Activity
From 06/20/2013 to 07/19/2013
07/19/2013
- 11:59 PM Revision 0356eebf (ceph): mon/PaxosService: update on_active() docs to clarify calling rules
- Signed-off-by: Sage Weil <sage@inktank.com>
- 11:57 PM Revision 6d326b84 (ceph): mon/OSDMonitor: discard failure waiters, info on shutdown
- This would prevent a leak, if we didn't assert before that in the
failure_reporter_t dtor.
Signed-off-by: Sage Weil ... - 11:39 PM Revision e4f2e3ec (ceph): mon/OSDMonitor: do not wait for readable in send_latest()
- send_latest() checks for readable and, if untrue, will wait before sending
out the latest OSDMap. This is completely... - 11:39 PM Revision 8371680b (ceph): mon: OSDMonitor: only thrash and propose if we are the leader
- 'thrash_map' is only set if we are the leader, so we would thrash and
propose the pending value if we are the leader.... - 11:39 PM Revision 6edec516 (ceph): Revert "mon/OSDMonitor: send_to_waiting() in on_active()"
- This reverts commit f06a124a7fa0717ef8c523408b31d814df57caca.
On peons, on_active() is only called when we *first* b... - 11:27 PM Revision 19bf7f7b (ceph): modified the workload for mixed-mon upgrade suite
- Signed-off-by: tamil <tamil.muthamizhan@inktank.com>
- 11:23 PM Revision 2795eb12 (ceph): Revert "mon: OSDMonitor: only thrash and propose if we are the leader"
- This reverts commit 5eac38797d9eb5a59fcff1d81571cff7a2f10e66.
- 11:22 PM Revision 0a996493 (ceph): Revert "mon/OSDMonitor: fix typo"
- This reverts commit d656aed599ee754646e16386ce5a4ab0117f2d6e.
- 11:01 PM rgw Bug #5691 (Fix Under Review): All bucket ACLs lost after upgrade
- Pushed fix to wip-5691.
- 07:12 PM rgw Bug #5691 (Resolved): All bucket ACLs lost after upgrade
- I upgraded from 0.66 to current next, 8c5e1db4fb76b5e1fcf6721ad210f143a571d7b8 (via c0e77c91b6c39998ef4e19a726db87b66...
- 10:09 PM Revision 8c5e1db4 (ceph): ceph_rest_api.py: remove unused imports
- Fixes: #5684
Signed-off-by: Dan Mick <dan.mick@inktank.com>
Reviewed-by: Sage Weil <sage@inktank.com> - 10:01 PM Revision ce46961e (ceph): ceph.in: better error message when daemon command returns nothing
- Fixes: #5683
signed-off-by: Dan Mick <dan.mick@inktank.com>
Reviewed-by: Sage Weil <sage@inktank.com> - 09:50 PM Revision 06ae53e2 (ceph): mon: improve osdmap subscription debug output
- Signed-off-by: Sage Weil <sage@inktank.com>
- 09:49 PM Revision d3902e2e (ceph): Merge remote-tracking branch 'gh/wip-stats' into next
- Reviewed-by: Samuel Just <sam.just@inktank.com>
- 09:44 PM Revision 633a1881 (ceph): Merge branch 'wip-centos-rgw'
- 09:44 PM Revision bd56af70 (ceph): Merge branch 'wip-centos-rgw'
- 09:42 PM Revision f0123db7 (ceph): rgw: turn off continue print on centos
- Also note that a ceph.conf setting is needed for that too
Signed-off-by: Josh Durgin <josh.durgin@inktank.com> - 09:36 PM Revision 29e7db26 (ceph): install: remove ceph-release rpm file after it's used
- No need for it anymore, it's just taking up space
Signed-off-by: Josh Durgin <josh.durgin@inktank.com> - 08:58 PM rgw Feature #5693 (Resolved): rgw: send data through copy operation to avoid timeout
- 08:50 PM Revision f4675dcb (ceph): test: switch the cls_replica_log tests to use a test fixture
- Instead of the silly macros.
Signed-off-by: Greg Farnum <greg@inktank.com> - 08:50 PM Revision 934ad881 (ceph): rgw: grab the old mtime when doing a user metadata put
- We're about to use this to enable (ick) a timestamp-based update
decision tree, to help in resolving disastrous doubl... - 08:29 PM Bug #5692 (Resolved): RADOS performance regression in 0.65
- Looks like after some narrowing down that we have what appears to be a pretty serious write performance regression st...
- 08:25 PM Revision bc1aca77 (ceph): Merge branch 'wip-rgw-next-2' into next
- Reviewed-by: Greg Farnum <greg@inktank.com>
- 08:21 PM Revision e4d2787b (ceph): rgw-admin: link / unlink should report errors
- Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
- 08:21 PM Revision 4e05786a (ceph): rgw: replace logic that compares regions
- The logic was a bit broken. Basically, we want to make sure
that region names are the same. However, if region name i... - 08:21 PM Revision d44082e4 (ceph): cls_rgw: quiet down verbose log message
- Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
- 08:21 PM Revision da8584f1 (ceph): rgw: remove extra unused param from RGWRados::get_attr()
- No user for the extra obj_version param.
Signed-off-by: Yehuda Sadeh <yehuda@inktank.com> - 08:21 PM Revision 4f4bdbd5 (ceph): rgw: fix bucket re-creation on secondary region
- We had a problem with bucket recreation, where we identified
that bucket has already existed, but missed the fact tha... - 08:21 PM Revision 2e518235 (ceph): rgw: forward x_amz_meta headers when forwarding a request
- Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
- 08:21 PM Revision 989a4d93 (ceph): rgw: adjust error for bucket removal on secondary region
- Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
- 08:21 PM Revision 85f3f09b (ceph): rgw: forward delete bucket request to master after removal
- We can only forward the bucket removal to the master if it was
successfully removed locally.
The master region has no... - 08:21 PM Revision 89ecba20 (ceph): rgw: remove s->objv_tracker
- was never initialized correctly anyway. It was only supposed to
be used for buckets, but it was never initialized in ... - 08:21 PM Revision 7cd0bd85 (ceph): rgw: bucket entry point object ver fixes
- Multiple fixes:
- sync master, secondary entry point ver on creation
- use correct entry point version when removin... - 08:21 PM Revision 0024e5aa (ceph): rgw: fix time parsing in replica log
- Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
- 08:21 PM Revision 30de0406 (ceph): mon: break con <-> session ref cycle in mon even if shutting down
- If we get a reset during shutdown, we should still break the cycle to avoid
tripping the valgrind leak detection. No... - 08:21 PM Revision 000d4d38 (ceph): mon: mark_down session by con, not addr
- We have the ConnectionRef here; use it. This avoids generating a spurious
RESET event for the connection.
Signed-of... - 08:21 PM Revision 11c47cc4 (ceph): client: mark_down by con
- We have the con handy; use it. This avoids generate a spurious RESET
event, which we do not need or do anything usef... - 08:21 PM Revision d1b83be1 (ceph): msgr: mark addr-based [lazy_]send_message and get_connection deprecated
- Signed-off-by: Sage Weil <sage@inktank.com>
- 08:21 PM Revision 0de70851 (ceph): mon/MonClient: fix small leak
- We need to delete the version_req_d here.
Signed-off-by: Sage Weil <sage@inktank.com> - 08:21 PM Revision 41c67e02 (ceph): osd: make ms_handle_reset debug more useful
- Signed-off-by: Sage Weil <sage@inktank.com>
Reviewed-by: Samuel Just <sam.just@inktank.com> - 08:21 PM Revision 561ac0b1 (ceph): osd: break con <-> session cycle when marking down old peers
- Signed-off-by: Sage Weil <sage@inktank.com>
Reviewed-by: Samuel Just <sam.just@inktank.com> - 08:21 PM Revision bfadcd2a (ceph): osd/ReplicatedPG: fix obc leak on invalid LIST_SNAPS op
- Signed-off-by: Sage Weil <sage@inktank.com>
Reviewed-by: Samuel Just <sam.just@inktank.com> - 08:21 PM Revision 8dcf0b19 (ceph): msgr: generate reset event on mark_down to addr (not con)
- If the caller is marking down an addr, they presumably don't have the
Connection* handy, so we should generate a rese... - 08:21 PM Revision 27868ca5 (ceph): msgr: update docs for mark_down, mark_down_all semantics
- * RESET events
* note that the reset detection only happens if it is enabled in the
policy.
Signed-off-by: Sage We... - 08:21 PM Revision 564075c9 (ceph): msg/SimpleMessenger: remove duplicated interface docs
- Document these in the interface, not the implementation; having two copies
clutters the header and invites them to ge... - 07:49 PM Revision 6ebb4863 (ceph): doc: Update syntax for rbd children
- The docs were incorrectly usins 'rbd snap children' when the actual
command is 'rbd children'.
Signed-off-by: Travis... - 07:49 PM Bug #5624 (Fix Under Review): osd: prepare_to_stop() segfaults if we get a signal during startup
- In wip-4624
- 06:32 PM Documentation #5690 (Closed): ceph "global options" should be documented somewhere
- There are a number of global options that virtually every Ceph tool supports; things like
-i/--id, -n/--name, -c/--c... - 06:27 PM Revision 0cc1cffd (ceph): Merge remote-tracking branch 'gh/next'
- 06:26 PM Revision d874b57e (ceph): sleep: new task
- 06:06 PM Bug #5672: osd: stuck in DeletingState::try_stop_deletion
- 03:43 PM Bug #5672: osd: stuck in DeletingState::try_stop_deletion
- osd/filestore lock loop, I'll have a fix shortly
- 03:09 PM Bug #5672: osd: stuck in DeletingState::try_stop_deletion
- ah, dispatch is stuck here:...
- 02:35 PM Bug #5672: osd: stuck in DeletingState::try_stop_deletion
- You're just noticing that osd4 dispatched messages that osd3 never received and faulted on the communication line for?
- 05:55 PM Bug #5683 (Resolved): ceph.in: error message on bad daemon command is confusing
- 02:09 PM Bug #5683 (Resolved): ceph.in: error message on bad daemon command is confusing
- When we get no data back from the socket because a bad command, the error message is confusing.
Easy to fix. - 05:55 PM Revision 8574b3cd (ceph): mon/PGMap: don't mangle stamp_delta in clear_delta()
- This is a delta, not a timestamp.
This triggered when a cluster is idle for 2* the mon_delta_reset_interval,
and req... - 05:54 PM Cleanup #5684 (Resolved): ceph_rest_api.py: remove unused imports
- 02:10 PM Cleanup #5684 (Fix Under Review): ceph_rest_api.py: remove unused imports
- ...
- 02:10 PM Cleanup #5684 (Resolved): ceph_rest_api.py: remove unused imports
- 05:45 PM Revision 99fa208f (ceph): osd: log PG state changes at level 5
- Signed-off-by: Sage Weil <sage@inktank.com>
Reviewed-by: Samuel Just <sam.just@inktank.com> - 05:39 PM Revision c549e628 (ceph): mon/PGMap: avoid negative pg stats when calculating rates
- We periodically see strange values come out of the estimated cluster
throughput and recovery rates. Pretty sure this... - 05:39 PM Revision 9292c18d (ceph): mon/PGMap: use signed values for calculated rates
- si_t (and friends) does not handle signed values, but at least we can
give the Formatters unmangled values. This sho... - 05:39 PM Revision 9f56a7b8 (ceph): PG, ReplicatedPG: pass a transaction down to ReplicatedPG::on_change
- Signed-off-by: Samuel Just <sam.just@inktank.com>
Reviewed-by: Sage Weil <sage@inktank.com> - 05:39 PM Revision 47516d9c (ceph): ReplicatedPG: track temp collection contents, clear during on_change
- We also assert in on_flushed() that the temp collection is actually
empty.
Fixes: #5670
Signed-off-by: Samuel Just <... - 05:31 PM Revision 2e09b6e0 (ceph): osd: add floor() method to pg/osd stat structs
- We often want to maintain a nonnegative value. We generalize
this to floors other than zero only because it makes th... - 05:31 PM Revision b4db5886 (ceph): osd: make pool_stat_t *log_size fields signed
- Signed-off-by: Sage Weil <sage@inktank.com>
- 05:22 PM CephFS Bug #5689 (Resolved): ceph-fuse crashed after upgrading from cuttlefish to dumpling
- when trying to run some workload after upgrading from cuttlefish v0.61.4 to v0.66, ceph-fuse crashed and also seeing ...
- 05:10 PM Revision 375c0d4e (ceph): Merge pull request #18 from alfredodeza/master
- Safer dict.get in queues.py
- 05:04 PM Revision 4bf4d92a (ceph): doc: radosgw needs 'mon = allow rw' to create its own pools
- Fixes: #5674
Signed-off-by: Sage Weil <sage@inktank.com>
Reviewed-by: Yehuda Sadeh <yehuda@inktank.com> - 04:59 PM Revision d3e53cd6 (ceph): mon/MonClient: better debugging on version requests
- From leak hunting, but useful.
Signed-off-by: Sage Weil <sage@inktank.com> - 04:20 PM rgw Cleanup #5124 (Resolved): rgw: use region instead of regions in radosgw-admin
- The only use of "regions" is that you can use it for "regions list". That also works if you do "region list" and ever...
- 04:17 PM Revision 23cde453 (ceph): Merge remote-tracking branch 'gh/next'
- 03:23 PM devops Documentation #5688 (Resolved): ceph-deploy: upgrade procedure has to be documented
- once bug#5686, bug #5687 are resolved, we need to document upgrade steps using ceph-deploy.
This is important for ... - 03:21 PM Revision c4d4f345 (ceph): doc/release-notes: fix typo
- Signed-off-by: Sage Weil <sage@inktank.com>
- 03:17 PM devops Bug #5686 (Rejected): ceph-deploy: install command should not restart the daemons when used for u...
- right now, we use install command to upgrade to a different ceph branch[ex: dumpling branch] on a running cluster[ex:...
- 03:09 PM Revision f0feabe8 (ceph): Message,OSD,PG: make Connection::features private
- Use has_feature() method too.
Signed-off-by: Samuel Just <sam.just@inktank.com>
Signed-off-by: Sage Weil <sage@inkta... - 03:09 PM Revision 053659d0 (ceph): msg/Pipe: work around incorrect features reported by earlier versions
- If we see a peer reporting features ~0ull, we know they are deluded in a
particular way and should infer what feature... - 02:48 PM Revision d1b47f40 (ceph): test: update cli test for radosgw-admin
- Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
- 02:36 PM Bug #2544 (Closed): Help text for "usage show" identical to "usage trim"
- Fixed last June
commit:c87c83f3184dab50823f74b5a30b553d65eedb65
- 02:20 PM Revision 7a7fca46 (ceph): Merge pull request #448 from kri5/wip-5416
- rgw: Adds --rgw-zone --rgw-region help text.
Reviewed-by: Yehuda Sadeh <yehuda@inktank.com> - 02:19 PM rgw Bug #5675 (Resolved): New rgw issues 301 redirect without Location header
- This should be resolved by commit:4e05786a58a1218e1b68d5fbcddefa7a72ac03ce
- 09:54 AM rgw Bug #5675 (Fix Under Review): New rgw issues 301 redirect without Location header
- 12:48 AM rgw Bug #5675 (Resolved): New rgw issues 301 redirect without Location header
- On my upgraded rgw running wip-5655 (c0e77c91b6c39998ef4e19a726db87b66850cf2c), i.e. two commits ahead 921a4aac8a8985...
- 02:17 PM Bug #5297: Slow requests after restarting an OSD (post peering)
- comment out clean_up_local()
- 02:17 PM Bug #5297 (Pending Backport): Slow requests after restarting an OSD (post peering)
- 02:14 PM Bug #5504 (Duplicate): osd stack on peereng for a long time
- see #5517
- 02:14 PM Bug #5301 (Can't reproduce): mon: leveldb crash in tcmalloc
- 02:14 PM Bug #5513 (Can't reproduce): osd crashes consistently after startup
- Missing map in osd store, possibly corrupt.
- 02:13 PM Bug #5401 (Can't reproduce): cuttlefish osd recovery slow
- 02:12 PM Bug #5596 (Can't reproduce): OSD keeps crashing while trying to rewight another OSD
- 02:02 PM Tasks #5584 (Resolved): measure peering performance
- 02:02 PM Feature #4462 (Resolved): api: RESTful client: implement remaining N-1 commands JSON or XML
- 02:02 PM Feature #4459 (Resolved): api: RESTful client: implement remaining commands JSON only
- 02:02 PM Feature #4460 (Resolved): api: RESTful client: demonstrate remaining N-1 commands JSON only
- 02:02 PM Tasks #5583 (Resolved): populate test cluster with rgw data
- 02:02 PM Tasks #5582 (Resolved): create large test cluster on burnupi and/or mira
- 02:02 PM Feature #4549 (Resolved): api: Incorporate and respond to comments on DRAFT REST API
- 01:55 PM Cleanup #2078 (Resolved): ceph tool: only output response data to stdout
- A crazy amount of iteration to get to this, but the CLI rewrite has resolved this issue, mostly
by outputting a blan... - 01:54 PM RADOS Bug #3693: crushtool compile fails with unhelpful message, diagnosis quite difficult
- Removing myself as owner; there's no way I'm going to be able to untangle boost::spirit in anything less than months
- 01:52 PM Feature #4463 (Resolved): api: RESTful client: demonstrate remaining N-1 commands JSON or XML
- commit:4cb0e062661a7190366753554d7202aa0168f265 et. seq.
- 01:51 PM devops Feature #5682 (Rejected): gitbuilder: create a way to push a branch to the front of the gitbuidle...
- 01:25 PM Bug #5473 (Resolved): osd/ReplicatedPG.cc: 1379: FAILED assert(0) in trim_object() on master, cut...
- 01:25 PM Bug #5507 (Resolved): osd: ENOENT on clone
- 01:25 PM Bug #5392 (Resolved): osd: unfound objects from thrashing
- 01:25 PM Bug #5269 (Resolved): osd: EEXIST on mkcoll
- 01:25 PM Bug #5631 (Resolved): osd/ReplicatedPG.cc: 3036: FAILED assert(iter)
- 01:24 PM Bug #5154 (Resolved): osd/SnapMapper.cc: 270: FAILED assert(check(oid))
- 01:24 PM Bug #5670 (Resolved): ceph_test_rados; FAILED assert(0)
- 01:21 PM rgw Feature #5680 (Resolved): rgw: teuthology: merge in changes to enable multiple regions
- 01:12 PM rgw Feature #5606 (In Progress): s3-tests: run in multi region setup
- 01:07 PM rgw Tasks #5586 (Resolved): rgw: build test plan
- 01:06 PM rgw Feature #4340: rgw: dr: data sync agent: implement full sync
- 01:05 PM rgw Feature #4339: rgw: multisite: metadata sync agent: update sync processing state
- 01:05 PM rgw Feature #4338: rgw: multisite: metadata sync agent: implement delta changes sync
- 01:05 PM rgw Feature #4337: rgw: multisite: metadata sync agent: implement full sync
- 01:05 PM rgw Cleanup #5558 (Resolved): rgw: modify certain radosgw-admin operations interface
- Merged in commit:7f67848
- 01:01 PM Revision 5263a6f7 (ceph): failed dict.get evaluates to None
- 12:59 PM Revision 65f49beb (ceph): safer get by falling back to a dictionary
- 12:52 PM rgw Bug #5674 (Resolved): Upgrade from 0.66 -> next: upgrade issues because of default caps
- doc fixes are merged. this isn't a new cap requirement. probably you were fine before because the .rgw* pools were ...
- 09:52 AM rgw Bug #5674 (Fix Under Review): Upgrade from 0.66 -> next: upgrade issues because of default caps
- 08:38 AM rgw Bug #5674: Upgrade from 0.66 -> next: upgrade issues because of default caps
- Attached it the log I kept from back before I modified my caps.
- 08:35 AM rgw Bug #5674: Upgrade from 0.66 -> next: upgrade issues because of default caps
- can you provide a log, so that we can identify what additional log message is missing. From what I can tell, other th...
- 12:41 AM rgw Bug #5674 (Resolved): Upgrade from 0.66 -> next: upgrade issues because of default caps
- Hi,
Due to bugfixes I'm testing, I upgraded all of my OSDs & mons to wip-5655 (c0e77c91b6c39998ef4e19a726db87b6685... - 11:59 AM rgw Bug #5676: swift nightly tests failing
- apparently I misread the log, the pool creation response arrives quickly.
- 11:27 AM rgw Bug #5676: swift nightly tests failing
- pool creation takes too long:...
- 08:19 AM rgw Bug #5676 (Resolved): swift nightly tests failing
- 11:46 AM devops Bug #4936: ceph-deploy fails to report errors
- change default behavior of ceph-deploy to be -v
- 11:45 AM devops Bug #4919 (Resolved): ceph-deploy: disk list doesn't properly display all the disks on a VM
- 11:45 AM devops Bug #5208: Debian Wheezy Needs the 'ca-certificates' package before you can wget the package keys
- do same for lsb-release and other dependencies
- 11:32 AM rgw Bug #5679 (Resolved): radosgw-agent: partial sync doesn't handle bucket.instance entries
- Just a silly typo.
commit:d693e6792dc74a2240ba5406417148e16bf1920d - 10:43 AM rgw Bug #5679 (Resolved): radosgw-agent: partial sync doesn't handle bucket.instance entries
- Just verified with latest version
- 11:20 AM Bug #5652: mon: occasional clock_skew test failure
- this still hard codes the defaults in teuthology.. a more robust way is to run
ceph-mon -i mon.$whatever --show-... - 09:50 AM devops Bug #5678 (Resolved): ceph-deploy mon destroy broken
- finally figured out why this keeps failing:
the mon replies immediately to 'mon destroy' command before it commits... - 09:11 AM devops Feature #5018: arm: ceph-deploy: push packages to ARM
- One possible problem that might be encountered is that upstream arm packages seem to have been moved from archive.ubu...
- 09:05 AM rbd Bug #5428: libceph: null deref in ceph_auth_reset
- focusing on teh warning leading up to this first: it looks like the socket callback is happening when the socket is i...
- 08:49 AM rbd Bug #5428: libceph: null deref in ceph_auth_reset
- hit this again, ubuntu@teuthology:/a/teuthology-2013-07-19_01:01:11-krbd-next-testing-basic/73011
- 08:48 AM CephFS Bug #3601: client: With multiple clients, file remove doesn't free up space
- Just make sure to check that the client doesn't have the file opened locally before trimming. ;)
- 08:17 AM CephFS Bug #3601: client: With multiple clients, file remove doesn't free up space
- Zheng Yan wrote:
> I think we can try trimming the unlinked inode when receiving cap message with "nlink == 0".
Y... - 01:07 AM CephFS Bug #3601: client: With multiple clients, file remove doesn't free up space
- I think we can try trimming the unlinked inode when receiving cap message with "nlink == 0".
- 08:46 AM Bug #5677 (Resolved): osd/OSD.cc: 5517: FAILED assert(_get_map_bl(epoch, bl))
- ...
- 08:13 AM Revision 76040d90 (ceph): rgw: Adds --rgw-zone --rgw-region help text.
- Signed-off-by: Christophe Courtaut <christophe.courtaut@gmail.com>
- 06:02 AM Revision 1b8d50ea (ceph): doc/release-notes: amend 0.61.5 release notes
- Upgrade all mons at once. See #5673
Signed-off-by: Sage Weil <sage@inktank.com> - 02:52 AM Revision 203bc59b (ceph): Merge branch 'wip-machine-based-workers' into next
- 02:37 AM Revision a2efc173 (ceph): Merge branch 'wip-machine-based-workers'
- 01:12 AM Revision a0cb40b4 (ceph): mon: decline to scrub when paxos is not active
- In f1ce8d7c955a2443111bf7d9e16b4c563d445712 we close a race between scrub
and paxos commit completion on the leader. ... - 01:05 AM Revision 4db5b930 (ceph): ceph: do not ignore osd leaks
- Signed-off-by: Sage Weil <sage@inktank.com>
- 01:04 AM Revision c9ba933b (ceph): mon/MonClient: fix small leak
- We need to delete the version_req_d here.
Signed-off-by: Sage Weil <sage@inktank.com> - 01:03 AM Revision 0ef5213d (ceph): Merge pull request #445 from ceph/wip-osd-leaks
- fix msgr issues causing osd leaks on shutdown
Reviewed-by: Greg Farnum <greg@inktank.com> - 12:10 AM Revision 4ed79429 (ceph): init-ceph: don't activate-all for vstart clusters
- Signed-off-by: Sage Weil <sage@inktank.com>
07/18/2013
- 11:53 PM Revision f9e9f9cb (ceph): mon/PGMonitor: fix 'pg map' output key names
- This got lost in a big file of fixes a while back. :/
Signed-off-by: Sage Weil <sage@inktank.com> - 11:53 PM CephFS Bug #5649: smbtorture test gets ebusy on kclient umount
- I manually ran the test 6 times, but couldn't reproduce the issue. I ran the tests using Fedora's samba4-test package...
- 11:04 PM Bug #5673 (Won't Fix): mon: upgrade from 0.61.4 to 0.61.5 breaks quorum
- the problem is that the old mons (0.61.4) are misinterpreting the feature bits of the new mons. the old ones do some...
- 09:30 PM Bug #5673 (Won't Fix): mon: upgrade from 0.61.4 to 0.61.5 breaks quorum
- maybe the feature bits head an inadvertant change? or the scrub feature bit?...
- 10:47 PM Bug #5671: pgs stuck in recovering, not progressing at all
- Ugh. This (9fe7806d4f1fe67fa10906df20cbed017321effe) isn't actually 2 commits behind wip-5655, it's just an earlier i...
- 06:53 PM Bug #5671 (Duplicate): pgs stuck in recovering, not progressing at all
- The osd 1 log indicates that it is running 9fe7806d4f1fe67fa10906df20cbed017321effe which is actually 2 commits behin...
- 05:52 PM Bug #5671: pgs stuck in recovering, not progressing at all
- OK, it gets even weirder.
1) Stop all 4 offending osds. ... - 05:36 PM Bug #5671: pgs stuck in recovering, not progressing at all
- cephdrop has 5671/ceph-osd.1.log.gz, --debug-ms=1 --debug-osd=20 --debug-filestore=20 from osd.1, one of the offendin...
- 05:35 PM Bug #5671 (Duplicate): pgs stuck in recovering, not progressing at all
- All OSDs run 0.66; 4 of them were upgraded to ea1c623406ac0761c6d63041dcc5ca849d1fa932, then I hit #5655. After upgra...
- 10:13 PM Revision 9ab539ea (ceph): PG: add perf counter for peering latency
- Signed-off-by: Samuel Just <sam.just@inktank.com>
Reviewed-by: Sage Weil <sage@inktank.com> - 10:05 PM Revision fd53d53a (ceph): msgr: mark addr-based [lazy_]send_message and get_connection deprecated
- Signed-off-by: Sage Weil <sage@inktank.com>
- 10:04 PM devops Feature #5016: ceph-deploy: gitbuilders for release packages
- RPM and Debian gitbuilders are working, and packages are pushed to gitbuilder.ceph.com.
- 10:02 PM Revision 8ff7d536 (ceph): client: mark_down by con
- We have the con handy; use it. This avoids generate a spurious RESET
event, which we do not need or do anything usef... - 10:02 PM Revision e1e0d505 (ceph): msgr: generate reset event on mark_down to addr (not con)
- If the caller is marking down an addr, they presumably don't have the
Connection* handy, so we should generate a rese... - 10:02 PM Revision 5e724179 (ceph): msgr: update docs for mark_down, mark_down_all semantics
- * RESET events
* note that the reset detection only happens if it is enabled in the
policy.
Signed-off-by: Sage We... - 10:02 PM Revision f7d0403f (ceph): msg/SimpleMessenger: remove duplicated interface docs
- Document these in the interface, not the implementation; having two copies
clutters the header and invites them to ge... - 10:02 PM Revision 3c81475b (ceph): mon: break con <-> session ref cycle in mon even if shutting down
- If we get a reset during shutdown, we should still break the cycle to avoid
tripping the valgrind leak detection. No... - 10:02 PM Revision e0067a85 (ceph): mon: mark_down session by con, not addr
- We have the ConnectionRef here; use it. This avoids generating a spurious
RESET event for the connection.
Signed-of... - 10:02 PM Revision c5226346 (ceph): osd/ReplicatedPG: fix obc leak on invalid LIST_SNAPS op
- Signed-off-by: Sage Weil <sage@inktank.com>
Reviewed-by: Samuel Just <sam.just@inktank.com> - 10:02 PM Revision d3748b23 (ceph): osd: break con <-> session cycle when marking down old peers
- Signed-off-by: Sage Weil <sage@inktank.com>
Reviewed-by: Samuel Just <sam.just@inktank.com> - 10:01 PM Revision 2428bfd6 (ceph): osd: make ms_handle_reset debug more useful
- Signed-off-by: Sage Weil <sage@inktank.com>
Reviewed-by: Samuel Just <sam.just@inktank.com> - 10:01 PM devops Feature #5018 (Resolved): arm: ceph-deploy: push packages to ARM
- Tested.
- 09:53 PM devops Feature #5523 (In Progress): libcurl 7.28+ packages
- Backported 7.29 version from Raring.
Wokring on automatically adding to the gitbuilder and release repos. - 09:08 PM Revision 921a4aac (ceph): cls_lock: fix duration test
- It's possible for us to just be really slow when getting the reply to the
first op or doing the second op, resulting ... - 09:08 PM Revision dd0246d2 (ceph): mds: tracedn should be NULL for LOOKUPINO/LOOKUPHASH reply
- Fixes: #5658
Signed-off-by: Yan, Zheng <zheng.z.yan@intel.com>
Reviewed-by: Sage Weil <sage@inktank.com> - 08:30 PM Revision 382d17ee (ceph): reorg kernel -> krbd + kcephfs
- Signed-off-by: Sage Weil <sage@inktank.com>
- 08:24 PM Revision f3f92fe2 (ceph): FileStore: add global replay guard for split, collection_rename
- In the event of a split or collection rename, we need to ensure that
we don't replay any operations on objects within... - 08:22 PM devops Bug #5345: ceph-disk: handle less common device names
- Hi Sage,
I think the first partition (data) was mounted. I should had noted that when after running zap, parted st... - 08:50 AM devops Bug #5345: ceph-disk: handle less common device names
- Jing Yuan Luke wrote:
> Hi Sage,
>
> I tried the methods you suggested (zap and prepare) on 2 identical servers w... - 02:38 AM devops Bug #5345: ceph-disk: handle less common device names
- Hi Sage,
I tried the methods you suggested (zap and prepare) on 2 identical servers with the same controller. Both... - 07:31 PM Revision 0985f8c3 (ceph): nuke: killall ceph-disk, too
- Signed-off-by: Sage Weil <sage@inktank.com>
- 07:26 PM Bug #5670: ceph_test_rados; FAILED assert(0)
- wip-5670, going to do some testing.
- 04:09 PM Bug #5670: ceph_test_rados; FAILED assert(0)
- Nope. Backfill might just skip an object leaving it treacherously in the temp directory. We'll have to track object...
- 03:53 PM Bug #5670: ceph_test_rados; FAILED assert(0)
- To elaborate, if the object had not been deleted, it would have had to be recovered or backfilled before IO could be ...
- 03:50 PM Bug #5670 (In Progress): ceph_test_rados; FAILED assert(0)
- 1) Primary starts pushing object foo to replica
2) replica creates tmp/foo
3) peering change, replica is out of act... - 03:36 PM Bug #5670 (Resolved): ceph_test_rados; FAILED assert(0)
ceph version 0.66-697-g921a4aa (921a4aac8a89850303233fe188998202e0ddfe0d)
2013-07-18T15:06:33.739 INFO:teutholog...- 07:17 PM Revision 723d691f (ceph): msg/Pipe: do not hold pipe_lock for verify_authorizer()
- We shouldn't hold the pipe_lock while doing the ms_verify_authorizer
upcalls.
Fix by unlocking a bit earlier, and ve... - 07:04 PM Revision 5f9a1d8a (ceph): Worker processes by machine type instead of teuthology branch.
- teuthology-suite and schedulewill now take --worker instead of
--branch. The branch is set by setting teuthology_bran... - 06:38 PM Revision 320032f6 (ceph): schedule_suite.sh: escape ceph-deploy overrides
- 06:21 PM Revision 2173d33d (ceph): ceph-deploy: support overrides
- Something like
overrides:
ceph-deploy:
foo: bar
Signed-off-by: Sage Weil <sage@inktank.com> - 06:10 PM Bug #5512 (Resolved): mon: missing full osdmaps after sync
- 01:57 PM Bug #5512: mon: missing full osdmaps after sync
- Greg Farnum wrote:
> Did we backport the update on commit versus round finish patch?
oh, that's right.. that's wh... - 01:48 PM Bug #5512: mon: missing full osdmaps after sync
- Did we backport the update on commit versus round finish patch?
- 01:46 PM Bug #5512: mon: missing full osdmaps after sync
- digging into why the scrub error came up on cuttlefish; this has been backported.
- 08:35 AM Bug #5512 (Pending Backport): mon: missing full osdmaps after sync
- 06:07 PM Bug #5201 (Resolved): osd: valgrind leaks
- 06:01 PM Bug #5672 (Resolved): osd: stuck in DeletingState::try_stop_deletion
- /a/dzafman-5154-7/71727
traced a random pg and found that osd.3 is having trouble talking to its peers:
from os... - 04:42 PM Revision a59493ec (ceph): doc/release-notes: add/fix changelog links
- Signed-off-by: Sage Weil <sage@inktank.com>
- 04:39 PM Revision d21d39e2 (ceph): doc/release-notes: v0.61.5
- Signed-off-by: Sage Weil <sage@inktank.com>
- 03:38 PM Revision 29c0252d (ceph): mon: fix off-by-one in check for when sync falls behind
- This is what e213b1bc25a212ffe42623c1d4b4eadf9f69319e intended to do
but managed to bungle by using >= instead of >.
... - 03:12 PM rgw Documentation #5669 (Resolved): Default site in Apache interferes with Gateway
- We have found that sometimes Apache installs with an active default site. While this default site is active, the gate...
- 02:52 PM CephFS Bug #5665: mds takeover too early causes new mds to shutdown
- Isn't this basically the MDS not getting to write all its startup state to disk?
Seems like maybe we should just pre... - 09:10 AM CephFS Bug #5665 (Duplicate): mds takeover too early causes new mds to shutdown
- after replay we get...
- 01:52 PM CephFS Bug #5658: kcepht: warning: fs/ceph/inode.c:1000 ceph_fill_trace+0x760/0x900 [ceph]()
- looks right, we'll see how it does tonight
- 01:25 PM rbd Bug #5668 (Resolved): librbd: flush perfcounter unused
- The aio_flush perf counter is incremented whenever an aio_flush is done, but the flush perf counter is never used. It...
- 11:51 AM Bug #5667 (Duplicate): osd/ReplicatedPG.cc: 4782: FAILED assert(obc->ssc)
- I don't have logs to confirm, but it's very likely that this was also caused by 5154.
- 10:56 AM Bug #5667 (Duplicate): osd/ReplicatedPG.cc: 4782: FAILED assert(obc->ssc)
- ...
- 11:38 AM Bug #5630 (Duplicate): osd/ReplicatedPG.cc: 7089: FAILED assert(r >= 0) in scan_range, cuttlefish
- 11:38 AM Bug #5630: osd/ReplicatedPG.cc: 7089: FAILED assert(r >= 0) in scan_range, cuttlefish
- My earlier comments were incorrect, the assert was after a getattr. Also, the second set of logs shows that it's 515...
- 11:18 AM Bug #5630: osd/ReplicatedPG.cc: 7089: FAILED assert(r >= 0) in scan_range, cuttlefish
- /a/dzafman-5154/70692
- 11:07 AM Bug #5655: Slow requests for 1h30 "currently waiting for missing objects"
- The bug is in v0.66 as well. You want commit:4255b5c2fb54ae40c53284b3ab700fdfc7e61748 on all of them (the problem her...
- 10:58 AM Bug #5655: Slow requests for 1h30 "currently waiting for missing objects"
- All the other OSDs run 0.66 but you seem to imply that it should be fixed there?
- 10:54 AM Bug #5655 (Resolved): Slow requests for 1h30 "currently waiting for missing objects"
- This was a bug in 0.65 and cuttlefish before 0.61.5, the feature set was getting computed as ~0 causing the new osds ...
- 03:53 AM Bug #5655: Slow requests for 1h30 "currently waiting for missing objects"
- I also restarted osd.3 & osd.4, which were upgraded directly from 0.66 to 07dfb6f4afac6b2156ad9d9e3489e3c601661c35 an...
- 09:32 AM Bug #5652 (Fix Under Review): mon: occasional clock_skew test failure
- 09:32 AM Bug #5652: mon: occasional clock_skew test failure
- pushed new patch to wip-5652 on teuthology
- 09:10 AM Bug #5652: mon: occasional clock_skew test failure
- Greg pointed me in the right direction and it looks like the 0.25 is coming from the base ceph.conf generate by teuth...
- 08:42 AM Bug #5652: mon: occasional clock_skew test failure
- Okay, there's a simple way to make the test independent from its 'max-skew' option, and that's to read the 'mon_clock...
- 08:28 AM Bug #5652: mon: occasional clock_skew test failure
- Okay, I accessed the cluster while one of my teuth jobs were running, and...
- 08:11 AM Bug #5652 (In Progress): mon: occasional clock_skew test failure
- The cause for the crash has been fixed on teuthology's wip-5652. It was caused by a missing key on a 'str'.format() ...
- 09:09 AM rgw Bug #5664 (Resolved): rgw: can't create containers on secondary region via swift
- With swift there's no way to specify location constraints for the new container. Solution is probably trivial (just s...
- 09:03 AM Bug #5663 (Won't Fix): ceph tool: crash (debian wheezy)
- On a current next branch....
- 08:48 AM Bug #5518: osd: marking single osd down makes others go down (cuttlefish)
- Hi Josh-
yes, this is fixed in the latest cuttlefish (0.61.5, about to send out the announcement now). commit:78f... - 02:16 AM Bug #5518: osd: marking single osd down makes others go down (cuttlefish)
- Hi Sage,
Was this resolved with a patch to Cuttlefish, that should make it in the next minor release?
Thanks.
... - 08:45 AM rbd Bug #5662 (Duplicate): kernel BUG at drivers/block/rbd.c:2133
- see #5454
- 01:07 AM rbd Bug #5662 (Duplicate): kernel BUG at drivers/block/rbd.c:2133
- Hi,
hit this on our cluster today.
Cluster info:
ceph version 0.61.3 (92b1e398576d55df8e5888dd1a9545ed3fd9953... - 05:37 AM devops Bug #5483 (In Progress): ceph-deploy --cluster othername mon create does not work
- 05:28 AM devops Bug #5489 (In Progress): ceph-deploy: mon destroy throws inappropriate message
- 05:03 AM Revision 59f3455e (ceph): Merge pull request #444 from ceph/wip-osd-latency
- osd: include op queue age histogram in osd_stat_t
Reviewed-by: Greg Farnum <greg@inktank.com>
Reviewed-by: Dan Mick ... - 04:36 AM Revision 07dfb6f4 (ceph): rgw: drop unused assignment
- rgw/rgw_rados.cc: In member function 'virtual int RGWPutObjProcessor_Atomic::handle_data(ceph::bufferlist&, off_t, vo...
- 04:33 AM Revision a395d55e (ceph): Merge remote-tracking branch 'gh/next'
- 03:59 AM Revision bfd60107 (ceph): Merge branch 'wip-machine-type'
- Reviewed-by: Sandon Van Ness <sandon@inktank.com>
- 03:54 AM Revision 99c40128 (ceph): lock: filter machine type for --list, --list-targets
- Signed-off-by: Sage Weil <sage@inktank.com>
- 03:54 AM Revision 608d8a20 (ceph): lock: make --summary list all machines by default
- Signed-off-by: Sage Weil <sage@inktank.com>
- 03:54 AM Revision 1d16a9ba (ceph): lock: drop machine-type default, but require for lock-many
- Signed-off-by: Sage Weil <sage@inktank.com>
- 01:21 AM Revision 82722efa (ceph): osd: include op queue age histogram in osd_stat_t
- This includes a simple power-of-2 histogram of op ages in the op queue
inside osd_stat_t. This can be used for a coa... - 01:21 AM Revision aa460c41 (ceph): mon: make 'health' warn about slow requests
- Currently we see slow request warnings go by in the cluster log, but they
are not reflected by 'ceph health'. Use th... - 01:18 AM Revision 2e216b54 (ceph): qa/workunits/cephtool/test.sh: test 'osd create <uuid>'
- Make sure it gives us back the same id.
Signed-off-by: Sage Weil <sage@inktank.com>
Reviewed-by: Dan Mick <dan.mick@... - 01:14 AM Revision 9e91395a (ceph): ceph.conf.template: enable osd debug verify stray on activate
- Signed-off-by: Samuel Just <sam.just@inktank.com>
- 01:11 AM Revision b41f1ba4 (ceph): PG: start flush on primary only after we process the master log
- Once we start serving reads, stray objects must have already
been removed. Therefore, we have to flush all operation... - 01:11 AM Revision 278c7b59 (ceph): ReplicatedPG: replace clean_up_local with a debug check
- Stray objects should have been cleaned up in the merge_log
transactions. Only on the primary have those operations
n...
07/17/2013
- 11:39 PM Revision 8ee10dc4 (ceph): v0.61.5
- 10:49 PM Bug #5655: Slow requests for 1h30 "currently waiting for missing objects"
- The revision I upgraded to the last hour and mentioned in my previous update (07dfb6f4afac6b2156ad9d9e3489e3c601661c3...
- 10:25 PM Bug #5655: Slow requests for 1h30 "currently waiting for missing objects"
- sounds like they are sending the new message types due to incorrect features... maybe because of the msgr stuff we ju...
- 10:21 PM Bug #5655: Slow requests for 1h30 "currently waiting for missing objects"
- I've had multiple /var/log across the cluster fill up from random OSDs that peer with osd.0 & osd.1 spewing "can't de...
- 10:04 PM Bug #5655: Slow requests for 1h30 "currently waiting for missing objects"
- This is now:...
- 03:27 PM Bug #5655: Slow requests for 1h30 "currently waiting for missing objects"
- I suspect a dropped message, this may be related to the stuck peering bug. Unfortunately, the relevant messages don'...
- 02:29 PM Bug #5655: Slow requests for 1h30 "currently waiting for missing objects"
- Cephdrop now has under /5655/
1) osd.0 tell debug-ms/osd/filestore 20 from earlier on (
2) ceph --asok $asok log du... - 02:05 PM Bug #5655 (Resolved): Slow requests for 1h30 "currently waiting for missing objects"
- This happened today after I upgraded & restarted osd.0 to next (d1501938f5d07c067d908501fc5cfe3c857d7281). I remember...
- 10:23 PM Revision 1a844112 (ceph): msgr: fix a typo/goto-cross from dd4addef2d
- We didn't build or review carefully enough!
Signed-off-by: Greg Farnum <greg@inktank.com>
Reviewed-by: Sage Weil <sa... - 10:09 PM CephFS Bug #5367: multiclient tests: kernel mount gets EPERM
- Zheng Yan wrote:
> kclient and MDS never return -EACCES. was ior executed with root privilege?
not root... - 10:09 PM CephFS Bug #5367: multiclient tests: kernel mount gets EPERM
- saw similar error on ceph-fuse:...
- 10:06 PM Bug #5644 (Resolved): radosgw-admin test fails on next
- 10:05 PM Bug #5505 (Resolved): mon health does not reflect slow requests
- 09:50 PM Revision ea1c6234 (ceph): Merge pull request #441 from ceph/wip-5626
- msgr fixes for lossless peer sessions
Reviewed-by: Greg Farnum <greg@inktank.com> - 09:39 PM Revision dd4addef (ceph): msgr: maintain list of accepting pipes
- New pipes exist in a sort of limbo before we know who the peer is and
add them to rank_pipe. Keep a list of them in ... - 09:39 PM Revision 687fe888 (ceph): msgr: close accepting_pipes from mark_down_all()
- We need to catch these pipes too, particularly when doing a rebind(),
to avoid them leaking through.
Signed-off-by: ... - 09:39 PM Revision ecab4bb9 (ceph): msg/Pipe: hold pipe_lock during important parts of accept()
- Previously we did not bother with locking for accept() because we were
not visible to any other threads. However, we... - 09:39 PM Revision 16568d9e (ceph): msg/Pipe: a bit of additional debug output
- Signed-off-by: Sage Weil <sage@inktank.com>
- 09:39 PM Revision 57bd6fd5 (ceph): osd: make 'from dead osd' message more informative
- I thought I saw some weirdness here.
Signed-off-by: Sage Weil <sage@inktank.com> - 09:38 PM Revision 994e2bf2 (ceph): msgr: adjust nonce on rebind()
- We can have a situation where:
- we have a pipe to a peer
- pipe goes to standby (on peer)
- we rebind to a new p... - 09:36 PM Revision 07a0860a (ceph): msgr: mark_down_all() after, not before, rebind
- If we are shutting down all old connections and binding to new ports,
we want to avoid a sequence like:
- close all... - 09:34 PM Revision 9f1c2726 (ceph): msg/Pipe: avoid creating empty out_q entry
- We need to maintain the invariant that all sub queues in out_q are never
empty. Fix discard_requeued_up_to() to avoi... - 09:34 PM Revision ad548e72 (ceph): msg/Pipe: unlock msgr->lock earlier in accept()
- Small cleanup. Nothing needs msgr->lock for the previously larger
window.
Signed-off-by: Sage Weil <sage@inktank.com> - 09:34 PM Revision 579d858a (ceph): msg/Pipe: assert lock is held in various helpers
- These all require that we hold pipe_lock.
Signed-off-by: Sage Weil <sage@inktank.com> - 09:12 PM Revision 0ebf23ce (ceph): ceph_mon: obtain backup monmap if store is marked with 'force_sync'
- Signed-off-by: Joao Eduardo Luis <joao.luis@inktank.com>
Reviewed-by: Sage Weil <sage@inktank.com> - 09:05 PM Revision c395687f (ceph): radosgw-admin: adapt task to recent changes
- Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
- 07:10 PM Bug #5154: osd/SnapMapper.cc: 270: FAILED assert(check(oid))
- 2013-07-17 15:14:10.588979 7f87c4c3a700 15 filestore(/var/lib/ceph/osd/ceph-0) _split_collection 3.0_head bits: 3
20... - 02:39 PM Bug #5154 (Need More Info): osd/SnapMapper.cc: 270: FAILED assert(check(oid))
- 07:07 PM CephFS Bug #5658: kcepht: warning: fs/ceph/inode.c:1000 ceph_fill_trace+0x760/0x900 [ceph]()
- the warning and hang should be related because the kclient ignores caps in the inode trace. the MDS waits forever whe...
- 06:49 PM CephFS Bug #5658: kcepht: warning: fs/ceph/inode.c:1000 ceph_fill_trace+0x760/0x900 [ceph]()
- Server::handle_client_lookup_ino() calls Server::reply_request() with a non-NULL tracedn. I think tracedn should be N...
- 06:05 PM CephFS Bug #5658 (Resolved): kcepht: warning: fs/ceph/inode.c:1000 ceph_fill_trace+0x760/0x900 [ceph]()
- i noticed this in dmesg:...
- 05:47 PM CephFS Bug #5649: smbtorture test gets ebusy on kclient umount
- ubuntu@teuthology:/var/lib/teuthworker/archive/teuthology-2013-07-17_01:01:03-fs-next-testing-basic/70473
- 05:07 PM Bug #5624: osd: prepare_to_stop() segfaults if we get a signal during startup
- ubuntu@teuthology:/a/sage-2013-07-17_16:50:00-osdleaks-wip-osd-leaks-testing-basic/70827
- 04:44 PM Revision d1501938 (ceph): mon/OSDMonitor: make 'osd pool mksnap ...' not expose uncommitted state
- We were returning success without waiting if the pending pool state had
the snap.
Signed-off-by: Sage Weil <sage@ink... - 04:36 PM Revision 56c5b835 (ceph): qa/workunits/cephtest/test.sh: put 'osd ls' before any 'osd create' tests
- A monc/mon connection fault or the dup command test flag may mean an extra
osd id is created that we isn't actually u... - 04:26 PM Revision ad9a1044 (ceph): mon: MonCommands: remove obsolete 'sync status' command
- Obsoleted by the sync refactor from
da0aff28ab478bcc3136715f92bc1af8d4b403c1
Signed-off-by: Joao Eduardo Luis <joao.... - 03:55 PM Bug #5432 (Resolved): msgr: bad locking mark_down_all
- ooof, just looked and it's gonna be like 20 patches. i'm thinking we skip it, unless the more important reconnect st...
- 03:26 PM Bug #5432: msgr: bad locking mark_down_all
- It's been 23 days; time to backport or decide we don't need to?
- 03:53 PM Feature #5648 (In Progress): mon should use 'module'/'perm' string in MonCommands.h for caps chec...
- 03:01 PM RADOS Feature #5657 (Resolved): monitor: deal with bad crush maps more gracefully
- ...
- 02:52 PM Bug #5517 (Pending Backport): osd: stuck peering on cuttlefish
- 10:59 AM Bug #5517: osd: stuck peering on cuttlefish
- ubuntu@teuthology:/a/teuthology-2013-07-17_01:00:18-rados-next-testing-basic/70345
- 02:52 PM Bug #5626 (Resolved): msg/Pipe.cc: 1012: FAILED assert(m)
- 02:39 PM Bug #5630 (Need More Info): osd/ReplicatedPG.cc: 7089: FAILED assert(r >= 0) in scan_range, cuttl...
- 12:05 PM Bug #5630: osd/ReplicatedPG.cc: 7089: FAILED assert(r >= 0) in scan_range, cuttlefish
- LFNIndex::list_subdirs and LFNIndex::list_objects cannot return -ENODATA (no hashed names)
=> HashIndex::get_path_co... - 11:43 AM Bug #5630: osd/ReplicatedPG.cc: 7089: FAILED assert(r >= 0) in scan_range, cuttlefish
- Might indicate a missing xattr on one of the collection subdirs.
- 11:40 AM Bug #5630: osd/ReplicatedPG.cc: 7089: FAILED assert(r >= 0) in scan_range, cuttlefish
- core says r was -61 (ENODATA).
- 02:39 PM Bug #5656 (Resolved): ./osd/OSDMap.h: 536: FAILED assert(pools.count(pg.pool()))
ceph version 0.66-672-gd150193 (d1501938f5d07c067d908501fc5cfe3c857d7281)
1: (OSD::_make_pg(std::tr1::shared_ptr...- 02:29 PM CephFS Bug #3601: client: With multiple clients, file remove doesn't free up space
- No progress yet; Inktank isn't able to devote much time to the filesystem right now.
See also #5039, which is the mo... - 06:47 AM CephFS Bug #3601: client: With multiple clients, file remove doesn't free up space
- Is there any progress on this? I've tried CephFS with actual dev-version 0.66 (b6b48dbefadb39419f126d0e62c035e0109060...
- 11:51 AM rbd Bug #5426: librbd: mutex assert in perfcounters::tinc in librbd::AioCompletion::complete()
- okay, not cuttlefish.
ubuntu@teuthology:/a/teuthology-2013-07-17_01:00:47-rbd-next-testing-basic/70375... - 11:26 AM devops Bug #5405: ceph-deploy: transient pushy exception on install
- once this his an upstream we need to build a package.
- 11:13 AM Bug #5631: osd/ReplicatedPG.cc: 3036: FAILED assert(iter)
- get_omap_iterator relies on lfn_find, while getattr relies on lfn_open. The latter might return attrs from an hobjec...
- 10:46 AM Bug #5631 (In Progress): osd/ReplicatedPG.cc: 3036: FAILED assert(iter)
- 10:38 AM Bug #5507: osd: ENOENT on clone
- This could be explained by a failure to resurrect a parent pg causing recovery on an hobject to perform writes on an ...
- 10:38 AM Bug #5392: osd: unfound objects from thrashing
- This could be explained by a failure to resurrect a parent pg causing recovery on an hobject to perform writes on an ...
- 10:37 AM Bug #5625 (Pending Backport): osd: scrub finds 20 missing objects on cuttlefish
- 10:37 AM Bug #5625: osd: scrub finds 20 missing objects on cuttlefish
- This could be explained by a failure to resurrect a parent pg causing recovery on an hobject to perform writes on an ...
- 10:36 AM Bug #5473: osd/ReplicatedPG.cc: 1379: FAILED assert(0) in trim_object() on master, cuttlefish
- This could be explained by a failure to resurrect a parent pg causing recovery on an hobject to perform writes on an ...
- 10:22 AM Bug #5069 (Resolved): monitor crashed during mon thrash in nightlies
- this should have been fixed with the recent mon sync and other changes
- 09:42 AM Documentation #5653 (Resolved): Pool quota needs documentation created
- There is currently none that I have been able to locate.
- 09:42 AM Bug #5652 (Resolved): mon: occasional clock_skew test failure
- ...
- 09:07 AM rgw Documentation #5651: rgw: secret features need documentation
- Those flags are part of radosgw-admin utility.
- 09:05 AM rgw Documentation #5651 (Resolved): rgw: secret features need documentation
- There may be other new features, but specifically, --gen-secret and --secret should be added to man pages, along with...
- 06:07 AM Revision 39654aa0 (ceph): Merge pull request #423 from dalgaaf/wip-da-SCA-pylint
- ceph.in, ceph_argparse.py: Fix some pylint issues
Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>
Reviewed-by... - 04:03 AM Revision e8c58d3e (ceph): rados: use 'ms inject intenral delays' during thrashing
- This tries to help hit unusual races in the messenger.
Signed-off-by: Sage Weil <sage@inktank.com> - 12:41 AM Revision 053ba3a2 (ceph): Merge branch 'master' of github.com:ceph/teuthology
- 12:41 AM Revision a24aaa41 (ceph): added overrides for ceph-deploy
- Signed-off-by: tamil <tamil.muthamizhan@inktank.com>
- 12:33 AM Revision 884fa2fc (ceph): OSD::_try_resurrect_pg: fix cur/pgid confusion
- This bug prevented resurrection of ancestor pgs where
necessary.
Fixes: #5269
This may result in pg A being created ... - 12:21 AM Revision 7e16b72d (ceph): mon/AuthMonitor: make 'auth del ...' idempotent
- Signed-off-by: Sage Weil <sage@inktank.com>
- 12:16 AM Revision f129d174 (ceph): qa/workunits/cephtool/test.sh: mds cluster_down/up are idempotent
- As of d45429b81ab9817284d6dca98077cb77b5e8280f; fix the test.
Signed-off-by: Sage Weil <sage@inktank.com> - 12:16 AM Revision b15513fc (ceph): workunit: set CEPH_CLI_TEST_DUP_COMMAND
- This will make the CLI do every mon command twice and make sure they both
succeed. This catches problems with mon co... - 12:14 AM Revision 51352677 (ceph): added conf section to ceph-deploy task
- Signed-off-by: tamil <tamil.muthamizhan@inktank.com>
07/16/2013
- 11:58 PM Revision f2fa01e2 (ceph): ceph: send successful commands twice with CEPH_CLI_TEST_DUP_COMMAND
- Monitor commands need to be idempotent. This helps us test this by
simply issuing any successful command a second ti... - 11:26 PM Revision d45429b8 (ceph): mon/MDSMonitor: make 'mds cluster_{up,down}' idempotent
- Signed-off-by: Sage Weil <sage@inktank.com>
- 11:10 PM Revision 9c4a0307 (ceph): osdmaptool: fix cli tests
- From the HASHPSPOOL change in acbc2f0bc0b4266125403aebb28e6e3a2365394d.
Signed-off-by: Sage Weil <sage@inktank.com> - 10:52 PM Revision 1ec26b8e (ceph): Merge branch 'wip-ceph-disk' into next
- Reviewed-by: Gary Lowell <gary.lowell@inktank.com>
Tested-by: Jing Yuan Luke <jyluke@gmail.com> - 10:51 PM Revision 35d3f2d8 (ceph): ceph-disk: refactor list_[all_]partitions
- Make these methods work in terms of device *names*, not paths, and fix up
the only direct list_partitions() caller to... - 10:51 PM Revision 3359aaed (ceph): ceph-disk: use get_dev_name() helper throughout
- This is more robust than the broken split trick.
Signed-off-by: Sage Weil <sage@inktank.com> - 10:51 PM Revision 5b031e10 (ceph): ceph-disk: reimplement is_partition() using /sys/block
- Signed-off-by: Sage Weil <sage@inktank.com>
- 10:51 PM Revision 2ea8fac4 (ceph): ceph-disk: use /sys/block to determine partition device names
- Not all devices are basename + number; some have intervening character(s),
like /dev/cciss/c0d1p2.
Signed-off-by: Sa... - 10:51 PM Revision e0401591 (ceph): ceph-disk: add get_dev_name, path helpers
- Signed-off-by: Sage Weil <sage@inktank.com>
- 10:36 PM Revision d656aed5 (ceph): mon/OSDMonitor: fix typo
- From 5eac38797d9eb5a59fcff1d81571cff7a2f10e66
Signed-off-by: Sage Weil <sage@inktank.com> - 10:35 PM Revision d90683fd (ceph): osd/OSDMonitor: make 'osd pool rmsnap ...' not racy/crashy
- Ensure that the snap does in fact exist before we try to remove it. This
avoids a crash where a we get two dup rmsna... - 10:30 PM Revision 1999fa2c (ceph): ObjectStore: add omap_rmkeyrange to dump
- Signed-off-by: Samuel Just <sam.just@inktank.com>
Reviewed-by: Sage Weil <sage@inktank.com> - 10:30 PM Revision 44c39177 (ceph): OSD: add perfcounter tracking messages delayed pending a map
- Signed-off-by: Samuel Just <sam.just@inktank.com>
Reviewed-by: Sage Weil <sage@inktank.com> - 10:29 PM Revision d9e0e789 (ceph): FileStore: add a perf counter for time spent acquiring op queue throttle
- Signed-off-by: Samuel Just <sam.just@inktank.com>
Reviewed-by: Sage Weil <sage@inktank.com> - 10:24 PM Revision 62d9983b (ceph): Merge branch 'wip-4779' into next
- Reviewed-by: Sage Weil <sage@inktank.com># Please enter a commit message to explain why this merge is necessary,
- 10:17 PM Revision c449a8b3 (ceph): Merge pull request #439 from yehudasa/wip-rgw-next
- rgw: quiet down ECANCELED on put_obj_meta()
Reviewed-by: Greg Farnum <greg@inktank.com> - 10:14 PM Revision 39bffac6 (ceph): ceph-disk: rely on /dev/disk/by-partuuid instead of special-casing jour...
- This was necessary when ceph-disk-udev didn't create the by-partuuid (and
other) symlinks for us, but now it is fragi... - 10:14 PM Revision 4d9d0ffb (ceph): mon/OSDMonitor: return error if we can't set the new bucket's name
- Signed-off-by: Sage Weil <sage@inktank.com>
Reviewed-by: Dan Mick <dan.mick@inktank.com> - 10:13 PM Revision 466d0f5f (ceph): crush: return EINVAL on invalid name from {insert,update,create_or_move...
- Signed-off-by: Sage Weil <sage@inktank.com>
Reviewed-by: Dan Mick <dan.mick@inktank.com> - 10:13 PM Revision 93fc07c1 (ceph): crush: add is_valid_crush_name() helper
- [A-Za-z0-9-_.]+
Signed-off-by: Sage Weil <sage@inktank.com>
Reviewed-by: Dan Mick <dan.mick@inktank.com> - 10:08 PM Revision 5eac3879 (ceph): mon: OSDMonitor: only thrash and propose if we are the leader
- 'thrash_map' is only set if we are the leader, so we would thrash and
propose the pending value if we are the leader.... - 10:05 PM Messengers Bug #5508 (Can't reproduce): msg/SimpleMessenger.cc: 230: FAILED assert(!cleared)
- 10:03 PM Bug #5466 (Can't reproduce): raring tarball builds mysteriously fail make check
- this hasn't happened since this has been migrated to lxc
- 09:55 PM Bug #5628 (Can't reproduce): admin_socket test failure on cuttlefish
- 09:52 PM Revision b19ec576 (ceph): mon/MDSMonitor: make 'ceph mds remove_data_pool ...' idempotent
- Signed-off-by: Sage Weil <sage@inktank.com>
- 09:49 PM Revision f06a124a (ceph): mon/OSDMonitor: send_to_waiting() in on_active()
- The send_latest() helper may put a message in the waiting_for_map list
if we are not readable, but currently send_to_... - 09:49 PM Revision ba28c7cc (ceph): mon/OSDMonitor: clean up waiting_for_map messages on shutdown
- Do not leak these.
Fixes: #5643
Signed-off-by: Sage Weil <sage@inktank.com>
Reviewed-by: Greg Farnum <greg@inktank.c... - 09:41 PM Bug #5505: mon health does not reflect slow requests
- ideas:
- pg_stat_t could include a concise histogram of op latencies
- once the ops *do* complete, the pg stats w... - 09:08 PM Revision 72d4351e (ceph): rgw: quiet down ECANCELED on put_obj_meta()
- Fixes: #5439
ECANCELED there means that we lost in a race to write the object. We
should treat it as a successful wr... - 09:07 PM Bug #5269: osd: EEXIST on mkcoll
- 01:21 PM Bug #5269: osd: EEXIST on mkcoll
- Andrey saw this on latest cuttlefish, see #5637
- 10:49 AM Bug #5269: osd: EEXIST on mkcoll
- ...
- 08:44 PM Bug #5644: radosgw-admin test fails on next
- Looking at it some more. As part of the geo metadata sync I changed it so that when we link a bucket to a user we don...
- 12:57 PM Bug #5644: radosgw-admin test fails on next
- That one fails on radosgw-admin bucket link operation. We made some changes to the way user is linked with the bucket...
- 11:05 AM Bug #5644 (In Progress): radosgw-admin test fails on next
- 10:46 AM Bug #5644 (Resolved): radosgw-admin test fails on next
- last 2 days
teuthology-2013-07-16_01:00:53-rgw-next-testing-basic/68506 - 08:41 PM Revision acbc2f0b (ceph): osd: do not enable HASHPSPOOL pool feature by default
- This was added in kernel 3.9 and should not yet be enabled by default.
Signed-off-by: Sage Weil <sage@inktank.com> - 08:15 PM Revision 64379e70 (ceph): ceph-disk: rely on /dev/disk/by-partuuid instead of special-casing jour...
- This was necessary when ceph-disk-udev didn't create the by-partuuid (and
other) symlinks for us, but now it is fragi... - 08:09 PM Revision 3706dbbf (ceph): PendingReleaseNotes: formatted ceph CLI output and ceph-rest-api
- Signed-off-by: Dan Mick <dan.mick@inktank.com>
- 06:53 PM Bug #5634 (Fix Under Review): auth startup reports "ObjectNotFound" when keyring file is unreadable
- 05:33 PM Revision 363d5428 (ceph): mon: Monitor: StoreConverter: clearer debug message on 'needs_conversio...
- The previous debug message outputted the function's name, as often our
functions do. This was however a source of be... - 05:33 PM Revision 0ea89760 (ceph): mon: Monitor: do not reopen MonitorDBStore during conversion
- We already open the store on ceph_mon.cc, before we start the conversion.
Given we are unable to reproduce this every... - 05:33 PM Revision ad1392f6 (ceph): mon: Monitor: StoreConverter: clearer debug message on 'needs_conversio...
- The previous debug message outputted the function's name, as often our
functions do. This was however a source of be... - 05:32 PM Revision e752c40c (ceph): mon: Monitor: StoreConverter: sanitize 'store' pointer on init
- We are supposed to have umount'ed the store and set the pointer to NULL.
We should not tolerate any other case on ini... - 05:31 PM Revision 036e6739 (ceph): mon: Monitor: do not reopen MonitorDBStore during conversion
- We already open the store on ceph_mon.cc, before we start the conversion.
Given we are unable to reproduce this every... - 05:22 PM Bug #5626 (Fix Under Review): msg/Pipe.cc: 1012: FAILED assert(m)
- 11:05 AM Bug #5626 (In Progress): msg/Pipe.cc: 1012: FAILED assert(m)
- 05:22 PM Bug #5517 (Fix Under Review): osd: stuck peering on cuttlefish
- 04:33 PM Revision 38691e7f (ceph): Merge pull request #438 from yehudasa/wip-rgw-next
- Fix an issue with bucket placements and with listing on new installations.
Reviewed-by: Greg Farnum <greg@inktank.com> - 03:54 PM devops Bug #5345 (Pending Backport): ceph-disk: handle less common device names
- 03:49 PM devops Bug #5345: ceph-disk: handle less common device names
- it sounds like the partition table didn't actually get written, or some other problem between your kernel and parted....
- 03:52 PM Bug #5637: OSD crash upon pool creation
- Sorry if it was unclear: dedicated parameter(s) for path of the omap and meta collections.
- 03:39 PM Bug #5637 (Rejected): OSD crash upon pool creation
- Andrey Korolyov wrote:
> Please excuse me, I am horribly wrong - we`re using our own patch which is affecting collec... - 03:33 PM Bug #5637: OSD crash upon pool creation
- Please excuse me, I am horribly wrong - we`re using our own patch which is affecting collections too and it was a gen...
- 01:30 PM Bug #5637: OSD crash upon pool creation
- Andrey, what version is the osd running? the sha1 in the log (43c453982c37a55defc5d9ffcfd8cf8a60755f24) is not in ce...
- 01:23 PM Bug #5637: OSD crash upon pool creation
- 01:23 PM Bug #5637: OSD crash upon pool creation
- actually, no:...
- 01:20 PM Bug #5637 (Duplicate): OSD crash upon pool creation
- 01:19 PM Bug #5637: OSD crash upon pool creation
- this looks like a dup of #5269
- 01:47 AM Bug #5637: OSD crash upon pool creation
- Here you go, log was truncated before pool creation to avoid excessive gzip attachment.
- 03:45 PM CephFS Bug #5031 (Resolved): mds/MDCache.cc: 5221: FAILED assert(reconnected_snaprealms.empty())
- 03:33 PM Bug #5336 (Can't reproduce): osd crash triggered by 'rbd rm ...'
- 03:19 PM rgw Bug #5439 (Resolved): rgw: 500 returned on s3readwrite
- Fixed, commit:72d4351ea5a470051e428ffc5531acfc7d1c7b6f
- 03:18 PM Bug #4779 (Resolved): The ceph command and crushtool have differing views on valid characters for...
- 03:06 PM Revision d873c292 (ceph): rgw/rgw_rest_s3.cc: fix dereferencing after deallocation
- Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>
- 03:06 PM Revision e45a9bb1 (ceph): rgw/rgw_op.cc: reduce scope of local 'ret' integer variable
- Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>
- 03:05 PM Revision c484150b (ceph): rgw/rgw_rados.cc: use static_cast instead of C-Style cast
- Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>
- 02:50 PM Bug #5643 (Resolved): mon: new memory leaks
- 12:34 PM Bug #5643: mon: new memory leaks
- 10:45 AM Bug #5643 (Resolved): mon: new memory leaks
- ubuntu@teuthology:/a/teuthology-2013-07-16_01:00:53-rgw-next-testing-basic/68508 leaked an MOSDAlive, lots of other f...
- 02:23 PM devops Bug #5405: ceph-deploy: transient pushy exception on install
- I've opened a pull request for pushy (https://github.com/axw/pushy/pull/42) that should solve this problem.
What h... - 09:59 AM devops Bug #5405: ceph-deploy: transient pushy exception on install
- This is very hard to replicate, but after adding a bit of verbosity in pushy, I managed to narrow it down to the Push...
- 02:00 PM CephFS Bug #5649 (Can't reproduce): smbtorture test gets ebusy on kclient umount
- ...
- 01:34 PM Feature #5648 (Resolved): mon should use 'module'/'perm' string in MonCommands.h for caps checking
- 'perm' was added to MonCommands.h for controlling GET vs PUT endpoints in ceph-rest-api;
however, it's also aimed at... - 01:12 PM rbd Bug #5647 (Resolved): krbd: EBlACKLIST osd reply resulting in an oops on 3.9
- ...
- 12:20 PM Revision cc49d3fe (ceph): rgw/rgw_rest.cc: fix malloc memory leak
- Free with malloc allocated memory in error case before return.
Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de> - 12:20 PM Revision 7a2e8a6f (ceph): rgw/rgw_bucket.cc: return result of function calls
- Return the result of rgw_(un)link_bucket()/ from
RGWBucketMetadataHandler::put() to signal errors correctly to
the fu... - 12:19 PM Revision 7769b01d (ceph): rgw_bucket.cc: don't return bool if return value is int
- Fix minor style issue: if the function return value is an integer
don't return bool values.
Signed-off-by: Danny Al-... - 12:18 PM Revision 64b512c0 (ceph): rgw/rgw_bucket.cc: remove unused local variable
- Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>
- 12:18 PM Revision a937a12d (ceph): rgw_admin.cc: fix same expression on both sides of '||'
- Remove twice listed checks, get checks in alphabetical order,
added comment above the checks to point out to keep it ... - 12:18 PM Revision 5ea4c5e1 (ceph): cls_rgw_client.cc: reduce scope of local variable
- Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>
- 10:53 AM rbd Bug #5391: krbd: crash in rbd_obj_request_create -> strlen
- again in ubuntu@teuthology:/a/teuthology-2013-07-16_01:01:08-kernel-next-testing-basic/68607
dump attached - 10:36 AM rbd Feature #5635: RBD Kernel Client Incompatibility leads to Kernel Hang
- - kenrel should not hang
- we should probably not make pspool the default (and it should happen *after* we switch to... - 10:35 AM Bug #5640 (Resolved): mon: FAILED assert(!db->create_and_open(std::cerr)) on 0.61.4
- 08:50 AM Bug #5640 (Fix Under Review): mon: FAILED assert(!db->create_and_open(std::cerr)) on 0.61.4
- pushed wip-5640
- 08:25 AM Bug #5640: mon: FAILED assert(!db->create_and_open(std::cerr)) on 0.61.4
- *sigh* we already have opened the db at that point. I'll push the fix as soon as I finish reviewing the surrounding ...
- 08:02 AM Bug #5640: mon: FAILED assert(!db->create_and_open(std::cerr)) on 0.61.4
- Yeah, from the strace it sure looks a lot like a race between an open(store.db/LOCK), and a fcntl(F_SETLK, F_UNLCK) f...
- 07:39 AM Bug #5640: mon: FAILED assert(!db->create_and_open(std::cerr)) on 0.61.4
- it sounds like *any* lines like...
- 07:17 AM Bug #5640 (Resolved): mon: FAILED assert(!db->create_and_open(std::cerr)) on 0.61.4
- I've been unable to reproduce this, but it kept happening on a user's cluster while trying to upgrade from bobtail to...
- 10:06 AM Feature #5642: Show usable space via CLI/API
- I am thinking in the context of alerts, knowing when any pool in the cluster in running out of usable space has value...
- 10:00 AM Feature #5642: Show usable space via CLI/API
- The reason we haven't done this yet is because it's a hard UI problem. We need to define how to present the informati...
- 09:42 AM Feature #5642 (New): Show usable space via CLI/API
- In addition to showing the total raw disk capacity available, it should be possible to show the usable disk capacity ...
- 09:41 AM rbd Bug #3493: libvirt: storage pool does not work with cephx disabled
- that was quick :)
- 06:21 AM rbd Bug #3493 (Resolved): libvirt: storage pool does not work with cephx disabled
- I just got merged upstream
http://libvirt.org/git/?p=libvirt.git;a=commitdiff;h=d58c8478443d49c6e702bbb2c56a567ef2... - 05:27 AM rbd Bug #3493: libvirt: storage pool does not work with cephx disabled
- I completely forgot about this one, my bad!
I just submitted a patch on the libvirt list to fix this. - 09:37 AM Bug #5627 (Resolved): mon: scrub mismatch - osdmap mismatch
- 02:13 AM Bug #5627: mon: scrub mismatch - osdmap mismatch
- Thanks.
- 07:33 AM Bug #5641 (Resolved): occasional crush_ops.sh failure
- teuthology-2013-07-15_20:00:18-rados-cuttlefish-testing-basic/68030...
- 04:07 AM Revision a81848ff (ceph): big: reenable big cluster
- small and medium are passing.
- 03:28 AM Feature #5421 (Resolved): mon: add formatter option for various mon commands
- 02:08 AM Revision e3d9084c (ceph): Created tasktest to test sequential and parallel tasks.
- Added sequential task and parallel task.
Changed _run_one_task to run_one_task (now called by new tasks too).
Fix #4... - 01:43 AM Revision 408014ee (ceph): rgw: handle ENOENT when listing bucket metadata entries
- Just return success (with an empty list)
Signed-off-by: Yehuda Sadeh <yehuda@inktank.com> - 12:04 AM Revision 47696d2b (ceph): calling mon destroy command after mds create
- Signed-off-by: tamil <tamil.muthamizhan@inktank.com>
07/15/2013
- 10:49 PM Revision eef44582 (ceph): rgw: fix bucket placement assignment
- When we set bucket.instance meta, we need to set
the correct bucket placement to the bucket (according to
the specifi... - 10:33 PM Revision 595c09df (ceph): messages/MClientReconnect: clear data when encoding
- The MClientReconnect puts everything in the data payload portion of
the message and nothing in the front portion. Th... - 10:15 PM Revision 39e5a2a4 (ceph): OSD: add config option for peering_wq batch size
- Large peering_wq batch sizes may excessively delay
peering messages resulting in unreasonably long
peering. This may... - 10:03 PM Revision 5c7b1e11 (ceph): fs: run everything on btrfs, not /
- 09:29 PM Revision b46930c9 (ceph): mon: make report pure json
- Put the crc in the status string and drop the header and footer. If users
want to capture it,
ceph report 2>&1 > fo... - 09:23 PM Revision 40a8bbdc (ceph): Merge remote-tracking branch 'gh/wip-mon-report' into next
- 09:05 PM Revision daf76723 (ceph): ceph: drop --threshold hack for 'pg dump_stuck'
- We can live with the incompatibility here; the hack is currently
not working anyway (see #5623).
Signed-off-by: Sage... - 08:48 PM Revision 4282971d (ceph): msg/Pipe: be a bit more explicit about encoding outgoing messages
- Signed-off-by: Sage Weil <sage@inktank.com>
- 08:47 PM Revision 314cf046 (ceph): messages/MClientReconnect: clear data when encoding
- The MClientReconnect puts everything in the data payload portion of
the message and nothing in the front portion. Th... - 08:46 PM Revision 6a524c35 (ceph): Merge pull request #436 from ceph/wip-mon-fixes
- Wip mon fixes
Reviewed-by: Greg Farnum <greg@inktank.com> - 08:42 PM Revision eac559f4 (ceph): mon: drop win_election() _reset() kludge and strengthen assertions
- This is only there for the benefit of win_standalone_election(), but it
doesn't need it, it clutters the code, and we... - 08:42 PM Revision 34f76bd9 (ceph): mon: set forwarded message recv stamp
- Set it to the stamp of the MForward that carried us. One could argue
we really want the original receive stamp on th... - 08:42 PM Revision 5c310107 (ceph): mon/Paxos: make all handle_accept paths go via out label
- Signed-off-by: Sage Weil <sage@inktank.com>
- 08:42 PM Revision c711203c (ceph): mon/Paxos: separate proposal commit from the end of the round
- Each commit should match with exactly one proposal; finish it when we
actually commit it and make sensible asserts.
... - 08:42 PM Revision 7666c33a (ceph): mon/PaxosService: assert not proposing in propose_pending
- Drop the useless active check after the assert, too.
Signed-off-by: Sage Weil <sage@inktank.com> - 08:42 PM Revision 1fc85acd (ceph): mon/PaxosService: consolidate resetting in restart()
- We had duplicated code in election_finished() and restart(), and it was
incomplete. Put it all in restart() only (th... - 08:42 PM Revision 3ea984cd (ceph): mon: once sync full is chosen, make sure we don't change our mind
- It is possible for a sequence like:
- probe
- first probe reply has paxos trim that indicates a full sync is
ne... - 08:42 PM Revision c67d50b8 (ceph): mon: set peon state to electing if other mons call an election
- Previously we would call mon->reset() and set various flags (like
exited_quorum timestamp), but the state would remai... - 08:20 PM Revision 2bf95e5a (ceph): Merge branch 'next'
- 07:54 PM Revision f1ce8d7c (ceph): mon: fix scrub vs paxos race: refresh on commit, not round completion
- Consider:
- paxos starts a commit N+1
- a majority of the peers ack it
- paxos::commit() writes N+1 it to disk
... - 07:20 PM Revision e996e9be (ceph): Merge pull request #437 from kri5/wip-fix-typo-rgw
- rgw: Fix typo in rgw_user.cc
Reviewed-by: Yehuda Sadeh <yehuda@inktank.com> - 06:36 PM Bug #5225 (Closed): arm: rbd fsx test failed on the arm set up
- The test now works using the correct image size.
- 06:17 PM Revision 1b525815 (ceph): Merge remote-tracking branch 'origin/wip-rgw-warnings' into next
- Conflicts:
src/test/test_rgw_admin_log.cc
src/test/test_rgw_admin_meta.cc
src/test/test_rgw_admin_opstate.cc
Revi... - 05:37 PM Revision 346d9f42 (ceph): rgw_admin: fix gc list encoding
- Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
- 05:37 PM Revision a722fb71 (ceph): rgw: fix bucket instance json encoding
- Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
- 05:24 PM Revision 791d51eb (ceph): Merge pull request #434 from gregsfortytwo/next
- test_rgw: fix a number of unsigned/signed comparison warnings
Reviewed-by: Yehuda Sadeh <yehuda@inktank.com> - 05:05 PM Revision 8c178e0d (ceph): mon: once sync full is chosen, make sure we don't change our mind
- It is possible for a sequence like:
- probe
- first probe reply has paxos trim that indicates a full sync is
ne... - 05:05 PM Revision 55ff523e (ceph): doc: Fixed link in Calxeda repo instruction.
- Signed-off-by: John Wilkins <john.wilkins@inktank.com>
- 05:02 PM Revision aa60f940 (ceph): mon: once sync full is chosen, make sure we don't change our mind
- It is possible for a sequence like:
- probe
- first probe reply has paxos trim that indicates a full sync is
ne... - 04:58 PM Revision 6d3e13f0 (ceph): rgw: fix more warnings
- test/test_rgw_admin_opstate.cc: In member function 'int admin_log::test_helper::extract_input(int, char**)':
warning:... - 04:39 PM Bug #5639 (Resolved): librados: rados_monitor_log can drop messages
- if we get a monitor session reset, and reconnect, we may have missed messages. a fix probably requires some tweaking...
- 03:43 PM Revision 1dfd06df (ceph): test_rgw: fix a number of unsigned/signed comparison warnings
- Signed-off-by: Greg Farnum <greg@inktank.com>
- 03:29 PM Bug #5637: OSD crash upon pool creation
- the osd log a fewlines up should tell us what the error was that it got back from the fs. can you attach the last few...
- 03:24 PM Bug #5637 (Rejected): OSD crash upon pool creation
- db2bb270e93ed44f9252d65d1d4c9b36875d0ea5 introduces some deadly behavior - every osd in the cluster crashing with SIG...
- 03:16 PM Bug #5623 (Resolved): pg dump_stuck command broke
- 02:57 PM rbd Bug #5636 (Resolved): krbd: crash in image refresh
- dumpall is attached...
- 02:15 PM CephFS Bug #4565: MDS/client: issue decoding MClientReconnect on MDS
- commit:314cf046b0b787ca69665e8751eab6fe7adb4037 in the next branch.
- 01:48 PM CephFS Bug #4565 (Resolved): MDS/client: issue decoding MClientReconnect on MDS
- 09:43 AM CephFS Bug #4565: MDS/client: issue decoding MClientReconnect on MDS
- Greg, please review.
- 02:08 PM rbd Feature #5635: RBD Kernel Client Incompatibility leads to Kernel Hang
- Client kernel version:
linux-image-3.8.0-19-generic from the ubuntu 64bit precise repository. - 02:06 PM rbd Feature #5635 (Resolved): RBD Kernel Client Incompatibility leads to Kernel Hang
- Using Ceph 0.66 and trying to mount RBD via the kernel client in kernels prior to 3.9 causes the following:
[ 486... - 01:50 PM Bug #5634: auth startup reports "ObjectNotFound" when keyring file is unreadable
- from_ceph_context() ignores any error and continues searching the list, and returns
only false on failure. It shoul... - 12:31 PM Bug #5634 (New): auth startup reports "ObjectNotFound" when keyring file is unreadable
- If the client.admin keyring is, say, 600/root, 'ceph -s' fails with...
- 11:47 AM Revision 01ec4f12 (ceph): rgw: Fix typo in rgw_user.cc
- Signed-off-by: Christophe Courtaut <christophe.courtaut@gmail.com>
- 11:39 AM Revision fe13f88c (ceph): doc: Modifies keyring path in radosgw config page.
- 11:17 AM Bug #5627: mon: scrub mismatch - osdmap mismatch
- this is probblay introduced while it was runnign older code (the sha1 you posted is only a day old). as the osdmaps ...
- 09:33 AM Bug #5627: mon: scrub mismatch - osdmap mismatch
- Monitors - 0.61.4-72-g6af0ed9 (6af0ed9bc4cc955f8c30ad9dc6e9095599f323d0)
osds - not restarted, hmm... looks like 0.6... - 08:58 AM Bug #5627: mon: scrub mismatch - osdmap mismatch
- what version exactly are you running?
- 04:54 AM Bug #5627 (Resolved): mon: scrub mismatch - osdmap mismatch
- Cluster (standard 3 nodes) HEALTH_OK. "ceph scrub" say "scrub mismatch". Mismatched numbers:
mon.0: ... osdmap=1682 ... - 10:12 AM Bug #5621 (Resolved): mon: paxos fc lc aren't update when sync_start clears store
- 09:20 AM Bug #5621: mon: paxos fc lc aren't update when sync_start clears store
- a281c9f88f0705c9427deb1ef786073933603de2 looks good to me. Reviewed-by on that patch.
- 09:46 AM rgw Support #5632 (New): Unable to use Swift when creating/updating custom container metadata
- Please see support issue 402: https://inktank.zendesk.com/agent/#/tickets/402
While using Swift, attempting to add... - 09:43 AM CephFS Bug #5418 (Need More Info): kceph: crash in remove_session_caps
- 09:26 AM Bug #5631 (Resolved): osd/ReplicatedPG.cc: 3036: FAILED assert(iter)
- ...
- 09:24 AM Bug #5154: osd/SnapMapper.cc: 270: FAILED assert(check(oid))
- ubuntu@teuthology:/a/teuthology-2013-07-15_01:00:16-rados-next-testing-basic/67483
- 09:15 AM Bug #5630 (Duplicate): osd/ReplicatedPG.cc: 7089: FAILED assert(r >= 0) in scan_range, cuttlefish
- ...
- 09:10 AM Bug #5628 (Can't reproduce): admin_socket test failure on cuttlefish
- 2 days in a row,
ubuntu@teuthology:/a/teuthology-2013-07-14_20:00:13-rados-cuttlefish-testing-basic/67114
and
... - 04:21 AM Revision 382ab4a2 (ceph): nfs: run osds on btrfs
- I hit an ext4 xattr hang; avoid that.
- 02:16 AM devops Bug #5345: ceph-disk: handle less common device names
- Hi Sage,
I had run partprobe and after probably 10 minutes, I still not seeing the second partition in /proc/parti...
07/14/2013
- 11:59 PM Revision 4ae00d78 (ceph): fs: disable ceph-deploy test
- Leave this to teh ceph-deploy suite
Signed-off-by: Sage Weil <sage@inktank.com> - 11:37 PM Revision cf9571c1 (ceph): test_rgw_admin_meta: fix warnings
- test/test_rgw_admin_meta.cc: In member function 'int admin_meta::test_helper::extract_input(int, char**)':
warning: t... - 11:36 PM Revision bc3088c9 (ceph): cls_rgw: fix warning
- cls/rgw/cls_rgw.cc: In function 'int get_obj_vals(cls_method_context_t, const string&, const string&, int, std::map, ...
- 11:33 PM Bug #5596: OSD keeps crashing while trying to rewight another OSD
- There was no crash dump the OSD process was just gone.But i've also seen one with heartbeat suicide.
- 04:02 PM Bug #5596 (Need More Info): OSD keeps crashing while trying to rewight another OSD
- 04:02 PM Bug #5596: OSD keeps crashing while trying to rewight another OSD
- i don't see an actual crash dump here.. is there a stack trace in the log indicating where it crashed? it's rare tha...
- 11:32 PM Revision d277edd2 (ceph): Merge branch 'wip-hadoop-doc'
- Signed-off-by: Noah Watkins <noahwatkins@gmail.com>
- 11:31 PM Revision a805958f (ceph): doc: document new hadoop config options
- Signed-off-by: Noah Watkins <noahwatkins@gmail.com>
- 11:31 PM Revision 6ef48c7e (ceph): doc: update Hadoop docs with plugin download
- Signed-off-by: Noah Watkins <noahwatkins@gmail.com>
- 11:24 PM Revision 9baa6680 (ceph): Makefile: cls_rgw needs cls_rgw_types linked in now too
- Signed-off-by: Sage Weil <sage@inktank.com>
- 11:20 PM Revision 6b7d9f2b (ceph): mon: include some (basic) auth info in report
- Nothing privileged!
Signed-off-by: Sage Weil <sage@inktank.com> - 11:16 PM Revision 56c36fa9 (ceph): mon: include paxos info in report
- Signed-off-by: Sage Weil <sage@inktank.com>
- 11:12 PM Revision bf7c4022 (ceph): mon: move quorum out of monmap
- Signed-off-by: Sage Weil <sage@inktank.com>
- 11:12 PM Revision 89ced1c2 (ceph): mon: include service first_committed in report
- Signed-off-by: Sage Weil <sage@inktank.com>
- 10:32 PM Revision e22f46fe (ceph): Merge pull request #429 from llonchj/hypertable_changes
- Hypertable changes
- 10:25 PM Bug #5517 (In Progress): osd: stuck peering on cuttlefish
- triggered the assert with ubuntu@teuthology:/a/sage-bug-5626-b/67249 on osd.2
osd.0 marks down osd.2 (looks like r... - 09:41 PM Bug #5512 (Resolved): mon: missing full osdmaps after sync
- 09:39 PM devops Bug #5345: ceph-disk: handle less common device names
- Hi Luke- have you had a chance to see if a 'partprobe /dev/cciss/c0d1' makes the journal parition appear?
- 09:34 PM Bug #5581 (Resolved): fix big suite failures
- 09:34 PM Bug #5401: cuttlefish osd recovery slow
- sorry, just can't prioritize this right now. not getting similar reports from other users, and need to focus on maki...
- 09:30 PM Bug #5626: msg/Pipe.cc: 1012: FAILED assert(m)
- - we are not detecting resets
- down osd is talking to a surviving osd, repeatedly marked down
- at some point, we ... - 04:30 PM Bug #5626 (Resolved): msg/Pipe.cc: 1012: FAILED assert(m)
- this is in the reconnect resent seq code that i just changed.
job was... - 09:28 PM Revision 629a9d77 (ceph): Use mon_host instead of mon_addr in ceph_conf
- Signed-off-by: Jordi Llonch <llonchj@gmail.com>
- 09:26 PM Revision 9ca4733d (ceph): hypertable recent version prototyping includes bool verify in length an...
- Signed-off-by: Jordi Llonch <llonchj@gmail.com>
- 04:28 PM Bug #5473: osd/ReplicatedPG.cc: 1379: FAILED assert(0) in trim_object() on master, cuttlefish
- 0> 2013-07-14 04:56:57.644145 7f12094ad700 -1 osd/ReplicatedPG.cc: In function 'ReplicatedPG::RepGather* Replica...
- 03:55 PM CephFS Bug #4565 (Fix Under Review): MDS/client: issue decoding MClientReconnect on MDS
- 08:33 AM CephFS Bug #5418: kceph: crash in remove_session_caps
- ...
- 08:28 AM Bug #5495 (Can't reproduce): ceph-mon and minus character in hostname
- Joao Luis wrote:
> Sage Weil wrote:
> > that second strace shows it hitting an unrelated assert on db->create_and_o... - 06:02 AM Bug #5495: ceph-mon and minus character in hostname
- Sage Weil wrote:
> that second strace shows it hitting an unrelated assert on db->create_and_open()... joao?
We'... - 08:18 AM rbd Bug #5426: librbd: mutex assert in perfcounters::tinc in librbd::AioCompletion::complete()
- ...
- 08:07 AM Bug #5625 (Resolved): osd: scrub finds 20 missing objects on cuttlefish
- job was...
- 04:11 AM Revision 6af0ed9b (ceph): mon: do not scrub if scrub is in progress
- This prevents an assert from unexpected scrub results from the previous
scrub on the leader.
Backport: cuttlefish
Si... - 04:11 AM Revision 5642155b (ceph): messages/MPGStats: do not set paxos version to osdmap epoch
- The PaxosServiceMessage version field is meant for client-coordinated
ordering of messages when switching between mon... - 04:09 AM Revision 06c65988 (ceph): osd/OSDmap: fix OSDMap::Incremental::dump() for new pool names
- The name is always present when pools are created, but not when they are
modified. Also, a name may be present with ...
07/13/2013
- 09:56 PM Revision df45b167 (ceph): Makefile: build cls_rgw even if we're not building radosgw
- Signed-off-by: Sage Weil <sage@inktank.com>
- 09:51 PM Revision 35ef8732 (ceph): Makefile: fix cls_rgw linkage
- Broken by 0c83b5fec233b7fc63205233403e7df32139d039.
Signed-off-by: Sage Weil <sage@inktank.com> - 09:31 PM Revision 65824071 (ceph): mon/PaxosService: prevent reads until initial service commit is done
- Do not process reads (or, by PaxosService::dispatch() implication, writes)
until we have committed the initial servic... - 09:19 PM Revision 5c3ff337 (ceph): client: send all request put's through put_request()
- Make sure all MetaRequest reference put's go through the same path that
releases inode references, including all of t... - 09:19 PM Revision 1df78ad7 (ceph): client: fix remaining Inode::put() caller, and make method psuedo-private
- Not sure I can make this actually private and make Client::put_inode() a
friend method (making all of Client a friend... - 09:19 PM Revision fea024cc (ceph): client: use put_inode on MetaRequest inode refs
- When we drop the request inode refs, we need to use put_inode() to ensure
they get cleaned up properly (removed from ... - 09:11 PM Revision 5d596a46 (ceph): ceph_manager: drop -t arg prefix for pg dump_stuck
- This is no longer needed, and ugly to support.
Signed-off-by: Sage Weil <sage@inktank.com> - 09:02 PM Revision cda17fd1 (ceph): Makefile: fix cls_refcount linkage
- Broken by d0bee5d85c1c862639450bad69c0ad20a98ef5c9.
Fixes: #5622
Signed-off-by: Sage Weil <sage@inktank.com> - 08:29 PM Revision fb95b631 (ceph): big: enable medium cluster
- Signed-off-by: Sage Weil <sage@inktank.com>
- 08:28 PM Revision 9be09798 (ceph): fs: disable maxfid smbtorture test
- This fails occasionally, unsure why.
Signed-off-by: Sage Weil <sage@inktank.com> - 08:26 PM Revision 8d9165a0 (ceph): qa/workunits/rbd/simple_big: fix unmap
- Signed-off-by: Sage Weil <sage@inktank.com>
- 08:25 PM Revision 15e3c9a4 (ceph): qa/workunits/fs/test_o_trunc.sh: fix path
- Signed-off-by: Sage Weil <sage@inktank.com>
- 08:00 PM Revision efe5b67b (ceph): mon/Paxos: bootstrap peon too if monmap updates
- If we get a monmap update, the leader bootstraps. Peons should do the
same.
Signed-off-by: Sage Weil <sage@inktank.... - 08:00 PM Revision f37c36e8 (ceph): mon/PaxosService: remove redundant check
- _active checks is_active() and rescheduled itself already.
Signed-off-by: Sage Weil <sage@inktank.com>
Reviewed-by: ... - 08:00 PM Revision d4876c53 (ceph): mon/PaxosService: do not prepare new pending if still proposing
- The _active callback can get called while are already proposing. If
that happens, we should not prepare a fresh new ... - 02:31 PM Bug #5515 (Resolved): mon: paxos allows reads before initial commit is done
- 02:29 PM rbd Bug #5464: krbd: modifying mapped image also modifies snapshot
- Sage Weil wrote:
> backport patch updated with mainline commit sha1 and sent to stable@, just waiting for it to land... - 02:29 PM rbd Feature #5465: openstack: cinder: support resize with rbd
- NOTE: the Linux kernel fix for #5464 erroneously points here; it should have references #5464.
- 02:19 PM CephFS Bug #5381 (Resolved): ceph-fuse: stuck with disconnected inodes on shutdown
- 02:18 PM Bug #5519 (Resolved): mon/osd: trimming of old maps based on last_epoch_clean is broken by design
- 02:12 PM Bug #5623 (Fix Under Review): pg dump_stuck command broke
- fixed teuthology to not pass the arg. let's just drop this hack to support the old usage.
- 09:05 AM Bug #5623 (Resolved): pg dump_stuck command broke
- ...
- 02:03 PM Bug #5622 (Resolved): osd: cls_refcount unit test failures
- commit:cda17fd1a9b7ba00ad85c9fff6e95434de3cf1a4
- 09:02 AM Bug #5622 (Resolved): osd: cls_refcount unit test failures
- teuthology-2013-07-13_01:00:18-rados-next-testing-basic, many runs...
- 09:07 AM Bug #5624 (Resolved): osd: prepare_to_stop() segfaults if we get a signal during startup
- we probably need to delay the signal installation?...
- 08:11 AM Bug #5621 (Resolved): mon: paxos fc lc aren't update when sync_start clears store
- fix is in wip-mon-fixes, needs review
ubuntu@teuthology:/a/teuthology-2013-07-12_20:00:05-rados-cuttlefish-testing... - 05:04 AM Revision bf4f8024 (ceph): mon/PaxosService: fix trim completion
- Do not call C_Committed after trim or else we will prematurely clear
the bool proposing, propose something again usin... - 04:42 AM Revision 741757a3 (ceph): ceph_argparse.py: allow valid char RE arg to CephString
- Change badchars to goodchars (no one was using badchars); allow
goodchars to be a RE character class of valid charact... - 04:42 AM Revision 0d66c9eb (ceph): MonCommands.h: use new validation for crush names (CephString goodchars)
- Signed-off-by: Dan Mick <dan.mick@inktank.com>
- 04:42 AM Revision 089dfe8e (ceph): ceph_argparse: ignore prefix mismatches, but quit if non-prefix
- I don't know what I was thinking; this was always the right validation
algorithm, and I broke it trying to simplify.
... - 04:20 AM Revision da4c749a (ceph): ceph_argparse.py: validate's 3rd arg is not verbose, it's partial
- Signed-off-by: Dan Mick <dan.mick@inktank.com>
- 03:56 AM Revision b9b9dd13 (ceph): Remove btrfs-tools and xfsprogs install step.
- We already install btrfs-tools and xfsprogs with ceph-qa-chef
Doing it here was just causing problems on non-ubuntu
d... - 03:54 AM Revision db4ad507 (ceph): Allow Overrides for ceph-deploy task
- This is needed so we can set the ceph branch for ceph-deploy
to use via the main yaml which is created via the suite
... - 03:24 AM Revision bd9cf10c (ceph): Workaround repopriority of yum local repo.
- Because of issues with package name differences vps are
setup to use repopriority and our local repo (which has
some ... - 03:20 AM Revision e3a07421 (ceph): Don't install kernels on non-ubuntu VPS
- Signed-off-by: Sandon Van Ness <sandon@inktank.com>
- 03:19 AM Revision 8b05c8cc (ceph): Re-create guest if it doesn't come up right.
- In some rare cases (mainly centos/rhel after creating the
guest with downburst it does not come up right. It
gets a k... - 03:16 AM Revision 4f478ccf (ceph): Add description option to lock.lock()
- For some reason lock_many() has a description but lock()
does not. This was useful in my testing of unlocking and
re-... - 03:13 AM Revision f16ecb99 (ceph): Use ceph.com mirror instea of github for ceph-qa-chef on VPS.
- Github.com is unreliable so use the ceph mirror instead.
Signed-off-by: Sandon Van Ness <sandon@inktank.com> - 03:10 AM Revision 4dbef128 (ceph): List IP address in orchestra.run output.
- Figuring out which machines output is coming from when things
are being executed on multiple machines can be a huge p... - 03:08 AM Revision d7a20ba2 (ceph): Fix Missed parenthesis.
- Signed-off-by: Sandon Van Ness <sandon@inktank.com>
- 03:05 AM Revision da7483a7 (ceph): VM: Use mac addresses from DB instead of randomizing.
- In order to make IP addresses less likely to change and to allow
a smaller DHCP pool to be used I generated static MA...
07/12/2013
- 11:38 PM Revision b5659b5e (ceph): Merge branch 'wip-wsgi' into next
- * wip-wsgi:
ceph-rest-api: separate into module and front-end for WSGI deploy
ceph-rest-api: make main program be... - 11:37 PM Revision cc109888 (ceph): ceph-rest-api: separate into module and front-end for WSGI deploy
- To deploy ceph-rest-api within a WSGI server (apache/mod_wsgi,
nginx/uwsgi, etc.), there needs to be an importable (.... - 11:21 PM Revision 495ee108 (ceph): msg/Pipe: fix RECONNECT_SEQ behavior
- Calling handle_ack() here has no effect because we have already
spliced sent messages back into our out queue. Inste... - 10:51 PM Revision 48a29598 (ceph): mon: make ancestor arg for 'osd crush rm|remove <name>' optional
- It was there for unlink, but not rm or remove.
Signed-off-by: Sage Weil <sage@inktank.com>
Reviewed-by: Dan Mick <da... - 10:50 PM Revision c0845a97 (ceph): mon: AuthMonitor: don't try to auth import a null buffer
- Hangs result if 'ceph auth import' is attempted without -i.
Check for this case and return error status. Also,
updat... - 10:50 PM Revision 60a19e74 (ceph): PendingReleaseNotes: ceph -s --format=json output change
- Signed-off-by: Sage Weil <sage@inktank.com>
- 10:44 PM Revision 62ae39ec (ceph): mon: be smarter about calculating last_epoch_clean lower bound
- We need to take PGs whose mapping has not changed in a long time into
account. For them, the pg state will indicate ... - 10:44 PM Revision da725852 (ceph): osd: report pg stats to mon at least every N (=500) epochs
- The mon needs a moderately accurate last_epoch_clean value in order to trim
old osdmaps. To prevent a PG that hasn't... - 10:44 PM Revision 757af91b (ceph): osd: fix warning
- From 653e04a79430317e275dd77a46c2b17c788b860b
Backport: cuttlefish, bobtail
Signed-off-by: Sage Weil <sage@inktank.c... - 10:21 PM Revision 65af2538 (ceph): Merge remote-tracking branch 'gh/wip-mon-sync-2' into cuttlefish
- Reviewed-by: Joao Eduardo Luis <joao.luis@inktank.com>
Reviewed-by: Greg Farnum <greg@inktank.com> - 10:18 PM Revision 0ee70083 (ceph): ceph.conf: enable old message assert
- If this triggers, the RECONNECT_SEQ feature is broken (and
maybe we've caught #5517). - 09:29 PM Bug #2895 (Resolved): cli: non-existent command returns confusing error message
- This is completely changed with the new CLI. I think the error messages are better.
- 09:27 PM Bug #4599 (Resolved): ceph auth import -i <file> is broken
- 02:40 PM Bug #4599: ceph auth import -i <file> is broken
- 09:25 PM Revision f67e437b (ceph): Merge branch 'wip-5371-next' into next
- Reviewed-by: Samuel Just <sam.just@inktank.com>
- 09:21 PM Revision 5dd1d4b1 (ceph): test: idempotent filestore test failure
- Remove obsolete use of collection_move()
Allow operations to be skipped if random selections don't make sense
Track t... - 09:18 PM Revision c70216a2 (ceph): Revert "test_filestore_idempotent: use obj name from source coll add"
- This reverts commit 0bc731ea93ae198afcdcba285751a6201c3de8de.
- 09:17 PM Revision 0a3c902c (ceph): Revert "test_filestore_idempotent: make newly created objects globally ...
- This reverts commit cdf5785c5e8b744f35ea34b2deb2a71edf8b9580.
- 08:33 PM Revision eccbdaa8 (ceph): ceph-object-corpus: added cuttlefish objects
- Signed-off-by: Sage Weil <sage@inktank.com>
- 08:31 PM Revision 97efc72d (ceph): rgw: fix a typo
- Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
- 08:06 PM Revision e390f44b (ceph): Merge remote-tracking branch 'gh/wip-corpus' into next
- Rgw bits Reviewed-by: Yehuda Sadeh <yehuda@inktank.com>
- 07:43 PM Revision a44854cd (ceph): qa/workunits/cls: remove new rgw tests
- These need extra arguments to connect to rgw...
Signed-off-by: Sage Weil <sage@inktank.com> - 07:30 PM Bug #4779: The ceph command and crushtool have differing views on valid characters for bucket names
- 07:30 PM Bug #4779: The ceph command and crushtool have differing views on valid characters for bucket names
- Decision is to allow only [a-zA-Z0-9-_.] in crush names. We can either add
'legalchars' to CephString and set it fo... - 07:17 PM Bug #5503: osd: ceph --admin-daemon interface doesn't handle spaces in names
- Sorry, David, I just noticed this was assigned to me, but I'm not clear on what exactly is breaking. Looks like some...
- 06:38 PM Revision b7c549db (ceph): msgr: add 'ms die on old message' to help catch reconnect seq issues
- Signed-off-by: Sage Weil <sage@inktank.com>
- 06:36 PM Revision 9f5cae44 (ceph): mon: structure 'ceph status' json output
- Return the summary info in a structured for instead of just mimicing the
legacy plaintext format.
Signed-off-by: Sag... - 06:34 PM Revision 2dcab51f (ceph): mon: fix copy/paste error
- From rebase/squash into 96c48835d19923b467ced3032948af6c2b7959f2.
Signed-off-by: Sage Weil <sage@inktank.com> - 06:28 PM Revision e0baf636 (ceph): Merge branch 'wip-mon-fixes' into next
- Reviewed-by: Joao Eduardo Luis <joao.luis@inktank.com>
- 06:28 PM Revision 2dc8688a (ceph): mon: stash latest state when flagging force_sync
- Store our latest state when we set the force_sync flag. This is important
because we will clear the store the next t... - 06:28 PM Revision cf9d1e1a (ceph): mon: clean up check_fsid() a bit
- Use the uuid_d type instead of kludging string comparisons.
Signed-off-by: Sage Weil <sage@inktank.com> - 06:28 PM Revision 96c48835 (ceph): mon: factor out stashing latest monmap
- Signed-off-by: Sage Weil <sage@inktank.com>
- 06:14 PM Revision 0c83b5fe (ceph): cls_rgw: add missing type tests
- Signed-off-by: Sage Weil <sage@inktank.com>
- 06:14 PM Revision d0bee5d8 (ceph): cls_refcount: add type tests
- Signed-off-by: Sage Weil <sage@inktank.com>
- 06:13 PM Revision a04525be (ceph): mon: fix sync_start doc
- Signed-off-by: Sage Weil <sage@inktank.com>
- 06:12 PM Revision e213b1bc (ceph): mon: fix recent paxos sync from 0 when first_committed == 1
- This is okay; allow it.
Signed-off-by: Sage Weil <sage@inktank.com> - 06:12 PM Revision e019dd04 (ceph): mon: fix off-by-one: no need to reapply previous last_committed after sync
- The old last_committed is already committed; don't reapply. This also fixes
the case where lc was 0 (i.e., we did get... - 05:36 PM Revision b8fad059 (ceph): doc/dev: add corpus.rst
- Structure of the corpus dir, how to add to it.
Signed-off-by: Sage Weil <sage@inktank.com> - 05:16 PM Revision 8e83a3be (ceph): Merge remote-tracking branch 'gh/next'
- 04:48 PM Revision 17c29957 (ceph): mon: add encoding tests for MonitorDBStore data types
- Signed-off-by: Sage Weil <sage@inktank.com>
- 04:48 PM Revision 0800ea1c (ceph): osd: add encoding tests for PushOp, PullOp, PushReplyOp
- Signed-off-by: Sage Weil <sage@inktank.com>
- 04:48 PM Revision 5139d0ce (ceph): mds: add encoding test for inode_backtrace_t; note incompat in corpus
- Signed-off-by: Sage Weil <sage@inktank.com>
- 04:48 PM Revision 8579d807 (ceph): mon: add encoding tests for newish MMonSync, MMonScrub
- Signed-off-by: Sage Weil <sage@inktank.com>
- 04:48 PM Revision 4a3241c5 (ceph): osd: reenable pg_query_t test
- This was disabled back in 4ce17cca2e924b2ef9565d5c345cd602bcc6dd98; not
sure why.
Signed-off-by: Sage Weil <sage@ink... - 04:48 PM Revision 3e4a2911 (ceph): osd/OSDmap: fix OSDMap::Incremental::dump() for new pool names
- The name is always present when pools are created, but not when they are
modified. Also, a name may be present with ... - 04:48 PM Revision afb814f5 (ceph): ceph-object-corpus: new objects for almost-dumpling
- Signed-off-by: Sage Weil <sage@inktank.com>
- 04:48 PM Documentation #5618: radosgw pool size guidelines
- Probably .rgw, .rgw.buckets, .rgw.buckets.index?
- 04:43 PM Documentation #5618 (Resolved): radosgw pool size guidelines
- It's somewhat easy to use the default pool sizes which would torpedo performance. The docs should call out which poo...
- 04:35 PM rbd Bug #5615: lock ops are not re-sent when cluster gets marked un-full
- oh right. in that case the op was probably sent and then dropped by the osd. objecter only pauses ops marked as wri...
- 04:08 PM rbd Bug #5615: lock ops are not re-sent when cluster gets marked un-full
- If it's stopping the op because the cluster is marked FULL then it ought to be able to know the op needs to be sent o...
- 03:39 PM rbd Bug #5615: lock ops are not re-sent when cluster gets marked un-full
- i bet this affects all class calls/execs, because objecter doesn't know if it is a read or a write. we may need to r...
- 03:33 PM rbd Bug #5615 (Duplicate): lock ops are not re-sent when cluster gets marked un-full
- ...
- 04:25 PM Bug #5495 (Need More Info): ceph-mon and minus character in hostname
- 03:27 PM Bug #5495: ceph-mon and minus character in hostname
- that second strace shows it hitting an unrelated assert on db->create_and_open()... joao?
- 04:22 PM devops Bug #5617 (Resolved): ceph-deploy: log ceph version to ceph.log
- As part of ceph-deploy install command, it would be would be nice to log ceph version to ceph.log.
we talked about... - 04:20 PM Bug #5371 (Resolved): idempotent filestore test failure
- 08:48 AM Bug #5371 (Fix Under Review): idempotent filestore test failure
- 03:52 PM Revision f28aa7e1 (ceph): Merge pull request #426 from dalgaaf/wip-da-SCA-cppcheck-5
- Fix some typical SCA issues from cppcheck
Reviewed-by: Sage Weil <sage@inktank.com> - 03:44 PM Bug #5616 (Resolved): repeated mon elections
- Repeatable on the mira103 cluster. Add a 4k pg pool, split to 8k, and then increase pgp_num to 8k. Will reproduce o...
- 03:22 PM Bug #5502 (Resolved): mon: long-running sync will restart (cuttlefish)
- 02:07 PM Feature #5614 (Resolved): mon: enable moving pools to HASHPSPOOL mode
- Right now there is no way to switch the CRUSH calculations a pool uses to include the PS. We want to be able to turn ...
- 01:31 PM Bug #5401: cuttlefish osd recovery slow
- Still seeing those SLOW requests and client crashes on cuttlefish. ANy Ide what to try? i've no expanded the cluster ...
- 12:40 PM Revision 6379f37d (ceph): test_rgw_admin_meta.cc: prefer prefix ++operator for non-trivial iterator
- Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>
- 12:39 PM Revision 329e467b (ceph): cls/rgw/cls_rgw.cc: use empty() instead of checking for size() == 0
- Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>
- 12:21 PM Revision 734c8245 (ceph): test_rgw_admin_log.cc: prefer prefix ++operator for non-trivial iterator
- Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>
- 12:18 PM Revision 88ba6515 (ceph): test_rgw_admin_opstate.cc: prefer prefix ++operator for non-trivial ite...
- Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>
- 12:15 PM Revision 028f6437 (ceph): test_rgw_admin_meta.cc: prefer prefix ++operator for non-trivial iterator
- Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>
- 10:10 AM Revision cccc09a4 (ceph): rgw/rgw_rest_opstate.cc: prefer prefix ++operator for non-trivial iterator
- Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>
- 10:08 AM Revision 22e0dbbd (ceph): rgw/rgw_rados.cc: use empty() instead of checking for size() > 0
- Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>
- 10:06 AM Revision ea0fcfac (ceph): cmdparse.cc: catch exception by reference
- Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>
- 09:40 AM rgw Bug #5347 (Resolved): rgw: bucket marker should include original zone name
- 08:42 AM devops Bug #5345: ceph-disk: handle less common device names
- what is strange is that parted showed 2 partitoins but cat /proc/partitions only showed 1. is it still in that state...
- 01:25 AM Revision 1f8fe709 (ceph): ceph-rest-api: Missing packaging (binary in .spec, manpage in both)
- Signed-off-by: Dan Mick <dan.mick@inktank.com>
- 12:17 AM Revision 3d25f461 (ceph): ceph-rest-api: make main program be "shell" around WSGI guts
- Move WSGI guts to pybind/ceph_rest_api.py for importability
into WSGI server frameworks.
Signed-off-by: Dan Mick <da...
07/11/2013
- 11:30 PM Revision cb0f8318 (ceph): ceph.in: output even a blank status line in compat mode
- Fixes: #5594
Signed-off-by: Dan Mick <dan.mick@inktank.com>
Reviewed-by: Sage Weil <sage@inktank.com> - 10:54 PM Revision 3a5cdb1e (ceph): added upgrade task for mixed mon test
- Signed-off-by: tamil <tamil@tamil-VirtualBox.(none)>
- 08:47 PM devops Bug #5345: ceph-disk: handle less common device names
- Hi Sage,
The code get through without any error, but I think somewhere in prepare_journal or related failed to set... - 01:00 PM devops Bug #5345: ceph-disk: handle less common device names
- ok, fixed the typo and redi the partition naming code.. try again?
thanks! - 04:36 AM devops Bug #5345: ceph-disk: handle less common device names
- Hi Sage,
I think there is a typo in line 327:
Traceback (most recent call last):
File "/usr/sbin/ceph-disk",... - 08:32 PM Bug #5594 (Resolved): ceph cli in compat mode should always add status+newline, even when status ...
- Fixed as noted
commit:cb0f831834ec9f3b38e5fbef87c1a9583e24f8bb - 06:42 PM Revision 93b3e63f (ceph): msg/Message: use old footer for encoded message dump
- This avoids the need for a conditional decoding check on ceph-dencoder,
and makes it match up with what encode_messag... - 06:25 PM Fix #5612 (Resolved): down mons prevents osd hosts from booting properly
- need to play with this and see what exactly the behavior is with both sysvinit and upstart. in particular, the osd s...
- 04:57 PM rgw Feature #5611 (New): rgw: swift GET request for object with custom metadata should show custom me...
- per customer:
According to [[http://docs.openstack.org/api/openstack-object-storage/1.0/content/special-metadata-a... - 04:45 PM rgw Feature #5610 (New): rgw: Creating a container with ACL metadata does not retain metadata when us...
- Per customer:
Using a PUT request to create a container with ACL metadata creates the container without any ACL pe... - 04:21 PM Revision cf8f16d7 (ceph): rgw: handle bucket removal by system user on master region
- intra region bucket creation, deletion and copy object to and from remote regions validated
Signed-off-by: Babu Shan... - 02:24 PM Revision 3b110dbc (ceph): ceph_argparse.py: define some self.* in __init__
- Fix for "Attribute 'xy' defined outside __init__": add __init__
function to class and set some default values for the... - 02:04 PM Revision 60c9ca62 (ceph): ceph_argparse.py: fix dangerous default value [] as arg
- Fix valid pylint warning about "Dangerous default value [] as argument".
Signed-off-by: Danny Al-Gaaf <danny.al-gaaf... - 01:56 PM Revision a049e290 (ceph): ceph_argparse.py: remove unnecessary semicolon
- Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>
- 01:54 PM Revision a28d5d64 (ceph): ceph_argparse.py: fix some ArgumentValid() commands
- Fix raise call for ArgumentValid(): self has no member t, use
t directly. Moved one check in the correct block to avo... - 01:51 PM Documentation #5608 (Resolved): Add ARM Options to Hardware Recommendations
- On http://ceph.com/docs/master/install/hardware-recommendations/
update Minimum Hardware Recommendations and Prod... - 01:49 PM rgw Feature #5607 (Resolved): s3-tests: new tests for multi region
Test 1: create bucket master, try to access on secondary
- create bucket on master region
- try to access buc...- 01:48 PM rgw Feature #5606 (Resolved): s3-tests: run in multi region setup
- The s3-tests functional tests suite should be extended to support multiple regions. The master / secondary endpoints ...
- 01:46 PM rgw Feature #5605 (Resolved): rgw: teuthology tests to check bucket issues in multi region env
- Test 2a: test bucket listing on secondary
1. create a bucket on master region
(actually using boto)
$ s3 creat... - 01:44 PM rgw Feature #5604 (Resolved): rgw: teuthology tests to check various user creation issues on multi re...
- Test 1 setup
1. Create users on master region: user1, user2
radosgw-admin user create --rgw-region=master --uid... - 01:43 PM rgw Feature #5603 (Resolved): rgw: teuthology test to check secondary region creation
- Test 0b: secondary region creation
- create secondary region + zone (logs turned off)
- update region map
- r... - 01:41 PM rgw Feature #5602 (Resolved): rgw: teuthology task to test default region as master region
- 0. Test region setup
Test 0a: check that default gateway acts as master
- create user on first gateway (intend... - 01:38 PM Revision 3c4ae839 (ceph): ceph_argparse.py: add missing spaces after comma
- Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>
- 01:37 PM Revision c2a63349 (ceph): ceph_argparse.py: add missing spaces around operator
- Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>
- 01:35 PM Revision 3278295d (ceph): ceph_argparse.py: fix bad indentation
- Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>
- 12:02 PM Revision 707c39dd (ceph): ceph.in: add missing spaces around operator
- Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>
- 12:01 PM Revision f714c01b (ceph): ceph.in: add missing spaces after comma
- Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>
- 11:59 AM Revision b82473ad (ceph): ceph.in: remove unnecessary semicolons
- Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>
- 11:52 AM devops Bug #5599 (Resolved): ceph-disk: prepare should issue a partprobe on the journal device too
- Currently, after preparing a disk, we issue a partprobe on the prepared device (ceph data).
When using a separate jo... - 11:52 AM Bug #5495: ceph-mon and minus character in hostname
- Forgot to mention that this user was attempting to upgrade from bobtail to cuttlefish.
- 11:40 AM Bug #5495: ceph-mon and minus character in hostname
- A user was able to reproduce this reliably enough to get an strace out of it. Attached.
- 09:44 AM Bug #5392: osd: unfound objects from thrashing
- trying to reproduce with logging.
- 04:55 AM Revision 5a6251c8 (ceph): ceph-rest-api: connect to cluster with configured client name
- Spotted by Noah Watkins
Signed-off-by: Dan Mick <dan.mick@inktank.com> - 04:04 AM Bug #5596: OSD keeps crashing while trying to rewight another OSD
- If i keep reweighting the next OSD on the SAME (this) host crashes.
- 04:03 AM Bug #5596 (Can't reproduce): OSD keeps crashing while trying to rewight another OSD
- I've added 8 new OSDs with a weight of 0. I then tried to reweight them by adding 0.01 in each turn to the weight of ...
- 03:58 AM Revision 17595f34 (ceph): mon: OSDMonitor: osd pool get: move to preproc, add formatted output
- Move 'osd pool get' handling to preprocess_command()
It's a read operation; there's no need for it to be in prepare_c... - 03:58 AM Revision e75db713 (ceph): mon: OSDMonitor: formatter output on 'osd map'
- Signed-off-by: Joao Eduardo Luis <joao.luis@inktank.com>
Reviewed-by: Sage Weil <sage@inktank.com> - 03:58 AM Revision c87f55b5 (ceph): CompatSet, MDSMap: formatted dumps must avoid numeric tags
- Tags made of numbers only confuse XML parsers. Avoid them.
Signed-off-by: Dan Mick <dan.mick@inktank.com> - 03:58 AM Revision 45dc5731 (ceph): OSDMonitor: return failure for 'osd lost' without safety switch
- If user doesn't supply --yes-i-really-mean-it, fail the operation
Signed-off-by: Dan Mick <dan.mick@inktank.com> - 03:58 AM Revision 4cb0e062 (ceph): Add 'ceph-rest-api'
- ceph-rest-api is a Python WSGI module for accessing the Ceph cluster.
It supports most of the commands supported by t... - 03:58 AM Revision 6cafb0e3 (ceph): ceph-rest-api tests
- Patterned after cephtool/test.sh, with some deeper validation of
output format and contents (because structured outpu... - 03:58 AM Revision e66d011c (ceph): cephtool/test.sh cleanup
- 1) don't pass subshell to expect_false
2) osd pause/unpause don't take arguments (apply to all OSDs)
Signed-off-by: ... - 03:58 AM Revision 07590ae8 (ceph): cephtool/test.sh: sync status no longer supported
- da0aff28ab478bcc3136715f92bc1af8d4b403c1 removed 'ceph sync status';
test script needs to stop trying it
Signed-off-... - 03:58 AM Revision e4d063b0 (ceph): ceph-rest-api manpage
- Signed-off-by: Dan Mick <dan.mick@inktank.com>
- 03:58 AM Revision 04fbc523 (ceph): mon: Monitor: 'health detail --format=json' should only output JSON
- Fixes: #5526
Signed-off-by: Joao Eduardo Luis <joao.luis@inktank.com> - 03:58 AM Revision a15d9327 (ceph): mon: formatted output for 'fsid'
- Signed-off-by: Dan Mick <dan.mick@inktank.com>
- 03:58 AM Revision 300f42b8 (ceph): mon: Monitor: support multiple formatters on some status functions
- Commands such as 'mon_status', 'quorum_status', 'sync_status' and
'sync_force' didn't support other formatter besides... - 03:47 AM Revision c6d1dba6 (ceph): test/encoding/import-genearted: fix
- Dest dir was wrong, cp shortcut check was wrong.
Signed-off-by: Sage Weil <sage@inktank.com> - 03:40 AM Revision 3cf8ae01 (ceph): mds: make rename_rollback::drec not assert on unexpected d_type
- Signed-off-by: Sage Weil <sage@inktank.com>
- 03:33 AM Revision 846382b3 (ceph): mds: fix rename_rollback::drec cctor
- Signed-off-by: Sage Weil <sage@inktank.com>
- 03:25 AM Revision db68975e (ceph): cls/rgw: init op in rgw_cls_obj_complete_op ctor
- Turned up by the encoding/decoding unit tests.
Signed-off-by: Sage Weil <sage@inktank.com> - 02:02 AM Revision 83bcde34 (ceph): mon,auth: AuthMonitor, KeyRing: add Formatter-dumps of auth info
- Signed-off-by: Dan Mick <dan.mick@inktank.com>
auth: KeyRing: encode_formatted() receives a label as first argument
... - 02:02 AM Revision 7049efca (ceph): mon: AuthMonitor: early out on bad entity name
- Signed-off-by: Dan Mick <dan.mick@inktank.com>
Reviewed-by: Sage Weil <sage@inktank.com> - 02:02 AM Revision 7da0320c (ceph): mon: AuthMonitor: dump formatted and plaintext key on 'get-key'
- Signed-off-by: Joao Eduardo Luis <joao.luis@inktank.com>
- 02:02 AM Revision b18554a6 (ceph): mon: AuthMonitor: formatted output to 'auth get-or-create[-key]'
- Signed-off-by: Joao Eduardo Luis <joao.luis@inktank.com>
- 02:02 AM Revision e76b05b9 (ceph): mon: AuthMonitor: improve error reporting for import
- Signed-off-by: Dan Mick <dan.mick@inktank.com>
- 02:02 AM Revision f8ee4a74 (ceph): auth: cephx: KeyServer: list secrets to formatter or plaintext
- Signed-off-by: Joao Eduardo Luis <joao.luis@inktank.com>
- 02:02 AM Revision db71a380 (ceph): mon: add formatted output for several commands:
- pg stat
pg map
osd stat
osd lspools
osd blacklist ls
osd create
osd getmaxosd
mds compat ... - 02:02 AM Revision 989be66c (ceph): rados.py: allow long-form names of 'warn' and 'err' as watch levels
- Signed-off-by: Dan Mick <dan.mick@inktank.com>
- 02:02 AM Revision c9b54d75 (ceph): common: Formatter: add flush to bufferlist
- Signed-off-by: Joao Eduardo Luis <joao.luis@inktank.com>
- 02:02 AM Revision bfa2284d (ceph): MonCommands: add new fields: modulename, perms, availability
- To help optimize the REST API, we need to know whether the commands
are read (GET) or write (PUT/POST). However, we ... - 02:02 AM Revision 0f2ebc6e (ceph): ceph.in, ceph_argparse.py: move descsort() to ceph_argparse
- Signed-off-by: Dan Mick <dan.mick@inktank.com>
- 02:02 AM Revision a5621625 (ceph): ceph CLI: valid() no longer returns bool, but just exception
- The type validation's valid() method was using a combination of
return code and exception to really indicate the same... - 02:02 AM Revision 2e2bc6d0 (ceph): ceph_argparse.py cleanup: simplify some code, error on unused args
- Signed-off-by: Dan Mick <dan.mick@inktank.com>
- 02:02 AM Revision 92d5eb08 (ceph): auth: Crypto: add encode_formatted/plaintext() functions
- Signed-off-by: Joao Eduardo Luis <joao.luis@inktank.com>
- 01:42 AM Revision 7f678480 (ceph): Merge pull request #420 from gregsfortytwo/master
- radosgw-admin help command naming and text improvements
Reviewed-by: Yehuda Sadeh <yehuda@inktank.com> - 01:33 AM Revision 1ad2127f (ceph): PGMonitor: fix stat comparison in prepare_pg_stats
- Signed-off-by: Samuel Just <sam.just@inktank.com>
Reviewed-by: Sage Weil <sage@inktank.com> - 01:00 AM Revision c0a78081 (ceph): rgw: use different daemon name and apache module paths
- Signed-off-by: Josh Durgin <josh.durgin@inktank.com>
07/10/2013
- 11:51 PM Revision e18a6b81 (ceph): Added task yaml for partial-upgrade of osds
- Signed-off-by: tamil <tamil.muthamizhan@inktank.com>
- 10:06 PM Revision 180c7b45 (ceph): Get device-by-path by looking for it instead of assuming 3rd entry.
- On some systems (virtual machines so far) the device-by-path entry
from udevadm is not always in the same spot so ins... - 10:06 PM Revision e537699b (ceph): Get device-by-path by looking for it instead of assuming 3rd entry.
- On some systems (virtual machines so far) the device-by-path entry
from udevadm is not always in the same spot so ins... - 09:55 PM Revision 836e35f8 (ceph): Get device-by-path by looking for it instead of assuming 3rd entry.
- On some systems (virtual machines so far) the device-by-path entry
from udevadm is not always in the same spot so ins... - 09:51 PM Bug #4599: ceph auth import -i <file> is broken
- Fix pushed to wip-4599; David, can you have a look?
- 09:50 PM Bug #4599 (Fix Under Review): ceph auth import -i <file> is broken
- 01:34 PM Bug #4599: ceph auth import -i <file> is broken
- OK. I believe it never worked with stdin, but David points out:
1) the help could be more clear about that, and
... - 09:28 PM Revision 348acecb (ceph): Merge branch 'wip-5492-gl' into next
- Re-instate fix for #5492 with corrections.
Signed-off-by: Gary Lowell <gary.lowell@inktank.com> - 08:44 PM Revision a120d81d (ceph): Makefile.am: fix ceph_sbindir
- This reinstates the fix for the ceph_sbindir from commit
352f362567bf270d0896fb7573df4ae5139a56fb, with corrections
f... - 08:27 PM Revision 5deeb7b0 (ceph): Merge branch 'next'
- 08:25 PM Revision dcb42ca3 (ceph): test_rados: don't use assert_greater
- It's only present in very recent versions of nose, and doesn't provide
much beyond a simple assert.
Signed-off-by: J... - 08:14 PM Revision d8fe4392 (ceph): Python Ioctx.get_xattrs does not take xattr_name
- Fixes bug 5528
Since it returns all xattrs, it does not take an xattr_name. Also,
add unit tests for the Object xatt... - 08:09 PM devops Bug #5345: ceph-disk: handle less common device names
- aha, i see the problem. pushed a fix.. can you see if it works now?
(thanks!) - 07:48 PM devops Bug #5345: ceph-disk: handle less common device names
- Hi Sage,
Here is what I got:
ceph-disk -v prepare /dev/cciss/c0d1
DEBUG:ceph-disk:dev /dev/cciss/c0d1 name is ... - 06:40 PM Revision 804314b8 (ceph): Merge remote-tracking branch 'gh/cuttlefish' into wip-mon-sync-2
- 06:35 PM Bug #5517: osd: stuck peering on cuttlefish
- I traced another instance of getting old messages (the only one in this run), but unfortunately it didn't capture the...
- 02:33 PM Bug #5517: osd: stuck peering on cuttlefish
- kick test 3 for pg 16.1403 is another instance of op_tp hung filling out inc osd map messages.
- 08:56 AM Bug #5517: osd: stuck peering on cuttlefish
- These are the oplogs after turning up log levels and marking osd.2 down:
12.16 ceph mon tell \* injectargs '--debu... - 06:32 PM Revision bc291d3f (ceph): osd: fix warning
- From 653e04a79430317e275dd77a46c2b17c788b860b
Backport: cuttlefish, bobtail
Signed-off-by: Sage Weil <sage@inktank.com> - 06:10 PM devops Bug #5479: Append our built packages with some sort of inktank/ceph identifier
- Well if the modified/backported version needs to be utilized for ceph stuff. I think the point is so when someone add...
- 06:07 PM devops Bug #5479 (In Progress): Append our built packages with some sort of inktank/ceph identifier
- We're doing a bunch of backports now, do you think the ceph tag is needed on unmodified packages ?
- 06:03 PM Revision 78f22663 (ceph): osd: limit number of inc osdmaps send to peers, clients
- We should not send an unbounded number of inc maps to our peers or clients.
In particular, if a peer is not contacted... - 06:02 PM Revision 653e04a7 (ceph): osd: limit number of inc osdmaps send to peers, clients
- We should not send an unbounded number of inc maps to our peers or clients.
In particular, if a peer is not contacted... - 05:19 PM Revision b36338be (ceph): messages/MPGStats: do not set paxos version to osdmap epoch
- The PaxosServiceMessage version field is meant for client-coordinated
ordering of messages when switching between mon... - 05:19 PM Revision 5c8a5634 (ceph): mon: do not populate MMonCommand paxos version field
- The field is not used or useful since the monitor does not even look
at it (in Monitor::handle_command()). Avoid pop... - 05:09 PM Revision 2ea6c4cc (ceph): rgw-admin: use 'get' instead of 'show' or info for geo stuff
- These all output data that you can put back in via 'set', so they're
a 'get' rather than something else.
Signed-off-... - 05:09 PM Revision 997c6f4e (ceph): rgw-admin: add some of the newer flags and specify their use
- Signed-off-by: Greg Farnum <greg@inktank.com>
- 04:43 PM rgw Bug #5595 (Resolved): object has a Content-Type, but its content_type property is not shown in Sw...
- This is in regards to a support request found here: https://inktank.zendesk.com/agent/#/tickets/396
An object (thu... - 04:34 PM Revision c124aa09 (ceph): Merge pull request #419 from dalgaaf/wip-da-fix-build-rados-classes
- ceph.spec.in: add missing %{_libdir}/rados-classes/libcls_* files
- 04:30 PM Revision d36ec78b (ceph): qa/workunits/cls/test_cls_rgw: add new rgw class tests
- Signed-off-by: Sage Weil <sage@inktank.com>
Reviewed-by: Yehuda Sadeh <yehuda@inktank.com> - 04:21 PM Revision 6ddaa26e (ceph): Merge pull request #418 from dalgaaf/wip-da-fix-build-ceph-test
- Fix build and packaging of ceph-test package
Reviewed-by: Sage Weil <sage@inktank.com> - 04:20 PM Revision bee4c72f (ceph): Merge pull request #417 from dalgaaf/wip-da-fix-CRYPTO_LIBS
- Makefile.am: fix build, use $(CRYPTO_LIBS)
Reviewed-by: Sage Weil <sage@Inktank.com> - 04:16 PM Revision 54ee2dc8 (ceph): rgw: Fix return value for swift user not found
- http://tracker.ceph.com/issues/1779 fixes #1779
Adjust the return value from rgw_get_user_info_by_swift call
in RGW_... - 04:12 PM Revision 2e79f73b (ceph): ceph.spec.in: add missing %{_libdir}/rados-classes/libcls_* files
- Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>
- 03:55 PM Revision ff5f25eb (ceph): debian/ceph-test.install: add missing files
- Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>
- 03:53 PM Revision 640d138c (ceph): ceph.spec.in: add some more ceph_test_cls_rgw_* files
- Add new ceph_test_cls_rgw_* files to ceph-test package section.
Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de> - 03:52 PM Revision dfe3c217 (ceph): src/Makefile.am: rename cls_test_rgw_*
- Rename files from cls_test_rgw_* to ceph_test_cls_rgw_* to fit the current
naming scheme.
Signed-off-by: Danny Al-Ga... - 03:51 PM Revision 78bae33f (ceph): ceph.spec.in: add some ceph_test_cls_* files
- Add some installed but not packaged ceph_test_cls_* files to
the ceph-test package.
Signed-off-by: Danny Al-Gaaf <da... - 03:49 PM Bug #5594 (Resolved): ceph cli in compat mode should always add status+newline, even when status ...
- The new CLI outputs the status message to stdout for compatibility with the old tool, but it
only does so if the sta... - 03:15 PM Revision dc99a23d (ceph): Makefile.am: fix build, use $(CRYPTO_LIBS)
- Use $(CRYPTO_LIBS) instead of -lcryptopp to work also with
nss crypto lib.
Signed-off-by: Danny Al-Gaaf <danny.al-ga... - 02:49 PM Bug #5371 (In Progress): idempotent filestore test failure
- I ran this test twice on Mira and twice on Plana nodes against a Jul 9, 2013 master branch (00ae543b3e32f89d906a0e934...
- 02:45 PM Bug #5492 (Resolved): scripts installing into /usr/usr/sbin (with --prefix=/usr)
- Thanks Danny. I tested $(exec_prefix)$(sbindir) on rpm and debian builds and it looks like does the right thing.
- 06:37 AM Bug #5492: scripts installing into /usr/usr/sbin (with --prefix=/usr)
- See my comment at git pull request #389: "... this patch will cause trouble on e.g. SUSE and also Debian. The files g...
- 01:49 PM Bug #5528 (Resolved): Python bindings Object.get_xattrs() requires unnecessary xattr_name
- Fixed by commit:d8fe4392604c45de6e7dd257d7f6a3b3af1fba0a
- 11:30 AM Bug #5518 (Resolved): osd: marking single osd down makes others go down (cuttlefish)
- 10:35 AM Bug #5518: osd: marking single osd down makes others go down (cuttlefish)
- hmm, yeah.. how about 100? that's also the max # of maps we will shove in a single MOSDMap message. otherwise ok?
- 10:02 AM Bug #5518: osd: marking single osd down makes others go down (cuttlefish)
- We will usually have more than the most recent 25 in cache, maybe a larger default?
- 11:06 AM rgw Cleanup #5558 (Fix Under Review): rgw: modify certain radosgw-admin operations interface
- Simple enough change. Passed my local make check.https://github.com/ceph/ceph/pull/420
- 09:28 AM rgw Cleanup #5558: rgw: modify certain radosgw-admin operations interface
- Just renamed them all to "get" since the output corresponds to what you give "set", and I want to keep the command na...
- 09:06 AM rgw Feature #3074 (Resolved): radosgw needs --help support
- 08:54 AM rgw Bug #1779 (Resolved): rgw: swift auth returns wrong error code when unexisting user is given
- 02:51 AM rgw Bug #1779: rgw: swift auth returns wrong error code when unexisting user is given
- As yehuda told me on the mailing list, the fix might not be backported.
Should we close the issue? - 08:52 AM rbd Bug #5426: librbd: mutex assert in perfcounters::tinc in librbd::AioCompletion::complete()
- both of these are cuttlefish.. maybe something that hasn't been backported?
- 08:52 AM rbd Bug #5426: librbd: mutex assert in perfcounters::tinc in librbd::AioCompletion::complete()
- ...
- 06:50 AM rbd Bug #5591: rbd-fuse crashes repeatedly under light load
- In case it is important, I am running an io generator against 20 * 1024MB RBD images in parallel, made available to t...
- 06:12 AM rbd Bug #5591 (Won't Fix): rbd-fuse crashes repeatedly under light load
- The rbd-fuse client crashes when I do any kind of sustained IO to its devices (as little as 10 IOPS) for more than a ...
- 04:55 AM Revision 47852c26 (ceph): mon/OSDMonitor: make 'osd crush rm ...' slightly more idempotent
- This is a manual backport of 18a624fd8b90d9959de51f07622cf0839e6bd9aa.
Do not return immediately if we are looking at... - 01:52 AM Revision 42b9ea95 (ceph): misc: move system type checking to a generic location
- Signed-off-by: Josh Durgin <josh.durgin@inktank.com>
- 01:51 AM Revision 7845848c (ceph): s3tests: fix client configurations that aren't dictionaries
- They're always used as dictionaries later on.
Signed-off-by: Josh Durgin <josh.durgin@inktank.com> - 01:51 AM Revision 343a42c0 (ceph): s3tests: fix client configurations that aren't dictionaries
- They're always used as dictionaries later on.
Signed-off-by: Josh Durgin <josh.durgin@inktank.com> - 01:38 AM rgw Feature #4365: rgw: allow setting wildcard keystone role
- Should we have wildcard support in role (Example: adm*, service*, *-staff), or a wildcard role to accept to tell rgw ...
- 12:12 AM Revision 8fd87093 (ceph): Merge pull request #400 from ceph/wip-mon-newsync
- simpler mon sync
Reviewed-by: Greg Farnum <greg@inktank.com>
Reviewed-by: Joao Eduardo Luis <joao.luis@inktank.com>
07/09/2013
- 11:24 PM Revision 51dbc389 (ceph): Merge branch 'wip-4982-4983-oloc-rebase'
- fixes: #4982
fixes: #4983
Reviewed-by: Samuel Just <sam.just@inktank.com>
Reviewed-by: Sage Weil <sage@inktank.com>
... - 11:20 PM Revision 1c22bdb7 (ceph): restart rgw after upgrade for rgw tests
- Signed-off-by: tamil <tamil.muthamizhan@inktank.com>
- 10:34 PM Revision aa7448cd (ceph): Merge pull request #415 from ceph/rgw-next
- 10:12 PM Bug #5581: fix big suite failures
- current run: osd.2 has stale pgs, flush_pg_stats hung. vapre a.yaml.out. didn't have logging on.
- 03:22 PM Bug #5581 (Resolved): fix big suite failures
- 09:12 PM Revision 00ae543b (ceph): mon: do not scrub if scrub is in progress
- This prevents an assert from unexpected scrub results from the previous
scrub on the leader.
Backport: cuttlefish
Si... - 09:12 PM Revision 8638fb64 (ceph): unittest_pglog: fix unittest
- This was broken by the pg_stat_t::reported cleanup.
Signed-off-by: Sage Weil <sage@inktank.com> - 09:10 PM Revision b10848e2 (ceph): Merge branch 'master' into wip-4982-4983-oloc-rebase
- 09:09 PM devops Bug #5345: ceph-disk: handle less common device names
- Jing Yuan Luke wrote:
> Hi Sage,
>
> I had the following error:
>
> root@yyy:~/ceph-configure# ceph-deploy -v ... - 07:03 PM devops Bug #5345: ceph-disk: handle less common device names
- Hi Sage,
I had the following error:
root@yyy:~/ceph-configure# ceph-deploy -v osd prepare xxx:cciss/c0d1
Prepa... - 09:09 PM Revision 8c501651 (ceph): osd: Add the ability to set capabilities on namespaces
- Parse namespace spec in osd caps and use in is_match()
Add test cases to unit test
feature: #4983 (OSD: namespaces p... - 09:09 PM Revision f03cdf55 (ceph): osd: Clean-up redundant use of object_locator_t
- Remove locator arg from get_object_context()/find_object_context()
Remove locator from object_info_t but retain encod... - 09:09 PM Revision bad317b9 (ceph): osd: Add namespace to dump_watchers output
- Signed-off-by: David Zafman <david.zafman@inktank.com>
- 09:09 PM Revision c4870143 (ceph): librados/misc.cc: reverse offset and length on write call
- Signed-off-by: Samuel Just <sam.just@inktank.com>
- 09:09 PM Revision 69af60db (ceph): librados: Add operate()/operate_read() log messages
- Signed-off-by: David Zafman <david.zafman@inktank.com>
- 09:09 PM Revision 94ccfac4 (ceph): rados: Add namespace arg (--namespace, -N) to rados command
- Signed-off-by: David Zafman <david.zafman@inktank.com>
- 09:09 PM Revision fddfdebf (ceph): test: Add namespace test cases to librados tests
- Signed-off-by: David Zafman <david.zafman@inktank.com>
- 09:09 PM Revision 6eb88a13 (ceph): test: Add namespace list test cases to librados test
- Signed-off-by: David Zafman <david.zafman@inktank.com>
- 09:09 PM Revision 71295640 (ceph): osd: Prepare for nspace match with simpler is_match_all()
- Signed-off-by: David Zafman <david.zafman@inktank.com>
- 09:09 PM Revision e761e4e5 (ceph): librados, os, osd, osdc, test: Add support for client specified namespaces
- Add rados_ioctx_namespace_set_key() and librados::IoCtx::namespace_set_key()
Add namespace to admin-daemon operations... - 09:05 PM Bug #5518 (Fix Under Review): osd: marking single osd down makes others go down (cuttlefish)
- 05:58 PM Bug #5518: osd: marking single osd down makes others go down (cuttlefish)
- ...
- 08:43 PM Revision ae866426 (ceph): Merge branch 'wip-mon-osdmap-trim'
- Reviewed-by: Samuel Just <sam.just@inktank.com>
- 08:42 PM Revision 449283f8 (ceph): mon/OSDMonitor: allow osdmap trimming to be forced via a config option
- In certain cases the admin may know that it is safe to trim old osdmaps but
a bug or other issue is preventing the Mo... - 08:42 PM Revision da81228c (ceph): osd: report pg stats to mon at least every N (=500) epochs
- The mon needs a moderately accurate last_epoch_clean value in order to trim
old osdmaps. To prevent a PG that hasn't... - 08:42 PM Revision cc0006de (ceph): mon: be smarter about calculating last_epoch_clean lower bound
- We need to take PGs whose mapping has not changed in a long time into
account. For them, the pg state will indicate ... - 08:42 PM Revision 8e075146 (ceph): osd: change pg_stat_t::reported from eversion_t to a pair of fields
- This rarely represents an actual eversion_t as the epoch and seq values are
bumped semi-independently to ensure it is... - 07:24 PM Revision 18a624fd (ceph): mon/OSDMonitor: make 'osd crush rm ...' slightly more idempotent
- This particular failure is easily triggered by the crush_ops.sh
workunit. Make it a bit less likely to fail.
Signed... - 06:45 PM Revision c5157dde (ceph): doc/release-notes: v0.66
- Signed-off-by: Sage Weil <sage@inktank.com>
- 06:12 PM Revision 3a1c316a (ceph): Add mon create and destroy with an optional argument mon_initial_members
- Signed-off-by: tamil <tamil.muthamizhan@inktank.com>
- 06:10 PM Revision 8bdd86a1 (ceph): Merge branch 'wip-mon-trim'
- Reviewed-by: Joao Eduardo Luis <joao.luis@inktank.com>
- 06:09 PM Revision da248a9e (ceph): mon: make service trim_to stateless
- Call get_trim_to() when we need to know how much to trim (if any), and
calculate it then. No need to keep this in a ... - 06:09 PM Revision d97f3142 (ceph): mon/PaxosService: simplify paxos_service_trim_min check
- Signed-off-by: Sage Weil <sage@inktank.com>
- 06:09 PM Revision cab8eeec (ceph): mon/PaxosService: move paxos_service_trim_max into caller, clean up
- Signed-off-by: Sage Weil <sage@inktank.com>
- 06:09 PM Revision 44db2ac5 (ceph): mon/PaxosService: inline trim()
- This is now trivial; pull it into the caller.
Signed-off-by: Sage Weil <sage@inktank.com> - 06:09 PM Revision 8799872d (ceph): mon/PaxosService: update docs a bit
- Signed-off-by: Sage Weil <sage@inktank.com>
- 06:09 PM Revision 886b637b (ceph): mon/PaxosService: pass trim target into encode_trim()
- This will help us in a few patches...
Signed-off-by: Sage Weil <sage@inktank.com> - 06:05 PM Revision 8b866d2e (ceph): mon: preserve last_committed_floor across sync
- Add a paranoid check to prevent us from forgetting how far ahead our
last_committed was when we sync. This prevents ... - 06:05 PM Revision ccceeee5 (ceph): mon/Paxos: remove unnecessary trim enable/disable
- The sync no longer cares if we trim Paxos versions as we go, as long as we
don't trim so fast that we fall behind bet... - 06:05 PM Revision af3b49f6 (ceph): mon/Paxos: move consistent check into Paxos::init()
- Signed-off-by: Sage Weil <sage@inktank.com>
- 06:05 PM Revision 45907dc1 (ceph): mon: add --force-sync startup option
- Signed-off-by: Sage Weil <sage@inktank.com>
- 06:05 PM Revision 9fc4e4f3 (ceph): mon/MonitorDBStore: drop unused single prefix synchronizer
- Signed-off-by: Sage Weil <sage@inktank.com>
- 06:05 PM Revision a4d0ccf6 (ceph): mon: rev the internal mon protocol
- This captures the new sync.
Signed-off-by: Sage Weil <sage@inktank.com> - 06:05 PM Revision bbfb5b41 (ceph): mon: drop single-use is_sync_on_going() check
- Signed-off-by: Sage Weil <sage@inktank.com>
- 06:05 PM Revision 1f5e6c22 (ceph): mon: no need to reset sync state on restart
- If we are in or forcing a sync, we can leave these there until the sync
completes.
Signed-off-by: Sage Weil <sage@in... - 06:05 PM Revision da0aff28 (ceph): mon: implement a simpler sync
- The previous sync implementation was highly stateful and very complex.
This made it very hard to understand and to de... - 06:05 PM Revision aa33bc88 (ceph): mon/Paxos: config min paxos txns to keep separately
- We were using paxos_max_join_drift to control the minimum number of
paxos transactions to keep around. Instead, make... - 05:42 PM Revision f326c4dc (ceph): mon/PGMonitor: cleanup: use const strings for pgmap prefixes
- Signed-off-by: Sage Weil <sage@inktank.com>
- 05:04 PM Bug #5517: osd: stuck peering on cuttlefish
- slider.ops.newdream.net:~samuelj/log_files/
- 03:23 PM Bug #5517: osd: stuck peering on cuttlefish
- osd.227 queries log on osd.21:
2013-07-08 11:49:10.723242 7f4afac64700 1 -- 10.81.144.107:6800/13176 --> 10.81.158.... - 04:32 PM rgw Bug #5522 (Resolved): rgw: use of select for waiting on curl sockets
- We now use curl_multi_wait() when available. When it's not available we force a timeout to the select() so that we do...
- 04:30 PM Feature #4983 (Resolved): OSD: namespaces pt 2 (caps)
- 12:39 PM Feature #4983: OSD: namespaces pt 2 (caps)
- 04:29 PM Feature #4982 (Resolved): OSD: namespaces pt 1 (librados/osd, not caps)
- 04:21 PM rgw Bug #5348 (Resolved): rgw: missing copy constraints checks for inter region user object copy
- 04:21 PM rgw Bug #5344 (Resolved): rgw: make list of bucket placement pools index configurable
- 04:19 PM rgw Feature #2169 (Resolved): rgw: api to control bucket placement
- 04:19 PM rgw Feature #2169: rgw: api to control bucket placement
- This is solved in Dumpling. It will be possible to set up placement targets for each region. At the zone level these ...
- 04:19 PM Revision 5faa4ac1 (ceph): rgw: warn on the lack of curl_multi_wait()
- Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
- 04:13 PM rgw Bug #5357 (Resolved): rgw: set and retrieve intra-region copy operation state
- 04:09 PM rgw Feature #3991 (Resolved): rgw: dr: region mgt changes: define datastructures
- 04:09 PM rgw Feature #3990 (Resolved): rgw: dr: implement new version objclass
- 04:09 PM rgw Feature #3989 (Resolved): rgw: dr: region mgt changes: radosgw admin changes
- 04:09 PM rgw Feature #3988 (Resolved): rgw: dr: region mgt changes: define/implement internal API
- 04:09 PM rgw Feature #3987 (Resolved): rgw: dr: region mgt changes: extend json parser with json decoder
- 04:08 PM rgw Feature #5134 (Resolved): rgw: RESTful api for datalog
- 04:08 PM rgw Feature #5133 (Resolved): rgw: RESTful api to lock/unlock mdlog
- 04:05 PM devops Feature #5091 (Resolved): google-perftools for arm
- 03:49 PM rgw Tasks #5586 (Resolved): rgw: build test plan
- 03:38 PM rgw Feature #5341 (Resolved): rgw: keep state for cross-rgw copy operations
- 03:38 PM rgw Feature #5352 (Resolved): rgw: metadata get should also dump mtime
- 03:38 PM rgw Feature #5349 (Resolved): rgw: intra-region object copy
- 03:38 PM rgw Documentation #5166: rgw: dr: async repl and DR documentation
- 03:38 PM rgw Feature #4310 (Resolved): rgw: multisite: radosgw changes: copy across regions
- 03:38 PM rgw Feature #5354 (Resolved): rgw: intra-region object copy should also set mtime on object
- 03:38 PM rgw Feature #5353 (Resolved): rgw: metadata put should apply mtime if set
- 03:38 PM rgw Feature #4098 (Resolved): rgw: multi-site: Global Bucket Namespace
- 03:38 PM rgw Feature #4334 (Resolved): rgw: dr: bucket index log API: implement RESTful API
- 03:38 PM rgw Feature #4333 (Resolved): rgw: multisite: metadata-changes log: implement RESTful API
- 03:38 PM rgw Feature #4329 (Resolved): rgw: dr: updated buckets log: RESTful API
- 03:38 PM rgw Feature #5008 (Resolved): rgw: bucket metadata changes should be reflected in mdlog
- 03:38 PM rgw Feature #4745 (Resolved): rgw: radosgw-admin command to stat object
- 03:37 PM rgw Feature #5358 (Resolved): rgw: RESTful api for intra-region copy state
- 03:37 PM rgw Feature #4613 (Resolved): Allow bucket data to reside in a separate pool to object data
- 03:37 PM rgw Feature #4330 (Resolved): rgw: dr: updated buckets log: radosgw-admin changes
- 03:37 PM rgw Feature #5417 (Resolved): rgw: separate bucket metadata object into pointer object and instance o...
- 03:37 PM rgw Feature #4328 (Resolved): rgw: dr: updated buckets log: tie into internal bucket changes tracker
- 03:37 PM rgw Feature #4327 (Resolved): rgw: dr: updated buckets log: create internal API
- 03:37 PM rgw Feature #4346 (Resolved): rgw: dr: bucket index objclass: changes
- 03:36 PM rgw Feature #4336 (Resolved): rgw: dr: sync processing state: implement internal RESTful API
- 03:35 PM rgw Feature #5406 (Resolved): rgw: a RESTful api to dump region map
- 03:35 PM rgw Feature #5408 (Resolved): rgw: turn off dr/geo logging
- 03:35 PM rgw Feature #4347 (Resolved): rgw: dr: bucket index objclass: fetch changes log
- 03:35 PM rgw Cleanup #5558 (In Progress): rgw: modify certain radosgw-admin operations interface
- 10:23 AM rgw Cleanup #5558 (Resolved): rgw: modify certain radosgw-admin operations interface
- We need to be more consistent with operations that read, write, list, and remove data:
region info ... - 03:34 PM rgw Feature #4309 (Resolved): rgw: multisite: metadata objects versioning
- 03:34 PM rgw Feature #4311 (Resolved): rgw: dr: radosgw changes: internal bucket changes tracker
- 03:34 PM rgw Feature #4331 (Resolved): rgw: multisite: metadata-changes log: create internal API
- 03:34 PM rgw Feature #4312 (Resolved): rgw: multisite: log metadata changes
- 03:34 PM rgw Feature #4332 (Resolved): rgw: multisite: metadata-changes log: tie into metadata update operations
- 03:31 PM Tasks #5585 (Resolved): test large scale exapnsion and contraction
- 03:30 PM Tasks #5584 (Resolved): measure peering performance
- 03:29 PM Tasks #5583 (Resolved): populate test cluster with rgw data
- 03:29 PM Tasks #5582 (Resolved): create large test cluster on burnupi and/or mira
- 03:18 PM Bug #5507: osd: ENOENT on clone
- 03:17 PM Messengers Bug #5508 (Need More Info): msg/SimpleMessenger.cc: 230: FAILED assert(!cleared)
- 03:17 PM Bug #5519 (Pending Backport): mon/osd: trimming of old maps based on last_epoch_clean is broken b...
- 03:14 PM Feature #3984 (Resolved): api: Send Out DRAFT REST API for Review
- 03:14 PM Feature #3983 (Resolved): api: create initial DRAFT REST API Design
- 03:12 PM Feature #3273: mon: simple dm-crypt key management
- Piston expressed interest in hooking this up to a PKI system they are looking at. Waiting on details from them.
- 03:11 PM Feature #5421 (In Progress): mon: add formatter option for various mon commands
- 03:10 PM Fix #5278 (Resolved): osd: smarter recovery for small objects
- 01:51 PM Bug #4599: ceph auth import -i <file> is broken
- Did it ever work from stdin? I don't see any code in the original tool that would
do that... - 01:40 PM Bug #5526 (Resolved): ceph health detail is broken for formatted output
- LGTM.
- 09:32 AM Bug #5526 (Fix Under Review): ceph health detail is broken for formatted output
- wip-rest commit commit:f810626a857dc34ff23d823d1b700488ff1798e8
- 12:32 PM Bug #5492 (In Progress): scripts installing into /usr/usr/sbin (with --prefix=/usr)
- This fix was temporarily reverted. The ceph build process for debian packages relies on binaries being found in /usr...
- 12:05 PM Bug #5524 (Resolved): df shows incorrect disk usage/size for cephfs mount
- 12:03 PM Bug #5524: df shows incorrect disk usage/size for cephfs mount
- I am using '3.8.0-19-generic' on Ubuntu 13.04.
If this was fixed in 3.9 then it is not really an issue going for... - 09:35 AM Bug #5524: df shows incorrect disk usage/size for cephfs mount
- what kernel version?
this was fixed in 3.9 or so - 12:04 PM Bug #5392: osd: unfound objects from thrashing
- ubuntu@teuthology:/var/lib/teuthworker/archive/teuthology-2013-07-08_20:00:15-rados-cuttlefish-testing-basic/59715
... - 11:47 AM Bug #5557 (Duplicate): /dev/disk/by-path not generated for ATA disks in later versions of udev
- the wip-ceph-disk no longer uses the by-path directory at all... see #5354
- 09:50 AM Bug #5557 (Duplicate): /dev/disk/by-path not generated for ATA disks in later versions of udev
- The version of udev in Ubuntu Saucy and higher does not generate /dev/disk/by-path entries for ATA disks.
This ude... - 11:25 AM devops Bug #4924: ceph-deploy: gatherkeys fails on raring (cuttlefish)
- Actually, this looks like it is being caused by the monitor crashing, and ceph-create-keys not being able to connect ...
- 11:19 AM devops Bug #4924: ceph-deploy: gatherkeys fails on raring (cuttlefish)
- I am seeing this same problem. I am using the latest master version of ceph-deploy, and the target node is Ubuntu 12....
- 09:37 AM CephFS Feature #5520: osdc: should handle namespaces
- This is needed before librbd or cephfs can use namespaces
- 09:14 AM rgw Bug #5478 (Resolved): rgw: fix init script for rgw daemon name
- 12:46 AM rgw Bug #5478: rgw: fix init script for rgw daemon name
- Should be closed as http://tracker.ceph.com/projects/ceph/repository/revisions/8b4cb8f37266183fe4e1925d07e50703e520f2...
- 08:13 AM Bug #5528 (Resolved): Python bindings Object.get_xattrs() requires unnecessary xattr_name
- get_xattrs() returns all xattrs for an object, but Object.get_xattrs() requires an xattr_name. This differs than Ioct...
- 07:35 AM Revision 76e79266 (ceph): rgw: fix args parsing
- Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
- 07:06 AM Revision 3c89a19c (ceph): os: Add missing pool to hobject_t:::dump() and hobject_t::decode()
- Signed-off-by: David Zafman <david.zafman@inktank.com>
- 07:06 AM Revision 9dd60a63 (ceph): TestRados: Output error for improper usage instead of Floating Point Ex...
- Signed-off-by: David Zafman <david.zafman@inktank.com>
- 07:06 AM Revision 72c27a30 (ceph): librados, osdc: Refactor IoCtxImpl to use operate()/operate_read()
- Add ObjectOperation::write() that includes len instead of using bufferlist length
Have selfmanaged_snap_rollback_obje... - 07:06 AM Revision b1b188a5 (ceph): os: Remove unused hobject_t::set_filestore_key()
- Signed-off-by: David Zafman <david.zafman@inktank.com>
- 07:06 AM Revision 30c951cc (ceph): osd: Fix object_locator_t::get_pool() return type
- Signed-off-by: David Zafman <david.zafman@inktank.com>
- 07:05 AM Revision 3931bfa6 (ceph): psim.cc: Fix comment on how to create .ceph_osdmap
- Signed-off-by: David Zafman <david.zafman@inktank.com>
- 07:05 AM Revision 7efbf5da (ceph): librados: Fix lock names
- Signed-off-by: David Zafman <david.zafman@inktank.com>
- 07:05 AM Revision 313b7a1f (ceph): os: Code conformance in LFNIndex.cc
- Signed-off-by: David Zafman <david.zafman@inktank.com>
- 07:04 AM Revision 395262cd (ceph): rgw: call appropriate curl calls for waiting on sockets
- If libcurl supports curl_multi_wait() then use it, otherwise
use select() and force a timeout, even if it has been di... - 06:43 AM Revision 73c2a3dc (ceph): configure.ac: detect whether libcurl supports curl_multi_wait()
- Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
- 06:17 AM Revision d4ca36b3 (ceph): Merge branch 'next'
- 05:41 AM Subtask #5510 (Fix Under Review): ObjectContext : replace ref with shared_ptr
- 12:47 AM Subtask #5510 (In Progress): ObjectContext : replace ref with shared_ptr
- 05:22 AM rgw Feature #3074: radosgw needs --help support
- This issue should be closed.
- 05:22 AM Revision b325d170 (ceph): lock: fix typo
- 05:17 AM Revision 8ff62ae4 (ceph): Merge remote-tracking branch 'gh/next'
- 05:17 AM Revision d08b6d6d (ceph): mon/PaxosService: prevent reads until initial service commit is done
- Do not process reads (or, by PaxosService::dispatch() implication, writes)
until we have committed the initial servic... - 04:44 AM Revision 63fe8635 (ceph): mon/PaxosService: unwind should_trim()
- Inline the single-caller helper. This will help us in a moment...
Signed-off-by: Sage Weil <sage@inktank.com> - 04:41 AM Revision d600dc93 (ceph): mon/PaxosService: unwind service_should_trim() helper
- Nobody overloads it; put it inline in should_trim().
Signed-off-by: Sage Weil <sage@inktank.com> - 04:41 AM Revision 6aa02304 (ceph): mon/MDSMonitor: remove unnecessary service_should_trim()
- We never set_trim_to(), so this is unnecessary.
Signed-off-by: Sage Weil <sage@inktank.com> - 04:40 AM Revision b71a0096 (ceph): mon/OSDMonitor: remove dup service_should_trim() implementation
- This matches the parent.
Signed-off-by: Sage Weil <sage@inktank.com> - 04:38 AM Revision 39b71c58 (ceph): mon/PaxosService: trim periodically instead of via propose_pending
- We want to trim old states even if there is no update activity. For
example, if a long-running rebalance finishes al... - 04:33 AM Revision 2f8ff2de (ceph): mon/PaxosService: reorder definitions
- Signed-off-by: Sage Weil <sage@inktank.com>
- 04:33 AM Revision 50ffe324 (ceph): mon/PaxosService: uninline should_trim()
- Signed-off-by: Sage Weil <sage@inktank.com>
- 04:10 AM Revision bfc26c65 (ceph): mon/OSDMonitor: fix base case for loading full osdmap
- Right after cluster creation, first_committed is 1 and latest stashed in 0,
but we don't have the initial full map ye... - 01:11 AM Revision f1b4398d (ceph): Merge branch 'master' of https://github.com/ceph/ceph
- 01:11 AM Revision 5edc1ff7 (ceph): doc: Added Ceph Object Storage installation instructions for CentOS/RHE...
- Signed-off-by: John Wilkins <john.wilkins@inktank.com>
- 12:49 AM Revision b536935f (ceph): mon/MonitorDBStore: expose get_chunk_tx()
- Allow users get the transaction unencoded.
Signed-off-by: Sage Weil <sage@inktank.com> - 12:49 AM Revision ca54efd6 (ceph): mon: sync all service prefixes, including pgmap_*
- This was just recently broken with the merge of the pgmap changes.
Signed-off-by: Sage Weil <sage@inktank.com> - 12:49 AM rgw Bug #5374: Avoid relying on keystone's admin token
- The patch above needs review.
Please change ticket status. - 12:47 AM rgw Bug #1779: rgw: swift auth returns wrong error code when unexisting user is given
- I need some advice here, if it need some backport, and if so where?
- 12:47 AM Revision 43fa7aab (ceph): mon/OSDMonitor: fix base case for loading full osdmap
- Right after cluster creation, first_committed is 1 and latest stashed in 0,
but we don't have the initial full map ye... - 12:10 AM Subtask #5527 (Resolved): unit tests for common/sharedptr_registry.hpp
- "work in progress":https://github.com/dachary/ceph/tree/wip-5527
07/08/2013
- 11:54 PM Revision d18fe740 (ceph): VM: Use mac addresses from DB instead of randomizing.
- In order to make IP addresses less likely to change and to allow
a smaller DHCP pool to be used I generated static MA... - 11:53 PM Revision 0e93dd93 (ceph): Merge branch 'wip-small-object-recovery'
- Conflicts:
src/include/ceph_features.h
Reviewed-by: Sage Weil <sage@inktank.com>
Fixes: #5278 - 11:52 PM Bug #5526 (Resolved): ceph health detail is broken for formatted output
- ceph health detail appends plain-text detail output whether or not a formatter is requested
- 11:43 PM Revision ae1b2e97 (ceph): ReplicatedPG: add handlers for MOSDPG(Push|Pull|PushReply)
- Signed-off-by: Samuel Just <sam.just@inktank.com>
- 11:43 PM Revision ad65de40 (ceph): ReplicatedPG: send compound messages to enlightened peers
- Signed-off-by: Samuel Just <sam.just@inktank.com>
- 11:43 PM Revision eec86b8d (ceph): OSD: convert handle_push to use PushOp
- Signed-off-by: Samuel Just <sam.just@inktank.com>
- 11:43 PM Revision c41d4dc4 (ceph): ReplicatedPG: send pushes en mass in recover_replicas, recover_backfill
- This way, the pushes might be later merged into a smaller number of
messages.
Signed-off-by: Samuel Just <sam.just@i... - 11:43 PM Revision 54e5f642 (ceph): ReplicatedPG: send pulls en masse in recover_primary
- Signed-off-by: Samuel Just <sam.just@inktank.com>
- 11:43 PM Revision 175c0777 (ceph): ReplicatedPG: split handle_push_reply out of sub_op_push_reply
- Signed-off-by: Samuel Just <sam.just@inktank.com>
- 11:43 PM Revision c56f16d4 (ceph): ReplicatedPG: split handle_pull out of sub_op_pull
- Signed-off-by: Samuel Just <sam.just@inktank.com>
- 11:43 PM Revision 264dbf3f (ceph): messages/,osd_types: add messages for Push, PushReply, Pull
- Signed-off-by: Samuel Just <sam.just@inktank.com>
- 11:43 PM Revision c0bd831a (ceph): OSD: add handlers for MOSDPG(Push|PushReply|Pull)
- MOSDPG(Push|PushReply|Pull|SubOp|SubOpReply) need the
same thing checked prior to queueing the op, so they
share a te... - 11:43 PM Revision 0f51b60c (ceph): ReplicatedPG: submit_push_data must take recovery_info as non-const
- Signed-off-by: Samuel Just <sam.just@inktank.com>
- 11:43 PM Revision 31e19a64 (ceph): ReplicatedPG: _committed_pushed_object don't pass op
- Add a separate callback to handle marking the event and
the stats.
Signed-off-by: Samuel Just <sam.just@inktank.com> - 11:43 PM Revision 82cb922e (ceph): ReplicatedPG: split send_push into build_push_op and send_push_op
- Signed-off-by: Samuel Just <sam.just@inktank.com>
- 11:43 PM Revision a4984328 (ceph): ReplicatedPG: pass a PushOp into handle_pull_response
- This is the first step toward packaging multiple
pushes/pulls into a single message.
Signed-off-by: Samuel Just <sam... - 11:11 PM Revision 24f90b83 (ceph): mon: implement simple 'scrub' command
- Compare all keys within the sync'ed prefixes across members of the quorum
and compare the key counts and CRC for inco... - 11:11 PM Revision 7fb3804f (ceph): mon: fix osdmap stash, trim to retain complete history of full maps
- The current interaction between sync and stashing full osdmaps only on
active mons means that a sync can result in an... - 11:11 PM Revision 95ef961d (ceph): Pipe: use uint64_t not unsigned when setting features
- Fixes: #5497
Signed-off-by: Samuel Just <sam.just@inktank.com>
Reviewed-by: Sage Weil <sage@inktank.com>
Reviewed-by:... - 11:11 PM Revision c2b38291 (ceph): ceph_features.h: declare all features as ULL
- Otherwise, the first 32 get |'d together as ints. Then, the result
((int)-1) is sign extended to ((long long int)-1)... - 11:11 PM Revision 926f723c (ceph): Elector.h: features are 64 bit
- Fixes: #5497
Signed-off-by: Samuel Just <sam.just@inktank.com>
Reviewed-by: Sage Weil <sage@inktank.com>
Reviewed-by:... - 10:45 PM Revision b6b48dbe (ceph): v0.66
- 10:34 PM Revision a9906641 (ceph): mon: implement simple 'scrub' command
- Compare all keys within the sync'ed prefixes across members of the quorum
and compare the key counts and CRC for inco... - 10:04 PM Revision afd6c7d8 (ceph): mon: fix osdmap stash, trim to retain complete history of full maps
- The current interaction between sync and stashing full osdmaps only on
active mons means that a sync can result in an... - 10:01 PM Revision 0a13124c (ceph): rados/mon_thrash: inject random delays in mon messages
- More useful than osd delays for this collection
- 10:01 PM Revision 5295406d (ceph): rados: expand mon thrashing tests
- - test sync (full and recent)
- test multiple mons down
- test other cluster sizes (3 and 9)
Signed-off-by: Sage Wei... - 10:01 PM Revision dbf12a31 (ceph): rados/monthrash: refactor a bit
- Signed-off-by: Sage Weil <sage@inktank.com>
- 10:01 PM Revision 94ae48cf (ceph): rados/monthrash: reduce matrix size a bit
- we don't need so many osd workloads
- 09:52 PM Revision c3b68493 (ceph): osd/osd_types: fix pg_stat_t::dump for last_epoch_clean
- Backport: cuttlefish
Signed-off-by: Sage Weil <sage@inktank.com>
(cherry picked from commit 69a55445439fce0dd6a3d32ff... - 09:52 PM Revision 09d258b7 (ceph): client: remove O_LAZY
- The once-upon-a-time unique O_LAZY value I chose forever ago is now
O_NOATIME, which means that some clients are choo... - 09:49 PM Revision dd1e6d45 (ceph): Revert "Makefile: fix ceph_sbindir"
- This reverts commit 352f362567bf270d0896fb7573df4ae5139a56fb.
Reverting this commit because it causes problems with ... - 09:16 PM Bug #5515 (Pending Backport): mon: paxos allows reads before initial commit is done
- 10:48 AM Bug #5515 (Fix Under Review): mon: paxos allows reads before initial commit is done
- 09:31 AM Bug #5515 (In Progress): mon: paxos allows reads before initial commit is done
- 09:14 AM Bug #5515 (Resolved): mon: paxos allows reads before initial commit is done
- ubuntu@teuthology:/a/teuthology-2013-07-07_20:00:17-rados-cuttlefish-testing-basic/58476
got an auth failure befor... - 09:02 PM Revision f07d2167 (ceph): rgw: fix bucket link
- Bucket link was assuming the bucket head object was holding the
bucket acl, which is not true anymore.
Signed-off-by... - 08:35 PM Revision b55e455e (ceph): Merge pull request #395 from kri5/wip-vstart-documentation
- doc: Add a page to document vstart.sh script
- 08:19 PM Revision eec903a6 (ceph): doc: Fix env variables in vstart.sh documentation
- Signed-off-by: Christophe Courtaut <christophe.courtaut@gmail.com>
- 07:55 PM Revision 69a55445 (ceph): osd/osd_types: fix pg_stat_t::dump for last_epoch_clean
- Backport: cuttlefish
Signed-off-by: Sage Weil <sage@inktank.com> - 07:45 PM Revision 3cbf6573 (ceph): Merge pull request #403 from ceph/wip-olazy
- merge: O_LAZY flag removal
Reviewed-by: Greg Farnum <greg@inktank.com - 07:23 PM Revision 9f8bfb4b (ceph): Merge pull request #397 from kri5/wip-5478
- rgw: Add explicit messages in radosgw init script
Reviewed-by: Yehuda Sadeh <yehuda@inktank.com> - 06:43 PM Revision 784ce258 (ceph): Merge pull request #406 from kri5/wip-3074
- rgw: Add --help support to radosgw
Reviewed-by: Yehuda Sadeh <yehuda@inktank.com> - 06:39 PM rgw Documentation #5525 (Resolved): Radosgw 'add the ceph keyring entries' section should be updated ...
- When using ceph-deploy a key file named 'ceph.client.admin.keyring ' is generated.
When attempting to follow the s... - 06:27 PM Bug #5524 (Resolved): df shows incorrect disk usage/size for cephfs mount
- After mounting a cephfs volume (ceph version 0.61.4) and filling it with 11GB of .wav files I used 'df' to list the m...
- 06:24 PM Revision 94afedf0 (ceph): client: remove O_LAZY
- The once-upon-a-time unique O_LAZY value I chose forever ago is now
O_NOATIME, which means that some clients are choo... - 05:55 PM CephFS Bug #5036: `ls` hangs on random folder
- was your mds complied from the newest source code? was there mds restart before you saw the hang? if there was, the b...
- 01:03 PM CephFS Bug #5036: `ls` hangs on random folder
- Yan,
Even after rebuilding my 3.10 kernel with the missing fix (libceph: call r_unsafe_callback when unsafe reply ... - 05:54 PM Revision e9d19b38 (ceph): common/crc32c: skip cpu detection incantation on not x86_64
- On i386 this fails to build with
common/crc32c-intel.c: In function 'ceph_have_crc32c_intel':
error: common/crc32c-i... - 05:44 PM Revision 0471b719 (ceph): Merge pull request #407 from dachary/wip-5487
- unit tests for ObjectContext read/write locks
Reviewed-by: Samuel Just <sam.just@inktank.com> - 05:40 PM devops Feature #5523 (Resolved): libcurl 7.28+ packages
- We need to provide libcurl 7.28 or newer packages for relevant architectures that don't have it. It's needed for rgw ...
- 05:40 PM Revision 497c71e3 (ceph): Merge pull request #17 from ceph/wip-mon-thrash
- mon thrash improvements
Reviewed-by: Josh Durgin <josh.durgin@inktank.com> - 05:39 PM Revision a0457492 (ceph): mon_thrasher: add pause/unpause of mons to thrashing
- This adds an additional element of laggyness to the cluster which should
cause mons to call new elections.
Signed-of... - 05:39 PM Revision fdebf40f (ceph): daemon-helper: send arbitrary signals via stdin
- Each byte written to stdin will be interpreted as a signal.
Signed-off-by: Sage Weil <sage@inktank.com> - 05:39 PM Revision 121b1b9f (ceph): mon_thrash: optionally scrub after each iteration (default true)
- Signed-off-by: Sage Weil <sage@inktank.com>
- 05:14 PM Revision 956fafc7 (ceph): qa/workunits/rbd/simple_big.sh: don't ENOSPC every time
- Set the count on the initial dd so we don't always ENOSPC.
Signed-off-by: Sage Weil <sage@inktank.com> - 04:58 PM Revision d423cf8c (ceph): qa/workunits/rbd/kernel.sh: move modprobe up
- Needs to happen before cleanup.
Signed-off-by: Sage Weil <sage@inktank.com> - 04:56 PM Revision 672f51be (ceph): qa/workunits/fs/test_o_trunc.sh: fix .sh to match new bin location
- To match 83f308962c53eec10db9e496987a9e4be7c87e9b.
Signed-off-by: Sage Weil <sage@inktank.com> - 04:49 PM devops Bug #5369 (Resolved): fedora18: sysvinit doesn't start mon on reboot
- Doing a chkconfig --list wasn't listing the network manager so it threw me off a bit but it was indeed enabled. I tri...
- 04:37 PM Revision eb567beb (ceph): Merge pull request #389 from ceph/wip-5492
- Makefile: fix ceph_sbindir
Reviewed-by: Mark Kirkwood <mark.kirkwood@catalyst.net.nz> - 04:33 PM Revision a02f2510 (ceph): mon: remove bad assert about monmap version
- It is possible to start a sync when our newest monmap is 0. Usually we see
e0 from probe, but that isn't always publ... - 04:23 PM rgw Bug #5522: rgw: use of select for waiting on curl sockets
- Solution that we discussed is to modify code to use curl_multi_wait() and to provide backported packages for the rele...
- 04:22 PM rgw Bug #5522 (Resolved): rgw: use of select for waiting on curl sockets
- We should use curl_multi_wait() instead. Main issue is that this function only available in more recent libcurl packa...
- 03:32 PM Feature #5521 (Duplicate): Enhance PGLS or new op to list all namespace/objects in a pool.
Sage pull request comment:
allow a 'any namespace' flag for PGLS.. maybe a different op code? (PGLSALL, or PGLS_...- 03:10 PM Bug #5512 (Pending Backport): mon: missing full osdmaps after sync
- 02:52 PM Bug #5492: scripts installing into /usr/usr/sbin (with --prefix=/usr)
- the proposed fix breaks the deb builds
- 10:58 AM Bug #5492 (Resolved): scripts installing into /usr/usr/sbin (with --prefix=/usr)
- 02:47 PM CephFS Feature #5520 (Rejected): osdc: should handle namespaces
As a follow on to 4982/4983 we should implement namespace handling in the ObjectCacher.- 02:45 PM Revision 7b7f752c (ceph): unit tests for ObjectContext read/write locks
- unit tests for the ObjectContext methods ondisk_write_lock,
ondisk_write_unlock, ondisk_read_lock and ondisk_read_unl... - 01:36 PM Bug #5519 (Fix Under Review): mon/osd: trimming of old maps based on last_epoch_clean is broken b...
- 01:00 PM Bug #5519 (Resolved): mon/osd: trimming of old maps based on last_epoch_clean is broken by design
- consider a cluster with one pg whose mapping does not change for 10000 epochs, while other pgs recover and then go cl...
- 12:21 PM Bug #5518 (Resolved): osd: marking single osd down makes others go down (cuttlefish)
- Settings:
- paxos propose interval = 1
- debug ms = 1
- debug osd = 20
Log:
07:57: Cluster health OK.
07... - 12:18 PM Bug #5517 (Resolved): osd: stuck peering on cuttlefish
- Settings:
- paxos propose interval = 1
- debug ms = 1
- debug osd = 20
- debug mon = 20
Log:
11:45: star... - 11:33 AM devops Bug #5345: ceph-disk: handle less common device names
- Hi Luke, Tomas,
Are you able to test the latest version in this branch? https://raw.github.com/ceph/ceph/wip-ceph... - 11:20 AM rgw Bug #5516 (Resolved): rgw: update bucket relink teuthology test
- In teuthology/task/radosgw-admin.py there is a test for relinking a bucket from one user to another. Here is the code...
- 09:40 AM Revision 6f1653a6 (ceph): rgw: Add --help support to radosgw
- http://tracker.ceph.com/issues/3074 fixes #3074
The support of --help option is added through this patch.
By now, it... - 09:33 AM Messengers Bug #5508: msg/SimpleMessenger.cc: 230: FAILED assert(!cleared)
- 09:32 AM Bug #5509 (Resolved): mon/Monitor.cc: 1395: FAILED assert(latest_monmap.epoch > 0)
- 09:02 AM RADOS Bug #5514: mon: can get stuck in tight loop with old rotating keys
- Isn't this going to clear itself up as soon as the mon generates new keys anyway?
Besides only being visible to peop... - 08:58 AM RADOS Bug #5514 (New): mon: can get stuck in tight loop with old rotating keys
- osd sees rotating keys are out of date, requests new ones
mon returns out of date keys
osd loops
this happened t... - 04:56 AM Revision a80ab93d (ceph): big: disable all but small until it passes
- 04:20 AM Revision 8bc50626 (ceph): Merge branch 'next'
- 04:19 AM Revision 85a1d6cc (ceph): mon: remove bad assert about monmap version
- It is possible to start a sync when our newest monmap is 0. Usually we see
e0 from probe, but that isn't always publ...
07/07/2013
- 08:40 PM Bug #5482: cephx: verify_reply coudln't decrypt with error: error decoding block for decryption
- The kernel version doesn't matter if you are just running the ceph userspace.
If you are mounting cephfs via the k... - 07:42 PM Bug #5482: cephx: verify_reply coudln't decrypt with error: error decoding block for decryption
- Sage Weil wrote:
> I actually meant 0.56.6, but you can move to 0.61.x (cuttelfish) too. See ceph.com/docs/master i... - 06:37 PM Revision 823bf385 (ceph): Revert "rbd: reeanble iozone test, just to see"
- This reverts commit 81862ad907e9ba94f0a2dfc9ed365b68f305cd8d.
Yeah, still broken. - 09:03 AM Bug #5513 (Can't reproduce): osd crashes consistently after startup
- I can't get one of my OSDs to start up, it gives the following log output (with debug osd = 20): https://pastee.org/u...
- 08:36 AM Bug #5512 (Resolved): mon: missing full osdmaps after sync
- full osdmaps are occasionally part of paxos transactions. if one of these overlaps with a sync, the synced monitor w...
- 06:03 AM Subtask #5510 (Fix Under Review): ObjectContext : replace ref with shared_ptr
07/06/2013
- 07:43 AM Linux kernel client Feature #190: krbd: DISCARD support
- These seems like a pretty important feature to me. Without this feature the storage necessary for a highly active (b...
- 07:29 AM Feature #5511 (Duplicate): rados.py support for object locking
- It seems like an easy task to expose to the python bindings the following functions:
rados_lock_exclusive()
rados_l... - 05:27 AM Subtask #5510 (Resolved): ObjectContext : replace ref with shared_ptr
- "work in progress":https://github.com/dachary/ceph/tree/wip-5510
"take 1":https://github.com/dachary/ceph/commit/a50...
07/05/2013
- 09:30 PM Revision 389f65e7 (ceph): mon_thrash: use _ instead of - consistently
- Signed-off-by: Sage Weil <sage@inktank.com>
- 09:30 PM Revision 9b22d388 (ceph): mon_thrash: fix more naming
- Signed-off-by: Sage Weil <sage@inktank.com>
- 06:24 PM Revision 288b5444 (ceph): kernel: simple_1tb -> simple_big
- 1TB is too big for plana
- 06:24 PM Revision 3f5a9623 (ceph): qa: write a somewhat <1tb image
- 1TB is enough to fill up 6 plana osds. And it takes forever. Write less.
Signed-off-by: Sage Weil <sage@inktank.com> - 06:20 PM Revision 54aa797a (ceph): qa/workunits/rbd/kernel.sh: modprobe rbd
- Signed-off-by: Sage Weil <sage@inktank.com>
- 06:18 PM Revision 84d76517 (ceph): test_o_trunc.sh moved
- 06:17 PM Revision 83f30896 (ceph): qa: move test_o_trunc.sh into fs dir
- Signed-off-by: Sage Weil <sage@inktank.com>
- 06:16 PM Revision 507a4ec8 (ceph): qa: move fs test binary into workunits dir so teuthology can build it
- Teuthology does a make in the workunits dir, so move this in there.
Signed-off-by: Sage Weil <sage@inktank.com> - 06:04 PM Revision a84e6d18 (ceph): mds/MDSTable: gracefully suicide on EBLACKLIST
- Signed-off-by: Sage Weil <sage@inktank.com>
Reviewed-by: Greg Farnum <greg@inktank.com> - 05:07 PM Revision 6ad9fe17 (ceph): mon/Paxos: fix sync restart
- If we have a sync going, and an election intervenes, the client will
try to continue by sending a new start_chunks re... - 05:06 PM Revision c5812b1c (ceph): mon: uninline _trim_enable and Paxos::trim_{enable,disable} so we can d...
- Signed-off-by: Sage Weil <sage@inktank.com>
- 04:21 PM Messengers Bug #5508 (In Progress): msg/SimpleMessenger.cc: 230: FAILED assert(!cleared)
- 11:26 AM Messengers Bug #5508 (Resolved): msg/SimpleMessenger.cc: 230: FAILED assert(!cleared)
- ...
- 04:04 PM Bug #5509 (Fix Under Review): mon/Monitor.cc: 1395: FAILED assert(latest_monmap.epoch > 0)
- 03:45 PM Bug #5509 (Resolved): mon/Monitor.cc: 1395: FAILED assert(latest_monmap.epoch > 0)
- ...
- 12:41 PM Revision 8b4cb8f3 (ceph): rgw: Add explicit messages in radosgw init script
- http://tracker.ceph.com/issues/5478 fixes #5478
Signed-off-by: Christophe Courtaut <christophe.courtaut@gmail.com> - 11:47 AM Bug #5502 (Fix Under Review): mon: long-running sync will restart (cuttlefish)
- 11:40 AM Bug #5495 (Can't reproduce): ceph-mon and minus character in hostname
- 08:40 AM Bug #5507 (Resolved): osd: ENOENT on clone
- ...
- 08:39 AM Bug #5445: random osd EPERM on journal
- teuthology-2013-07-05_01:00:13-rados-master-testing-basic 55351: and 55360:
- 06:44 AM Revision d09ce3df (ceph): rgw: fix rgw_remove_bucket()
- function was referring bucket info object directly, instead of going
through helper functions, which is now a must.
... - 01:58 AM Bug #5504: osd stack on peereng for a long time
- ceph version 0.56.6 (95a0bda7f007a33b0dc7adf4b330778fa1e5d70c)
6 mons at at 6 servers
156 osds: 156 up, 156 in
648... - 12:09 AM Revision 6fbcbd7f (ceph): mon/Paxos: increase paxos max join drift
- A value of 10 is too aggressive for large, long-running syncs. 100 is
about 2 minutes of activity at most, which shou... - 12:09 AM Revision f3a51fa3 (ceph): mon/Paxos: configure minimum paxos txns separately
- We were using paxos_max_join_drift to control the minimum number of
paxos transactions to keep around. Instead, make... - 12:09 AM Revision 1156721f (ceph): mon: include any new paxos commits in each sync CHUNK message
- We already take note of the paxos version when we begin the sync. As
sync progresses and there are new paxos commits...
07/04/2013
- 11:55 PM rgw Feature #5506 (Resolved): rgw: use Keystone to authenticate S3 requests
- The idea is that there should be an alternative way to authenticate S3 requests. Currently we handle the S3 authentic...
- 09:31 PM Bug #5497: ceph features mis-reported
- I'm pretty sure that a pre-fix OSD is going to report that it supports all features to a post-fix OSD — right? That's...
- 08:57 PM Bug #5497: ceph features mis-reported
- the new command syntax (json instead of vector<string), and new MonCap encoding. the old protocol won't ever see a >...
- 07:49 PM Bug #5497: ceph features mis-reported
- What was the mon protocol change covering (I missed it going in), and don't we need to preserve compatibility across ...
- 07:52 PM RADOS Feature #5280: osd/client: messages should be tagged with the earliest sane map
- Couldn't we pretty easily on the client side just keep track of the last map which changed the osd states or the pg_t...
- 07:17 PM Revision 40672219 (ceph): mon/MonitorDBStore: expose get_chunk_tx()
- Allow users get the transaction unencoded.
Signed-off-by: Sage Weil <sage@inktank.com> - 04:00 PM Revision 972d22e2 (ceph): Adding new Wireshark dissector. This is loosely based on the original
- dissector, it has been re-worked to bring into line with Wireshark
coding guidelines that aim to ensure portability. ... - 12:08 PM Revision a793e203 (ceph): doc: Add a page to document vstart.sh script
- Signed-off-by: Christophe Courtaut <christophe.courtaut@gmail.com>
- 11:54 AM Bug #5505 (Resolved): mon health does not reflect slow requests
- 06:56 AM rgw Bug #5416: --help output needs --rgw-zone option
- I must be blind because I can't find it :-) Would you be so kind as to paste the github URL where it can be found ?
- 05:59 AM Bug #5504 (Duplicate): osd stack on peereng for a long time
- PGs from one sometimes osd stacks on peereng for a long time.
--
ceph health details
HEALTH_WARN 3 pgs peering; ... - 05:08 AM Revision 81862ad9 (ceph): rbd: reeanble iozone test, just to see
- See #2689
- 05:04 AM Revision c14847c3 (ceph): .gitignore: cls_test_*
- Signed-off-by: Sage Weil <sage@inktank.com>
- 04:56 AM Revision 998a5134 (ceph): fs: add o_trunc test on ceph-fuse
- Signed-off-by: Sage Weil <sage@inktank.com>
- 04:55 AM Revision 89cd717d (ceph): kernel: test o_trunc
- Signed-off-by: Sage Weil <sage@inktank.com>
- 04:53 AM Revision 22227cd1 (ceph): qa: add O_TRUNC test
- From: Yan, Zheng <yan.zheng@intel.com>
Simple reproducer for #5453, modified to run for a finite number of
iteration... - 04:43 AM Revision 46b7fc2e (ceph): radosgw-admin: fix cli test
- Signed-off-by: Sage Weil <sage@inktank.com>
- 02:23 AM Bug #5495: ceph-mon and minus character in hostname
- Sage Weil wrote:
> what version is this?
This are the official 0.61.4 packages from ceph.com
> can you strace ... - 02:07 AM Revision d54932cb (ceph): Fix VM issues.
- Fix of #5494 although bad description. Instead of adding a wait
the code used to detect if the guest was back up is f... - 01:08 AM Revision a0b1be99 (ceph): rgw: fix type encoding
- due to bad merge
Signed-off-by: Yehuda Sadeh <yehuda@inktank.com> - 12:15 AM Revision db2bb270 (ceph): mon: enable leveldb cache by default
- 256 is not as large as the upstream 512 MB, but will help signficiantly and
be less disruptive for existing cuttlefis... - 12:09 AM Revision d30fc4bd (ceph): Merge remote-tracking branch 'origin/master' into wip-rgw-geo
- Conflicts:
src/Makefile.am
src/include/rados/librados.hpp
src/rgw/rgw_admin.cc
src/rgw/rgw_bucket.cc
src/rgw/rgw... - 12:00 AM Bug #5401: cuttlefish osd recovery slow
- OK reverting them to the defaults seems to work fine too.
07/03/2013
- 11:57 PM Revision 123f676e (ceph): mon/Paxos: make 'paxos trim disabled max versions' much much larger
- 108000 is about 3 hours if paxos is going full-bore (1 proposal/second).
That ought to be pretty safe. Otherwise, we... - 11:56 PM Revision 71ebfe7e (ceph): mon/Paxos: make 'paxos trim disabled max versions' much much larger
- 108000 is about 3 hours if paxos is going full-bore (1 proposal/second).
That ought to be pretty safe. Otherwise, we... - 11:49 PM Bug #5401: cuttlefish osd recovery slow
- Sorry solution is the wrong word - minimize the effect (it'sstill there but not as much as before)
- 11:47 PM Bug #5401: cuttlefish osd recovery slow
- Nobody - i was just helpless and tried to find a solution for me.
- 03:35 PM Bug #5401: cuttlefish osd recovery slow
- What prompted you to set 'filestore min sync interval' and 'filestore max sync interval'? The defaults are 0.01 and ...
- 01:33 AM Bug #5401: cuttlefish osd recovery slow
- Sage - any news / ideas?
- 11:23 PM Revision ab93696e (ceph): mon: be less chatty about discarding messages
- Signed-off-by: Sage Weil <sage@inktank.com>
- 11:17 PM Revision 03393c0d (ceph): mon: do not reopen MonitorDBStore during startup
- level doesn't seem to like this when it races with an internal compaction
attempt (see below). Instead, let the stor... - 10:42 PM Bug #5503 (Resolved): osd: ceph --admin-daemon interface doesn't handle spaces in names
The simple parsing mechanism string_to_vec() uses spaces to separate arguments. There is no processing of quotes. ...- 10:37 PM Revision e8b42a69 (ceph): osd/OSDMap: handle case where some new osds have hb_front and others don't
- Do not assume that because at least one OSD has an hb_front addr that they
all do, or else we will end up assigning g... - 10:37 PM Revision 81343f1d (ceph): osd: clear hb_front if it was previously non-NULL and is now NULL
- If we have a real addr for hb_front for a given osd and then a new map
has the osd coming up without an hb_front, we ... - 10:36 PM Feature #4982: OSD: namespaces pt 1 (librados/osd, not caps)
- 10:27 PM Revision e960e1bb (ceph): Merge branch 'master' of https://github.com/ceph/ceph
- 10:26 PM Revision e0da832a (ceph): doc: Added write caps. Required for auto-creating pools.
- Signed-off-by: John Wilkins <john.wilkins@inktank.com>
- 10:24 PM Revision 01d3e094 (ceph): osd: fix race when queuing recovery ops
- Previously we would sample how many ops to start under the lock, drop it,
and start that many. This is racy because ... - 10:13 PM devops Bug #5211 (Resolved): ceph-disk prepare: list_partitions() shouldn't return disks
- 10:12 PM Bug #5330 (Resolved): ceph daemon <name> ... broken
- commit:88f73c5a6363eebe5d51d2636bc1328fec40dd2c
- 10:11 PM CephFS Bug #4850 (Resolved): ceph-fuse: disconnected inode on shutdown with fsstress + mds thrashing
- i think this is resolved now...
- 10:02 PM RADOS Feature #5280: osd/client: messages should be tagged with the earliest sane map
- i started coding this up for the objecter but realized it doesn't really help.. when we initiate a io we calculate th...
- 09:58 PM Feature #5137 (Resolved): osd: magically fall back to leveldb for xattrs
- 09:58 PM Feature #4983 (Fix Under Review): OSD: namespaces pt 2 (caps)
- 09:56 PM CephFS Bug #5453 (Resolved): kclient: multiple_rsync tee output partially zeroed
- thanks, added this to the test suite
- 06:32 PM CephFS Bug #5453: kclient: multiple_rsync tee output partially zeroed
- my reproducer...
- 10:31 AM CephFS Bug #5453: kclient: multiple_rsync tee output partially zeroed
- patch is in testing branch (tho i'm tracking down a different regression in that branch).
btw, yan, were you able ... - 06:03 AM CephFS Bug #5453: kclient: multiple_rsync tee output partially zeroed
- Sage Weil wrote:
> in combination with the mds s/wrlock/xlock/ change?
the kclient patch fixes this issue alone. ... - 09:46 PM devops Bug #5345 (Fix Under Review): ceph-disk: handle less common device names
- 07:05 AM devops Bug #5345 (In Progress): ceph-disk: handle less common device names
- Thanks, Luke--that was exactly the info I needed!
- 12:22 AM devops Bug #5345: ceph-disk: handle less common device names
- Hi Sage,
Just tried the ceph-disk as per your suggestion, however I found the following error:
ceph-deploy osd ... - 09:46 PM Bug #5497 (Resolved): ceph features mis-reported
- i don't think we need to backport this.. the mons have a protocol version change and can't talk from cuttlefish -> 0....
- 01:22 PM Bug #5497 (Pending Backport): ceph features mis-reported
- 11:26 AM Bug #5497 (Resolved): ceph features mis-reported
- I observed some osd daemons with 32 bit ~0 as features while others have 64 bit ~0.
- 08:58 PM Revision daee9dbe (ceph): ObjectStore,Context: add register_on_complete
- Signed-off-by: Samuel Just <sam.just@inktank.com>
- 08:58 PM Revision b0e339ce (ceph): ObjectStore: only register non-null contexts
- Signed-off-by: Samuel Just <sam.just@inktank.com>
- 08:58 PM Revision 194e858c (ceph): osd: do not use temp_coll for single-step pushes
- If we are recovering an object in a single step, there is no need to
write it to temp and then move it. Avoiding tha... - 08:58 PM Revision d87ae0a2 (ceph): osd_types: add PushOp, PushReplyOp PullOp
- Signed-off-by: Samuel Just <sam.just@inktank.com>
- 08:23 PM Revision a825ae7a (ceph): rgw: remove a few warnings
- Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
- 08:21 PM Revision 4255b5c2 (ceph): ceph_features.h: declare all features as ULL
- Otherwise, the first 32 get |'d together as ints. Then, the result
((int)-1) is sign extended to ((long long int)-1)... - 08:21 PM Revision 3564e304 (ceph): Elector.h: features are 64 bit
- Fixes: #5497
Signed-off-by: Samuel Just <sam.just@inktank.com>
Reviewed-by: Sage Weil <sage@inktank.com>
Reviewed-by:... - 08:21 PM Revision bc3e2f09 (ceph): Pipe: use uint64_t not unsigned when setting features
- Fixes: #5497
Signed-off-by: Samuel Just <sam.just@inktank.com>
Reviewed-by: Sage Weil <sage@inktank.com>
Reviewed-by:... - 08:20 PM Revision 14ab4301 (ceph): Merge pull request #393 from ceph/wip-crc
- use sse4.2 crc32c instruction
- 08:17 PM Revision dc29c7a4 (ceph): rgw: renamed a few unit tests
- Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
- 08:14 PM Revision db63965a (ceph): ceph.spec.in: Fix file name typo
- Signed-off-by: Gary Lowell <gary.lowell@inktank.com>
- 08:04 PM Revision c2873c1d (ceph): rgw: make rgw_cls_complete_op backward compatible
- Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
- 07:20 PM Revision 08dd35d9 (ceph): common: autoselect crc32c based on cpu features
- If the CPu supposts SSE4.2, use the crc32c instructions. Use the magic
incantation from who knows where to do this. ... - 07:20 PM Revision 7c59288d (ceph): Initial Intel SSE4 crc32c implementation.
- Signed-off-by: Mark Nelson <mark.nelson@inktank.com>
- 07:05 PM Bug #5502: mon: long-running sync will restart (cuttlefish)
- the other odd behavior we saw: a mon that was in quorum was restarted, and had to sync.
- 05:58 PM Bug #5502 (Resolved): mon: long-running sync will restart (cuttlefish)
- on a largish cluster we observe that a ~20 minute mon sync restarts after it finishes. it looks like a problem with ...
- 06:38 PM Revision 20e3abcc (ceph): rgw: make rgw_bucket_dir_entry backward compatible
- Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
- 06:19 PM CephFS Bug #5036: `ls` hangs on random folder
- I suggest trying test branch ceph-client. At least two bugs that can cause hang like this have been fixed.
- 04:08 PM CephFS Bug #5036: `ls` hangs on random folder
- ...
- 02:25 PM CephFS Bug #5036: `ls` hangs on random folder
- Milosz, I think you've run into #2019, which I've reopened. Quan might have seen the same issue but there's not the r...
- 02:15 PM CephFS Bug #5036: `ls` hangs on random folder
- I'm experiencing the same issue here when trying to ls one directory in our ceph cluster on nodes. Using both vanilla...
- 06:13 PM Revision 4fb782c3 (ceph): mds: fix O_TRUNC locking
- When truncating a file, we should xlock the corresponding filelock.
(revoke any Fw caps from clients).
[note from sw... - 06:11 PM Revision cd44a623 (ceph): Makefile: include rbdmap in dist tarball
- Signed-off-by: Sage Weil <sage@inktank.com>
- 04:59 PM Revision 813d96ff (ceph): ceph: don't check leaks on client.* (i.e., radosgw)
- ...until we fix them. This way we can see other valgrind issues.
- 03:44 PM Bug #5460 (Resolved): v0.61.3 -> v0.65 upgrade: new OSDs mark old as down
- commit:e8b42a6998653bde488502097eaa0a2fb834d964
- 01:02 PM Bug #5460: v0.61.3 -> v0.65 upgrade: new OSDs mark old as down
- pushed a patch to paravoid-test branch; can you give it a try? the best theory i have is that at one point an osd st...
- 01:48 AM Bug #5460: v0.61.3 -> v0.65 upgrade: new OSDs mark old as down
- I started osd.0 with 0.65-188-g946a838 and --debug-ms 20 --debug-osd 20. I did have the same errors, although this ti...
- 03:16 PM Bug #5495 (Need More Info): ceph-mon and minus character in hostname
- what versino is this?
can you strace -f ceph-mon and attach that output? that'll give a better hint as to where t... - 03:56 AM Bug #5495 (Can't reproduce): ceph-mon and minus character in hostname
- It looks like ceph-mon does not cope with a - in the hostname:
# /usr/bin/ceph-mon --cluster=office -i test-uplink... - 02:53 PM CephFS Bug #2019: mds: CInode::filelock stuck in sync->mix
- it's a kclient bug, probably already fixed by 'libceph: call r_unsafe_callback when unsafe reply is received'
- 02:31 PM CephFS Bug #2019: mds: CInode::filelock stuck in sync->mix
- I'm into this as part of bug: #5036...
- 02:27 PM CephFS Bug #2019: mds: CInode::filelock stuck in sync->mix
- See #5036.
- 02:45 PM Bug #5500: ceph CLI should validate, reject bad daemon commands
- I should note that the admin socket already supports "get_command_descriptions" so this change is
all in the CLI. - 02:42 PM Bug #5500 (Resolved): ceph CLI should validate, reject bad daemon commands
- The CLI rewrite has the ability to validate commands before sending; it does this with mon and
osd commands, but not... - 02:03 PM Revision 2f9afb86 (ceph): Merge pull request #390 from ksperis/rbdmap.init-2
- init-rbdmap install
Reviewed-by: Sage Weil <sage@inktank.com> - 01:31 PM devops Bug #5499: ceph-deploy --cluster clustername osd prepare fails
- ssh root@node "ceph-disk-prepare --cluster office -- /path/to/mountpoint"
suceeds - 01:30 PM devops Bug #5499 (Resolved): ceph-deploy --cluster clustername osd prepare fails
- ceph-deploy --cluster clustername osd prepare node:/path/to/mountpoint fails with
ceph-disk-prepare -- /path/to/mo... - 12:48 PM CephFS Fix #5498 (New): client: report actual file count instead of object count
- Right now we fill in the statvfs struct's f_files member with the number of files. Instead we should use the mount po...
- 12:16 PM rgw Bug #5374: Avoid relying on keystone's admin token
- A pull request has been made for this issue https://github.com/ceph/ceph/pull/392
Might need testing. - 11:32 AM devops Bug #5369: fedora18: sysvinit doesn't start mon on reboot
- sandon, can you make sure our images or chef or whatever is updated to avoid this pitfall?
- 11:32 AM devops Bug #5369: fedora18: sysvinit doesn't start mon on reboot
- this is a network.service vs NetworkManager problem.. the $network LSB line allegedly doesn't wait for networkmanager...
- 09:45 AM devops Bug #5496: Unable to install librados rpm on Fedora 18
- This looks like it's an issue with fedora moving the libraries into a ceph-libs package. If we don't want to change ...
- 09:40 AM devops Bug #5496 (In Progress): Unable to install librados rpm on Fedora 18
- 04:30 AM devops Bug #5496 (Resolved): Unable to install librados rpm on Fedora 18
- I am unable to install librados library due to conflict in existing package (ceph-libs) already installed on fedora 1...
- 09:38 AM Bug #5492 (Fix Under Review): scripts installing into /usr/usr/sbin (with --prefix=/usr)
- 09:33 AM CephFS Bug #5458: mds: standby-replay -> replay takeover does not handle racing expire/trim
- Saw this backtrace again at /a/teuthology-2013-07-02_01:00:48-fs-next-testing-basic/52429/teuthology.log
- 08:56 AM rgw Bug #1779 (Pending Backport): rgw: swift auth returns wrong error code when unexisting user is given
- Do we want to backport this anywhere?
- 06:35 AM rgw Bug #1779: rgw: swift auth returns wrong error code when unexisting user is given
- This issue should be closed as https://github.com/ceph/ceph/pull/385 was merged.
- 07:30 AM rgw Bug #5416: --help output needs --rgw-zone option
- That's for both radosgw, radosgw-admin. This specific option is already upstream (in cuttlefish) iirc.
- 07:28 AM rgw Bug #5324 (Resolved): radosgw-admin --help missing the --shard-id option
- 12:21 AM rgw Bug #5324: radosgw-admin --help missing the --shard-id option
- This issue should be closed.
Fixed by https://github.com/ceph/ceph/commit/f571bd38f6c6aff861566e14a44570291eb959a2 - 07:09 AM Bug #5482 (Can't reproduce): cephx: verify_reply coudln't decrypt with error: error decoding bloc...
- I actually meant 0.56.6, but you can move to 0.61.x (cuttelfish) too. See ceph.com/docs/master in the upgrade sectio...
- 12:41 AM Bug #5482: cephx: verify_reply coudln't decrypt with error: error decoding block for decryption
- Sage Weil wrote:
> i haven't seen this particular message before. 0.56.3 is a bit out of date, though; please upgra... - 12:40 AM Bug #5482: cephx: verify_reply coudln't decrypt with error: error decoding block for decryption
- ok, i'll try to upgrade.
BTW, if i upgrade from 0.56.3 to 0.61.4, my data stored on OSD would be loss? or how to bac... - 07:06 AM rgw Bug #5228 (Duplicate): radosgw-admin bucket list no longer shows all buckets
- 12:31 AM rgw Bug #5228: radosgw-admin bucket list no longer shows all buckets
- Seems to be a duplicate of http://tracker.ceph.com/issues/5455
- 06:51 AM rgw Bug #5455: radosgw-admin buckets list regression
- As discussed here https://github.com/ceph/ceph/pull/384, this issue should be closed.
- 05:10 AM Revision 352f3625 (ceph): Makefile: fix ceph_sbindir
- From Denis Kaganovich <mahatma@eu.by>:
At least in some cases, scripts (ceph-disk*, ceph-create-keys)
installing int... - 04:29 AM Revision 99411714 (ceph): rgw: keep max_marker on bucket index
- and expose it via the REST interface
Signed-off-by: Yehuda Sadeh <yehuda@inktank.com> - 02:43 AM Revision 9376ce4f (ceph): Merge remote-tracking branch 'gh/next'
- Conflicts:
src/mon/OSDMonitor.cc - 01:59 AM Revision 2eea41a4 (ceph): Merge remote-tracking branch 'origin/wip-rgw-geo-2' into wip-rgw-geo
- 01:26 AM Revision f571bd38 (ceph): Merge pull request #382 from kri5/wip-5324
- rgw: admin: Add --shard-id option to help
Reviewed-by: Yehuda Sadeh <yehuda@inktank.com> - 01:13 AM Revision 24699cce (ceph): Merge branch 'wip-rgw-geo-2' into wip-rgw-geo
- Conflicts:
src/rgw/rgw_admin.cc
src/rgw/rgw_common.cc
src/rgw/rgw_common.h
src/rgw/rgw_rest.cc
src/rgw/rgw_rest_... - 01:13 AM Revision ede7b5a5 (ceph): upgrade/rgw: use bobtail s3tests
- 01:09 AM Revision 86c380bd (ceph): upgrade/rgw: mds restart order doesn't matter
- 01:08 AM Revision 80b0156a (ceph): upgrade/rgw-double: start rgw
- 01:01 AM Revision 0da7da82 (ceph): unitests: fix compilation
- Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
- 12:54 AM Bug #5481: Ceph OSD Connection refused
- Sage Weil wrote:
> if the ceph-osd daemon isn't running, other daemons will get connection refused. look in the log... - 12:10 AM Revision 9942efde (ceph): mon: dead code removal
- Remove code for 'mds cluster_fail', 'osd tell', and auth_usage()
Signed-off-by: Dan Mick <dan.mick@inktank.com>
Revi...
07/02/2013
- 11:55 PM Revision 09e9fef0 (ceph): Merge branch 'wip-ceph-disk'
- Passed ceph-deploy suite.
Reviewed-by: Dan Mick <dan.mick@inktank.com> - 11:30 PM Revision 5c053b10 (ceph): rgw: handle new rest operations
- md + data logs, get shard info
Signed-off-by: Yehuda Sadeh <yehuda@inktank.com> - 11:15 PM Revision fcc423d6 (ceph): rgw: RESTful api for shards info
- For both mdlog, datalog
Signed-off-by: Yehuda Sadeh <yehuda@inktank.com> - 10:38 PM Revision 19828cb8 (ceph): cls_log: store info in header, new op to retrieve it
- Keep max_timestamp, max_marker. Add a new op to retrieve header.
Signed-off-by: Yehuda Sadeh <yehuda@inktank.com> - 10:02 PM Revision 288088d2 (ceph): cls_replica_log: implement decode_json for the types
- Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
- 10:02 PM Revision e10005c5 (ceph): rgw: rework replica log + RESTful api
- Rather than making the caller code handle encode/decode/dump etc,
we introduce an RGWReplicaBounds struct which the R... - 10:02 PM Revision b3f6423e (ceph): rgw: replica log RESTful api uses bucket instance id
- Modified RESTful api for replica log to use bucket instance id
instead of bucket name.
Signed-off-by: Yehuda Sadeh <... - 10:02 PM Revision 7db133dd (ceph): ceph_json: add a utime_t decoder
- This needs to be cleaned up a bit as we don't actually want to
import rgw_common into ceph_json (we'll need to pull t... - 09:43 PM Revision 0143acc4 (ceph): sysvinit, upstart: handle symlinks to dirs in /var/lib/ceph/*
- Match a symlink to a dir, not just dirs. This fixes the osd case of e.g.,
creating an osd in /data/osd$id in which c... - 09:43 PM Revision 87c98e92 (ceph): sysvinit, upstart: handle symlinks to dirs in /var/lib/ceph/*
- Match a symlink to a dir, not just dirs. This fixes the osd case of e.g.,
creating an osd in /data/osd$id in which c... - 09:28 PM rgw Feature #5406 (Fix Under Review): rgw: a RESTful api to dump region map
- 09:00 PM Revision e082f124 (ceph): ceph-disk: make is_held() smarter about full disks
- Handle the case where the device is a full disk. Make the partition
check a bit more robust (don't make assumptions ... - 09:00 PM Revision cb97338b (ceph): ceph-disk: handle /dev/foo/bar devices throughout
- Assume the last component is the unique device name, even if it appears
under a subdir of /dev.
Signed-off-by: Sage ... - 08:43 PM Revision a4805efc (ceph): osdc/Objecter: resend command map version checks on reconnect
- We already do this for Ops and LingerOps, but missed this when we added
CommandOps to the mix. The result is that an... - 08:06 PM Revision 653bc2de (ceph): common: move parse_time into utime_t header
- Signed-off-by: Greg Farnum <greg@inktank.com>
- 07:53 PM devops Bug #5345: ceph-disk: handle less common device names
- Hi Jing,
As far as I can tell the current ceph-disk supports these device names, but as I mentioned I don't have a... - 07:26 PM devops Bug #5345: ceph-disk: handle less common device names
- I had similar problem as Thomas but mine are HP Blades (more specifically BL460 G1) with P200i controllers but I susp...
- 07:39 PM Revision fc7d6227 (ceph): RESTful implementation to dump regionmap implementation
- Signed-off-by: Babu Shanmugam <anbu@enovance.com>
- 07:39 PM Revision b6c83665 (ceph): key_value_store: fixup errno handling
- Don't use "errno" as a variable name, and include <errno.h>
instead of some of the specific paths.
Signed-off-by: Gr... - 07:39 PM Revision 370d1752 (ceph): rgw: remove get_bound_info from RGWReplicaLogger
- It doesn't have any users and we're about to expose a wrapper class instead.
Signed-off-by: Greg Farnum <greg@inktan... - 07:38 PM Revision 641bd2fa (ceph): RESTful API implementation for replica_log
- Signed-off-by: Babu Shanmugam <anbu@enovance.com>
- 06:32 PM Revision 08cf4c0d (ceph): Merge pull request #386 from ceph/wip-osd-xattr
- automatically enable xattrs in omap; make size limits well defined
Reviewed-by: Samuel Just <sam.just@inktank.com> - 06:17 PM Revision 43f60c30 (ceph): rbd: remove kernel rbd test from rbd suite
- It's already present in the kernel suite.
- 06:03 PM Revision da1fb93f (ceph): mds: man page: Fixed cut & paste error
- 05:53 PM Revision 32673bd0 (ceph): Merge pull request #388 from dachary/master
- set object_info_t pool of an ObjectContext if it is undefined or bad
Reviewed-by: Samuel Just <sam.just@inktank.com> - 05:40 PM Revision a4cd6317 (ceph): set object_info_t pool of an ObjectContext if it is undefined or bad
- When reading object_info_t from an existing object attribute, the pool
may be < 0 and should be set to the pool conta... - 05:08 PM Revision 7e878bcc (ceph): rgw: add RGWFormatter_Plain allocation to sidestep cranky strlen()
- Valgrind complains about an invalid read when we don't pad the allocation,
and because it is inlined we can't whiteli... - 04:57 PM CephFS Bug #5453: kclient: multiple_rsync tee output partially zeroed
- btw the mds change passed 11 iterations before it stopped because of a chef/network hiccup.
- 09:42 AM CephFS Bug #5453: kclient: multiple_rsync tee output partially zeroed
- in combination with the mds s/wrlock/xlock/ change?
- 02:53 PM Revision 0efcccdc (ceph): Move rbdmap file to /etc/ceph
- Signed-off-by: Laurent Barbe <laurent@ksperis.com>
- 02:50 PM Revision ffe70458 (ceph): install rules for init-rbdmap
- Signed-off-by: Laurent Barbe <laurent@ksperis.com>
- 02:43 PM Bug #5490 (Resolved): osd won't start after ceph-deploy on Debian 7 - fix
- commit:87c98e92d1375c8bc76196bbbf06f677bef95e64
- 02:07 PM Bug #5490 (Fix Under Review): osd won't start after ceph-deploy on Debian 7 - fix
- wip-5490
fixed the same thing in the upstart scripts. nice catch, thanks! - 02:31 PM Bug #5493 (Resolved): objecter: hang on osd_command to nonexistent target
- 11:43 AM Bug #5493 (Resolved): objecter: hang on osd_command to nonexistent target
- job was teuthology-2013-07-02_01:00:14-rados-next-testing-basic 52253
- 01:27 PM Revision 6ff2b16b (ceph): rgw-admin: Adds option --shard-id to help.t
- Signed-off-by: Christophe Courtaut <christophe.courtaut@gmail.com>
- 01:24 PM Revision 3a20b069 (ceph): rgw: admin: Add --shard-id option to help
- http://tracker.ceph.com/issues/5324 fixes #5324
Explains operations dependency,
which is more important than the sema... - 10:48 AM Bug #5460 (Need More Info): v0.61.3 -> v0.65 upgrade: new OSDs mark old as down
- Can you set 'debug ms = 20' and 'debug osd = 20', reproduce the problem, and attach teh log on the marked-down osd an...
- 09:31 AM devops Bug #5369: fedora18: sysvinit doesn't start mon on reboot
- (06:25:53 PM) mbiebl: systemctl enable NetworkManager-wait-online.service
here's the full exhcnage:... - 09:30 AM CephFS Support #5491 (Rejected): Use of RBD kernel module
- Questions like this should be addressed to the ceph-users list, please. :)
http://ceph.com/resources/mailing-list-irc/ - 04:02 AM CephFS Support #5491 (Rejected): Use of RBD kernel module
- I would be grateful if you could explain a couple of items as these are not immediately clear to me from the Ceph doc...
- 08:53 AM Bug #5492 (Resolved): scripts installing into /usr/usr/sbin (with --prefix=/usr)
- At least in some cases, scripts (ceph-disk*, ceph-create-keys) installing into /usr/usr/sbin. This happened while "ce...
- 06:58 AM rgw Feature #3074: radosgw needs --help support
- The -d option of radosgw seems to come from global_init function, which is located under src/global/global_init.cc, a...
- 05:53 AM rgw Bug #5402: rgw compilation problem on wip-rgw-geo-2 branch
- Fixed by https://github.com/ceph/ceph/commit/cfc1f2ee1f4345a88d7d72f33f6f9b838fa134ce
- 04:00 AM Revision 08123b6a (ceph): Merge pull request #387 from ceph/wip-5346
- rgw: add RGWFormatter_Plain allocation to sidestep cranky strlen()
Reviewed-by: Yehuda Sadeh <yehuda@inktank.com> - 01:21 AM Revision 49ff63b1 (ceph): rgw: add RGWFormatter_Plain allocation to sidestep cranky strlen()
- Valgrind complains about an invalid read when we don't pad the allocation,
and because it is inlined we can't whiteli... - 12:06 AM Revision 94742d4c (ceph): doc: Minor fix.
- Signed-off-by: John Wilkins <john.wilkins@inktank.com>
07/01/2013
- 11:23 PM Revision e80eb486 (ceph): kernel: specify fs (and use raw disks) for rbd-nomount tests
- Otherwise simple 1tb test fills up the root disks. :(
- 11:19 PM Bug #5460: v0.61.3 -> v0.65 upgrade: new OSDs mark old as down
- I rechecked and confirm that all the old ones run 0.61.3. I've stopped the new 0.65+ ones since then, but you can see...
- 05:46 PM Bug #5460: v0.61.3 -> v0.65 upgrade: new OSDs mark old as down
- I updated a cluster from 0.61 to the same 0.65 SHA1 version.
$ ceph --version
ceph version 0.65-188-g946a838 (946... - 11:09 PM Revision 7120f22f (ceph): upgrade: ignore mds restart order on rbd, rados tests
- Not relevant to rbd; put these in the fs upgrade matrix.
- 11:05 PM Bug #5490 (Resolved): osd won't start after ceph-deploy on Debian 7 - fix
- I created a cluster with ceph-deploy on a set of Debian 7 VMs for testing. Each VM has a physical device/partition at...
- 10:55 PM Revision 2ad4ff69 (ceph): rgw: metadata, data logs can be turned on / off
- Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
- 10:20 PM Revision 935c2784 (ceph): mon: Paxos: update first_committed on first paxos proposal
- We were adding this update to a transaction that would only be applied
on the leader. The peons would never see a fi... - 10:19 PM devops Bug #5489: ceph-deploy: mon destroy throws inappropriate message
- test setup: burnupi05, burnupi21, burnupi63
burnupi21 is the one, stuck up at ceph-create-keys. - 10:19 PM devops Bug #5489 (Resolved): ceph-deploy: mon destroy throws inappropriate message
- I tried to build 3 node cluster using ceph-deploy on centos , but forgot to disable iptables service on one of the no...
- 09:44 PM CephFS Bug #5453: kclient: multiple_rsync tee output partially zeroed
- patch "ceph: fix pending vmtruncate race" should fix the issue.
- 09:30 PM Revision 847465ff (ceph): librados: fix test warning on 32-bit platforms
- Signed-off-by: Sage Weil <sage@inktank.com>
- 09:23 PM rgw Bug #5346 (Resolved): rgw: invalid read from RGWFormatter_Plain::write_data
- Sage pushed a fix at commit:49ff63b1750789070a8c6fef830c9526ae0f6d9f
- 06:22 PM rgw Bug #5346 (Fix Under Review): rgw: invalid read from RGWFormatter_Plain::write_data
- 03:39 PM rgw Bug #5346: rgw: invalid read from RGWFormatter_Plain::write_data
- using a trivial implemention of strlen avoids this. unfortunately we can't whitelist the glibc strlen call because i...
- 09:21 PM Revision 93ae1523 (ceph): radosgw-admin: add missing quote
- 09:21 PM Revision c0bf24d7 (ceph): radosgw-admin: test 'bucket list' command (all buckets)
- Verifies fix for #5455
Signed-off-by: Sage Weil <sage@inktank.com> - 09:08 PM Revision e7fa5fca (ceph): rgw: move radosgw-admin test into rgw suite
- 08:45 PM Revision 1b578a80 (ceph): librados: add test for large and many xattrs
- Verify that we can set large and large numbers of attrs on an object.
Signed-off-by: Sage Weil <sage@inktank.com> - 08:45 PM Revision 6d90dad4 (ceph): os/FileStore: automatically enable 'filestore xattr use omap' as needed
- Automatically enable the 'filestore xattr use omap' option if the fs
does not appear to handle large xattrs on its ow... - 08:42 PM Revision 8694d296 (ceph): osd/PGLog: populate log_keys_debug from read_old_log()
- Fixes: #5470
Signed-off-by: Sage Weil <sage@inktank.com>
Reviewed-by: Samuel Just <sam.just@inktank.com> - 06:36 PM rbd Bug #5468 (Resolved): qemu: reports snapshot not supported on rpms
- Josh built some rpm packages by adding a qemu patch and we tested it. Works fine now.
The packages are at http://w... - 06:01 PM Revision cfc1f2ee (ceph): rgw: modify RGWBucketInfo::creation_time encoding / decoding
- Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
- 05:27 PM Revision ca614028 (ceph): mds: warn on unconnected snap realms
- When there are more than one active MDS, restarting MDS triggers
assertion "reconnected_snaprealms.empty()" quite oft... - 05:00 PM rgw Feature #5408 (Fix Under Review): rgw: turn off dr/geo logging
- 04:49 PM rbd Bug #5488: librbd: deadlock in image refresh
- looks like bh_write is holding cache_lock (== ObjectCacher::lock) and trying ot take snap_read, while ictx_refresh() ...
- 04:19 PM rbd Bug #5488 (Resolved): librbd: deadlock in image refresh
- ...
- 04:06 PM Subtask #5487 (In Progress): Factor out ObjectContext / ReplicatedPG::object_contexts
- 12:52 PM Subtask #5487 (Closed): Factor out ObjectContext / ReplicatedPG::object_contexts
- * "read/write locks unit tests":https://github.com/ceph/ceph/pull/407
- 04:05 PM rbd Bug #5464: krbd: modifying mapped image also modifies snapshot
- backport patch updated with mainline commit sha1 and sent to stable@, just waiting for it to land in the stable tree.
- 03:22 PM Bug #5484 (Resolved): mon/Paxos.cc: 51: FAILED assert(err == 0) (reapply_all_versions())
- commit:935c27842e4549b32c1c8721e178e9f3a10f2ad4
- 10:58 AM Bug #5484: mon/Paxos.cc: 51: FAILED assert(err == 0) (reapply_all_versions())
- ...
- 10:11 AM Bug #5484 (Resolved): mon/Paxos.cc: 51: FAILED assert(err == 0) (reapply_all_versions())
- ubuntu@teuthology:/a/teuthology-2013-07-01_01:00:09-rados-master-testing-basic/51450$ cat summary.yaml
description:... - 02:56 PM Subtask #5433: Factor out the ReplicatedPG object replication and client IO logic as a PGBackend ...
- Sam's comments:
* It is better to do the changes within the same file to better read the diffs
* The changes are in... - 02:44 PM rgw Bug #5402 (Resolved): rgw compilation problem on wip-rgw-geo-2 branch
- 02:29 PM rgw Bug #5455 (Resolved): radosgw-admin buckets list regression
- teuthology.git commit:c0bf24d770162fd2babfb806aacda4c78b23288d
- 02:08 PM Revision 1fbd50b1 (ceph): Merge pull request #385 from kri5/wip-1779
- rgw: Fix return value for swift user not found
Reviewed-by: Yehuda Sadeh <yehuda@inktank.com> - 01:42 PM Bug #5470 (Resolved): osd/PGLog.h: 199: FAILED assert(log.log.size() == log_keys_debug.size())
- 10:01 AM Bug #5470: osd/PGLog.h: 199: FAILED assert(log.log.size() == log_keys_debug.size())
- Looks right to me.
- 12:57 PM Revision 4089001d (ceph): rgw: Fix return value for swift user not found
- http://tracker.ceph.com/issues/1779 fixes #1779
Adjust the return value from rgw_get_user_info_by_swift call
in RGW_... - 12:36 PM CephFS Feature #5486 (Resolved): kclient: make it work with selinux
- see #5477 for the latest failed attempt
- 12:34 PM CephFS Bug #5477 (Resolved): Unable to create files on CephFS on Fedora 18 using kernel module
- 12:19 PM CephFS Bug #5477: Unable to create files on CephFS on Fedora 18 using kernel module
- Many thanks for the responses Sage and Greg. You were right - once I disabled SElinux this worked.
Chris - 10:51 AM CephFS Bug #5485 (Can't reproduce): failed cifs mount
- In teuthology, logs at /a/teuthology-2013-07-01_01:00:46-fs-master-testing-basic/51619...
- 09:50 AM Bug #5154: osd/SnapMapper.cc: 270: FAILED assert(check(oid))
- ubuntu@teuthology:/a/teuthology-2013-07-01_01:00:09-rados-master-testing-basic/51502/remote$ less ceph-osd.4.log
... - 09:47 AM Bug #5482 (Need More Info): cephx: verify_reply coudln't decrypt with error: error decoding block...
- i haven't seen this particular message before. 0.56.3 is a bit out of date, though; please upgrade to 0.56.6 to get ...
- 09:45 AM Bug #5481 (Rejected): Ceph OSD Connection refused
- ceph-users is a better forum for these sorts of questions. thanks!
- 09:43 AM Bug #5481: Ceph OSD Connection refused
- if the ceph-osd daemon isn't running, other daemons will get connection refused. look in the log for the down osd to...
- 06:04 AM rgw Bug #1779: rgw: swift auth returns wrong error code when unexisting user is given
- Here comes a fix for this issue.
https://github.com/ceph/ceph/pull/385 - 05:16 AM devops Bug #5483: ceph-deploy --cluster othername mon create does not work
- It looks like othe ceph-deploy commands also fail with a cluster name different than "ceph".
- 04:51 AM devops Bug #5483 (Resolved): ceph-deploy --cluster othername mon create does not work
- ceph-deploy --cluster othername mon create nodename
does not work. It creates a directory /var/lib/ceph/mon/ceph-n... - 03:25 AM rgw Bug #5416: --help output needs --rgw-zone option
- Joe Buck wrote:
> For the ./radosgw-admin command that is.
On which branch the --rgw-zone exist? Is this radosgw-...
06/30/2013
- 12:40 PM Bug #5401: cuttlefish osd recovery slow
- The following seems to help:
- osd recovery max active = 2
- filestore min sync interval = 1
- filestore max sync ... - 04:00 AM Revision c73a24be (ceph): rgw: end-marker serves as last value, not as upperbound
- Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
- 03:59 AM Revision c4140303 (ceph): rgw: relax marker requirements for log apis
- Don't require start-time, start-marker.
Signed-off-by: Yehuda Sadeh <yehuda@inktank.com> - 03:39 AM Revision 064e92f0 (ceph): Makefile.am: some more makefile rules fixes
- Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
- 03:39 AM Revision ce7d816d (ceph): rgw: cls_log_entry has id field
- The id field can be used as the marker for log list / trim
operations.
Signed-off-by: Yehuda Sadeh <yehuda@inktank.com> - 03:39 AM Revision b21a41ab (ceph): rgw: update log APIs to use markers
- a few APIs were missing markers, fixed that.
Signed-off-by: Yehuda Sadeh <yehuda@inktank.com> - 03:38 AM Revision c2a56e52 (ceph): rgw: date string parsing also parses seconds fraction
- Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
- 12:41 AM Revision a6bbc91c (ceph): Makefile.am: some more makefile rules fixes
- Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
06/29/2013
- 11:13 PM Revision 26423b92 (ceph): rgw: date string parsing also parses seconds fraction
- Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
- 10:28 PM Revision 0335905d (ceph): rgw: fix cli test
- Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
- 10:28 PM Revision 5314e3fe (ceph): Makefile.am: add missing rule
- Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
- 07:46 PM Feature #5437: ceph-mon performance on ARM
- Downgraded to Normal, due to landing of patch. Keeping open as a placeholder for any future issues identified for imp...
- 05:35 PM CephFS Bug #5453: kclient: multiple_rsync tee output partially zeroed
- please check if the attached patch solves this issue
- 02:09 PM Bug #5401: cuttlefish osd recovery slow
- Just another point regarding my idea. If we have an osd having writes to avg 50 PGs/min then ceph now wants that this...
- 07:02 AM Bug #5460: v0.61.3 -> v0.65 upgrade: new OSDs mark old as down
- I've just tried ...
- 06:56 AM Bug #5482 (Can't reproduce): cephx: verify_reply coudln't decrypt with error: error decoding bloc...
- Hi all,
i found an OSD down, my env is ceph 0.56.3 , the part of log is below:
==== 47+0+0 (2545873131 0 0) 0x6... - 06:45 AM Bug #5481 (Rejected): Ceph OSD Connection refused
- Hi all,
i found an error in ceph osd log file.in my osd tree, there're 2 osd marked down, not running.
in osd... - 01:29 AM Revision 09e374f1 (ceph): osd: set maximum object attr size
- Make a well-defined maximum size of an object attribute. Since Linus has
a 64KB limit, and that is what we normally ...
06/28/2013
- 10:43 PM Revision e11f2588 (ceph): mon/PGMonitor: use post_paxos_update, not init, to refresh from osdmap
- We do two things here:
- make init an one-time unconditional init method, which is what the
health service expect... - 10:42 PM Revision 4d07fb01 (ceph): mon/PaxosService: add post_paxos_update() hook
- Some services need to update internal state based on other service's
state, and thus need to be run after everyone ha... - 10:03 PM Bug #5470 (Fix Under Review): osd/PGLog.h: 199: FAILED assert(log.log.size() == log_keys_debug.si...
- 09:47 PM Revision e56ef07e (ceph): rgw: initialize sent_header in RGWOp_Opstate_List
- Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
- 09:25 PM Revision e235958f (ceph): mds: log before respawning when standby-replay falls behind
- Call into an MDS method so that we can write to the log.
Signed-off-by: Sage Weil <sage@inktank.com> - 09:24 PM Revision 946a838c (ceph): Merge remote-tracking branch 'gh/wip-5381' into wip-mds
- Reviewed-by: Greg Farnum <greg@inktank.com>
- 08:17 PM Revision e75057fc (ceph): rgw-admin: restructure replicalog commands
- Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
- 07:21 PM Revision 87217e1e (ceph): client: send all request put's through put_request()
- Make sure all MetaRequest reference put's go through the same path that
releases inode references, including all of t... - 06:50 PM Revision 9af3b86b (ceph): client: fix remaining Inode::put() caller, and make method psuedo-private
- Not sure I can make this actually private and make Client::put_inode() a
friend method (making all of Client a friend... - 06:31 PM CephFS Bug #5453: kclient: multiple_rsync tee output partially zeroed
- i hit it after just a couple iterations of the teuthology test. i'll capture the osd log...
- 06:08 PM CephFS Bug #5453: kclient: multiple_rsync tee output partially zeroed
- I can't reproduce this locally. how difficult to reproduce this? what's the backend fs for osd?
- 06:26 PM Revision 12ab4f88 (ceph): Merge remote-tracking branch 'enovance/wip-rgw-geo-2-opstate' into wip-...
- Conflicts:
src/Makefile.am
src/rgw/rgw_admin.cc
src/rgw/rgw_common.h
Signed-off-by: Yehuda Sadeh <yehuda@inktank.... - 05:44 PM Revision 0742f919 (ceph): kernel: add instal, ceph tasks to rbd-nomount collection
- 05:32 PM CephFS Bug #5411: teuthology: bad object dereference
- IME that's what this kind of error from gevent/eventlet etc. means - once the thread exits in a certain abnormal way,...
- 03:28 PM CephFS Bug #5411: teuthology: bad object dereference
- Yeah, I am/somebody will need to spend some time digging into this when we have some time free. There's another issue...
- 03:24 PM CephFS Bug #5411: teuthology: bad object dereference
- I think this is just a symtom of the mds_thrasher crashing, but not logging the exception since this join happens bef...
- 03:36 PM rbd Bug #5480 (Can't reproduce): libceph: unexpected old state in con_sock_state_change
- This happened after running this job in a loop for a while:...
- 03:19 PM devops Bug #5479 (Resolved): Append our built packages with some sort of inktank/ceph identifier
- Seems like our packages should have something appended to them so we know they are the inktank/ceph ones. I ran into ...
- 02:58 PM rgw Feature #5406 (In Progress): rgw: a RESTful api to dump region map
- 02:58 PM rgw Feature #5408 (In Progress): rgw: turn off dr/geo logging
- 02:57 PM rgw Feature #4336 (Fix Under Review): rgw: dr: sync processing state: implement internal RESTful API
- 02:25 PM CephFS Bug #5381 (Pending Backport): ceph-fuse: stuck with disconnected inodes on shutdown
- commit:946a838cffa0927d1237489e8c2c143e87d66892
- 01:54 PM Revision 2e857a4d (ceph): librados: fix cmd OSDCommand test
- If we get ENXIO, buflen will be 0.
Signed-off-by: Sage Weil <sage@inktank.com> - 01:48 PM Bug #5401: cuttlefish osd recovery slow
- mhm retestet. Log looks different but the result is still the same. Strangely after processing around 100pgs the reco...
- 01:34 PM Bug #5401: cuttlefish osd recovery slow
- oops, it's 'osd recover clone overlap = false' (no y)
- 12:12 PM Bug #5401: cuttlefish osd recovery slow
- Oh i forgot to say that in last log i activated debug_filestore later on at 20:51:02. it was 0 before.
- 11:58 AM Bug #5401: cuttlefish osd recovery slow
- Log still contains:
2013-06-28 20:51:29.928755 7fe5c33ba700 20 filestore(/ceph/osd.3/) _do_clone_range 3260416~8192 ... - 11:52 AM Bug #5401: cuttlefish osd recovery slow
- new files here - still a lot of reading goes on /home/cephdrop/5401/overlap_false_ceph-osd.3.log.gz
- 11:44 AM Bug #5401: cuttlefish osd recovery slow
- all of the read activity is from using the clone when recovering snapped objects. can you see how things perform wit...
- 11:38 AM Bug #5401: cuttlefish osd recovery slow
- Just to note the log was done with (filestore_queue_max_ops = 50, filestore_queue_committing_max_ops = 50) I h...
- 11:28 AM Bug #5401: cuttlefish osd recovery slow
- OK Log uploaded to:
/home/cephdrop/5401/debug_filestore_20_ceph-osd.3.log.gz
The high read rates started at 20:22... - 07:29 AM Bug #5401: cuttlefish osd recovery slow
- Stefan Priebe wrote:
> I'm not really sure that this is problem cause it was working fine under bobtail. But i can't... - 07:14 AM Bug #5401: cuttlefish osd recovery slow
- I'm not really sure that this is problem cause it was working fine under bobtail. But i can't guarantee that the load...
- 07:02 AM Bug #5401: cuttlefish osd recovery slow
- It is a good idea. I have to think a bit more about how hard it would be to implement. Currently the pg_temp remapp...
- 06:47 AM Bug #5401: cuttlefish osd recovery slow
- thanks sage - i already played with the threading but that's not the problem. The problem is 2x 10-15MB/s reading whi...
- 06:22 AM Bug #5401: cuttlefish osd recovery slow
- Stefan Priebe wrote:
> While recovering i'm seeing two threads out of the OSD doing in parallel 10-15MB/s reading an... - 12:20 AM Bug #5401: cuttlefish osd recovery slow
- Generally i think i would be the best to optionally block/redirect client I/O on a recovering OSD. So it can recover ...
- 01:36 PM rgw Bug #5478 (Resolved): rgw: fix init script for rgw daemon name
- after installing and configuring radosgw on centos6, found that the naming convention is bit confusing for rgw daemon...
- 10:54 AM devops Bug #5405: ceph-deploy: transient pushy exception on install
- on centos 6.4,...
- 10:51 AM Linux kernel client Bug #5429: libceph: rcu stall, null deref in osd_reset->__reset_osd->__remove_osd
- plana72 still sitting in kdb.
- 10:51 AM Linux kernel client Bug #5429: libceph: rcu stall, null deref in osd_reset->__reset_osd->__remove_osd
- hit this again, ubuntu@teuthology:/a/teuthology-2013-06-28_01:01:07-kernel-master-testing-basic/48683
- 09:31 AM CephFS Bug #5250: ceph-mds 0.61.2 aborts on start
- Wow, that is a much simpler test case than I would expect to be required. I can reproduce with a single file and this...
- 02:24 AM CephFS Bug #5250: ceph-mds 0.61.2 aborts on start
This is all in the lab at present.
We have been doing some additional testing, and have now confirmed that this...- 09:23 AM CephFS Bug #5477: Unable to create files on CephFS on Fedora 18 using kernel module
- And you don't need any kernel support to run the Ceph daemons. You should also check the permissions — it's possible ...
- 06:51 AM CephFS Bug #5477: Unable to create files on CephFS on Fedora 18 using kernel module
- I suspect this is SElinux or something similar getting in the way...
- 04:47 AM CephFS Bug #5477 (Resolved): Unable to create files on CephFS on Fedora 18 using kernel module
- I have mounted a CephFS filesystem on a Fedora 18 system, which succeeds as follows:
[root@e8c4-dl360g7-03 ceph]# ... - 06:21 AM Revision 72ac281d (ceph): rgw: don't remove bucket index when removing a bucket
- Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
- 06:12 AM Revision 10a2b60f (ceph): rgw: bilog list by bucket instance (RESTful api)
- Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
- 06:08 AM Bug #5239: osd: Segmentation fault in ceph-osd / tcmalloc
- Yes, boost1.46 from Ubuntu does seem to make a difference. The last built has been running for 5 days now with far fe...
- 04:39 AM Revision 81bee648 (ceph): client: use put_inode on MetaRequest inode refs
- When we drop the request inode refs, we need to use put_inode() to ensure
they get cleaned up properly (removed from ... - 04:13 AM Revision 674bbabe (ceph): rgw: user param in access key is not manadatory
- for json decoding
Signed-off-by: Yehuda Sadeh <yehuda@inktank.com> - 04:13 AM Revision 977df778 (ceph): ceph_json: reset values
- Call the default constructor if entry not found. Also,
reset containers before inserting new data.
Signed-off-by: Ye... - 04:13 AM Revision 9d702e58 (ceph): rgw: advance ptr when sending data
- Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
- 03:49 AM Revision 4f2a194a (ceph): big: timebox test to two hours
- 03:47 AM Revision 25d7d965 (ceph): big: thrash, and whitelist wrongly marked me down
06/27/2013
- 11:31 PM Revision 13f9c184 (ceph): doc: Created an install page for Calxeda development packages.
- Signed-off-by: John Wilkins <john.wilkins@inktank.com>
- 11:30 PM Bug #5401: cuttlefish osd recovery slow
- While recovering i'm seeing two threads out of the OSD doing in parallel 10-15MB/s reading and 10-20Mb/s writing. The...
- 07:01 AM Bug #5401: cuttlefish osd recovery slow
- mikedawson told me on irc that he has the same problem - since upgrading from bobtail to cuttlefish
- 10:23 PM Revision 8f0fbc22 (ceph): Merge branch 'next'
- 10:21 PM Revision 90f5c448 (ceph): ceph-disk: s/else if/elif/
- Signed-off-by: Greg Farnum <greg@inktank.com>
Reviewed-by: Joao Luis <joao.luis@inktank.com>
(cherry picked from comm... - 10:21 PM Revision 9e604ee6 (ceph): ceph-disk: s/else if/elif/
- Signed-off-by: Greg Farnum <greg@inktank.com>
Reviewed-by: Joao Luis <joao.luis@inktank.com>
(cherry picked from comm... - 10:09 PM Revision d8300606 (ceph): Merge pull request #372 from ceph/wip-mon-pgmap
- Reviewed-by: Joao Eduardo Luis <joao.luis@inktank.com>
- 09:39 PM CephFS Bug #5381 (Fix Under Review): ceph-fuse: stuck with disconnected inodes on shutdown
- 09:22 AM CephFS Bug #5381: ceph-fuse: stuck with disconnected inodes on shutdown
- 09:08 PM Revision c22b941e (ceph): Update keys if they have changed before locking
- Signed-off-by: Sandon Van Ness <sandon@inktank.com>
- 06:05 PM Bug #4599: ceph auth import -i <file> is broken
- As of now ./ceph auth import -i appears to work, but reading from stdin doesn't:
dzafman@ubuntu:~/ceph/src$ ./ceph... - 05:03 PM Feature #4839 (Resolved): api: make new CLI send old version of commands to old monitors during u...
- 05:03 PM Feature #4315 (Resolved): api: create python CLI wrapper for ceph tool; read command descriptions...
- 05:03 PM Feature #4314 (Resolved): api: modify ceph tool to describe own commands
- 03:45 PM Feature #4200 (Resolved): mon: break pgmap into separate leveldb keys
- 03:25 PM devops Bug #5476 (Resolved): ceph-deploy: disk list command is broken on centos 6.4
- This error was actually in ceph-disk, not in ceph-deploy. Now fixed.
cuttlefish commit:90f5c448abeb127ae5a5528a79b... - 02:59 PM devops Bug #5476: ceph-deploy: disk list command is broken on centos 6.4
- test setup: burnupi63
- 02:59 PM devops Bug #5476 (Resolved): ceph-deploy: disk list command is broken on centos 6.4
- on the latest cuttlefish branch, disk list command is broken,
[ubuntu@burnupi63 ceph-deploy]$ sudo ceph -v
ceph v... - 02:42 PM Feature #5437: ceph-mon performance on ARM
- Latest wip-mon-pgmap tests got through PG creation without issue. The mons have been quite stable on ARM with 168 OS...
- 02:40 PM Revision b4ee7cac (ceph): changes to accomodate review comments on previous version
- Signed-off-by: Babu Shanmugam <anbu@enovance.com>
- 02:21 PM Bug #5424 (Resolved): mon/Paxos.cc: 549: FAILED assert(begin->last_committed == last_committed)
- 01:38 PM devops Bug #5475 (Closed): ceph-deploy deb install breaks on i386 due to missing python-pushy package
- We're not building python-pushy for i386, so attempts to apt-get install ceph-deploy fail:
> rturk@ubuntu:~$ sudo ... - 01:34 PM Bug #5473 (Resolved): osd/ReplicatedPG.cc: 1379: FAILED assert(0) in trim_object() on master, cut...
- ...
- 10:08 AM rgw Bug #5415: rgw: failing valgrind leak checks
- 41708: (401s) collection:verify clusters:fixed-2.yaml fs:btrfs.yaml msgr-failures:few.yaml tasks:rgw_s3tests.yaml val...
- 01:50 AM rgw Bug #5415: rgw: failing valgrind leak checks
- Sage Weil wrote:
> teuthology-2013-06-21_01:00:44-rgw-master-testing-basic/41708 and 41709
Hi, I'm willing to wor... - 08:03 AM Bug #5471: mon: do not join a quorum if quorum's version is lower than ours
- I have a simple patch for this that simply compares the quorum's version to our own paxos version and forces us to su...
- 07:58 AM Bug #5471 (Resolved): mon: do not join a quorum if quorum's version is lower than ours
- With p being the monitor's Paxos version, consider:
* A - p:100 (at time quorum was formed)
* B - p:100 (at time ... - 07:59 AM rgw Feature #4336 (In Progress): rgw: dr: sync processing state: implement internal RESTful API
- 07:57 AM rgw Feature #5358 (Fix Under Review): rgw: RESTful api for intra-region copy state
- 07:57 AM rgw Feature #4613 (Fix Under Review): Allow bucket data to reside in a separate pool to object data
- 07:45 AM rgw Bug #5324 (In Progress): radosgw-admin --help missing the --shard-id option
- 02:42 AM rgw Bug #5324: radosgw-admin --help missing the --shard-id option
- Hi,
Here comes a fix for this bug.
https://github.com/ceph/ceph/pull/382
Best - 07:38 AM rgw Feature #5420 (Rejected): rgw: integrate bucket metadata changes with bucket index log
- We're not going to do that, decided on a different approach.
- 07:37 AM rgw Feature #5417 (Fix Under Review): rgw: separate bucket metadata object into pointer object and in...
- 07:36 AM rgw Bug #5344 (Fix Under Review): rgw: make list of bucket placement pools index configurable
- 06:05 AM Bug #5470 (Resolved): osd/PGLog.h: 199: FAILED assert(log.log.size() == log_keys_debug.size())
- ...
- 05:39 AM Revision 8f1da8f3 (ceph): rgw: fix regionmap json decoding
- Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
- 05:28 AM Revision 20434dee (ceph): upgrade: move fs upgrade tests to a separate suite
- For now.. until we figure out why they keep failing!
- 05:19 AM Revision 68c013d7 (ceph): Merge remote-tracking branch 'gh/next'
- 05:15 AM Revision 61a04360 (ceph): Merge pull request #378 from ceph/wip-init-rbd
- Reviewed-by: Sage Weil <sage@inktank.com>
- 05:11 AM Revision cd7510f2 (ceph): qa/workunits/misc/multiple_rsync: put tee output in /tmp
- 2013-06-25T10:29:15.811 INFO:teuthology.task.workunit.client.0.err:+ rsync -auv --exclude local/ /usr/ usr.2
2013-06-... - 04:31 AM Revision 5c4bb463 (ceph): rgw: fix radosgw-admin buckets list
- Fixes: #5455
Backport: cuttlefish
This commit fixes a regression, where radosgw-admin buckets list
operation wasn't r... - 04:30 AM Revision e1f9fe58 (ceph): rgw: fix radosgw-admin buckets list
- Fixes: #5455
Backport: cuttlefish
This commit fixes a regression, where radosgw-admin buckets list
operation wasn't r... - 04:06 AM Revision fe663317 (ceph): Handle non-existent front interface in maps from older MONs
- Fix OSDService::get_con_osd_hb() to not try to get_connection() without front interface
Fix OSD::handle_osd_map() to ... - 03:41 AM Revision 867ead91 (ceph): qa/workunits/rbd/simple_1tb: add simple rbd read/write test on large image
- Motivated by #5454.
Signed-off-by: Sage Weil <sage@inktank.com> - 02:30 AM Revision 033f9ea6 (ceph): kernel: add rbd test to write a 1TB image
- Motivated by #5454
- 02:29 AM Revision 2bdbccf9 (ceph): marginal: restructure the multimds collection
- - two clients types (ceph-fuse and kclient)
- two clients in each test
- separate out tasks - 01:42 AM Revision b2fb4876 (ceph): ceph-disk: use unix lock instead of lockfile class
- The lockfile class relies on file system trickery to get safe mutual
exclusion. However, the unix syscalls do this f... - 01:36 AM Revision 26e7a6ff (ceph): ceph-disk: do not mount over an osd directly in /var/lib/ceph/osd/$clus...
- If we see a 'ready' file in the target OSD dir, do not mount our device
on top of it.
Among other things, this preve... - 01:28 AM Revision 8a17f33b (ceph): ceph-disk: do not mount over an osd directly in /var/lib/ceph/osd/$clus...
- If we see a 'ready' file in the target OSD dir, do not mount our device
on top of it.
Among other things, this preve... - 12:48 AM Revision 140e63b4 (ceph): ceph: disable logrotate
- This screwed up the log archival step at the end, and generally makes a
mess of automated runs.
Fixe: #5451 - 12:34 AM Revision 986185ca (ceph): mon/PGMonitor: avoid duplicating map_pg_create() effort on same maps
- If we have an election and refresh, but the osdmap does not change, there
is no need to recalculate the pg create map... - 12:09 AM Revision ca55c341 (ceph): cephtool/test.sh: add case for auth add with no caps
- Test case for failure in #5467. Supplying new auth info overwrites.
Signed-off-by: Dan Mick <dan.mick@inktank.com>
...
06/26/2013
- 11:30 PM Bug #5444 (Rejected): ceph df print error message
- 11:18 PM Revision bfed2d60 (ceph): MonCommands.h: auth add doesn't require caps (it can use -i <file>)
- This was a regression from the old behavior introduced by the
CLI rewrite.
Fixes: #5467
Signed-off-by: Dan Mick <dan... - 11:15 PM CephFS Bug #5381: ceph-fuse: stuck with disconnected inodes on shutdown
- this is sufficient to reproduce. i think this is a problem with unlinked inodes in the client cache not getting clea...
- 10:35 PM Bug #5460: v0.61.3 -> v0.65 upgrade: new OSDs mark old as down
- Faidon Liambotis wrote:
> I did have mons upgraded and restarted first, sorry for not mentioning this earlier. I'll ... - 09:21 PM Bug #5460: v0.61.3 -> v0.65 upgrade: new OSDs mark old as down
- I did have mons upgraded and restarted first, sorry for not mentioning this earlier. I'll try the workaround neverthe...
- 09:07 PM Bug #5460 (Resolved): v0.61.3 -> v0.65 upgrade: new OSDs mark old as down
- commit:fe6633172ea10b9f95c6d59bccdac01651195f25
- 08:08 PM Bug #5460: v0.61.3 -> v0.65 upgrade: new OSDs mark old as down
- Note the workaround here is to upgrade and restart Mons first.
- 07:10 PM Bug #5460 (Fix Under Review): v0.61.3 -> v0.65 upgrade: new OSDs mark old as down
- Proposed fix in wip-5460.
In older release the osdmap did not specify a front-side interface. A couple of places ... - 02:02 AM Bug #5460 (Resolved): v0.61.3 -> v0.65 upgrade: new OSDs mark old as down
- I tried upgrading my v0.61.3 cluster to v0.65 today. All of the new (v0.65) OSDs are marking all of the old (v0.61.3)...
- 10:08 PM CephFS Bug #5453: kclient: multiple_rsync tee output partially zeroed
- putting the tee'd file in /tmp fixes the problem, implying this is a kclient/cephfs bug of some sort. moving this in...
- 09:35 PM rgw Bug #5455: radosgw-admin buckets list regression
- committed, but still need the radosgw-admin.py test before we close the bug
- 07:39 PM Revision d1d90284 (ceph): Merge branch 'next'
- 07:28 PM Revision 71f3e56d (ceph): Makefile.am: fix libglobal.la race with ceph_test_cors
- ceph_test_cors had libglobal.la in its _LDFLAGS macro definition;
it should have been in _LDADD. Moreover, things us... - 06:55 PM rbd Bug #5469 (Resolved): qemu-io: segfault when tried IO with invalid arguments
- tested this on rhel 6.3 and rhel6.4
tried qemu-io to perform IO on the rbd image by writing in a pattern and readi... - 06:42 PM devops Bug #5390 (Resolved): ceph-deploy osd create hangs
- 06:40 PM Bug #5424: mon/Paxos.cc: 549: FAILED assert(begin->last_committed == last_committed)
- 06:39 PM rbd Bug #5464: krbd: modifying mapped image also modifies snapshot
- 03:24 PM rbd Bug #5464: krbd: modifying mapped image also modifies snapshot
- This was caused by bf0d5f503dc11d6314c0503591d258d60ee9c944, which was first included in 3.9.
- 03:20 PM rbd Bug #5464 (Fix Under Review): krbd: modifying mapped image also modifies snapshot
- The workunit that tests this case (rbd/kernel.sh) succeeds on both 3.9 and 3.10-rc7 with the patches in wip-snapc-3.9...
- 11:08 AM rbd Bug #5464 (Resolved): krbd: modifying mapped image also modifies snapshot
- http://permalink.gmane.org/gmane.comp.file-systems.ceph.devel/15815
- 06:36 PM Bug #5445 (Resolved): random osd EPERM on journal
- 06:28 PM Bug #5445 (Pending Backport): random osd EPERM on journal
- 07:52 AM Bug #5445 (Fix Under Review): random osd EPERM on journal
- pushed wip-5445.
this normally wouldn't happen, except that teuthology does not define fsid in the ceph.conf, so c... - 05:13 PM Bug #5467: ceph auth add <entity> -i <file> broken
- Added simple test at Sage's suggestion in commit:ca55c3416e02398991c9789b1590d721fbca212d
- 04:21 PM Bug #5467 (Resolved): ceph auth add <entity> -i <file> broken
- commit:bfed2d60a59bf69863989ca3bc108079c1d37f4f
- 03:48 PM Bug #5467 (Fix Under Review): ceph auth add <entity> -i <file> broken
- 03:17 PM Bug #5467 (In Progress): ceph auth add <entity> -i <file> broken
- 03:16 PM Bug #5467 (Resolved): ceph auth add <entity> -i <file> broken
- In the Ceph CLI rewrite, I broke "auth add" without caps. auth add currently requires
both an entity name and at le... - 03:30 PM rbd Bug #5468 (Resolved): qemu: reports snapshot not supported on rpms
- Tested qemu on the following packages "http://ceph.com/packages/qemu-kvm/redhat/x86_64/qemu-img-0.12.1.2-2.355.el6.2....
- 03:03 PM Revision 1747a387 (ceph): kernel: reorg kernel rbd tests; add a few missing scripts
- Move all the simple workunits into a separate collection, and test with
and without msgr failure injection.
There we... - 02:38 PM Bug #5466 (Can't reproduce): raring tarball builds mysteriously fail make check
- I've noticed this intermittently: make check on raring tarball builds fails like this:...
- 01:55 PM Revision d2f38118 (ceph): mon/Paxos: update first_committed only from paxos
- Do not touch the in-memory first_committed until the trim commits. This
avoids any possible confusion due to races a... - 01:55 PM Revision ac63b2e0 (ceph): mon/Paxos: clean up removal of pre-conversion paxos states
- Use a helper, independent of trim machinery, and call on leader, too.
Signed-off-by: Sage Weil <sage@inktank.com> - 01:55 PM Revision b8d04a2a (ceph): mon/PaxosService: rename scrub
- Make the name patch the one in Paxos.
Signed-off-by: Sage Weil <sage@inktank.com> - 01:55 PM Revision 516445be (ceph): mon/Paxos: simplify trim()
- Collapse all the trim methods into a single simple method.
Signed-off-by: Sage Weil <sage@inktank.com> - 01:55 PM Revision ea1f316e (ceph): mon: do not reopen MonitorDBStore during startup
- level doesn't seem to like this when it races with an internal compaction
attempt (see below). Instead, let the stor... - 01:55 PM Revision 13168698 (ceph): mon/PaxosService: add post_paxos_update() hook
- Some services need to update internal state based on other service's
state, and thus need to be run after everyone ha... - 01:55 PM Revision e635c478 (ceph): mon/PGMonitor: use post_paxos_update, not init, to refresh from osdmap
- We do two things here:
- make init an one-time unconditional init method, which is what the
health service expect... - 01:55 PM Revision 290ccde1 (ceph): mon/Paxos: set first_committed on first commit
- Signed-off-by: Sage Weil <sage@inktank.com>
- 01:27 PM Revision 5511daf3 (ceph): doc: public network statement needed on new monitors.
- When using ceph-deploy to create a new monitor on a host that is not
in the initial set of hosts defined by the ceph-... - 11:32 AM Bug #5401: cuttlefish osd recovery slow
- Hi sage very strange. I tried your suggestions but also setting recovery active to 20 and priority to 60 does not hel...
- 11:32 AM rbd Feature #5465 (Resolved): openstack: cinder: support resize with rbd
- Cinder now has an api for growing volumes. Add support for it to the rbd driver
- 10:09 AM Bug #5272 (Duplicate): Updating ceph from 0.61.2 to 0.61.3 obviously changes tunables of existing...
- This was issues with the MDS doing heavy reads off of the OSDs. See #4405 and the related caching issues.
- 06:28 AM Bug #5195 (Resolved): "ceph-deploy mon create" fails when adding additional monitors
- Updated documentation to add a note about needing the public network statement.
- 05:55 AM Revision af00f733 (ceph): rgw: automatic pool creation for placement pools
- With the new pools configuration, now we auto create the
pools when needed (through bucket creation). Also, make
sure... - 04:25 AM Revision 2d2aa00e (ceph): mon/PGMonitor: store PGMap directly in store, bypassing PaxosService st...
- Instead of encoding incrementals and periodically dumping the whole encoded
PGMap, instead store everything in a rang... - 04:25 AM Revision d31ed950 (ceph): mon/PaxosService: allow paxos service writes while paxos is updating
- In commit f985de28f86675e974ac7842a49922a35fe24c6c I mistakenly made
is_writeable() false while paxos was updating du... - 04:25 AM Revision 11e03253 (ceph): mon/Paxos: drop unnecessary last_committed loads
- Drop (apparently) ad-hoc refreshes of last_committed from the store.
These are unnecessary and confusing.
Signed-off... - 04:25 AM Revision ad9c2948 (ceph): mon/Paxos: assert that the store gives us back what we just wrote
- In bug #5424 I observed leveldb failing internally and then returning
bad info. We then hit a random/confusing asser... - 04:25 AM Revision e93730b7 (ceph): mon: enable leveldb cache by default
- 512 MB sounds reasonable to me.
Signed-off-by: Sage Weil <sage@inktank.com> - 04:25 AM Revision fe365339 (ceph): mon/Paxos: never write first_committed except during trim
- The trimming is handled by proposing transactions. Do not confuse matters
by writing (incorrect) first_committed val... - 03:54 AM Feature #5461 (Rejected): ceph-disk-prepare should support LVM volumes
- It is very convenient at least for testing purposes to configure osds on LVM volumes. Unfortunately ceph-disk-prepare...
- 03:06 AM Revision 7a2566c6 (ceph): rgw: remove test placement info
- Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
- 02:48 AM Revision ccb3dd5a (ceph): mds: fix underwater dentry cleanup
- If the underwater dentry is a remove link, we shouldn't mark the
inode clean
Signed-off-by: Yan, Zheng <zheng.z.yan@... - 02:18 AM Revision 224130c9 (ceph): rgw (test): remove some warnings
- Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
- 01:47 AM rbd Bug #5454: krbd: assertion failure in rbd_img_obj_callback()
- Hi Sage,
yes, dd was the program, bs=1024k
Regards
Andreas - 12:59 AM Revision 1b162ce6 (ceph): rgw: initialize user system flag
- Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
06/25/2013
- 11:05 PM Revision 00bc3a84 (ceph): rbd/rgw upgrade suites to test upgrade from bobtail to cuttlefish to next
- Signed-off-by: tamil <tamil.muthamizhan@inktank.com>
- 10:30 PM Revision 7681c58e (ceph): rgw: log in the same shard for bucket entry point and instance
- We'd like to have bucket entry point and instance info at the same
log shard, so that we can process them in order.
... - 10:21 PM Bug #5445: random osd EPERM on journal
- oh, the test in question isn't mounting a drive, but is storing the data directly in /var/lib/ceph/osd/ceph-$id. the...
- 09:35 PM Bug #5445 (In Progress): random osd EPERM on journal
- this happens on tasks that don't use all available disks. a previous job with ceph-deploy leaves behind osd disks, s...
- 09:59 PM Revision d4e39a76 (ceph): rgw: unlink/link don't always update entry point
- Some operations already update the entry point, so no
need to do it again.
Signed-off-by: Yehuda Sadeh <yehuda@inkta... - 09:14 PM Revision 5680fa1e (ceph): doc/release-notes: v0.65
- Signed-off-by: Sage Weil <sage@inktank.com>
- 09:00 PM Revision 6673b2d3 (ceph): rgw: tie metadata put to bucket link/unlink
- and lot's of constifying
Signed-off-by: Yehuda Sadeh <yehuda@inktank.com> - 09:00 PM Revision 5c3df085 (ceph): cls_rgw: cleanup
- Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
- 08:45 PM Revision 70be76b2 (ceph): Merge branch 'next'
- 08:40 PM Revision 3020c5ea (ceph): mon/Elector: cancel election timer if we bootstrap
- If we short-circuit and bootstrap, cancel our timer. Otherwise it will
go off some time later when we are in who kno... - 08:40 PM Revision 305f0c50 (ceph): mon: cancel probe timeout on reset
- If we are probing and get (say) an election timeout that calls reset(),
cancel the timer. Otherwise, we assert later... - 07:56 PM Revision 82db84be (ceph): rgw: some more internal api cleanups
- Use of rgw_bucket when referring to the bucket instance,
use bucket name when referring to the bucket entry point.
Al... - 07:49 PM Revision ee03afaa (ceph): dump_stuck: fix test
- The mon-osd-report-timeout setting shouldn't be there! We will set the
other item explicitly, and remove both from t... - 07:48 PM CephFS Bug #5450 (Resolved): mds: failed CDir::_fetched() assert
- nice! cherry-picked to commit:ccb3dd5ad5533ca4e9b656b4e3df31025a5f2017
- 07:08 PM CephFS Bug #5450: mds: failed CDir::_fetched() assert
- 0.61.4-5-gd572cf6 ? probably already fixed by commit:81d073fecb (mds: fix underwater dentry cleanup)
- 10:20 AM CephFS Bug #5450 (Resolved): mds: failed CDir::_fetched() assert
- ...
- 07:48 PM Revision d5acff28 (ceph): rados: dump-stick: remove configs from yaml; let task set them
- The mon osd report timeout was wrong; see #5440.
- 07:33 PM Bug #5459: ceph-mon failure using wip-mon-pgmap on ARM
- this is almost certainly the max_open_files limit. add max open files = 16384 to ceph.conf and restart the mons.
- 07:28 PM Bug #5459 (Resolved): ceph-mon failure using wip-mon-pgmap on ARM
- This happened on the mixed burnupi/calxeda cluster with wip-mon-pginfo. leveldb caching was set to 256MB....
- 07:19 PM CephFS Bug #5418: kceph: crash in remove_session_caps
- Zheng Yan wrote:
> I still don't figure out that root cause of the crash, infinite loop in iterate_session_caps(), B... - 07:01 PM CephFS Bug #5418: kceph: crash in remove_session_caps
- I still don't figure out the cause of the crash, infinite loop in iterate_session_caps(), BUG_ON(session->s_nr_caps >...
- 01:00 PM CephFS Bug #5418: kceph: crash in remove_session_caps
- ubuntu@teuthology:/a/teuthology-2013-06-25_01:00:47-kernel-next-testing-basic/45603
- 07:19 PM Revision c4be5a70 (ceph): rgw: unlink bucket from user on metadata rm bucket:< bucket>
- Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
- 06:09 PM Revision 86c73c94 (ceph): rgw: fixes to object versioning tracking
- There are a few different cases for setting the object version.
Either we need to create a new version, or we need to... - 06:05 PM Revision 8bd31d42 (ceph): rgw: filter read xattrs
- We're only interested in object xattrs that have specific rgw.user
prefix.
Signed-off-by: Yehuda Sadeh <yehuda@inkta... - 06:03 PM Revision 422bb6d0 (ceph): rgw: add str_startswith()
- useful util
Signed-off-by: Yehuda Sadeh <yehuda@inktank.com> - 06:01 PM CephFS Bug #5458 (Duplicate): mds: standby-replay -> replay takeover does not handle racing expire/trim
- not sure this is the right diagnosis since i only looked at this briefly, but:...
- 05:57 PM Revision 0e1612b3 (ceph): Merge pull request #380 from dachary/wip-4907
- get_xattr() can return more than 4KB
Reviewed-by: Josh Durgin <josh.durgin@inktank.com> - 05:36 PM Revision 8db289f2 (ceph): cls_ver: rename version xattr, add some more logging
- Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
- 05:30 PM Bug #5336: osd crash triggered by 'rbd rm ...'
- no word from Florian on the ML; downgrading to high.
- 05:28 PM Bug #5401: cuttlefish osd recovery slow
- hmm, i don't see anything awry here. the rops count does go over the max because we get client io requests to missing...
- 03:15 PM Bug #5401: cuttlefish osd recovery slow
- OK this is a complete log from starting a stopped OSD, recover and then shut it down again. This one was done using t...
- 02:28 PM Bug #5401: cuttlefish osd recovery slow
- debug osd = 20
debug ms = 1
thanks! - 01:57 PM Bug #5401: cuttlefish osd recovery slow
- Will test soon. Can also provide a complete log. Just debug osd 20? everything else can be 0?
- 01:35 PM Bug #5401: cuttlefish osd recovery slow
- pushed another wip-5401 and resolves the race more thoroughly. tested lightly and it behaves. but i haven't been ab...
- 01:11 PM Bug #5401: cuttlefish osd recovery slow
- Stefan, can you capture an osd log that covers the entire period from before recovery starts up until when it slows/s...
- 11:43 AM Bug #5401: cuttlefish osd recovery slow
- thanks sage i tested wip-5401 branch no change ;-(
- 11:21 AM Bug #5401: cuttlefish osd recovery slow
- Yes see comment #7 (log file uploaded to cephdrop folder 5401/ceph-osd.44.log.gz).
Yes peering is already fully co... - 11:19 AM Bug #5401: cuttlefish osd recovery slow
- i've pushed wip-5021 which makes a small optimization that will improve recovery for small objects. it's not well te...
- 11:18 AM Bug #5401: cuttlefish osd recovery slow
- Did you put the logs in cephdrop?
I'm mostly curious to see if we can figure out where the disk thrashing is coming ... - 11:14 AM Bug #5401: cuttlefish osd recovery slow
- Maybe what's missing. The backtraces i provided and the logs are form that timeframe. Also it's the timeframe i'm see...
- 09:59 AM Bug #5401: cuttlefish osd recovery slow
- Stefan Priebe wrote:
> First the slow requests and VM hangs are results of a heavily stressed disk. I see a MASSIVE ... - 12:03 AM Bug #5401: cuttlefish osd recovery slow
- Yes generally this check needs to be fixed. But it does not solve the problem.
To be a bit more specific.
First... - 05:15 PM Revision 12678a10 (ceph): Merge pull request #379 from dachary/wip-5312
- skip TEST(EXT4StoreTest, _detect_fs) if DISK or MOUNTPOINT are undefined
Reviewed-by: Sage Weil <sage@inktank.com> - 05:07 PM Bug #5239: osd: Segmentation fault in ceph-osd / tcmalloc
- latest is that the precise version of boost appears to have resolved this.. so far.
- 04:57 PM Revision 29e14baf (ceph): mon/PaxosService: s/get_version()/get_last_committed()/
- Avoid aliasing simple accessors; use a single name instead. Also, function
name overloading will throw a wrench in t... - 04:57 PM Revision 0fd776da (ceph): mon/PGMap: make int type explicit
- We get away with this because int is 32-bits on x86_64 and i386 both, but
we should be explicit anyway!
Signed-off-b... - 04:57 PM Revision 0d73eb4d (ceph): mon/PGMonitor: drop some dead code
- Signed-off-by: Sage Weil <sage@inktank.com>
- 04:57 PM Revision 950c0f35 (ceph): mon/PaxosService: move upgrade_format() machinery into PaxosService
- We originally did this in AuthMonitor, but it is perfect for PGMonitor too,
so make it generic.
Signed-off-by: Sage ... - 04:57 PM Revision c8f79369 (ceph): mon/AuthMonitor: start at format 1 (latest) for new clusters
- Signed-off-by: Sage Weil <sage@inktank.com>
- 04:47 PM rgw Bug #5422 (Resolved): object deletion should log the object tag
- fixed in the wip-rgw-geo-2 branch.
- 04:46 PM rgw Bug #5441 (Resolved): bucket creation does not log the tag
- this should be fixed now, reopen if you see it again.
- 04:43 PM rbd Bug #5454: krbd: assertion failure in rbd_img_obj_callback()
- Hi Andreas,
How are you writing to the image? dd with bs=1M or something? - 12:15 PM rbd Bug #5454 (Resolved): krbd: assertion failure in rbd_img_obj_callback()
- when working on a large rbd image, I am hitting an
assertion failure in rbd kernel module; the
assertion is in rbd_... - 04:19 PM Revision c2d517ef (ceph): v0.65
- 04:16 PM devops Bug #5345: ceph-disk: handle less common device names
- can you please try the version of ceph-disk in the wip-ceph-disk branch? it has a bunch of changes to be smarter abo...
- 03:24 PM devops Bug #5345: ceph-disk: handle less common device names
- I have several HP dl180's with p400 raid controllers. This is all standard hardware.
The disk paths are enumerated... - 02:15 PM rgw Bug #5455 (Resolved): radosgw-admin buckets list regression
- With bobtail, it was possible to list every rgw buckets on the cluster with:
# radosgw-admin buckets list
[
"a... - 02:10 PM Revision 3016f46f (ceph): get_xattr() can return more than 4KB
- Instead of failing if the attribute to be returned is larger than 4KB,
double the buffer size each time librados.rado... - 01:59 PM Bug #5438 (Resolved): mon/Monitor.cc: 1888: FAILED assert(is_probing() || is_synchronizing())
- 01:09 PM Revision 6e320a1b (ceph): skip TEST(EXT4StoreTest, _detect_fs) if DISK or MOUNTPOINT are undefined
- The TEST(EXT4StoreTest, _detect_fs) test is meant to be run from
qa/workunits/filestore/filestore.sh, after the ext4 ... - 12:50 PM Bug #5440 (Resolved): osd: marked down due to no pgstats reports
- broken test + test yaml, fixed in teuthology.git and ceph-qa-suite.git
- 08:55 AM Bug #5440: osd: marked down due to no pgstats reports
- ubuntu@teuthology:/a/teuthology-2013-06-25_01:00:06-rados-next-testing-basic/45417
in mon log, osd msgs suddenly s... - 12:13 PM CephFS Bug #5453 (In Progress): kclient: multiple_rsync tee output partially zeroed
- 12:09 PM CephFS Bug #5453 (Resolved): kclient: multiple_rsync tee output partially zeroed
- latest run:...
- 11:55 AM devops Bug #5452 (Resolved): ceph-deploy: purge command reports unsupported platform on fedora18
- commit f67486632cb40a34d089c9dcfa371383a5f3ab9c
fix worked fine! - 11:15 AM devops Bug #5452 (Resolved): ceph-deploy: purge command reports unsupported platform on fedora18
- while purge command does uninstall on rhel and centos, it reports, "unsupported platform" for fedora18.
We need to... - 10:59 AM Bug #4907 (Resolved): rados python bindings: get_xattr() uses a fixed 4k buffer
- commit:3016f46f53d4701ead1e30f2a3d67a39ca0050f8
- 07:15 AM Bug #4907 (Fix Under Review): rados python bindings: get_xattr() uses a fixed 4k buffer
- proposed patch : https://github.com/ceph/ceph/pull/380
- 10:49 AM Feature #4982 (Fix Under Review): OSD: namespaces pt 1 (librados/osd, not caps)
- 10:15 AM Bug #5312 (Resolved): Skip EXT4StoreTest._detect_fs test if DISK or MOUNTPOINT environment variab...
- 06:14 AM Bug #5312 (Fix Under Review): Skip EXT4StoreTest._detect_fs test if DISK or MOUNTPOINT environmen...
- ready for rewiew : https://github.com/ceph/ceph/pull/379
- 05:52 AM Bug #5312: Skip EXT4StoreTest._detect_fs test if DISK or MOUNTPOINT environment variables not set
- TEST (EXT4StoreTest, _detect_fs) was introduced by https://github.com/ceph/ceph/commit/574051f8da0a30073a7d5da880878e...
- 09:31 AM Bug #5449 (Can't reproduce): osd crash immediately after booting up
- Debian 7.1 (wheezy), kernel 3.2.41-2+deb7u2, ceph 0.61.4. I have three osds and only osd.1 crashes just after booting...
- 09:27 AM Bug #5444: ceph df print error message
- this is the fix:
$ git tag --contains 0c2b738d8d07994fee4c73dd076ac9364a64bdb2
v0.64
maybe your ceph-mon daemo... - 12:15 AM Bug #5444: ceph df print error message
- ceph version 0.64-639-g37a2017 (37a20174fd22a79938ba9c93046e8830f4a3306f)
- 06:43 AM Revision 63e81afe (ceph): rgw: multiple fixes related to metadata, bucket creation
- Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
- 04:59 AM Revision a8f601d5 (ceph): ceph-disk: make list_partition behave with unusual device names
- When you get device names like sdaa you do not want to mistakenly conclude that
sdaa is a partition of sda. Use /sys... - 04:48 AM Revision a4ddf704 (ceph): Add rc script for rbd map/unmap
- Init script for mapping/unmapping rbd device on startup and shutdown.
On start, map rbd dev according to /etc/rbdmap,... - 04:23 AM Revision 586c68f5 (ceph): ceph-fuse: create finisher threads after fork()
- The ObjectCacher and MonClient classes both instantiate Finisher
threads. We need to make sure they are created *aft... - 04:07 AM Revision 7c9dee01 (ceph): mon/PaxosService: drop unused {get,set,put}_version(prefix, a, bl)
- Signed-off-by: Sage Weil <sage@inktank.com>
- 04:07 AM Revision 6060268f (ceph): mon/PaxosService: some whitespace
- Signed-off-by: Sage Weil <sage@inktank.com>
- 04:07 AM Revision b28bd787 (ceph): mon/PaxosService: drop unused last_accepted_name
- Signed-off-by: Sage Weil <sage@inktank.com>
- 04:07 AM Revision 6429cb1d (ceph): mon/PaxosService: remove unused exists_version() variant
- Signed-off-by: Sage Weil <sage@inktank.com>
- 04:07 AM Revision 6721122f (ceph): mon/MonmapMonitor: avoid exists_version() helper
- We are the only user; open-code it.
Signed-off-by: Sage Weil <sage@inktank.com> - 04:07 AM Revision 78b7ec7d (ceph): mon/PaxosService: drop unused helpers
- Signed-off-by: Sage Weil <sage@inktank.com>
- 04:07 AM Revision 78fefe65 (ceph): mon: make PaxosService::get_value() int return type 64-bit
- Backport: cuttlefish
Signed-off-by: Sage Weil <sage@inktank.com> - 04:07 AM Revision c47f271d (ceph): mon: fix mkfs monmap cleanup
- exists_key(a,b) was looking for "monmap/mkfs/monmap".
Signed-off-by: Sage Weil <sage@inktank.com> - 04:07 AM Revision 83c49be3 (ceph): mon/PaxosService: remove mkfs helpers
- Keep it simple. These are one-liners.
Signed-off-by: Sage Weil <sage@inktank.com> - 04:07 AM Revision 872f4d5f (ceph): mon/PaxosService: simplify full helpers, drop single-use helper
- We are the only caller for get_version(prefix, name), so move it inline
and drop it. Also rename full_version_name t... - 04:07 AM Revision 1d913d20 (ceph): mon/OSDMOnitor: use provided get_version_full()
- Signed-off-by: Sage Weil <sage@inktank.com>
- 04:00 AM Revision cedb6063 (ceph): ceph-deploy: no need for ceph-fuse for non fs workunits
- 04:00 AM Revision 2df1e209 (ceph): ceph-deploy: install correct branch for rados api tests
- Fixes: #5442
Signed-off-by: Sage Weil <sage@inktank.com> - 03:41 AM Revision 37a20174 (ceph): Merge remote-tracking branch 'gh/next'
- 01:51 AM Revision 9ae0ec83 (ceph): mon/Elector: cancel election timer if we bootstrap
- If we short-circuit and bootstrap, cancel our timer. Otherwise it will
go off some time later when we are in who kno... - 01:49 AM Revision 1c890f5c (ceph): client: fix warning
- client/Client.cc: In member function 'virtual void Client::ms_handle_remote_reset(Connection*)':
warning: client/Clie... - 01:12 AM Revision 03d3be3e (ceph): mon: cancel probe timeout on reset
- If we are probing and get (say) an election timeout that calls reset(),
cancel the timer. Otherwise, we assert later... - 01:05 AM rbd Feature #1790: rbd: have a way of establishing configured mappings at boot time
- For the record : "Add rc script for rbd map/unmap":https://github.com/ceph/ceph/commit/a4ddf704868832e119d7949e96fe35...
- 12:59 AM Revision c3b97591 (ceph): mon/AuthMonitor: ensure initial rotating keys get encoded when create_i...
- The create_initial() method may get called multiple times; make sure it
will unconditionally generate new/initial rot... - 12:59 AM Revision 0cc826c3 (ceph): init-radosgw.sysv: remove -x debug mode
- Fixes: #5443
Signed-off-by: Sage Weil <sage@inktank.com>
(cherry picked from commit 31d6062076fdbcd2691c07a23b381b26a... - 12:58 AM Revision 521fdc2a (ceph): mon/AuthMonitor: ensure initial rotating keys get encoded when create_i...
- The create_initial() method may get called multiple times; make sure it
will unconditionally generate new/initial rot... - 12:44 AM Revision 3791a1e5 (ceph): osd: tolerate racing threads starting recovery ops
- We sample the (max - active) recovery ops to know how many to start, but
do not hold the lock over the full duration,... - 12:42 AM Revision 31d60620 (ceph): init-radosgw.sysv: remove -x debug mode
- Fixes: #5443
Signed-off-by: Sage Weil <sage@inktank.com> - 12:27 AM Revision 7e41c103 (ceph): rgw: bucket entry point contains owner, other fix
- We now keep the owner at the bucket entry point. Also, reading
bucket info at RGWCreateBucket::execute(), because pre...
06/24/2013
- 11:18 PM Revision a6b65fe8 (ceph): Merge pull request #15 from ceph/wip-ulimits
- Reviewed-by: Warren Usui <warren.usui@inktank.com>
- 11:18 PM Revision 3da94551 (ceph): Merge pull request #15 from ceph/wip-ulimits
- Reviewed-by: Warren Usui <warren.usui@inktank.com>
- 11:16 PM Revision 02b0b4a9 (ceph): mon/PaxosService: not active during paxos UPDATING_PREVIOUS
- Treat this as an extension of the recovery process, e.g.
RECOVERING -> ACTIVE
or
RECOVERING -> UPDATING_PREVIOUS -... - 11:16 PM Revision e1ac7c6c (ceph): mon/PaxosService: allow paxos service writes while paxos is updating
- In commit f985de28f86675e974ac7842a49922a35fe24c6c I mistakenly made
is_writeable() false while paxos was updating du... - 11:16 PM Revision 4d57c12f (ceph): common/pick_addresses: behave even after internal_safe_to_start_threads
- ceph-mon recently started using Preforker to working around forking issues.
As a result, internal_safe_to_start_threa... - 11:16 PM Revision cf75478d (ceph): mon: safely signal bootstrap from MonmapMonitor::update_from_paxos()
- Signed-off-by: Sage Weil <sage@inktank.com>
(cherry picked from commit dc83430124a5fd37573202a4cc0986c3c03739ef) - 11:16 PM Revision fd1769cb (ceph): mon/Paxos: go active *after* refreshing
- The update_from_paxos() methods occasionally like to trigger new activity.
As long as they check is_readable() and is... - 11:16 PM Revision 57c89291 (ceph): mon: simplify Monitor::init_paxos()
- Signed-off-by: Sage Weil <sage@inktank.com>
(cherry picked from commit e832e76a4af04b091c806ad412bcfd0326d75a2d) - 11:16 PM Revision 35745cba (ceph): mon/PaxosService: simplify readable check
- Recast this in terms of the paxos check and our additional conditions,
which match wait_for_readable().
Signed-off-b... - 11:16 PM Revision 9d7c40e3 (ceph): mon/PaxosService: simplify is_writeable
- Recast this in terms of paxos check + our conditions, and make it
match wait_for_writeable().
Signed-off-by: Sage We... - 11:16 PM Revision 10d41200 (ceph): mon/Paxos: cleanup: drop unused PREPARING state bit
- This is never set when we block, and nobody looks at it.
Signed-off-by: Sage Weil <sage@inktank.com>
(cherry picked ... - 11:16 PM Revision c43b1f4d (ceph): mon/Paxos: not readable when LOCKED
- If we are re-proposing a previously accepted value from a previous quorum,
we should not consider it readable, becaus... - 11:16 PM Revision c6d5dc4d (ceph): mon: simplify states
- - make states mutually exclusive (an enum)
- rename locked -> updating_previous
- set state prior to begin() to simpl... - 11:16 PM Revision 34acc5a3 (ceph): mon: explicitly refresh_from_paxos() when leveldb state changes
- Instead of opportunistically calling each service's update_from_paxos(),
instead explicitly refresh all in-memory sta... - 11:16 PM Revision 1d866250 (ceph): mon: remove unnecessary update_from_paxos calls
- The refresh() will do this when the state changes; no need to
opportunistically call this method all of the time.
Si... - 11:16 PM Revision 265212a7 (ceph): mon: no need to refresh from _active
- The refresh is done explicitly by the monitor, independent of the more
fragile PaxosService callbacks.
Signed-off-by... - 11:16 PM Revision 054e96d9 (ceph): mon/PaxosService: cache {first,last}_committed
- Refresh the in-memory values when we are told the on-disk paxos state
may have changed.
Signed-off-by: Joao Eduardo ... - 11:16 PM Revision 6ac58cd9 (ceph): mon/Paxos: do paxos refresh in finish_proposal; and refactor
- Do the paxos refresh inside finish_proposal, ordered *after* the leader
assertion so that MonmapMonitor::update_from_... - 11:15 PM Revision eb86eebe (ceph): common/pick_addresses: behave even after internal_safe_to_start_threads
- ceph-mon recently started using Preforker to working around forking issues.
As a result, internal_safe_to_start_threa... - 09:50 PM Revision cad8cf58 (ceph): Add python-argparse to dependencies (for pre-2.7 systems)
- Signed-off-by: Dan Mick <dan.mick@inktank.com>
- 09:01 PM Bug #5442 (Resolved): ceph-deploy: rados test failed in the nightlies
- qa suite commit:2df1e209d3649a46aaf01c6cd961cb96638fe1f6
- 11:50 AM Bug #5442 (Resolved): ceph-deploy: rados test failed in the nightlies
- logs: ubuntu@teuthology:/a/teuthology-2013-06-24_01:01:13-ceph-deploy-master-testing-basic/44159...
- 08:47 PM Revision 9a9c941d (ceph): Merge pull request #376 from dalgaaf/wip-da-SCA-cppcheck-3
- Reviewed-by: Sage Weil <sage@inktank.com>
- 08:42 PM Bug #5431 (Resolved): osd: dump_stuck test fails with ENXIO
- 08:35 PM Bug #5444 (Need More Info): ceph df print error message
- what version is this? the units were off by a factor of 1000. this was recently fixed..
- 07:16 PM Bug #5444 (Rejected): ceph df print error message
- root@ubuntu:/media# ceph df
GLOBAL:
SIZE AVAIL RAW USED %RAW USED
3903M 2267M 1367M ... - 08:04 PM Revision 046e3b71 (ceph): debian, rpm: remove python-lockfile dependency
- As for 2a4953b697a3464862fd3913336edfd7eede2487 ceph-disk no longer uses
this.
Signed-off-by: Sage Weil <sage@inktan... - 07:54 PM Revision 4474a0cc (ceph): mon/AuthMonitor: make initial auth include rotating keys
- This closes a very narrow race during mon creation where there are no
service keys.
Fixes: #5427
Signed-off-by: Sage... - 07:31 PM Bug #5445 (Can't reproduce): random osd EPERM on journal
- ubuntu@teuthology:/a/teuthology-2013-06-24_18:48:34-rados-cuttlefish-testing-basic/45071...
- 07:25 PM Revision 8cc36895 (ceph): Merge remote-tracking branch 'gh/next'
- 06:58 PM rgw Bug #5441: bucket creation does not log the tag
- previous tag: a tag for the read_version
- 06:57 PM rgw Bug #5441: bucket creation does not log the tag
- Not sure what you mean:
That's the tag:... - 11:46 AM rgw Bug #5441 (Resolved): bucket creation does not log the tag
- In the wip-rgw-geo-2 branch, the mdlog entries for a bucket creation do not include the tag.
The tag should be incl... - 06:52 PM Bug #5438 (Pending Backport): mon/Monitor.cc: 1888: FAILED assert(is_probing() || is_synchronizin...
- 11:39 AM Bug #5438 (Resolved): mon/Monitor.cc: 1888: FAILED assert(is_probing() || is_synchronizing())
- ...
- 06:01 PM Revision 31422521 (ceph): Merge pull request #16 from ceph/wip-5431
- Reviewed-by: Warren Usui <warren.usui@inktank.com>
- 05:58 PM Revision f046dab8 (ceph): mds: do not assume segment list is non-empty in standby_trim_segments
- If we restart standby replay shortly after startup, before we actually have
any segments, we an trigger a segfault he... - 05:54 PM CephFS Bug #5381 (Need More Info): ceph-fuse: stuck with disconnected inodes on shutdown
- 01:03 PM CephFS Bug #5381: ceph-fuse: stuck with disconnected inodes on shutdown
- next time we see this (or any other ceph-fuse hsutdown hang), grab teh logs manually via scp before nuking, and note ...
- 05:49 PM rgw Bug #5443 (Resolved): el6: rgw init script has debug mode on
- commit:31d6062076fdbcd2691c07a23b381b26abc59f65
- 05:17 PM rgw Bug #5443 (Resolved): el6: rgw init script has debug mode on
- rgw init script that comes with the rpm packages for centos has debug mode on.
bash -x is set. It would be nice to... - 05:26 PM rgw Bug #5323: trim data log lists dates as optional, enforced as required in the current code
- I wonder if they should be optional. It's easy to purge the entire log by mistake.
- 05:20 PM Revision 7ef921c8 (ceph): Merge pull request #374 from ceph/wip-5427
- Reviewed-by: Joao Eduardo Luis <joao.luis@inktank.com>
- 04:39 PM Bug #5401 (In Progress): cuttlefish osd recovery slow
- that fix looks correct, but i'm not sure it is the problem here.. if i'm reading this right, blowing past the max wil...
- 04:25 PM Bug #5407 (Resolved): mon: is_writeable doesn't match wait_for_writeable on cuttlefish
- bah, well, ended up just backporting the rest of the series from master. in the long run that will be easier to main...
- 06:21 AM Bug #5407 (Fix Under Review): mon: is_writeable doesn't match wait_for_writeable on cuttlefish
- pushed to wip-mon-cuttlefish.
- 04:24 PM Bug #5205 (Resolved): mon: FAILED assert(ret == 0) on config's set_val_or_die() from pick_address...
- 12:53 PM Bug #5205 (Fix Under Review): mon: FAILED assert(ret == 0) on config's set_val_or_die() from pick...
- 04:24 PM Bug #5195: "ceph-deploy mon create" fails when adding additional monitors
- yeah, let's update the docs ("the new monitor needs to know what address to bind to") and close this bug. Yay!
- 04:22 PM Bug #5195: "ceph-deploy mon create" fails when adding additional monitors
- After clean-up and re-installation of wip-5195 it worked.
2013-06-24 19:13:52.853138 7f3f92b9e700 0 log [INF] : m... - 01:51 PM Bug #5195: "ceph-deploy mon create" fails when adding additional monitors
- I ran a test with the wip-5195 branch. That fixed the issue with the assert.
There now seems to be a authenticati... - 12:53 PM Bug #5195 (Fix Under Review): "ceph-deploy mon create" fails when adding additional monitors
- 10:55 AM Bug #5195: "ceph-deploy mon create" fails when adding additional monitors
- Gary Lowell wrote:
> With the public_network statement added to ceph.conf it look slike it gets further gut hits an ... - 10:20 AM Bug #5195: "ceph-deploy mon create" fails when adding additional monitors
- With the public_network statement added to ceph.conf it look slike it gets further gut hits an assert. Stack trace a...
- 01:29 PM Revision ab6ccbe2 (ceph): test/librados/cmd.cc: use static_cast instead of C-Style cast
- Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>
- 01:27 PM rgw Feature #4613: Allow bucket data to reside in a separate pool to object data
- Implemented this one for dumpling. Made it possible to define different data and index pools for a bucket, and give u...
- 01:24 PM Revision 626d9eed (ceph): opstate RESTful API implementation with unit test application
- Signed-off-by: Babu Shanmugam <anbu@enovance.com>
- 01:24 PM Revision 79b5a486 (ceph): osdc/Objecter.cc: use static_cast instead of C-Style cast
- Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>
- 01:19 PM Revision e7602a1e (ceph): mon/MonClient.cc: use static_cast instead of C-Style cast
- Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>
- 12:54 PM Bug #5427 (Resolved): mon: could not get service secret for auth subsystem
- 10:20 AM Bug #5427 (Pending Backport): mon: could not get service secret for auth subsystem
- 10:01 AM Bug #5427: mon: could not get service secret for auth subsystem
- Looked at it yesterday, pull request 374. I believe it to be correct.
- 09:26 AM Bug #5427: mon: could not get service secret for auth subsystem
- Joao - please review.
- 12:34 PM Revision de4a7645 (ceph): common/cmdparse.cc: reduce scope of local variable 'pos'
- Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>
- 12:29 PM Revision b485a3e6 (ceph): common/cmdparse.cc: remove unused variable
- Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>
- 12:24 PM Revision a92a720e (ceph): osd/OSD.cc: prefer prefix ++operator for non-trivial iterator
- Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>
- 12:18 PM Revision 835315b6 (ceph): OSDMonitor.cc: prefer prefix ++operator for non-trivial iterator
- Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>
- 12:14 PM Revision c700db09 (ceph): mon/MonCap.cc: use empty() instead of if(size())
- Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>
- 11:50 AM Revision 4ab5bf6f (ceph): common/cmdparse.cc: prefer prefix ++operator for non-trivial iterator
- Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>
- 11:44 AM Bug #5154: osd/SnapMapper.cc: 270: FAILED assert(check(oid))
- just hit this on job...
- 11:43 AM Bug #5440 (Resolved): osd: marked down due to no pgstats reports
- 2013-06-24T02:04:34.124 INFO:teuthology.task.ceph.mon.b.err:2013-06-24 02:04:37.762017 7fe7e462b700 -1 mon.b@0(leader...
- 11:40 AM rgw Bug #5439 (Resolved): rgw: 500 returned on s3readwrite
- teuthology@/a/teuthology-2013-06-24_01:00:42-rgw-master-testing-basic/44036:...
- 10:58 AM CephFS Bug #5333 (Resolved): mds: segfault in MDLog::standby_trim_segments
- done, commit:f046dab88fcfeda23391bcd694abc65ff1ed8cd8
- 10:12 AM CephFS Bug #5333 (Pending Backport): mds: segfault in MDLog::standby_trim_segments
- I saw this crash under teuthology in the next branch as well; can we put it there?
- 10:44 AM CephFS Bug #5411: teuthology: bad object dereference
- #5333 is what I was referring to. There's a whole string of failures which are hitting both that and this.
- 10:08 AM CephFS Bug #5411: teuthology: bad object dereference
- Josh, I went back and looked at the first instance (/a/teuthology-2013-06-18_01\:00\:37-fs-next-testing-basic/38877/)...
- 10:05 AM CephFS Bug #5411: teuthology: bad object dereference
- Happened again...
- 09:45 AM CephFS Bug #5411: teuthology: bad object dereference
- If you look at the message from the first exception, it says the mds failed:...
- 09:36 AM CephFS Bug #5382: mds: failed objecter assert on shutdown
- /a/teuthology-2013-06-23_20:00:47-fs-cuttlefish-testing-basic/43843/teuthology.log
- 09:10 AM CephFS Bug #5250: ceph-mds 0.61.2 aborts on start
- Unfortunately this is an area where CephFS needs some hardening and some recovery tools — part of why we don't recomm...
- 05:49 AM CephFS Bug #5250: ceph-mds 0.61.2 aborts on start
- We have fit a very similar problem with V0.61.2. We are unable to start any MDS daemons following testing that involv...
- 09:09 AM Bug #5432 (Pending Backport): msgr: bad locking mark_down_all
- the crash was from the earlier changes that are only in master, and this whole series is just to fix Connection and *...
- 08:17 AM Feature #5437 (Resolved): ceph-mon performance on ARM
- Scaling tests with 168 OSDs show bottlenecks when ceph-mon is running on ARM processors. Requests are not being proc...
- 06:30 AM Subtask #2659 (Can't reproduce): mon: Single-Paxos: ceph tool -w subscriptions not being updated
- 06:28 AM Subtask #2621 (Resolved): mon: Single-Paxos: synchronize the MonitorDBStore of oblivious monitor
- 06:26 AM Bug #5294 (Closed): mon upgrade issue 0.61.2 -> 0.61.3
- From ceph-users, it appears the monitors were just taking a long, long time compacting a 500GB+ store (which explains...
- 06:18 AM Revision 1e3161a8 (ceph): rgw: add a system request param to select op bucket instance
- Added rgwx-bucket-instance request param that selects the actual bucket
instance to use for this request.
Signed-off... - 05:42 AM Revision 134d08a9 (ceph): Merge pull request #375 from ceph/wip-msgr
- misc msgr fixes
Reviewed-by: Greg Farnum <greg@inktank.com> - 05:34 AM Revision 76228caf (ceph): rgw: data log contains bucket instance info in key
- Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
- 04:58 AM Revision b338524f (ceph): Revert "upgrade: add double-hop bobtail -> cuttlefish -> next fs tests"
- This reverts commit fbdb7774bb381ff1426650c47cdb36282c4c77d2.
We don't care about bobtail fs. - 04:58 AM Revision 56f9b29e (ceph): upgrade: add rados-double suite (bobtail -> cuttlefish -> next)
- 04:53 AM Revision 2917fd4a (ceph): upgrade/rados: refactor, expand matrix, start with cuttlefish
- 04:44 AM Revision e672d6b2 (ceph): upgrade/fs: make single hop cuttlefish -> next (not bobtail -> next)
- The bobtail ceph-fuse is buggy and these tests keep failing e.g. on umount,
but we don't really care about fixing the... - 04:44 AM Revision 5dd137ac (ceph): rgw: format bucket.instance meta entries nicely
- Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
- 04:41 AM Revision 827d2f81 (ceph): upgrade: refactor single-hop fs tests
- 04:41 AM Revision fbdb7774 (ceph): upgrade: add double-hop bobtail -> cuttlefish -> next fs tests
- 04:16 AM Revision 4e90c5ec (ceph): rgw: use new instance handler for bucket operations
- Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
- 04:00 AM Revision 71869c4b (ceph): rgw: create meta handler for bucket instance
- Create utility functions for reading writing bucket entry
point and bucket instance. Add a separate meta handler for
... - 03:44 AM Revision 947eba1a (ceph): rados: fix multiclient tests
- Each client (not run) gets its own pool!
Signed-off-by: Sage Weil <sage@inktank.com> - 02:42 AM Revision 00973dfd (ceph): rgw: put_bucket_info() uses objv_tracker in bucket info
- Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
- 01:52 AM Revision dab57ef8 (ceph): rgw: keep objv_tracker on bucket info
- This ties it to the specific object instance info.
Signed-off-by: Yehuda Sadeh <yehuda@inktank.com> - 01:09 AM Revision 57dc7362 (ceph): msgr: clear_pipe+queue reset when replacing lossy connections
- We already handle the lossless replacement and lossy fault paths, but
not the lossy replacement. This fixes an asser...
06/23/2013
- 11:21 PM Revision e6e1df69 (ceph): dump_stuck: fix race with osd start
- Occasionally we don't wait long enough for the osd to start and
mark itself up. Keep trying until flush succeeds.
F... - 11:20 PM Bug #5195: "ceph-deploy mon create" fails when adding additional monitors
- Yeah,I think as things currently stand though the Mon looks for that option to be defined. We can probably fix it in...
- 10:55 PM Bug #5195: "ceph-deploy mon create" fails when adding additional monitors
- Sage Weil wrote:
> oh, right. in this case i think teh thing to do is add 'public network = 1.2.3.0/24' or whatever... - 10:03 PM Bug #5195: "ceph-deploy mon create" fails when adding additional monitors
- oh, right. in this case i think teh thing to do is add 'public network = 1.2.3.0/24' or whatever to the ceph.conf so...
- 10:43 PM Bug #5432: msgr: bad locking mark_down_all
- Merged into master with commit:134d08a9654f66634b893d493e4a92f38acc63cf. Does wip-msgr need any backports? I think th...
- 03:10 PM Bug #5432 (Fix Under Review): msgr: bad locking mark_down_all
- wip-msgr fixes this already, needs review!
- 11:03 AM Bug #5432 (Resolved): msgr: bad locking mark_down_all
- ...
- 10:38 PM Revision e053d171 (ceph): rgw: refactor policy reading
- reuse bucket info, no need to reread it again.
Signed-off-by: Yehuda Sadeh <yehuda@inktank.com> - 10:12 PM CephFS Bug #5021: ceph-fuse: crash on traceless reply
- btw wip-5021 still hasn't merged because it failed the smbtorture test. i'll rebase on master and retest to see wher...
- 10:11 PM devops Bug #5193: RHEL6 does not ship with xfsprogs
- xfs support is ramping up for newer 6.x rhel releases, so think this goes away on its own...
- 10:10 PM Revision 5fc1dabf (ceph): msgr: clear_pipe inside pipe_lock on mark_down_all
- Observed a segfault in rebind -> mark_down_all -> clear_pipe -> put that
may have been due to a racing thread clearin... - 10:10 PM Revision afafb87e (ceph): msgr: clear con->pipe inside pipe_lock on mark_down
- We need to do this under protection of the pipe_lock.
Signed-off-by: Sage Weil <sage@inktank.com> - 10:10 PM Revision ec612a5b (ceph): msg/Pipe: goto fail_unlocked on early failures in accept()
- Instead of duplicating an incomplete cleanup sequence (that does not
clear_pipe()), goto fail_unlocked and do the cle... - 10:10 PM Revision 9586305a (ceph): msgr: reaper: make sure pipe has been cleared (under pipe_lock)
- All paths to pipe shutdown should have cleared the con->pipe reference
already. Assert as much.
Also, do it under p... - 10:09 PM CephFS Bug #5105 (Duplicate): mds/CInode.cc: 1996: FAILED assert(auth_pins >= 0)
- #4832
- 10:06 PM CephFS Bug #5333 (Resolved): mds: segfault in MDLog::standby_trim_segments
- commit:abd0ff64e108b7670a062b3fa39baaf3d3e48fb3
- 08:39 PM Revision d74cdad5 (ceph): rgw: fix policy read
- policy reading used to have exact same logic for both buckets
and objects, however, we now need to read bucket attrib... - 08:12 PM Revision ad640672 (ceph): rgw: init member variable
- Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
- 08:12 PM Revision b09d799f (ceph): rgw: encode bucket info only after setting a flag
- Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
- 04:30 PM CephFS Bug #5430 (Duplicate): newfs makes ceph-mds segfault in suicide
- #5432
- 10:57 AM CephFS Bug #5430: newfs makes ceph-mds segfault in suicide
- ...
- 10:52 AM CephFS Bug #5430 (Duplicate): newfs makes ceph-mds segfault in suicide
- ...
- 04:25 PM Revision cd98eb0c (ceph): mon/AuthMonitor: make initial auth include rotating keys
- This closes a very narrow race during mon creation where there are no
service keys.
Fixes: #5427
Signed-off-by: Sage... - 04:22 PM Bug #5431 (Fix Under Review): osd: dump_stuck test fails with ENXIO
- https://github.com/ceph/teuthology/pull/16
- 11:00 AM Bug #5431 (Resolved): osd: dump_stuck test fails with ENXIO
- ...
- 04:15 PM Revision 13dbe9d1 (ceph): enable-coredump -> adjust-ulimits
- and set max_files to be big, too!
- 04:15 PM Revision 253cc98d (ceph): enable-coredump -> adjust-ulimits
- and set max_files to be big, too!
- 03:53 PM Revision 9b2dfb75 (ceph): mon: do not leak no_reply messages
- I think I assumed no_reply() was releasing the references, but it is
not. Which is better, since send_reply() doesn'... - 03:53 PM Revision ad12b0d6 (ceph): mon: fix leak of MOSDFailure messages
- We need to discard/cancel/free the failure report messages before we
cancel a report out. Assert in the dtor to ensu... - 01:30 PM Subtask #5433 (Rejected): Factor out the ReplicatedPG object replication and client IO logic as a...
- "work in progress":https://github.com/dachary/ceph/tree/wip-5433
h3. Moving code PG <=> ReplicatedPG
Prior to d... - 12:20 PM Subtask #5085 (Rejected): PG::merge_log should not have side effects other than on the log & miss...
- It's probably too early in the process to do that kind of enhancement / modification.
- 10:37 AM Bug #5427 (Fix Under Review): mon: could not get service secret for auth subsystem
- 09:26 AM Bug #5427: mon: could not get service secret for auth subsystem
- the leader never ticked while paxos was healthy, i think because of the clock skew.
see wip-5427 - 08:59 AM Bug #5427 (Resolved): mon: could not get service secret for auth subsystem
- ...
- 10:24 AM Linux kernel client Bug #5429: libceph: rcu stall, null deref in osd_reset->__reset_osd->__remove_osd
- leaving plana56 in kdb
- 10:23 AM Linux kernel client Bug #5429 (Resolved): libceph: rcu stall, null deref in osd_reset->__reset_osd->__remove_osd
- ...
- 10:15 AM rbd Bug #5428: libceph: null deref in ceph_auth_reset
- leaving plana09 in kdb
- 10:12 AM rbd Bug #5428: libceph: null deref in ceph_auth_reset
- first guess was a shutdown race, but ceph_monc_stop() is flushing the msgr wq. also, no other threads appear to be i...
- 10:02 AM rbd Bug #5428 (Can't reproduce): libceph: null deref in ceph_auth_reset
- ...
- 08:51 AM rbd Bug #5426 (Resolved): librbd: mutex assert in perfcounters::tinc in librbd::AioCompletion::comple...
- ...
- 08:15 AM Revision c3260b27 (ceph): rgw: initial work to separate bucket name and instance
- Bucket info now resides in an instance specific object. The
old bucket info object now contains a pointer to the new
... - 08:15 AM Revision c103031e (ceph): rgw: initialize system flag in RGWUserInfo
- Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
- 08:08 AM rbd Bug #5425: krbd: xfstest 89 hang, 'read_partial_message skipping long message'
- ...
- 08:03 AM rbd Bug #5425 (Resolved): krbd: xfstest 89 hang, 'read_partial_message skipping long message'
- ...
- 08:04 AM Revision 7e3493f8 (ceph): rgw: various object put processor fixes
- Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
- 06:26 AM Revision 5f43fa81 (ceph): rgw: fix broken obj put
- Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
06/22/2013
- 05:28 PM Revision 1aca370e (ceph): debian: ceph-common requires matching version of python-ceph
- If they skew the ceph_argparse.py module may be missing.
Signed-off-by: Sage Weil <sage@inktank.com> - 02:15 PM Bug #5401: cuttlefish osd recovery slow
- This one seems to "hide" the problem:
commit 9fe5611fdd7374654ad58185fa3988e216c52f08
Author: Stefan Priebe <s.prie... - 02:04 PM Bug #5401: cuttlefish osd recovery slow
- max may also be negative but there is only a check for == 0 in OSD.cc
- 01:03 PM Bug #5401: cuttlefish osd recovery slow
- It seems locking and unlocking the mutex isn't working correctly. So multiple threads seem to higher recovery_ops_act...
- 12:31 PM Bug #5401: cuttlefish osd recovery slow
- g_conf->osd_recovery_max_active is 5 but on a freshly started osd i'm seeing log messages like these ...2013-06-22 .....
- 06:10 AM Bug #5401: cuttlefish osd recovery slow
- It did defer the recovery again and again. Some example log output:
2013-06-22 15:07:20.187878 7f8c3f49c700 15 osd... - 04:11 AM Bug #5401: cuttlefish osd recovery slow
- I could hide this problem by higher "osd recovery delay start" => 120 but then the overall recovery time is very high...
- 12:04 PM Bug #5424: mon/Paxos.cc: 549: FAILED assert(begin->last_committed == last_committed)
- Greg Farnum wrote:
> Shouldn't that cause LevelDB to block or throw an error or something? I'm not quite sure how it... - 10:22 AM Bug #5424: mon/Paxos.cc: 549: FAILED assert(begin->last_committed == last_committed)
- Shouldn't that cause LevelDB to block or throw an error or something? I'm not quite sure how it leads to us not readi...
- 09:53 AM Bug #5424 (Resolved): mon/Paxos.cc: 549: FAILED assert(begin->last_committed == last_committed)
- all peons died with teh above assert. the leader did this:...
- 04:59 AM Revision d7af5e14 (ceph): rgw: handle bucket creation with specified placement pool
- Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
- 01:46 AM Revision b89d7420 (ceph): Merge branch 'next'
- Conflicts:
src/ceph.in - 01:39 AM Revision 94eada40 (ceph): Add header comments and Inktank copyrights to ceph.in/ceph_argparse.py
- Signed-off-by: Dan Mick <dan.mick@inktank.com>
- 01:39 AM Revision 67a3c1e4 (ceph): ceph.in: rip out reusable code to pybind/ceph_argparse.py
- Signed-off-by: Dan Mick <dan.mick@inktank.com>
Conflicts:
src/ceph.in
06/21/2013
- 11:50 PM Bug #5195 (In Progress): "ceph-deploy mon create" fails when adding additional monitors
- The problem occurs when a monitor is added on a host that was not in the initial list of cluster members.
Sequence... - 11:22 PM Revision c1198d68 (ceph): debian: update postinst, prerm hooks
- This syncs up the hooks with the latest master versions. In particular,
do not blindly stop/restart daemons on packa... - 11:16 PM Revision 2fcbf2ba (ceph): rgw: user configuration for bucket placement
- Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
- 11:06 PM Revision e5e924ca (ceph): rgw: data structures for new data/index placement rules
- Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
- 10:53 PM Revision d39d648e (ceph): Merge remote-tracking branch 'remotes/origin/wip-sandon-cephdeploy'
- 10:52 PM Revision c4272a17 (ceph): ceph: even shinier
- Signed-off-by: Sage Weil <sage@inktank.com>
Reviewed-by: Dan Mick <dan.mick@inktank.com> - 10:50 PM Revision 34ef2f24 (ceph): ceph: do not busy-loop on ceph -w
- Signed-off-by: Sage Weil <sage@inktank.com>
Reviewed-by: Dan Mick <dan.mick@inktank.com> - 10:27 PM Revision a6142a64 (ceph): Merge pull request #14 from clee/cleanup
- Clean up nested-if logic
Reviewed-by: Sage Weil <sage@inktank.com> - 09:53 PM Revision 27912e58 (ceph): librados: make cmd test tolerate NXIO for osd commands
- The cluster may be thrashing underneath us; tolerate NXIO in case the OSD
is currently down.
Signed-off-by: Sage Wei... - 09:45 PM Revision f6270a7f (ceph): Wipe out existing id_rsa.pub and id_rsa before pushing ssh keys
- A very simple change. Just touch a file first (to create it if it
doesn't yet exist so the delete doesn't error out) ... - 09:25 PM Revision dcd27531 (ceph): Merge remote-tracking branch 'gh/wip-mds'
- Reviewed-by: Sage Weil <sage@inktank.com>
- 09:23 PM Revision abd0ff64 (ceph): mds: do not assume segment list is non-empty in standby_trim_segments
- If we restart standby replay shortly after startup, before we actually have
any segments, we an trigger a segfault he... - 06:53 PM Revision 241ad073 (ceph): rgw: make replica log object name configurable
- Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
- 06:53 PM Revision d572cf6f (ceph): mds: fix iterator invalidation for backtrace removal
- - Don't increment before we dereference!
- We need to update the iterator before we delete the item.
This code is ch... - 06:50 PM Revision 8e4c292e (ceph): rados: whitelist 'had wrong client addr' for cephtool test
- Recently added stuff to cephtool/test.sh workunit that triggers this.
fd769c0f21aab47cb7faacf7b6122e5288f1156e - 06:47 PM Revision 469900e6 (ceph): RGWReplicaBucketLogger: store bucket replica logs in the .logs pool
- Signed-off-by: Greg Farnum <greg@inktank.com>
- 05:53 PM Revision 409422be (ceph): Merge branch 'wip-rgw-geo-2' into wip-rgw-geo
- Conflicts:
src/test/cli/radosgw-admin/help.t - 05:47 PM rgw Bug #5422 (Resolved): object deletion should log the object tag
- object tags are used to tell one instance of an object from another with the same name (to differentiate a deleted ob...
- 05:07 PM Bug #5414 (Resolved): qa jobs failing with /var/lib/ceph/osd/*/journal root-owned from prior clus...
- 04:30 PM Bug #5414: qa jobs failing with /var/lib/ceph/osd/*/journal root-owned from prior clusters
- 10:29 AM Bug #5414 (Resolved): qa jobs failing with /var/lib/ceph/osd/*/journal root-owned from prior clus...
- this is causing various runs to fail. so far i see it on upgrade runs. maybe the package update is triggering somet...
- 04:35 PM Revision 50957772 (ceph): osd: init test_ops_hook
- CID 1019628 (#1 of 1): Uninitialized pointer field (UNINIT_CTOR)
2. uninit_member: Non-static class member "test_ops_... - 04:35 PM Revision 17d2745f (ceph): osd: initialize OSDService::next_notif_id
- CID 1019627 (#1 of 1): Uninitialized scalar field (UNINIT_CTOR)
2. uninit_member: Non-static class member "next_notif... - 03:20 PM Revision 3bebbc09 (ceph): mds: rev protocol
- Commit 18b9e63b4df643e1f2fb8f17416089e5d970bf60 changed the OTW lock
encoding.
Signed-off-by: Sage Weil <sage@inktan... - 03:19 PM Revision ded2e84f (ceph): mds: kill Server::handle_client_lookup_hash()
- Server::handle_client_lookup_ino() is more simple and robust. Use it
to handle both LOOKUPHASH and LOOKUINO requests.... - 03:19 PM Revision 2147c4e3 (ceph): mds: use "open-by-ino" helper to handle LOOKUPINO request
- Fixes #3541
Signed-off-by: Yan, Zheng <zheng.z.yan@intel.com> - 02:33 PM rgw Feature #4335 (Resolved): rgw: dr: sync processing state: define datastructures
- Oooh! Oooh! Merged into the integration branch.
- 02:15 PM Feature #5421: mon: add formatter option for various mon commands
- I've done "auth export". The tricky bit is understanding the naming and levels of containers (that end up unnamed in...
- 02:13 PM Feature #5421 (Resolved): mon: add formatter option for various mon commands
- 02:14 PM Feature #4983: OSD: namespaces pt 2 (caps)
- 02:08 PM Feature #3983 (In Progress): api: create initial DRAFT REST API Design
- 02:08 PM Feature #4463 (In Progress): api: RESTful client: demonstrate remaining N-1 commands JSON or XML
- 02:07 PM Feature #4462 (In Progress): api: RESTful client: implement remaining N-1 commands JSON or XML
- 02:07 PM Feature #4459 (In Progress): api: RESTful client: implement remaining commands JSON only
- 02:07 PM Feature #4460 (In Progress): api: RESTful client: demonstrate remaining N-1 commands JSON only
- 02:07 PM Fix #5278 (In Progress): osd: smarter recovery for small objects
- 02:06 PM Feature #4457 (Resolved): api: add JSON schema/output protocol to rados.py
- 02:06 PM Feature #4458 (Resolved): api: RESTful client: prototype 1 command JSON only
- 02:06 PM Feature #4461 (Resolved): api: RESTful client: prototype 1 command JSON or XML
- 02:06 PM Feature #4547 (Resolved): api: implement self-description for --admin-daemon commands
- 02:04 PM Feature #4548 (Resolved): api: implement self-description for osd/mon tell commands
- 02:04 PM Feature #4455 (Resolved): api: move '--format' into just another command argument
- 01:58 PM rgw Feature #5420 (Rejected): rgw: integrate bucket metadata changes with bucket index log
- for the sake of correctness and robustness
- 01:43 PM devops Feature #5013 (In Progress): build internal openstack + ceph cluster out of some burnupi
- 01:43 PM devops Feature #5214 (In Progress): Kernel gitbuilders for rpm distros
- 12:48 PM Feature #5419 (New): cephtool: sanitize extra args before configuring cluster handle
- The '--admin-socket' argument is not a valid ceph tool argument. The option one would actually want is '--admin-daemo...
- 12:02 PM CephFS Bug #5418: kceph: crash in remove_session_caps
- kdb dumpall attached
- 12:02 PM CephFS Bug #5418 (Resolved): kceph: crash in remove_session_caps
- ...
- 11:33 AM rgw Feature #5417 (Resolved): rgw: separate bucket metadata object into pointer object and instance o...
- Instead of having a single bucket metadata object, we'll separate into a bucket 'head' object that will point at the ...
- 11:07 AM rgw Bug #5416: --help output needs --rgw-zone option
- For the ./radosgw-admin command that is.
- 11:07 AM rgw Bug #5416 (Resolved): --help output needs --rgw-zone option
- The new --rgw-zone option needs to be added to the --help output.
- 10:30 AM rgw Bug #5415 (Resolved): rgw: failing valgrind leak checks
- teuthology-2013-06-21_01:00:44-rgw-master-testing-basic/41708 and 41709
- 10:13 AM rgw Feature #5356 (Rejected): rgw: RESTful api for bucket upstream zone + marker info
- As with #5353, we're not doing it at the moment, not clear if we really need it.
- 10:10 AM rgw Feature #5355 (Rejected): rgw: get and set bucket upstream zone + marker info
- We're not doing it atm. Discussed it with Greg, and we think that the replica log already covers this info, so there'...
- 09:36 AM Bug #5413 (Resolved): osd: valgrind issue in watch code (cuttlefish?)
- commit:17d2745f095e7bb640dece611d7824d370ea3b81
- 08:17 AM Bug #5413 (Resolved): osd: valgrind issue in watch code (cuttlefish?)
- teuthology-2013-06-20_20:00:11-rados-cuttlefish-testing-basic/41401...
- 09:25 AM rbd Feature #4550: Create Qemu+RBD rpm package for RHEL+CentOS 6.3 on ceph.com
- These packages have not been through QA yet.
- 09:13 AM rbd Feature #4550: Create Qemu+RBD rpm package for RHEL+CentOS 6.3 on ceph.com
- Great, thanks Gary. Have these been QAd?
- 04:57 AM Revision 24e59b43 (ceph): rgw: buffer atomic put handler
- Since we tied the atomic put handler to libcurl output
data, which uses much smaller chunks, we need to buffer
data, ... - 04:57 AM Revision 02de43ad (ceph): rgw: tie opstate into intra-region copy operations
- Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
06/20/2013
- 11:13 PM Bug #5301: mon: leveldb crash in tcmalloc
- Maciej Galkiewicz wrote:
> debian wheezy (7.0)
ok now it sounds a lot like #5239. i'm not able to reproduce this... - 11:04 PM Revision b7763af4 (ceph): big: enable big cluster (~50 node)
- 11:01 PM Feature #3273 (Need More Info): mon: simple dm-crypt key management
- http://marc.info/?l=ceph-devel&m=137179443405614&w=2
- 10:55 PM Revision e97a2c86 (ceph): Merge remote-tracking branch 'yan/wip-mds' into wip-mds
- 10:40 PM rbd Feature #4550: Create Qemu+RBD rpm package for RHEL+CentOS 6.3 on ceph.com
- Following the packaging discussions, the redhat packages were respun with the latest redhat sources + the ceph rados ...
- 10:14 PM Revision 69e1a912 (ceph): ceph.in: fix ^C handling in watch (trap exception in while, too)
- Signed-off-by: Dan Mick <dan.mick@inktank.com>
- 10:14 PM Revision 31d221c3 (ceph): ceph.in: remove some TAB chars
- Signed-off-by: Dan Mick <dan.mick@inktank.com>
- 10:04 PM Revision 29f6f277 (ceph): ceph: --version as well as -v
- Signed-off-by: Sage Weil <sage@inktank.com>
- 09:51 PM rgw Feature #4340 (In Progress): rgw: dr: data sync agent: implement full sync
- 09:51 PM rgw Feature #5358 (In Progress): rgw: RESTful api for intra-region copy state
- 09:50 PM rgw Feature #5356 (In Progress): rgw: RESTful api for bucket upstream zone + marker info
- 09:50 PM rgw Feature #5341 (Fix Under Review): rgw: keep state for cross-rgw copy operations
- 09:42 PM Revision 234fab09 (ceph): Merge branch 'wip-rgw-geo-enovance' into wip-rgw-geo-2
- 09:41 PM Revision 6d6dc42e (ceph): fixup "radosgw-admin: new commands to look at and clean up replica logs"
- 09:33 PM CephFS Fix #5399: timestamp changes on replayed mds request (pjd link 71)
- probably need to extend the replayed request message to include the timestamps we got for the inode and dir so that t...
- 09:33 PM CephFS Fix #5399: timestamp changes on replayed mds request (pjd link 71)
- - we send a create to mds
- get an ack, but it isn't journaled
- pjd stats the mtime/ctime/ec.
- mds restarts
- w... - 09:12 PM CephFS Bug #5290: mds: crash whilst trying to reconnect
- ubuntu@teuthology:/var/lib/teuthworker/archive/teuthology-2013-06-20_13:32:57-fs-master-testing-basic/41231
logs in ... - 09:10 PM Revision e4ef5c6f (ceph): cls_replica_log: add the actual class
- This class can be applied to any object, and will keep track of
the bounds, matching times, and matching sets of in-p... - 09:10 PM Revision 29df88af (ceph): cls_replica_log: add client with user documentation
- Signed-off-by: Greg Farnum <greg@inktank.com>
- 09:10 PM Revision 36514731 (ceph): cls_replica_log: add tests.
- We do several good updates, a bad update, a good delete,
a bad delete, several good gets, and a bad get.
Signed-off-... - 09:10 PM Revision ab79ba46 (ceph): cls_replica_log: integrate with RGWRados
- We introduce an implementation class RGWReplicaLogger, and two user
classes RGWReplicaObjectLogger (for the data/meta... - 09:10 PM Revision 59bfdd00 (ceph): radosgw-admin: new commands to look at and clean up replica logs
- Signed-off-by: Greg Farnum <greg@inktank.com>
- 09:10 PM Revision 22a02e95 (ceph): cls_replica_log: add ops for new class
- Signed-off-by: Greg Farnum <greg@inktank.com>
- 09:10 PM Revision d1c95943 (ceph): cls_replica_log: add types for new class
- Signed-off-by: Greg Farnum <greg@inktank.com>
- 08:42 PM Revision 70b54671 (ceph): Clean up nested-if logic
- 08:38 PM Revision 0deb6d40 (ceph): rgw: lock related modifications
- 1. zone-id inclusion in lock/unlock for mdlog and datalog
2. renewal of lock if the locker request lock again
3. modi... - 08:32 PM Revision 3b4c11b2 (ceph): rgw: add max-entries, marker for log operations
- Signed-off-by: Babu Shanmugam <anbu@enovance.com>
Conflicts:
src/rgw/rgw_admin.cc
src/rgw/rgw_rest_log.cc - 07:30 PM Revision 4bf5b732 (ceph): Merge remote-tracking branch 'gh/next'
- 07:26 PM Revision 714f2128 (ceph): osdc: re-calculate truncate_size for strip objects
- Feed truncate_size through the striping algorithm so that it reflects the
correct per-object offset (as opposed to th... - 07:19 PM Revision ebb46c45 (ceph): qa/workunits/misc/multiple_rsync.sh: wtf
- 2013-06-15T12:55:29.808 INFO:teuthology.task.workunit.client.0.err:+ rsync -auv --exclude local/ /usr/ usr.1
2013-06-... - 06:45 PM CephFS Bug #5333 (Fix Under Review): mds: segfault in MDLog::standby_trim_segments
- wip-5333
this looks like a simple matter of not crashing if the segment list is empty. that at least covers this ... - 12:53 PM CephFS Bug #5333: mds: segfault in MDLog::standby_trim_segments
- Just a note: maybe we missed a spot, but I remember doing a re-read head object, retry journal read whenever we get a...
- 12:47 PM CephFS Bug #5333: mds: segfault in MDLog::standby_trim_segments
- ubuntu@teuthology:/var/lib/teuthworker/archive/teuthology-2013-06-20_01:00:49-fs-next-testing-basic/40965
with ful... - 06:30 PM Revision ffdb7236 (ceph): mon: more fix dout use in sync_requester_abort()
- Signed-off-by: Sage Weil <sage@inktank.com>
(cherry picked from commit d60534b8f59798feaeeaa17adba2a417d7777cbf) - 06:29 PM Revision 38ddae04 (ceph): mon: fix raw use of *_dout in sync_requester_abort()
- Signed-off-by: Sage Weil <sage@inktank.com>
(cherry picked from commit 8a4ed58e39b287fd8667c62b45848487515bdc80) - 06:28 PM Revision fd769c0f (ceph): qa/workunits/cephtool/test.sh: fix and cleanup several tests
- Signed-off-by: Sage Weil <sage@inktank.com>
- 06:23 PM Revision f420e5c6 (ceph): mon: drop deprecated 'stop_cluster'
- Signed-off-by: Sage Weil <sage@inktank.com>
- 06:23 PM Revision 4977b88a (ceph): mds: make 'mds compat rm_*compat' idempotent
- Signed-off-by: Sage Weil <sage@inktank.com>
- 06:15 PM CephFS Bug #5380 (Resolved): osdc/Filer.cc: 163: FAILED assert(probe->known_size[p->oid] <= shouldbe)
- 12:30 PM CephFS Bug #5380: osdc/Filer.cc: 163: FAILED assert(probe->known_size[p->oid] <= shouldbe)
- 06:11 PM Revision 4a038d6d (ceph): mon: make 'log ...' command wait for commit before reply
- Previously we would just dump the command argument to our local log client
and reply immediately, which could lose th... - 06:09 PM Revision 483aa74f (ceph): test: update cli test
- Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
- 06:09 PM Revision 008bdc5e (ceph): radosgw_admin: fixup cli test
- Signed-off-by: Greg Farnum <greg@inktank.com>
- 06:04 PM Revision 7b5ec743 (ceph): cls_rgw: skip namespaced entries when listing bucket
- We added a namespace to the entries, and sadly there can be
valid entries beyond it. Previously we assumed that once ... - 06:04 PM Revision 5de54f6a (ceph): a/workunits/cephtool/test.sh: --no-log-to-stderr when examining stderr
- We can get random messages to stderror from socket reconnects and such;
discard those if we are looking at stderr in ... - 05:42 PM Bug #5412 (Resolved): doc bug: incorrect reference to monitor quorum requirements
- http://ceph.com/docs/master/rados/deployment/ceph-deploy-mon/...
- 04:46 PM Revision d60534b8 (ceph): mon: more fix dout use in sync_requester_abort()
- Signed-off-by: Sage Weil <sage@inktank.com>
- 03:36 PM Revision 8a4ed58e (ceph): mon: fix raw use of *_dout in sync_requester_abort()
- Signed-off-by: Sage Weil <sage@inktank.com>
- 03:08 PM Revision c8af85d4 (ceph): doc/release-notes: v0.61.4
- Signed-off-by: Sage Weil <sage@inktank.com>
- 02:56 PM Bug #4004 (Can't reproduce): Intermittent kernel build failures
- 02:52 PM Bug #4004: Intermittent kernel build failures
- Closing since we haven't seen any problems for a couple months.
- 02:55 PM devops Cleanup #5106 (Resolved): ceph_deploy: install/compile error on wheezy
- The was happening do to syntax in the test programs that wasn't supported on Python 2.6. Not shipping the test direc...
- 02:51 PM Bug #2176 (Resolved): dependencies not checked by autoconf
- We've got all the current dependencies in the configure.ac checks and in the rpm or debian requirements.
- 02:42 PM CephFS Bug #5411 (Resolved): teuthology: bad object dereference
- ...
- 01:30 PM CephFS Fix #5268: mds: fix/clean up file size/mtime recovery code
- See also #4485.
- 01:30 PM CephFS Feature #4485: Improve "needsrecover" handling
- See also #5268.
- 01:24 PM CephFS Feature #1693 (In Progress): libcephfs: Support TRIM (hole punching)
- See "[PATCH] Ceph-fuse: Punch hole support" from Li Wang.
- 01:17 PM CephFS Feature #3541 (In Progress): mds: robust ino lookup using file backpointers
- A bunch of this got done, but Sage isn't sure if the client -> LOOKUPINO messages are wired up to that infrastructure...
- 12:58 PM Feature #4929: Erasure encoded placement group
- the pad is only archived for so long, keep a "pad backup":http://pad.ceph.com/p/Erasure_encoding_as_a_storage_backend...
- 11:31 AM Bug #5409 (Resolved): mon: log command does not wait for commit
- 10:43 AM Bug #5409 (Resolved): mon: log command does not wait for commit
- the mon replies immediately, and may lose the msg if it restarts....
- 10:35 AM rgw Feature #5408 (Resolved): rgw: turn off dr/geo logging
- 10:26 AM Bug #5407 (Resolved): mon: is_writeable doesn't match wait_for_writeable on cuttlefish
- fixes in master.. need a minimal cuttlefish backport.
- 09:58 AM rgw Feature #5406 (Resolved): rgw: a RESTful api to dump region map
- 09:42 AM devops Bug #5405 (Resolved): ceph-deploy: transient pushy exception on install
- ...
- 08:37 AM devops Feature #5403 (Resolved): make ceph.com repos mirrorable
- 08:01 AM Fix #5388: osd: localized reads (from replicas) reordered wrt writes
- Since disabling localized reads I've not seen the problem occur, so thanks :)
- 07:32 AM Bug #5401: cuttlefish osd recovery slow
- Full backtrace (while recovering):
http://pastebin.com/raw.php?i=DWGHiNP6
2nd full backtrace:
http://pastebin.co... - 07:18 AM Bug #5401: cuttlefish osd recovery slow
- Not sure if this helps:
# /etc/init.d/ceph stop osd.24; sleep 15; /etc/init.d/ceph start osd.24; sleep 10; inotif... - 06:59 AM Bug #5401: cuttlefish osd recovery slow
- Some more information. While recoverig i see nearly no CPU load. If i look at the disk activity i see a HUGE amount o...
- 06:53 AM Bug #5401: cuttlefish osd recovery slow
- Lowering osd recovery max active makes it even more worth as the over all recovery takes longer. So it's not the I/O ...
- 01:01 AM Bug #5401 (Can't reproduce): cuttlefish osd recovery slow
- While the peering is fine now (Bug #5232) (latest upstream/cuttlefish) even without wip_cuttlefish_compact_on_startup...
- 06:59 AM Bug #5292: mon: monitor crashing due to not being in the monmap (no monmap to be in)
- Thanks.
- 06:56 AM Bug #5292 (Resolved): mon: monitor crashing due to not being in the monmap (no monmap to be in)
- You hit #5205 -- not the same issue, thus closing this ticket again.
- 06:48 AM Bug #5292: mon: monitor crashing due to not being in the monmap (no monmap to be in)
- Here you go. Please let me know if you need anything else.
Jeff - 06:42 AM Bug #5292 (Need More Info): mon: monitor crashing due to not being in the monmap (no monmap to be...
- Okay, can you post the monitor's logs with 'debug mon = 20' ?
- 02:46 AM Bug #5292: mon: monitor crashing due to not being in the monmap (no monmap to be in)
- I did a reboot, just to make sure :-(
# ceph -v
ceph version 0.61.4 (1669132fcfc27d0c0b5e5bb93ade59d147e23404)
... - 06:48 AM rgw Bug #5402 (Resolved): rgw compilation problem on wip-rgw-geo-2 branch
- The wip-rgw-geo-2 branch does not compile from a2cf14fe27a2da54e44b12a373b15b29c89d31b9.
In fact the method encode... - 03:02 AM Revision 1fccfd8b (ceph): Merge pull request #367 from ceph/wip-ceph-cli
- Reviewed-by: Dan Mick <dan.mick@inktank.com>
- 02:54 AM Revision 84444d0b (ceph): Merge remote-tracking branch 'upstream/next'
- 02:53 AM Revision c39b99cd (ceph): FileStore: handle observers in constructor/destructor
- Signed-off-by: Samuel Just <sam.just@inktank.com>
Reviewed-by: Sage Weil <sage@inktank.com> - 02:44 AM Revision 02115830 (ceph): init-radosgw: use radosgw --show-config-value to get config values
- This way we get the correct default values, as reflected by radosgw
itself.
Signed-off-by: Sage Weil <sage@inktank.com> - 02:44 AM Revision 716fc350 (ceph): ceph-conf: make --show-config-value reflect daemon defaults
- We want DAEMON defaults, but we don't want global_init to do anything else
daemonish like print a banner or mkdir /va... - 02:44 AM Revision 88f73c5a (ceph): ceph: fix ceph-conf call to get admin socket path for 'daemon <name> ...'
- Signed-off-by: Sage Weil <sage@inktank.com>
- 02:15 AM Revision cf3bc251 (ceph): FileStore: apply changes after disabling m_filestore_replica_fadvise
- Signed-off-by: Samuel Just <sam.just@inktank.com>
Reviewed-by: Dan Mick <dan.mick@inktank.com>
(cherry picked from co... - 02:14 AM Revision ed8b0e65 (ceph): FileStore: apply changes after disabling m_filestore_replica_fadvise
- Signed-off-by: Samuel Just <sam.just@inktank.com>
Reviewed-by: Dan Mick <dan.mick@inktank.com> - 12:58 AM Revision 2a4953b6 (ceph): ceph-disk: use unix lock instead of lockfile class
- The lockfile class relies on file system trickery to get safe mutual
exclusion. However, the unix syscalls do this f... - 12:55 AM Revision cb6bc95a (ceph): Merge remote-tracking branch 'gh/next'
- 12:06 AM Fix #5232: osd: slow peering due to pg log rewrites
- I'm fridudad ;-) Peering works fine but recovery does not. In my initial text of this tracker i also mentioned recove...
Also available in: Atom