Project

General

Profile

Activity

From 06/01/2016 to 06/30/2016

06/30/2016

06:55 PM Bug #15653: crush: low weight devices get too many objects for num_rep > 1
Adam Emerson
10:52 AM Bug #16553 (New): Removing Writeback Cache Tier Does not clean up Incomplete_Clones
It seem that the issue related to bug #8882 still happen on Infernalis (v9.2.1). After removing the writeback cache b... Lazuardi Nasution

06/28/2016

03:17 AM Bug #16500 (Resolved): ceph_erasure_code_benchmark parameter checking error for LRC plugin
There's no parameter set could pass the parameter checking in cep_erasure_code_benchmark while using LRC.... Ting-Yi Lin

06/24/2016

10:27 PM Bug #15653: crush: low weight devices get too many objects for num_rep > 1
I just need to put in a special case for LIST, (I don't know if I should just not bother with TREE) test it some, and... Adam Emerson

06/18/2016

03:56 PM Bug #16385: rados bench seq and rand tests do not work if op_size != object_size
See pull request wip-krehm-16385 for code fix. Anonymous
11:45 AM Bug #16385 (Fix Under Review): rados bench seq and rand tests do not work if op_size != object_size
rados bench write correctly creates objects whose object_size is a multiple of op_size, but radios bench seq and rand... Anonymous

06/17/2016

06:37 PM Bug #16379 (Closed): [ERROR ] "ceph auth get-or-create for keytype admin returned -1
on master branch, the test has been failing due to below issue
http://teuthology.ovh.sepia.ceph.com/teuthology/teu...
Vasu Kulkarni
01:44 PM Bug #16365 (Resolved): Better network partition detection
We had a situation where due to hardware issues our network was getting lossy for one OSD node, without the Ceph moni... Peter Sabaini
01:37 AM Documentation #16356 (Resolved): doc: manual deployment of ceph monitor needs fix
http://docs.ceph.com/docs/master/install/manual-deployment/
while setting up ceph-mon
step 8 in the doc says to c...
Tamilarasi muthamizhan

06/16/2016

09:26 AM Bug #16258: ceph audit logs are not logging to ceph.audit.log if we specify "mon cluster log file...
naga venkata bokka wrote:
> actually it is documentation defect, it should be stated clearly under this section http...
naga venkata bokka
09:25 AM Bug #16258: ceph audit logs are not logging to ceph.audit.log if we specify "mon cluster log file...
actually it is documentation defect, it should be stated clearly under this section http://docs.ceph.com/docs/jewel/r... naga venkata bokka

06/14/2016

04:12 AM Bug #16279 (Closed): assert(objiter->second->version > last_divergent_update) failed
We build a cluster with 2 osds,
write 100G files, and shutdown the power during deleting the files.
After the machi...
huang jun

06/13/2016

01:20 PM Bug #16236: cache/proxied ops from different primaries (cache interval change) don't order proper...
/a/sage-2016-06-11_06:29:35-rados-jewel---basic-smithi/251518
(at least ceph_test_rados got ERANGE.. presumably it...
Sage Weil
10:06 AM Bug #16258 (New): ceph audit logs are not logging to ceph.audit.log if we specify "mon cluster lo...
In ceph.conf if we specify option "mon cluster log file" both audit and cluster logs are going into the same file cep... naga venkata bokka
04:47 AM Bug #16177 (New): leveldb horrendously slow
Kefu Chai
04:46 AM Bug #16177: leveldb horrendously slow
Adam,
seems we are throttling when exporting the pglog to the OSD (yeah, it's a bug, imo!). you can disable it by ...
Kefu Chai

06/11/2016

09:51 AM Backport #16239 (Resolved): 'ceph tell osd.0 flush_pg_stats' fails in rados qa run
https://github.com/ceph/ceph/pull/15475 Kefu Chai
12:04 AM Bug #16236: cache/proxied ops from different primaries (cache interval change) don't order proper...
Ok, the first sequence of ops came from osd.1, the second from osd.3. The bug is that the queues do not guarantee or... Samuel Just

06/10/2016

11:09 PM Bug #16236: cache/proxied ops from different primaries (cache interval change) don't order proper...
2016-06-10 04:45:58.553823 7fad3adb1700 15 osd.1 pg_epoch: 143 pg[1.3( v 134'1051 (110'825,134'1051] local-les=124 n=... Samuel Just
11:02 PM Bug #16236: cache/proxied ops from different primaries (cache interval change) don't order proper...
Same interval from
2016-06-10 04:45:49.866540 7fad38dad700 10 osd.1 136 dequeue_op 0x7fad5f96e400 prio 63 cost 64308...
Samuel Just
10:42 PM Bug #16236: cache/proxied ops from different primaries (cache interval change) don't order proper...
2016-06-10 04:42:41.619711 7fad4f7df800 0 osd.1 0 using 0 op queue with priority op cut off at 196.
I guess we're...
Samuel Just
10:40 PM Bug #16236: cache/proxied ops from different primaries (cache interval change) don't order proper...
> 2016-06-10 04:45:49.866540 7fad38dad700 10 osd.1 136 dequeue_op 0x7fad5f96e400 client.4123.0:3612 1.df123c4b
wai...
Samuel Just
10:35 PM Bug #16236: cache/proxied ops from different primaries (cache interval change) don't order proper...
Samuel Just wrote:
> The interesting sequence is:
> 2016-06-10 04:45:37.605757 7fad47760700 15 osd.1 134 enqueue_op...
Samuel Just
10:20 PM Bug #16236: cache/proxied ops from different primaries (cache interval change) don't order proper...
The interesting sequence is:
2016-06-10 04:45:37.605757 7fad47760700 15 osd.1 134 enqueue_op 0x7fad5f96e400 client.4...
Samuel Just
10:16 PM Bug #16236: cache/proxied ops from different primaries (cache interval change) don't order proper...
sjust@teuthology:/a/samuelj-2016-06-09_19:31:27-rados-wip-16211-jewel-distro-basic-smithi/249150/remote$ grep -o '201... Samuel Just
10:13 PM Bug #16236 (Won't Fix): cache/proxied ops from different primaries (cache interval change) don't ...
Came from finish_promote, writes happened between first and second copy_get which caused the version to change. Samuel Just

06/09/2016

04:51 PM Bug #16177: leveldb horrendously slow
Since this export/import process is being so slow (now running for a week), and the inodes I'm working on extracting ... Adam Tygart
09:43 AM Bug #16177: leveldb horrendously slow
Hi Adam,
We're looking into why it appears to be spending so much time doing writes and why the write back throttl...
Brad Hubbard

06/08/2016

12:26 PM Bug #16177: leveldb horrendously slow
This is what I used when compressing the osd:
tar -c -z --xattrs --selinux --acls --preserve -f /tmp/ceph-osd.16.tar...
Adam Tygart
06:36 AM Bug #16177 (Need More Info): leveldb horrendously slow
Kefu Chai
05:47 AM Bug #16177: leveldb horrendously slow
i will try to reproduce this with ceph-osd.16.tar.gz, following the steps at http://thread.gmane.org/gmane.comp.file-... Kefu Chai
05:25 AM Bug #16177: leveldb horrendously slow
Adam. when did you collected the trace using "poormansprofiler". i think it was collected when you were importing the... Kefu Chai

06/07/2016

12:26 PM Bug #16177: leveldb horrendously slow
And, to add to that, 5 and a half hours of profiling using both methods:
http://people.cis.ksu.edu/~mozes/perftrac...
Adam Tygart
04:57 AM Bug #16177: leveldb horrendously slow
How about a ... Adam Tygart
04:06 AM Bug #16177: leveldb horrendously slow
Hey Adam,
I don't know a lot about ceph_objectstore_tool but to get the ball rolling my naive approach would be to...
Brad Hubbard
03:47 AM Bug #16177 (Closed): leveldb horrendously slow
Recently ran into an issue using cephfs where loading pg data on osd start (with some lightning fast ssds) took long ... Adam Tygart

06/06/2016

08:16 PM Feature #16172 (New): osd,librados: deprecate unbounded omap_get etc operations
These return an unbounded amount of data (all omap keys). Deprecate these, and make the client-side switch to positi... Sage Weil

06/03/2016

07:09 AM Bug #16133 (New): rados bench: ./common/Mutex.h:96: void Mutex::_pre_unlock(): Assertion `nlock >...
[root@robert xiucai]# rados bench -p ssd 10 seq
2016-06-03 06:38:36.066873 7f6059968700 0 -- :/578428433 >> 10.0.10...
kid hualing
 

Also available in: Atom