| *** mikecmpbll has quit IRC | 00:18 | |
| *** psachin has joined #openstack-swift | 02:46 | |
| *** gkadam has joined #openstack-swift | 03:41 | |
| *** viks__ has joined #openstack-swift | 03:51 | |
| *** hseipp has joined #openstack-swift | 06:43 | |
| *** pcaruana has joined #openstack-swift | 06:49 | |
| *** ccamacho has joined #openstack-swift | 06:50 | |
| *** cbartz has joined #openstack-swift | 06:53 | |
| *** links has joined #openstack-swift | 06:59 | |
| *** rcernin has quit IRC | 07:31 | |
| *** d0ugal has joined #openstack-swift | 07:54 | |
| *** mikecmpbll has joined #openstack-swift | 10:18 | |
| *** pcaruana has quit IRC | 10:32 | |
| *** pcaruana has joined #openstack-swift | 10:32 | |
| *** hseipp has quit IRC | 11:03 | |
| *** threestrands has quit IRC | 11:05 | |
| *** gkadam has quit IRC | 11:23 | |
| *** gkadam has joined #openstack-swift | 11:23 | |
| *** hseipp has joined #openstack-swift | 12:03 | |
| tdasilva | good morning | 12:15 |
|---|---|---|
| *** gkadam has quit IRC | 12:51 | |
| *** links has quit IRC | 12:57 | |
| *** cbartz has quit IRC | 13:00 | |
| *** cbartz has joined #openstack-swift | 13:17 | |
| *** vinsh has joined #openstack-swift | 13:46 | |
| *** psachin has quit IRC | 13:50 | |
| *** vinsh_ has quit IRC | 13:50 | |
| *** vinsh has quit IRC | 14:16 | |
| *** ianychoi has quit IRC | 14:36 | |
| *** ianychoi has joined #openstack-swift | 14:42 | |
| *** pcaruana has quit IRC | 15:29 | |
| notmyname | good morning | 15:42 |
| *** ccamacho is now known as ccamacho|brb | 16:10 | |
| *** cbartz has quit IRC | 16:13 | |
| *** hseipp has quit IRC | 16:26 | |
| *** pcaruana has joined #openstack-swift | 16:26 | |
| *** itlinux has joined #openstack-swift | 16:29 | |
| *** jlviva-viva is now known as jlvillal | 16:40 | |
| openstackgerrit | Tim Burke proposed openstack/swift master: Move legacy-swift-dsvm-functional job in-tree https://review.openstack.org/589270 | 16:57 |
| openstackgerrit | Tim Burke proposed openstack/swift master: Add IPv6 DSVM tests https://review.openstack.org/591497 | 16:57 |
| *** openstackgerrit has quit IRC | 17:04 | |
| *** gyee has joined #openstack-swift | 18:27 | |
| *** gyee has quit IRC | 18:28 | |
| *** gyee has joined #openstack-swift | 18:28 | |
| *** openstackgerrit has joined #openstack-swift | 18:32 | |
| openstackgerrit | Merged openstack/swift master: import zuul job settings from project-config https://review.openstack.org/596210 | 18:32 |
| *** itlinux has quit IRC | 18:54 | |
| *** itlinux has joined #openstack-swift | 19:11 | |
| *** pcaruana has quit IRC | 19:13 | |
| *** itlinux has quit IRC | 19:16 | |
| openstackgerrit | Merged openstack/swift master: minor updates to changelog and release notes https://review.openstack.org/595260 | 19:40 |
| *** geaaru has joined #openstack-swift | 21:33 | |
| samueldmq | hello! | 21:36 |
| samueldmq | I'm talking to swift and getting 200 OK and 'NoSuchKey' as body on an object that does not exist | 21:37 |
| samueldmq | is that a known behavior or is the cloud I'm talking to really broken? | 21:37 |
| samueldmq | #link http://paste.openstack.org/show/728877/ example using openstacksdk | 21:37 |
| DHE | it should behave properly with a 404 response | 21:37 |
| samueldmq | exactly, not sure why I'm hitting a 200 in that case | 21:38 |
| *** rcernin has joined #openstack-swift | 21:51 | |
| notmyname | samueldmq: NoSuchKey is an S3 api error response | 22:18 |
| notmyname | so it sounds like you're talking to an endpoint with the s3 api compat layer enabled | 22:18 |
| samueldmq | notmyname: kk but what about the 200 OK, is that expected within the s3 api compat layer? | 22:29 |
| * samueldmq looks for docs on that | 22:29 | |
| notmyname | on the container write? | 22:34 |
| samueldmq | notmyname: GET on an object that does not exist | 22:36 |
| *** zaitcev has joined #openstack-swift | 22:36 | |
| *** ChanServ sets mode: +v zaitcev | 22:36 | |
| samueldmq | notmyname: http://paste.openstack.org/show/728877/ | 22:36 |
| notmyname | it would be much easier to debug if you could show the actual API calls being made, eg using curl instead of an SDK that's wrapping the calls | 22:37 |
| notmyname | eg i don't actually see a 200 response code in that paste | 22:37 |
| samueldmq | notmyname: you're very right, I saw that by debugging myself with sdk. I will redo with curl | 22:40 |
| samueldmq | notmyname: there it is | 22:51 |
| samueldmq | #link http://paste.openstack.org/show/728883/ | 22:51 |
| notmyname | samueldmq: that does indeed look odd | 22:54 |
| notmyname | the x-mesh-debug-req-id header hints that there may be something in between the client and the cluster? | 22:54 |
| notmyname | and are you expecting to use the s3 api? | 22:55 |
| samueldmq | notmyname: I don't know what my provider is doing. tbh I just wanted to use the swift apis :-) | 22:56 |
| notmyname | ah ok | 22:56 |
| samueldmq | I am writing some functional tests to validate -sdk against a lot of clouds ... and just stumbled on this | 22:56 |
| samueldmq | perhaps I can ask my provider what the they're doing | 22:57 |
| notmyname | then confirm that the storage url you're getting back from auth is ok. maybe they're returning separate swift and s3 endpoints? | 22:57 |
| notmyname | the "/swift/" in the URL is also somewhat unusual in my experience | 22:58 |
| notmyname | assuming the endpoint from auth is ok, then everything I've seen in the pastebin look fine. you'll need to take it up iwth your provider | 22:58 |
| notmyname | I'd be able to help more if you had control over the cluster | 22:58 |
| *** mikecmpbll has quit IRC | 23:00 | |
| samueldmq | notmyname: jsut checked the service catalog and yes, that's the url I get back from it | 23:03 |
| samueldmq | that's the only public endpoint for swift I can see | 23:04 |
| mattoliverau | Most radosgw deploys I've seen tend to have a /swift in the url. So you might not actually be hitting something that isn't using a 100% Swift API. Can you run an /info on it. | 23:04 |
| samueldmq | notmyname: I will take it up with them, that's a huge help already, appreciated it! | 23:04 |
| mattoliverau | lol, double negative, lets try again, you might be hitting something that isn't 100% compliant with the Swift API | 23:05 |
| notmyname | mattoliverau: oh, interesting | 23:05 |
| mattoliverau | ie. they might be running a Ceph Radosgw end point. | 23:05 |
| samueldmq | hmmm | 23:05 |
| samueldmq | let me see a /info | 23:06 |
| mattoliverau | what does a swift stat look like? | 23:06 |
| mattoliverau | what account name comes back, in most radogw stats on the account I do doesn't give me the account name in the output. though maybe that's just a mis configuration on my end (I'm no a ceph expert). But it's used here at work dev cloud builds sometimes and it confused the hell out of me | 23:07 |
| samueldmq | #link http://paste.openstack.org/show/728885/ | 23:09 |
| samueldmq | for the /info request | 23:09 |
| notmyname | yeah, that's suspicious: "version":"12.2.6" | 23:10 |
| notmyname | swift just released 2.19.0 | 23:10 |
| samueldmq | oh they're far far ahead :-) | 23:10 |
| * notmyname starts to wonder what's in the magical swift v12 | 23:10 | |
| notmyname | it's the future!! | 23:11 |
| samueldmq | hehe this is getting interesting. I'll poke the apis and see what we're gonna implement next | 23:11 |
| notmyname | lol | 23:12 |
| mattoliverau | lol | 23:12 |
| samueldmq | I can't run the swift stat thing it's a public cloud | 23:15 |
| mattoliverau | 12.2.6 looks like it might be Ceph Luminous (well I quick google search). So yeah, radosgw's swift api endpoint. | 23:16 |
| samueldmq | 12.2.6 is clearly a ceph version though. thye might be doing the radosgw endpoint thing you mentioned. even though I don't know what kind of sourcery is that | 23:16 |
| samueldmq | but seems like puttin swift API on top of base ceph? | 23:16 |
| mattoliverau | they're just have their own implementation of the Swift API. | 23:16 |
| notmyname | samueldmq: yeah, they do that. at least some of it | 23:16 |
| samueldmq | which returns 200 on GET /OBJECT_THAT_DOES_NOT_EXIST ! awesome | 23:17 |
| mattoliverau | lol, yeah nice :P | 23:17 |
| samueldmq | notmyname: mattoliverau: awesome, I appreciate your time helping to debug. this was interesting | 23:19 |
| notmyname | np | 23:19 |
| mattoliverau | samueldmq: nps :) | 23:19 |
| samueldmq | #protip TIL when you implement an API, keep the all the contracts consistent. | 23:20 |
| mattoliverau | +100 | 23:21 |
| mattoliverau | oh and good morning swift channel :) | 23:22 |
| notmyname | oh hi mattoliverau :-) | 23:22 |
| timburke | fwiw, the json response is also very strange to me -- swift3/s3api does xml... | 23:25 |
| *** Jeffrey4l has quit IRC | 23:26 | |
| *** Jeffrey4l has joined #openstack-swift | 23:28 | |
| *** kei-ichi has joined #openstack-swift | 23:30 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!