Project

General

Profile

Activity

From 08/27/2020 to 09/25/2020

09/25/2020

10:09 PM Bug #46658: Ceph-OSD nautilus/octopus memory leak ?
Hi @Igor,
Thank you for your help :) To answer your questions :
1) First of all what's the current setting for o...
Christophe Hauquiert

09/23/2020

05:31 PM Bug #45765: BlueStore::_collection_list causes huge latency growth pg deletion
Thanks for your help Igor.
> Could you please raise rocksdb_delete_range_threshold to e.g. 1 billion. This wouldn...
Andy Allan
01:39 PM Backport #46350 (In Progress): octopus: ObjectStore/StoreTestSpecificAUSize.SyntheticMatrixCompre...
Nathan Cutler
08:14 AM Backport #47195: octopus: Default value for 'bluestore_volume_selection_policy' is wrong
This update was made using the script "backport-resolve-issue".
backport PR https://github.com/ceph/ceph/pull/37092
m...
Nathan Cutler

09/22/2020

05:58 PM Bug #47053 (Resolved): Default value for 'bluestore_volume_selection_policy' is wrong
Igor Fedotov
05:57 PM Backport #47195 (Resolved): octopus: Default value for 'bluestore_volume_selection_policy' is wrong
Igor Fedotov
01:21 PM Backport #47195: octopus: Default value for 'bluestore_volume_selection_policy' is wrong
Nathan Cutler wrote:
> https://github.com/ceph/ceph/pull/37092
merged
Yuri Weinstein

09/21/2020

09:24 PM Bug #45765: BlueStore::_collection_list causes huge latency growth pg deletion
Andy Allan wrote:
> I'm experiencing a similar problem with our brand new octopus cluster (default settings, radosgw...
Igor Fedotov
07:55 PM Bug #45765: BlueStore::_collection_list causes huge latency growth pg deletion
I'm experiencing a similar problem with our brand new octopus cluster (default settings, radosgw only, 18 NVMe OSDs, ... Andy Allan

09/20/2020

09:25 AM Bug #47551 (Fix Under Review): Some structs aren't bound to mempools properly
Igor Fedotov
12:54 AM Bug #47551 (In Progress): Some structs aren't bound to mempools properly
Igor Fedotov
12:54 AM Bug #47551 (Resolved): Some structs aren't bound to mempools properly
bluestore_shared_blob_t is dynamically allocated but isn't configured for mempool
bluestore_blob_use_tracker_t might...
Igor Fedotov
12:07 AM Bug #46658: Ceph-OSD nautilus/octopus memory leak ?
@Christophe - given that excessive memory consumption occurs on image export (i.e. reading) I guess this might be rel... Igor Fedotov

09/19/2020

09:25 PM Bug #46658: Ceph-OSD nautilus/octopus memory leak ?
I just upgraded from 15.2.4 to 15.2.5 and still facing the same memory problem :/ Christophe Hauquiert

09/18/2020

04:42 PM Bug #47211 (Resolved): nautilus: unrecognised rocksdb_option crashes osd process while starting t...
While running with --resolve-parent, the script "backport-create-issue" noticed that all backports of this issue are ... Nathan Cutler

09/17/2020

07:03 PM Backport #47194: nautilus: Default value for 'bluestore_volume_selection_policy' is wrong
This update was made using the script "backport-resolve-issue".
backport PR https://github.com/ceph/ceph/pull/37091
m...
Nathan Cutler
06:44 PM Backport #47521: nautilus: unrecognised rocksdb_option crashes osd process while starting the osd
This update was made using the script "backport-resolve-issue".
backport PR https://github.com/ceph/ceph/pull/37055
m...
Nathan Cutler
06:42 PM Backport #47521 (Resolved): nautilus: unrecognised rocksdb_option crashes osd process while start...
Nathan Cutler
06:41 PM Backport #47521 (Resolved): nautilus: unrecognised rocksdb_option crashes osd process while start...
https://github.com/ceph/ceph/pull/37055 Nathan Cutler
06:40 PM Bug #47211 (Pending Backport): nautilus: unrecognised rocksdb_option crashes osd process while st...
Nathan Cutler

09/16/2020

07:03 PM Backport #47194 (Resolved): nautilus: Default value for 'bluestore_volume_selection_policy' is wrong
Igor Fedotov
03:52 PM Backport #47194: nautilus: Default value for 'bluestore_volume_selection_policy' is wrong
Nathan Cutler wrote:
> https://github.com/ceph/ceph/pull/37091
merged
Yuri Weinstein

09/15/2020

08:09 PM Bug #47475 (Fix Under Review): Compressed blobs lack checksums
Igor Fedotov
07:20 PM Bug #47475: Compressed blobs lack checksums
As a result corrupted data might reach compressor header decoder which asserts/throws exception Igor Fedotov
07:19 PM Bug #47475 (Resolved): Compressed blobs lack checksums
Looks like a bug in BlueStore::_do_allocate_data() func.
Igor Fedotov
05:49 PM Bug #46490: osds crashing during deep-scrub
Lawrence,
I'm not sure that it's enough to remove object to fix such misreferences. This still leaves inconsistent e...
Igor Fedotov
05:32 PM Bug #47211 (Resolved): nautilus: unrecognised rocksdb_option crashes osd process while starting t...
Igor Fedotov
08:09 AM Bug #46124 (Fix Under Review): Potential race condition regression around new OSD flock()s
Kefu Chai
08:06 AM Bug #46124: Potential race condition regression around new OSD flock()s
i think we also need to include https://github.com/ceph/ceph/pull/37153 for properly returning the right errno. Kefu Chai
06:38 AM Bug #46124: Potential race condition regression around new OSD flock()s
per https://repo.or.cz/glibc.git/blob/HEAD:/sysdeps/posix/flock.c,... Kefu Chai
02:33 AM Bug #47453 (New): checksum failures lead to assert on OSD shutdown in lab tests
... Greg Farnum

09/14/2020

09:19 PM Bug #47443 (Fix Under Review): Hybrid allocator might cause duplicate admin socket command regist...
Igor Fedotov
02:32 PM Bug #47443 (Resolved): Hybrid allocator might cause duplicate admin socket command registration.
This is rather not a big deal for Octopus+ releases which has https://github.com/ceph/ceph/pull/30217
They use a sin...
Igor Fedotov
07:46 PM Bug #47446 (New): No snap trim progress after removing large snapshots
I've issues with cleaning up a cluster with a relatively large CephFS (201M objects, 372 TiB), after snapshots where ... Patrick Salecker
03:54 PM Bug #46124 (New): Potential race condition regression around new OSD flock()s
per the discussion in https://github.com/ceph/ceph/pull/34566, i am reopening this ticket. Kefu Chai
03:31 PM Bug #46124: Potential race condition regression around new OSD flock()s
I believe this bug was not fixed properly; the added commit "just retries 3 times".
This makes success more likely...
Niklas Hambuechen
02:04 PM Bug #46124 (Pending Backport): Potential race condition regression around new OSD flock()s
Igor Fedotov
03:32 PM Bug #47211: nautilus: unrecognised rocksdb_option crashes osd process while starting the osd
https://github.com/ceph/ceph/pull/37055 merged Yuri Weinstein
06:11 AM Backport #46714 (Resolved): nautilus: Rescue procedure for extremely large bluefs log
This update was made using the script "backport-resolve-issue".
backport PR https://github.com/ceph/ceph/pull/36930
m...
Nathan Cutler

09/13/2020

06:49 AM Bug #47271: ceph version 14.2.10-OSD fails
1) Is this happening to multiple OSDs?-yes all the osds has all so the same hdd module
2) Are OSDs able to start...
cephuser2345 user

09/12/2020

02:15 PM Bug #46490: osds crashing during deep-scrub
Neither the new gcc version, nor a new kernel version (upgraded form 4.14 to 5.4) were able to solve this issue for u... Lawrence Smith

09/11/2020

10:08 AM Bug #47174: [BlueStore] Pool/PG deletion(space reclamation) is very slow
Serg D wrote:
> it looks similar to problem https://tracker.ceph.com/issues/47044
yep, indeed
Igor Fedotov
08:46 AM Bug #47174: [BlueStore] Pool/PG deletion(space reclamation) is very slow
Thank you, Igor for looking into this one.
FYI, as this is continuously reproduced this time again we had deleted ...
Vikhyat Umrao
09:13 AM Bug #46886 (Resolved): upgrade/nautilus-x-master: bluefs mount failed to replay log: (14) Bad ad...
Kefu Chai

09/10/2020

06:36 PM Backport #47195 (In Progress): octopus: Default value for 'bluestore_volume_selection_policy' is ...
Neha Ojha
06:34 PM Backport #47194 (In Progress): nautilus: Default value for 'bluestore_volume_selection_policy' is...
Neha Ojha

09/09/2020

11:35 AM Backport #45683 (In Progress): mimic: Large (>=2 GB) writes are incomplete when bluefs_buffered_i...
https://github.com/ceph/ceph/pull/37056 Igor Fedotov
10:22 AM Bug #47174: [BlueStore] Pool/PG deletion(space reclamation) is very slow
it looks similar to problem https://tracker.ceph.com/issues/47044 Serg D

09/08/2020

09:17 PM Bug #47174 (In Progress): [BlueStore] Pool/PG deletion(space reclamation) is very slow
Igor Fedotov
09:15 PM Bug #47211 (Fix Under Review): nautilus: unrecognised rocksdb_option crashes osd process while st...
Missed backport, specific to Nautilus. Igor Fedotov
08:54 PM Bug #47271: ceph version 14.2.10-OSD fails
Could you please answer the following questions:
1) Is this happening to multiple OSDs?
2) Are OSDs able to sta...
Igor Fedotov

09/07/2020

08:10 PM Bug #46575 (Resolved): os/bluestore: simplify Onode pin/unpin logic.
While running with --resolve-parent, the script "backport-create-issue" noticed that all backports of this issue are ... Nathan Cutler
06:44 AM Bug #47330 (Closed): ceph-osd can't start when CURRENT file does not end with newline or content ...

When the machine is shutdown or reboot by IPMI, some ceph-osds cannot be started. The error message is as follo...
daolong zhang

09/06/2020

10:19 AM Backport #46584 (Resolved): octopus: os/bluestore: simplify Onode pin/unpin logic.
This update was made using the script "backport-resolve-issue".
backport PR https://github.com/ceph/ceph/pull/36795
m...
Nathan Cutler

09/02/2020

10:46 PM Backport #46714 (In Progress): nautilus: Rescue procedure for extremely large bluefs log
https://github.com/ceph/ceph/pull/36930
merged
Neha Ojha
12:28 PM Bug #47271 (Closed): ceph version 14.2.10-OSD fails
Hi
we have updated ceph from version 14.2.9 to version.
14.2.10 and since then we are getting osd crash and the o...
cephuser2345 user
12:43 AM Bug #47243 (Duplicate): bluefs _allocate failed then assert
-4> 2020-09-01T21:55:35.451+0800 ffffb5f50500 3 rocksdb: [le/block_based/filter_policy.cc:584] Using legacy Bloo... chunsong feng

08/31/2020

11:57 AM Backport #47213 (In Progress): nautilus: BlueFS volume selector assert
Igor Fedotov
11:55 AM Backport #47213: nautilus: BlueFS volume selector assert
https://github.com/ceph/ceph/pull/36909 Igor Fedotov
11:45 AM Backport #47213 (Resolved): nautilus: BlueFS volume selector assert
https://github.com/ceph/ceph/pull/36909 Igor Fedotov
11:06 AM Bug #47211 (Resolved): nautilus: unrecognised rocksdb_option crashes osd process while starting t...
while upgrading cluster from 14.2.9 to 14.2.11 unrecognised (legacy?) rocksdb_option crashes osd process while starti... Pietari Hyvärinen
10:43 AM Bug #43538 (Pending Backport): BlueFS volume selector assert
Igor Fedotov
09:10 AM Bug #43538: BlueFS volume selector assert
I'm seeing this killing almost every osd one by one ( during 48h cycle) on one of our production clusters after upgra... Pietari Hyvärinen

08/28/2020

02:38 PM Backport #47195 (Resolved): octopus: Default value for 'bluestore_volume_selection_policy' is wrong
https://github.com/ceph/ceph/pull/37092 Nathan Cutler
02:37 PM Backport #47194 (Resolved): nautilus: Default value for 'bluestore_volume_selection_policy' is wrong
https://github.com/ceph/ceph/pull/37091 Nathan Cutler
02:37 PM Bug #47069 (Resolved): os/bluestore: dump onode that has too many spanning blobs
While running with --resolve-parent, the script "backport-create-issue" noticed that all backports of this issue are ... Nathan Cutler
10:36 AM Bug #46124 (Resolved): Potential race condition regression around new OSD flock()s
Igor Fedotov

08/27/2020

06:31 PM Bug #47174: [BlueStore] Pool/PG deletion(space reclamation) is very slow
- Uploaded debug log file and osd.190.perf.dump - ... Vikhyat Umrao
06:19 PM Bug #47174: [BlueStore] Pool/PG deletion(space reclamation) is very slow
- dump_mempools:... Vikhyat Umrao
05:54 PM Bug #47174: [BlueStore] Pool/PG deletion(space reclamation) is very slow
I have selected one OSD for providing the debug log information. perf top from the same OSD.
ceph 84323 ...
Vikhyat Umrao
05:29 PM Bug #47174 (Resolved): [BlueStore] Pool/PG deletion(space reclamation) is very slow
Version - 14.2.8 was also reproduced in 12.2.12
- We use cosbench to fill the cluster - obviously for the RGW work...
Vikhyat Umrao
05:21 PM Backport #46584: octopus: os/bluestore: simplify Onode pin/unpin logic.
Igor Fedotov wrote:
> https://github.com/ceph/ceph/pull/36795
merged
Yuri Weinstein
02:29 PM Bug #47053 (Pending Backport): Default value for 'bluestore_volume_selection_policy' is wrong
Kefu Chai
02:15 PM Bug #44731 (Closed): Space leak in Bluestore
Doesn't look like a Ceph issue. Neha Ojha
02:11 PM Bug #46490 (Need More Info): osds crashing during deep-scrub
Neha Ojha
02:03 PM Bug #46994 (Need More Info): 14.2.11 OSD crash BlueFS.cc: 1662: FAILED ceph_assert(r == 0)
Neha Ojha
10:56 AM Backport #47070 (Resolved): nautilus: os/bluestore: dump onode that has too many spanning blobs
This update was made using the script "backport-resolve-issue".
backport PR https://github.com/ceph/ceph/pull/36756
m...
Nathan Cutler
 

Also available in: Atom