Activity
From 05/14/2017 to 06/12/2017
06/12/2017
- 08:46 PM Bug #20211: rgw: bucket index not syncing when rados cluster abnormally
- fang yuxiang wrote:
> Yehuda Sadeh wrote:
> > I don't understand what step (4) is. What do you mean fail bucket ind... - 08:35 PM Backport #20260 (In Progress): hammer: Increase verbosity level on RGWObjManifest line
- 08:34 PM Backport #20260 (Rejected): hammer: Increase verbosity level on RGWObjManifest line
- https://github.com/ceph/ceph/pull/15379
- 08:34 PM Backport #20269 (Resolved): kraken: wrong object size after copy of uncompressed multipart objects
- https://github.com/ceph/ceph/pull/16179
- 08:34 PM Backport #20268 (Resolved): kraken: get wrong content when download object with specific range wh...
- https://github.com/ceph/ceph/pull/16178
- 08:34 PM Backport #20263 (Resolved): kraken: "datalog trim" can't work as expected
- https://github.com/ceph/ceph/pull/16175
- 08:34 PM Backport #20262 (Resolved): jewel: "datalog trim" can't work as expected
- https://github.com/ceph/ceph/pull/16299
- 08:34 PM Backport #20261 (Resolved): kraken: 'radosgw-admin usage show' listing 0 bytes_sent/received
- https://github.com/ceph/ceph/pull/16174
- 08:32 PM Bug #16682 (Resolved): ceph jewel failed create s3 tpye subuser from admin rest api
- 08:11 PM Bug #20259: Increase verbosity level on RGWObjManifest line
- hammer backport pr: https://github.com/ceph/ceph/pull/15379
- 08:10 PM Bug #20259 (Resolved): Increase verbosity level on RGWObjManifest line
- https://github.com/ceph/ceph/pull/7285
- 06:53 PM Backport #19786 (Resolved): jewel: ceph jewel failed create s3 tpye subuser from admin rest api
- 06:52 PM Bug #11239 (Resolved): civetweb defaults to libssl.so and libcrypto.so when versions not passed
- 06:42 PM Feature #20235 (Fix Under Review): vstart.sh fails to start more than one radosgw process
- 01:08 PM Feature #20235: vstart.sh fails to start more than one radosgw process
- Pull request https://github.com/ceph/ceph/pull/15632
- 12:47 PM Feature #20235 (In Progress): vstart.sh fails to start more than one radosgw process
- 05:17 PM Bug #17888 (Resolved): rgw:fix for deleting objects name beginning and ending with underscores of...
- 02:40 PM Bug #20071 (Pending Backport): wrong object size after copy of uncompressed multipart objects
- 02:33 PM Bug #20255 (New): rgw: Size of copied object is wrong in the index
- When copying an object that is longer than @max_head_size@ (currently 4194304 bytes), the size of the copied object i...
- 02:24 PM Bug #20234: rgw: shadow objects are sometimes not removed
- More debugging shows that the pure copy operation produces the correct object, it is the put_acl call, which s3cmd do...
- 12:59 PM Bug #20234: rgw: shadow objects are sometimes not removed
- Some more data:
This only seems to affect the S3 API, objects copied via the Swift API do not seem to be affected.... - 01:17 PM Bug #20100 (Pending Backport): get wrong content when download object with specific range when co...
- 12:46 PM Bug #17164: Orphan data gets leaked on Bucket deletion
- Reviving this old thread by a colleague, the PR submitted earlier [1] had failed teuthology rgw run, due to radosgw-a...
- 10:48 AM Support #20253 (New): Ceph RGW Users (rgw keystone implicit tenants)
- Apologies, I am not 100% sure if this expected behaviour or a bug, Happy to re-write as feature request if needed.
... - 06:26 AM Bug #20247 (Fix Under Review): get bucket policy return wrong format result
- 06:26 AM Bug #20247: get bucket policy return wrong format result
- https://github.com/ceph/ceph/pull/15586
- 02:08 AM Bug #20247 (Resolved): get bucket policy return wrong format result
- hi, when i get bucket policy of bucket test3...
- 06:12 AM Bug #20252 (Fix Under Review): RGW:RGWPutBucketPolicy error when set BucketPolicy again without d...
- 05:30 AM Bug #20252 (Resolved): RGW:RGWPutBucketPolicy error when set BucketPolicy again without delete pr...
- hi,when i create a bucket and set policy to the bucket,use
s3cmd setpolicy 2-referpolicy s3://test1
it set poli... - 05:44 AM Bug #20218 (Closed): rgw: some problem set bucket policy in master branch
- Closed as per joke lee's comment.
- 03:28 AM Bug #20251: rgw: meta sync thread crash at RGWMetaSyncShardCR
- I found this issue was caused by the lease_stack reference count, and will open a pr for this
- 03:26 AM Bug #20251 (Resolved): rgw: meta sync thread crash at RGWMetaSyncShardCR
- radosgw crash in meta sync as below:
ceph version 10.2.3-124-g82a9117 (82a9117651e68c5b843b364353081da1b284475f)
... - 02:53 AM Bug #20250 (Rejected): RGW instance can't startup successfully in Ubuntu 16.04
- cat /etc/issue
Ubuntu 16.04.2 LTS \n \l
one step in /etc/init.d/radosgw failed:
. /lib/lsb/init-functions
r... - 02:11 AM Bug #20248 (Resolved): bucket policy IfExists in wrong position
- as http://docs.aws.amazon.com/IAM/latest/UserGuide/reference_policies_elements.html metioned
IfExists to the end of ...
06/10/2017
- 10:03 PM Bug #20244: multisite: log_meta on secondary zone causes continuous loop of metadata sync
- The radosgw-admin commands for zone/zonegroup generally prevent this from happening, but it was observed after an upg...
- 09:59 PM Bug #20244 (Resolved): multisite: log_meta on secondary zone causes continuous loop of metadata sync
- If log_meta is set on a secondary zone, it will add another mdlog entry for everything it syncs from the master zone....
06/09/2017
- 09:05 PM Backport #20088: valgrind reports leak in RGWRemoteMetadataCR
- h3. original description
similar leak report on RGWReadRemoteMDLogShardInfoCR as well... - 08:40 PM Bug #20231 (Fix Under Review): rgw: can't download large random file when compression enable
- 07:03 PM Bug #20231: rgw: can't download large random file when compression enable
- previous pr was closed carelessly.
-https://github.com/ceph/ceph/pull/15605-
I think that one must use s3 multi... - 05:47 AM Bug #20231: rgw: can't download large random file when compression enable
- -https://github.com/ceph/ceph/pull/15591-
- 05:43 AM Bug #20231 (Resolved): rgw: can't download large random file when compression enable
- one can reproduce issue by steps(using tools s3cmd):
1. create a bucket with compression
2. use dd to create a big ... - 06:04 PM Bug #20210: rgw: fix input/output err when using radosgw-admin bi list
- Yehuda Sadeh wrote:
> Does that still happen, even after: https://github.com/ceph/ceph/pull/14981 ?
sorry, I thin... - 05:39 PM Bug #20211: rgw: bucket index not syncing when rados cluster abnormally
- Yehuda Sadeh wrote:
> I don't understand what step (4) is. What do you mean fail bucket index add op.
I consider ... - 05:13 PM Bug #20211: rgw: bucket index not syncing when rados cluster abnormally
- Yehuda Sadeh wrote:
> I don't understand what step (4) is. What do you mean fail bucket index add op.
1. rados cl... - 02:05 PM Bug #20037: rgw log includes zero byte sometimes
- https://github.com/ceph/ceph/pull/15599
- 01:17 PM Bug #20037 (In Progress): rgw log includes zero byte sometimes
- seems like we're doing a str. assign where bufferlist::c_Str() itself would be sufficient
- 12:56 PM Feature #20235 (Fix Under Review): vstart.sh fails to start more than one radosgw process
- When running something like...
- 12:53 PM Bug #20234: rgw: shadow objects are sometimes not removed
- Verified that this also affects Jewel. But before committing a fix, better make sure that http://tracker.ceph.com/iss...
- 12:13 PM Bug #20234 (Resolved): rgw: shadow objects are sometimes not removed
- Testing on master, after copying an object, then deleting the copy and finally deleting the original object, the asso...
- 12:18 PM Bug #20079 (Closed): rgw jewel: valgrind failure in receive_header
- 12:18 PM Bug #20079: rgw jewel: valgrind failure in receive_header
- fixed with https://github.com/ceph/ceph/pull/15382
- 06:27 AM Bug #20232 (New): can not set more than one refer on bucket policy
- when i set ...
- 06:22 AM Bug #20218: rgw: some problem set bucket policy in master branch
- sorry, this is my problem,i use git apply to cherry pick some code, and after i reclone the code ,this no NoSuchBucke...
06/08/2017
- 08:28 PM Bug #19870: rgw: bytes_send and bytes_recv in the msg of usage show returning is 0 in master branch
- cbodley has a fix for a similar problem with swift:
https://github.com/ceph/ceph/pull/14042
I believe it may be ... - 06:39 PM Bug #18355 (In Progress): when call the multiparts upload API TransferManager.upload() by s3 java...
- 06:38 PM Bug #18437: error log's error code is not right in data sync
- The error code there should probably be changed to signed int. Should be trivial to fix.
- 06:34 PM Bug #20107: rgw: Truncated objects
- Ah, no, no different pools involved.
And no, with the volume of requests we are dealing with, we are already at ab... - 06:30 PM Bug #20107: rgw: Truncated objects
- Do you have a reference to that bug? I haven't been able to find anything that looked like it would match. Also, the ...
- 06:21 PM Bug #20107: rgw: Truncated objects
- The bug I mentioned was dealing with copying objects that resided on different pools. Do you have a scenario where th...
- 06:15 PM Bug #20107: rgw: Truncated objects
- A copy operation should increase the tail's refcount. The garbage collection will try to remove it by dropping the re...
- 06:34 PM Bug #20128 (Resolved): rgw: teuthology failure in the rgw task (configured via rgw_snaps.yaml)
- 06:33 PM Bug #20166: RGW: Bad error handling when tail object is missing
- Do you have any proposed change?
- 06:27 PM Bug #20190 (Pending Backport): "datalog trim" can't work as expected
- 06:23 PM Bug #20205: rgw: s3 api, need to return error NoSuchVersion when object does not exist and versio...
- https://github.com/ceph/ceph/pull/15528
- 06:23 PM Bug #20206: rgw: delete null versioned object doesn't return x-amz-version-id http header
- https://github.com/ceph/ceph/pull/15528
- 06:06 PM Bug #20210: rgw: fix input/output err when using radosgw-admin bi list
- Does that still happen, even after: https://github.com/ceph/ceph/pull/14981 ?
- 06:03 PM Bug #20211: rgw: bucket index not syncing when rados cluster abnormally
- I don't understand what step (4) is. What do you mean fail bucket index add op.
- 05:56 PM Bug #18596: rgw: object owner should not be able to download the obj without container permission...
- Not quite. There are a set of permissions that will do that, however, if these permissions are not set then users wil...
- 03:49 PM Bug #17779 (Fix Under Review): rgw: s3 API does not honor rgw_keystone_implicit_tenants when keys...
- https://github.com/ceph/ceph/pull/15572
- 02:12 PM Bug #20225 (Resolved): 'radosgw-admin usage show' listing 0 bytes_sent/received
- Fixed in master with: https://github.com/ceph/ceph/pull/14042
- 07:15 AM Bug #20220 (Resolved): mismatch Signature in swift formpost
- When use the feature of swift FormPost correctly according to the doc, it will also return "FormPost: Invalid Signatu...
- 05:47 AM Backport #19847 (In Progress): jewel: multisite: incremental metadata sync does not properly adva...
- 01:19 AM Bug #20218 (Closed): rgw: some problem set bucket policy in master branch
- i test set bucket policy in master branch ceph cluster, i test RGW Bucket Policies #14307 in my ceph cluster, but i s...
06/07/2017
- 05:46 PM Backport #19847: jewel: multisite: incremental metadata sync does not properly advance to next pe...
- https://github.com/ceph/ceph/pull/15556
- 05:35 PM Bug #20112 (Resolved): Use boost::regex instead of std::regex for rgw_iam_policy.cc and rgw_role.cc
- 01:19 PM Bug #20211 (Fix Under Review): rgw: bucket index not syncing when rados cluster abnormally
- 08:25 AM Bug #20211: rgw: bucket index not syncing when rados cluster abnormally
- https://github.com/ceph/ceph/pull/15545
- 08:15 AM Bug #20211 (Fix Under Review): rgw: bucket index not syncing when rados cluster abnormally
- Bucket index syncing Issue occur in our environment.
I think one can reproduce it by steps:
1. deploy multisit... - 01:19 PM Bug #20210 (Fix Under Review): rgw: fix input/output err when using radosgw-admin bi list
- 07:53 AM Bug #20210: rgw: fix input/output err when using radosgw-admin bi list
- https://github.com/ceph/ceph/pull/15543
- 07:45 AM Bug #20210 (Fix Under Review): rgw: fix input/output err when using radosgw-admin bi list
- radosgw-admin bi list report input/output err;
one can reproduce it by steps:
1.deploy a multisite environment and ... - 12:56 PM Bug #20107: rgw: Truncated objects
- I found an example that is recent enough so that there are still some relevant logs available. Here an object got cop...
- 09:22 AM Bug #20107: rgw: Truncated objects
- It seems this is happening when an object is copied, the copy has the same prefix for shadow objects as the original ...
- 12:06 PM Backport #20088: valgrind reports leak in RGWRemoteMetadataCR
- Note the master commit is already in kraken:...
- 12:02 PM Backport #20088 (Resolved): valgrind reports leak in RGWRemoteMetadataCR
- 11:34 AM rgw-testing Bug #20214 (New): Add a test for multisite with 5G object
- 11:34 AM rgw-testing Bug #20213 (New): Add test for POST Content-MD5 header
- 07:15 AM Bug #19870: rgw: bytes_send and bytes_recv in the msg of usage show returning is 0 in master branch
- I may have found the teuthology tests. in ceph, master branch, "qa/tasks/radosgw_admin.py" and "qa/tasks/radosgw_adm...
- 06:50 AM Bug #19432: rgw: folders starting with "_" underscore are not in bucket index
- Sorry for the delay.
I have reproduced it on master with Daniel Biazus instructions.
I could not reproduce it wit...
06/06/2017
- 10:29 PM Bug #20206 (New): rgw: delete null versioned object doesn't return x-amz-version-id http header
- 09:27 PM Bug #20205 (New): rgw: s3 api, need to return error NoSuchVersion when object does not exist and ...
- 06:36 PM Bug #17964: rgw: multipart parts on versioned bucket create versioned bucket index entries
- https://github.com/ceph/ceph/pull/14500
- 06:34 PM Bug #16355 (In Progress): radosgw-admin command, adding the tenants can not view the user usage
- 06:29 PM Bug #19194 (In Progress): after s3cmd put file to bucket, 'radosgw-admin usage show --categories=...
- 06:28 PM Bug #19219 (Resolved): Data sync problem in multisite
- 06:28 PM Bug #19587 (In Progress): rgw: put lifecycle configuration fails if Prefix is not set
- 06:24 PM Bug #19285: rgw: user quota did not work well on multipart upload
- 05:57 PM Bug #19739 (Fix Under Review): Content-MD5 header is not validated with POST uploads
- 05:57 PM Bug #19870: rgw: bytes_send and bytes_recv in the msg of usage show returning is 0 in master branch
- while we're at it, we need to make sure that there's sufficient teuthology coverage for this issue. I think we do tes...
- 05:53 PM Bug #20177: RGW lifecycle not expiring objects due to permissions on lc pool
- Some notes (cribbed from email by Dan Gryniewicz):
The lifecycle config looks sane; Dan, do you see anything here?... - 03:52 PM Backport #19847: jewel: multisite: incremental metadata sync does not properly advance to next pe...
- i'll take this one - i expect to have a pr ready today
- 03:21 PM Bug #20201 (Resolved): radosgw refuses upload when Content-Type missing from POST policy
- When doing an S3 upload, if Content-Type is missing from the policy part of the upload, then radosgw refuses the uplo...
- 01:53 PM Bug #20190 (Fix Under Review): "datalog trim" can't work as expected
- 07:56 AM Bug #20190: "datalog trim" can't work as expected
- https://github.com/ceph/ceph/pull/15507
- 07:43 AM Bug #20190 (Resolved): "datalog trim" can't work as expected
- The radosgw-admin command datalog trim can't remove the log as expected.
It returns "ERROR: trim_entries(): (61) No ... - 10:01 AM Backport #19723 (In Progress): kraken: rgw_file: introduce rgw_lookup type hints
- 10:00 AM Backport #20195 (In Progress): kraken: rgw_file: restore (corrected) fix for dir "partial match"...
- 09:59 AM Backport #20195 (Resolved): kraken: rgw_file: restore (corrected) fix for dir "partial match" (r...
- https://github.com/ceph/ceph/pull/13871
- 07:47 AM Backport #19144 (In Progress): kraken: rgw_file: FHCache residence check should be exhaustive
- 02:21 AM Bug #20112 (Fix Under Review): Use boost::regex instead of std::regex for rgw_iam_policy.cc and r...
- https://github.com/ceph/ceph/pull/15498
06/05/2017
- 09:02 AM Backport #19767 (In Progress): jewel: rgw: Delete non-empty bucket in slave zonegroup
- 03:25 AM Feature #20179: rgw: subuser usage log statistics
- https://github.com/ceph/ceph/pull/15064
- 03:24 AM Feature #20179 (Fix Under Review): rgw: subuser usage log statistics
- current usage log statistics only support user level,this feature will expand usage log statistics to subuser level.
06/04/2017
- 07:28 PM Bug #20177 (Resolved): RGW lifecycle not expiring objects due to permissions on lc pool
- see http://lists.ceph.com/pipermail/ceph-users-ceph.com/2017-June/018367.html
- 07:12 PM Backport #19613 (Need More Info): jewel: multisite: rest api fails to decode large period on 'per...
- 07:12 PM Backport #19613: jewel: multisite: rest api fails to decode large period on 'period commit'
- This backport depends on https://github.com/ceph/ceph/commit/35375b75160 - marked the PR DNM until we determine wheth...
- 01:14 PM Bug #20176 (Fix Under Review): rgw bugfix:s3 aws v2 signature priority between header['X-Amz-Date...
- 01:01 PM Bug #20176 (Resolved): rgw bugfix:s3 aws v2 signature priority between header['X-Amz-Date'] and h...
- as http://docs.aws.amazon.com/AmazonS3/latest/dev/RESTAuthentication.html Time Stamp Requirement section metioned
... - 09:37 AM Backport #19772 (In Progress): jewel: rgw: swift: disable revocation thread under certain circums...
- 09:35 AM Backport #20155 (In Progress): jewel: fix: rgw crashed caused by shard id out of range when listi...
- 07:54 AM Backport #19758: jewel: multisite: after CreateBucket is forwarded to master, local bucket may us...
- Casey Bodley wrote:
> https://github.com/ceph/ceph/pull/15010 merged, please include in backports
@Casey: That co... - 07:53 AM Backport #20159 (In Progress): jewel: rgw_file: RGWLibFS::get_inst() not stable
- 07:50 AM Backport #20157 (Resolved): jewel: rgw_file: handle chunked readdir
- 07:49 AM Bug #19624: rgw_file: handle chunked readdir
- @Matt - https://github.com/ceph/ceph/pull/14458 was backported to jewel by https://github.com/ceph/ceph/pull/14653 . ...
- 07:46 AM Backport #20142 (In Progress): jewel: rgw_file: unlink assert unsafe in multi-gateway environments
- 07:44 AM Backport #20138 (In Progress): jewel: rgw_file: attempt to unlink non-empty bucket could lead to ...
- 07:43 AM Backport #20144 (In Progress): jewel: rgw_file: cannot delete bucket w/uxattrs
- 07:28 AM Bug #20064 (Pending Backport): allow empty data_extra_pool in zone placement
- 07:24 AM Bug #20064 (Need More Info): allow empty data_extra_pool in zone placement
- The master commit changes a line in the function "rgw_get_obj_data_pool" which was added by 9c1b828e838. That commit ...
- 07:28 AM Backport #20136 (Need More Info): jewel: allow empty data_extra_pool in zone placement
- The master commit changes a line in the function "rgw_get_obj_data_pool" which was added by 9c1b828e838. That commit ...
- 07:27 AM Backport #20137 (Need More Info): kraken: allow empty data_extra_pool in zone placement
- The master commit changes a line in the function "rgw_get_obj_data_pool" which was added by 9c1b828e838. That commit ...
- 07:06 AM Backport #20029 (In Progress): jewel: multisite: radosgw-admin period commands cant use --remote ...
06/03/2017
- 06:34 PM Backport #20020 (In Progress): jewel: multisite: RGWPeriodPusher CRThread startup races with RGWH...
- 06:26 PM Backport #20018 (In Progress): jewel: multisite: memory leak on failed lease in RGWDataSyncShardCR
- 06:22 PM Backport #20003 (In Progress): jewel: rgw_file: v3 write timer does not close open handles
- 06:21 PM Backport #19847 (Need More Info): jewel: multisite: incremental metadata sync does not properly a...
- Non-trivial RGW multisite backport. Abhi, want to have a go at it?
- 06:16 PM Backport #19844 (Need More Info): jewel: Add custom user data support in bucket index
- 06:16 PM Backport #19844: jewel: Add custom user data support in bucket index
- Ordinarily we don't backport features to stable releases, but assigning to Pavan if he wants to have a go at it.
- 06:09 PM Backport #19838 (In Progress): jewel: reduce log level of 'storing entry at' in cls_log
- 06:07 PM Backport #19775 (In Progress): jewel: multisite: realm rename does not propagate to other clusters
- 06:05 PM Backport #19768 (Need More Info): jewel: [multisite] operating bucket's acl&cors is not restricte...
- Needs https://github.com/ceph/ceph/pull/15260 once it's merged.
- 06:05 PM Backport #19768 (In Progress): jewel: [multisite] operating bucket's acl&cors is not restricted o...
- 04:37 PM Bug #18208: rgw: when uploading the objects continuesly in the versioned bucket, some objects wil...
- Zhandong Guo wrote:
> Do we have plan to patch this in 10.2.8?
This is probably missing the cutoff, and will be i... - 04:36 PM Backport #19765 (In Progress): jewel: rgw: when uploading the objects continuesly in the versione...
- 04:17 PM Backport #19158 (Resolved): jewel: RGW health check errors out incorrectly
- 04:16 PM Backport #18903 (Need More Info): jewel: rgw: first write also tries to read object
- caused tests to fail
- 04:15 PM Backport #19764 (In Progress): jewel: set latest object's acl failed
- 04:13 PM Backport #19758 (Need More Info): jewel: multisite: after CreateBucket is forwarded to master, lo...
- 04:12 PM Backport #19758: jewel: multisite: after CreateBucket is forwarded to master, local bucket may us...
- Need More Info (see comment in master issue)
- 04:10 PM Backport #19758 (In Progress): jewel: multisite: after CreateBucket is forwarded to master, local...
- 04:09 PM Bug #19745: multisite: after CreateBucket is forwarded to master, local bucket may use different ...
- Casey Bodley wrote:
> https://github.com/ceph/ceph/pull/15010 merged, please include in backports
@Casey: That co... - 02:40 PM Backport #19724 (Resolved): jewel: RGW S3 v4 authentication issue with X-Amz-Expires
- 02:36 PM Backport #19722 (Resolved): jewel: rgw_file: introduce rgw_lookup type hints
- 02:30 PM Backport #19720 (In Progress): jewel: rgw_file: fix size and (c|m)time unix attrs in write_finish
- 02:21 PM Backport #19143 (Resolved): jewel: rgw_file: FHCache residence check should be exhaustive
- 02:19 PM Backport #19169 (Resolved): jewel: rgw_file: allow setattr on placeholder (common_prefix) direct...
- 02:19 PM Backport #19167 (Resolved): jewel: rgw_file: RGWReaddir (and cognate ListBuckets request) don't ...
- 02:19 PM Backport #19165 (Resolved): jewel: rgw_file: "exact match" invalid for directories, in RGWLibFS:...
- 02:18 PM Backport #19161 (Resolved): jewel: rgw_file: fix marker computation
- 02:16 PM Backport #19615 (In Progress): jewel: multisite: bucket zonegroup redirect not working
- 02:14 PM Backport #19613 (In Progress): jewel: multisite: rest api fails to decode large period on 'period...
- 12:41 PM Backport #19532 (Need More Info): jewel: radosgw-admin: unable to set global quotas in RGWPeriodC...
- 12:34 PM Backport #19532 (In Progress): jewel: radosgw-admin: unable to set global quotas in RGWPeriodConfig
06/02/2017
- 12:35 PM Bug #20037: rgw log includes zero byte sometimes
- Similar things happen in the output of @radosgw-admin object stat@, some strings are shown with a trailing @\u0000@:
... - 12:05 PM Bug #20166 (New): RGW: Bad error handling when tail object is missing
- When doing an S3 GET on an object, it may happen that one of the tail objects does not exist. When radosgw notices th...
- 07:37 AM Backport #20160 (Rejected): kraken: rgw_file: RGWLibFS::get_inst() not stable
- 07:37 AM Backport #20159 (Resolved): jewel: rgw_file: RGWLibFS::get_inst() not stable
- https://github.com/ceph/ceph/pull/15459
- 07:37 AM Backport #20158 (Resolved): kraken: rgw_file: handle chunked readdir
- https://github.com/ceph/ceph/pull/13871
- 07:37 AM Backport #20157 (Resolved): jewel: rgw_file: handle chunked readdir
- https://github.com/ceph/ceph/pull/14653
- 07:36 AM Backport #20156 (Resolved): kraken: fix: rgw crashed caused by shard id out of range when listing...
- https://github.com/ceph/ceph/pull/16173
- 07:36 AM Backport #20155 (Resolved): jewel: fix: rgw crashed caused by shard id out of range when listing ...
- https://github.com/ceph/ceph/pull/15465
- 07:36 AM Backport #20147 (Resolved): kraken: rgw: 'gc list --include-all' command infinite loop the first ...
- https://github.com/ceph/ceph/pull/16139
- 07:36 AM Backport #20146 (Resolved): jewel: rgw: 'gc list --include-all' command infinite loop the first 1...
- https://github.com/ceph/ceph/pull/15719
- 07:36 AM Backport #20145 (Rejected): kraken: rgw_file: cannot delete bucket w/uxattrs
- 07:36 AM Backport #20144 (Resolved): jewel: rgw_file: cannot delete bucket w/uxattrs
- https://github.com/ceph/ceph/pull/15459
- 07:36 AM Backport #20143 (Rejected): kraken: rgw_file: unlink assert unsafe in multi-gateway environments
- 07:36 AM Backport #20142 (Resolved): jewel: rgw_file: unlink assert unsafe in multi-gateway environments
- https://github.com/ceph/ceph/pull/15459
- 07:35 AM Backport #20139 (Rejected): kraken: rgw_file: attempt to unlink non-empty bucket could lead to hang
- 07:35 AM Backport #20138 (Resolved): jewel: rgw_file: attempt to unlink non-empty bucket could lead to hang
- https://github.com/ceph/ceph/pull/15459
- 07:35 AM Backport #20137 (Rejected): kraken: allow empty data_extra_pool in zone placement
- 07:35 AM Backport #20136 (Rejected): jewel: allow empty data_extra_pool in zone placement
06/01/2017
- 03:57 PM Backport #20088 (In Progress): valgrind reports leak in RGWRemoteMetadataCR
- 12:12 AM Bug #20128: rgw: teuthology failure in the rgw task (configured via rgw_snaps.yaml)
- this might solve it:
https://github.com/ceph/ceph/pull/15400
- 12:08 AM Bug #20128 (Resolved): rgw: teuthology failure in the rgw task (configured via rgw_snaps.yaml)
- See the failure here: http://pulpito.ceph.com/sage-2017-05-31_18:45:30-rados-wip-sage-testing---basic-smithi/1248627
...
05/31/2017
- 03:22 PM Bug #18087 (Resolved): rgw: LCWorker's worktime is not the same as config rgw_lifecycle_work_time.
- 02:26 PM Backport #20015 (In Progress): kraken: multisite: bi_list() decode failures
- https://github.com/ceph/ceph/pull/15384
- 01:35 PM Backport #20014 (In Progress): jewel: multisite: bi_list() decode failures
- https://github.com/ceph/ceph/pull/15383
- 01:32 PM Backport #20088: valgrind reports leak in RGWRemoteMetadataCR
- wip jewel pr: https://github.com/ceph/ceph/pull/15382
- 12:05 PM Bug #20121 (Fix Under Review): rgw_file: missing lock protection for readdir against gc
- 09:27 AM Bug #20121: rgw_file: missing lock protection for readdir against gc
- https://github.com/ceph/ceph/pull/15329
- 09:26 AM Bug #20121: rgw_file: missing lock protection for readdir against gc
- The state.mtx protects state.events accessed both from readdir and gc.
- 09:26 AM Bug #20121 (Resolved): rgw_file: missing lock protection for readdir against gc
- The state.mtx protects state.events accessed both from readdir and gc.
- 07:18 AM Feature #19510: per-object storage class
- Chang Liu wrote:
> Hi, Is there anyone else doing this? let me know please.
Hi~ Liu Chang:
I'm investing to br... - 02:58 AM Bug #20112: Use boost::regex instead of std::regex for rgw_iam_policy.cc and rgw_role.cc
- I only found it because it bit me too. Seems like it's a big club :)
- 02:24 AM Bug #20112: Use boost::regex instead of std::regex for rgw_iam_policy.cc and rgw_role.cc
- i confirm this issue. i also found that that gcc 4.8's regex implementation was buggy and incomplete the hard way.
... - 01:36 AM Bug #20100: get wrong content when download object with specific range when compression was enabled
- the previous description is not proper.
In fact, this issue exists in s3 download with specific range for a multip...
05/30/2017
- 11:35 PM Bug #20112: Use boost::regex instead of std::regex for rgw_iam_policy.cc and rgw_role.cc
- please dont just do a search-and-replace. Put a regex in the ceph:: namespace so we can map it to the right thing dep...
- 11:08 PM Bug #20112 (Resolved): Use boost::regex instead of std::regex for rgw_iam_policy.cc and rgw_role.cc
- src/rgw/rgw_iam_policy.cc and src/rgw/rgw_role.cc include code that uses the std::regex header.
Per https://bugzi... - 08:09 PM Bug #17934 (Fix Under Review): rgw: /info lacks swift.max_meta_count
- https://github.com/ceph/ceph/pull/15369
- 03:29 PM Bug #20098 (Fix Under Review): rgw: connection reset or radosgw crashed when download large zero ...
- https://github.com/ceph/ceph/pull/15335
- 03:00 PM Bug #20100 (Fix Under Review): get wrong content when download object with specific range when co...
- https://github.com/ceph/ceph/pull/15323
- 12:39 PM Bug #20048 (Resolved): Ceph no longer builds for s390x after Beast submodule enabled
- 12:07 PM Bug #20107 (Resolved): rgw: Truncated objects
- On a large cluster, we are seeing some objects being truncated at 512kb size. The behaviour when trying to access the...
05/27/2017
- 08:32 AM Bug #20101: swift slo/dlo download err when enable compression
- seems that compression is not supported for swift slo/dlo
- 03:47 AM Bug #20101 (New): swift slo/dlo download err when enable compression
- swift slo/dlo object can't be downloaded.
issue can be reproduced with belown steps:
ceph version 10.2.3-62-g5f00... - 03:04 AM Bug #20100 (Resolved): get wrong content when download object with specific range when compressio...
- create a bucket with compression enabled
Given a 32MB file, do steps:
1.divided the file into 4 parts
2.uplo... - 02:40 AM Bug #20098: rgw: connection reset or radosgw crashed when download large zero object with compres...
- large zero object has a large compression rate, even 4M compressed data can decompress several GB data. Handle so muc...
- 02:36 AM Bug #20098 (Resolved): rgw: connection reset or radosgw crashed when download large zero object w...
large zero object has a large compression rate, even 4M compressed data can decompress several GB data. Handle so m...- 12:47 AM Backport #19806 (Resolved): jewel: APIs to support Ragweed suite
- 12:30 AM Feature #20094: RFW: make civetweb max request size configurable to allow larger s3 object metadata
- PR was submitted to upstream civetweb and has been merged: https://github.com/civetweb/civetweb/pull/465
05/26/2017
- 07:29 PM Backport #20078 (In Progress): rgw jewel valgrind failure: make_params
- 02:00 PM Backport #20078: rgw jewel valgrind failure: make_params
- https://github.com/ceph/ceph/pull/15312
- 01:59 PM Backport #20078: rgw jewel valgrind failure: make_params
- Since this issue is already fixed in Kraken & master, it only appears in Jewel, need to copy the issue to backport tr...
- 03:34 PM Bug #20096 (New): rgw: jewel valgrind failure RGWHTTPManager::add_request
- we're allocating a rgw_http_req_data and cleaning up on failures, not exactly for successful cases?...
- 02:34 PM Bug #20064 (Pending Backport): allow empty data_extra_pool in zone placement
- 02:23 PM Feature #20094: RFW: make civetweb max request size configurable to allow larger s3 object metadata
- Initial pull request at https://github.com/ceph/civetweb/pull/20, was requested to submit to upstream civetweb.
- 02:23 PM Feature #20094 (New): RFW: make civetweb max request size configurable to allow larger s3 object ...
- Proposes to turn civetweb's macro for MAX_REQUEST_SIZE=16k into a configuration variable.
- 04:57 AM Bug #20048 (Fix Under Review): Ceph no longer builds for s390x after Beast submodule enabled
- https://github.com/ceph/ceph/pull/15225
05/25/2017
- 10:31 PM Backport #20088 (Resolved): valgrind reports leak in RGWRemoteMetadataCR
- https://github.com/ceph/ceph/pull/15382
- 10:03 PM Bug #20085 (New): valgrind failure on curl_multi_init
- could be a false +ve as well, we do a `curl_mutli_cleanup` in the destructor...
- 08:40 PM Backport #17546: jewel: rgw: bucket resharding
- See #17756 for the backport that was actually merged.
- 08:39 PM Backport #17546 (Rejected): jewel: rgw: bucket resharding
- 07:01 PM Bug #20082: rgw_file: bucket uxattrs not restored correctly (because set on "<bucket>/")
- Further explication: this reproduced as unix attrs set on buckets 'changing back' to their default values after a re...
- 06:59 PM Bug #20082 (In Progress): rgw_file: bucket uxattrs not restored correctly (because set on "<bucke...
- This turns out to be the corollary of the earlier bug in removing buckets w/uxattrs. Rather than conditionally put t...
- 06:45 PM Fix #19732: fix: rgw crashed caused by shard id out of range when listing data log
- *master PR*: https://github.com/ceph/ceph/pull/14389
- 05:55 PM Backport #19736 (Resolved): radosgw/s3 chunked transfer encodings and fast_forward_request
- 05:46 PM Bug #19870: rgw: bytes_send and bytes_recv in the msg of usage show returning is 0 in master branch
- duplicated of http://tracker.ceph.com/issues/19194
- 05:46 PM Bug #19214 (Pending Backport): rgw_file: RGWLibFS::get_inst() not stable
- https://github.com/ceph/ceph/pull/13832
- 05:43 PM Bug #19624 (Pending Backport): rgw_file: handle chunked readdir
- https://github.com/ceph/ceph/pull/14653
https://github.com/ceph/ceph/pull/14458 - 05:36 PM Bug #20045 (Pending Backport): rgw_file: cannot delete bucket w/uxattrs
- https://github.com/ceph/ceph/pull/15210
- 05:34 PM Bug #20061 (Pending Backport): rgw_file: attempt to unlink non-empty bucket could lead to hang
- https://github.com/ceph/ceph/pull/15246
- 05:34 PM Bug #20075 (Fix Under Review): rgw: multipart copy-part encounts 404 error by using S3 java sdk
- 08:34 AM Bug #20075: rgw: multipart copy-part encounts 404 error by using S3 java sdk
- Fixed:https://github.com/ceph/ceph/pull/15283
- 03:55 AM Bug #20075 (Resolved): rgw: multipart copy-part encounts 404 error by using S3 java sdk
- Testing multipart copy-part feature by using S3 java sdk(version:1.10.69) got 404 No Such Key. RGW log detail are as...
- 05:22 PM Bug #20079 (Closed): rgw jewel: valgrind failure in receive_header
- failure when adding the headers to the out_headers map ...
- 03:42 PM Backport #20078 (Resolved): rgw jewel valgrind failure: make_params
- https://github.com/ceph/ceph/pull/15312
- 11:38 AM Backport #19098 (Resolved): RGW/openssl fix for autoconf logic problem in Ubuntu Xenial
- 11:04 AM Backport #19003 (Resolved): jewel: civetweb defaults to libssl.so and libcrypto.so when versions ...
05/24/2017
- 06:32 PM Bug #17632: byte ranges on slo/dlo objects broken in master
- is this still a problem?
- 06:20 PM Bug #18977 (Fix Under Review): rgw: end_marker parameter doesn't work on Swift container's listing
- https://github.com/ceph/ceph/pull/15273
- 04:21 PM Backport #17546 (Resolved): jewel: rgw: bucket resharding
- Fixed in https://github.com/ceph/ceph/pull/13341
- 04:06 PM Bug #16968 (Closed): synchonized bucket can't be used in other zonegroup
- 03:39 PM Bug #20073 (New): valgrind failures jewel: possible memory corruption in StreamRWReq::get_resource
- seen at http://pulpito.ceph.com/abhi-2017-05-23_12:16:16-rgw-wip-jewel-backports-mon-sortbitwise-distro-basic-smithi/...
- 12:57 PM Bug #20071 (Resolved): wrong object size after copy of uncompressed multipart objects
- when compression is disabled, the object's accounted_size is set to the head object's size, rather than the size from...
- 08:27 AM Bug #16888: [multisite] operating bucket's acl&cors is not restricted on slave zone
- cors should be forward to master
https://github.com/ceph/ceph/pull/15260 - 03:14 AM Bug #20066 (New): rgw_admin:change the owner of a bucket, and the stats don't decline from the ol...
- when the bucket is unlinked, the stats don't decline from the old user.
05/23/2017
- 10:18 PM Bug #19956: radosgw Segmentation fault when service is restarted during lifecycle processing
- ...
- 09:18 PM Bug #19956: radosgw Segmentation fault when service is restarted during lifecycle processing
- I got a coredump + symbols for this. let me know if you want the dump, here is the stack with full symbols:
#0 0x... - 10:05 PM Bug #20047: rgw_file: unlink assert unsafe in multi-gateway environments
- PR to master: https://github.com/ceph/ceph/pull/15216
- 06:30 PM Bug #20047 (Pending Backport): rgw_file: unlink assert unsafe in multi-gateway environments
- 08:45 PM Bug #20064 (Resolved): allow empty data_extra_pool in zone placement
- RGWZoneParams::get_head_data_pool() and rgw_get_obj_data_pool() both use placement.data_extra_pool directly instead o...
- 07:08 PM Bug #20061 (Resolved): rgw_file: attempt to unlink non-empty bucket could lead to hang
- An early return in new unlink bucket num_entries check is
missing a conditional unlock and unref. - 05:55 PM Fix #19732 (Pending Backport): fix: rgw crashed caused by shard id out of range when listing data...
- 05:54 PM Bug #19978 (Pending Backport): rgw: 'gc list --include-all' command infinite loop the first 1000 ...
05/22/2017
- 11:23 PM Bug #20048 (Resolved): Ceph no longer builds for s390x after Beast submodule enabled
- https://github.com/ceph/ceph/pull/14273 introduces a regression in that it breaks the s390x build.
The immediate c... - 09:43 PM Bug #20047 (Resolved): rgw_file: unlink assert unsafe in multi-gateway environments
- The assert that unlinked objects are really unlinked is not safe for production use, since the operation is only loca...
- 06:51 PM Bug #20045 (Resolved): rgw_file: cannot delete bucket w/uxattrs
- If a setattr (e.g., chown) has been performed on a bucket, then like any directory it has a hidden object storing its...
- 06:08 PM Bug #20044 (New): Creating a zone does not add the realm ID to the zone's config
- When you create a zone, the output shows the realm id (which,
I have to assume, is supposed to be stored in the zone... - 05:10 PM Bug #19961: luminous: rgw: leaked _shadow_ zero size tail objects if whole object size is multipl...
- Fixed in v12.0.3
- 02:55 PM Feature #19917: radosgw access log is lacking useful information
- Pushed some initial code for the other information items at https://github.com/cloudbau/ceph/tree/wip-rgw-logging, he...
- 02:35 PM Bug #20037 (Resolved): rgw log includes zero byte sometimes
- The log line looks like this:...
- 11:38 AM Bug #16293: Multipart objects are broken when using S3 and SWIFT API at the same time
- I tested it also on:
radosgw --version
ceph version 10.2.6 (656b5b63ed7c43bd014bcafd81b001959d5f089f)
and we hav... - 10:37 AM Backport #20007 (In Progress): jewel: rgw: remove unnecessary 'error in read_id for object name: ...
- 10:28 AM Backport #20007 (Resolved): jewel: rgw: remove unnecessary 'error in read_id for object name: def...
- https://github.com/ceph/ceph/pull/15197
- 10:29 AM Backport #20031 (Resolved): kraken: rgw: Swift's at-root features (/crossdomain.xml, /info, /heal...
- https://github.com/ceph/ceph/pull/16168
- 10:29 AM Backport #20030 (Rejected): kraken: multisite: radosgw-admin period commands cant use --remote in...
- 10:29 AM Backport #20029 (Rejected): jewel: multisite: radosgw-admin period commands cant use --remote in ...
- https://github.com/ceph/ceph/pull/15458
- 10:28 AM Backport #20021 (Rejected): kraken: multisite: RGWPeriodPusher CRThread startup races with RGWHTT...
- 10:28 AM Backport #20020 (Resolved): jewel: multisite: RGWPeriodPusher CRThread startup races with RGWHTTP...
- https://github.com/ceph/ceph/pull/15457
- 10:28 AM Backport #20019 (Rejected): kraken: multisite: memory leak on failed lease in RGWDataSyncShardCR
- 10:28 AM Backport #20018 (Resolved): jewel: multisite: memory leak on failed lease in RGWDataSyncShardCR
- https://github.com/ceph/ceph/pull/15457
- 10:28 AM Backport #20015 (Resolved): kraken: multisite: bi_list() decode failures
- https://github.com/ceph/ceph/pull/15384
- 10:28 AM Backport #20014 (Resolved): jewel: multisite: bi_list() decode failures
- https://github.com/ceph/ceph/pull/15383
- 10:28 AM Backport #20004 (Rejected): kraken: rgw_file: v3 write timer does not close open handles
- 10:28 AM Backport #20003 (Resolved): jewel: rgw_file: v3 write timer does not close open handles
- https://github.com/ceph/ceph/pull/15456
05/19/2017
- 02:49 PM Bug #19988 (Resolved): RGW: can't stack compression and encryption filters
- New compression and encryption filters do not stack, but logically could do so. Users must choose between at rest da...
- 02:32 PM Feature #19917: radosgw access log is lacking useful information
- This https://github.com/civetweb/civetweb/commit/02992037dde3433285790f03ab632d4a8625d3b9
- 02:31 PM Feature #19917: radosgw access log is lacking useful information
- This patch in civetweb upstream seems to be fixing issue #1.
- 07:45 AM Bug #19894: Write fails on larger files on nfs-ganesha-rgw
- Matt Benjamin wrote:
> Hi Stein,
>
> I did a by-hand verify of current v10.2.7 stable and upstream nfs-ganesha v2... - 12:41 AM Bug #19978 (Resolved): rgw: 'gc list --include-all' command infinite loop the first 1000 items
- https://github.com/ceph/ceph/pull/12774
05/18/2017
- 06:18 PM Bug #17465 (Resolved): multisite: coroutine deadlock in RGWMetaSyncCR after ECANCELED errors
- both backports are resolved
- 07:30 AM Bug #16098: client.radosgw.* vs. client.rgw.*; incompabilities and confusion in code and document...
- I was using the Ubuntu LTS version pre-systemd at that time. So if the init scripts are not maintained, I guess the p...
- 07:26 AM Bug #16098: client.radosgw.* vs. client.rgw.*; incompabilities and confusion in code and document...
- With the advent of systemd, the sysvinit scripts are no longer maintained. Which operating system+version are you using?
05/17/2017
- 04:00 PM Bug #16650 (Resolved): qa: rgw_roundtrip fail
- The PR merged so I assume this is done?
- 03:14 PM Bug #19965 (New): RGW handling of combined If-None-Match and If-Modified-Since request is not RFC...
- If you look at the rfc2616 (section 14.26), my interpretation is that is that the result should be as follow :
<pr... - 11:37 AM Bug #19961 (New): luminous: rgw: leaked _shadow_ zero size tail objects if whole object size is m...
If s3 object size is multible of rgw_max_chunk_size
then after deletion and after gc completes
last _shadow_ obj...- 11:17 AM Bug #19959 (Resolved): luminous: rgw: s3 user stats not updated after radosgw-admin bucket rm --p...
- luminous branch, v12.0.2
After deletion of bucket with radosgw-admin with --purge-objects (with deletion of objects)... - 04:47 AM Bug #19956: radosgw Segmentation fault when service is restarted during lifecycle processing
- Yikes, same fault also happened when I pushed a new period map, which caused the 'realm reloader' to trigger. That ma...
- 04:25 AM Bug #19956 (Resolved): radosgw Segmentation fault when service is restarted during lifecycle proc...
- ceph version 11.2.0 (f223e27eeb35991352ebc1f67423d4ebc252adb7),
1. Trigger a lifecycle gc run
2. Restart the ser... - 04:31 AM Bug #19906: Random Segmentation fault thread_name:civetweb-worker
- I suspect this is the same as my crash: http://tracker.ceph.com/issues/19704 - though your offsets are different bec...
- 03:36 AM Documentation #19953 (Duplicate): multisite.rst: update pool names in Pools section and describe ...
- 02:47 AM Documentation #19953 (Duplicate): multisite.rst: update pool names in Pools section and describe ...
- 02:31 AM Bug #19951 (Resolved): rgw:SLO download response header "X-Object-Meta-Static-Large-Object" has 4...
- https://github.com/ceph/ceph/pull/15045
05/16/2017
- 04:41 PM Bug #19920: test_cls_rgw.sh FAILED ClsSDK.TestSDKCoverageReplay (2035 ms)
- also in http://qa-proxy.ceph.com/teuthology/teuthology-2017-05-15_05:00:21-smoke-master-testing-basic-vps/1178707/teu...
- 02:53 PM Bug #19932 (Pending Backport): rgw_file: v3 write timer does not close open handles
05/15/2017
- 11:06 PM Bug #19932 (Fix Under Review): rgw_file: v3 write timer does not close open handles
- https://github.com/ceph/ceph/pull/15097
- 09:37 PM Bug #19932 (Resolved): rgw_file: v3 write timer does not close open handles
- When a v3 write timer expires, it completes the current write as required, but does not clear the FLAG_OPEN or FLAG_S...
- 08:58 AM Bug #19927 (New): rgw-multisite: slave zone can not redirect website configure request to metadat...
hi,
when i make website configure request to slave zone rgw, and get website configure from master zone,but i ca...- 06:39 AM Tasks #16784 (Fix Under Review): RADOSGW feature requires FCGI which no longer has an upstream so...
Also available in: Atom