Project

General

Profile

Activity

From 02/09/2016 to 03/09/2016

03/09/2016

11:10 PM Backport #15032 (In Progress): infernalis: QEMU VM hangs talking to RBD via librbd
Jason Dillaman
09:52 PM Backport #15032 (Resolved): infernalis: QEMU VM hangs talking to RBD via librbd
https://github.com/ceph/ceph/pull/8010 Nathan Cutler
11:09 PM Backport #15033 (In Progress): hammer: QEMU VM hangs talking to RBD via librbd
Jason Dillaman
09:52 PM Backport #15033 (Resolved): hammer: QEMU VM hangs talking to RBD via librbd
https://github.com/ceph/ceph/pull/8011 Nathan Cutler
10:37 PM Bug #14988: QEMU VM hangs talking to RBD via librbd
*master PR*: https://github.com/ceph/ceph/pull/8009 (reproducer -- confirms not an issue in Jewel) Jason Dillaman
09:52 PM Bug #14988 (Pending Backport): QEMU VM hangs talking to RBD via librbd
Note that there is no master PR for this issue. Jason is going to do the fix in infernalis and then backport it from ... Nathan Cutler
09:46 PM Bug #14988: QEMU VM hangs talking to RBD via librbd
Shouldn't be an issue in Jewel due to new asynchronous internals Jason Dillaman
09:26 PM Bug #14988 (In Progress): QEMU VM hangs talking to RBD via librbd
Jason Dillaman
09:13 PM Bug #14988: QEMU VM hangs talking to RBD via librbd
If an AIO request results in an image refresh due to a new snapshot creation, the cache will be flushed. When this o... Jason Dillaman
10:02 PM Bug #14789 (Resolved): Truncate of non-existent object results in object map flagged as exists
Josh Durgin
09:55 PM Bug #15034 (Resolved): Possible inconsistent IO when using configured for multiple op threads
When 'rbd_op_threads' is set to a value higher than 1 (defaults to 1), it's quite possible for an aio flush op to com... Jason Dillaman
06:11 PM Bug #13122 (Resolved): rbd: stripingv2 feature bit not set when unnecessary for striping params
This was fixed sometime in the recent rbd cli refactor. Josh Durgin
04:56 PM Subtask #15029 (Resolved): Use watch/notify to alert to specific changes in the rbd_mirroring object
This will eliminate to constantly poll the object for updates -- instead, it would only need to pull the object after... Jason Dillaman
04:32 PM Bug #15028 (Resolved): Deep-flatten incorrectly updates fast-diff stats
Flattening a clone of an empty image with fast-diff and deep-flatten enabled results in all objects being flagged as ... Jason Dillaman
03:24 PM Fix #15026 (New): use derr instead of cerr in rbd cli actions
... Ilya Dryomov
09:14 AM Bug #15007: Cannot rebuild object map in in-flight IO
Due to the upgrade to infernalis it would be fine to get the ability for live enabling the object map. Jonas +1 Tim Förster
08:01 AM Bug #15007: Cannot rebuild object map in in-flight IO
Jason Dillaman wrote:
> > Object Map Rebuild: 99% complete...failed.
> > rbd: rebuilding object map failed: (16) De...
Jonas Keidel
02:09 AM Fix #3588 (Resolved): rbd.py's clone should take stripe parms, call rbd_clone2
Fixed as part of the new image create api use in commit:c3be44ed8f9d99d7257156f2e0bf7ef6ea1b5761 Josh Durgin

03/08/2016

06:21 PM Bug #14908 (Fix Under Review): Journal replay needs to conditionally refresh the image between ev...
*PR*: https://github.com/ceph/ceph/pull/7978 Jason Dillaman
04:52 PM Bug #14908 (In Progress): Journal replay needs to conditionally refresh the image between events
Jason Dillaman
04:58 PM Bug #14986 (Closed): rbd export/export-diff with needs to protect against unintended journal replay
The journal actually won't be replayed since export/export-diff open the image read-only. This prevents the journal ... Jason Dillaman
04:45 PM Bug #14953 (Rejected): can add 'rbd lock' to format 1 images
Jason Dillaman
04:43 PM Bug #14953: can add 'rbd lock' to format 1 images
@Vasu, the "rbd lock"-related commands are to be considered as advisory locks. There is no enforcement within librbd... Jason Dillaman
04:41 PM Bug #14988 (Need More Info): QEMU VM hangs talking to RBD via librbd
Josh Durgin
04:19 PM Bug #14780 (Resolved): librbd: TaskFinisher lifetime no longer matches ImageWatcher
Josh Durgin
02:50 PM Bug #15007 (Resolved): Cannot rebuild object map in in-flight IO
> Object Map Rebuild: 99% complete...failed.
> rbd: rebuilding object map failed: (16) Device or resource busy
Jason Dillaman

03/07/2016

04:52 PM Documentation #15000 (Resolved): Need better documentation to describe RBD image features
Jason Dillaman
01:18 PM Bug #14988: QEMU VM hangs talking to RBD via librbd
@Roy,
From the backtrace, Ceph is waiting for a cache flush to complete. The thread responsible for handling cac...
Jason Dillaman

03/06/2016

08:11 PM Bug #14988: QEMU VM hangs talking to RBD via librbd
... Jason Dillaman
06:28 PM Bug #14988: QEMU VM hangs talking to RBD via librbd
@Roy: Upgrading is recommended anyway, but before you take that step you can try setting "rbd_non_blocking_aio" to fa... Nathan Cutler
06:01 PM Bug #14988: QEMU VM hangs talking to RBD via librbd
I can try with a newer Ceph, but #13726 seems to be dependant upon RBD cache being disabled but it should be enabled ... Roy Keene

03/05/2016

10:49 PM Bug #14988: QEMU VM hangs talking to RBD via librbd
@Roy: Can you reproduce this on 0.94.6? If not, it is probably http://tracker.ceph.com/issues/13726 Nathan Cutler
08:44 PM Bug #14988 (Resolved): QEMU VM hangs talking to RBD via librbd
I'm having a repeatable hang in QEMU (v2.4.1) talking to a Ceph (v0.95.4) RBD via librbd.
I've acquired a core fil...
Roy Keene
12:32 AM Bug #14986 (Closed): rbd export/export-diff with needs to protect against unintended journal replay
If journaling is enabled on an image, reading from the image HEAD will result in a journal replay. If the image was ... Jason Dillaman

03/04/2016

09:33 PM Bug #14984 (Resolved): rbdmap.service should include RBDMAPFILE environment variable
/usr/bin/rbdmap requires the RBDMAPFILE variable to be set. I've worked around this on my machines by adding Environm... Paul Rawson
03:28 PM Subtask #14981 (Resolved): rbd-mirror daemon systemd / upstart scripts
Jason Dillaman

03/03/2016

03:48 PM Subtask #14869 (Fix Under Review): journal library should re-use thread pool and timer
*PR*: https://github.com/ceph/ceph/pull/7906 Jason Dillaman
09:17 AM Subtask #13296 (In Progress): Mirroring API + rbd CLI updates for images
Ricardo Dias
01:34 AM Subtask #13296: Mirroring API + rbd CLI updates for images
Correct, the description is from an old design. We are using http://pad.ceph.com/p/rbd_mirror_daemon to track the cu... Jason Dillaman

03/02/2016

10:06 PM Subtask #13296: Mirroring API + rbd CLI updates for images
Jason Dillaman wrote:
> A new "mirroring enabled" field should be added to the RBD image header. Mirroring should n...
Ricardo Dias
05:37 PM Subtask #13296: Mirroring API + rbd CLI updates for images
Jason Dillaman
05:43 PM Bug #14953 (Rejected): can add 'rbd lock' to format 1 images
this is on hammer 0.94.5, i think its deprecated now recently but still want to document it
1) 2 node ceph cluste...
Vasu Kulkarni
05:37 PM Subtask #14869 (In Progress): journal library should re-use thread pool and timer
Jason Dillaman
05:35 PM Subtask #14413 (In Progress): Pull full image from remote peer when out-of-sync
Jason Dillaman
05:19 PM Subtask #14663 (Fix Under Review): Integrate librbd with journal tag allocation
*PR*: https://github.com/ceph/ceph/pull/7884 Jason Dillaman
04:33 AM Subtask #14414 (Fix Under Review): Add new "exclusive lock released" journal event to librbd
https://github.com/ceph/ceph/pull/7818 Josh Durgin
04:30 AM Subtask #14161 (Resolved): Serialize snap create, snap rollback, and resize operations for journa...
https://github.com/ceph/ceph/pull/7019 Josh Durgin
04:23 AM Subtask #14419 (Resolved): RBD mirroring settings should be moved to new pool object
https://github.com/ceph/ceph/pull/7620 Josh Durgin
04:21 AM Feature #2468 (Resolved): librbd: provide a way for a user to flush and invalidate the cache
Josh Durgin
04:18 AM Bug #14822 (Resolved): Journal replay needs to support replaying successful maintenance ops
Josh Durgin
04:18 AM Cleanup #14809 (Resolved): ImageWatcher unwatch and flush should use async librados methods
Josh Durgin
04:17 AM Tasks #14550 (Resolved): Deprecate format v1 RBD images
Josh Durgin
04:16 AM Tasks #14551 (Resolved): Update RBD default features
Josh Durgin

03/01/2016

09:52 PM Feature #14909 (New): Journal replay should support multiple tag class streams
Jason Dillaman
04:33 PM Subtask #14937 (Resolved): Cloned images must only be bootstrapped after parent image successfull...
Jason Dillaman

02/29/2016

06:24 PM Tasks #14551 (Fix Under Review): Update RBD default features
*PR*: https://github.com/ceph/ceph/pull/7846
*PR*: https://github.com/ceph/ceph-qa-suite/pull/841
*PR*: https://git...
Jason Dillaman
04:28 PM Tasks #14551 (In Progress): Update RBD default features
Jason Dillaman
05:26 PM Backport #13757: rbd: pure virtual method called
It was rejected because firefly is EOL. Loïc Dachary
05:22 PM Backport #13757: rbd: pure virtual method called
This problem was reported in the OpenStack gate for ceph:
https://bugs.launchpad.net/ubuntu/+source/ceph/+bug/1551...
James Page
02:09 AM Tasks #14550 (In Progress): Deprecate format v1 RBD images
Jason Dillaman

02/28/2016

08:34 PM Cleanup #14809: ImageWatcher unwatch and flush should use async librados methods
*PR*: https://github.com/ceph/ceph/pull/7839 Jason Dillaman
05:47 PM Cleanup #14809 (In Progress): ImageWatcher unwatch and flush should use async librados methods
Jason Dillaman
05:46 PM Feature #14909 (In Progress): Journal replay should support multiple tag class streams
Jason Dillaman

02/26/2016

07:22 PM Feature #14909 (New): Journal replay should support multiple tag class streams
Assume an journal layout like:
Object 0: Object 1:
tag=0 (class 0), entry=0 tag=2 (class 0...
Jason Dillaman
07:15 PM Bug #14908 (Resolved): Journal replay needs to conditionally refresh the image between events
If a previously replayed maintenance operation requires an image refresh, perform the refresh before processing more ... Jason Dillaman
04:42 PM Subtask #13298 (Resolved): Journal registered client entry should store bufferlist payload
Jason Dillaman
06:03 AM Backport #14863 (Resolved): hammer: test/crypto.cc:161:43: error: ‘ceph_clock_now’ was not declar...
Loïc Dachary
02:54 AM Feature #14264 (Resolved): async rbd_open() option
Haomai Wang

02/25/2016

08:04 PM Subtask #14869 (Resolved): journal library should re-use thread pool and timer
This will allow the RBD mirror daemon to not need to allocate two threads per image. Jason Dillaman
04:40 PM Feature #14867 (Resolved): add object map verifier to RBD CLI
Add rbd object-map verify to the RBD CLI that reads the object map and compares it with a rebuilt one from the availa... Douglas Fuller
03:35 PM Bug #14822 (Fix Under Review): Journal replay needs to support replaying successful maintenance ops
*PR*: https://github.com/ceph/ceph/pull/7785 Jason Dillaman
12:06 PM Backport #14863: hammer: test/crypto.cc:161:43: error: ‘ceph_clock_now’ was not declared in this ...
https://github.com/ceph/ceph/commit/fb4b6c52d8e31e8dedfa8aecc16c389f0e7d86cf was partially included as part of https:... Loïc Dachary
12:04 PM Backport #14863 (Resolved): hammer: test/crypto.cc:161:43: error: ‘ceph_clock_now’ was not declar...
https://github.com/ceph/ceph/pull/7801 Loïc Dachary
06:07 AM Feature #14264: async rbd_open() option
I think the status can be changed to resolved. Mykola Golub

02/24/2016

04:23 PM Bug #14822 (In Progress): Journal replay needs to support replaying successful maintenance ops
Jason Dillaman
02:23 PM Bug #14847 (Resolved): teuthology: test_librbd_python crashes
Jason Dillaman
11:10 AM Bug #14789: Truncate of non-existent object results in object map flagged as exists
https://github.com/ceph/ceph/pull/7772 Xinxin Shu
11:09 AM Bug #14789 (Fix Under Review): Truncate of non-existent object results in object map flagged as e...
Xinxin Shu
06:10 AM Bug #14483 (Resolved): Flattening an rbd image with a running Qemu instances causes librbd worker...
Loïc Dachary
06:08 AM Backport #14611 (Resolved): hammer: Flattening an rbd image with a running Qemu instances causes ...
Loïc Dachary
06:03 AM Backport #14611 (In Progress): hammer: Flattening an rbd image with a running Qemu instances caus...
Loïc Dachary
06:08 AM Backport #14620 (Resolved): hammer: unsafe use of libnss PK11_GetBestSlot()
Loïc Dachary
06:03 AM Backport #14620 (In Progress): hammer: unsafe use of libnss PK11_GetBestSlot()
Loïc Dachary

02/23/2016

04:53 PM Bug #14595 (Resolved): Race between releasing exclusive lock and incoming write ops
This is not an issue within jewel due to the new block-writes ability and exclusive lock state machine. Jason Dillaman
04:30 PM Bug #14846 (In Progress): teuthology: helgrind warnings
Jason Dillaman
01:14 PM Bug #14846 (Closed): teuthology: helgrind warnings
http://teuthology.ovh.sepia.ceph.com/teuthology/teuthology-2016-02-22_16:00:01-rbd-master---basic-openstack/3178/teut... Jason Dillaman
04:29 PM Bug #14847 (Fix Under Review): teuthology: test_librbd_python crashes
*PR*: https://github.com/ceph/ceph/pull/7761 Jason Dillaman
01:15 PM Bug #14847 (Resolved): teuthology: test_librbd_python crashes
http://teuthology.ovh.sepia.ceph.com/teuthology/teuthology-2016-02-22_16:00:01-rbd-master---basic-openstack/3200/teut... Jason Dillaman

02/20/2016

05:29 AM Backport #14824: hammer: rbd and pool quota do not go well together
Original description: A partial backport to fix the OSD crash was completed already (#13335). This ticket is to adapt... Loïc Dachary
05:26 AM Bug #12018 (Pending Backport): rbd and pool quota do not go well together
Loïc Dachary

02/19/2016

09:01 PM Bug #14085 (Can't reproduce): TestInternal.ShrinkFlushesCache librbd/test_internal.cc:686: Value ...
Jason Dillaman
08:58 PM Subtask #14412 (Resolved): Track epoch and period for mirrored images
Jason Dillaman
07:28 PM Backport #14824 (Rejected): hammer: rbd and pool quota do not go well together
Josh Durgin
04:07 PM Bug #14822 (Resolved): Journal replay needs to support replaying successful maintenance ops
For example:
journal snap create start
snap create
journal snap create finish
<crash>
-- on replay --
snap cr...
Jason Dillaman

02/18/2016

09:45 PM Cleanup #14809 (Resolved): ImageWatcher unwatch and flush should use async librados methods
See https://github.com/ceph/ceph/pull/7649 for new API Jason Dillaman
05:55 AM Bug #10773: dumpling, hammer, jewel: qemu/tests/qemu-iotests/077 fails
http://pulpito.ceph.com/loic-2016-02-16_22:05:19-rbd-hammer-backports---basic-multi/12638/... Loïc Dachary

02/17/2016

04:37 PM Bug #14789 (Resolved): Truncate of non-existent object results in object map flagged as exists
This won't affect the proper read of data but it will incorrectly affect the disk usage calculations (and it introduc... Jason Dillaman
02:54 AM Cleanup #14781 (New): Reuse ImageCtx::op_work_queue for TaskFinisher's finisher thread
Convert any remaining tasks to be fully async so that the extra thread can be removed from librbd. Jason Dillaman
01:19 AM Bug #14780 (Fix Under Review): librbd: TaskFinisher lifetime no longer matches ImageWatcher
https://github.com/ceph/ceph/pull/7667 Josh Durgin
01:13 AM Bug #14780 (Resolved): librbd: TaskFinisher lifetime no longer matches ImageWatcher
Since notify handling was made async from the librados threads in commit:d898995b0e3ea301b1325f68a0532d57afa3c816 tes... Josh Durgin

02/12/2016

01:37 PM Bug #14373 (Resolved): Cannot reliably create snapshot after freezing QEMU IO
Nathan Cutler
01:37 PM Bug #14030 (Resolved): rbd merge-diff doesn't properly handle >2GB diffs
Nathan Cutler
01:35 PM Bug #13491 (Resolved): misdirected op in rbd balance-reads test
Nathan Cutler
05:42 AM Backport #14610 (Resolved): infernalis: Flattening an rbd image with a running Qemu instances cau...
Loïc Dachary
05:38 AM Backport #14610 (In Progress): infernalis: Flattening an rbd image with a running Qemu instances ...
Loïc Dachary
05:40 AM Backport #14062 (Resolved): infernalis: rbd merge-diff doesn't properly handle >2GB diffs
Loïc Dachary
05:38 AM Backport #14062 (In Progress): infernalis: rbd merge-diff doesn't properly handle >2GB diffs
Loïc Dachary
05:40 AM Backport #14542 (Resolved): infernalis: Cannot reliably create snapshot after freezing QEMU IO
Loïc Dachary
05:38 AM Backport #14542 (In Progress): infernalis: Cannot reliably create snapshot after freezing QEMU IO
Loïc Dachary
05:36 AM Backport #13733 (Resolved): misdirected op in rbd balance-reads test
Loïc Dachary

02/11/2016

04:32 PM Subtask #14739 (Can't reproduce): Support deleting a journal when its being replayed by the RBD m...
Attempting to delete an image which is being mirrored will not work if the RBD mirroring daemon is watching the journ... Jason Dillaman
04:26 PM Feature #14738 (Resolved): Optionally unregister "laggy" journal clients
Support an optional configuration setting for the maximum number of object sets a journal client can be behind before... Jason Dillaman
04:37 AM Bug #14280 (Resolved): cls_rbd: object_map_save should enable checksums
Loïc Dachary
04:37 AM Bug #14164 (Resolved): [ FAILED ] TestLibRBD.SnapRemoveViaLockOwner
Loïc Dachary
02:12 AM Tasks #13959: add python3-specific packages for rbd and rados python bindings
The conversion of rbd.py to cython makes this a little more complex - we'll need to build separate versions of the rb... Josh Durgin

02/10/2016

01:18 AM Bug #14147 (Closed): coredumps in rbd-jewel-distro-basic-mira
no longer happening on jewel Josh Durgin
01:16 AM Bug #8385: RBD / QEMU Crash: Invalid fastbin entry (free)
This may be related to http://tracker.ceph.com/issues/14620 Josh Durgin

02/09/2016

03:40 PM Backport #14620: hammer: unsafe use of libnss PK11_GetBestSlot()
Maybe http://tracker.ceph.com/issues/14367 is the corresponding issue ? Loïc Dachary
06:34 AM Backport #14620 (Fix Under Review): hammer: unsafe use of libnss PK11_GetBestSlot()
Loïc Dachary
06:22 AM Backport #14620 (In Progress): hammer: unsafe use of libnss PK11_GetBestSlot()
Loïc Dachary
06:21 AM Backport #14620: hammer: unsafe use of libnss PK11_GetBestSlot()

PK11_GetBestSlot() does not appear to be threadsafe, and can be called from multiple threads, leading to a crash si...
Loïc Dachary
06:14 AM Bug #14476 (Resolved): rbd: TaskFinisher::cancel should remove event from SafeTimer
Loïc Dachary
06:14 AM Feature #13633 (Resolved): Verify self-managed snapshot functionality on image create
Loïc Dachary
06:14 AM Bug #14452 (Resolved): rbd-replay does not check for EOF and goes to endless loop
Loïc Dachary
05:57 AM Backport #14142 (Resolved): infernalis: Verify self-managed snapshot functionality on image create
Loïc Dachary
05:57 AM Backport #14199 (Resolved): infernalis: [ FAILED ] TestLibRBD.SnapRemoveViaLockOwner
Loïc Dachary
05:57 AM Backport #14321 (Resolved): infernalis: cls_rbd: object_map_save should enable checksums
Loïc Dachary
 

Also available in: Atom