Activity
From 11/21/2013 to 12/20/2013
12/20/2013
- 11:44 PM Revision 3510400d (ceph): compat: enable lseek64 alias
- Signed-off-by: Noah Watkins <noahwatkins@gmail.com>
- 11:44 PM Revision cf9569f3 (ceph): statfs: include headers for statfs structs
- Signed-off-by: Noah Watkins <noahwatkins@gmail.com>
- 11:43 PM Revision 9d72c4f8 (ceph): libcephfs: ignore missing offset64 definition
- on apple.
Signed-off-by: Noah Watkins <noahwatkins@gmail.com> - 11:20 PM Revision 7b98862d (ceph): Merge pull request #977 from ceph/wip-kill-raid4
- osd: remove remaining instances of raid4 pool types (never implemented)
Reviewed-by: Loic Dachary <loic@dachary.org> - 10:26 PM Revision 6afaf486 (ceph): Merge pull request #975 from BCLibCoop/bclibcoop/rgw_cors
- RGW: CORS use the correct headers for checking, and validate headers as lowercase where needed
Reviewed-by: Yehuda S... - 10:19 PM Revision d9299fd4 (ceph): Merge pull request #979 from dachary/wip-wrapped-vstart-errors
- vstart_wrapped_tests must fail if one test fail
Reviewed-by: Sage Weil <sage@inktank.com> - 09:50 PM Revision cf53d8a6 (ceph): rgw: abstract RGWProcess
- RGWProcess is a generic request handler that sends request to the
workqueue for processing.
Add RGWFCGXProcess specia... - 09:50 PM Revision d9a1ff38 (ceph): rgw: add some debug output
- Dump request environment when starting processing request. We used to do
it before, but it was recently dropped as pa... - 09:41 PM Revision 7bacc413 (ceph): Merge pull request #818 from ceph/wip-rgw-standalone-2
- Wip rgw standalone 2
Reviewed-by: Sage Weil <sage@inktank.com> - 09:14 PM Revision cdc178fb (ceph): Revert "Enable libs3 support for debian packages"
- This reverts commit 8814265f0888f8091a7d83a900ffd6b65ae77f34.
Or not! This adds a build-time dependency which none ... - 09:00 PM Revision 1e238e6f (ceph): mon: pool create will not fail if the type differs
- It looked like it worked because the wrapper hide the error. The failing
tests are commented out so that the other te... - 08:53 PM Revision 5f1957de (ceph): doc/release-notes: v0.67.5
- Signed-off-by: Sage Weil <sage@inktank.com>
- 08:48 PM Revision e13e4562 (ceph): Merge pull request #166 from ceph/wip-lockspell-wusui
- Fix spelling error in comment.
- 08:39 PM Revision 18107814 (ceph): Fix spelling error in teuthology/task/locktest.py comment
- 08:13 PM Revision 98a15253 (ceph): unittests: fail if one test fail
- vstart_wrapped_tests must return on error if one of the tests
fail.
Signed-off-by: Loic Dachary <loic@dachary.org> - 07:28 PM Revision a913ded2 (ceph): v0.72.2
- 06:48 PM Revision f86c3c07 (ceph): Merge pull request #460 from toabctl/build-depends
- Enable libs3 support for debian packages
Reviewed-by: Sage Weil <sage@inktank.com> - 06:29 PM Revision eb7ed58c (ceph): Merge pull request #916 from ceph/port/buffer
- buffer: use int64_t instead of loff_t
Reviewed-by: Sage Weil <sage@inktank.com> - 04:21 PM Revision 9ab947c7 (ceph): buffer: use int64_t instead of loff_t
- Because portability.
Signed-off-by: Noah Watkins <noahwatkins@gmail.com> - 04:01 PM Revision f8ce69cb (ceph): Add ability to mark jobs as 'dead'
- Signed-off-by: Zack Cerza <zack.cerza@inktank.com>
- 03:25 PM Bug #6797: ceph osd out does not migrate properly
- Arguably the most common use case is drive replacement, and Ceph documentation recommends setting noout flag when you...
- 02:52 PM Subtask #7048: mon: erasure crush rule vs pool process
- 02:23 PM Subtask #7048 (Resolved): mon: erasure crush rule vs pool process
- - "change 'rep' to 'replicated'":https://github.com/ceph/ceph/pull/981
- "create 'osd_pool_default_crush_rule_{repli... - 02:25 PM rgw Fix #7043: RGW CORS fixes for Access-Control-Request-Headers
- Merged with master, still pending on merge to dumpling, emperor.
- 02:17 PM Feature #6435 (Rejected): EC: [link] review and test PGBackend
- 02:14 PM Feature #6231 (Resolved): buffer: cache crc in buffer::raw (or similar) to avoid recalculation fo...
- 02:03 PM Feature #6835 (In Progress): EC: ec pgs will need to be able to specify temp primaries other than...
- wip-temp-primary branch
- 02:02 PM Feature #6831 (In Progress): EC: Adapt ReplicatedPG read path to handle async reads
- 01:43 PM rgw Feature #6195 (Resolved): rgw: test full sync (with large object)
- 01:32 PM devops Feature #7047 (Resolved): rhel7: build process for rbd.ko, ceph.ko kernel modules
- 01:31 PM devops Feature #7046 (Resolved): rhel7: gitbuilder for rbd.ko, ceph.ko kernel modules
- 01:18 PM Bug #6685: osd/ReplicatedPG.cc: 8345: FAILED assert(0 == "erroneously present object")
- ubuntu@teuthology:/a/teuthology-2013-12-19_23:00:07-rados-next-distro-basic-plana/9508
- 01:10 PM Fix #6763 (Resolved): crushtool: don't warn so harshly when enabling CRUSH_TUNABLES
- 01:03 PM rgw Feature #6870 (Fix Under Review): rgw: user level quota: new user objclass
- 12:42 PM Fix #6977: objecter: ENOSPC returned by pool quota will cause IO errors for non-cached rbd images
- One way to fix this is to treat it similar to the full flag, by putting an overquota flag on a pool in the osdmap, an...
- 11:24 AM Revision e04d7b87 (ceph): Merge pull request #933 from dachary/wip-erasure-code-benchmark
- osd: erasure code benchmark tool
Reviewed-by: Andreas Peters <andreas.joachim.peters@cern.ch>
Reviewed-by: Christoph... - 11:15 AM Revision 81dee1b6 (ceph): osd: erasure code benchmark workunit
- Display benchmark results for the default erasure code plugins, in a tab
separated CSV file. The first two column con... - 11:15 AM Revision 42b4fe14 (ceph): osd: erasure code benchmark is installed is part of ceph-test
- Add to the packaging for RPMs and DEBs
Signed-off-by: Loic Dachary <loic@dachary.org> - 11:15 AM Revision 03693ace (ceph): osd: git ignore erasure code benchmark binary
- Signed-off-by: Loic Dachary <loic@dachary.org>
- 11:15 AM Revision a619fe98 (ceph): osd: erasure code benchmark tool
- Implement the ceph_erasure_code_benchmark utility to:
* load an erasure code plugin
* loop over the encode/decode f... - 10:28 AM Revision ff9455ba (ceph): osd: conditionally disable dlclose of erasure code plugins
- When profiling, tools such as valgrind --tool=callgrind require that the
dynamically loaded libraries are not dlclose... - 10:28 AM Revision c7d8ba7b (ceph): osd: better performances for the erasure code example
- The XOR based example is ten times slower than it could because it uses
the buffer::ptr[] operator. Use a temporary c... - 10:28 AM Revision a36bc5fa (ceph): osd: set erasure code packet size default to 2048
- As shown in
https://www.usenix.org/legacy/events/fast09/tech/full_papers/plank/plank_html/
under "Impact of the Packe... - 09:50 AM rbd Bug #6174 (Can't reproduce): osdc/ObjectCacher.cc: 526: FAILED assert(i->empty()) on cuttlefish f...
- haven't seen this on recent code; won't fix cuttlefish at this point.
- 09:49 AM rbd Fix #6079 (Resolved): libceph: osd_client does not handle PAUSERD or PAUSEWR or FULL flags in osdmap
- 09:49 AM rbd Fix #5048 (Resolved): krbd: limit of ~230 mapped images at once
- 05:47 AM Revision 018164ad (ceph): rados/thrash: add small objects ceph_test_rados workload
- With small objects we can generate many more ops more quickly and stress
the cluster a bit differently.
Signed-off-b... - 05:24 AM Revision 8879e439 (ceph): osd: Fix assert which doesn't apply when compat_mode on
- Signed-off-by: David Zafman <david.zafman@inktank.com>
Reviewed-by: Samuel Just <sam.just@inktank.com>
(cherry picked... - 05:23 AM Revision 0bd5cb65 (ceph): Add backward comptible acting set until all OSDs updated
- Add configuration variable to override compatible acting set handling.
Later we'll check the osdmap that all OSDs are... - 03:47 AM Documentation #4117: Document various language bindings better
- Here's another, maintained ruby binding: https://github.com/netskin/ceph-ruby
- 01:58 AM CephFS Bug #6914: Accessing specific file on CephFS causes client hang
- Actually waiting for this to be sorted out takes a while. So to be able to make use of cephfs I did the following:
... - 12:40 AM Revision 8d31f71b (ceph): osd/ReplicatedPG: fix promote cancellation
- The canceling caller cleans up the blocked objects for us; we simply need
to bail out early.
Signed-off-by: Sage Wei... - 12:40 AM Revision 14037983 (ceph): osd/ReplicatedPG: EBUSY on cache-evict when watchers are present
- Linger operations will follow the object to the cache pool when the pool
overlay process is set. If we evict the obj... - 12:40 AM Revision 028bb0d1 (ceph): osd/osd_types: include num_objects_dirty, num_whiteouts in object_stat_...
- Signed-off-by: Sage Weil <sage@inktank.com>
- 12:40 AM Revision c561d5ea (ceph): osd/ReplicatedPG: track dirty, whiteout stat counts
- These counts will be useful (even necessary!) for the cache agent, and are
generally interesting to the admin as well... - 12:40 AM Revision bc893f55 (ceph): osd/ReplicatedPG: fix undirty on clean object
- Return success, but do not screw up the stats.
Signed-off-by: Sage Weil <sage@inktank.com> - 12:40 AM Revision 2a9c6fcd (ceph): vstart.sh: go faster
- Signed-off-by: Sage Weil <sage@inktank.com>
- 12:40 AM Revision 923bff1d (ceph): osd/ReplicatedPG: drop RepGather::ondone callback
- We kick the blocked contexts in the completion path of process_copy_chunk(),
after we have take the RWWRITE obc lock.... - 12:40 AM Revision 2fda4c01 (ceph): osd/ReplicatedPG: fix locking for promote
- After we get the copy-from data and unblock the obc, we still need to take
the RWWRITE lock on the object for the dur... - 12:40 AM Revision 0d2d6a5f (ceph): osd/ReplicatedPG: debug: add an assert for copy-get
- Signed-off-by: Sage Weil <sage@inktank.com>
- 12:40 AM Revision 1af6723c (ceph): osd/ReplicatedPG: debug: clean up oi printout
- Signed-off-by: Sage Weil <sage@inktank.com>
- 12:40 AM Revision 99cee55c (ceph): osd/osd_types: debug: include size in object_info_t operator<<
- Signed-off-by: Sage Weil <sage@inktank.com>
- 12:40 AM Revision 1dcbb663 (ceph): ceph_test_rados_api_tier: fix HitSet* test names
- Signed-off-by: Sage Weil <sage@inktank.com>
- 12:40 AM Revision 9ed6679a (ceph): ceph_test_rados: test cache_flush, cache_try_flush, cache_evict
- Signed-off-by: Sage Weil <sage@inktank.com>
- 12:39 AM Revision ffdaa5f4 (ceph): vstart.sh: --cache <pool> to set up pool cache(s) on startup
- Signed-off-by: Sage Weil <sage@inktank.com>
- 12:39 AM Revision ba2f9e29 (ceph): osd/osd_types: include user_version in operator<< object_info_t
- Signed-off-by: Sage Weil <sage@inktank.com>
- 12:39 AM Revision 54f0c60c (ceph): osd/osd_types: make object_info_t::dump() dump user_version
- Backport: emperor
Signed-off-by: Sage Weil <sage@inktank.com> - 12:39 AM Revision 2d5a7e2c (ceph): osd/ReplicatedPG: uninline CopyFromCallback, PromoteCallback
- Signed-off-by: Sage Weil <sage@inktank.com>
- 12:39 AM Revision bc051041 (ceph): osd/ReplicatedPG: handle ECANCELED in C_CopyFrom, C_Flush
- Signed-off-by: Sage Weil <sage@inktank.com>
- 12:39 AM Revision 927b0e60 (ceph): osd/ReplicatedPG: fix user_version preservation for copy_from
- In the process of fixing this for flush, we break promote, so we need to
adjust them both here. Basic strategy: do n... - 12:39 AM Revision ad3b4666 (ceph): osd/ReplicatedPG: implement cache-flush, cache-try-flush
- Implement a rados operation that will flush a dirty object in the cache
tier by writing it back to the base tier.
Si... - 12:39 AM Revision 71cd4a22 (ceph): rados: add cache-flush, cache-evict, cache-flush-evict-all commands
- Signed-off-by: Sage Weil <sage@inktank.com>
- 12:39 AM Revision ea519b48 (ceph): qa/workunits/rados: test cache-{flush,evict,flush-evict-all}
- Signed-off-by: Sage Weil <sage@inktank.com>
- 12:39 AM Revision 1bde88f8 (ceph): qa/workunits/rados: rename cache pool tests
- Signed-off-by: Sage Weil <sage@inktank.com>
- 12:39 AM Revision 57e91455 (ceph): qa/workunits/rados/test_cache_pool.sh: fixes
- Signed-off-by: Sage Weil <sage@inktank.com>
12/19/2013
- 10:43 PM Revision e312048a (ceph): Allow passing multiple job_ids
- Signed-off-by: Zack Cerza <zack.cerza@inktank.com>
- 10:37 PM Revision edaec9a8 (ceph): osd: Fix assert which doesn't apply when compat_mode on
- Signed-off-by: David Zafman <david.zafman@inktank.com>
Reviewed-by: Samuel Just <sam.just@inktank.com> - 10:36 PM CephFS Bug #5805 (Can't reproduce): mds: crash loading sessionmap
- probably fixed by commit 40613b700b
- 10:32 PM Revision eeeb6267 (ceph): For teuthology-kill, s/suite/run/
- Signed-off-by: Zack Cerza <zack.cerza@inktank.com>
- 10:32 PM Revision 220779c8 (ceph): Implement single-job killing
- Signed-off-by: Zack Cerza <zack.cerza@inktank.com>
- 10:27 PM Revision 228358ff (ceph): Merge pull request #165 from ceph/wip-7042-fix-wusui
- Do not run local handling fix if local parameter is not found.
- 10:20 PM Revision 37815b76 (ceph): Do not run local handling fix if local parameter is not found.
- Fixes: 7042
Signed-off-by: Warren Usui <warren.usui@inktank.com> - 10:10 PM Revision a60ac919 (ceph): v0.67.5
- 09:28 PM Revision ac16a9df (ceph): osd: remove remaining instances of raid4 pool types (never implemented)
- Signed-off-by: Sage Weil <sage@inktank.com>
- 06:17 PM Revision 30876e69 (ceph): Merge pull request #941 from ceph/wip-6028
- #6028 : ensure that erasure coded pools don't work until the osds can handle it
Reviewed-by: Sage Weil <sage@inktank... - 05:27 PM Revision dbcef31b (ceph): Merge pull request #156 from ceph/teuthology-doc-hadoop-wusui
- Added docstrings. Cleaned up code (broke up long lines, removed unused
- 05:24 PM Revision 006c0311 (ceph): Merge pull request #164 from ceph/wip-rados
- rados: add in more (optional) op types
- 05:23 PM Revision d70d1ad7 (ceph): Merge pull request #160 from ceph/wip-fix-5149-wusui
- Added handling of a 'local' option inside install.py which specifies
- 04:29 PM Revision b014c718 (ceph): Catch every exception here, for now.
- Signed-off-by: Zack Cerza <zack.cerza@inktank.com>
- 04:29 PM Revision 9a29c3ef (ceph): Log calls to teuthology-report more verbosely
- Signed-off-by: Zack Cerza <zack.cerza@inktank.com>
- 04:24 PM Bug #7044 (Can't reproduce): Segmentation fault rados suite on master
ubuntu@teuthology:/a/teuthology-2013-12-18_23:00:02-rados-master-testing-basic-plana/8649
{description: 'rados/t...- 04:14 PM Revision 0d2d937e (ceph): Merge pull request #973 from ceph/wip-mds-coverity2
- mds: fixes for coverity scan
Reviewed-by: Sage Weil <sage@inktank.com> - 04:09 PM Revision 40a48de6 (ceph): mds: fix Resetter locking
- ceph-mds --reset-journal didn't work; it would deadlock waiting for
the osdmap. Comparing the init code in the Dumpe... - 04:05 PM Revision 2cf9fa94 (ceph): Merge pull request #969 from yuyuyu101/wip-7040
- Fix segmentation fault when handler is NULL pointer
Reviewed-by: Sage Weil <sage@inktank.com> - 04:03 PM rgw Fix #7043 (Resolved): RGW CORS fixes for Access-Control-Request-Headers
- I was trying to use the RGW CORS support, and found that it flubbed handling of the Access-Control-Request-Headers he...
- 02:39 PM Revision eb12c81f (ceph): Merge pull request #972 from dachary/wip-build-depends
- packaging: revert adding argparse and uuidgen
- 02:33 PM Revision 087fe57f (ceph): packaging: revert adding argparse and uuidgen
- Because the gitbuilder build environment does not read the build depends
from the packages and needs to be updated in... - 02:27 PM Revision 4f0918af (ceph): Merge pull request #971 from dachary/wip-build-depends
- packaging: make check needs argparse and uuidgen
Reviewed-by: Christophe Courtaut <christophe.courtaut@gmail.com> - 01:57 PM Revision 812e59b2 (ceph): Merge pull request #970 from dachary/wip-autogen
- autogen: test compare strings with != not -ne
Reviewed-by: Christophe Courtaut <christophe.courtaut@gmail.com> - 12:57 PM Revision 82725385 (ceph): packaging: make check needs argparse and uuidgen
- make check runs vstart.sh to setup a cluster from source and it misses
the python-argparse library as well as uuidgen... - 12:40 PM Revision f193925a (ceph): autogen: test compare strings with != not -ne
- Signed-off-by: Loic Dachary <loic@dachary.org>
- 10:54 AM Revision 98af37d0 (ceph): Fix segmentation fault when handler is NULL pointer
- Signed-off-by: Haomai Wang <haomaiwang@gmail.com>
- 10:51 AM devops Bug #7036 (Resolved): qemu-img package not found
- This is fixed and tested for CentOS/RHEL in ceph-qa-chef commit:
bd5d62496b26f3cc1ad67e1e5c575f01e32587e7
- 10:31 AM Messengers Feature #6411 (In Progress): msgr: move iov + sendmsg into bufferlist::sendmsg method
- 10:31 AM Revision b2a80004 (ceph): Merge pull request #966 from dachary/ceph-master
- crush: silence error messages in unit tests
Reviewed-by: Joao Eduardo Luis <joao.luis@inktank.com> - 10:28 AM Revision 27f4d1f6 (ceph): Merge pull request #968 from ceph/wip-crush
- make check is happy and the change is non controversial
Reviewed-by: Loic Dachary <loic@dachary.org> - 10:15 AM Feature #6028 (Resolved): EC: [link] ensure that erasure coded pools don't work until the osds ca...
- 09:32 AM Bug #7040 (Resolved): MemStore::queue_transactions access NULL point bug
- 02:51 AM Bug #7040 (Resolved): MemStore::queue_transactions access NULL point bug
- see src/os/memstore.cc:651
When commit b8884e01a05fb65c67e643acd60653598960cd78 introduce TPhandle argument but no... - 09:26 AM rgw Bug #6899 (Resolved): rgw: delete bucket on secondary zone returns success, but bucket is not del...
- cherry-picked in commit:8cd33e3a8ebf7c2aa796ec9f92d6b554c39ff705
- 09:23 AM rgw Feature #6513: rgw: dr: Service scripts for meta/data sync agents
- radosgw-agent needs a config file to start up. The script is testing for the presecence of a default named config fi...
- 09:07 AM Linux kernel client Feature #294 (Resolved): honor PAUSE* flags in osdmap
- 09:07 AM Linux kernel client Feature #27 (Resolved): ACLs
- 09:07 AM Linux kernel client Feature #591 (Resolved): implement FALLOC_FL_PUNCH_HOLE
- 09:02 AM Linux kernel client Cleanup #2130 (Rejected): ceph: xattr: complete cleanups following review
- 09:01 AM Linux kernel client Feature #3282 (Resolved): add new 'performance' kernel gitbuilder
- 06:27 AM Revision 44aacaed (ceph): mds: fixes for coverity scan
- Signed-off-by: Yan, Zheng <zheng.z.yan@intel.com>
- 05:22 AM Revision 0497db49 (ceph): crush/mapper: finish adding choose_local_[fallback_]tries
- Didn't finish this in d129e09e57fbc61cfd4f492e3ee77d0750c9d292.
Signed-off-by: Sage Weil <sage@inktank.com> - 05:22 AM Revision ea3a0bb8 (ceph): crush/mapper: attempts -> tries
- Signed-off-by: Sage Weil <sage@inktank.com>
- 05:22 AM Revision 8b38f10b (ceph): crush/mapper: fix crush_choose_firstn comment
- Signed-off-by: Sage Weil <sage@inktank.com>
- 05:22 AM Revision 99f41de3 (ceph): vstart.sh: NOTE, not WARNING, to make gitbuilder happy
- Signed-off-by: Sage Weil <sage@inktank.com>
- 02:21 AM Revision bdeaa841 (ceph): osd: OSDMap: add 'get_up_osds()' function
- Signed-off-by: Joao Eduardo Luis <joao.luis@inktank.com>
- 02:21 AM Revision c996f669 (ceph): mon: OSDMonitor: add optional 'pool type' arg to 'osd pool create'
- Allow specifying 'rep', 'raid4' and 'erasure'.
Only allow setting type 'erasure' if all up osds support erasure codes... - 02:21 AM Revision 250b4462 (ceph): pybind: test_ceph_argparse: test 'ceph osd pool create' with pool type
- Signed-off-by: Joao Eduardo Luis <joao.luis@inktank.com>
- 02:21 AM Revision b3ee598f (ceph): qa: workunit: cephtool: test osd pool create with erasure type
- Signed-off-by: Joao Eduardo Luis <joao.luis@inktank.com>
- 02:20 AM Revision 5756c05a (ceph): ceph_osd: add EC to OSD's supported features used by the messenger
- Fixes: 6028
Signed-off-by: Joao Eduardo Luis <joao.luis@inktank.com> - 02:19 AM Revision 59ad6da4 (ceph): osd: OSD: reflect OSDMap EC flag being set by setting on-disk feature
- If OSDMap has the EC feature set, then update our superblock to
reflect as such, making our on-disk format incompatib... - 02:19 AM Revision 8ac84db0 (ceph): mon: OSDMonitor: handle osd features on boot
- Add the osd's features to the osd's extra info field in the OSDMap
so we can track which OSDs are able to deal with E... - 02:14 AM Revision 73992d2e (ceph): osd: OSDMap: check for erasure pools when getting features
- If we have a pool with type Erasure, then we consider we require
CEPH_FEATURE_OSD_ERASURE_CODES.
Signed-off-by: Joao... - 02:14 AM Revision 31743d50 (ceph): osd: OSDMap: add 'features' bit mask field to osd_xinfo_t
- And make sure we dump it on osd_xinfo_t::dump().
Signed-off-by: Joao Eduardo Luis <joao.luis@inktank.com> - 02:14 AM Revision bfc86a82 (ceph): include/ceph_features: add CEPH_FEATURES_OSD_ERASURE_CODES
- Signed-off-by: Joao Eduardo Luis <joao.luis@inktank.com>
- 02:14 AM Revision 178f6841 (ceph): osd: OSD: add binary compat feature for Erasure Codes
- Signed-off-by: Joao Eduardo Luis <joao.luis@inktank.com>
- 01:06 AM Revision 668092fb (ceph): Merge branch 'wip-dzfix'
- 01:03 AM CephFS Bug #6914: Accessing specific file on CephFS causes client hang
- searching for the nodes was done like this:
[rolf@ceph-mon01 ~]$ bzip2 -d < /tmp/mds-20131219.dump.bz2 | grep 1000... - 12:59 AM CephFS Bug #6914: Accessing specific file on CephFS causes client hang
- Also note the MDS log:
2013-12-18 17:03:43.775909 7f5e99e39700 1 mds.0.62 active_start
2013-12-18 17:03:43.78239... - 12:59 AM Revision 20b79982 (ceph): Increase timeout by 50% in thrashers/mapgap.yaml
- Seen in /a/teuthology-2013-12-17_23:00:03-rados-next-distro-basic-plana/7188
Signed-off-by: David Zafman <david.zafm... - 12:52 AM Revision 5d8795d9 (ceph): moved samba suite a level up [suites/fs/samba to suites/samba]
- Signed-off-by: tamil <tamil.muthamizhan@inktank.com>
12/18/2013
- 10:32 PM Revision fe136845 (ceph): added execute permission to the script
- Signed-off-by: tamil <tamil.muthamizhan@inktank.com>
- 10:23 PM Revision d9e33ea7 (ceph): rgw workunit to test bucket quota
- Signed-off-by: tamil <tamil.muthamizhan@inktank.com>
- 10:22 PM Revision 085ad889 (ceph): task to test rgw bucket quota
- Signed-off-by: tamil <tamil.muthamizhan@inktank.com>
- 09:12 PM Revision 7587ee52 (ceph): rgw: don't return data within the librados cb
- Fixes: #7030
The callback is running within a single Finisher thread, thus we
shouldn't block there. Append read data... - 09:12 PM Revision 8745ba63 (ceph): rgw: fix use-after-free when releasing completion handle
- Backport: emperor, dumpling
Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
Reviewed-by: Sage Weil <sage@inktank.com... - 09:11 PM Revision 451381b4 (ceph): rgw: don't return data within the librados cb
- Fixes: #7030
The callback is running within a single Finisher thread, thus we
shouldn't block there. Append read data... - 09:11 PM Revision b570c801 (ceph): rgw: fix use-after-free when releasing completion handle
- Backport: emperor, dumpling
Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
Reviewed-by: Sage Weil <sage@inktank.com... - 09:11 PM Revision c8890ab2 (ceph): rgw: fix use-after-free when releasing completion handle
- Backport: emperor, dumpling
Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
Reviewed-by: Sage Weil <sage@inktank.com> - 09:10 PM Revision d6a4f6ad (ceph): rgw: don't return data within the librados cb
- Fixes: #7030
The callback is running within a single Finisher thread, thus we
shouldn't block there. Append read data... - 08:38 PM Revision 031be568 (ceph): Use saucy gitbuilder for arm package checking.
- Some-how missed it checks both sha1 and package version file
and package version was still the quantal gitbuilder whi... - 07:41 PM Revision 5320db57 (ceph): rados: add in more (optional) op types
- Signed-off-by: Sage Weil <sage@inktank.com>
- 07:30 PM Revision e6ad4d4a (ceph): osd: make obc copyfrom blocking generic
- Signed-off-by: Sage Weil <sage@inktank.com>
- 07:26 PM Revision 8dec2b27 (ceph): librados, osd: add flags to COPY_FROM
- If we initiate a COPY_FROM as part of a FLUSH operation, we will need to
set a flag so that the read-side of the copy... - 06:42 PM Revision 25db0de1 (ceph): Merge pull request #926 from yuyuyu101/wip/6950
- Move PerfCounter from ObjectStore.h to FileStore.h
Reviewed-by: Sage Weil <sage@inktank.com> - 06:41 PM Revision e54489da (ceph): Merge pull request #961 from dachary/wip-erasure-code-alignment
- erasure-code: tests must use aligned buffers
Reviewed-by: Sage Weil <sage@inktank.com> - 06:24 PM Revision f32a36d2 (ceph): Merge pull request #967 from ceph/wip-7026
- osd/ReplicatedPG: fix hit_set_setup() on_activate()
Reviewed-by: Greg Farnum <greg@inktank.com> - 06:00 PM Revision 1bb46c0e (ceph): Merge pull request #955 from ceph/wip-crush-2
- crush: make set_chooseleaf_tries work with firstn chooseleaf, too
Reviewed-by: Samuel Just <sam.just@inktank.com>
Re... - 05:57 PM Revision 202d1f7b (ceph): Merge pull request #964 from apeters1971/wip-arch-sse2
- ARCH: add variable for sse2 register
Reviewed-by: Loic Dachary <loic@dachary.org> - 05:52 PM Revision e624e166 (ceph): crush: silence error messages in unit tests
- The error messages are intentional when error conditions are
created. They will create false positive in the gitbuild... - 05:43 PM Feature #7039: Check other OSD CEPH_FEATURE_OSD_ERASURE_CODES to enable acting set changes
Look through want and backfill vectors and check
osd_xinfo_t& xi = osdmap.get_xinfo(i);
- 05:38 PM Feature #7039 (Resolved): Check other OSD CEPH_FEATURE_OSD_ERASURE_CODES to enable acting set cha...
In PG::choose_acting() we need to test other OSDs CEPH_FEATURE_OSD_ERASURE_CODES bit in osdmap in order to determin...- 05:06 PM Bug #6905 (Duplicate): nightlies: failed to become clean before timeout expired
- 04:24 PM Bug #6905: nightlies: failed to become clean before timeout expired
My previous comment indicating that a clean timeout occurred while recovery was still gong on applied to /a/teuthol...- 04:14 PM Bug #6905: nightlies: failed to become clean before timeout expired
- David Zafman wrote:
> I've run the particular yaml file with increased OSD debugging multiple times. It does not re... - 04:04 PM Bug #6905: nightlies: failed to become clean before timeout expired
I've run the particular yaml file with increased OSD debugging multiple times. It does not reproduce. I do notice...- 09:34 AM Bug #6905: nightlies: failed to become clean before timeout expired
- /a/teuthology-2013-12-17_23:00:03-rados-next-distro-basic-plana/7188
- 05:05 PM Revision 94149703 (ceph): ARCH: adding SSE2 flag to arch-test
- 04:57 PM Revision fd5f4026 (ceph): Merge pull request #965 from ksperis/rbdmap.upstart
- upstart: add rbdmap script
Reviewed-by: Sage Weil <sage@inktank.com> - 04:35 PM Revision 88365c29 (ceph): crush: expand info about tunables that we dump
- Include the tunables profile, and flags indicating whether it is optimal,
legacy, or requires certain features.
Sign... - 04:35 PM Revision 7e0c84b8 (ceph): mon/OSDMonitor: 'osd crush show-tunables'
- Signed-off-by: Sage Weil <sage@inktank.com>
- 04:35 PM Revision d129e09e (ceph): crush: add set_choose_local_[fallback_]tries steps
- This alls all of the tunables to be overridden by a specific rule.
Signed-off-by: Sage Weil <sage@inktank.com> - 04:35 PM Revision d0f14dfa (ceph): mon: warn if crush has non-optimal tunables
- Allow warning to be disabled via ceph.conf. Link to the docs from the
warning detail. Add a section to the docs spe... - 03:00 PM Revision d76188d7 (ceph): Merge pull request #952 from kri5/master
- vstart: Update apache conf to run against apache 2.4
Reviewed-by: Loic Dachary <loic@dachary.org> - 02:54 PM Revision 30f8aa1d (ceph): vstart: Update apache conf to run against apache 2.4
- Signed-off-by: Christophe Courtaut <christophe.courtaut@gmail.com>
- 02:53 PM devops Bug #7036 (In Progress): qemu-img package not found
- See Bug #7004. I removed the old version of qemu-kvm. The new packages are:
qemu-guest-agent-0.12.1.2-2.41... - 12:34 PM devops Bug #7036 (Resolved): qemu-img package not found
- All ceph-deploy tests that depend on this package are failing because that package is not found.
Example log outpu... - 02:22 PM Linux kernel client Bug #6984: RBD volume not mountable after creating 8 or more snapshots
- Apparently this is because the kernel client can't tolerate arbitrary-length replies from the OSD to different kinds ...
- 01:55 PM CephFS Bug #6914: Accessing specific file on CephFS causes client hang
- Can you upload the cache dump? How did you search through it for those inodes? The only way I can think of to make a ...
- 08:42 AM CephFS Bug #6914: Accessing specific file on CephFS causes client hang
- Some simularity with http://tracker.ceph.com/issues/6087? We use the Emperor release so that should be resolved.
- 04:50 AM CephFS Bug #6914: Accessing specific file on CephFS causes client hang
- After forcing ceph to use the original mds again by stopping the new mds and using ceph -w we see:
2013-12-18 13:37:... - 04:18 AM CephFS Bug #6914: Accessing specific file on CephFS causes client hang
- We added an extra mds on another host and turned the primary mds off. It took a while (8 minutes I think) but after t...
- 01:20 PM Revision b86d450a (ceph): upstart: add rbdmap script
- Upstart script for mapping / unmapping rbd device based on /etc/ceph/rbdmap file.
It does not mount or unmount filesy... - 01:13 PM rgw Bug #7030 (Resolved): rgw: can't read objects > 512k
- 10:28 AM rgw Bug #7030 (Resolved): rgw: can't read objects > 512k
- Deadlock on the fastcgi socket when with high load when trying to read objects > 512k. Another symptom is having slow...
- 12:31 PM rbd Bug #6480: librbd crashed qemu-system-x86_64
- Igor Lukyanov wrote:
> They say there was a bug in malloc in libc 2.15 (we have the same major version)
>
> https... - 12:30 PM rbd Bug #6480: librbd crashed qemu-system-x86_64
- They say there was a bug in malloc in libc 2.15 (we have the same major version)
https://jira.mongodb.org/browse/S... - 12:21 PM rbd Bug #6480: librbd crashed qemu-system-x86_64
- In addition to comment posted by Andrey:
The disk workload of the VM was very peaky: http://imgur.com/SLFdJvv
VM cr... - 11:36 AM rbd Bug #6480: librbd crashed qemu-system-x86_64
- The same on cuttlefish. Attached graphs are for the problematic VM for different periods.
- 11:06 AM rgw Bug #7031 (Rejected): rgw multiregion test fails under valgrind
- 10:58 AM rgw Bug #7031 (Rejected): rgw multiregion test fails under valgrind
- see e.g. http://pulpito.ceph.com/compare/?branch=master&suite=rgw
- 11:05 AM Feature #7035 (Resolved): cachepool: agent: blocking eviction (when cache pool is 'full')
- 11:05 AM Feature #7034 (Resolved): cachepool: agent: basic eviction
- 11:04 AM Feature #7033 (Resolved): cachepool: agent: trivial flushing
- 11:03 AM Fix #6990 (Resolved): osd crash when running mixed versions of dumpling and master
- 10:23 AM Bug #7026 (Resolved): LibRadosTier.HitSetWrite failure with thrashing
- Looks good, merged to master. commit:7e4a8004880ffecd89edbfc9c62c4c00605aaf13
- 09:13 AM Bug #7026 (Fix Under Review): LibRadosTier.HitSetWrite failure with thrashing
- 10:16 AM Revision e4537d31 (ceph): ARCH: add variable for sse2 register
- 09:32 AM Bug #7018 (Duplicate): osd/ReplicatedPG.cc: 1439: FAILED assert(is_replica())
- #6990
- 09:32 AM Bug #7019 (Duplicate): osd/PG.cc: 4517: FAILED assert(!flushed)
- #6990
- 07:31 AM devops Bug #6962 (Resolved): libleveldb1 fails to install for cuttlefish testing basic plana test
- This is no longer an issue. Probably a single machine was stuck with this and cause this to come up again. Closing fo...
- 02:18 AM Revision 0d217cf9 (ceph): qa/workunits/cephtool/test.sh: clean up our client.xx.keyring
- Signed-off-by: Sage Weil <sage@inktank.com>
- 02:00 AM CephFS Bug #6930: Folders report "Not a folder" in Kernel 3.13-rc1 and -rc2
- After compiling and testing current kernel from linus git, I can confirm that the changes merged by linus yesterday f...
- 12:12 AM CephFS Bug #6930 (Resolved): Folders report "Not a folder" in Kernel 3.13-rc1 and -rc2
- 12:12 AM CephFS Bug #6930: Folders report "Not a folder" in Kernel 3.13-rc1 and -rc2
- this went into linus' tree yesterday, commit:a5905a92
12/17/2013
- 11:58 PM CephFS Bug #6930: Folders report "Not a folder" in Kernel 3.13-rc1 and -rc2
- Found a hint that this error might not be directly related to the latest ceph-client merge into linux mainline, as Ma...
- 11:47 PM CephFS Bug #6930: Folders report "Not a folder" in Kernel 3.13-rc1 and -rc2
- I'm encountering the very same issue when trying to mount cephfs (0.72.1) from a kernel-client using linus' plain lin...
- 11:58 PM Revision 7e4a8004 (ceph): osd/ReplicatedPG: fix hit_set_setup() on_activate()
- Didn't read that if carefully the first time around!
Signed-off-by: Sage Weil <sage@inktank.com> - 07:46 PM Revision 434dce1f (ceph): Merge pull request #960 from ceph/wip-6990
- Add backward comptible acting set until all OSDs updated
Reviewed-by: Samuel Just <sam.just@inktank.com> - 07:43 PM Revision 19cff890 (ceph): Add backward comptible acting set until all OSDs updated
- Add configuration variable to override compatible acting set handling.
Later we'll check the osdmap that all OSDs are... - 07:26 PM Revision b1530679 (ceph): erasure-code: tests must use aligned buffers
- The underlying code assumes the memory buffer is aligned on a long
boundary which is not always the case. Using buffe... - 06:46 PM Revision 2e5a461e (ceph): Merge pull request #953 from dachary/wip-qa-suite
- use qa/workunits/cephtool/test.sh as a unittest
Reviewed-by: Sage Weil <sage@inktank.com> - 06:25 PM Revision 0680a136 (ceph): bump the OSDs to three per mon/host
- Signed-off-by: Alfredo Deza <alfredo.deza@inktank.com>
(cherry picked from commit 8e543213bdafbec557c1bf5c962d87e49d6... - 05:03 PM Revision f5d32a33 (ceph): mds: drop unused find_ino_dir
- Remove all traces of find_ino_dir, it is no longer used.
Signed-off-by: Alexandre Oliva <oliva@gnu.org> - 05:02 PM Revision c60a3644 (ceph): Fix typo in #undef in ceph-dencoder
- Signed-off-by: Alexandre Oliva <oliva@gnu.org>
Signed-off-by: Sage Weil <sage@inktank.com> - 04:53 PM Revision 9e456555 (ceph): qa: add ../qa/workunits/cephtool/test.sh to unittests
- Signed-off-by: Loic Dachary <loic@dachary.org>
- 04:51 PM Revision 526e2528 (ceph): Merge pull request #957 from ceph/wip-rbd-coverity
- rbd: make coverity happy
Reviewed-by: Sage Weil <sage@inktank.com> - 04:46 PM Revision c1eb55c6 (ceph): qa: vstart wrapper helper for unittests
- Primarily useful to run scripts from qa/workunits as part of make check.
vstart_wrapper.sh starts a vstart.sh cluste... - 03:42 PM Revision 0edbda20 (ceph): rbd: make coverity happy
- A recent coverity run found two "defects" in rbd.cc:
** CID 1138367: Time of check time of use (TOCTOU)
/rbd.cc: 2... - 03:10 PM Bug #7026 (Resolved): LibRadosTier.HitSetWrite failure with thrashing
- 12:48 PM Revision d93881f3 (ceph): vstart/stop: use pkill instead of killall
- killall fails to kill all OSDs when called as a oneliner. Replace with a
loop using pkill that retries until there ar... - 12:47 PM Revision ae56cef3 (ceph): qa: recursively remove .gcno and .gcda
- Instead of removing them only in the current directory. Leftovers
prevent running make check-coverage properly becaus... - 09:43 AM Bug #6803 (Can't reproduce): rados test failing in the nightlies on next branch
- 09:42 AM Bug #6982 (Duplicate): osd crashed when running mixed versions of dumpling and master
- #7019
- 06:05 AM devops Bug #5283 (New): Ceph-deploy can't handle /dev/disk/by-* device paths
- 05:46 AM Revision b082c094 (ceph): crushtool: reorg test-map-* cli tests
- Signed-off-by: Sage Weil <sage@inktank.com>
- 05:35 AM Revision 41d5186f (ceph): crush/CrushWrapper: minor cleanup on tunables helpers
- Signed-off-by: Sage Weil <sage@inktank.com>
- 05:33 AM Revision 685c6950 (ceph): crush/mapper: generalize descend_once
- The legacy behavior is to make the normal number of tries for the
recursive chooseleaf call. The descend_once tunabl... - 01:10 AM Revision 6f431200 (ceph): ceph_test_rados_api_tier: fix HitSetTrim vs split, too
- Signed-off-by: Sage Weil <sage@inktank.com>
- 01:03 AM Revision 00f436c1 (ceph): Merge pull request #904 from ceph/wip-mds-cluster2
- Wip mds cluster2
Reviewed-by: Sage Weil <sage@inktank.com> - 12:52 AM Revision c5bccfef (ceph): ceph_test_rados_api_tier: fix HitSetRead test race with split
- Recalculate the hash on each iteration in case we are racing with split.
Fixes: #7013
Signed-off-by: Sage Weil <sage... - 12:32 AM devops Bug #6962: libleveldb1 fails to install for cuttlefish testing basic plana test
- I verified that leveldb was in the precise cuttlefish repo. Is this problem still happening ?
- 12:31 AM Revision 94da54ff (ceph): Merge pull request #954 from ceph/wip-7009
- mon: move supported_commands fields, methods into Monitor, and fix leak
Reviewed-by: Greg Farnum <greg@inktank.com> - 12:09 AM Revision 7e618c93 (ceph): mon: move supported_commands fields, methods into Monitor, and fix leak
- We were leaking the static leader_supported_mon_commands. Move this into
the class so that we can clean up in the de... - 12:08 AM devops Bug #7004 (In Progress): qemu: rhel and centos qemu packages should depend on librbd
- I'm guessing that part of the problem is that the previous backports of the qemu-kvm packages were still in the ceph ...
12/16/2013
- 11:11 PM rgw Feature #6513 (In Progress): rgw: dr: Service scripts for meta/data sync agents
- 09:40 PM Bug #7019 (Duplicate): osd/PG.cc: 4517: FAILED assert(!flushed)
- ...
- 09:38 PM Bug #7018: osd/ReplicatedPG.cc: 1439: FAILED assert(is_replica())
- ubuntu@teuthology:/a/teuthology-2013-12-16_10:06:55-upgrade:upgrade-parallel-next-testing-basic-plana/5079
ubuntu@te... - 09:36 PM Bug #7018 (Duplicate): osd/ReplicatedPG.cc: 1439: FAILED assert(is_replica())
- ...
- 08:27 PM Revision deded446 (ceph): mongoose: update submodule
- switched to a different source repository, added .gitignore
Signed-off-by: Yehuda Sadeh <yehuda@inktank.com> - 08:22 PM Revision a0eb1a8e (ceph): Use shell=True to call teuthology-report
- Signed-off-by: Zack Cerza <zack.cerza@inktank.com>
- 07:34 PM Revision c22ee528 (ceph): Catch OSError if script isn't in $PATH
- Signed-off-by: Zack Cerza <zack.cerza@inktank.com>
- 06:27 PM Revision ef10a5c2 (ceph): rgw: fix memory leak
- Remove frontend object on shutdown.
Signed-off-by: Yehuda Sadeh <yehuda@inktank.com> - 05:43 PM Revision 420fff62 (ceph): Revert "Use path when calling teuthology-report. …"
- This reverts commit e4b5ab811e954a5b134d413aeb338805b5e3441d.
- 05:27 PM Revision 1597d4e9 (ceph): Merge pull request #951 from ceph/wip-linux-version
- common: introduce get_linux_version()
Reviewed-by: Sage Weil <sage@inktank.com> - 05:18 PM CephFS Bug #6613: samba is crashing in teuthology
- teuthology claims smbd crashed with signal SIGTERM after sending SIGTERM to smbd. No error in the log, no coredump fi...
- 04:57 PM Revision 6696ab64 (ceph): FileJournal: switch to get_linux_version()
- For the purposes of FileJournal::_check_disk_write_cache(), use
get_linux_version(), which is based on uname(2), inst... - 04:57 PM Revision 824b3d8e (ceph): FileJournal: use pclose() to close a popen() stream
- In FileJournal::_check_disk_write_cache(), use pclose() instead of
fclose() to close a stream, created by popen().
S... - 04:57 PM Revision a2babe27 (ceph): configure: break up AC_CHECK_HEADERS into one header-file per line
- Break up AC_CHECK_HEADERS macro into one header-file per line so it's
easier to read and make changes.
Signed-off-by... - 04:57 PM Revision fcf6e987 (ceph): common: introduce get_linux_version()
- get_linux_version() returns a version of the currently running kernel,
encoded as in int, and is contained in common/... - 04:53 PM Bug #7013 (Resolved): rados.sh failure (LibRadosMisc.HitSetTrim)
- 02:06 PM Bug #7013 (Resolved): rados.sh failure (LibRadosMisc.HitSetTrim)
- 2013-12-15T23:38:06.965 INFO:teuthology.orchestra.run.err:[10.214.132.28]: dumped all in format json
2013-12-15T23:3... - 04:30 PM Bug #7015 (Duplicate): valgrind issues in mon (I think based on the output below)
- #7009
- 02:24 PM Bug #7015 (Duplicate): valgrind issues in mon (I think based on the output below)
- 2013-12-16T01:18:19.009 INFO:teuthology.task.ceph:Checking for errors in any valgrind logs...
2013-12-16T01:18:19.01... - 04:30 PM Bug #7009 (Resolved): mon: leaking memory from elector mon commands
- 04:08 PM Bug #7009 (Fix Under Review): mon: leaking memory from elector mon commands
- 03:30 PM Bug #7009 (In Progress): mon: leaking memory from elector mon commands
- 02:25 PM Bug #7014: rados: stuck degraded, possibly related to acting_backfill changes
- Another option would be to reproduce with logging. If you catch it before it gets cleaned up, it should be pretty ob...
- 02:22 PM Bug #7014 (Can't reproduce): rados: stuck degraded, possibly related to acting_backfill changes
- End of ceph.log:
2013-12-15 23:51:19.781079 mon.0 10.214.131.3:6789/0 2397 : [INF] pgmap v1364: 213 pgs: 212 active+... - 11:31 AM Bug #6751: Pool 'df' statistics go bad after changing PG count
- I spoke with Sam about this when after it came in and the bug is an obvious result of doing fixed-cost pg splits: the...
- 11:26 AM CephFS Bug #7012 (Duplicate): smbd crash during cifs + dbench
- 08:34 AM CephFS Bug #7012 (Duplicate): smbd crash during cifs + dbench
- ...
- 09:56 AM rbd Bug #5186 (Won't Fix): krbd: mapping same image produces ambiguous /dev file
- 09:30 AM Bug #6902 (Can't reproduce): upgrade-parallel test failed in the nightlies
- 09:29 AM Bug #6781 (Can't reproduce): timed out waiting for recovery - probably ceph command hang
- let's wait for this on a real branch
- 09:28 AM Bug #6938 (Resolved): client full handling races with OSDs having newer maps
- 08:32 AM CephFS Bug #7011 (New): ENOTEMPTY on ceph-fuse + snaptest-? test
- ...
- 07:55 AM Bug #6988 (Fix Under Review): pg query doesn't show current 'problem' stats
- 07:27 AM Revision fae569d3 (ceph): Rename filestore_perf_t to objectstore_perf_t
- The name "filestore_perf_t" isn't suitable for other backends, so we
need to rename it.
Signed-off-by: Haomai Wang <... - 07:27 AM Revision 015e9819 (ceph): Move PerfCounter from ObjectStore.h to FileStore.h
- The anonymous enum which used by FileStore as PerfCounter is only can be
used by FileStore, so we need to drive out O... - 06:24 AM Revision b6d1d8f1 (ceph): mds: finish opening sessions even if import aborted
- Signed-off-by: Yan, Zheng <zheng.z.yan@intel.com>
- 06:24 AM Revision 802df76f (ceph): mds: properly update mdsdir's authority during recovery
- dirfrag of mdsdir doesn't inherit its parent inode's authority.
Signed-off-by: Yan, Zheng <zheng.z.yan@intel.com> - 06:24 AM Revision 43f7268f (ceph): mds: properly set dirty flag when journalling import
- Signed-off-by: Yan, Zheng <zheng.z.yan@intel.com>
- 06:24 AM Revision 4526d13a (ceph): mds: fix stale session handling for multiple mds
- Don't add new caps to stale session when importing inodes. Don't
touch session when importing caps because it confuse... - 06:24 AM Revision 80005f1e (ceph): mds: fix discover path race
- When C_MDC_RetryDiscoverPath executed, we may have already become
auth mds of base
Signed-off-by: Yan, Zheng <zheng.... - 05:45 AM Revision a680ea7b (ceph): osd/ReplicatedPG: update pg stat fields when they are queried
- When we query a PG, fill in (most of) the same fields that get
populated when the stats are reported to the mon.
Fix... - 05:44 AM Revision ebb3ad9f (ceph): osd/PG: move some pg stat update into a helper
- When we publish PG stats to the monitor, we fill in a bunch of fields
with information about degraded objects, scrub ... - 05:43 AM Revision 05274f33 (ceph): osd: include peer_info in pg query
- This can be very useful when diagnosing peering problems.
Signed-off-by: Sage Weil <sage@inktank.com> - 05:16 AM Revision 58d68995 (ceph): Merge pull request #947 from dachary/wip-6824
- mon: set ceph osd (down|out|in|rm) error code on failure
Reviewed-by: Sage Weil <sage@inktank.com> - 04:15 AM Revision ef902ee0 (ceph): mds: send info of imported caps back to the exporter (rename)
- use MMDSSlaveRequest::OP_FINISH slave request to send information
of rename imported caps back to the exporter. This ... - 04:15 AM Revision 4fdeb00d (ceph): mds: include counterpart's information in cap import/export messages
- when exporting indoes with client caps, the importer sends cap import
messages to clients, the exporter sends cap exp... - 04:15 AM Revision 6a565881 (ceph): mds: re-send cap exports in resolve message.
- For rename operation that changes inode's authority, if master mds
of the operation crashed, inode's original auth md... - 04:15 AM Revision 9dc52ff0 (ceph): mds: send cap import messages to clients after importing subtree succeeds
- When importing subtree, the importer sends cap import messages to clients
before the import subtree operation is cons... - 04:15 AM Revision 05b192fa (ceph): mds: simplify how to export non-auth caps
- Introduce a new flag in cap import message. If client finds the flag
is set, it releases exporter's caps (send releas... - 04:15 AM Revision d0b744a1 (ceph): client: handle session flush message
- Signed-off-by: Yan, Zheng <zheng.z.yan@intel.com>
- 04:15 AM Revision 71d1eb37 (ceph): mds: add CEPH_FEATURE_EXPORT_PEER and bump the protocal version
- Signed-off-by: Yan, Zheng <zheng.z.yan@intel.com>
- 04:15 AM Revision 5fdcc568 (ceph): mds: fix bug in MDCache::open_ino_finish
- It's wrong to erase open_ino_info_t after finishing contexts, because
MDCache::open_ino() can be called again when fi... - 04:15 AM Revision 87ca2604 (ceph): mds: fix sending resolve message
- need to send resolve message when mds is in reconnect state
Signed-off-by: Yan, Zheng <zheng.z.yan@intel.com> - 04:15 AM Revision e1818692 (ceph): mds: don't request CEPH_CAP_PIN from auth mds
- avoid triggering assert(in->get_loner() >= 0 && in->mds_caps_wanted.empty())
in Locker::file_xsyn()
Signed-off-by: Y... - 04:15 AM Revision 1536e814 (ceph): mds: check lock state before eval_gather
- Locker::eval_gather() can dispatch requests, which may change other
locks' states.
Signed-off-by: Yan, Zheng <zheng.... - 04:15 AM Revision e6c4d32e (ceph): mds: waiting for slave reuqest to finish
- If MDS receives a client request, but find there is an existing
slave request. It's possible that other MDS forwarded... - 04:15 AM Revision f134c772 (ceph): mds: avoid allocating MDRequest::More when cleanup request
- Signed-off-by: Yan, Zheng <zheng.z.yan@intel.com>
- 04:15 AM Revision 65259796 (ceph): mds: include inode version in auth mds' lock messages
- encode inode version in auth mds' lock messages, so that version
of replica inodes get updated. This is important bec... - 04:15 AM Revision 77515b7a (ceph): mds: increase cap sequence when sharing max size
- For case:
- client voluntarily releases some caps through cap update message
- mds shares the new max by sending ca... - 04:15 AM Revision d00ec791 (ceph): mds: flush session messages before exporting caps
- Following sequence of events can happen when exporting inodes:
- client sends open file request to mds.0
- mds.0 han... - 04:15 AM Revision ff8b9ac3 (ceph): mds: send info of imported caps back to the exporter (export dir)
- Introduce a new class Capability::Import and use it to send information
of imported caps back to the exporter. This i... - 04:15 AM Revision 85171fd6 (ceph): mds: send info of imported caps back to the exporter (cache rejoin)
- Use cache rejoin ack message to send information of rejoin imported
caps back to the exporter. Also move the code tha... - 04:15 AM Revision 5a902a0e (ceph): mds: unify nonce type
- MDSCacheObject::replica_nonce is defined as __s16, but nonce type
in MDSCacheObject::replica_map is int. This mismatc... - 04:15 AM Revision 0aed0d48 (ceph): mds: handle cache rejoin corner case
- A recovering MDS may receives strong cache rejoin from a survivor,
then the survivor restarts, the recovering MDS rec... - 04:15 AM Revision 498d5c49 (ceph): mds: process delayed expire if exporting dir cancelled in warnning state
- we may add delayed expire when exporting dir is in warnning state
Signed-off-by: Yan, Zheng <zheng.z.yan@intel.com> - 04:15 AM Revision 2fea08b5 (ceph): mds: merge delayed cache expire
- Signed-off-by: Yan, Zheng <zheng.z.yan@intel.com>
- 04:15 AM Revision 913f7fd8 (ceph): mds: fix empty directory check
- Since commit 310032ee81(fix mds scatter_writebehind starvation), rdlock
a scatter lock does not always propagate dirt... - 04:15 AM Revision bd561772 (ceph): mds: re-send discover if want_xlocked becomes true
- If want_xlocked becomes true, we can not rely on previously sent discover
because it's likely the previous discover i... - 04:15 AM Revision 31f5b027 (ceph): mds: fix rename notify
- commit 1d86f77edf (mds: fix cross-authorty rename race) introduced
rename notify, but it puts the code in wrong brack... - 04:15 AM Revision 3ac08860 (ceph): mds: avoid issuing caps when inode is frozen
- Signed-off-by: Yan, Zheng <zheng.z.yan@intel.com>
- 04:15 AM Revision b2a13700 (ceph): mds: re-issue caps after importing inode
- After importing inode, the issued caps can be less than the caps
client wants. So always re-issue caps after importin... - 04:15 AM Revision d8440c4c (ceph): mds: avoid leaving bare-bone dirfrags in the cache
- Signed-off-by: Yan, Zheng <zheng.z.yan@intel.com>
- 04:15 AM Revision b7d78918 (ceph): mds: keep dentry lock in sync state
- unlike locks of other types, dentry lock in unreadable state can
block path traverse, so it should be in sync state a... - 04:15 AM Revision ab93aa59 (ceph): mds: freeze tree deadlock detection.
- there are two situations that result freeze tree deadlock.
- mds.0 authpins an item in subtree A
- mds.0 sends req... - 04:15 AM Revision 9471fdc6 (ceph): mds: put import/export related states together
- Current code uses several STL maps to record import/export related
states. A map lookup is required for each state ac... - 04:15 AM Revision 0344d9af (ceph): mds: rework stale import/export message detection
- Current code uses import state to detect obsolete import/export messages.
it does not work for the case: cancel a sub... - 02:25 AM Bug #6992: OSD assert fails after it found it was marked as down by monitor during high system lo...
- Then I think this may be a multi thread issue. In SimpleMessenger::rebind(), it calls accepter.rebind() before mark_d...
- 01:08 AM Bug #6824 (Resolved): Removal of an OSD that is not down should set non-successful status code
- merged https://github.com/ceph/ceph/commit/15b8616b13a327701c5d48c6cb7aeab8fcc4cafc
- 12:57 AM Revision edc4224d (ceph): Merge remote-tracking branch 'gh/wip-hitset'
- Reviewed-by: Greg Farnum <greg@inktank.com>
Conflicts:
src/common/config_opts.h
src/osd/ReplicatedPG.cc
src/osdc/... - 12:23 AM Revision f192a600 (ceph): Revert "common/Formatter: add newline to flushed output if m_pretty"
- This reverts commit d6146b0d915f1420b5e76f7037f656460c314461.
As Yehuda points out, this does not properly handle ca... - 12:22 AM Revision c7b44d66 (ceph): Revert "common: fix perf_counters unittests for trailing newline in m_p...
- This reverts commit ba5572397c0e48378b0a0e556db1b2c02756617e.
12/15/2013
- 10:06 PM Revision 31507c90 (ceph): qa: test for error when ceph osd rm is EBUSY
- http://tracker.ceph.com/issues/6824 fixes #6824
Signed-off-by: Loic Dachary <loic@dachary.org> - 09:21 PM Bug #7009 (Resolved): mon: leaking memory from elector mon commands
- ...
- 08:57 PM Feature #6805 (Resolved): mon: find a way to properly extend/change mon commands without breaking...
- 08:52 PM Revision 31b60bfd (ceph): rgw: Fix CORS allow-headers validation
- This fix is needed because Ceph presently validates CORS headers in a
case-sensitive manner. Keeps a local cache of l... - 08:45 PM Revision 15b8616b (ceph): mon: set ceph osd (down|out|in|rm) error code on failure
- Instead of always returning true, the error code is set if at least one
operation fails.
EINVAL if the OSD id is inv... - 08:45 PM Revision cb352484 (ceph): qa: silence cephtool tests cleanup
- The file removal installed to be triggered when the script stops must
not fail if the file does not exist.
Signed-of... - 08:45 PM Revision f9cfa24a (ceph): qa: add function name and line number to cephtool output
- Signed-off-by: Loic Dachary <loic@dachary.org>
- 08:45 PM Revision 4b9a41aa (ceph): qa: make cephtool test imune to pool size
- instead of assuming the pool size is 2, query it and increment it to
test for pool set data size. It allows to run th... - 07:48 PM Bug #6751 (New): Pool 'df' statistics go bad after changing PG count
- (Weird, thought I had updated this at the time but apparently not)
Yeah, the values persisted for tens of minutes ... - 08:07 AM Bug #6751 (Need More Info): Pool 'df' statistics go bad after changing PG count
- 07:46 PM Revision 6a7edab2 (ceph): rgw: Clarify naming of case-change functions
- It is not clear that the lowercase_http_attr & uppercase_http_attr
functions replace dashes with underscores. Rename ... - 07:43 PM Revision 2abacd96 (ceph): rgw: Look at correct header about headers for CORS
- The CORS standard dictates that preflight requests are made with the
Access-Control-Request-Headers header containing... - 06:24 PM Revision 80c6c54d (ceph): Merge pull request #716 from ceph/wip-formatter-newlines
- common/Formatter: add newline to flushed output if m_pretty
- 06:23 PM Revision 3862ad8f (ceph): Merge pull request #943 from dachary/wip-formatter-newlines
- common: fix perf_counters unittests for trailing newline in m_pretty
- 06:18 PM Revision 550adb82 (ceph): Merge pull request #942 from sstock/master
- Add -n option to mount.ceph, feature 7006
Reviewed-by: Sage Weil <sage@inktank.com> - 05:49 PM Revision e37467b7 (ceph): Add -n option to mount.ceph. Required by autofs when /etc/mtab is a li...
- Signed-off-by: Steve Stock <steve@technolope.org>
- 04:41 PM Revision 11065b5a (ceph): Merge pull request #937 from christian-marie/master
- Document librados's rados_write's behaviour in reguards to return value.
- 04:40 PM Revision 25838f3b (ceph): Merge pull request #924 from dachary/wip-erasure-doc
- doc: update erasure code development doc
- 04:40 PM Revision 62a7d9c7 (ceph): Merge pull request #946 from dachary/wip-80-column
- osd: format test_osd_types.cc to 80 columns
- 04:40 PM Revision caf59635 (ceph): Merge pull request #945 from dachary/wip-6981
- ceph-disk: zap needs at least one device
Reviewed-by: Sage Weil <sage@inktank.com> - 04:39 PM Revision 89dd0206 (ceph): Merge pull request #944 from dachary/wip-6679
- common: fix rare race condition in Throttle unit tests
Reviewed-by: Sage Weil <sage@inktank.com> - 04:32 PM Revision 9c71d97b (ceph): Merge pull request #948 from dachary/wip-6736-1
- mon: typo s/degrated/degraded/
Backport: emperor, dumpling - 04:15 PM Revision aa365e4b (ceph): mon: typo s/degrated/degraded/
- http://tracker.ceph.com/issues/6736 refs #6736
Signed-off-by: Loic Dachary <loic@dachary.org> - 04:07 PM Documentation #7007 (Closed): Missing page for "OS Recommendations"
- This page linked from different places, the one that I saw was http://ceph.com/docs/master/rados/deployment/preflight...
- 03:23 PM Revision 5741bfe9 (ceph): osd: format test_osd_types.cc to 80 columns
- Signed-off-by: Loic Dachary <loic@dachary.org>
- 02:34 PM Revision 07888ef3 (ceph): ceph-disk: zap needs at least one device
- If given no argument, ceph-disk zap should display the usage instead of
silently doing nothing. Silence can be confus... - 02:23 PM Bug #6679 (Resolved): throttle: transient unit test failure
- 05:44 AM Bug #6679 (Fix Under Review): throttle: transient unit test failure
- 05:36 AM Bug #6679: throttle: transient unit test failure
- "work in progress":https://github.com/ceph/ceph/pull/944
- 05:04 AM Bug #6679: throttle: transient unit test failure
- Reproduced with...
- 02:22 PM Bug #6981 (Resolved): ceph-disk zap with no disk arg succeeds with no error
- 06:39 AM Bug #6981 (Fix Under Review): ceph-disk zap with no disk arg succeeds with no error
- "proposed fix":https://github.com/ceph/ceph/pull/945
- 01:31 PM Revision e57239e9 (ceph): common: fix rare race condition in Throttle unit tests
- The thread created to test Throttle race conditions updates a value (
throttle.get_current() ) that is tested by the ... - 01:30 PM Revision 938f22ca (ceph): common: format Throttle test to 80 columns
- Signed-off-by: Loic Dachary <loic@dachary.org>
- 12:24 PM Revision ba557239 (ceph): common: fix perf_counters unittests for trailing newline in m_pretty
- Signed-off-by: Loic Dachary <loic@dachary.org>
- 11:26 AM Revision c744aec6 (ceph): Merge pull request #929 from kazhang/add-pkg-config
- add apt-get install pkg-config for ubuntu server
Reviewed-by: Loic Dachary <loic@dachary.org>
Reviewed-by: Sage Weil... - 10:17 AM Feature #7006 (Resolved): mount.ceph: add -n option (no mtab update)
- commit:e37467b7bf33f963ace18e21df35884a86109cc7
- 08:17 AM Bug #6736 (Fix Under Review): Bugs in per pool IOPs/recovery statistics
- "typo fix":https://github.com/ceph/ceph/pull/948
- 08:17 AM Bug #6736: Bugs in per pool IOPs/recovery statistics
- 07:56 AM Bug #6824 (Fix Under Review): Removal of an OSD that is not down should set non-successful status...
- "work in progress":https://github.com/ceph/ceph/pull/947
- 07:55 AM Bug #6824: Removal of an OSD that is not down should set non-successful status code
- 06:58 AM Bug #6894 (Resolved): 'ceph osd dump' produces malformed JSON if pool snapshots are present
- "matching pull request":https://github.com/ceph/ceph/pull/871
- 05:49 AM Bug #6598 (Can't reproduce): osd crash after recreating pool with same name (cuttlefish + bobtail?)
- 04:11 AM Bug #6787: upstart is restarting daemons which we want to be dead
- #6789 also happens because a mon is restarted automatically by upstart. It's not the only reason though but it seems ...
12/14/2013
- 03:14 PM Revision e4b5ab81 (ceph): Use path when calling teuthology-report. …
- The 'teuthology-report' command is probably not going to exist
in $PATH so get the location of the running command an... - 02:02 PM Feature #7006: mount.ceph: add -n option (no mtab update)
- Created https://github.com/ceph/ceph/pull/942 which solves the problem I was having with automount, works now:
<pr... - 01:33 PM Feature #7006 (Resolved): mount.ceph: add -n option (no mtab update)
- While experimenting with automounting a ceph file system I discovered automount assumes the -n option is supported. ...
- 12:35 AM Revision d1e63b3c (ceph): ceph_test_rados: test is_dirty, undirty
- Signed-off-by: Sage Weil <sage@inktank.com>
- 12:35 AM Revision 87547bde (ceph): ceph_test_rados: debug: include exists|dne in update_object_version
- Signed-off-by: Sage Weil <sage@inktank.com>
- 12:35 AM Revision 81279e3b (ceph): osd/ReplicatedPG: rename invalidate_forward
- Signed-off-by: Sage Weil <sage@inktank.com>
- 12:35 AM Revision f86d6e77 (ceph): osd/ReplicatedPG: debug: improve maybe_handle_cache() handling
- Signed-off-by: Sage Weil <sage@inktank.com>
- 12:35 AM Revision f50389d7 (ceph): ceph_test_rados_api_tier: test undirty on non-existent object
- Signed-off-by: Sage Weil <sage@inktank.com>
- 12:35 AM Revision 697151ea (ceph): osd/osd_types: fix operator<< on copy-get operation
- This was missed in 15c8267e34aaba7a6d1d316b22519982a997f5a0.
Signed-off-by: Sage Weil <sage@inktank.com> - 12:35 AM Revision 0dc59af9 (ceph): osd/ReplicatedPG: fix promote: set oi.size
- Signed-off-by: Sage Weil <sage@inktank.com>
- 12:35 AM Revision c0e4ed34 (ceph): osd/ReplicatedPG: more verbose heading for process_copy_chunk
- Signed-off-by: Sage Weil <sage@inktank.com>
- 12:35 AM Revision ea088fae (ceph): osd/osd_types: operator<< for ObjectContext::RWState
- Signed-off-by: Sage Weil <sage@inktank.com>
- 12:35 AM Revision 3d9c4997 (ceph): osd: rename IGNORE_OVERLAY -> IGNORE_CACHE
- This is about skipping cache logic, not the tier pool overlay property.
Signed-off-by: Sage Weil <sage@inktank.com> - 12:35 AM Revision 067536c2 (ceph): osdc/Objecter: add CEPH_OSD_FLAG_IGNORE_OVERLAY flag
- If the flag is set, send the op to the pool specified and ignore the
overlay. Note that this obsoletes the global Ob... - 12:35 AM Revision 42d6af1c (ceph): osd/ReplicatedPG: use IGNORE_OVERLAY flag for copy-from
- No need to use the Objecter-wide setting now.
Signed-off-by: Sage Weil <sage@inktank.com> - 12:35 AM Revision 330a1305 (ceph): osdc/Objecter: remove honor_cache_redirects global flag
- We can do this on a per-op basic with CEPH_OSD_FLAG_IGNORE_OVERLAY.
Signed-off-by: Sage Weil <sage@inktank.com> - 12:35 AM Revision 41be4feb (ceph): librados: seek during object iteration
- Add ability to reset iterator to a specific hash position. For now, we
just truncate this to the current PG. In the... - 12:35 AM Revision 14f76cc2 (ceph): ceph_test_rados: fix CopyFromOp locking
- Signed-off-by: Sage Weil <sage@inktank.com>
- 12:35 AM Revision 10c9be34 (ceph): osd/ReplicatedPG: allow osds to issue writes to osds
- We asserted that the client was not an OSD years ago when we separated out
the client and cluster networks. Now, we ... - 12:35 AM Revision 36bbcf8e (ceph): osd/ReplicatedPG: drop unnecessary temp vars in execute_ctx()
- Both of these are pulled out of ctx->obs, which is not updated until the
very end; use that instead!
Signed-off-by: ... - 12:35 AM Revision 5e547f87 (ceph): osd/ReplicatedPG: be consistent about ctx->obs vs ctx->obc->obs
- Just for consistency (ctx->obs =- &ctx->obc->obs).
Signed-off-by: Sage Weil <sage@inktank.com>
Conflicts:
src/osd... - 12:35 AM Revision 3ef73106 (ceph): osd/ReplicatedPG: add SKIPRWLOCKS flag
- Flush puts us in an conundrum:
- the flush eventually writes, behaving like a write
- writes take the write lock a... - 12:35 AM Revision 56ad14ec (ceph): osd/ReplicatedPG: split off finish_ctx from execute_ctx
- The second part of execute_ctx() is doing some somewhat generic work to
make the prepared updates in the ctx apply, u... - 12:35 AM Revision 66263bb6 (ceph): osd/ReplicatedPG: use get_next_version() in finish_promote
- Signed-off-by: Sage Weil <sage@inktank.com>
- 12:35 AM Revision ca86656e (ceph): osd/ReplicatedPG: use finish_ctx for finish_promote
- Use the common code here to avoid duplicating this logic.
Signed-off-by: Sage Weil <sage@inktank.com> - 12:35 AM Revision 90eb1ec1 (ceph): osd/ReplicatedPG: set ctx->obc in simple_repop_create
- Strangely nobody hss needed this yet, but we will shortly.
Signed-off-by: Sage Weil <sage@inktank.com> - 12:35 AM Revision 78df1c37 (ceph): osd/ReplicatedPG: set object_info and snapset xattrs on promote
- For the normal write path, prepare_transaction() handles this for us. In
this case, we need to do it explicitly.
Si... - 12:35 AM Revision 387e224a (ceph): librados: add cache_flush(), cache_try_flus(), cache_evict() methods
- Not yet implemented by the OSD.
Signed-off-by: Sage Weil <sage@inktank.com> - 12:35 AM Revision 181cb8e8 (ceph): librados, osd: add IGNORE_OVERLAY flag
- Add a flag that will make the OSD bypass the cache overlay logic. This is
needed in order to handle operations like ... - 12:35 AM Revision 85282319 (ceph): osd/osd_types: introduce helper for osd op flags -> string conversion
- Signed-off-by: Sage Weil <sage@inktank.com>
Conflicts:
src/osd/osd_types.h - 12:35 AM Revision 8b9b7136 (ceph): librados: add an aio_operate that takes a write and flags
- Until now you could only pass flags to read operations.
Signed-off-by: Sage Weil <sage@inktank.com> - 12:35 AM Revision 4c014edd (ceph): osd/ReplicatedPG: implement cache_evict
- Signed-off-by: Sage Weil <sage@inktank.com>
- 12:35 AM Revision 0b81ff68 (ceph): osd/ReplicatedPG: skip promote for DELETE
- If an op starts with DELETE there is no need to promote the old content
from the base tier. Note that this only work... - 12:35 AM Revision 20d149e1 (ceph): osd/ReplcatedPG: maybe_handle_cache style
- Signed-off-by: Sage Weil <sage@inktank.com>
- 12:35 AM Revision be29f47a (ceph): ceph_test_rados_api_tier: rename tests
- Signed-off-by: Sage Weil <sage@inktank.com>
- 12:35 AM Revision 0b7b16d7 (ceph): ceph_test_rados_api_tier: add simple promote-on-read test
- Signed-off-by: Sage Weil <sage@inktank.com>
- 12:35 AM Revision 0b085b17 (ceph): osd/ReplicatedPG: create whiteout on promote ENOENT
- If we try to fetch an object from the base tier and it is not present, we
can create a whiteout object.
Signed-off-b... - 12:35 AM Revision e0a49698 (ceph): osd/ReplicatedPG: ENOENT when deleting a whiteout
- Signed-off-by: Sage Weil <sage@inktank.com>
- 12:35 AM Revision 2aea631c (ceph): ceph_test_rados_api_tier: verify delete creates whiteouts
- Signed-off-by: Sage Weil <sage@inktank.com>
- 12:35 AM Revision fabc6ba1 (ceph): osd/ReplicatedPG: set whiteout in cache pool on delete
- If we delete an object in the cache pool, set the whiteout flag instead of
removing the on-disk object.
Signed-off-b... - 12:35 AM Revision fd8f7d29 (ceph): osd/ReplicatedPG: clear whiteout when writing into cache tier
- If we have a whiteout object and then write over it, clear the whiteout
flag.
Signed-off-by: Sage Weil <sage@inktank... - 12:35 AM Revision dd079e2a (ceph): osd/ReplicatedPG: handle is_whiteout in do_osd_ops()
- Most of the time we handle whiteouts by returning ENOENT before we even
get this far. However, for a mixed read/write... - 12:35 AM Revision 0699fc5c (ceph): ReplicatedPG: copy: don't return from finish_copyfrom
- The return value is meaningless; nothing in this function can fail.
Signed-off-by: Greg Farnum <greg@inktank.com>
Re... - 12:35 AM Revision b371dd8b (ceph): ReplicatedPG: promote: first draft pass at doing object promotion
- This is not yet at all complete -- among other things, it will
retry forever on any object which doesn't exist in the... - 12:35 AM Revision d15aedbd (ceph): ReplicatedPG: promote: add the OpRequest to the Callback
- This way we can do stuff to it, and we're about to.
Signed-off-by: Greg Farnum <greg@inktank.com>
Reviewed-by: Sage ... - 12:35 AM Revision 5fa08fb8 (ceph): ReplicatedPG: promote: handle failed promotes
- If we get an error back, reply to the client directly and remove
the op which triggered promotion from our blocked op... - 12:35 AM Revision 0caa02c5 (ceph): workunits: check errors propagate on cache pools in caching_redirects.sh
- Signed-off-by: Greg Farnum <greg@inktank.com>
Reviewed-by: Sage Weil <sage@inktank.com> - 12:35 AM Revision 920c0bff (ceph): workunits: break down cache pool tests to be more precise; expand some
- Signed-off-by: Greg Farnum <greg@inktank.com>
Reviewed-by: Sage Weil <sage@inktank.com> - 12:35 AM Revision 4a29b22e (ceph): osd/ReplicatedPG: move r<0 handling into finish_promote()
- Let logic in header, and will let us handle ENOENT with a whiteout.
Signed-off-by: Sage Weil <sage@inktank.com> - 12:35 AM Revision 654d8c33 (ceph): osd/ReplicatedPG: UNDIRTY is not a user_modify
- Signed-off-by: Sage Weil <sage@inktank.com>
- 12:35 AM Revision 66f2e748 (ceph): osd/ReplicatedPG: use simple_repop_{create,submit} for finish_promote
- Signed-off-by: Sage Weil <sage@inktank.com>
- 12:35 AM Revision 325aae36 (ceph): ReplicatedPG: copy: switch out the CopyCallback interface
- The tuple was already unwieldy with 4 members; I didn't want to add
more. Instead, create a new CopyResults struct wh... - 12:13 AM Revision a3e50b09 (ceph): rgw: rework framework configuration
- The framework config code is now more generic and ready for multiple
config options. Different frameworks are comma s... - 12:09 AM Revision b7946ff4 (ceph): doc: Added additional comments on placement targets and default placement.
- Signed-off-by: John Wilkins <john.wilkins@inktank.com>
- 12:08 AM Revision 902f19c2 (ceph): doc: Updates to federated config.
- Reverted Emperor versionadded to Dumpling as it gets backported.
Added default index and bucket pools to pool creatio... - 12:02 AM Revision 29cc7229 (ceph): test_ipaddr: add another unit test
- Was checking something for kbader.
Signed-off-by: Sage Weil <sage@inktank.com> - 12:02 AM Revision 026b724b (ceph): osd/ReplicatedPG: drop unused hit_set_start_stats
- Signed-off-by: Sage Weil <sage@inktank.com>
- 12:01 AM Revision 3d768d23 (ceph): osd/ReplicatedPG: maintain stats for the hit_set_* objects
- We also make hit_set.current_info reflect only the on-disk 'current', not
anything that is not persisted.
Signed-off...
12/13/2013
- 10:54 PM Revision 9814b93a (ceph): osd/ReplicatedPG: set object_info_t, SnapSet on hit_set objects
- These are first-class user-visible rados objects and need these attrs.
Signed-off-by: Sage Weil <sage@inktank.com> - 10:50 PM Revision dabd5d6e (ceph): vstart.sh: --hitset <pool> <type>
- Signed-off-by: Sage Weil <sage@inktank.com>
- 10:44 PM Revision 9bd1ebe9 (ceph): rgw: Allow wildcard in supported keystone roles.
- http://tracker.ceph.com/issues/4365 fixes #4365
Signed-off-by: Christophe Courtaut <christophe.courtaut@gmail.com>
(... - 10:42 PM Revision ec086df1 (ceph): ObjectStore: add ObjectStore::get_cur_stats and impl in FileStore
- Signed-off-by: Samuel Just <sam.just@inktank.com>
(cherry picked from commit ebde89d5602536b4bc651737e4280fdfb6634c32) - 10:42 PM Revision 455f28a7 (ceph): OSDMonitor: add 'osd perf' command to dump recent osd perf information
- Signed-off-by: Samuel Just <sam.just@inktank.com>
(cherry picked from commit 94c3f29a32cbf169d896015da6765febd3c724e0) - 10:42 PM Revision 0eae05f3 (ceph): rgw: turn swift COPY into PUT
- Fixes: #6606
The swift COPY operation is unique in a sense that it's a write
operation that has its destination not s... - 10:42 PM Revision 0edee8e9 (ceph): rgw: get rid of req_state.bucket_name
- No need for this field, as we already have req_state.bucket_name_str.
This saves us some memory allocation / freeing ... - 10:42 PM Revision 5df2f35e (ceph): rgw: init src_bucket_name, src_object in the S3 handler
- Be consistent and initialize these fields also in the S3 case.
Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
(che... - 10:42 PM Revision e52b9b76 (ceph): Validate S3 tokens against Keystone
- - Added config option to allow S3 to use Keystone auth
- Implemented JSONDecoder for KeystoneToken
- RGW_Auth_S3::aut... - 10:42 PM Revision b37c5d20 (ceph): rgw: do not use UNPRIVILEGED_DAEMON option
- This is sort of a backport for 4f403c26dc0048ad63e20d20369fa86bfb31c50e
in that we get the same behavior change, but ... - 10:42 PM Revision 5bf7e7be (ceph): osd_types: add osd_perf_stat_t type
- Signed-off-by: Samuel Just <sam.just@inktank.com>
(cherry picked from commit 95f3353a6ab3a4dd2bf28eaca7122487942e913e) - 06:51 PM Revision 8ee1febf (ceph): Merge pull request #2 from ceph/wip-6947
- use 3 OSDs in fs tests
- 06:49 PM Revision d4edaec8 (ceph): use 3 OSDs in fs tests
- Signed-off-by: Alfredo Deza <alfredo.deza@inktank.com>
- 06:49 PM Revision 1c008302 (ceph): Merge remote-tracking branch 'gh/wip-objecter-full-2'
- Reviewed-by: Sage Weil <sage@inktank.com>
- 06:40 PM Revision 8e4b5bf8 (ceph): Merge pull request #936 from ceph/wip-rbd-single-major
- rbd: support for single-major device number allocation scheme
Reviewed-by: Josh Durgin <josh.durgin@inktank.com> - 06:03 PM Revision e7652e6b (ceph): Merge pull request #932 from ceph/wip-6979
- replace sgdisk subprocess calls with a helper
Reviewed-by: Sage Weil <sage@inktank.com> - 05:58 PM Revision d5ac7365 (ceph): Merge remote-tracking branch 'gh/next'
- 05:57 PM Revision 5bb04763 (ceph): test/libcephfs: release resources before umount
- Fixes: #6742
Signed-off-by: Yan, Zheng <zheng.z.yan@intel.com>
Reviewed-by: Sage Weil <sage@inktank.com> - 05:25 PM Revision 7cb815f7 (ceph): Merge pull request #162 from jcsp/fsid-conf
- Fix FSID not being set in ceph.conf
- 05:24 PM Revision 02e0a1e9 (ceph): Merge pull request #161 from jcsp/ssh-config
- Respect .ssh/config when opening SSH connections
- 05:06 PM Revision 897dfc11 (ceph): use the new get_command helper in check_call
- Signed-off-by: Alfredo Deza <alfredo@deza.pe>
- 04:48 PM Revision 2e2b8feb (ceph): Skip the 'dead' report on old branches
- Signed-off-by: Zack Cerza <zack.cerza@inktank.com>
- 04:41 PM devops Bug #7004 (Rejected): qemu: rhel and centos qemu packages should depend on librbd
- The newest version in ceph-extras will dynamically disable whatever isn't supported, so any version bobtail or later ...
- 03:40 PM Revision a4213059 (ceph): rbd: switch to strict_strtol for major parsing
- Use common/strict_strtol, which actually parses integers in a proper
way, instead of atoi for parsing /sys/bus/rbd/de... - 03:40 PM Revision 462b3898 (ceph): rbd: match against whole disks on unmap
- Currently the way 'rbd unmap' translates a user-provided block device
into an rbd id is it matches the major number o... - 03:40 PM Revision 784cc894 (ceph): rbd: match against both major and minor on unmap on newer kernels
- As described in commit "rbd: match against wholedisk device numbers on
unmap", currently we only match against major ... - 03:40 PM Revision 8a473bcc (ceph): rbd: add support for single-major device number allocation scheme
- With the preparatory commits ("rbd: match against wholedisk device
numbers on unmap" and "rbd: match against both maj... - 03:40 PM Revision eae8531e (ceph): rbd: modprobe with single_major=Y on newer kernels
- On kernels that support it, and if 'rbd map' is given a chance to
modprobe, turn on single-major device number alloca... - 03:33 PM rgw Feature #6999: Make RGW_MAX_PUT_SIZE configurable
- A customer requests that we consider making this parameter larger. A configurable parameter might be a good compromis...
- 03:27 PM rgw Feature #6999: Make RGW_MAX_PUT_SIZE configurable
- This is a part of the S3 spec, and anybody using larger objects should probably be using multi-part upload. It's obvi...
- 03:14 PM rgw Feature #6999 (Resolved): Make RGW_MAX_PUT_SIZE configurable
- /src/rgw/rgw_common.h has a #define for RGW_MAX_PUT_SIZE which limits uploads to 5GB. Please consider this for a conf...
- 03:01 PM Documentation #6774: Documentation: osd scrub load threshold incorrect.
- Hi Tyler,
I just looked at rados/configuration/osd-config-ref.rst.
Looks like you have already fixed the wording.... - 01:58 PM Documentation #6998 (Closed): Add hyperlink for paxos
- 01:44 PM Documentation #6998 (Closed): Add hyperlink for paxos
- Add hyperlink to wikipedia entry for Paxos algorithm.
- 10:43 AM Bug #6992 (Need More Info): OSD assert fails after it found it was marked as down by monitor duri...
- 08:56 AM Bug #6992: OSD assert fails after it found it was marked as down by monitor during high system lo...
- Do you have complete logs of this? I don't think this is the full story.
If an osd is marked down, it will bind t... - 10:15 AM CephFS Bug #6742: failed libcephfs_interface_tests (LibCephFS.ReaddirRCB hangs)
- commit:5bb04763de38cedafbe3dfa76e67f2b617bb5938
- 09:57 AM CephFS Bug #6742 (Resolved): failed libcephfs_interface_tests (LibCephFS.ReaddirRCB hangs)
- 10:02 AM Bug #6796 (Resolved): ceph mons interpretting pg splits very wrong
- 10:02 AM Bug #6979 (Resolved): ceph-disk breaks when sgdisk is not available
- 09:43 AM Bug #6938 (Fix Under Review): client full handling races with OSDs having newer maps
- 05:31 AM devops Bug #6962: libleveldb1 fails to install for cuttlefish testing basic plana test
- It looks like this issue re-appeared.
Log output:... - 02:25 AM Revision 24a048b7 (ceph): Document librados's rados_write's behaviour in reguards to return value.
- 02:14 AM Revision a865fece (ceph): osd/ReplicatedPG: debug: improve hit_set func banners
- Signed-off-by: Sage Weil <sage@inktank.com>
- 02:13 AM Revision b6871cf8 (ceph): osd/ReplicatedPG: do not update current_last_update on activate
- Don't update this when we apply the log to our in-memory hitset! We should
only update this when we persist somethin... - 12:04 AM Revision 36c0344f (ceph): Use saucy gitbuilder when grabbing sha1 for arm.
- Old quantal gitbuilders are gone until hardware comes back. Use
the new saucy gitbuilders instead.
12/12/2013
- 11:33 PM Revision 966dad54 (ceph): Make sure to report all results.
- If a just-finished job was using a teuthology branch not known to
contain the reporting feature, then report the job ... - 11:14 PM Bug #6992: OSD assert fails after it found it was marked as down by monitor during high system lo...
- https://github.com/ceph/ceph/pull/938
- 10:40 PM Bug #6992 (Resolved): OSD assert fails after it found it was marked as down by monitor during hig...
- If the OSD load is high and heartbeat message is not replied timely, it may be marked as down by monitor. In this cas...
- 11:00 PM Revision 7f135ec9 (ceph): Enable reporting of single jobs
- (also switch to docopt)
Signed-off-by: Zack Cerza <zack.cerza@inktank.com> - 09:45 PM Revision 3d23b9b2 (ceph): Remove the child's stderr completely
- Signed-off-by: Zack Cerza <zack.cerza@inktank.com>
- 09:34 PM Revision 9ff4d4a4 (ceph): Fix FSID not being set in ceph.conf
- Symptom was that 'ceph --admin-daemon... config get fsid'
returned zeros, while correct fsid was present in cluster m... - 05:47 PM Revision 625f479b (ceph): When starting a job, tell paddles it's running
- Signed-off-by: Zack Cerza <zack.cerza@inktank.com>
- 05:42 PM Revision 31b0823d (ceph): Merge pull request #934 from cernceph/wip-rgw-ulimit
- radosgw: increase nofiles ulimit on sysvinit machines
- 05:41 PM Revision 500de8b2 (ceph): Merge pull request #935 from ceph/wip-vstart-memstore
- vstart.sh: add --memstore option
- 05:31 PM Revision bcde2003 (ceph): vstart.sh: add --memstore option
- for setting memstore backed osds
Signed-off-by: Yehuda Sadeh <yehuda@inktank.com> - 04:22 PM Fix #6990: osd crash when running mixed versions of dumpling and master
- oh yes Greg, you are right. this happened on the dumpling osds and not on the master ones.
- 03:46 PM Fix #6990: osd crash when running mixed versions of dumpling and master
- And I'm pretty sure this is fallout from some of David's no-acting-set-for-backfillers changes.
- 03:45 PM Fix #6990: osd crash when running mixed versions of dumpling and master
- Tamil, I think you're backwards about which OSDs crash? That backtrace says .67.4, and the master branch OSDs don't c...
- 03:39 PM Fix #6990 (Resolved): osd crash when running mixed versions of dumpling and master
- steps to reproduce:
1.running a cluster of 2 nodes with dumpling version of ceph.
2. upgrade only the osds on the... - 04:16 PM Revision a9334a1c (ceph): use the absolute path for executables if found
- Signed-off-by: Alfredo Deza <alfredo@deza.pe>
- 03:51 PM Linux kernel client Feature #6991 (Resolved): support CEPH_FEATURE_OSDMAP_ENC
- The userspace code is switching the OSDMap encoding format going forward, in particular prompted by what's needed for...
- 03:26 PM Revision 43561f79 (ceph): remove trailing semicolon
- Signed-off-by: Alfredo Deza <alfredo@deza.pe>
- 01:53 PM Revision a33c95f1 (ceph): radosgw: increase nofiles ulimit on sysvinit machines
- Clusters with many OSDs require a higher nofiles ulimit than the RHEL default. Increase it.
Tested-by: Dan van der S... - 01:44 PM Bug #6988 (Resolved): pg query doesn't show current 'problem' stats
- pg <pgid> query shows 0 for the four stats
num_objects_missing_on_primary
num_object_copies
num_objects_unfound
... - 01:09 PM devops Feature #6020: radosgw-apache opinionated package
- 10:45 AM devops Feature #6986 (Closed): Build packages containing RBD and CephFS kernel modules for installation ...
- RHEL7 beta is now available. We need to provide two packages - one for RBD and one for CephFS - which provide prebuil...
- 09:29 AM Feature #6028 (In Progress): EC: [link] ensure that erasure coded pools don't work until the osds...
- 04:07 AM Linux kernel client Bug #6984 (Resolved): RBD volume not mountable after creating 8 or more snapshots
- Currently using Ceph RBD volumes for SAMBA shares. Each volume is regularly snapshotted and the snapshots are mounted...
- 02:07 AM Revision a7f87f3a (ceph): Longer timeout after sync/reboot.
- With only a 5 second sleep via ssh and python it looks like a
race-condition was sometimes hitting where it would thi... - 12:13 AM Revision 71cefc29 (ceph): doc/release-notes: sort
- meh
Signed-off-by: Sage Weil <sage@inktank.com> - 12:11 AM Revision ee3173d9 (ceph): doc/release-notes: fix indentation; sigh
- Signed-off-by: Sage Weil <sage@inktank.com>
12/11/2013
- 11:59 PM Revision 3abc1894 (ceph): doc/release-notes: v0.73
- Signed-off-by: Sage Weil <sage@inktank.com>
- 11:39 PM Revision 03429d1e (ceph): PendingReleaseNotes: note CRUSH and hashpspool default changes
- Signed-off-by: Sage Weil <sage@inktank.com>
- 11:37 PM Revision 1504b961 (ceph): Merge pull request #930 from ceph/wip-hashpspool
- enable hashpspool by default
Reviewed-by: Samuel Just <sam.just@inktank.com> - 11:33 PM Revision 1c4759ce (ceph): mon: OSDMonitor: receive CephInt on 'osd pool set' instead on CephString
- This partially reverts 2fe0d0d9 in order to allow Emperor monitors to
forward mon command messages to Dumpling monito... - 11:33 PM Revision 1983ed80 (ceph): mon: OSDMonitor: drop cmdval_get() for unused variable
- We don't ever use any value as a float, so just drop obtaining it. This
makes it easier to partially revert 2fe0d0d9... - 11:33 PM Revision c044b218 (ceph): Partial revert "mon: osd pool set syntax relaxed, modify unit tests"
- This reverts commit 08327fed8213a5d24cd642e12b38a171b98924cb, except
for the hashpspool bit. We switched back to an ... - 11:29 PM Revision bb50276f (ceph): Revert "Partial revert "mon: osd pool set syntax relaxed, modify unit t...
- This reverts commit e80ab94bf44e102fcd87d16dc11e38ca4c0eeadb.
We accept non-CephInt arguments again, now that we've ... - 11:29 PM Revision 0cd36e05 (ceph): mon/OSDMonitor: take 'osd pool set ...' value as a string again
- We ran into problems before when we made this a string because a mixed
cluster of mons might forward a client request... - 09:41 PM Revision f3ce07c6 (ceph): Respect .ssh/config when opening SSH connections
- This handles that case where your private key is
in a non-default location that you're pointing
to in ~/.ssh/config. - 09:37 PM Revision a447fb7d (ceph): osd/OSDMonitor: accept 'osd pool set ...' value as string
- Newer monitors take this as a CephString. Accept that so that if we are
mid-upgrade and get a forwarded message usin... - 09:27 PM Revision 72a304ac (ceph): Merge pull request #925 from ceph/wip-mon-api
- Merge in changes to unify the API presented by the monitors and handle changes gracefully.
(Upgrade tests) Tested-by... - 08:41 PM Revision e19e3801 (ceph): replace sgdisk subprocess calls with a helper
- Signed-off-by: Alfredo Deza <alfredo@deza.pe>
- 07:19 PM Revision 4b6d7214 (ceph): osd: enable HASHPSPOOL by default
- Much like the CRUSH tunables, this first appears in kernel v3.9.
Unlike the CRUSH tunables, it does not appear in Ce... - 06:12 PM Revision 2bfd34ac (ceph): mon: by default, warn if some members of the quorum are "classic"
- Signed-off-by: Greg Farnum <greg@inktank.com>
- 06:12 PM Revision fb47d540 (ceph): mon: if we're the leader, don't validate command matching
- Classic-format commands never match our leader command set!
Signed-off-by: Greg Farnum <greg@inktank.com> - 05:06 PM Bug #6979: ceph-disk breaks when sgdisk is not available
- This was covered in the email list, but figured it would make sense to put in here too: It seems likely that this can...
- 01:22 PM Bug #6979 (Fix Under Review): ceph-disk breaks when sgdisk is not available
- Pull request opened: https://github.com/ceph/ceph/pull/932
- 08:08 AM Bug #6979 (Resolved): ceph-disk breaks when sgdisk is not available
- ...
- 03:41 PM Bug #6982 (Duplicate): osd crashed when running mixed versions of dumpling and master
- have a cluster up and running on dumpling, upgrading only the osds on one node to master branch followed by thrashing...
- 02:19 PM Bug #6903 (Resolved): nightlies: upgrade tests hung on ceph quorum status
- local tests passed.
- 02:05 PM Feature #5999: EC: [link] OSD internals must work in terms of cpg_t
- Apparently it's impractical to make progress on this until 5990, 5994, wip-hitset and wip-cache are merged.
- 10:17 AM Feature #5999 (In Progress): EC: [link] OSD internals must work in terms of cpg_t
- 02:05 PM Bug #6981 (Resolved): ceph-disk zap with no disk arg succeeds with no error
- ...which is a bit terrifying. It should at least warn about the missing arg.
- 11:17 AM Linux kernel client Bug #6809: 3.11 kernel panic: Workqueue: ceph-msgr con_work
- I too am having similar issues to this on kernel 3.11.(7|9|11) with Fedora 19.
I've tested it across Ceph versions... - 10:56 AM Subtask #5858: Backfill should be able to handle multiple backfill peers
- "work in progress":https://github.com/ceph/ceph/pull/931
- 07:45 AM Revision 0eb784b6 (ceph): Added handling of a 'local' option inside install.py which specifies
- a local directory containing deb or rpm files to be installed.
Fixes: 5149
Signed-off-by: Warren Usui <warren.usui@i... - 02:28 AM Revision e6200578 (ceph): add apt-get install pkg-config for ubuntu server
- Signed-off-by: Kai Zhang <kaizh.pub@gmail.com>
- 01:08 AM Revision b8884e01 (ceph): MemStore: update for the new ObjectStore interface
- 68fdcfa1cc249af859400a2ce4590fefbb2f525b changed the ObjectStore
interface in the 'next' branch, which was merged int...
12/10/2013
- 10:48 PM Revision b3acff1d (ceph): Use continue, not break
- Fixes a bug where not all pids were being collected
Signed-off-by: Zack Cerza <zack.cerza@inktank.com> - 10:48 PM Revision 4a6e47cd (ceph): Tweak logic for pid lookup
- Signed-off-by: Zack Cerza <zack.cerza@inktank.com>
- 10:25 PM Revision 77145f1b (ceph): Fix indentation
- Signed-off-by: Zack Cerza <zack.cerza@inktank.com>
- 09:00 PM Revision e5a02c33 (ceph): Merge branch 'next'
- 08:25 PM Revision b66902b6 (ceph): Merge pull request #927 from dachary/wip-crush-test
- crush: remove crushtool test leftover
- 07:35 PM Revision 8ac1da8e (ceph): crush: remove crushtool test leftover
- Signed-off-by: Loic Dachary <loic@dachary.org>
- 07:19 PM Revision 57574fef (ceph): Don't show child's stderr, but show archive path
- Signed-off-by: Zack Cerza <zack.cerza@inktank.com>
- 07:10 PM Revision 6bd63a1e (ceph): Merge pull request #920 from dachary/wip-man
- man: Ceph is also an object store
Reviewed-by: Sage Weil <sage@inktank.com> - 06:23 PM Revision ec609cac (ceph): Elector: use monitor's encoded command sets instead of our own
- Signed-off-by: Greg Farnum <greg@inktank.com>
- 06:14 PM Revision 85a024a6 (ceph): Merge pull request #865 from ceph/wip-doc-build-cluster
- Wip doc build cluster
- 06:09 PM Revision e223e534 (ceph): Monitor: encode and expose mon command sets
- Signed-off-by: Greg Farnum <greg@inktank.com>
- 05:56 PM Fix #6977 (Resolved): objecter: ENOSPC returned by pool quota will cause IO errors for non-cached...
- The client has no idea why it's getting ENOSPC, and block layers turn this into EIO. The only way clients can detect ...
- 05:34 PM Revision 420a2f15 (ceph): man: update man/ from doc/man/8
- As explained in admin/manpage-howto.txt
Signed-off-by: Loic Dachary <loic@dachary.org> - 05:33 PM Revision 8d60cd1a (ceph): man: Ceph is also an object store
- Replace
Ceph distributed file system
with
Ceph distributed storage system
to help reduce the idea that Ceph... - 05:06 PM Revision d6504740 (ceph): Merge pull request #923 from dachary/wip-crush-test
- CrushTester patches and documentation
Reviewed-by: Sage Weil <sage@inktank.com> - 04:56 PM Revision faaf5463 (ceph): os/MemStore: do on_apply_sync callback synchronously
- We can easily deadlock if we put this in the Finisher thread behind other
work; do it synchronously!
Signed-off-by: ... - 04:06 PM Revision 339b7c47 (ceph): Add debug statements
- Signed-off-by: Zack Cerza <zack.cerza@inktank.com>
- 04:02 PM Revision 971216b5 (ceph): Merge pull request #159 from ceph/wip-cache
- rados: allow existing pool(s) to be used
- 02:26 PM devops Bug #6966 (Resolved): ceph-disk: prepare --dmcrypt failing
- ...
- 02:18 PM devops Bug #6962 (Resolved): libleveldb1 fails to install for cuttlefish testing basic plana test
There is no leveldb package in the cuttlefish repo. The leveldb shared library is only needed for more recent rel...- 07:26 AM devops Bug #6962 (Resolved): libleveldb1 fails to install for cuttlefish testing basic plana test
- Log file: http://qa-proxy.ceph.com/teuthology/teuthology-2013-12-09_19:01:30-ceph-deploy-cuttlefish-testing-basic-pla...
- 02:04 PM Messengers Fix #6965 (In Progress): messenger: Pipe should not hold pipe lock while doing message encode and...
- Right now we hold the pipe lock while running the message encode and cephx steps in Pipe::writer. We don't need to, a...
- 09:38 AM Feature #5994: EC: [link] Backfill should be able to handle multiple backfill peers
- 04:55 AM Revision d8ad51ee (ceph): v0.73
- 04:53 AM Revision 990b2b5d (ceph): ceph_test_rados_api_tier: make HitSetWrite handle pg splits
- Signed-off-by: Sage Weil <sage@inktank.com>
- 03:15 AM rgw Bug #6961 (Resolved): create bucket add's new .bucket.meta file every time
- ---
root@vm-1:/vol0/ceph/osd# find | grep meta | grep test1 | wc -l
44
root@vm-1:/vol0/ceph/osd# s3 -u create test... - 01:05 AM Revision b86aefc2 (ceph): restructured upgrade suites to be under one folder called upgrade.
- Signed-off-by: tamil <tamil.muthamizhan@inktank.com>
- 12:02 AM Revision 6c856a2e (ceph): rados: allow existing pool(s) to be used
- Signed-off-by: Sage Weil <sage@inktank.com>
12/09/2013
- 11:43 PM Feature #6960 (Closed): DBObjectMap: refactor interface
- 1. expose underlying KeyValueDB transactions to caller, so they can bundle several DBObjectMap ops together and captu...
- 11:37 PM Revision 2266eeb3 (ceph): ceph.conf: put 2x command in [global]
- so that osdmaptool sees it.
Signed-off-by: Sage Weil <sage@inktank.com> - 11:34 PM Revision a6f4d71c (ceph): Elector: keep a list of classic mons instead of each mon's commands
- We aren't actually using the sets, so don't bother keeping them.
Signed-off-by: Greg Farnum <greg@inktank.com> - 10:57 PM Revision 48b8ba4a (ceph): Create a DateTime object from the timestamp
- Signed-off-by: Zack Cerza <zack.cerza@inktank.com>
- 10:42 PM Revision 5ea5018d (ceph): Make -a optional
- Signed-off-by: Zack Cerza <zack.cerza@inktank.com>
- 10:42 PM Revision 025ab366 (ceph): Add missing req: psutil
- 09:16 PM Revision 3d6feb4b (ceph): Merge pull request #151 from ceph/wip-distro-kernel
- Wip distro kernel
- 09:01 PM Revision d7289f75 (ceph): Auto-restart
- If /tmp/teuthology-restart-workers is newer than the running process,
restart.
Signed-off-by: Zack Cerza <zack.cerza... - 09:01 PM Revision 33a3600f (ceph): Merge pull request #158 from ceph/wip-nuke
- make nuke behave
- 08:10 PM Revision fbc4f990 (ceph): crush: remove scary message string
- The string is no longer used and can be removed.
Signed-off-by: Loic Dachary <loic@dachary.org> - 08:10 PM Revision 20263dd3 (ceph): crush: add unitest for crushtool --show-bad-mappings
- Signed-off-by: Loic Dachary <loic@dachary.org>
- 08:10 PM Revision a888a57f (ceph): crush: implement --show-bad-mappings for indep
- Support the presence of ITEM_NONE device numbers in the indep mapping as
proof of a bad mapping. Implement the associ... - 08:10 PM Revision 472f495e (ceph): crush: document the --test mode of operations
- Signed-off-by: Loic Dachary <loic@dachary.org>
- 07:42 PM Revision 1b80f4aa (ceph): nuke: ignore exceptions while issuing reboot command
- I'm seeing failed tasks (and nuke) leak machines. It looks like we are
getting an exception on the '... reboot -f -n... - 07:42 PM Revision 478ecc30 (ceph): Remove unused variable.
- Signed-off-by: Sandon Van Ness <sandon@inktank.com>
- 07:35 PM Revision ce8ff0a3 (ceph): Added additional comments.
- Signed-off-by: Sandon Van Ness <sandon@inktank.com>
- 07:26 PM Revision d33df28c (ceph): Elector: transmit local api on election win, accept leader's on loss
- If we're the leader, just point to our local set. Disseminating these
will let peons advertise the full command set s... - 07:26 PM Revision cb51b1ed (ceph): Monitor: disseminate leader's command set instead of our own
- Signed-off-by: Greg Farnum <greg@inktank.com>
- 07:26 PM Revision 3cb58f74 (ceph): Monitor: validate incoming commands against the leader's set too
- Then check against our own, and forward if we don't recognize it
or for some reason don't match.
Signed-off-by: Greg... - 07:26 PM Revision ba673be3 (ceph): Monitor: import MonCommands.h from original Dumpling and expose it
- If the Elector doesn't receive a set of commands from the elected leader, it
assumes the monitor is "classic" and use... - 07:26 PM Revision f1ccdb41 (ceph): Elector: share local command set when deferring
- We're about to use this at a basic level, to identify when we have
"classic" monitors in-quorum, but could also do so... - 07:26 PM Revision ea86444f (ceph): Monitor: Elector: share the classic command set if we have a classic mon
- The leader now checks to see if any monitors did not provide their
command set, and if so, shares the list of "classi... - 07:26 PM Revision f9329036 (ceph): Monitor: pull command mapping out of _allowed_command()
- We want to be able to validate commands against both the leader and
local command sets, so make that functionality ge... - 07:26 PM Revision 8025fb33 (ceph): messages: make room for passing supported monitor commands in MMonElection
- We're going to use this space to let leader tell everybody what
commands it supports.
Signed-off-by: Greg Farnum <gr... - 07:16 PM Revision 7d000e34 (ceph): Merge pull request #918 from ceph/port/misc
- Misc portability patches
Reviewed-by: Sage Weil <sage@inktank.com> - 06:03 PM Bug #6903: nightlies: upgrade tests hung on ceph quorum status
- thats right, modified the test.
my local test run got stuck up with sync issues :(
will close the bug once the test... - 04:58 PM Documentation #6959 (Resolved): Diagrams showing even numbers of Monitors
- The diagrams used here http://ceph.com/docs/master/radosgw/federated-config/#inter-region-metadata-replication all sh...
- 04:28 PM Revision 4c5f7ba8 (ceph): Merge pull request #922 from dachary/wip-crush-choose-tries
- crush: fix map->choose_tries boundary test
Reviewed-by: Sage Weil <sage@inktank.com> - 02:19 PM Revision b8bdbd92 (ceph): doc: update erasure code development doc
- With a link to the tracker issue implementing the new indep mode.
Signed-off-by: Loic Dachary <loic@dachary.org> - 01:07 PM CephFS Bug #6957 (Duplicate): LibCephFS.ReaddirRCB hangs
- 12:54 PM CephFS Bug #6957 (Duplicate): LibCephFS.ReaddirRCB hangs
- several runs including ubuntu@teuthology:/var/lib/teuthworker/archive/teuthology-2013-12-07_23:01:29-fs-next-testing-...
- 12:24 PM Bug #6955 (Won't Fix): ceph-disk should set the guid correctly when re-using a partition
- If an existing ceph data partition is re-used, the partition GUID is not reset when ceph-disk prepares it. As a conse...
- 11:27 AM Feature #6805 (Fix Under Review): mon: find a way to properly extend/change mon commands without ...
- PR: https://github.com/ceph/ceph/pull/925
I haven't looked at backporting it to emperor yet, although it should be... - 10:22 AM Bug #6945 (Resolved): btrfs: async snap ioctl deadlock
- 04:15 AM Bug #6945: btrfs: async snap ioctl deadlock
- This is a regression that was introduced in 3.12-rc1 and has been fixed
in 3.12-rc6, commit 1bda19eb73d68b304148e672... - 10:11 AM Documentation #6682 (In Progress): Adjustments to the Ceph Quick installation guide
- wip-doc-quick-start has the updates. Requires review by Alfredo.
- 09:57 AM Revision 41152a63 (ceph): crush: --show-utilization* implies --show-statistics
- --show-utilization* outputs only if --show-statistics is set, which is
confusing. Instead of failing, set --show-stat... - 08:29 AM Bug #6685: osd/ReplicatedPG.cc: 8345: FAILED assert(0 == "erroneously present object")
- /a/teuthology-2013-12-07_23:00:19-rados-next-testing-basic-plana/135872
also possibly
ubuntu@teuthology:/a/teut... - 08:07 AM devops Feature #6365 (Resolved): Package up newer version of Qemu for RHEL & EL6.3 and 6.4
- 08:05 AM devops Feature #5775 (Resolved): create qemu rbd package for rhel 6.5 - qemu-rbd
- 08:02 AM Bug #3569: Monitor & OSD failures when an OSD clock is wrong
- I wasn't able to reproduce this bug at Ceph v0.72.1. Increasing memory was insignificant and stopped after a while.
- 06:21 AM Revision 3dcbf460 (ceph): encoding: fix [encode|decode]_array_nohead
- We want to actually encode each element and keep it, rather than
writing each one at the position after the array end... - 06:21 AM Revision ac69a012 (ceph): MonCommand: support encode/decode
- Signed-off-by: Greg Farnum <greg@inktank.com>
- 06:21 AM Revision dca5383f (ceph): MonCommand: add operator== and operator!=
- Signed-off-by: Greg Farnum <greg@inktank.com>
- 06:21 AM Revision 4cd5c3bf (ceph): Monitor: expose local monitor commands to other compilation units
- Signed-off-by: Greg Farnum <greg@inktank.com>
- 06:21 AM Revision dcb0a4f3 (ceph): Monitor: add a separate leader_supported_commands
- This isn't used yet, but will be shortly.
Signed-off-by: Greg Farnum <greg@inktank.com> - 05:44 AM Subtask #6900 (Resolved): crush: fix indep mode
- "merged":https://github.com/ceph/ceph/pull/869
- 01:35 AM CephFS Bug #6914: Accessing specific file on CephFS causes client hang
- At first my response would be: there's no /sys/kernel/debug/ceph dir. But later it became clear that debugfs needs to...
12/08/2013
- 09:17 PM Revision c928f077 (ceph): crush: output --show-bad-mappings on err
- Instead of using stdout so that it displays well when used in
conjunction with --show-statistics
Signed-off-by: Loic... - 09:17 PM Revision 7482d62f (ceph): crush: add CrushTester accessors
- Signed-off-by: Loic Dachary <loic@dachary.org>
- 04:00 PM Revision 5e0722fa (ceph): crush: fix map->choose_tries boundary test
- CrushWrapper::start_choose_profile allocates map->choose_tries with
choose_total_tries elements. When crush_choose_fi... - 04:59 AM Revision 94da2153 (ceph): Merge pull request #869 from ceph/wip-crush
- crush changes for erasure coding
Reviewed-by: Loic Dachary <loic@dachary.org>
Reviewed-by: Samuel Just <sam.just@ink... - 02:07 AM Revision 3b39a8a9 (ceph): librbd: rename howmany to avoid conflict
- A howmany macro exists on some platforms in standard headers, but there
really isn't any sort of standard that I've f... - 02:07 AM Revision ad3825c6 (ceph): TrackedOp: remove unused private variable
- Signed-off-by: Noah Watkins <noahwatkins@gmail.com>
- 02:07 AM Revision ef4061f0 (ceph): librbd: remove unused private variable
- Signed-off-by: Noah Watkins <noahwatkins@gmail.com>
12/07/2013
- 10:01 PM Revision 096f9b32 (ceph): Merge pull request #917 from ceph/port/compat
- compat: define replacement TEMP_FAILURE_RETRY
Reviewed-by: Sage Weil <sage@inktank.com> - 10:00 PM Revision 96068bfa (ceph): Merge pull request #919 from ceph/port/fdatasync
- wbthrottle: use feature check for fdatasync
Reviewed-by: Sage Weil <sage@inktank.com> - 09:20 PM Revision a2766063 (ceph): ceph.conf: default to 2x
- A bunch of our tests rely on this; they need to be fixed
before we can run at 3x.
Signed-off-by: Sage Weil <sage@ink... - 08:59 PM rbd Bug #6548: tgt: Kernel panic putting zpool on iSCSI LUN using bs_rbd
- The problem seems to have gone away after updating most of the software and retesting. Versions now under test:
C... - 08:06 PM CephFS Bug #6914: Accessing specific file on CephFS causes client hang
- Rolf Fokkens wrote:
> [inode 10000013022 [2,head] /disco/herbert/disco-data/output/video-conversion auth v34 ap=6+0 ... - 08:00 PM CephFS Bug #6914: Accessing specific file on CephFS causes client hang
- Rolf Fokkens wrote:
> Well, actually now the client hang seems to be gone. Still the MDS seems to attribute the lock... - 06:37 PM Revision 539fe261 (ceph): wbthrottle: use feature check for fdatasync
- Checking for fdatasync uses the same approach as the qemu configure
script. The relevant commit is d1722a27f552a22561... - 06:24 PM Revision 60a25093 (ceph): rados_sync: remove unused private variable
- Signed-off-by: Noah Watkins <noahwatkins@gmail.com>
- 06:24 PM Revision 663da61c (ceph): rados_sync: fix mismatched tag warning
- Signed-off-by: Noah Watkins <noahwatkins@gmail.com>
- 06:24 PM Revision 43c16767 (ceph): mon: check for sys/vfs.h existence
- Signed-off-by: Noah Watkins <noahwatkins@gmail.com>
- 06:22 PM Revision c99cf265 (ceph): make: increase maximum template recursion depth
- With clang on OSX spirit blows up without this.
Signed-off-by: Noah Watkins <noahwatkins@gmail.com> - 06:18 PM Revision e2be0991 (ceph): compat: define replacement TEMP_FAILURE_RETRY
- Not all platforms have it.
Signed-off-by: Noah Watkins <noahwatkins@gmail.com> - 02:11 PM Bug #6903: nightlies: upgrade tests hung on ceph quorum status
- the test is broken: it is waiting for a quorum of [a,b] but it's getting [a,b,c] because there is no incompatibility ...
- 01:14 PM Bug #6313: dumpling: FAILED assert(latest->is_update()) from recover_primary()
- it's a delete:...
- 01:06 PM Bug #6313: dumpling: FAILED assert(latest->is_update()) from recover_primary()
- on dumpling:...
- 09:33 AM rgw Bug #6951 (Resolved): rgw: bucket re-create fails when cache is disabled
- 06:12 AM Bug #6950 (Resolved): refactor OSD awareness of FileStore to make the ObjectStore backend configu...
- In order to make ease for different backend such as keyvaluestore, we need to extract awareness of FileStore codes:
... - 01:42 AM Revision c0a43275 (ceph): nuke: fix sync before reboot timeout
- If you do 'timeout 5 sync' and sync hangs, timeout will block trying to
kill it.
Instead, just background sync, wait... - 12:56 AM Revision a52ef1df (ceph): Merge remote-tracking branch 'gh/wip-fix-3x'
- Reviewed-by: Josh Durgin <josh.durgin@inktank.com>
- 12:55 AM Revision 0386095e (ceph): Merge remote-tracking branch 'gh/wip-fix-tunables'
- Reviewed-by: Josh Durgin <josh.durgin@inktank.com>
- 12:24 AM Revision 3b3cbf52 (ceph): crush/CrushCompiler: make current set of tunables 'safe'
- We can reenable this error the next time we add new tunables.
Signed-off-by: Sage Weil <sage@inktank.com> - 12:24 AM Revision 4eb8891d (ceph): crush/CrushCompiler: start with legacy tunables when compiling
- Ensure that a crush file always compiled deterministically, even though
the default values for *new* maps has changed... - 12:24 AM Revision 8535ceda (ceph): crushtool: remove scary tunables messages
- Signed-off-by: Sage Weil <sage@inktank.com>
- 12:22 AM Revision e8fdef21 (ceph): crush: add indep data set to cli tests
- This will help us catch things if we break the mapping.
Signed-off-by: Sage Weil <sage@inktank.com> - 12:22 AM Revision 564de6ea (ceph): osdmaptool: fix cli tests for 3x
- Signed-off-by: Sage Weil <sage@inktank.com>
- 12:21 AM Revision 6704be68 (ceph): osd: default to 3x replication
- 3x is the recommendation; it should be the default too.
Signed-off-by: Sage Weil <sage@inktank.com> - 12:10 AM Revision 308e4f9d (ceph): Merge pull request #913 from dachary/wip-crush-unittest
- CrushWrapper::move_bucket unittest and minor fixes
Reviewed-by: Sage Weil <sage@inktank.com> - 12:03 AM Revision 8d0180b1 (ceph): objecter: don't take extra throttle budget for resent ops
- These ops have already taken their budget in the original op_submit().
It will be returned via put_op_budget() when t...
12/06/2013
- 11:48 PM Revision 38647f76 (ceph): Revert "osd: default to 3x replication"
- This reverts commit cb26fbde52f31b449af60acce3ced34e593d6e1e.
Fix unit tests and do integration tests first; this ma... - 11:31 PM Revision cbeb1f45 (ceph): crush: detach_bucket must test item >= 0 not > 0
- Since detach_bucket is a private helper solely used by move_bucket which
contains another ( correct ) safeguard, the ... - 11:27 PM Revision 2cd73f9d (ceph): crush: remove obsolete comments from link_bucket
- Probably copy/pasted from move_bucket.
Signed-off-by: Loic Dachary <loic@dachary.org> - 11:21 PM Revision e00324b2 (ceph): crush: remove redundant code from move_bucket
- The following was introduced in 2012 by a2d0cff1b071bed84ac439e4fcf9ddfb936f89c8
// un-set the device name so we c... - 11:20 PM Revision 8ef80a4c (ceph): crush: unittest CrushWrapper::move_bucket
- Signed-off-by: Loic Dachary <loic@dachary.org>
- 10:45 PM Revision 865880b5 (ceph): Merge pull request #888 from ceph/wip-crush-tunables
- default to bobtail-era crush tunables.
Reviewed-by: Joao Eduardo Luis <joao.luis@inktank.com> - 10:38 PM Revision 650f896c (ceph): Merge pull request #903 from ceph/wip-memstore
- memstore: reference ObjectStore backend
Reviewed-by: Samuel Just <sam.just@inktank.com> - 10:37 PM Revision a6d66f9c (ceph): common/bloom_filter: fix copy ctor
- We should not delete[] an uninitialized pointer.
Signed-off-by: Sage Weil <sage@inktank.com> - 10:37 PM Revision 01f3ff72 (ceph): osd/HitSet: fix dump() of fpp
- Signed-off-by: Sage Weil <sage@inktank.com>
- 10:37 PM Revision bf96a7ea (ceph): osd/HitSet: fix copy ctor
- Signed-off-by: Sage Weil <sage@inktank.com>
- 10:37 PM Revision f0cfd229 (ceph): ceph_test_rados_api_tier: make HitSetRead test less noisy
- Signed-off-by: Sage Weil <sage@inktank.com>
- 10:37 PM Revision 3ea9230a (ceph): osd/HitSet: fix sealed initialization in Params ctor
- Signed-off-by: Sage Weil <sage@inktank.com>
- 10:37 PM Revision 456daf2a (ceph): ceph_test_rados_api_tier: add HitSet trim test
- Signed-off-by: Sage Weil <sage@inktank.com>
- 10:37 PM Revision 01cbbfaa (ceph): ceph_test_rados_api_tier: HitSetRead -> HitSetWrite
- This way it will pass despite thrashing.
Signed-off-by: Sage Weil <sage@inktank.com> - 10:37 PM Revision 638b2744 (ceph): ceph_test_rados_api_tier: add HitSetRead
- Verify that the HitSet reflects a read (and never written) object.
Signed-off-by: Sage Weil <sage@inktank.com> - 10:37 PM Revision a72094d5 (ceph): osd/HitSet: take Params as const ref to avoid confusion about ownership
- Signed-off-by: Sage Weil <sage@inktank.com>
- 10:37 PM Revision 737533f2 (ceph): osd: prevent zero BloomHitSet fpp
- Signed-off-by: Sage Weil <sage@inktank.com>
- 10:37 PM Revision 146e6aa7 (ceph): osd/ReplicatedPG: archive hit_set if it is old and not full
- This matches the condition under which we call _persist().
Signed-off-by: Sage Weil <sage@inktank.com> - 10:37 PM Revision c365cca4 (ceph): osd/HitSet: track BloomHitSet::Params fpp in micros, not as a double
- ...and store it as a 32-bit value, so that it actually works!
Signed-off-by: Sage Weil <sage@inktank.com> - 10:37 PM Revision 1d0af14a (ceph): test: add a HitSet unit test
- Signed-off-by: Greg Farnum <greg@inktank.com>
- 10:37 PM Revision c0eb95b8 (ceph): mds/Capability: no copying
- Signed-off-by: Sage Weil <sage@inktank.com>
- 10:37 PM Revision 1c107d3c (ceph): ceph-dencoder: add 'copy' command to test operator=
- Signed-off-by: Sage Weil <sage@inktank.com>
- 10:37 PM Revision c941e829 (ceph): test/encoding/check-generated: test copy ctor, operator=
- Signed-off-by: Sage Weil <sage@inktank.com>
- 10:37 PM Revision 916313c3 (ceph): osd/ReplicatedPG: put hit_set objects in a configurable namespace
- Signed-off-by: Sage Weil <sage@inktank.com>
- 10:37 PM Revision 4a743fb6 (ceph): osd/ReplicatedPG: trim old hit_set objects on persist
- Any time we persist a hit_set object, take the opportunity to remove any
old ones that we don't want any more.
Note ... - 10:37 PM Revision 1e94e27f (ceph): ceph_test_rados: hit hit_set_{list,get} rados operations
- This will do a list, and then get a random HitSet.
Signed-off-by: Sage Weil <sage@inktank.com> - 10:37 PM Revision 3c2d2d76 (ceph): pg_hit_set_info_t: remove unused size, target_size members
- Signed-off-by: Greg Farnum <greg@inktank.com>
- 10:37 PM Revision fa76d5e4 (ceph): ReplicatedPG: do not seal() HitSets until we're done with them
- We don't want to seal HitSets just because we're writing a
snapshot to disk; it potentially shrinks the in-memory one... - 10:37 PM Revision 5da12858 (ceph): osd/ReplicatedPG: apply log to new HitSet to capture writes after peering
- Signed-off-by: Sage Weil <sage@inktank.com>
- 10:37 PM Revision 41e0f970 (ceph): osd/HitSet: make pg_pool_t and Params operator<< less parenthetical
- pool 0 'data' rep size 2 min_size 1 crush_ruleset 0 object_hash rjenkins pg_num 8 pgp_num 8 last_change 8 owner 0 cra...
- 10:37 PM Revision 68c44cbb (ceph): mon/OSDMonitor: non-zero default bloom fpp
- Signed-off-by: Sage Weil <sage@inktank.com>
- 10:37 PM Revision e8ef7249 (ceph): osd/osd_types: include pg_hit_set_history_t in pg_info_t
- Track metadata about the currently accumulating HitSet as well as
previously archived ones in the pg_info_t. This wi... - 10:37 PM Revision db3fd115 (ceph): osd: add hit_set_* parameters to pg_pool_t
- Add pool properties to control what type of HitSet we want to use, along with
some (mostly generic) parameters.
Sign... - 10:37 PM Revision b92f4311 (ceph): mon/OSDMonitor: set hit_set fields
- Signed-off-by: Sage Weil <sage@inktank.com>
Signed-off-by: Greg Farnum <greg@inktank.com> - 10:37 PM Revision a97129f1 (ceph): osd/ReplicatedPG: add basic HitSet tracking
- Signed-off-by: Sage Weil <sage@inktank.com>
Signed-off-by: Greg Farnum <greg@inktank.com> - 10:37 PM Revision 904859e9 (ceph): osd/ReplicatedPG: use vectorized osd_op outdata for pg ops
- This lets us put PGLS in a compound operation. Nothing does that yet, but
this would allow it.
Despite appearances, ... - 10:37 PM Revision 0c43b778 (ceph): librados, osd: list and get HitSets via librados
- Signed-off-by: Sage Weil <sage@inktank.com>
Signed-off-by: Greg Farnum <greg@inktank.com> - 10:37 PM Revision a0cfbfd7 (ceph): librados: create new ceph_test_rados_api_tier target
- Move the dirty/undirty test to it, and add one for HitSets.
Signed-off-by: Sage Weil <sage@inktank.com>
Signed-off-b... - 10:37 PM Revision 4b5ab3f1 (ceph): osdc/Objecter: expose methods for getting object hash position and pg
- Signed-off-by: Sage Weil <sage@inktank.com>
Reviewed-by: Greg Farnum <greg@inktank.com> - 10:37 PM Revision 82859068 (ceph): librados: expose methods for calculating object hash position
- Signed-off-by: Sage Weil <sage@inktank.com>
Reviewed-by: Greg Farnum <greg@inktank.com> - 10:37 PM Revision 0b0d1e8e (ceph): librados: add wait_for_latest_osdmap()
- There are times when users may need to make sure the client has the
latest osdmap, for example after sending a mon co... - 10:37 PM Revision 9776e97a (ceph): osd/PG: factor out get_next_version()
- Signed-off-by: Sage Weil <sage@inktank.com>
Signed-off-by: Greg Farnum <greg@inktank.com> - 10:37 PM Revision 0b9874cd (ceph): osd/ReplicatedPG: factor out simple_repop_{create,submit} helpers
- This makes it easier to create repops correctly, and should help
prevent bugs like the one we remove here in process_... - 10:37 PM Revision c01b183d (ceph): osd_types: add generic HitSet type with bloom and explicit implementations
- Track a set of hash values, either explicitly or using a bloom_filter. Hide
the implementation and allow us to transp... - 10:37 PM Revision b5ea4700 (ceph): common/bloom_filter: fix operator=
- Signed-off-by: Sage Weil <sage@inktank.com>
- 10:37 PM Revision a430525c (ceph): osd/osd_types: add pg_hit_set_{info,history}_t
- Signed-off-by: Sage Weil <sage@inktank.com>
- 10:37 PM Revision bc7ace2e (ceph): librados: constify inbl command args
- Signed-off-by: Sage Weil <sage@inktank.com>
Reviewed-by: Greg Farnum <greg@inktank.com> - 10:37 PM Revision 468fffa5 (ceph): librados: document aio_flush()
- Signed-off-by: Sage Weil <sage@inktank.com>
Reviewed-by: Greg Farnum <greg@inktank.com> - 10:37 PM Revision d2963c0a (ceph): librados: add mon_command to C++ API
- This way librados users can execute monitor commands.
Signed-off-by: Sage Weil <sage@inktank.com>
Reviewed-by: Greg ... - 10:37 PM Revision 0d4ea9f7 (ceph): encoding: allow users to specify a different compatv after encoding
- This way we can set the compatv preferentially depending on whether
we've actually encoded new information or not.
S... - 10:37 PM Revision d692da34 (ceph): osd/osd_types: add explicit hash to object_locator_t
- Instead of hashing the object name or key, we allow the hash position to be
provided explicitly.
Signed-off-by: Sage... - 10:37 PM Revision 76e0b88f (ceph): osd/OSDMap: use new object_locator_t::hash to place object in a pg
- The hash value, if provided, becomes the ps (placement seed) portion of the
pg_t, skipping any hashing of the object ... - 10:37 PM Revision 92879f77 (ceph): osd: capture hashing of objects to hash positions/pgs in pg_pool_t
- The hashing is dependent on pool properties; capture (more of) it in a
method instead of having it in OSDMap.
Signed... - 10:37 PM Revision fb49065f (ceph): mon/MonClient: constify inbl command args
- Signed-off-by: Sage Weil <sage@inktank.com>
Reviewed-by: Greg Farnum <greg@inktank.com> - 10:37 PM Revision a29d4fc3 (ceph): osdc/Objecter: constify inbl command args
- Signed-off-by: Sage Weil <sage@inktank.com>
Reviewed-by: Greg Farnum <greg@inktank.com> - 10:36 PM Revision ef0f255a (ceph): osdc/Objecter: reimplement list_objects
- Return to caller at the end of each PG. This allows the caller to look at
the [pg_]hash_position and get something m... - 10:36 PM Revision d2e6cc63 (ceph): librados: add get_pg_hash_position to determine pg while listing objects
- Signed-off-by: Sage Weil <sage@inktank.com>
- 10:36 PM Revision eff932c6 (ceph): osdc/Objecter: stick bl inside ListContext
- This is simpler and less error-prone.
Signed-off-by: Sage Weil <sage@inktank.com> - 10:36 PM Revision 8e5803ab (ceph): osdc/Objecter: factor pg_read out of list_objects code
- This will get used later for other ops against PGs (instead of objects).
Signed-off-by: Sage Weil <sage@inktank.com> - 10:36 PM Revision dd8c9398 (ceph): osdc/Objecter: separate explicit pg target from current target
- The pgid field is used to store the pg the op mapped to. We were just
setting it directly for PGLS. Instead, fill i... - 10:36 PM Revision 9381b693 (ceph): osdc/Objecter: drop redundant condition
- We are inside an if (response_size) block.
Signed-off-by: Sage Weil <sage@inktank.com> - 10:36 PM Revision bffcca6a (ceph): osd/osd_types: make pref optional in pg_t constructor
- We don't use preferred placements any more, so this will
make it easier to start dropping references to it in new cod... - 10:33 PM Revision 3caf3eff (ceph): rbd: check write return code during bench-write
- This is allows rbd-bench to detect http://tracker.ceph.com/issues/6938
when combined with rapidly changing the mon os... - 10:33 PM Revision e32874fc (ceph): objecter: resend all writes after osdmap loses the full flag
- Now that the osd does not respond if it gets a map with the full flag
set first, clients need to resend all writes.
... - 10:33 PM Revision 4111729d (ceph): osd: drop writes when full instead of returning an error
- There's a race between the client and osd with a newly marked full
osdmap. If the client gets the new map first, it ... - 10:25 PM Revision 1d5427a7 (ceph): Merge pull request #907 from ceph/wip-3x
- osd: default to 3x replication
- 10:24 PM Revision caa0e22e (ceph): crush: CHOOSE_LEAF -> CHOOSELEAF throughout
- This aligns the internal identifier names with the user-visible names in
the decompiled crush map language.
Signed-o... - 10:24 PM Revision c8530194 (ceph): crush: add feature CRUSH_V2 for new indep mode and SET_*_TRIES rule steps
- Signed-off-by: Sage Weil <sage@inktank.com>
- 10:24 PM Revision e632a79b (ceph): PendingReleaseNotes: note change of CRUSH indep mode in release notes
- Signed-off-by: Sage Weil <sage@inktank.com>
- 10:24 PM Revision 384f01df (ceph): crush/mapper: dump indep partial progression for debugging
- ...if DEBUG_INDEP is #defined.
Signed-off-by: Sage Weil <sage@inktank.com> - 10:24 PM Revision cb88763c (ceph): crush/mapper: fix up the indep tests
- Fix indentation.
Simplify+fix the changed vs moved calculation.
Use the new SET_CHOOSE_LEAF_TRIES command.
Signed-of... - 10:24 PM Revision 64aeded5 (ceph): crush/mapper: apply chooseleaf_tries to firstn mode too
- Parameterize the attempts for the _firstn choose method, and apply the
rule-specified tries count to firstn mode as w... - 10:24 PM Revision d1b97462 (ceph): crush/mapper: add SET_CHOOSE_TRIES rule step
- Since we can specify the recursive retries in a rule, we may as well also
specify the non-recursive tries too for com... - 10:24 PM Revision 09ce7a2b (ceph): crush/CrushWrapper: set chooseleaf_tries to 5 for 'simple' indep rules
- When making a generic indep rule, set the recursive retry to 5. This gives
better overall results.
Signed-off-by: S... - 10:24 PM Revision 03911b07 (ceph): crush/CrushCompiler: [de]compile set_choose[leaf]_tries rule step
- Signed-off-by: Sage Weil <sage@inktank.com>
- 10:24 PM Revision 431a13eb (ceph): osd/OSDMap: fix feature calculation for CACHEPOOL
- We need to include the faeture in the mask.
Signed-off-by: Sage Weil <sage@inktank.com> - 10:15 PM Revision 580cf5f6 (ceph): Merge pull request #886 from ceph/wip-6922
- Fix some pg_num change return codes and make them more resistant to mis-use
Reviewed-by: Sage Weil <sage@inktank.com> - 08:35 PM Revision 63755c42 (ceph): Merge pull request #909 from dachary/wip-crush-unittest
- more CrushWrapper unittest
- 07:40 PM Revision 000c59a9 (ceph): crush: unittest remove useless c->create()
- Signed-off-by: Loic Dachary <loic@dachary.org>
- 07:40 PM Revision dc095214 (ceph): crush: unittest CrushWrapper::check_item_loc
- Signed-off-by: Loic Dachary <loic@dachary.org>
- 07:40 PM Revision b8190180 (ceph): crush: unittest use const instead of define
- And reduce the depth of the hierarchy because three levels of buckets
capture the same cases as four levels.
Signed-... - 07:40 PM Revision 16ac5904 (ceph): crush: unittest s/std::string/string/
- Signed-off-by: Loic Dachary <loic@dachary.org>
- 07:40 PM Revision 09938e64 (ceph): crush: unittest CrushWrapper::update_item
- Signed-off-by: Loic Dachary <loic@dachary.org>
- 07:40 PM Revision 4e26cc0d (ceph): crush: unittest CrushWrapper::get_immediate_parent
- Signed-off-by: Loic Dachary <loic@dachary.org>
- 07:24 PM Revision 516788d1 (ceph): Merge remote-tracking branch 'origin/next'
- 07:10 PM Revision 7a9a088d (ceph): rgw: fix reading bucket policy in RGWBucket::get_policy()
- Fixes: 6940
Backport: dumpling, emperor
We changed the way we keep the bucket policy, and we shouldn't try to
access... - 06:35 PM Revision cb26fbde (ceph): osd: default to 3x replication
- 3x is the recommendation; it should be the default too.
Signed-off-by: Sage Weil <sage@inktank.com> - 05:36 PM Bug #6945 (Resolved): btrfs: async snap ioctl deadlock
- ...
- 05:00 PM Bug #6944 (Resolved): objecter: localized read for missing objects polls replicas instead of wait...
- Replicas do not have the machinery to handle waiting for a missing object, so they return -EAGAIN when localized read...
- 04:29 PM Revision f4c16236 (ceph): Merge pull request #901 from dachary/wip-crush-unittest
- crush: check for invalid names in loc[]
Reviewed-by: Sage Weil <sage@inktank.com> - 04:27 PM Documentation #6465: admin/build-doc should have some kind of build check for broken links
- Today in #ceph: (04:25:48 PM) angdraug: at http://ceph.com/docs/master/ the link to Ceph Block Device is broken: poi...
- 02:41 PM Feature #6805 (In Progress): mon: find a way to properly extend/change mon commands without break...
- Current plan: Leader will share its command map with peons. Peons will advertise the leader's set, validate incoming ...
- 02:18 PM Revision bec1ac19 (ceph): Merge pull request #157 from ceph/wip-watchdog
- Implement a watchdog for queued jobs
- 02:15 PM Bug #6922 (Pending Backport): mon: do not allow unreasonable PG splitting
- 01:33 PM devops Feature #6365 (In Progress): Package up newer version of Qemu for RHEL & EL6.3 and 6.4
- 01:30 PM devops Bug #6941 (Duplicate): IOError when using ceph-deploy admin.
- 12:41 PM devops Bug #6941 (Duplicate): IOError when using ceph-deploy admin.
- vagrant@ceph-node0:~/ceph-cluster$ ceph-deploy admin ceph-node0 ceph-node1 ceph-node2 ceph-node3
[ceph_deploy.cli][I... - 01:18 PM rgw Feature #3454 (Fix Under Review): Support temp URLs for Swift API
- 01:15 PM rgw Feature #6877: tool to trim metadata, data, and bucket index logs
- 01:15 PM rgw Feature #6872: rgw: user level quota: async update of bucket info in user objclass
- 01:15 PM rgw Feature #6871: rgw: user level quota: core functionality
- 01:12 PM rbd Bug #5425: krbd: xfstest 89 hang, 'read_partial_message skipping long message'
- 01:12 PM rbd Fix #6079 (In Progress): libceph: osd_client does not handle PAUSERD or PAUSEWR or FULL flags in ...
- 01:08 PM rbd Fix #6079: libceph: osd_client does not handle PAUSERD or PAUSEWR or FULL flags in osdmap
- 01:10 PM rbd Fix #5048 (In Progress): krbd: limit of ~230 mapped images at once
- 01:08 PM rgw Feature #6870 (In Progress): rgw: user level quota: new user objclass
- 11:03 AM rgw Bug #6940 (Resolved): rgw: reading bucket policy through bucket restful api fails
- 10:51 AM Bug #6903: nightlies: upgrade tests hung on ceph quorum status
- logs: ubuntu@teuthology:/a/teuthology-2013-12-05_19:40:03-upgrade-parallel-master-testing-basic-plana/132263
- 08:43 AM Revision aedbc99f (ceph): crush: check for invalid names in loc[]
- Add the is_valid_crush_loc helper to test for invalid crush names in
insert_item and update_item, before performing a... - 07:13 AM Revision aa63d673 (ceph): os/MemStore: implement reference 'memstore' backend
- This is (as near to) a trivial ObjectStore backend for the OSD as we can
get at the moment. Everything is stored in ... - 07:13 AM Revision fe03ad28 (ceph): osd: queue pg deletion after on_removal txn
- The removal is normally so slow that these don't really race, but they
could.
Signed-off-by: Sage Weil <sage@inktank... - 02:15 AM Revision 80fb336b (ceph): Merge pull request #900 from ceph/wip-mon-mds-trim
- mon: MDSMonitor: trim versions and let PaxosService decide whether to propose
We were not trimming mdsmap versions a... - 01:06 AM Revision c98c1043 (ceph): tools: ceph-kvstore-tool: output value contents to file on 'get'
- Signed-off-by: Joao Eduardo Luis <joao.luis@inktank.com>
- 01:06 AM Revision 47ee7970 (ceph): mon: ceph-kvstore-tool: get size of value for prefix/key
- Signed-off-by: Joao Eduardo Luis <joao.luis@inktank.com>
- 01:06 AM Revision 00048fe3 (ceph): mon: Have 'ceph report' print last committed versions
- Only for those services that weren't doing it.
Backport: dumpling
Backport: emperor
Signed-off-by: Joao Eduardo Lui... - 01:06 AM Revision cc643828 (ceph): mon: MDSMonitor: let PaxosService decide on whether to propose
- Signed-off-by: Joao Eduardo Luis <joao.luis@inktank.com>
12/05/2013
- 11:48 PM Revision 856f8344 (ceph): Implement a watchdog for queued jobs
- This continually posts the run's status to the results server, if
configured, at an interval defaulting to 600 second... - 11:33 PM Revision 58231460 (ceph): os/ObjectStore: make getattrs() pure virtual
- It is required.
Signed-off-by: Sage Weil <sage@inktank.com> - 09:05 PM Revision 11e26ee4 (ceph): s/true/1 and s/false/0
- Signed-off-by: tamil <tamil.muthamizhan@inktank.com>
- 05:47 PM Revision cf099415 (ceph): mon: MDSMonitor: implement 'get_trim_to()' to let the mon trim mdsmaps
- This commit also adds two options to the MDSMonitor:
- mon_max_mdsmap_epochs: the maximum amount of maps we'll kee... - 05:41 PM Revision 032a00bb (ceph): mon: MDSMonitor: remove reference to no-longer-used encode_trim()
- We weren't using it and it's no longer used by anyone anyway.
Signed-off-by: Joao Eduardo Luis <joao.luis@inktank.com> - 05:41 PM Revision 62fb4750 (ceph): mon: MDSMonitor: consider 'debug level' parameter on 'print_map()'
- The parameter was there, just not used. It does default to 7, so
existing callers are okay.
Signed-off-by: Joao Edu... - 05:41 PM Revision 3e845b56 (ceph): mon: MDSMonitor: print map on encode_pending() iff debug mon = 30+
- Signed-off-by: Joao Eduardo Luis <joao.luis@inktank.com>
- 05:37 PM Bug #6938 (Resolved): client full handling races with OSDs having newer maps
- Currently osds reply to writes with -ENOSPC when the full flag in the osd map is set. If a client with an older map w...
- 05:18 PM Revision 39ddd213 (ceph): Merge pull request #899 from dachary/wip-crush-unittest
- CrushWrapper::insert_item unittest and minor fixes
Reviewed-by: Sage Weil <sage@inktank.com> - 05:07 PM Revision b9bff8e8 (ceph): crush: remove redundant test in insert_item
- A year after the last modification of test to check if an item was added
twice to the same bucket, the subtree_contai... - 05:07 PM Revision ccc60145 (ceph): crush: CrushWrapper unit tests
- Covers all cases for the following methods. All but insert_item are trivial.
* insert_item
* set_item_name
* name_ex... - 05:06 PM Revision 8af75968 (ceph): crush: insert_item returns on error if bucket name is invalid
- A bucket name may be created as a side effect of insert_item. All names
in the loc argument are checked for validity ... - 04:06 PM devops Bug #6937 (Resolved): udev: OSD using dmcrypt aren't automatically started
- When a storage node using dmcrypt for OSDs is started, the OSDs do not get started.
Running ceph-disk activate afte... - 02:52 PM rgw Bug #6936 (Resolved): radosgw-admin allows same user to be created multiple times
- radosgw-admin user create allows same user to be created multiple times but with different access_key and secret_key....
- 01:51 PM RADOS Feature #6935 (New): Exceeding pool quota should be more graceful
- We should try to make failures more graceful or display warnings when quota is being approached. Currently for exampl...
- 01:09 PM Bug #6685 (In Progress): osd/ReplicatedPG.cc: 8345: FAILED assert(0 == "erroneously present object")
- log: ubuntu@teuthology:/a/teuthology-2013-12-04_23:00:01-rados-master-testing-basic-plana/130920
this seems to be ... - 12:26 PM devops Bug #5475 (Closed): ceph-deploy deb install breaks on i386 due to missing python-pushy package
- This is no longer happening and from ceph-deploy v1.3 we no longer depend on pushy
- 11:27 AM rbd Feature #6934 (New): Consistency check when importing RBD diff
- It would be very usefull to have some sort of meganism in place which would ensure that the image you're importing a ...
- 01:49 AM Revision 42119261 (ceph): A create_if_vm call was made more than once when a lock-many style lock
- was performed. This caused downburst to run twice, and the second
downburst fails as a result of the first downburst... - 01:36 AM Revision 207c910e (ceph): Merge branch 'teuthology-fix-downburst-yaml-wusui'
- 01:31 AM Revision 94f7dd1f (ceph): Implement --downburst-conf parameter for teuthology-lock.
- Load the appropriate yaml information when found (this formerly
did not work). Make sure teuthology --lock works wit... - 01:01 AM Revision f8b59659 (ceph): added fs:xfs to upgrade suites
- Signed-off-by: tamil <tamil.muthamizhan@inktank.com>
12/04/2013
- 10:46 PM Revision 3b8371a4 (ceph): os/ObjectStore: prevent copying
- Signed-off-by: Sage Weil <sage@inktank.com>
- 10:46 PM Revision a70200e3 (ceph): os/ObjectStore: pass cct to ctor
- Signed-off-by: Sage Weil <sage@inktank.com>
- 07:42 PM Revision 0dd7d298 (ceph): Merge pull request #892 from jpds/ceph-disk-journal-mbrtogpt
- Call --mbrtogpt on journal run of sgdisk should the drive require a GPT ...
Reviewed-by: Sage Weil <sage@inktank.com... - 05:08 PM devops Bug #6698 (Resolved): new osds added to the cluster not starting up due to crush lookup failure
- tried reproducing this issue by upgrading from dumpling to emperor and then trying to add osds to the cluster but it ...
- 03:52 PM devops Bug #6698: new osds added to the cluster not starting up due to crush lookup failure
- Sage suspects this may have been fixed by 1bb5aadbca80eb46d5393bb1ad68af3da025e157
- 04:17 PM Bug #6756: journal full hang on startup
- d8d27f13e11dcaefd3aa1c049b97c980283da575 was my first attempt.
Reverted due to bugginess in 703f9a09e2449712a99f0865... - 03:42 PM Revision ea600d0e (ceph): Merge pull request #782 from danchai/master
- ObjBencher: add rand_read_bench to support rand test in rados-bench
- 01:57 PM Revision 35011e0b (ceph): Call --mbrtogpt on journal run of sgdisk should the drive require a GPT...
- Signed-off-by: Jonathan Davies <jonathan.davies@canonical.com>
Reviewed-by: Sage Weil <sage@inktank.com>
Reviewed-by:... - 12:40 PM devops Tasks #6932 (Rejected): Create a tar.gz file with packages per-distro
- In issue #6639 ceph-deploy is using a local directory to push packages to a directory on remote hosts and point the p...
- 07:41 AM Revision cae10830 (ceph): ObjBencher: add rand_read_bench functions to support rand test in rados...
- Signed-off-by: Tengwei Cai <tengweicai@gmail.com>
- 06:46 AM Revision e8298592 (ceph): doc/rados/operations/crush: fix more
- Signed-off-by: Sage Weil <sage@inktank.com>
- 06:18 AM Revision 7709a10f (ceph): doc/rados/operations/crush: fix rst
- Signed-off-by: Sage Weil <sage@inktank.com>
- 05:41 AM Revision bd08d332 (ceph): OSD: allow project_pg_history to handle a missing map
- If we get a peering message for an old map we don't have, we
can throwit out: the sending OSD will learn about the ne... - 05:41 AM Revision 9875c8b1 (ceph): init, upstart: prevent daemons being started by both
- There can be only one init system starting a daemon. If there is a
host entry in ceph.conf for a daemon, sysvinit wou... - 02:16 AM Revision 8543d0ed (ceph): Added docstrings. Cleaned up code (broke up long lines, removed unused
- variable references, pep8 formatted most of the code (one set of long lines
remains), and changed some variable and m... - 01:49 AM CephFS Bug #6930 (Resolved): Folders report "Not a folder" in Kernel 3.13-rc1 and -rc2
- Seems to be a regression in the 3.13-rc1 and -rc2 kernels
When you mount a filesystem is mounts fine. dmesg is cle... - 01:31 AM Revision 5cc60996 (ceph): rbd: make default size larger for xfstests
- Test 167 runs out of space on newer kernels
Signed-off-by: Josh Durgin <josh.durgin@inktank.com> - 01:02 AM Revision 68fdcfa1 (ceph): FileSTore: do not time out threads while they're waiting for op throttle
- In order to support this, we pass TPHandle* through the ObjectStore
interface, and then if we have one we suspend the... - 12:39 AM Revision 8cf2a84a (ceph): Merge pull request #893 from jdurgin/wip-init-highlander
- init, upstart: prevent daemons being started by both
Reviewed-by: Sage Weil <sage@inktank.com> - 12:33 AM Revision c2e467b4 (ceph): init, upstart: prevent daemons being started by both
- There can be only one init system starting a daemon. If there is a
host entry in ceph.conf for a daemon, sysvinit wou... - 12:23 AM Bug #6301: ceph-osd hung by XFS using linux 3.10
- maybe the status should be changed from closed to something more appropriate, even if ceph isn't the culprit here ...
- 12:21 AM Bug #6301: ceph-osd hung by XFS using linux 3.10
- Loïc kindly generated this issue for me (lack of time, no open ID at the moment of the report... etc)
So it's obvi... - 12:05 AM Revision 7ff7cf25 (ceph): doc: Partially incorporated comments form Loic Dachary and Aaron Ten Clay.
- Signed-off-by: John Wilkins <john.wilkins@inktank.com>
12/03/2013
- 11:28 PM Revision 5e34beb6 (ceph): init, upstart: prevent daemons being started by both
- There can be only one init system starting a daemon. If there is a
host entry in ceph.conf for a daemon, sysvinit wou... - 10:41 PM Revision 0b9bb544 (ceph): crush: add a few unit tests for INDEP mode
- Signed-off-by: Sage Weil <sage@inktank.com>
- 10:41 PM Revision 94350996 (ceph): crush/mapper: strip firstn conditionals out of crush_choose, rename
- Now that indep is handled by crush_choose_indep, rename crush_choose to
crush_choose_firstn and remove all the condit... - 10:41 PM Revision 9085286f (ceph): osd, crush: add 'erasure' pool/pg type
- Signed-off-by: Sage Weil <sage@inktank.com>
- 10:41 PM Revision 6dc68f4f (ceph): osd/OSDMap: do not shift result when removing nonexistent osds
- If it is a replicated pool, remove and shift to the left. For erasure
pools, replace nonexistent items with CRUSH_IT... - 10:41 PM Revision 14138372 (ceph): mon, crush: add mode to "osd crush rule create-simple ..."
- Add a mode (firstn or indep) to the create-simple command. Make it
optional and default to firstn (for compatiblity ... - 10:41 PM Revision d51a219a (ceph): osd/osd_types: pg_pool_t: fix /// -> ///< comments
- Signed-off-by: Sage Weil <sage@inktank.com>
- 10:41 PM Revision 1b567ee0 (ceph): crush/mapper: clarify numrep vs endpos
- Pass numrep (the width of the result) separately from the number of results
we want *this* iteration. This makes thi... - 10:41 PM Revision 2731d303 (ceph): crush/mapper: pass parent r value for indep call
- Pass down the parent's 'r' value so that we will sample different values in
the recursive call when the parent tries ... - 10:41 PM Revision c43c893b (ceph): crush/mapper: new SET_CHOOSE_LEAF_TRIES command
- Explicitly control the number of sample attempts, and allow the number of
tries in the recursive call to be explicitl... - 10:41 PM Revision 1cfe140b (ceph): crush: eliminate CRUSH_MAX_SET result size limitation
- This is only present to size the temporary scratch arrays that we put on
the stack. Let the caller allocate them as ... - 10:41 PM Revision b1d4dd4e (ceph): crush: return CRUSH_ITEM_UNDEF for failed placements with indep
- For firstn mode, if we fail to make a valid placement choice, we just
continue and return a short result to the calle... - 10:41 PM Revision 86e97803 (ceph): crush: use breadth-first search for indep mode
- Signed-off-by: Sage Weil <sage@inktank.com>
- 10:41 PM Revision 7967ae0e (ceph): crush/CrushWrapper: add dump_tree()
- This is based on the OSDMap::print_tree(), but is a bit simpler because
we do not have up/down information at this le... - 10:41 PM Revision 4551fee9 (ceph): crush/mapper: add note about r in recursive choose
- Signed-off-by: Sage Weil <sage@inktank.com>
- 08:56 PM Revision e94ca0bd (ceph): Merge pull request #891 from ceph/wip-filestore-remount
- allow various threaded services to be restarted to allow mount->umount->mount to work
Reviewed-by: Greg Farnum <greg... - 08:47 PM Revision 8f377cc1 (ceph): Merge pull request #879 from xinglin/coverity-fixes
- Coverity fixes
The rgw patch is
Reviewed-by: Yehuda Sadeh <yehuda@inktank.com>
The rest is
Reviewed-by: Sage Weil <... - 08:27 PM Revision cd6be292 (ceph): common/Finisher: allow finisher to be restarted
- Allow a start() after stop() by clearing the stop flag on a successful
stop.
Signed-off-by: Sage Weil <sage@inktank.... - 08:27 PM Revision 36505e85 (ceph): os/FileStore: allow mount after umount
- Clear the stop flag on successful shutdown so that we can start again
later.
Signed-off-by: Sage Weil <sage@inktank.... - 08:27 PM Revision 858a21b9 (ceph): common/WorkQueue: allow start() after stop()
- Signed-off-by: Sage Weil <sage@inktank.com>
- 07:41 PM Bug #6781: timed out waiting for recovery - probably ceph command hang
- So procedurally what's the right course here? Would the behavior have had anything to do with the branch you were te...
- 02:26 PM Bug #6781: timed out waiting for recovery - probably ceph command hang
- That sounds like a different bug then. We'd want logs (osd20, filestore20, ms1) to reproduce.
- 07:07 PM Revision 941875de (ceph): doc: Partially incorporated comments from Aaron Ten Clay.
- Signed-off-by: John Wilkins <john.wilkins@inktank.com>
- 07:02 PM devops Bug #6698 (Need More Info): new osds added to the cluster not starting up due to crush lookup fai...
- I guess I'm not sure what the expected behavior is here; do we think that the OSDs should start, but not do much, unt...
- 06:57 PM Revision d8ccd739 (ceph): OSDMonitor: use a different approach to prevent extreme multipliers on ...
- Signed-off-by: Greg Farnum <greg@inktank.com>
- 06:48 PM Revision f60f5fe9 (ceph): Merge pull request #843 from ceph/wip-osd-metadata
- feed osd info about os, kernel, memory, arch to the mons
Reviewed-by: Joao Eduardo Luis <joao.luis@inktank.com>
Revi... - 06:39 PM Revision 9770d5d8 (ceph): Merge pull request #860 from dachary/wip-6827
- Reviewed-by: Josh Durgin <josh.durgin@inktank.com>
- 06:19 PM Revision 9c8df349 (ceph): os/WBThrottle: added check of return value for posix_fadvise()
- check the return value of posix_fadvise() in WBThrottle::entry()
Signed-off-by: Xing Lin <xinglin@cs.utah.edu> - 06:19 PM Revision 0fd3cbf6 (ceph): rgw/rgw_op: added check of return value for rgw_unlink_bucket()
- checked the return value of rgw_unlink_bucket() in RGWCreateBucket::execute()
Signed-off-by: Xing Lin <xinglin@cs.ut... - 06:19 PM Revision dc06e23e (ceph): test/mon/workloadgen: added check of return value for statfs()
- check the return value of statfs() in update_osd_stat()
Signed-off-by: Xing Lin <xinglin@cs.utah.edu>
test/mon/work... - 06:18 PM Revision de097785 (ceph): test/filestore: add check of return values in StoreTest::SetUp
- add return value check for function calls, including mkdir, mkfs
and mount
Signed-off-by: Xing Lin <xinglin@cs.utah.... - 06:18 PM Revision 2884c817 (ceph): test/filestore/workloadgen: added check of return value for mkdir
- check return value of mkdir
Signed-off-by: Xing Lin <xinglin@cs.utah.edu>
test/filestore/workloadgen: fix return va... - 04:13 PM Revision 6ac9269d (ceph): Merge pull request #889 from dachary/wip-osd-metadata
- depends on #843 : CPU info + check osd metadata
- 02:39 PM Revision 602d5d53 (ceph): osd: include CPU info
- It will be handy when trying to figure out CPU bound problems / benchmarks.
Signed-off-by: Loic Dachary <loic@dachar... - 02:27 PM Bug #6756 (Need More Info): journal full hang on startup
- What's the status on this? Last comment was "Testing fix on wip-queueing2"...does that mean that's when this bug man...
- 02:18 PM Bug #6903: nightlies: upgrade tests hung on ceph quorum status
- logs: ubuntu@teuthology:/a/teuthology-2013-12-02_19:40:02-upgrade-parallel-next-testing-basic-plana/127541
log: ub... - 02:12 PM Revision f5025307 (ceph): pybind: syntax check osd metadata
- Signed-off-by: Loic Dachary <loic@dachary.org>
- 11:12 AM Bug #6827 (Resolved): ceph-disk hangs on blkid -s TYPE /dev/fd0
- "merged":https://github.com/ceph/ceph/commit/9770d5d84e26b03f99e29d21cfc5337912712414
- 10:02 AM devops Feature #6836: Get all Ceph installation package dependencies from EPEL etc, into either ceph or ...
- Can you also generate a list of the dependencies from the base-distribution that need to be installed for Ceph?
Some... - 09:36 AM Bug #6922: mon: do not allow unreasonable PG splitting
- Hrm, a customer has asked that we backport this patch to dumpling as well; is that sort of UI change appropriate on a...
- 08:31 AM Revision 65701978 (ceph): ceph-disk: blacklist /dev/fd0
- blkid -s TYPE /dev/fd0 has been verified to hang forever on a
H8DMR-82 supermicro motherboard running
3.8.0-33-gener... - 06:36 AM Revision bcb6bfd2 (ceph): test/ceph_decoder: add return value check for read_file()
- check the return value of read_file function call
Signed-off-by: Xing Lin <xinglin@cs.utah.edu> - 06:15 AM Revision 20507b5f (ceph): crush/CrushWrapper: default to the "new" bobtail-era tunables
- These are supported as of kernel version v3.9:
- chooseleaf_descend_once was added in 1604f48
- others were added i... - 06:12 AM Revision 07978bb2 (ceph): crush/CrushWrapper: explicitly set tunables on create, decode
- Set the tunables to the 'default' values on create.
Default to the legacy values when we decode, so that an old map ... - 05:57 AM Revision 8812c4e9 (ceph): rgw: don't error out on empty owner when setting acls
- Fixes: #6892
Backport: dumpling, emperor
s3cmd specifies empty owner field when trying to set acls on object
/ bucket... - 05:55 AM Revision 031b9ed5 (ceph): Merge pull request #877 from ceph/wip-6892
- rgw: don't error out on empty owner when setting acls
Reviewed-by: Sage Weil <sage@inktank.com> - 05:48 AM Revision 3f1352ff (ceph): doc/rados/operations/crush-map: describe crush locations, hook
- Signed-off-by: Sage Weil <sage@inktank.com>
- 05:46 AM Revision 396ee0d4 (ceph): add script/run-coverity
- Keep this in-tree.
Signed-off-by: Sage Weil <sage@inktank.com> - 02:04 AM Bug #6301: ceph-osd hung by XFS using linux 3.10
- I confirm, we had (several times since this cluster setup) a similar problem here. It seems related to some memory co...
- 02:01 AM Bug #6301: ceph-osd hung by XFS using linux 3.10
- Emmanuel Lacour ran into the same problem today and "sent a bug report":http://oss.sgi.com/pipermail/xfs/2013-Decembe...
- 01:59 AM Bug #6301: ceph-osd hung by XFS using linux 3.10
- This really is a XFS problem that should be reported as follows : http://xfs.org/index.php/XFS_FAQ#Q:_What_informatio...
- 01:36 AM Revision 283de99c (ceph): Merge pull request #887 from jdurgin/wip-peek-meta
- osd: read into correct variable for magic string
Reviewed-by: Sage Weil <sage@inktank.com> - 12:25 AM Revision 01ab8e63 (ceph): OSDMap: use unsigned instead of int for OSD counts
- These functions can never return <0, so use unsigned.
Signed-off-by: Greg Farnum <greg@inktank.com>
12/02/2013
- 11:26 PM Revision a5eda4fc (ceph): doc: Updated scrub defaults and added note about deep scrubbing.
- Signed-off-by: John Wilkins <john.wilkins@inktank.com>
- 11:16 PM Revision f11b380e (ceph): osd: read into correct variable for magic string
- 4d140a71a1a48081b449b7d8dde808eb6e74c6b2 refactored this and
introduced a bug. peek_meta() was accidentally reading i... - 11:13 PM Revision f57dad64 (ceph): OSDMonitor: prevent extreme multipliers on PG splits
- Fixes: #6922
Backport: emperor
Signed-off-by: Greg Farnum <greg@inktank.com> - 11:13 PM Revision 26c00c5b (ceph): OSDMonitor: return EEXIST if users try to do a PG merge
- This was returning a machine-code success if users tried to merge, and
was erroneously (but harmlessly) increasing th... - 10:02 PM Bug #6827: ceph-disk hangs on blkid -s TYPE /dev/fd0
- Sage Weil wrote:
> Loic Dachary wrote:
> > Motherboard "H8DMR-82":http://www.supermicro.com/Aplus/motherboard/Opter... - 10:02 PM Bug #6827: ceph-disk hangs on blkid -s TYPE /dev/fd0
- Loic Dachary wrote:
> Motherboard "H8DMR-82":http://www.supermicro.com/Aplus/motherboard/Opteron2000/MCP55/H8DMR-82.... - 09:56 PM rgw Fix #6892 (Resolved): rgw: ignore empty owner in set acl api
- 09:26 PM Revision 4fdc5d9e (ceph): workunits: use integers instead of true/false for hashpspool command
- Backport: emperor
Signed-off-by: Greg Farnum <greg@inktank.com>
Reviewed-by: Samuel Just <sam.just@inktank.com - 09:12 PM rgw Bug #6919: rgw: copy of some zero sized objects failure
- Zero sized objects that were created by multipart upload cannot be coppied. This only affects bobtail and cuttlefish.
- 12:49 PM rgw Bug #6919 (Resolved): rgw: copy of some zero sized objects failure
- 09:07 PM Revision cc71fae7 (ceph): Merge pull request #881 from ceph/wip-objectstore
- osd: clean up the osd / objectstore separation
Reviewed-by: Samuel Just <sam.just@inktank.com> - 08:52 PM Revision 7b392a81 (ceph): white listed "wrongly marked me down" to avoid false alarm in the night...
- Signed-off-by: tamil <tamil.muthamizhan@inktank.com>
- 07:16 PM rbd Bug #6926 (Resolved): rbd: diff output includes previously non-existent objects as zeroed extents
- For objects that exist in the current image, but did not in previous snapshots, extra 'zero' extents are listed.
T... - 06:49 PM Revision c77ce90f (ceph): doc: Fixed hyperlink to the manual installation section.
- Signed-off-by: John Wilkins <john.wilkins@inktank.com>
- 06:48 PM Revision 648f3bcd (ceph): doc: Added a link to get packages.
- Signed-off-by: John Wilkins <john.wilkins@inktank.com>
- 06:21 PM Revision 16b75768 (ceph): osd/OSDMap: fix typo and crush types helper
- Bah!
Signed-off-by: Sage Weil <sage@inktank.com> - 06:19 PM Revision 1242642d (ceph): Merge pull request #882 from ceph/wip-6699
- sysvinit, upstart: use df -P when weighting new OSDs
Reviewed-by: Noah Watkins <noah.watkins@inktank.com> - 06:10 PM Revision a4254f7f (ceph): Merge pull request #876 from javacruft/wip-missed-module
- Mark libcls_kvs as a module
Reviewed-by: Sage Weil <sage@inktank.com> - 05:59 PM Revision 5274b3ef (ceph): Merge pull request #841 from ceph/wip-crush-chassis
- osd/OSDMap: add 'chassis' to default type hierarchy
Reviewed-by: Loic Dachary <loic@dachary.org> - 05:24 PM Revision 78df4a14 (ceph): Merge pull request #883 from idryomov/wip-readme
- README: update a list of build deps for rpm-based systems
Reviewed-by: Sage Weil <sage@inktank.com> - 04:31 PM Revision 6298a576 (ceph): osd/OSDMap: add region, pdu, pod types while we are at it
- One use noted that they have a 'pdu' type in their type hierarchy that
typically spans multiple racks. Others are kn... - 04:28 PM Feature #6924 (New): Add a new Info state to ceph health
- Currently, a transient issue, such as a PG not active/clean or an OSD down, causes the ceph status to show a warning....
- 04:25 PM Revision 3a6707a1 (ceph): osd/OSDMap: add 'chassis' to default type hierarchy
- A chassis is usually bigger than a host but smaller than a rack. This will
be useful for a broad class of modern har... - 04:24 PM RADOS Feature #6923 (New): OSDMonitor: maintain pool->OSD and OSD->pool mappings
- It would be nice if the monitor had explicit mappings from OSDs->pools and back again. You can work this out by itera...
- 03:44 PM Bug #6922 (Fix Under Review): mon: do not allow unreasonable PG splitting
- https://github.com/ceph/ceph/pull/886
- 02:29 PM Bug #6922 (Resolved): mon: do not allow unreasonable PG splitting
- We've had two different kinds of issues with PG splits:
1) the incorrectly-interpreted PG values from the Dumpling->... - 03:36 PM devops Feature #6836 (Resolved): Get all Ceph installation package dependencies from EPEL etc, into eith...
- Same package list as for emperor (see #6590).
- 02:40 PM Bug #6920 (Resolved): kernel package missing in gitbuilder for ubuntu?
- There was a checkin on 11/28 that triggered a build creating sha1 5a698e686e3cfd2052b57ca00889d03bf361cc38. The kern...
- 02:10 PM Bug #6920: kernel package missing in gitbuilder for ubuntu?
- ...
- 02:09 PM Bug #6920 (Resolved): kernel package missing in gitbuilder for ubuntu?
- logs: ubuntu@teuthology:/a/teuthology-2013-11-27_19:35:21-upgrade-small-next-testing-basic-vps/120858
- 02:29 PM Fix #6908 (Rejected): OSD: do not do PG creates in-line with incoming request
- 01:59 PM rgw Bug #6911: rgw test failure on the arm set up
- log: ubuntu@teuthology:/a/teuthology-2013-12-02_04:10:14-rgw-master---basic-saya/127052
- 01:57 PM rgw Bug #6913: valgrind issues when running rgw tests
- logs: ubuntu@teuthology:/a/teuthology-2013-12-02_04:10:14-rgw-master---basic-saya/127068
- 01:17 PM Bug #6097: btrfs locking regression on async snap ioctl
- It looks like it's been fixed upstream. Commit
b1a06a4b574996692b72b742bf6e6aa0c711a948, went into 3.13-rc1. - 10:18 AM devops Bug #6699 (Resolved): sysvinit script setting incorrect OSD weights
- 09:54 AM RADOS Feature #6918 (New): Extensions to 'ceph osd metadata'
In reference to: https://github.com/ceph/ceph/pull/843
Some extra bits of information would be useful in the OSD...- 09:52 AM Revision 5e1fc147 (ceph): README: update a list of build deps for rpm-based systems
- A list of build dependencies for rpm-based OSes is missing a few items.
This fills in the gaps.
Signed-off-by: Ilya ... - 06:47 AM Revision f5ddb12a (ceph): Merge remote-tracking branch 'gh/next'
- 06:44 AM CephFS Bug #5036: `ls` hangs on random folder
- Herbert XYZ wrote:
> Our MDS log (with MDS in dubug mode) did output this:
> [inode 10000013022 [2,head] /path/to/f... - 02:42 AM CephFS Bug #5036: `ls` hangs on random folder
- Our MDS log (with MDS in dubug mode) did output this:
[inode 10000013022 [2,head] /path/to/frozen/file auth v34 ap=6... - 02:28 AM CephFS Bug #5036: `ls` hangs on random folder
- I would like to state that we also had this problem, but were unable to reproduce it. A folder got into a 'state' suc...
- 06:21 AM Revision e80ab94b (ceph): Partial revert "mon: osd pool set syntax relaxed, modify unit tests"
- This reverts commit 08327fed8213a5d24cd642e12b38a171b98924cb, except
for the hashpspool bit. We switched back to an ... - 05:36 AM CephFS Bug #6914: Accessing specific file on CephFS causes client hang
- As an addition to the problem Rolf is experiencing, please note this:
The 'ls' process got the 'state' D as report... - 05:03 AM Feature #4929: Erasure encoded placement group
- Add links to the Firefly summit to the description
12/01/2013
- 09:58 PM devops Feature #6836 (In Progress): Get all Ceph installation package dependencies from EPEL etc, into e...
- 09:58 PM devops Bug #6590 (Resolved): Ceph Package Dependencies not Included in Ceph Extras
11/30/2013
- 09:44 PM Revision f50371f0 (ceph): Merge remote-tracking branch 'gh/next'
- 09:41 PM Revision fd175ab9 (ceph): sysvinit, upstart: use df -P when weighting new OSDs
- This avoids parsing out the wrong value when a long device name makes
df wrap over two lines.
Fixes: #6699
Reported-... - 06:29 AM CephFS Bug #6914: Accessing specific file on CephFS causes client hang
- Well, actually now the client hang seems to be gone. Still the MDS seems to attribute the locks to the old client ID's.
- 06:25 AM CephFS Bug #6914 (Resolved): Accessing specific file on CephFS causes client hang
- When accessing a specific file (/disco/herbert/disco-data/output/video-conversion) the client hangs. Any kind of fsta...
- 06:28 AM Revision 41056e52 (ceph): osd: use ObjectStore::create() helper to create ObjectStore impl
- No more knowledge of FileStore!
Signed-off-by: Sage Weil <sage@inktank.com> - 06:28 AM Revision e0f20892 (ceph): osd: construct ObjectStore outside of OSD
- This lets ceph_osd.cc handle the config details and use it directly for
all of the random command-line stuff, elimina... - 06:28 AM Revision bf5a7b24 (ceph): os/ObjectStore: generalize the FileStore's target_version
- This kills the last FileStore reference in the OSD.
Signed-off-by: Sage Weil <sage@inktank.com> - 06:28 AM Revision 6bb42a0c (ceph): osd: move peek_journal_fsid() into ObjectStore (from OSD)
- This kills the last FileJournal reference in the OSD!
Signed-off-by: Sage Weil <sage@inktank.com> - 06:28 AM Revision 4d140a71 (ceph): os/ObjectStore: add {read,write}_meta
- Move these from the OSD. Use a generic implementation in ObjectStore that
hopefully all backends can share (so that ... - 06:28 AM Revision 6d92a8ef (ceph): osd: drop obsolete 'filestore' bool option
- Signed-off-by: Sage Weil <sage@inktank.com>
- 06:28 AM Revision 237d6b83 (ceph): os/ObjectStore: add static create() method
- Generic way to create an ObjectStore implementation of the required type,
so that users don't need to know anything a... - 06:26 AM Revision ea9dabf7 (ceph): mon/OSDMonitor: include osd metadata in 'ceph report'
- Signed-off-by: Sage Weil <sage@inktank.com>
- 06:26 AM Revision 2974eb5a (ceph): osd: include misc osd config info in metadata sent to the mon
- addr info, pathnames.
Signed-off-by: Sage Weil <sage@inktank.com> - 06:26 AM Revision c4f817f9 (ceph): mon/OSDMonitor: move osd metadata dump into a helper
- Signed-off-by: Sage Weil <sage@inktank.com>
- 06:25 AM Revision 51cd7508 (ceph): Merge pull request #838 from xinglin/boost-program-option-check
- autoconf: add check for the boost_program_options library
Reviewed-by: Noah Watkins <noah.watkins@inktank.com> - 06:24 AM Revision 444ddaef (ceph): Merge pull request #829 from ceph/port/detect-clang
- conf: use better clang detection
11/29/2013
- 08:05 AM rgw Bug #6913: valgrind issues when running rgw tests
- the following one appears to be similar.
logs: ubuntu@teuthology:/a/teuthology-2013-11-27_04:15:17-rgw-next---basi... - 07:57 AM rgw Bug #6913: valgrind issues when running rgw tests
- not sure, if it is a issue with the test or valgrind. filing this bug to just keep track of it.
- 07:56 AM rgw Bug #6913 (Duplicate): valgrind issues when running rgw tests
- logs: ubuntu@teuthology:/a/teuthology-2013-11-27_04:15:17-rgw-next---basic-saya/120829...
11/27/2013
- 11:17 PM Revision 838b6c83 (ceph): PG: don't query unfound on empty pgs
- When the replica responds, it responds with a notify
rather than a log, which the primary then ignores since
it is al... - 11:17 PM Revision 964c8e97 (ceph): PG: retry GetLog() each time we get a notify in Incomplete
- If for some reason there are no up OSDs in the history which
happen to have usable copies of the pg, it's possible th... - 10:45 PM Revision 8cd33e3a (ceph): rgw: bucket meta remove don't overwrite entry point first
- Fixes: #6056
When removing a bucket metadata entry we first unlink the bucket
and then we remove the bucket entrypoi... - 09:36 PM Revision 14cf4caf (ceph): rgw: don't error out on empty owner when setting acls
- Fixes: #6892
Backport: dumpling, emperor
s3cmd specifies empty owner field when trying to set acls on object
/ bucket... - 05:18 PM rbd Bug #6775 (Rejected): kvm backtrace on rbd task
- This looks unrelated to ceph. The machines can be nuked.
- 05:00 PM rbd Bug #6800 (Resolved): rbd/qemu-iotests.sh Failing on Arm.
- This fails due to using a different version of qemu, which has slightly different output.
cherry-picked removal from... - 04:12 PM Bug #6902 (Need More Info): upgrade-parallel test failed in the nightlies
- All the logs indicate is that an exit status of 1 came from ceph_test_rados. There are no other apparent messages in...
- 03:31 PM rgw Bug #6899: rgw: delete bucket on secondary zone returns success, but bucket is not deleted
- This is issue #6056 which was fixed post dumpling. Commit:0373d749cea7d9b532069ba8ebca2f005b2c9f59 fixes it, we'll ne...
- 03:26 PM rgw Bug #6911 (Won't Fix): rgw test failure on the arm set up
- logs: ubuntu@teuthology:/a/teuthology-2013-11-27_04:15:17-rgw-next---basic-saya/120813...
- 03:15 PM Bug #6910 (Pending Backport): don't query empty osds for unfound
- 03:15 PM Bug #6910 (Resolved): don't query empty osds for unfound
- 03:14 PM Bug #6909 (Resolved): Incomplete state should retry on Notify
- 02:07 PM Fix #6908 (Need More Info): OSD: do not do PG creates in-line with incoming request
- I'm incorrect about this; it's not a single transaction but rather one per PG. Will try and get more details.
- 02:02 PM Fix #6908 (Rejected): OSD: do not do PG creates in-line with incoming request
- We got a report that splitting a pool from 8 PGs to 2040 PGs caused a number of OSDs in the cluster to go down tempor...
- 01:36 PM rgw Fix #6892 (Fix Under Review): rgw: ignore empty owner in set acl api
- 09:06 AM Revision 7ec21ed8 (ceph): Mark libcls_kvs as a module
- This was missed by the changes in commit 909a733fe7119fc1bf123b3739e3acc6f2135b6a
Signed-off-by: James Page <james.p... - 08:36 AM rbd Feature #4550: Create Qemu+RBD rpm package for RHEL+CentOS 6.3 on ceph.com
- I know this is marked as resolved, but there does not appear to be a process to ensure packages are built against new...
11/26/2013
- 11:09 PM Revision ff60c364 (ceph): Merge pull request #859 from ceph/wip-pr781
- fix the bug ctypes.util.find_library to search for librados failed on...
- 11:00 PM Revision 909a733f (ceph): Merge pull request #864 from javacruft/wip-debian-feedback
- Misc fixes from current Debian packaging
Reviewed-by: Sage Weil <sage@inktank.com> - 10:59 PM Revision 17b495ab (ceph): Merge pull request #873 from ceph/wip-mds-assert
- mds: Add assertion to catch object mutation error
Reviewed-by: Sage Weil <sage@inktank.com> - 07:46 PM Bug #6905 (Duplicate): nightlies: failed to become clean before timeout expired
- this happened running rados test on master branch
logs: ubuntu@teuthology:/a/teuthology-2013-11-25_23:00:04-rados-... - 07:40 PM Revision e5357c0c (ceph): test: remove zero-copy read_fd test temporarily
- This fails since zero-copy from read_fd was temporarily disabled by
03d63c4b2d63b655924c5657637e85abdef40899.
Signed... - 07:36 PM Bug #6903 (Resolved): nightlies: upgrade tests hung on ceph quorum status
- logs: ubuntu@teuthology:/a/teuthology-2013-11-25_19:40:02-upgrade-parallel-next-testing-basic-plana/118237
- 07:35 PM Bug #6901 (Resolved): sha1 for emperor doesnt exist in gitbuilder path for fc18
- There was a comit to the emperor branch Monday 11/25 that triggered a rebuild. The build on the fedora18 gitbuilder ...
- 05:57 PM Bug #6901 (Resolved): sha1 for emperor doesnt exist in gitbuilder path for fc18
- there are nightly failures as http://gitbuilder.ceph.com/ceph-rpm-fc18-x86_64-basic/ref/emperor is missing.
Gary i... - 07:34 PM Bug #6902 (Can't reproduce): upgrade-parallel test failed in the nightlies
- the rados test failed when upgrading from dumpling to next branch.
logs: ubuntu@teuthology:/a/teuthology-2013-11-2... - 05:44 PM Revision bafb5c36 (ceph): doc: clarify crush rule create-simple and fix typos
- update the description of osd crush rule create-simple using terms that
match the keywords from the crushmap to avoid... - 05:44 PM Revision 371dc71e (ceph): doc: fix formatting typo in mon configuration
- Signed-off-by: Loic Dachary <loic@dachary.org>
Reviewed-by: Josh Durgin <josh.durgin@inktank.com> - 05:44 PM Revision 00ee9a52 (ceph): doc: improve ceph-mon usage, fix ceph tell examples
- The ceph-mon command usage is updated to document all of the ceph-mon
specific options.
The ceph tell usage examples... - 05:23 PM rgw Bug #6899: rgw: delete bucket on secondary zone returns success, but bucket is not deleted
- It looks like this is happening because the admin metadata API is reading the version of the bucket object, then call...
- 05:06 PM Revision 890bd5b3 (ceph): Merge pull request #862 from ceph/wip-6829
- Reviewed-by: Josh Durgin <josh.durgin@inktank.com>
- 05:02 PM Revision 8db8ed73 (ceph): Merge pull request #871 from dmick/wip-6894
- Reviewed-by: Josh Durgin <josh.durgin@inktank.com>
- 04:59 PM Revision 5254cbd3 (ceph): Merge pull request #872 from ceph/wip-5871
- mds: remove superfluous warning of releasing lease
Reviewed-by: Sage Weil <sage@inktank.com> - 03:59 PM Revision 57f9c252 (ceph): Merge pull request #874 from liewegas/wip-readme
- README: add yum command line
- 03:34 PM rbd Fix #5048: krbd: limit of ~230 mapped images at once
- A user on IRC had a problem with iostat related to this:
# iostat
[snip]
rbd3 0.02 0.07 ... - 03:19 PM Revision 4fa8f68f (ceph): README: add yum command line
- This makes it easy to cut and paste. Yay!
Signed-off-by: Sage Weil <sage@inktank.com> - 01:22 PM CephFS Bug #5871 (Resolved): client.4145 released lease on dn 1000000061b/f108 which dne
- by commit 09a4c1bbe9ff345fd2ab3e48ac7129ff9c1c7e0fb
- 12:06 PM Bug #6827: ceph-disk hangs on blkid -s TYPE /dev/fd0
- Motherboard "H8DMR-82":http://www.supermicro.com/Aplus/motherboard/Opteron2000/MCP55/H8DMR-82.cfm without a floppy dr...
- 10:28 AM Subtask #6900 (Fix Under Review): crush: fix indep mode
- 10:28 AM Subtask #6900 (Resolved): crush: fix indep mode
- 10:26 AM Feature #6832 (In Progress): EC: Adapt pg log to include information necessary for rollback
- 08:23 AM devops Bug #6701: ceph-deploy osd prepare on directory path fails: OSError: [Errno 18] Invalid cross-dev...
- Added a comment to that function, Hash: 2d9c452332d51f550abb2a189c1a3621a20c504a
- 07:54 AM devops Bug #6701 (Resolved): ceph-deploy osd prepare on directory path fails: OSError: [Errno 18] Invali...
- 08:13 AM devops Bug #6717 (Closed): Failure to remove samba packages
- 07:10 AM Revision b0dce8a0 (ceph): mds: Add assertion to catch object mutation error
- Signed-off-by: Yan, Zheng <zheng.z.yan@intel.com>
- 06:56 AM Revision 09a4c1bb (ceph): mds: remove superfluous warning of releasing lease
- When receiving the lease release message, it's possible that the lease
has already expired and the corresponding dent... - 06:37 AM Revision c409e362 (ceph): mon: osd dump should dump pool snaps as array, not object
- Fixes: #6894
Signed-off-by: Dan Mick <dan.mick@inktank.com> - 04:56 AM Revision 49a48ae8 (ceph): Merge branch 'wip-fix-teuth-tgt-wusui'
- 04:44 AM Revision 4c7dd504 (ceph): tgt and iscsi code need some minor fixes. Moved the settle call during
- simple read testing. In iscsi.py, generic_mkfs and generic_mount need
to be called from the main body of the task. ... - 03:35 AM Bug #6894: 'ceph osd dump' produces malformed JSON if pool snapshots are present
- Fix looks right to me
- 02:03 AM Revision 1804e136 (ceph): Merge pull request #835 from ceph/wip-6705
- qa: workunits: mon: ping.py: test 'ceph ping'
Reviewed-by: Josh Durgin <josh.durgin@inktank.com> - 01:12 AM Revision 29ef3d46 (ceph): Merge branch 'wip-zero-copy-bufferlist-last'
- Reviewed-by: Samuel Just <sam.just@inktank.com>
- 12:17 AM Bug #6897 (Duplicate): ceph osd crashed while running rados test
- Already fixed 6896 today from this run.
11/25/2013
- 11:31 PM Revision e75b2d58 (ceph): Merge pull request #154 from ceph/wip-multi-mtype
- Wip multi mtype
- 10:38 PM Revision 03d63c4b (ceph): buffer: turn off zero-copy reads for now
- Some users will need to be changed to handle getting the correct
length from bufferlist::read_fd.
Signed-off-by: Jos... - 09:59 PM Bug #6894 (Fix Under Review): 'ceph osd dump' produces malformed JSON if pool snapshots are present
- 09:32 PM Bug #6894: 'ceph osd dump' produces malformed JSON if pool snapshots are present
- This interested me, so I tried:...
- 10:33 AM Bug #6894 (Resolved): 'ceph osd dump' produces malformed JSON if pool snapshots are present
Example:...- 09:57 PM Revision 524f666e (ceph): rbd: Release resource before return
- Close file before return.
Fix coverity issue: CID 1019579
Signed-off-by: Li Wang <liwang@ubuntukylin.com>
Reported-b... - 09:57 PM Revision 784d1888 (ceph): mds: Release resource before return
- Close file before return.
Fix coverity issue: CID 1019571
Signed-off-by: Li Wang <liwang@ubuntukylin.com>
Reported-b... - 09:22 PM Revision d6c8830b (ceph): Merge pull request #868 from ceph/wip-6896
- osd: Remove bogus assert(active == acting.size())
Reviewed-by: Samuel Just <sam.just@inktank.com> - 09:20 PM Revision 66f51f82 (ceph): osd: Remove bogus assert(active == acting.size())
- We saw this assert because active is not correctly computed.
Remove assert and incorrectly computed active count.
We ... - 07:20 PM rgw Bug #6899 (Resolved): rgw: delete bucket on secondary zone returns success, but bucket is not del...
- This occured during a multi-zone sync test on dumpling. The sync agent requested a delete of the 'mybar' bucket to th...
- 06:02 PM Fix #6705 (Resolved): mon: test for ping
- commit:50868a528dd422c741e654c3f3e742be010ea1a0
- 06:00 PM Feature #6419 (Resolved): buffer: make write_fd use splice where possible
- commit:3f6fa05d7224f9b04d849c4a5309a3da01d3cb34
- 05:59 PM Feature #6417 (Resolved): buffer: make read use raw_pipe() where possible
- commit:445fb183906a56e3b412537f3756c38af667cb25 and commit:3f6fa05d7224f9b04d849c4a5309a3da01d3cb34
- 05:58 PM Feature #6416 (Resolved): buffer: create raw_pipe class; convert to buffer if inspected
- commit:5021b43e0a6ece07b8dd364ce227f60e4916f707
- 05:58 PM Feature #6415 (Resolved): buffer: instrument c_ptr() calls
- commit:75d4a72086bcfcab5c0d6834e72cbc6e473c801e
- 05:24 PM Revision 5ccc2ecd (ceph): SubmittingPatches: describe github pull-request process
- Signed-off-by: Sage Weil <sage@inktank.com>
- 05:15 PM Bug #6896 (Resolved): osd/PG.cc: 1302: FAILED assert(active == acting.size())
- 66f51f82d457f6d3170c47daed2ca3458b888df1
- 01:04 PM Bug #6896 (Fix Under Review): osd/PG.cc: 1302: FAILED assert(active == acting.size())
- 11:46 AM Bug #6896: osd/PG.cc: 1302: FAILED assert(active == acting.size())
So the "active" calculation is a minor existing bug in the way PG_STATE_DEGRADED is set. The assert() is wrong. I...- 11:22 AM Bug #6896: osd/PG.cc: 1302: FAILED assert(active == acting.size())
calc_acting() uses this to find backfill targets:...- 10:37 AM Bug #6896 (Resolved): osd/PG.cc: 1302: FAILED assert(active == acting.size())
- PG::activate() looks at actingbackfill and finds number of non-backfilling replicas. The count in "active" should be...
- 04:20 PM Revision ab05580a (ceph): Add missing stuff to clean target
- Sweepup a few bits a pieces that get left behind by make clean.
Signed-off-by: James Page <james.page@ubuntu.com> - 04:19 PM Revision ae46c38c (ceph): Correctly mark library modules
- Mark RADOS class libraries as module so that they will be automatically
ignored when stripping binaries.
Signed-off-... - 04:10 PM Revision 94ca1cc6 (ceph): ceph-object-corpus: revert accidental revert
- From 5df3024d7847950672169069b9dc8c8dcb876019
Signed-off-by: Sage Weil <sage@inktank.com> - 03:55 PM Revision a24f1326 (ceph): rgw: lower some debug message
- Fixes: #6084
Backport: dumpling, emperor
Reported-by: Ron Allred <rallred@itrefined.com>
Signed-off-by: Yehuda Sadeh... - 03:54 PM Revision 69e055f3 (ceph): rgw: lower some debug message
- Fixes: #6084
Backport: dumpling, emperor
Reported-by: Ron Allred <rallred@itrefined.com>
Signed-off-by: Yehuda Sadeh... - 03:53 PM Revision fd1ec0fa (ceph): Merge pull request #863 from ceph/wip-6804
- rgw: lower some debug message
Reviewed-by: Sage Weil <sage@inktank.com> - 12:06 PM Bug #6897 (Duplicate): ceph osd crashed while running rados test
- logs: ubuntu@teuthology:/a/teuthology-2013-11-24_23:00:03-rados-master-testing-basic-plana/117237...
- 10:36 AM Subtask #6895 (Rejected): omap needs to be disablable on a per-pool basis, this needs to be requi...
- 09:50 AM rgw Fix #6892 (Resolved): rgw: ignore empty owner in set acl api
- s3cmd sends an empty owner field when setting acls. Currently radosgw rejects this since it expects the owner to matc...
- 09:21 AM Bug #6777 (Can't reproduce): nightlies: gem dependency error
- 09:19 AM Revision c0297b43 (ceph): Changes suggested per review.
- Signed-off-by: Sandon Van Ness <sandon@inktank.com>
- 07:54 AM rgw Fix #6804 (Resolved): Overly verbose logging: "setting object write_tag=" - rgw_rados.cc
- 06:36 AM Bug #6807 (Resolved): Debian Wheezy Teuthology Ceph-deploy run failed.
- Merged into ceph-deploy's master branch with hash: 109040e
- 06:07 AM Revision b089adb8 (ceph): Merge pull request #861 from dachary/init-ceph
- init: fix typo s/{$update_crush/${update_crush/
11/24/2013
- 05:23 PM Bug #6884 (Resolved): cephtool/test.sh fails in the nightlies
- Yay. Merged into next yesterday after an email report of success.
- 07:41 AM Bug #6884: cephtool/test.sh fails in the nightlies
- the test passed when run with mon and client on the same node.
11/23/2013
- 01:09 AM Revision 0b14c155 (ceph): Merge pull request #832 from ceph/wip-flush-5855-review
- Reviewed-by: Samuel Just <sam.just@inktank.com>
Reviewed-by: David Zafman <david.zafman@inktank.com> - 12:14 AM Revision 30bc0e27 (ceph): safe_io: add functions for handling splice
- Like the other functions, these don't handle non-blocking I/O.
Signed-off-by: Josh Durgin <josh.durgin@inktank.com> - 12:14 AM Revision eb94b8fd (ceph): buffer: add an exception with an error code
- This allows e.g. raw buffer constructors to provide more specific
information about what failed, as well as a useful ... - 12:14 AM Revision 5021b43e (ceph): buffer: create raw pipe-based buffer
- This uses a pipe to reference kernel memory so we can use splice(2) to
avoid extra data copies. Take an fd in the fac... - 12:14 AM Revision 3f6fa05d (ceph): buffer: add methods to read and write using zero copy
- Create explicit methods for testing. Make buffer::list::write_fd() use
zero-copy if all the buffers support it. Don'... - 12:14 AM Revision be29b347 (ceph): buffer: attempt to size raw_pipe buffers
- Make sure the requested length is below the maximum pipe size for now,
since we're only using one pipe and splicing o... - 12:14 AM Revision 445fb183 (ceph): buffer: try to do zero copy in read_fd
- Leave the explicit read_fd_zero_copy around as well for testing.
Signed-off-by: Josh Durgin <josh.durgin@inktank.com> - 12:14 AM Revision 75d4a720 (ceph): buffer: enable tracking of calls to c_str()
- Track buffer::ptr::c_str() to catch internal calls that use it, like
buffer::ptr::cmp(). buffer::list::c_str() will b... - 12:13 AM Revision ebb261f9 (ceph): buffer: abstract raw data related methods
- Create a virtual function that returns the raw data instead of
accessing it directly, so raw buffers backed by pipes ... - 12:13 AM Revision b8518fea (ceph): buffer: remove unused raw::length() method
- This was accidentally introduced in
0c23a5624a80903fba7e635e8c44f38a79caf223
Signed-off-by: Josh Durgin <josh.durgin...
11/22/2013
- 11:49 PM Revision 740d406e (ceph): doc: Clarified description on scrub load threshold.
- fixes: #6735
Signed-off-by: John Wilkins <john.wilkins@inktank.com> - 11:48 PM Revision 8a3ed9be (ceph): doc: Fixed instruction.
- Signed-off-by: John Wilkins <john.wilkins@inktank.com>
- 11:16 PM Revision e8ccb91f (ceph): Merge branch 'master' of https://github.com/ceph/ceph
- 11:15 PM Revision fc5789da (ceph): doc: Added commentary to configure pg defaults. Clarified size commentary.
- Signed-off-by: John Wilkins <john.wilkins@inktank.com>
- 11:14 PM Revision dceaef43 (ceph): doc: PG splitting added to docs.
- fixes: #6727
Signed-off-by: John Wilkins <john.wilkins@inktank.com> - 11:13 PM Revision d39676b1 (ceph): doc: Took out "future" reference to namespaces.
- Signed-off-by: John Wilkins <john.wilkins@inktank.com>
- 11:03 PM Revision deec86c7 (ceph): Also catch httplib2.ServerNotFoundError
- Signed-off-by: Zack Cerza <zack.cerza@inktank.com>
- 08:59 PM Revision 9a55d892 (ceph): doc: Clarification of terms.
- fixes: #6749
Signed-off-by: John Wilkins <john.wilkins@inktank.com> - 06:52 PM Revision 53951fd5 (ceph): ReplicatedPG: test for missing head before find_object_context
- find_object_context doesn't return EAGAIN for a missing head.
I chose not to change that behavior since it might hide... - 05:33 PM Bug #6781: timed out waiting for recovery - probably ceph command hang
- 17:28:17.774153 ceph.log shows pgs bumped to 82 (from 72)
17:28:56.294849 last osd map change from ceph.log
17:5... - 04:44 PM Bug #6781: timed out waiting for recovery - probably ceph command hang
- please ignore the previous comment, looks like the test is still running...
- 04:24 PM Bug #6781: timed out waiting for recovery - probably ceph command hang
- ceph pg dump command hung, the test machines [plana33, plana18] hung on the nightlies are available, if someone wants...
- 05:22 PM Bug #6884: cephtool/test.sh fails in the nightlies
- Actually, it looks like mon.a and client.0 are on different hosts, and that would certainly prevent the admin socket ...
- 05:20 PM Bug #6884: cephtool/test.sh fails in the nightlies
- ...
- 05:11 PM Bug #6884: cephtool/test.sh fails in the nightlies
- when tested with wip-6884 for workunit and ceph branch, test failed. not sure why.
logs: ubuntu@mira055:/home/ubun... - 02:42 PM Bug #6884 (In Progress): cephtool/test.sh fails in the nightlies
- This is the change to the monitor API; it needs to get 0/1 instead of false/true.
- 02:37 PM Bug #6884 (Resolved): cephtool/test.sh fails in the nightlies
- logs: ubuntu@teuthology:/a/teuthology-2013-11-21_23:00:19-rados-next-testing-basic-plana/113055
This test actually... - 05:16 PM Feature #5991 (Resolved): EC: [link] Backfill peers should not be included in the acting set
- 86e4fd4f7a9848511e1bf9ae433fcd3a35a0818f
- 05:15 PM Subtask #5855 (Resolved): Backfill peers should not be included in the acting set
- merged as https://github.com/ceph/ceph/commit/86e4fd4f7a9848511e1bf9ae433fcd3a35a0818f
- 02:50 PM Subtask #5855: Backfill peers should not be included in the acting set
- 04:44 PM Feature #6419: buffer: make write_fd use splice where possible
- 04:43 PM Feature #6417: buffer: make read use raw_pipe() where possible
- 04:43 PM Feature #6416: buffer: create raw_pipe class; convert to buffer if inspected
- 04:43 PM Feature #6415: buffer: instrument c_ptr() calls
- 04:36 PM Feature #6735: scrub documentation needs to be improved (see below conversation or as Sam)
These 2 defaults in the doc don't match the defaults in the code:...- 03:51 PM Feature #6735 (Resolved): scrub documentation needs to be improved (see below conversation or as ...
- http://ceph.com/docs/master/rados/configuration/osd-config-ref/
- 03:48 PM Feature #6735 (In Progress): scrub documentation needs to be improved (see below conversation or ...
- 04:15 PM rbd Bug #5426: librbd: mutex assert in perfcounters::tinc in librbd::AioCompletion::complete()
- recent logs : ubuntu@teuthology:/a/teuthology-2013-11-21_23:01:11-rbd-next-testing-basic-plana/113257...
- 03:33 PM rgw Bug #6889 (Resolved): rgw: usage log: don't log system user operations
- 03:23 PM Documentation #6727 (Resolved): pg docs still imply pg_num must be set at creation, but we have s...
- http://ceph.com/docs/master/rados/operations/pools/#create-a-pool
http://ceph.com/docs/master/rados/operations/place... - 03:21 PM Subtask #6888 (Resolved): EC/Tiering: Disallow omap writes on pools which are backed by EC pools
- EC will not allow omap at all in the first cut. Support will require a kind of "omap_append" operation which only cr...
- 03:08 PM Revision b35fc1bc (ceph): rgw: lower some debug message
- Fixes: #6084
Backport: dumpling, emperor
Reported-by: Ron Allred <rallred@itrefined.com>
Signed-off-by: Yehuda Sadeh... - 03:06 PM Revision 561e7b0b (ceph): rgw: initialize RGWUserAdminOpState::system_specified
- Fixes: #6829
Backport: dumpling, emperor
We didn't init this member variable, which might cause that when
modifying u... - 02:27 PM devops Feature #6882 (Resolved): downburst: list available distro/versions
- Understanding what arguments to pass to teuthology --os-type --os-version, or to downburst --distro --distroversion, ...
- 02:01 PM Bug #6810 (In Progress): very high monitor memory usage after upgrade dumpling -> emperor
- 02:00 PM Bug #6806: mon: audit cmd_getval() calls to make sure they handle failures correctly
- 01:59 PM Fix #6780 (In Progress): monitor errors when checking for quorum status
- 01:39 PM rgw Feature #6193 (Resolved): rgw: DR: per bucket sync
- 01:39 PM rgw Feature #6192 (Resolved): rgw: DR: per object sync
- 01:39 PM rgw Feature #6191 (Resolved): rgw: DR: per bucket replica_log handling
- 01:38 PM rgw Feature #6190 (Resolved): rgw: DR: read list of buckets to do full sync on
- 01:38 PM rgw Feature #4341 (Resolved): rgw: dr: data sync agent: implement delta changes sync
- 01:38 PM rgw Feature #4340 (Resolved): rgw: dr: data sync agent: implement full sync
- 01:20 PM rbd Documentation #5006 (Resolved): doc: openstack configuration changes for havana
- 01:19 PM rgw Feature #6877 (Closed): tool to trim metadata, data, and bucket index logs
- This could be part of the sync agent and run automatically, or a separate tool with similar configuration.
Metadat... - 01:14 PM rgw Feature #4465 (Duplicate): Quotas per user/bucket - asynchronous
- 01:09 PM Documentation #6749 (Resolved): Documentation says CRUSH bucket weights are integers
- 01:04 PM rgw Feature #6195 (In Progress): rgw: test full sync (with large object)
- 01:03 PM rgw Feature #5170 (Resolved): RGW: Object restriping tool to fix large objects from argonaut.
- 11:17 AM Bug #6796: ceph mons interpretting pg splits very wrong
- Downgrading to Urgent, since only waiting on Backport
- 03:02 AM Bug #6796 (Pending Backport): ceph mons interpretting pg splits very wrong
- commit:337195f04653eed8e8f153a5b074f3bd48408998
- 11:06 AM rgw Feature #6874 (Resolved): rgw: user level quota: rest api
- 11:06 AM rgw Feature #6873 (Resolved): rgw: user level quota: radosgw-admin interface
- 11:05 AM rgw Feature #6872 (Resolved): rgw: user level quota: async update of bucket info in user objclass
- 11:05 AM Bug #6849 (Duplicate): Inconsistent action of ceph osd pool set commands (e.g. volume size)
- :)
If it persists after upgrading that node let us know, but I'm pretty sure it's #6796, which we'll have a point ... - 10:53 AM Bug #6849: Inconsistent action of ceph osd pool set commands (e.g. volume size)
- Well spotted, somehow one of the MON Nodes did not install Emperor.
Strange since it was reinstalled with puppet... ... - 10:43 AM Bug #6849: Inconsistent action of ceph osd pool set commands (e.g. volume size)
- Are you sure you're running all emperor monitors? (None of them are on dumpling?)
- 01:13 AM Bug #6849 (Duplicate): Inconsistent action of ceph osd pool set commands (e.g. volume size)
- Hi,
I am seeing some very weird behaviour with the osd commands.
Running on SL6.4 with Emperor (0.72.1)
Doing th... - 11:04 AM rgw Feature #6871 (Resolved): rgw: user level quota: core functionality
- A new rgw internal service that caches user level quota info. Keep track of user totals, provide info and updates it ...
- 11:00 AM rgw Feature #6870 (Resolved): rgw: user level quota: new user objclass
- New user objclass to manage user quota info
- 10:53 AM devops Bug #6698: new osds added to the cluster not starting up due to crush lookup failure
- 10:51 AM Bug #6758 (Resolved): clone_range missing head src
- 10:18 AM Bug #6834: nightlies: monitor crashed in emperor
- I was not able to reproduce this on my local cluster though.
looks like it did happen last night as well,
ubunt... - 07:26 AM Bug #6834: nightlies: monitor crashed in emperor
- A user came forward with high monitor memory usage on ticket #6810. I wonder if this is the same thing and somehow th...
- 10:11 AM rgw Feature #3454: Support temp URLs for Swift API
- 08:58 AM Bug #6807 (Fix Under Review): Debian Wheezy Teuthology Ceph-deploy run failed.
- Pull Request opened: https://github.com/ceph/ceph-deploy/pull/139
- 07:09 AM Bug #6807: Debian Wheezy Teuthology Ceph-deploy run failed.
- Just found out that before, we would just try to remove `/var/lib/ceph` and then we would check if that is still ther...
- 07:01 AM rgw Fix #6804 (Fix Under Review): Overly verbose logging: "setting object write_tag=" - rgw_rados.cc
- 06:56 AM rgw Bug #6829 (Fix Under Review): rgw: missing RGWUserAdminOpState::system_specified initialization
- 06:04 AM Revision f6b5acc0 (ceph): internal.py: nitty little spelling error
- Signed-off-by: Dan Mick <dan.mick@inktank.com>
- 05:59 AM rbd Bug #6851 (Resolved): rbd snap unprotect test failing
- Nightly tests for RBD are failing.
The specific one that keeps failing is:... - 05:46 AM Bug #6850 (Closed): mon: 'ceph health detail' with formatted output doesn't report low space on d...
- 03:01 AM Bug #6820 (Pending Backport): Bad commandline usage crashed my monitor
- commit:49d2fb71422fe4edfe5795c001104fb5bc8c98c3
- 02:56 AM Revision 203065c0 (ceph): Merge pull request #857 from ceph/wip-6796
- mon: OSDMonitor: receive CephInt on 'osd pool set' instead on CephString
Reviewed-by: Greg Farnum <greg@inktank.com> - 02:49 AM Revision b2ee9352 (ceph): PendingReleaseNotes: mention 6796 and 'ceph osd pool set' behavior change
- Signed-off-by: Joao Eduardo Luis <joao.luis@inktank.com>
- 02:49 AM Revision 7c6d43cf (ceph): doc: rados: operations: pools: document 'osd pool set foo hashpspool'
- Signed-off-by: Joao Eduardo Luis <joao.luis@inktank.com>
- 02:49 AM Revision 49d2fb71 (ceph): mon: OSDMonitor: don't crash if formatter is invalid during osd crush dump
- Code would assume a formatter would always be defined. If a 'plain'
formatter or even an invalid formatter were to be... - 02:49 AM Revision 337195f0 (ceph): mon: OSDMonitor: receive CephInt on 'osd pool set' instead on CephString
- This partially reverts 2fe0d0d9 in order to allow Emperor monitors to
forward mon command messages to Dumpling monito... - 02:48 AM Revision 7191bb2b (ceph): mon: OSDMonitor: drop cmdval_get() for unused variable
- We don't ever use any value as a float, so just drop obtaining it. This
makes it easier to partially revert 2fe0d0d9... - 01:43 AM Revision 50868a52 (ceph): qa: workunits: mon: ping.py: test 'ceph ping'
- Basic testing by forcing each monitor out of quorum at a time and making
sure they still reply to ping requests.
Fix...
11/21/2013
- 11:21 PM Revision f7af3e72 (ceph): Schedule-suite Use 'multi' tube for multiple types. Scheduling.
- Signed-off-by: Sandon Van Ness <sandon@inktank.com>
- 10:19 PM Revision c38eeec8 (ceph): Allow ability to use multi machine type deliminated by ,- \t.
- I was originally attempting a more complicated locking mechanism
but I think its almost as good to just have it attem... - 09:52 PM Revision 6b5aaf3e (ceph): doc: Minor updates to manual deployment document.
- Signed-off-by: John Wilkins <john.wilkins@inktank.com>
- 07:56 PM Revision d04f3a6a (ceph): Skip cluster() if use_existing_cluster is True
- Signed-off-by: Zack Cerza <zack.cerza@inktank.com>
- 05:51 PM Revision 3502d4f4 (ceph): init: fix typo s/{$update_crush/${update_crush/
- Signed-off-by: Dietmar Maurer <dietmar@proxmox.com>
Reviewed-by: Loic Dachary <loic@dachary.org> - 05:22 PM devops Feature #6836 (Resolved): Get all Ceph installation package dependencies from EPEL etc, into eith...
- 04:00 PM Bug #6796 (In Progress): ceph mons interpretting pg splits very wrong
- 03:40 PM Feature #6835 (Resolved): EC: ec pgs will need to be able to specify temp primaries other than ac...
- OSDMap interface must allow primariness to be specified seperately from acting set position for ec pools.
- 03:35 PM Bug #6834: nightlies: monitor crashed in emperor
- From this backtrace it looks like either there was a hardware problem, or the monitor was using so much memory it cou...
- 03:24 PM Bug #6834 (Can't reproduce): nightlies: monitor crashed in emperor
- logs: ubuntu@teuthology:/a/teuthology-2013-11-18_19:31:27-upgrade-parallel-next-testing-basic-plana/107772...
- 03:12 PM devops Feature #6310: Get Dumpling into CentOS Ceph repo
- Dumpling is not yet in the Centos repo.
I reran the mock build on centos 6.3 and 6.4 to verify that there should n... - 03:11 PM Bug #6833: `/etc/init.d/ceph status` occasionally exists silently
- Hmm. `/etc/init.d/ceph restart` is doing the same thing.
- 02:55 PM Bug #6833 (Can't reproduce): `/etc/init.d/ceph status` occasionally exists silently
- I have a cluster that got wedged somehow, and when I run `/etc/init.d/ceph status` it simply exits with status 0. Tha...
- 03:11 PM Bug #6820: Bad commandline usage crashed my monitor
- nevermind, got it.
- 01:59 PM Feature #5991: EC: [link] Backfill peers should not be included in the acting set
- 01:57 PM Feature #5990 (In Progress): EC: [link] Factor out the ReplciatedPG object replication and client...
- 01:39 PM devops Bug #6701: ceph-deploy osd prepare on directory path fails: OSError: [Errno 18] Invalid cross-dev...
- Yeah, I think a note is fine.
- 12:33 PM devops Bug #6701: ceph-deploy osd prepare on directory path fails: OSError: [Errno 18] Invalid cross-dev...
- `delete=True` is the default, and we are explicitly setting that flag to `delete=False` because of that reason.
Wo... - 01:38 PM Feature #6832 (Resolved): EC: Adapt pg log to include information necessary for rollback
- This feature includes adapting the ReplicatedBackend to allow xattr, append operations to be rollback-able. It also ...
- 01:35 PM Feature #6831 (Resolved): EC: Adapt ReplicatedPG read path to handle async reads
- This is a bit challenging because there might be a sequence of reads at different offsets within the transaction. Th...
- 12:40 PM Bug #6807 (In Progress): Debian Wheezy Teuthology Ceph-deploy run failed.
- 07:28 AM Bug #6807: Debian Wheezy Teuthology Ceph-deploy run failed.
- It sounds like there was an earlier problem with the test or a different failure — why is it trying to delete the cep...
- 07:21 AM Bug #6807: Debian Wheezy Teuthology Ceph-deploy run failed.
- It looks like the reason we were enforcing the *single file system* was because we might still have OSDs mounted (hen...
- 09:46 AM devops Feature #5282 (Closed): Get Dumpling into EPEL
- Ceph 0.67.3 dumpling is in the epel repository.
- 09:25 AM rgw Bug #6830 (Resolved): S3 CompleteMultipartUploadResult has empty ETag element
- RHEL 6.4, Ceph 0.67.
The S3 Complete Multipart Upload operation returns a result that looks like this:... - 09:12 AM Feature #6828: osd should not silently fail to start when journal partition has no UUID
- I now understand this is not a bug but a feature. It probably deserves a warning of some kind ? http://dachary.org/?p...
- 04:53 AM Feature #6828 (Rejected): osd should not silently fail to start when journal partition has no UUID
- Providing a journal that is a partition that does not have the expected journal UUID should trigger an error....
- 06:50 AM rgw Bug #6829 (Resolved): rgw: missing RGWUserAdminOpState::system_specified initialization
- By inspecting code I noticed that, which means that when modifying user configuration the system settings of the user...
- 03:29 AM Revision 29178d83 (ceph): doc: Cleanup of Add/Remove OSDs.
- Signed-off-by: John Wilkins <john.wilkins@inktank.com>
- 03:28 AM Revision 634295d1 (ceph): doc: Cleanup on Add/Remove monitor.
- Signed-off-by: John Wilkins <john.wilkins@inktank.com>
- 03:27 AM Revision de2bcd5d (ceph): doc: Added a manual deployment doc.
- Signed-off-by: John Wilkins <john.wilkins@inktank.com>
- 03:27 AM Revision 10b4bf6a (ceph): doc: Added manual deployment section to index.
- Signed-off-by: John Wilkins <john.wilkins@inktank.com>
- 03:16 AM Bug #6827 (Resolved): ceph-disk hangs on blkid -s TYPE /dev/fd0
- On a supermicro hardware running ubuntu 12.04.3 ceph-disk list hangs forever trying to *blkid -s TYPE /dev/fd0*
<pre... - 03:01 AM Bug #6826 (Duplicate): Non-equal performance of 'freshly joined' OSDs
- Was just a bare eye observation for a long time, but I`ll try to formalize it here.
For OSDs entered recently perfom... - 02:42 AM Revision 8e8cf30f (ceph): os, osd, tools: Add backportable compatibility checking for sharded obj...
- OSD
New CEPH_OSD_FEATURE_INCOMPAT_SHARDS
FileStore
NEW CEPH_FS_FEATURE_INCOMPAT_SHARDS
Add FSSuperblock with fe... - 02:38 AM Revision b493178f (ceph): Merge branch 'next'
- 02:38 AM Revision 35c91cac (ceph): os, osd, tools: Add backportable compatibility checking for sharded obj...
- OSD
New CEPH_OSD_FEATURE_INCOMPAT_SHARDS
FileStore
NEW CEPH_FS_FEATURE_INCOMPAT_SHARDS
Add FSSuperblock with fe... - 02:37 AM Revision f753d56a (ceph): test: use older names for module setup/teardown
- setUp and tearDown require nosetests 0.11, but 0.10.4 is the latest on
centos. Rename to use the older aliases, which... - 02:17 AM Revision 83ee6843 (ceph): osd: fix bench block size
- The command was declared to take 'size' in dumpling, but was trying to
read 'bsize' instead, so it always used the de... - 02:16 AM Revision 9e2d0d31 (ceph): osd: fix bench block size
- The command was declared to take 'size' in dumpling, but was trying to
read 'bsize' instead, so it always used the de... - 02:03 AM Revision bac37088 (ceph): Don't run qemu-iotests on arm as well
- Different versions of qemu don't work with the qemu-iotests we're
using. Just disable these for now.
Signed-off-by: ... - 02:03 AM Revision bf9434db (ceph): Merge pull request #153 from ceph/wip-6790
- Reviewed by Warren.
- 12:37 AM Revision c5a26b38 (ceph): Use shortened version in order to avoid revision/arch mishaps.
- Sometimes -X is added to package names which does not exist in the
/version file. Simply using the version string doe... - 12:06 AM Bug #6810: very high monitor memory usage after upgrade dumpling -> emperor
- I use ceph version 0.72-3-g5e1e02c (5e1e02c99b620fa4ffd2b455eb8e005b172fa05c), which is the "hotfix" for http://track...
Also available in: Atom