Unable to delete bucket - endless multipart uploads?
Hi All, We've been dealing with what seems to be a pretty annoying bug for a while now. We are unable to delete a customer's bucket that seems to have an extremely large number of aborted multipart uploads. I've had $(radosgw-admin bucket rm --bucket=pusulax --purge-objects) running in a screen session for almost 3 weeks now and it's still not finished; it's most likely stuck in a loop or something. The screen session with debug-rgw=10 spams billions of these messages: 2021-02-23 15:38:58.667 7f9b55704840 10 RGWRados::cls_bucket_list_unordered: got _multipart_04/d3/04d33e18-3f13-433c-b924-56602d702d60-31.msg.2~0DTalUjTHsnIiKraN1klwIFO88Vc2E3.meta[] 2021-02-23 15:38:58.667 7f9b55704840 10 RGWRados::cls_bucket_list_unordered: got _multipart_04/d7/04d7ad26-c8ec-4a39-9938-329acd6d9da7-102.msg.2~K_gAeTpfEongNvaOMNa0IFwSGPpQ1iA.meta[] 2021-02-23 15:38:58.667 7f9b55704840 10 RGWRados::cls_bucket_list_unordered: got _multipart_04/da/04da4147-c949-4c3a-aca6-e63298f5ff62-102.msg.2~-hXBSFcjQKbMkiyEqSgLaXMm75qFzEp.meta[] 2021-02-23 15:38:58.667 7f9b55704840 10 RGWRados::cls_bucket_list_unordered: got _multipart_04/db/04dbb0e6-dfb0-42fb-9d0f-49cceb18457f-102.msg.2~B5EhGgBU5U_U7EA5r8IhVpO3Aj2OvKg.meta[] 2021-02-23 15:38:58.667 7f9b55704840 10 RGWRados::cls_bucket_list_unordered: got _multipart_04/df/04df39be-06ab-4c72-bc63-3fac1d2700a9-11.msg.2~_8h5fWlkNrIMqcrZgNbAoJfc8BN1Xx-.meta[] This is probably the 2nd or 3rd time I've been unable to delete this bucket. I also tried running $(radosgw-admin bucket check --fix --check-objects --bucket=pusulax) before kicking off the delete job, but that didn't work either. Here is the bucket in question, the num_objects counter never decreases after trying to delete the bucket: [root@os5 ~]# radosgw-admin bucket stats --bucket=pusulax { "bucket": "pusulax", "num_shards": 144, "tenant": "", "zonegroup": "dbb69c5b-b33f-4af2-950c-173d695a4d2c", "placement_rule": "default-placement", "explicit_placement": { "data_pool": "", "data_extra_pool": "", "index_pool": "" }, "id": "c0d0b8a5-c63c-4c24-9dab-8deee88dbf0b.3209338.4", "marker": "c0d0b8a5-c63c-4c24-9dab-8deee88dbf0b.3292800.7", "index_type": "Normal", "owner": "REDACTED", "ver": "0#115613,1#115196,2#115884,3#115497,4#114649,5#114150,6#116127,7#114269,8#115220,9#115092,10#114003,11#114538,12#115235,13#113463,14#114928,15#115135,16#115535,17#114867,18#116010,19#115766,20#115274,21#114818,22#114805,23#114853,24#114099,25#114359,26#114966,27#115790,28#114572,29#114826,30#114767,31#115614,32#113995,33#115305,34#114227,35#114342,36#114144,37#114704,38#114088,39#114738,40#114133,41#114520,42#114420,43#114168,44#113820,45#115093,46#114788,47#115522,48#114713,49#115315,50#115055,51#114513,52#114086,53#114401,54#114079,55#113649,56#114089,57#114157,58#114064,59#115224,60#114753,61#114686,62#115169,63#114321,64#114949,65#115075,66#115003,67#114993,68#115320,69#114392,70#114893,71#114219,72#114190,73#114868,74#113432,75#114882,76#115300,77#114755,78#114598,79#114221,80#114895,81#114031,82#114566,83#113849,84#115155,85#113790,86#113334,87#113800,88#114856,89#114841,90#115073,91#113849,92#114554,93#114820,94#114256,95#113840,96#114838,97#113784,98#114876,99#115524,100#115686,101#112969,102#112156,103#112635,104#112732,105#112933,106#112412,107#113090,108#112239,109#112697,110#113444,111#111730,112#112446,113#114479,114#113318,115#113032,116#112048,117#112404,118#114545,119#112563,120#112341,121#112518,122#111719,123#112273,124#112014,125#112979,126#112209,127#112830,128#113186,129#112944,130#111991,131#112865,132#112688,133#113819,134#112586,135#113275,136#112172,137#113019,138#112872,139#113130,140#112716,141#112091,142#111859,143#112773", "master_ver": "0#0,1#0,2#0,3#0,4#0,5#0,6#0,7#0,8#0,9#0,10#0,11#0,12#0,13#0,14#0,15#0,16#0,17#0,18#0,19#0,20#0,21#0,22#0,23#0,24#0,25#0,26#0,27#0,28#0,29#0,30#0,31#0,32#0,33#0,34#0,35#0,36#0,37#0,38#0,39#0,40#0,41#0,42#0,43#0,44#0,45#0,46#0,47#0,48#0,49#0,50#0,51#0,52#0,53#0,54#0,55#0,56#0,57#0,58#0,59#0,60#0,61#0,62#0,63#0,64#0,65#0,66#0,67#0,68#0,69#0,70#0,71#0,72#0,73#0,74#0,75#0,76#0,77#0,78#0,79#0,80#0,81#0,82#0,83#0,84#0,85#0,86#0,87#0,88#0,89#0,90#0,91#0,92#0,93#0,94#0,95#0,96#0,97#0,98#0,99#0,100#0,101#0,102#0,103#0,104#0,105#0,106#0,107#0,108#0,109#0,110#0,111#0,112#0,113#0,114#0,115#0,116#0,117#0,118#0,119#0,120#0,121#0,122#0,123#0,124#0,125#0,126#0,127#0,128#0,129#0,130#0,131#0,132#0,133#0,134#0,135#0,136#0,137#0,138#0,139#0,140#0,141#0,142#0,143#0", "mtime": "2020-06-17 20:27:16.685833Z", "max_marker": "0#,1#,2#,3#,4#,5#,6#,7#,8#,9#,10#,11#,12#,13#,14#,15#,16#,17#,18#,19#,20#,21#,22#,23#,24#,25#,26#,27#,28#,29#,30#,31#,32#,33#,34#,35#,36#,37#,38#,39#,40#,41#,42#,43#,44#,45#,46#,47#,48#,49#,50#,51#,52#,53#,54#,55#,56#,57#,58#,59#,60#,61#,62#,63#,64#,65#,66#,67#,68#,69#,70#,71#,72#,73#,74#,75#,76#,77#,78#,79#,80#,81#,82#,83#,84#,85#,86#,87#,88#,89#,90#,91#,92#,93#,94#,95#,96#,97#,98#,99#,100#,101#,102#,103#,104#,105#,106#,107#,108#,109#,110#,111#,112#,113#,114#,115#,116#,117#,118#,119#,120#,121#,122#,123#,124#,125#,126#,127#,128#,129#,130#,131#,132#,133#,134#,135#,136#,137#,138#,139#,140#,141#,142#,143#", "usage": { "rgw.multimeta": { "size": 0, "size_actual": 0, "size_utilized": 97009704, "size_kb": 0, "size_kb_actual": 0, "size_kb_utilized": 94737, "num_objects": 3592952 } }, "bucket_quota": { "enabled": false, "check_on_raw": false, "max_size": -1, "max_size_kb": 0, "max_objects": -1 } } We're running 14.2.16 on our RGWs and OSD nodes. Anyone have any ideas? Is it possible to target this bucket via rados directly to try and delete it? I'm weary of doing stuff like that though. Thanks in advance. - Dave Monschein
I wonder if this will be resolved by: https://github.com/ceph/ceph/pull/39358 <https://github.com/ceph/ceph/pull/39358> Deleting a bucket invokes an unordered listing, so that the objects in the bucket can be removed. There was a bug that caused this to loop back the objects. Eric
On Feb 23, 2021, at 3:56 PM, David Monschein <monschein@gmail.com> wrote:
Hi All,
We've been dealing with what seems to be a pretty annoying bug for a while now. We are unable to delete a customer's bucket that seems to have an extremely large number of aborted multipart uploads. I've had $(radosgw-admin bucket rm --bucket=pusulax --purge-objects) running in a screen session for almost 3 weeks now and it's still not finished; it's most likely stuck in a loop or something. The screen session with debug-rgw=10 spams billions of these messages:
...
We're running 14.2.16 on our RGWs and OSD nodes. Anyone have any ideas? Is it possible to target this bucket via rados directly to try and delete it? I'm weary of doing stuff like that though. Thanks in advance.
- Dave Monschein
Hi Dave, On Tuesday, February 23rd, 2021 at 21:56, David Monschein <monschein@gmail.com> wrote:
We're running 14.2.16 on our RGWs and OSD nodes. Anyone have any ideas? Is it possible to target this bucket via rados directly to try and delete it?
I'm weary of doing stuff like that though. Thanks in advance.
I just had the exact same issue. After trying to delete the bucket with `radosgw-admin bucket rm --purge-objects --bucket=tenant/cursed-bucket` and making absolutely no progress over the weekend, I figured I would try and abort all those multipart uploads first. I used `s3cmd multipart s3://cursed-bucket` to list the multipart uploads, and `s3cmd abortmp s3://cursed-bucket/upload/object <upload-id>` to abort each of them. I had to run that in a loop because `s3cmd multipart s3://cursed-bucket` kept returning more multipart uploads (because of pagination, I assume); after a while, I was left with only a dozen multipart uploads that seemingly couldn't be removed that way, so I stopped the loop. At that point, I ran `radosgw-admin bucket rm --purge-objects --bucket=tenant/cursed-bucket` again, and this time it started spitting out ``` 2021-03-05 11:02:10.254 7f881dbab42d 0 abort_bucket_multiparts WARNING : aborted 680 incomplete multipart uploads 2021-03-05 11:02:57.354 7f881dbab42d 0 abort_bucket_multiparts WARNING : aborted 1360 incomplete multipart uploads 2021-03-05 11:03:35.137 7f881dbab42d 0 abort_bucket_multiparts WARNING : aborted 1860 incomplete multipart uploads [...] ``` and I could see the object count decreasing in `radosgw-admin bucket stats --bucket=tenant/cursed-bucket`. After a few hours, the bucket was finally deleted! If you haven't yet managed to delete your bucket, give this a try and let me know if it worked for you too. And if you did figure out another way to delete your bucket in the meantime, I'd be interested to know how you did it as well. Cheers, -- Ben
participants (3)
-
Benoît Knecht
-
David Monschein
-
J. Eric Ivancich