| *** rcernin has quit IRC | 00:23 | |
| *** rcernin has joined #openstack-swift | 00:23 | |
| *** rcernin has quit IRC | 01:04 | |
| *** rcernin has joined #openstack-swift | 01:05 | |
| *** rcernin has quit IRC | 01:05 | |
| *** edausq has quit IRC | 01:48 | |
| *** rcernin has joined #openstack-swift | 03:19 | |
| *** psachin has joined #openstack-swift | 03:31 | |
| *** rcernin has quit IRC | 04:20 | |
| *** rcernin has joined #openstack-swift | 04:22 | |
| *** evrardjp has quit IRC | 04:33 | |
| *** evrardjp has joined #openstack-swift | 04:33 | |
| *** rcernin has quit IRC | 04:38 | |
| *** threestrands has joined #openstack-swift | 04:40 | |
| *** rcernin has joined #openstack-swift | 04:43 | |
| *** adriant has joined #openstack-swift | 06:07 | |
| adriant | I've run into a bug around the handling of the 'info' api with the Swiftclient (and the OpenStackSDK) when talking to RGW, and I'm not really sure how I should be handling this bug... | 06:10 |
|---|---|---|
| adriant | https://docs.ceph.com/docs/master/radosgw/keystone/ < mentions deploying RGW and serving it in the keystone catalog at: http://radosgw.example.com:8080/swift/v1 | 06:11 |
| adriant | which means the info API ends up at: http://radosgw.example.com:8080/swift/info | 06:11 |
| adriant | but the swiftclient AND the sdk both create the url to that as: http://radosgw.example.com:8080/info | 06:11 |
| adriant | which for a real Swift url (without the 'swift' part) would work fine, but doesn't for RGW | 06:12 |
| adriant | Is this the kind of thing we should be fixing in the Swiftclient/SDK somehow, or is this just RGW nonsense? | 06:15 |
| adriant | timburke, not sure how to solve this nicely, or where the 'bug' really lands :( | 06:19 |
| *** mvkr has quit IRC | 06:21 | |
| mattoliverau | adriant: this is RGW configuration. We don't put a /swift before the /v1 but ceph does by default, but you can turn it off with: | 06:31 |
| mattoliverau | rgw swift url prefix | 06:31 |
| mattoliverau | and set it to nothing or '' | 06:31 |
| mattoliverau | or rather '/' | 06:32 |
| mattoliverau | but if you do that, it can effect RGWs s3api. | 06:32 |
| mattoliverau | adriant: but yeah it has nothing to do with the Swift api, it's just ceph rgw nonsense. | 06:33 |
| *** mvkr has joined #openstack-swift | 06:33 | |
| *** mvkr has quit IRC | 07:19 | |
| *** mvkr has joined #openstack-swift | 07:32 | |
| *** rcernin has quit IRC | 07:36 | |
| *** openstackstatus has quit IRC | 07:39 | |
| *** openstackstatus has joined #openstack-swift | 07:40 | |
| *** ChanServ sets mode: +v openstackstatus | 07:40 | |
| *** rpittau|afk is now known as rpittau | 07:46 | |
| *** ccamacho has joined #openstack-swift | 07:51 | |
| *** mikecmpbll has joined #openstack-swift | 08:05 | |
| *** jv_ has quit IRC | 08:15 | |
| *** ccamacho has quit IRC | 08:23 | |
| *** jv_ has joined #openstack-swift | 08:30 | |
| *** rcernin has joined #openstack-swift | 08:40 | |
| *** rcernin has quit IRC | 08:40 | |
| *** dtantsur|afk is now known as dtantsur | 09:05 | |
| *** tkajinam has quit IRC | 09:05 | |
| *** threestrands has quit IRC | 10:05 | |
| *** rpittau is now known as rpittau|bbl | 10:20 | |
| *** mikecmpbll has quit IRC | 12:07 | |
| *** mikecmpbll has joined #openstack-swift | 12:09 | |
| *** rpittau|bbl is now known as rpittau | 12:22 | |
| *** jv_ has quit IRC | 12:56 | |
| adriant | mattoliverau, thought so, it felt like something we can/should solve at the deployment layer... and yeah I'm not sure how the s3api would work if you did get rid of the prefix :/ | 13:04 |
| adriant | mattoliverau, would/should we add some logic in the swiftclient or openstacksdk to better handle the building of a url for 'info' ? | 13:06 |
| adriant | maybe take the base keystone url, and strip everything after (and including) '/v1' ? | 13:06 |
| adriant | and then add the '/info' part? Which would still work with a Swift prefix, or potentially any deployment that decides to deploy Swift with a prefix. | 13:10 |
| *** jv_ has joined #openstack-swift | 13:19 | |
| *** tkajinam has joined #openstack-swift | 13:37 | |
| *** zaitcev has joined #openstack-swift | 13:38 | |
| *** ChanServ sets mode: +v zaitcev | 13:38 | |
| *** psachin has quit IRC | 13:44 | |
| openstackgerrit | Clay Gerrard proposed openstack/swift master: Test for Versioned Object If-Match https://review.opendev.org/730414 | 13:49 |
| *** dtantsur is now known as dtantsur|brb | 13:53 | |
| *** rpittau is now known as rpittau|brb | 14:10 | |
| *** mikecmpbll has quit IRC | 14:11 | |
| *** mikecmpbll has joined #openstack-swift | 14:15 | |
| *** tkajinam has quit IRC | 14:36 | |
| *** rpittau|brb is now known as rpittau | 14:39 | |
| *** dtantsur|brb is now known as dtantsur | 14:50 | |
| clayg | is there a canonical way to copy a HeaderKeyDict - it's surprising to me that it doesn't support .copy directly since it inherits from dict | 14:56 |
| clayg | oh, wth, a HeaderEnvironProxy isn't a HeaderKeyDict anyway | 14:59 |
| openstackgerrit | Clay Gerrard proposed openstack/swift master: versioning: Have versioning symlinks make pre-auth requests to reserved container https://review.opendev.org/724393 | 15:10 |
| *** ccamacho has joined #openstack-swift | 15:28 | |
| *** dtantsur is now known as dtantsur|afk | 15:34 | |
| *** rpittau is now known as rpittau|afk | 15:51 | |
| *** coreycb has quit IRC | 16:51 | |
| *** zaitcev has quit IRC | 16:51 | |
| *** coreycb has joined #openstack-swift | 16:58 | |
| timburke | clayg, hey! i think https://review.opendev.org/#/c/724393/ is going to fix up a tempurl/object-versioning bug i just noticed, too! | 17:13 |
| patchbot | patch 724393 - swift - versioning: Have versioning symlinks make pre-auth... - 4 patch sets | 17:13 |
| timburke | adriant, sounds like https://bugs.launchpad.net/python-swiftclient/+bug/1712358 -- there's a fix on master (https://github.com/openstack/python-swiftclient/commit/947c09f); i've been meaning to do a swiftclient release... | 17:15 |
| openstack | Launchpad bug 1712358 in python-swiftclient "Capability discovery endpoint detection is hardcoded" [Undecided,Fix released] | 17:15 |
| clayg | Yeah that change’ll be pretty good. | 17:19 |
| timburke | clayg, confirmed that the patch allows you to make usable tempurls for the current version, where master always 401s -- want me to write up another bug for that, or mention it as a comment on https://bugs.launchpad.net/swift/+bug/1880013 ? | 17:40 |
| openstack | Launchpad bug 1880013 in OpenStack Object Storage (swift) "ACLs don't work with x-versions-enable: true" [Medium,In progress] | 17:40 |
| clayg | if it's a client observable bug that's going to exist in a published version of swift it's super helpful to get an upstream bug IMHO | 17:50 |
| clayg | i wasn't aware of the issue | 17:50 |
| clayg | if we have a bug # and know the change that fixes it we should update the commit message if we can | 17:50 |
| timburke | 👍 that was my gut feeling, too | 17:50 |
| clayg | oh... I see now | 17:51 |
| clayg | ok, we could just update the title to something like "alternatate authed" or even just "acls & tempurls" | 17:52 |
| clayg | "ACLs & tempurls don't work with ... " actually seem fine | 17:52 |
| clayg | timburke: it doesn't need to be a separate bug | 17:52 |
| openstackgerrit | Alexandre Lécuyer proposed openstack/swift feature/losf: New key format for objects in the index-server https://review.opendev.org/723609 | 17:58 |
| *** zaitcev has joined #openstack-swift | 18:01 | |
| *** ChanServ sets mode: +v zaitcev | 18:01 | |
| openstackgerrit | Merged openstack/swift master: Test for Versioned Object If-Match https://review.opendev.org/730414 | 18:04 |
| *** irclogbot_0 has quit IRC | 18:26 | |
| timburke | clayg, well, now i'm confused -- we already *have* container tempurl func tests? https://github.com/openstack/swift/blob/master/test/functional/test_object_versioning.py#L2638 | 18:29 |
| timburke | but then, why was i getting 401s?? | 18:29 |
| clayg | uhhh... wtf | 18:30 |
| *** irclogbot_3 has joined #openstack-swift | 18:30 | |
| clayg | well and how was was the container scoped tempurl auth callback granting access to the reserved container 😞 | 18:31 |
| clayg | software is so complicated | 18:31 |
| timburke | nvm, i'm being dumb: https://github.com/openstack/swift/blob/master/test/functional/test_object_versioning.py#L2715 | 18:33 |
| timburke | apparently i fixed up the test expectations when i first submitted the patch, i just forgot | 18:35 |
| clayg | well, so but then that test should fail now? | 18:36 |
| tdasilva | rledisez: can you share a link to that etherpad you had detailing the performance work you were doing? | 18:47 |
| rledisez | tdasilva: I wrote https://etherpad.opendev.org/p/swift-profiling and (probably not the one you're looking for) https://etherpad.opendev.org/p/swift-replacing-hashes.pkl | 18:49 |
| rledisez | I'm pretty sure I had one about MD5 too, but I can't find it | 18:49 |
| timburke | https://etherpad.openstack.org/p/swift-md5-optimisation | 18:54 |
| timburke | fwiw, i made a public calendar for the meeting times for next week: https://calendar.google.com/calendar/ical/sfi876p12h1dg0lnltvrr37rng%40group.calendar.google.com/public/basic.ics | 19:05 |
| openstackgerrit | Tim Burke proposed openstack/swift master: dsvm: Run domain_remap tests with tempauth https://review.opendev.org/730957 | 19:55 |
| openstackgerrit | Tim Burke proposed openstack/swift master: dsvm: Run service-user tests under tempauth https://review.opendev.org/730958 | 19:55 |
| *** irclogbot_3 has quit IRC | 20:22 | |
| *** irclogbot_0 has joined #openstack-swift | 20:25 | |
| *** renich has joined #openstack-swift | 20:27 | |
| openstackgerrit | Merged openstack/swift master: Breakup reclaim into batches https://review.opendev.org/727876 | 20:36 |
| openstackgerrit | Merged openstack/swift master: fix s3api functional tests https://review.opendev.org/571021 | 20:36 |
| *** renich_ has joined #openstack-swift | 20:57 | |
| *** renich has quit IRC | 20:58 | |
| *** renich_ is now known as renich | 20:58 | |
| tdasilva | rledisez, timburke: thanks! I was actually looking for the swift-profiling one | 21:13 |
| *** ccamacho has quit IRC | 21:34 | |
| openstackgerrit | Tim Burke proposed openstack/swift master: swift-get-nodes: Unquote paths before working with them https://review.opendev.org/724141 | 21:57 |
| openstackgerrit | Tim Burke proposed openstack/swift master: Ussuri contrib docs community goal https://review.opendev.org/726992 | 22:06 |
| openstackgerrit | Tim Burke proposed openstack/swift master: prefactor: Pass s3api Config to S3Requests https://review.opendev.org/729091 | 22:39 |
| openstackgerrit | Tim Burke proposed openstack/swift master: s3api: Add config option to return 429s on ratelimit https://review.opendev.org/729093 | 22:39 |
| *** tkajinam has joined #openstack-swift | 22:57 | |
| *** rcernin has joined #openstack-swift | 23:05 | |
| *** threestrands has joined #openstack-swift | 23:17 | |
| mattoliverau | morning | 23:22 |
| openstackgerrit | Tim Burke proposed openstack/swift master: dsvm: Use devstack's s3api "service" https://review.opendev.org/731004 | 23:29 |
| openstackgerrit | Tim Burke proposed openstack/swift master: Skip tests when only changing README.rst, CONTRIBUTING.rst, etc. https://review.opendev.org/731007 | 23:43 |
| openstackgerrit | Tim Burke proposed openstack/swift master: Simplify wsgify() https://review.opendev.org/731008 | 23:46 |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!