Activity
From 05/16/2019 to 06/14/2019
06/14/2019
- 05:40 PM Bug #40368 (Fix Under Review): [rbd-mirror] image sync can crash when updating progress
- 03:11 PM Bug #40368 (Resolved): [rbd-mirror] image sync can crash when updating progress
- The invocation of "m_copied_objects.top()" is not guarded so it might be invoked on an empty heap, resulting in an as...
- 03:25 PM Feature #40071 (Fix Under Review): [rbd] mirror image enable/disable should enable/disable journa...
- 03:20 PM Bug #39447 (Pending Backport): [tests] "rbd" teuthology suite has no coverage of "rbd diff"
06/13/2019
- 10:17 PM Feature #40072: [journal] tweak config defaults to improve small-IO performance
- Skip last commit (enabling writethrough mode) during backport
- 10:16 PM Feature #40072 (Fix Under Review): [journal] tweak config defaults to improve small-IO performance
- 05:27 PM Feature #40071 (In Progress): [rbd] mirror image enable/disable should enable/disable journaling
- 07:26 AM Support #40302: rbd-mirror: how to ensure the data consistency in secondary cluster
- Hi Jason,
After a second thought, I think that's not a good idea. We really should not wish an application-consis...
06/12/2019
- 05:09 PM Support #24953: Bad performance when enable journaling feature for rbd
- Hi Jason,
As mentioned, we cooked a patchset to implement the idea about this. In general, there are some patches ... - 04:36 PM Support #40302 (New): rbd-mirror: how to ensure the data consistency in secondary cluster
- When we are using rbd-mirror, we found a problem when we are going to start service on secondary image, where device-...
- 10:00 AM Backport #39675 (Resolved): nautilus: [test] possible race condition in rbd-nbd disconnect
- 09:38 AM Backport #40293 (Resolved): nautilus: rbd-nbd return correctly error message when no-match device.
- https://github.com/ceph/ceph/pull/29315
06/11/2019
- 09:47 PM Backport #39675: nautilus: [test] possible race condition in rbd-nbd disconnect
- Nathan Cutler wrote:
> https://github.com/ceph/ceph/pull/28273
merged - 08:24 PM Bug #39269: rbd-nbd return correctly error message when no-match device.
- @Jason: the status got changed to "Pending Backport" but the Backport field was empty, so I filled it provisionally w...
- 11:09 AM Bug #39269 (Pending Backport): rbd-nbd return correctly error message when no-match device.
- 08:17 PM Backport #40277 (Resolved): mimic: [object-map] resizing an image might result in an assert in 'O...
- https://github.com/ceph/ceph/pull/30796
- 08:16 PM Backport #40276 (Resolved): nautilus: [object-map] resizing an image might result in an assert in...
- https://github.com/ceph/ceph/pull/29246
- 08:16 PM Backport #40275 (Rejected): luminous: [object-map] resizing an image might result in an assert in...
- 08:16 PM Backport #40272 (Resolved): nautilus: [rbd-mirror] ensure tcmalloc is used if available
- https://github.com/ceph/ceph/pull/29193
- 05:02 PM Feature #40072 (In Progress): [journal] tweak config defaults to improve small-IO performance
- 01:39 PM Feature #40256 (Resolved): rbd bench: allow KB/SEC + MB/SEC + GB/SEC
- Currently the output of rbd bench only outputs bytes/s which is not easy to read when benchmarking in the scale of GB...
- 07:33 AM Backport #40122: nautilus: wrong format for rbd_mirror prometheus metrics
- https://github.com/ceph/ceph/pull/28485
- 07:09 AM Backport #40122 (In Progress): nautilus: wrong format for rbd_mirror prometheus metrics
- 07:11 AM Bug #39952 (Pending Backport): [object-map] resizing an image might result in an assert in 'Objec...
06/10/2019
- 09:52 PM Bug #40249 (Resolved): luminous: "test/librbd/test_librbd.cc:4729: Failure" in rados
- Run: http://pulpito.front.sepia.ceph.com/yuriw-2019-06-06_14:54:17-rados-wip-yuri-testing-2019-06-05-2303-luminous-di...
- 06:24 PM Bug #16811: [udev] /dev/rbd/<poolname>/<imagename> symlink should have an fsid in it
- The namespace ticket is in rbd, moved this one there as well.
- 06:02 PM Bug #40247 (Resolved): [udev] include image namespace in symlink path
- Currently, the "50-rbd" udev rule an "ceph-rbdnamer" do not understand RBD namespaces. This can result in two images ...
- 05:10 PM Bug #40110 (Pending Backport): [rbd-mirror] ensure tcmalloc is used if available
- 10:29 AM Backport #40237 (Resolved): nautilus: "profile rbd" OSD cap should add "class rbd metadata_list" ...
- https://github.com/ceph/ceph/pull/29328
- 10:28 AM Backport #40235 (Resolved): nautilus: [CLI]rbd: get positional argument error when using --image
- https://github.com/ceph/ceph/pull/28870
- 10:28 AM Backport #40234 (Resolved): mimic: [CLI]rbd: get positional argument error when using --image
- https://github.com/ceph/ceph/pull/29294
- 10:28 AM Backport #40233 (Resolved): luminous: [CLI]rbd: get positional argument error when using --image
- https://github.com/ceph/ceph/pull/29295
06/06/2019
- 04:18 PM Bug #40110 (Fix Under Review): [rbd-mirror] ensure tcmalloc is used if available
- 04:13 PM Bug #40110 (In Progress): [rbd-mirror] ensure tcmalloc is used if available
- 03:19 PM Bug #40095 (Pending Backport): [CLI]rbd: get positional argument error when using --image
- 03:18 PM Bug #39973 (Pending Backport): "profile rbd" OSD cap should add "class rbd metadata_list" cap by ...
06/04/2019
- 10:29 AM Backport #40122 (Resolved): nautilus: wrong format for rbd_mirror prometheus metrics
- https://github.com/ceph/ceph/pull/28485
06/03/2019
- 08:43 PM Bug #39977 (Pending Backport): wrong format for rbd_mirror prometheus metrics
- 02:09 PM Bug #40110 (Resolved): [rbd-mirror] ensure tcmalloc is used if available
- For small IO operations, tcmalloc is still markedly faster than the glibc allocator (w/ tcache).
06/01/2019
- 10:21 AM Backport #40105 (Resolved): nautilus: [test] qemu_dynamic_features workunit fails to disable fast...
- https://github.com/ceph/ceph/pull/28127
05/31/2019
- 02:32 PM Bug #39968 (Fix Under Review): [journal] add support for aligned appends
- 02:19 PM Bug #39946: [test] qemu_dynamic_features workunit fails to disable fast-diff+object-map
- Backport after (together with) backporting #39521
- 02:17 PM Bug #39946 (Pending Backport): [test] qemu_dynamic_features workunit fails to disable fast-diff+o...
- 12:56 PM Bug #39656 (Resolved): [test] devstack is broken (again)
- 12:56 PM Backport #39728 (Resolved): mimic: [test] devstack is broken (again)
- 12:55 PM Backport #39727 (Resolved): luminous: [test] devstack is broken (again)
- 11:58 AM Bug #39447: [tests] "rbd" teuthology suite has no coverage of "rbd diff"
- h3. original description
The "rbd" teuthology suite runs a workunit, "qa/workunits/rbd", which covers rbd snapsho... - 08:56 AM Bug #40095 (Fix Under Review): [CLI]rbd: get positional argument error when using --image
05/30/2019
- 11:54 PM Bug #40095 (Resolved): [CLI]rbd: get positional argument error when using --image
- ...
- 10:40 PM Bug #40094 (New): [caps] require min OSD release of luminous to use "profile rbd" caps
- A user prematurely switched existing caps to "profile rbd" prior to upgrading all OSDs. This resulted in in-use RBD i...
- 08:41 PM Backport #39728: mimic: [test] devstack is broken (again)
- Nathan Cutler wrote:
> https://github.com/ceph/ceph/pull/28129
merged - 08:38 PM Backport #39727: luminous: [test] devstack is broken (again)
- Nathan Cutler wrote:
> https://github.com/ceph/ceph/pull/28130
merged - 11:03 AM Backport #39674 (In Progress): mimic: [test] possible race condition in rbd-nbd disconnect
- https://github.com/ceph/ceph/pull/28310
- 09:16 AM Cleanup #39072 (Resolved): Drop "ceph_test_librbd_api" target
- 09:16 AM Backport #39423 (Resolved): nautilus: Drop "ceph_test_librbd_api" target
- 09:15 AM Bug #38706 (Resolved): flush skips requests in QOS throttler queue
- 09:14 AM Backport #38869 (Resolved): nautilus: flush skips requests in QOS throttler queue
- 09:13 AM Backport #39428 (Resolved): nautilus: 'rbd mirror status --verbose' will occasionally seg fault
- 09:12 AM Backport #39462 (Resolved): nautilus: [rbd-mirror] "bad crc in data" error when listing large pools
05/29/2019
- 10:10 PM Backport #39423: nautilus: Drop "ceph_test_librbd_api" target
- Nathan Cutler wrote:
> https://github.com/ceph/ceph/pull/28091
merged - 10:09 PM Backport #38869: nautilus: flush skips requests in QOS throttler queue
- Nathan Cutler wrote:
> https://github.com/ceph/ceph/pull/28120
merged - 10:09 PM Backport #39428: nautilus: 'rbd mirror status --verbose' will occasionally seg fault
- Nathan Cutler wrote:
> https://github.com/ceph/ceph/pull/28121
merged - 10:08 PM Backport #39462: nautilus: [rbd-mirror] "bad crc in data" error when listing large pools
- Nathan Cutler wrote:
> https://github.com/ceph/ceph/pull/28122
merged - 06:30 PM Feature #40072 (Resolved): [journal] tweak config defaults to improve small-IO performance
- Small IO throughput w/ journaling enabled is substantially worse than the expected worse-case of 2x the latency / 0.5...
- 06:15 PM Feature #40071 (Resolved): [rbd] mirror image enable/disable should enable/disable journaling
- It's confusing to many users that you need to enable/disable journaling independently from the RBD mirror image state...
- 01:39 PM Bug #39952 (Fix Under Review): [object-map] resizing an image might result in an assert in 'Objec...
- 01:12 PM Bug #39952 (In Progress): [object-map] resizing an image might result in an assert in 'ObjectMap:...
- 02:01 AM Backport #39675 (In Progress): nautilus: [test] possible race condition in rbd-nbd disconnect
- https://github.com/ceph/ceph/pull/28273
05/28/2019
- 08:29 AM Bug #39973 (Fix Under Review): "profile rbd" OSD cap should add "class rbd metadata_list" cap by ...
- 12:02 AM Bug #39973: "profile rbd" OSD cap should add "class rbd metadata_list" cap by default.
- *PR:*https://github.com/ceph/ceph/pull/28253
05/27/2019
- 10:58 AM Bug #40024 (Closed): Client(X86) to access Ceph
- 09:47 AM Bug #40024: Client(X86) to access Ceph
- Mykola Golub wrote:
> It is failing because the librbd thinks that the request offset 3198484480 (note, it is larger... - 08:03 AM Bug #40024 (Need More Info): Client(X86) to access Ceph
- It is failing because the librbd thinks that the request offset 3198484480 (note, it is larger 2G, max 32 bit signed)...
- 06:47 AM Bug #40024: Client(X86) to access Ceph
- qemu-kvm(X86_64) to access ARM's Ceph cluster, IO is error;
2019-05-24 14:38:29.504104 7f2ccd17b700 20 librbd::io::A... - 06:10 AM Bug #40024 (Closed): Client(X86) to access Ceph
05/23/2019
- 09:29 AM Bug #39257 (Resolved): [rbd-mirror] image replayer should periodically flush IO and commit positions
- 09:29 AM Backport #39288 (Resolved): nautilus: [rbd-mirror] image replayer should periodically flush IO an...
- 09:28 AM Bug #38661 (Resolved): deep cp a migration prepared image will results assert
- 09:28 AM Backport #39224 (Resolved): nautilus: deep cp a migration prepared image will results assert
- 09:26 AM Backport #39315 (Resolved): nautilus: krbd: fix rbd map hang due to udev return subsystem unordered
- 09:25 AM Backport #39729 (Resolved): nautilus: [test] devstack is broken (again)
- 09:25 AM Backport #39195 (Resolved): nautilus: Several race conditions are possible between io::ObjectRequ...
05/22/2019
- 07:49 PM Backport #39288: nautilus: [rbd-mirror] image replayer should periodically flush IO and commit po...
- Nathan Cutler wrote:
> https://github.com/ceph/ceph/pull/27937
merged - 04:25 PM Bug #38364 (Resolved): [sparsify] verify that image isn't using an EC data pool
- 04:25 PM Backport #39226 (Resolved): nautilus: [sparsify] verify that image isn't using an EC data pool
- 04:14 PM Backport #39226: nautilus: [sparsify] verify that image isn't using an EC data pool
- Nathan Cutler wrote:
> https://github.com/ceph/ceph/pull/27903
merged - 04:24 PM Bug #24668 (Resolved): [test] qemu-iotests tests fails under latest Ubuntu kernel
- 04:23 PM Backport #39541 (Resolved): nautilus: [test] qemu-iotests tests fails under latest Ubuntu kernel
- 04:13 PM Backport #39541: nautilus: [test] qemu-iotests tests fails under latest Ubuntu kernel
- Nathan Cutler wrote:
> https://github.com/ceph/ceph/pull/27988
merged - 04:16 PM Backport #39590 (Resolved): nautilus: qa/tasks/rbd_fio: fixed missing delimiter between 'cd' and ...
- 04:15 PM Backport #39224: nautilus: deep cp a migration prepared image will results assert
- Nathan Cutler wrote:
> https://github.com/ceph/ceph/pull/27882
merged - 04:11 PM Backport #39315: nautilus: krbd: fix rbd map hang due to udev return subsystem unordered
- Nathan Cutler wrote:
> https://github.com/ceph/ceph/pull/28019
merged - 04:11 PM Backport #39729: nautilus: [test] devstack is broken (again)
- Nathan Cutler wrote:
> https://github.com/ceph/ceph/pull/28128
merged - 04:10 PM Backport #39195: nautilus: Several race conditions are possible between io::ObjectRequest and io:...
- Nathan Cutler wrote:
> https://github.com/ceph/ceph/pull/28132
merged
05/21/2019
- 06:53 PM Bug #39703 (Resolved): 'qemu-img' seg faults on rbd_close under CentOS 7.5
- 04:30 PM Bug #39977 (Fix Under Review): wrong format for rbd_mirror prometheus metrics
- 10:19 AM Bug #39968 (In Progress): [journal] add support for aligned appends
05/20/2019
- 06:07 PM Bug #39977 (In Progress): wrong format for rbd_mirror prometheus metrics
- 10:12 AM Bug #39977 (Resolved): wrong format for rbd_mirror prometheus metrics
- When rbd_mirror is in use then all rbd_mirror metrics are incorrect. The output is in the format
ceph_rbd_mirror_POO... - 02:21 PM Bug #39980 (Fix Under Review): error from replay does not stored in rbd-mirror status
- 02:14 PM Bug #39980 (Resolved): error from replay does not stored in rbd-mirror status
- When a remote journal replay completes with an error, `ImageReplayer::handle_replay_complete` properly updates the st...
- 07:08 AM Bug #39973 (Resolved): "profile rbd" OSD cap should add "class rbd metadata_list" cap by default.
- Clients restricted to a namespace cannot currently open an image since it requires permission to read the pool metadata.
05/19/2019
- 10:07 AM Bug #39968 (Resolved): [journal] add support for aligned appends
- Journal entries are normally appended to the the end of the journal object. Unaligned appends can result in between 2...
05/17/2019
- 01:46 PM Bug #38706: flush skips requests in QOS throttler queue
- Jason wrote: "Backport not required for Mimic" and rejected the mimic backport issue. I am now deleting that issue be...
- 01:36 PM Backport #39501 (Need More Info): mimic: snapshot object maps can go inconsistent during copyup
- 01:36 PM Backport #39500 (Need More Info): luminous: snapshot object maps can go inconsistent during copyup
- 01:35 PM Backport #39499 (Need More Info): nautilus: snapshot object maps can go inconsistent during copyup
05/16/2019
- 03:01 PM Bug #39710: [exclusive lock] An image got more than one lock owners at the same time
- The tcmu-runner I tested at first was a continous integration build(not a formal release). If two tcmu-runners of thi...
- 02:52 PM Backport #38508 (In Progress): luminous: [rbd-mirror] LeaderWatcher stuck in loop if pool deleted
- 01:40 PM Backport #38509 (In Progress): mimic: [rbd-mirror] LeaderWatcher stuck in loop if pool deleted
- 12:55 PM Backport #37691: mimic: Image mirroring should be disabled when it is moved to trash
- PR: https://github.com/ceph/ceph/pull/28150
- 12:53 PM Backport #37692: luminous: Image mirroring should be disabled when it is moved to trash
- PR: https://github.com/ceph/ceph/pull/28149
- 12:43 PM Bug #39952 (Resolved): [object-map] resizing an image might result in an assert in 'ObjectMap::op...
- In 'librbd::operation::ResizeOperation', the IO is unblocked prior to resizing the object map in 'send_grow_object_ma...
- 12:52 AM Backport #38564 (In Progress): luminous: [librbd] race condition possible when validating RBD pool
Also available in: Atom