Project

General

Profile

Activity

From 08/30/2016 to 09/28/2016

09/28/2016

08:13 PM Bug #17355 (Pending Backport): rbd-mirror: potential crash during replay shut down
Mykola Golub
08:08 PM Subtask #17424 (Resolved): [data pool] update rbd CLI to permit "--data-pool" optionals
All CLI commands where an image can be created (create, clone, copy, import) should support an optional "--data-pool"... Jason Dillaman
08:06 PM Subtask #17423 (Resolved): [data pool] add new non-dynamic feature bit for data pool
* add new 'RBD_IMAGE_OPTION_DATA_POOL' image option to API
* new data pool feature would be set if data pool specifi...
Jason Dillaman
08:04 PM Subtask #17422 (Resolved): [data pool] cls_rbd updates for data pool
* new "get_data_pool" cls method against image header
* existing "create" cls method should optionally expect data p...
Jason Dillaman

09/27/2016

08:37 PM Feature #14738 (Pending Backport): Optionally unregister "laggy" journal clients
Jason Dillaman
12:37 PM Feature #14738 (Resolved): Optionally unregister "laggy" journal clients
Mykola Golub
08:31 PM Bug #17416 (Resolved): Improve resiliency of rbd-mirror stress test case
Jason Dillaman
02:35 PM Bug #15259: performance: journaling results in 4X slowdown when writes are not blocked by cache
*PR*: https://github.com/ceph/ceph/pull/10445 Jason Dillaman
02:35 PM Bug #15259 (Pending Backport): performance: journaling results in 4X slowdown when writes are not...
Jason Dillaman
12:32 PM Subtask #15688 (Resolved): Stub API integration between librbd and the new client-side cache
Mykola Golub
08:04 AM Bug #16988 (In Progress): 'rbd feature disable' against non-primary image results in hung CLI
Moved to resolved too quickly, should be done after PR#10980 is merged. Venky Shankar
08:03 AM Bug #16988 (Resolved): 'rbd feature disable' against non-primary image results in hung CLI
Tested with PR#10980, no cli hang was seen:
> ./bin/rbd --cluster cluster2 -p mirror feature disable r0 journaling...
Venky Shankar

09/26/2016

11:03 PM Cleanup #16130 (Pending Backport): Proxied operations shouldn't result in error messages if replayed
Ken Dreyer
11:03 PM Bug #16473: Demotion/promotion results in the image syncing again from the beginning
@Ken: most likely, yes -- but there isn't an explicit test case to verify that it is properly functioning. Jason Dillaman
10:56 PM Bug #16473: Demotion/promotion results in the image syncing again from the beginning
Jason, is this fixed by https://github.com/ceph/ceph/pull/10703 ? Ken Dreyer
10:58 PM Subtask #15108 (Pending Backport): Periodically update the sync point object number during sync
Ken Dreyer
08:39 PM Feature #17409 (Resolved): Optional separate data pool support for v2 images
In support of RBD images on EC pools, the image data block objects will need to be stored on a separate pool from the... Jason Dillaman
01:56 PM Bug #16056 (Pending Backport): RBD should restrict mirror enable/disable actions on parents/clones
@mykola: Yes, probably an oversight. Thanks. Jason Dillaman
12:47 PM Bug #16056: RBD should restrict mirror enable/disable actions on parents/clones
@dillaman Should we backport this to jewel? Mykola Golub
12:45 PM Bug #16056 (Resolved): RBD should restrict mirror enable/disable actions on parents/clones
Mykola Golub
12:43 PM Bug #16079 (Pending Backport): Enable/Disable of features is allowed even the features are alread...
Mykola Golub
12:22 PM Bug #16707 (Resolved): rbd-replay-prep doesn't record discard IO events
Mykola Golub
12:02 PM Bug #15764 (Fix Under Review): rbd-mirror bootstrap fails with -EEXIST when creating local image
PR: -https://github.com/ceph/ceph/pull/10896- (closed) Venky Shankar
11:56 AM Bug #16988: 'rbd feature disable' against non-primary image results in hung CLI
Jason, I think Mykola's PR https://github.com/ceph/ceph/pull/10980 would automatically fix this, right? Venky Shankar
10:34 AM Bug #17283 (Pending Backport): rbd-mirror: potential race condition results in heap corruption
Mykola Golub
09:59 AM Backport #17406 (Resolved): jewel: rbd-mirror: force-promoted image will remain R/O until local r...
https://github.com/ceph/ceph/pull/11433 Loïc Dachary
09:58 AM Backport #17405 (Resolved): jewel: Sporadic failure in TestMockJournal.ReplayOnDiskPostFlushError
https://github.com/ceph/ceph/pull/11433 Loïc Dachary
09:58 AM Backport #17404 (Resolved): jewel: update_features API needs to support backwards/forward compati...
https://github.com/ceph/ceph/pull/11462 Loïc Dachary

09/24/2016

05:14 PM Bug #16974 (Pending Backport): rbd-mirror: force-promoted image will remain R/O until local rbd-m...
Mykola Golub

09/23/2016

12:55 PM Bug #17317 (Pending Backport): Sporadic failure in TestMockJournal.ReplayOnDiskPostFlushError
Mykola Golub
10:23 AM Bug #17330 (Pending Backport): update_features API needs to support backwards/forward compatibility
Mykola Golub

09/22/2016

05:29 PM Bug #17182 (Resolved): "sysfs write failed" in smoke
Jason Dillaman
05:29 PM Backport #17375 (Resolved): jewel: "sysfs write failed" in smoke
Jason Dillaman
12:13 PM Backport #17375 (Resolved): jewel: "sysfs write failed" in smoke
https://github.com/ceph/ceph-qa-suite/pull/1182 Nathan Cutler
04:35 PM Backport #17384 (Resolved): jewel: helgrind: TestLibRBD.TestIOPP potential deadlock closing an im...
https://github.com/ceph/ceph/pull/11463 Nathan Cutler
02:24 PM Feature #17379 (New): rbd: journal: remove the image thread pool size limit
The current thread pool in ImageCtx.cc is initialized with a size of 1. This should be changed to use the config opti... Ricardo Dias
01:32 PM Bug #17198 (Pending Backport): helgrind: TestLibRBD.TestIOPP potential deadlock closing an image ...
Mykola Golub
12:12 PM Backport #17374 (Resolved): hammer: image.stat() call in librbdpy fails sometimes
https://github.com/ceph/ceph/pull/11949 Nathan Cutler
12:12 PM Backport #17373 (Resolved): jewel: image.stat() call in librbdpy fails sometimes
https://github.com/ceph/ceph/pull/11464 Nathan Cutler
10:50 AM Bug #17310 (Pending Backport): image.stat() call in librbdpy fails sometimes
Mykola Golub

09/21/2016

11:45 PM Bug #17199 (Won't Fix): helgrind: lock order warning with journal enabled + cache disabled
Helgrind is non-functional under master branch due to std::mutex -- no need to fix warnings Jason Dillaman
11:43 PM Bug #17201 (Won't Fix): helgrind: false positive lock order violation between journal and librbd
Helgrind is non-functional under master branch due to std::mutex -- no need to fix warnings. Jason Dillaman
11:43 PM Bug #17202 (Won't Fix): helgrind: lock order warning within AioCompletion
Helgrind is non-functional under master branch due to std::mutex -- no need to fix warnings. Jason Dillaman
11:43 PM Bug #17299 (Won't Fix): helgrind: async messenger fails to destruct spin lock
Helgrind is non-functional under master branch due to std::mutex -- no need to fix warnings. Jason Dillaman
11:43 PM Bug #17300 (Won't Fix): helgrind: boost intrusive_set data race with non-constant time size
Helgrind is non-functional under master branch due to std::mutex -- no need to fix warnings. Jason Dillaman
11:42 PM Bug #17301 (Won't Fix): helgrind: async messenger race condition warnings
Helgrind is non-functional under master branch due to std::mutex -- no need to fix warnings. Jason Dillaman
11:38 PM Bug #17355 (Fix Under Review): rbd-mirror: potential crash during replay shut down
*PR*: https://github.com/ceph/ceph/pull/11185 Jason Dillaman
12:15 PM Bug #17355 (Resolved): rbd-mirror: potential crash during replay shut down
If there is an in-flight event being committed to the journal as the replayer is being shut down, it's possible for a... Jason Dillaman
03:53 PM Bug #17182 (Pending Backport): "sysfs write failed" in smoke
Jason Dillaman
02:38 PM Bug #17182 (Fix Under Review): "sysfs write failed" in smoke
https://github.com/ceph/ceph-qa-suite/pull/1178 Ilya Dryomov
02:10 PM Bug #17182: "sysfs write failed" in smoke
Ah, sorry - smoke does require a recent kernel since jewel, so 5 is fine. Ilya Dryomov
11:52 AM Bug #17182 (In Progress): "sysfs write failed" in smoke
5 won't do it because it doesn't install a kernel here - it should be 1. I'll send a PR. Ilya Dryomov
01:42 PM Feature #17356 (Resolved): object-map: batch updates during trim operation
When removing or shrinking a very large image, currently each individual object removal could result in an individual... Jason Dillaman
11:08 AM Feature #16213 (Fix Under Review): Replicate dynamic feature updates
PR: https://github.com/ceph/ceph/pull/10980 Mykola Golub
11:07 AM Feature #16212 (Fix Under Review): Replicate image metadata settings
PR: https://github.com/ceph/ceph/pull/11168 Mykola Golub

09/20/2016

10:12 PM Bug #17182: "sysfs write failed" in smoke
Should be resolved with a modification like the following:... Jason Dillaman
09:56 PM Bug #17182: "sysfs write failed" in smoke
Also in http://qa-proxy.ceph.com/teuthology/teuthology-2016-09-20_05:00:02-smoke-master-testing-basic-vps/425655/teut... Yuri Weinstein
07:58 PM Backport #17340 (Resolved): jewel: exclusive_lock::AcquireRequest doesn't handle -ERESTART on ima...
https://github.com/ceph/ceph/pull/11465 Nathan Cutler
06:54 PM Bug #17330 (Fix Under Review): update_features API needs to support backwards/forward compatibility
*PR*: https://github.com/ceph/ceph/pull/11155 Jason Dillaman
06:44 PM Bug #17330 (Resolved): update_features API needs to support backwards/forward compatibility
Attempting to disable exclusive lock using a Jewel+ client against an Infernalis OSD will result in an error. Since I... Jason Dillaman
05:34 PM Bug #17317 (Fix Under Review): Sporadic failure in TestMockJournal.ReplayOnDiskPostFlushError
*PR*: https://github.com/ceph/ceph/pull/11153 Jason Dillaman
04:08 PM Bug #17317 (Resolved): Sporadic failure in TestMockJournal.ReplayOnDiskPostFlushError
... Jason Dillaman
02:28 PM Bug #17199 (In Progress): helgrind: lock order warning with journal enabled + cache disabled
Jason Dillaman
02:21 PM Bug #17198 (Fix Under Review): helgrind: TestLibRBD.TestIOPP potential deadlock closing an image ...
*PR*: https://github.com/ceph/ceph/pull/11152 Jason Dillaman
01:39 PM Bug #17198 (In Progress): helgrind: TestLibRBD.TestIOPP potential deadlock closing an image with ...
Jason Dillaman
02:14 PM Bug #17296 (Closed): [teuthology] failure in cli-integration/rbd/formatted-output.t
Jewel test branch accidentally used against master branch codebase. Jason Dillaman
01:37 PM Bug #17296 (In Progress): [teuthology] failure in cli-integration/rbd/formatted-output.t
Jason Dillaman
11:39 AM Bug #17310 (Fix Under Review): image.stat() call in librbdpy fails sometimes
*PR*: https://github.com/ceph/ceph/pull/11148 Jason Dillaman
11:19 AM Bug #17310 (In Progress): image.stat() call in librbdpy fails sometimes
Jason Dillaman
10:18 AM Bug #17310 (Resolved): image.stat() call in librbdpy fails sometimes
I'm running a simple test script to reproduce this:... Jens Harbott

09/19/2016

06:12 PM Bug #17301 (Won't Fix): helgrind: async messenger race condition warnings
Jason Dillaman
05:35 PM Bug #17300 (Won't Fix): helgrind: boost intrusive_set data race with non-constant time size
Jason Dillaman
05:34 PM Bug #17299 (Won't Fix): helgrind: async messenger fails to destruct spin lock
Jason Dillaman
05:28 PM Bug #14846 (Closed): teuthology: helgrind warnings
Will track and fix as individual tracker tickets as helgrind warnings are always a moving target. Jason Dillaman
12:32 PM Bug #17296 (Closed): [teuthology] failure in cli-integration/rbd/formatted-output.t
... Jason Dillaman

09/18/2016

02:38 PM Backport #17290 (New): jewel: ImageWatcher: use after free within C_UnwatchAndFlush
Xiaoxi Chen
02:37 PM Backport #17290 (Duplicate): jewel: ImageWatcher: use after free within C_UnwatchAndFlush
Xiaoxi Chen
02:36 PM Backport #17290 (Resolved): jewel: ImageWatcher: use after free within C_UnwatchAndFlush
https://github.com/ceph/ceph/pull/11466 Xiaoxi Chen

09/17/2016

02:22 PM Bug #17289 (Pending Backport): ImageWatcher: use after free within C_UnwatchAndFlush
Mykola Golub
12:32 PM Bug #17289 (Fix Under Review): ImageWatcher: use after free within C_UnwatchAndFlush
*PR*: https://github.com/ceph/ceph/pull/11112 Jason Dillaman
12:29 PM Bug #17289 (Resolved): ImageWatcher: use after free within C_UnwatchAndFlush
Jason Dillaman

09/16/2016

06:24 PM Bug #17202 (In Progress): helgrind: lock order warning within AioCompletion
Jason Dillaman
02:13 PM Bug #16974 (Fix Under Review): rbd-mirror: force-promoted image will remain R/O until local rbd-m...
*PR*: https://github.com/ceph/ceph/pull/11090#issuecomment-247567977 Jason Dillaman

09/15/2016

10:38 PM Bug #17283 (Fix Under Review): rbd-mirror: potential race condition results in heap corruption
*PR*: https://github.com/ceph/ceph/pull/11104 Jason Dillaman
10:14 PM Bug #17283 (Resolved): rbd-mirror: potential race condition results in heap corruption
Changes introduced in commit 4df913d1 might result in concurrent writes to rbd::mirror::ImageReplayer::m_event_entry. Jason Dillaman
03:15 PM Feature #16780 (Fix Under Review): rbd-mirror: use sparse read during image sync
*PR*: https://github.com/ceph/ceph/pull/11005 Jason Dillaman

09/13/2016

10:37 PM Bug #17182 (Need More Info): "sysfs write failed" in smoke
Ceph bug? Zack Cerza

09/12/2016

03:17 PM Bug #16717 (Resolved): "[ FAILED ] TestLibRBD.TestCreateLsDeletePP" in upgrade:client-upgrade-j...
*Jewel PR*: https://github.com/ceph/ceph-qa-suite/pull/1173 Jason Dillaman
03:01 PM Bug #16717 (Pending Backport): "[ FAILED ] TestLibRBD.TestCreateLsDeletePP" in upgrade:client-u...
I think we still need to backport https://github.com/ceph/ceph-qa-suite/pull/1172 to jewel branch. AFAIR teuthology r... Mykola Golub
12:51 PM Bug #16717 (Resolved): "[ FAILED ] TestLibRBD.TestCreateLsDeletePP" in upgrade:client-upgrade-j...
Jason Dillaman
11:33 AM Bug #16717 (Fix Under Review): "[ FAILED ] TestLibRBD.TestCreateLsDeletePP" in upgrade:client-u...
Mykola Golub
11:32 AM Bug #16717: "[ FAILED ] TestLibRBD.TestCreateLsDeletePP" in upgrade:client-upgrade-jewel-distro...
PR: https://github.com/ceph/ceph-qa-suite/pull/1172 Mykola Golub
02:01 PM Backport #17265 (In Progress): jewel: Possible deadlock race condition between image close and li...
Xiaoxi Chen
01:52 PM Backport #17265 (Resolved): jewel: Possible deadlock race condition between image close and libra...
https://github.com/ceph/ceph/pull/11466 Xiaoxi Chen
09:10 AM Bug #17254 (Pending Backport): Possible deadlock race condition between image close and librados ...
Mykola Golub
09:10 AM Bug #17227 (Pending Backport): exclusive_lock::AcquireRequest doesn't handle -ERESTART on image::...
Mykola Golub
07:40 AM Backport #17262 (Resolved): jewel: rbd-nbd IO hang
https://github.com/ceph/ceph/pull/11467 Loïc Dachary
07:40 AM Backport #17261 (Resolved): jewel: Potential seg fault when blacklisting a client
Loïc Dachary

09/11/2016

01:10 PM Bug #17227 (Fix Under Review): exclusive_lock::AcquireRequest doesn't handle -ERESTART on image::...
*PR*: https://github.com/ceph/ceph/pull/11044 Jason Dillaman

09/10/2016

08:25 AM Bug #17251 (Pending Backport): Potential seg fault when blacklisting a client
Mykola Golub

09/09/2016

02:59 PM Bug #17254 (Fix Under Review): Possible deadlock race condition between image close and librados ...
*PR*: https://github.com/ceph/ceph/pull/11037 Jason Dillaman
02:37 PM Bug #17254 (Resolved): Possible deadlock race condition between image close and librados shutdown
librbd::ImageWatcher::C_UnwatchAndFlush instantiates a librados::Rados class to invoke aio_watch_flush. However, it's... Jason Dillaman
01:34 PM Bug #16921 (Pending Backport): rbd-nbd IO hang
Mykola Golub
12:47 PM Bug #16974 (In Progress): rbd-mirror: force-promoted image will remain R/O until local rbd-mirror...
*NOTE*: blocker for upstream cinder RBD replication driver feature Jason Dillaman
12:33 PM Bug #17251 (Fix Under Review): Potential seg fault when blacklisting a client
*PR*: https://github.com/ceph/ceph/pull/11034 Jason Dillaman
12:31 PM Bug #17251 (Resolved): Potential seg fault when blacklisting a client
Jason Dillaman
09:55 AM Bug #16717: "[ FAILED ] TestLibRBD.TestCreateLsDeletePP" in upgrade:client-upgrade-jewel-distro...
Consistently happens:... Loïc Dachary

09/08/2016

03:51 PM Bug #16921: rbd-nbd IO hang
Spurious read request from kernel via rbd-nbd that raced with the resize resulted in the cache not being able to inva... Jason Dillaman
02:00 PM Backport #17241 (In Progress): jewel: "*** Caught signal" in krbd
Loïc Dachary
01:59 PM Backport #17241 (Resolved): jewel: "*** Caught signal" in krbd
https://github.com/ceph/ceph/pull/11018 Loïc Dachary
01:59 PM Backport #17243 (Resolved): jewel: Deadlock in several librbd teuthology test cases
https://github.com/ceph/ceph/pull/12890 Loïc Dachary
01:59 PM Backport #17242 (Resolved): jewel: ImageWatcher: double unwatch of failed watch handle
Loïc Dachary
01:57 PM Bug #17223 (Pending Backport): "*** Caught signal" in krbd
Loïc Dachary
01:57 PM Bug #17223: "*** Caught signal" in krbd
https://github.com/ceph/ceph/pull/9517
https://github.com/ceph/ceph/pull/9080
https://github.com/ceph/ceph/pull/9123
Loïc Dachary

09/07/2016

05:43 PM Bug #16921: rbd-nbd IO hang
2016-09-07 17:15:50.931644 7f59527fc700 -1 librbd::AioCompletion: 0x7f5938000cc0 fail: (22) Invalid argument
2016-09...
Jason Dillaman
02:55 PM Bug #16921: rbd-nbd IO hang
... Jason Dillaman
02:49 AM Bug #17227 (Resolved): exclusive_lock::AcquireRequest doesn't handle -ERESTART on image::RefreshR...
http://qa-proxy.ceph.com/teuthology/jdillaman-2016-09-06_13:37:26-rbd-wip-jd-testing---basic-smithi/402468/teuthology... Jason Dillaman

09/06/2016

08:44 PM Bug #17223 (In Progress): "*** Caught signal" in krbd
Ilya Dryomov
05:44 PM Bug #17223: "*** Caught signal" in krbd
Most of the coredumps here are probably #16030 material. Need to backport that fix to jewel. Ilya Dryomov
04:32 PM Bug #17223 (Resolved): "*** Caught signal" in krbd
This is jewel 10.2.3 release
Run: http://pulpito.front.sepia.ceph.com/yuriw-2016-09-05_15:56:10-krbd-master-testin...
Yuri Weinstein
03:23 PM Bug #17219 (Resolved): rbd-nbd: disallow mapping images >2TB in size
The nbd kernel driver uses a hard-coded 512 byte sector size, resulting in a 2TB block size limit for nbd devices. Jason Dillaman
12:38 PM Bug #17210 (Pending Backport): ImageWatcher: double unwatch of failed watch handle
Mykola Golub
02:02 AM Backport #16952 (In Progress): hammer: ceph 10.2.2 rbd status on image format 2 returns "(2) No s...
Wei-Chung Cheng

09/05/2016

06:37 PM Bug #16921 (Fix Under Review): rbd-nbd IO hang
*PR*: https://github.com/ceph/ceph/pull/10985 Jason Dillaman
02:19 PM Bug #16921: rbd-nbd IO hang
Appears to be related to teothology's fixed PID mapping:
2016-09-05 10:18:24.479834 7fddfbc3ce00 -1 open: failed t...
Jason Dillaman
02:02 PM Bug #16921 (In Progress): rbd-nbd IO hang
Jason Dillaman
01:58 AM Feature #15632 (Resolved): Expose librbd API methods to directly acquire and release the exclusiv...
Jason Dillaman

09/04/2016

11:43 PM Feature #16780: rbd-mirror: use sparse read during image sync
all right. Anonymous
11:43 PM Feature #17178: [OpenStack Glance] update store driver to sparsify rbd-backed images
I know,we should implement the similar logic of rbd CLI's import tool in glance store. I can help on it. Anonymous
02:50 PM Bug #17210 (Fix Under Review): ImageWatcher: double unwatch of failed watch handle
*PR*: https://github.com/ceph/ceph/pull/10974 Jason Dillaman
02:46 PM Bug #17210 (Resolved): ImageWatcher: double unwatch of failed watch handle
http://pulpito.ceph.com/jdillaman-2016-09-03_07:55:53-rbd-wip-jd-testing---basic-smithi/397723
Crash witnessed aft...
Jason Dillaman

09/02/2016

07:01 PM Bug #17188 (Pending Backport): Deadlock in several librbd teuthology test cases
Mykola Golub
12:24 PM Feature #16780: rbd-mirror: use sparse read during image sync
@Tianqing Li: essentially no harm to keep it enabled always -- librbd already utilizes sparse reads for every read re... Jason Dillaman
03:58 AM Feature #16780: rbd-mirror: use sparse read during image sync
should we make it configurable or always do sparse read? Anonymous
12:23 PM Feature #17178: [OpenStack Glance] update store driver to sparsify rbd-backed images
@Tianqing Li: there is logic within the rbd CLI's import tool to skip fully zeroed blocks. We could optionally put so... Jason Dillaman
09:57 AM Feature #17178: [OpenStack Glance] update store driver to sparsify rbd-backed images
Is the python rbd binding support skipping fully zeroed blocks? I found use cli, it will, but python binding is not, ... Anonymous
12:11 PM Bug #17201: helgrind: false positive lock order violation between journal and librbd
Also:
2016-09-01T20:32:00.355 INFO:tasks.workunit.client.0.vpm041.stderr:==21289== Thread #21: lock order "0x7CF10...
Jason Dillaman
12:07 PM Bug #17201 (Won't Fix): helgrind: false positive lock order violation between journal and librbd
2016-09-01T20:25:58.599 INFO:tasks.workunit.client.0.vpm041.stderr:==21289== Thread #21: lock order "0x7CA9070 before... Jason Dillaman
12:09 PM Bug #17202 (Won't Fix): helgrind: lock order warning within AioCompletion
2016-09-01T20:31:50.916 INFO:tasks.workunit.client.0.vpm041.stderr:==21289== Thread #14: lock order "0x7C99258 before... Jason Dillaman
11:58 AM Bug #17199 (Won't Fix): helgrind: lock order warning with journal enabled + cache disabled
http://qa-proxy.ceph.com/teuthology/trociny-2016-09-01_18:30:34-rbd-wip-mgolub-testing---basic-vps/395472/teuthology.... Jason Dillaman
11:40 AM Bug #17198 (Resolved): helgrind: TestLibRBD.TestIOPP potential deadlock closing an image with rea...
http://qa-proxy.ceph.com/teuthology/trociny-2016-09-01_18:31:51-rbd-wip-mgolub-testing---basic-vps/395528/teuthology.... Jason Dillaman

09/01/2016

10:07 PM Bug #17195 (Resolved): There seems to be a thread waiting indefinitely in krbd.cc
We recently encountered this issue where rbd map hangs. From the thread trace we identified this piece of code:
...
Abhi N
01:58 AM Bug #17188 (Fix Under Review): Deadlock in several librbd teuthology test cases
*PR*: https://github.com/ceph/ceph/pull/10945 Jason Dillaman

08/31/2016

10:19 PM Bug #17188 (Resolved): Deadlock in several librbd teuthology test cases
http://pulpito.ceph.com/jdillaman-2016-08-29_07:51:01-rbd-wip-jd-testing---basic-smithi/ Jason Dillaman
03:16 PM Bug #17182 (Resolved): "sysfs write failed" in smoke
Run: http://pulpito.ceph.com/teuthology-2016-08-31_05:00:01-smoke-master-testing-basic-vps/
Job: 394034
Logs: http:...
Yuri Weinstein
05:27 AM Feature #17178 (In Progress): [OpenStack Glance] update store driver to sparsify rbd-backed images
If a task flow isn't being utilized to automatically convert qcow2 images to raw images on the glance backend, the us... Jason Dillaman
 

Also available in: Atom