM rgw Feature #2869 (Fix Under Review): rgw: expand date format support
- 12:59 PM rgw Bug #2877 (Fix Under Review): rgw: ETag parsing in complete multipart upload should xml decode ETag
- 12:59 PM rgw Bug #2878 (Fix Under Review): rgw: chunked encoding for POST requests (e.g., complete multipart u...
- 12:59 PM rgw Bug #2879 (Fix Under Review): rgw: xml parser doesn't work correctly with escape sequences
- 11:50 AM Bug #2902 (Resolved): common lib tries to open literal ~/.ceph/ceph.conf
- ...
- 11:48 AM Bug #2901 (Resolved): librados-config should not read ceph.conf
- ...
- 11:38 AM Bug #2900 (Resolved): ceph fuse crashed
- Logs: ubuntu@teuthology: /a/teuthology-2012-07-27_19:00:07-regression-master-testing-gcov/1581
Core file: /a/teuthol... - 11:09 AM Bug #2897 (Resolved): ceph fuse error segfault
- ...
- 11:03 AM devops Feature #2780 (Closed): gitbuilder: move to vercoi, redo deployment if feasible
- 10:52 AM Bug #2823: osd: out of order ACKs
- Log location: ubuntu@teuthology:/a/teuthology-2012-08-01_19:00:04-regression-master-testing-gcov/4196
ubuntu@teuth... - 10:50 AM Bug #2823: osd: out of order ACKs
- (10:46:42 AM) tamil.muthamizhan@newdream.net: 4196: (1138s) collection:rados-thrash clusters:6-osd-3-machine.yaml fs:...
- 10:45 AM Bug #2823 (New): osd: out of order ACKs
- 10:10 AM Bug #2887: pjd open/08.t failed test 2
- recent logs: ubuntu@teuthology:/a/teuthology-2012-08-01_19:00:04-regression-master-testing-gcov/4126
- 10:00 AM Bug #2896 (Won't Fix): ceph pg dump has empty hb_out field
- I was looking at "ceph pg dump" output today on a patched argonaut build and saw that while all the osd stat outputs ...
08/01/2012
- 06:31 PM Bug #2895: cli: non-existent command returns confusing error message
- ...and ceph osd map rbd/rbd_info returns "unknown command map', which is just wrong;
the problem is the argument nee... - 05:33 PM Bug #2895 (Resolved): cli: non-existent command returns confusing error message
- 'ceph osd crush get' returns 'unknown command crush', instead of the full command.
http://www.spinics.net/lists/ce... - 05:28 PM Feature #2894 (Resolved): cli: help command for ceph subsystems
- To make commands and their usage discoverable and easy to look up, each subsystem could provide a help command
that ... - 04:50 PM Bug #2887: pjd open/08.t failed test 2
- Also, ubuntu@teuthology:/a/teuthology-2012-07-31_19:00:04-regression-master-testing-gcov/3654
- 04:47 PM Bug #2887: pjd open/08.t failed test 2
- Also, ubuntu@teuthology:/a/teuthology-2012-08-01_00:01:38-regression-next-testing-basic/3784
- 04:43 PM Bug #2887: pjd open/08.t failed test 2
- Recent logs: ubuntu@teuthology:/a/teuthology-2012-08-01_02:00:04-regression-testing-testing-basic/3909
- 01:51 PM Bug #2887 (Resolved): pjd open/08.t failed test 2
- pjd open/08.t failed test 2 on both ceph-fuse and kclient.
Logs:- ubuntu@teuthology:/a/teuthology-2012-07-31_02:00... - 04:34 PM devops Feature #2893 (Closed): crowbar: Nested virtualization for running OpenStack in vercoi vm
- 04:34 PM devops Feature #2893 (Closed): crowbar: Nested virtualization for running OpenStack in vercoi vm
- 04:30 PM Bug #2892 (Resolved): ceph health detail kills monitor
- Executed the following:
ubuntu@burnupi30:~$ sudo ceph health detail
Wait awhile and ceph will start to output t... - 03:10 PM Bug #2891 (Can't reproduce): heap profiler hangs when trying to start it up on the mon
- We tried to turn heap profiling on the mon (congress), however the last thing we see in the logs is the message that ...
- 02:24 PM Bug #2890 (Resolved): monitor: "recognize" heap commands
- The monitor accepts the standard heap profiler commands, but it tells the user it doesn't due to not setting return c...
- 01:56 PM devops Feature #2889 (Closed): crowbar: script for easily packaging ceph barclamp
- 01:53 PM devops Feature #2888 (Closed): crowbar: Make VM disk IO cache writes for performance
- 01:38 PM devops Documentation #2886 (Rejected): doc: crush location tricks, ceph.conf, automatic host=
- - how it autoupdates on osd startup
- how hosts won't migrate from container to another automatically - 01:37 PM devops Feature #2885 (Resolved): doc: mon initial members requirements, functioning, admin steps to take
- 01:36 PM devops Feature #2884 (Rejected): doc: osd hotplugging
- 01:34 PM devops Feature #2883 (Rejected): chef: union lists and maps in env vs node ceph.conf json
- As an admin, I want to specify in environment
"osd crush location": {
"datacenter": "westcoast",
}
an... - 01:33 PM devops Documentation #2882 (Rejected): doc: chef environment ceph.conf content tricks
- 01:32 PM devops Feature #2881 (Rejected): doc: chef cookbook better README, internal structure, assumptions
- 01:29 PM devops Feature #2704 (Closed): sepia: Use ``names`` as resolver on plana, burnupi, vercoi
- dnscache01 and dnscache02 are happily serving anything that uses DHCP to get its configuration.
There may be stati... - 01:20 PM devops Feature #2880 (Rejected): chef: use get-or-create instead of get-or-create-key
- ceph.git commit 4551808fa00b812fee6e0c196fd333eca0b06de9 adds "ceph auth get-or-create". Switch to using it in ceph-c...
- 01:10 PM rgw Bug #2877: rgw: ETag parsing in complete multipart upload should xml decode ETag
- There are two different issues here. The first one is that we don't remove the quotes when comparing the etags. The s...
- 12:55 PM rgw Bug #2879 (Resolved): rgw: xml parser doesn't work correctly with escape sequences
- e.g., when providing data with """, the entity is getting clobbered.
07/31/2012
- 09:44 PM Bug #2873 (Fix Under Review): Stack trace thrown when using obsync
- 06:18 PM Bug #2873: Stack trace thrown when using obsync
- Figured out what the problem is, it appears that on L111, it should go from being...
- 11:27 AM Bug #2873 (Resolved): Stack trace thrown when using obsync
- ...
- 03:38 PM RADOS Bug #2874: apparent CRUSH mapping failure
- check if setting the tunables all to 0 makes it go away
- 11:40 AM RADOS Bug #2874 (Resolved): apparent CRUSH mapping failure
- While doing crowbar tests, I created a 3-OSD cluster (on separate VMs) that ended up with 6 degraded PGs....
- 03:36 PM rgw Bug #2504 (In Progress): rgw: use multiple notifications objects
- 03:35 PM rgw Bug #2878 (Resolved): rgw: chunked encoding for POST requests (e.g., complete multipart uploads)
- We shouldn't require length passed for these requests.
- 03:28 PM rgw Bug #2877 (Resolved): rgw: ETag parsing in complete multipart upload should xml decode ETag
- Should be able to accept both:...
- 03:08 PM Bug #2876 (Resolved): mon: pg stuck peering (for example) broken?
- ...
- 02:01 PM Bug #2875 (Resolved): osd: pg stuck in GetLog
- we weren't checking if newest_update_osd went down (it could be outside the prior set)
- 12:43 PM Linux kernel client Bug #2573 (Resolved): libceph: many "socket closed" messages
- I was seeing this too, but with the latest code and all (knock wood) the races closed I'm not anymore. Going to opti...
- 11:49 AM Bug #2846 (Resolved): Malformed keyring file causes kernel null pointer deref on "rbd map"
- userland fixes applied to stable, next.
thanks! - 11:42 AM Bug #2846: Malformed keyring file causes kernel null pointer deref on "rbd map"
- kernel patch is in testing branch.
- 06:23 AM Subtask #2805 (Fix Under Review): mon: Single-Paxos: Sync: Create a test unit to verify the corre...
- 06:22 AM Subtask #2805: mon: Single-Paxos: Sync: Create a test unit to verify the correctness of the whole...
- Currently available tests:
* Removing keys:
> * Using both the whole-space iterator and the whole-space snapshot ...
07/30/2012
- 06:46 PM Linux kernel client Bug #2868: kclient: crash in __kick_osd_requests -> __reset_osd -> __remove_osd
- hoping this was the messenger locking stuff, let's see if it pops up again
- 06:45 PM rbd Bug #2715 (Resolved): krbd: spinlock wrong CPU
- 06:45 PM Linux kernel client Bug #2867 (Resolved): kclient: crash from ffsb in con_work -> kernel_sendmsg
- 06:45 PM Linux kernel client Bug #2392: First read of symlink after ceph filesystem mounted gives error
- 04:52 PM rbd Bug #2872 (Resolved): RBD resize command allows image size -1
- Ceph Version : 0.48
Resize rbd image to size -1 allows rbd image to be resized to 15 Exabytes, which is incorrect.... - 03:52 PM rbd Bug #2871 (Resolved): rbd export command hangs when trying to export an image of size 0 to a loca...
- Ceph Version: 0.48
Steps followed:
1. create a rbd image of size 1000 mb in rbd pool
2. resize the rbd image t... - 10:52 AM Bug #2866 (Resolved): osd: pg stuck with unfound
- commit:9e5d4e61a73343397e67e918e87f1e6dcb8ec72d and commit:7b9d37c662313929b52011ddae47cc8abab99095
- 10:51 AM Bug #2860 (Resolved): osd: stuck waiting for pg acting set to change
- commit:bae837010b6b486011b06dd97664fb54c3f3ff44 and commit:96feca450c5505a06868bc012fe998a03371b77f
- 09:14 AM Bug #2819: krbd: lockup on large writes, msgr fault injection
- i'm unable to reproduce this on a real kernel.. it only happens on uml.
here is a full backtrace:... - 08:01 AM Bug #2638 (Resolved): mon: make pool ops idempotent
- 08:01 AM Bug #2830 (Duplicate): [argonaut] osd/OSD.cc: 3906: FAILED assert(_get_map_bl(epoch, bl))
07/29/2012
- 09:31 PM Linux kernel client Bug #2688 (Duplicate): lockup on ffsb + thrashing
- 09:31 PM Linux kernel client Bug #2260 (Resolved): libceph: null pointer dereference at try_write+0x638+0xfb0
- this is either #2867, or a similar issue that is since resolved.
- 09:28 PM Linux kernel client Bug #2790 (Duplicate): libceph: crash in read_partial_message_section on ffsb
- 09:24 PM Linux kernel client Bug #2867: kclient: crash from ffsb in con_work -> kernel_sendmsg
- *sigh of relief*
- 08:22 PM Linux kernel client Bug #2867: kclient: crash from ffsb in con_work -> kernel_sendmsg
- This appears to be a regression, so it is effectively blocking sending the pull request to Linus.
07/28/2012
- 05:52 PM Feature #2280 (Resolved): improve gitbuilder infrastructure
- 05:50 PM RADOS Subtask #2792 (Fix Under Review): mon: require tunable feature bit if current osdmap uses non-def...
- 03:49 PM rgw Feature #2869 (Resolved): rgw: expand date format support
- should be able to parse the following:
Sat, 28 Jul 2012 20:35:55 UTC
Which uses UTC instead of GMT. - 03:30 PM Feature #2477 (Fix Under Review): rados bench cleanup
- 03:30 PM Feature #1783 (Fix Under Review): osd: scrub incrementally across hash range using MOSDPGScan
- 07:37 AM Linux kernel client Bug #2868 (Resolved): kclient: crash in __kick_osd_requests -> __reset_osd -> __remove_osd
- ...
07/27/2012
- 05:52 PM Linux kernel client Bug #2867 (Resolved): kclient: crash from ffsb in con_work -> kernel_sendmsg
- ...
- 05:18 PM Bug #2866 (Fix Under Review): osd: pg stuck with unfound
- 04:29 PM Bug #2866 (Resolved): osd: pg stuck with unfound
- on congress, observed pg stuck with unfound objects. kicking peering (marking primary down) resolved it.
in testi... - 05:15 PM Bug #2860 (Fix Under Review): osd: stuck waiting for pg acting set to change
- 03:17 PM Bug #2860: osd: stuck waiting for pg acting set to change
- i can reproduce this with:...
- 12:39 PM Bug #2860 (Resolved): osd: stuck waiting for pg acting set to change
- ...
- 03:26 PM rbd Bug #2865 (Resolved): rbd import fails for directory but creates rbd image
- Ceph Version: 0.48
Created a local directory t_dir.
when tried to import directory t_dir to rbd/rbd_image, it rep... - 02:36 PM rgw Bug #2864 (Won't Fix): rados leaves behind references to old buckets
- As this behavior can only be reproduced through deleting objects directly through rados, and not radosgw-admin or API...
- 02:09 PM rgw Bug #2864 (Won't Fix): rados leaves behind references to old buckets
- It is possible to create an inconsistent state by following this procedure:
1. create a bucket through an API call... - 01:58 PM Bug #2824 (Resolved): ceph-fuse; hang mounting with ms failures
- 01:46 PM CephFS Bug #2863 (Resolved): client: does not tolerate traceless replies from mds
- In at least one case (_create's _mknod) we do not tolerate a (write) reply from the mds with no trace. This happens ...
- 01:21 PM rbd Bug #2862 (Resolved): CLI: rbd create command throws inappropriate error messages
- Ceph Version: 0.48
When tried a few negative test cases using "rbd create command", found that the command display... - 12:57 PM rbd Bug #2861 (Won't Fix): CLI: rbd create command requires validation for image-name
- Ceph version: 0.48
When trying to create a rbd image, the image name seems to accept empty string and special char... - 11:45 AM Bug #2462: osd/PG.cc: 402: FAILED assert(log.head >= olog.tail && olog.head >= log.tail)
- just swa this on congress during a huge crush restructure:...
- 11:31 AM rgw Tasks #2859 (New): Make add subuser in radosgw-admin idempotent
- Currently, attempting to create a subuser that already exists returns an error; it has been suggested that this behav...
- 11:25 AM Bug #2858: mon: osd id parsing returns 0 when passed 'osd.1234'
- Not sure exactly what scenario you're looking at here or what the bug is, but there are lots of places in the monitor...
- 11:08 AM Bug #2858 (Resolved): mon: osd id parsing returns 0 when passed 'osd.1234'
- 10:54 AM Bug #2752: Setting large maxosd kills all mons
- Thanks Yehuda!
- 10:51 AM Feature #2857 (Resolved): compile non-production builds with -fno-omit-frame-pointer
- This will let us get much more useful profiling data out of various tools with relatively minimal CPU overhead.
- 08:17 AM Bug #2856 (Resolved): osd: bound size of transactions trimming old osdmaps
- The monitor can arbitrarily advance it's oldest map. The osd should avoid sending down an arbitrarily large transacti...
07/26/2012
- 10:36 PM Bug #2830 (Need More Info): [argonaut] osd/OSD.cc: 3906: FAILED assert(_get_map_bl(epoch, bl))
- this may duplicate #2843.. sadly didn't take note of the osd id :(
- 10:34 PM Bug #2837 (Resolved): osd: past_interval calculation inefficient
- 10:34 PM Bug #2849 (Resolved): osd: past_intervals not shared on backfill restart
- 04:25 PM Bug #2849 (Resolved): osd: past_intervals not shared on backfill restart
- peer info value is clobbered by backfill block prior to the dne() check in PG::activate()
this explains a lot! - 06:08 PM rbd Subtask #2855 (Closed): krbd: copy-up on write to clone
- 06:07 PM rbd Subtask #2854 (Closed): krbd: write path
- verify the target object exists in write requests. if we fail with ENOENT, trigger a copy-up.
- 06:07 PM rbd Tasks #2853 (Resolved): krbd: read path
- 06:06 PM rbd Subtask #2852 (Closed): krbd: open parent on open
- 06:05 PM rbd Feature #2851 (Duplicate): krbd: RBD layering support
- Kernel client should support all the layering functionality of the usermode client.
- 06:00 PM rbd Feature #2850 (Duplicate): libceph: support multi-operation transactions
- 03:23 PM Bug #2752 (Resolved): Setting large maxosd kills all mons
- Fixed, commit:5601ae27d6daf167dd83b3fc91b7b9591ca0cea6.
- 12:28 PM Bug #2848 (Won't Fix): OSDMap: pool_id is 64-bit, but pool_max is 32-bit
- A large number of pools will overflow pool_max before using the full range of pool ids.
- 12:26 PM Linux kernel client Cleanup #2847 (Resolved): libceph: osdmap definition is out of date
- In particular, pool_id is an int instead of a 64-bit integer. There are probably other important differences as well.
- 10:53 AM rbd Feature #2562 (In Progress): librbd: open parent images, read path, write path
- 10:46 AM rbd Feature #2726 (In Progress): krbd: clean up bio_pair leak/whatever
- Guangliang Zhao sent a patch to fix that, however, I had some concerns about it, and I'm waiting for him to respond t...
- 08:26 AM Bug #2846: Malformed keyring file causes kernel null pointer deref on "rbd map"
- Ok, I finally know the failing path.
So when you call add_key with an invalid payload, it will be parsed by ceph_k... - 08:07 AM Bug #2846: Malformed keyring file causes kernel null pointer deref on "rbd map"
- Damnit ... first it didn't take the formatting and second I pasted the wrong code :p...
- 08:05 AM Bug #2846: Malformed keyring file causes kernel null pointer deref on "rbd map"
- wrt to kernel crash, here's a minimal test case that will crash any machine that has rbd module loaded (works as user...
- 04:28 AM Bug #2846: Malformed keyring file causes kernel null pointer deref on "rbd map"
- I was pointing to a keyring file directly that happened to start with an empty line. So in rbd.cc, the function read_...
- 03:53 AM Bug #2846 (Resolved): Malformed keyring file causes kernel null pointer deref on "rbd map"
- Reported by Sylvain Munaut ("tnt" on OFTC):
(12:30:27) tnt: Is mounting a RBD on a machine that has an OSD suppose... - 02:32 AM Bug #2845 (Resolved): mkcephfs hasn't learned about new default keyring locations in argonaut
- In 0.48, when running @mkcephfs@ in a @cephx@ authentication enabled cluster, the per-daemon keys for MDSs and OSDs a...
07/25/2012
- 09:54 PM Bug #2843 (Can't reproduce): filestore: replay failure on xfs
- congress osd.328 crashed with...
- 05:55 PM Bug #2842: mon: health detail lists pgs multiple times
- This and #2827 may be related?
- 05:52 PM Bug #2842 (Won't Fix): mon: health detail lists pgs multiple times
- ...
- 05:16 PM rgw Bug #1855: Creation of a subuser that appears to own an s3 key is possible, and removing the subu...
- This bug can be reproduced by using the following options
./radosgw-admin -c {'ceph.conf'} --rgw-socket-path=/tmp... - 05:16 PM rgw Bug #2841 (Resolved): rgw: fix usage trim
- looking at the code, it seems that we don't encode the user in usage-trim (and also encode the wrong structure).
- 04:52 PM CephFS Bug #2187: pjd chown/00.t failed test 97
- 2012-07-23T19:16:10.185 INFO:teuthology.task.workunit.client.0.out:not ok 43
2012-07-23T19:16:10.186 INFO:teuthology... - 04:51 PM CephFS Bug #2187: pjd chown/00.t failed test 97
- Latest log: ubuntu@teuthology:/a/teuthology-2012-07-23_19:00:03-regression-master-testing-gcov/16530
- 04:30 PM Feature #2840 (Resolved): mon: $mon_data/cluster_fsid file
- maybe written/verified by mkfs!
- 04:23 PM rgw Feature #2839 (Resolved): rgw: garbage collection
- Provide a garbage collection mechanism, along the lines of what was described in a post to the mailing list.
- 04:20 PM rgw Bug #2652: Segmentation fault in rest-bench
- is it still happening?
- 04:20 PM rgw Bug #2665: rest-bench hangs periodically
- is that still happening?
- 04:10 PM devops Feature #2574 (Resolved): crowbar: use data disks automatically, journal inside data directory
- There were bugs and the history was wrecked by github pull requests again, so I redid some commits, but this function...
- 03:45 PM rgw Feature #2039 (Rejected): rgw: keep more than one bucket marker object
- That's not the case anymore. We use the unique client id and a running counter instead.
- 02:08 PM Bug #2838 (Resolved): mon: json version of 'osd tree'
- 01:52 PM Bug #2824: ceph-fuse; hang mounting with ms failures
- 01:52 PM Bug #2835 (Resolved): osd: do not send alive/upthru until booted
- 01:52 PM Bug #2836 (Resolved): osd: boot condition check incorrect
- 10:52 AM Bug #2836 (Resolved): osd: boot condition check incorrect
- commit:5979351ef3d3d03bced9286f79cbc22524c4a8de
- 11:04 AM Bug #2837 (Resolved): osd: past_interval calculation inefficient
- It is still possible for osds to get pgs without past intervals and need to recalculate them, and that calculation ca...
Also available in: Atom