| *** littleidea has quit IRC | 00:01 | |
| *** bencherian has joined #openstack-dev | 00:02 | |
| *** littleidea has joined #openstack-dev | 00:05 | |
| *** e1mer has joined #openstack-dev | 00:15 | |
| *** e1mer has quit IRC | 00:27 | |
| *** jog0 has joined #openstack-dev | 00:32 | |
| *** jog0 has quit IRC | 00:34 | |
| *** asalkeld has quit IRC | 00:36 | |
| *** asalkeld has joined #openstack-dev | 00:36 | |
| *** b1rkh0ff has quit IRC | 00:43 | |
| *** winston-d has joined #openstack-dev | 01:01 | |
| *** dtroyer is now known as dtroyer_zzz | 01:04 | |
| *** winston-d has quit IRC | 01:09 | |
| *** jog0 has joined #openstack-dev | 01:14 | |
| *** jog0 has quit IRC | 01:15 | |
| *** e1mer has joined #openstack-dev | 01:20 | |
| *** dtroyer_zzz is now known as dtroyer | 01:25 | |
| *** dtroyer is now known as dtroyer_zzz | 01:35 | |
| *** andrewsben is now known as andrewsben_zz | 01:48 | |
| *** hugokuo has joined #openstack-dev | 01:54 | |
| *** littleidea has quit IRC | 02:05 | |
| *** littleidea has joined #openstack-dev | 02:06 | |
| *** e1mer has quit IRC | 02:19 | |
| *** shang has joined #openstack-dev | 02:24 | |
| *** e1mer has joined #openstack-dev | 02:27 | |
| *** shang has quit IRC | 02:34 | |
| *** e1mer has quit IRC | 02:34 | |
| *** dtroyer_zzz is now known as dtroyer | 02:35 | |
| *** timello has joined #openstack-dev | 02:36 | |
| *** shang has joined #openstack-dev | 02:36 | |
| *** asalkeld has quit IRC | 02:42 | |
| *** timello has quit IRC | 02:45 | |
| *** asalkeld has joined #openstack-dev | 02:46 | |
| *** pixelbeat has quit IRC | 02:50 | |
| *** andrewsben_zz is now known as andrewsben | 02:54 | |
| *** dtroyer is now known as dtroyer_zzz | 03:04 | |
| notmyname | mtaylor: jeblair: can you check on jenkins for swift? seems I can't trigger a build with my +2/+1 on https://review.openstack.org/#/c/7838/ | 03:08 |
|---|---|---|
| *** mikal has quit IRC | 03:11 | |
| *** mikal has joined #openstack-dev | 03:13 | |
| notmyname | mtaylor: jeblair: FWIW "Build step 'Publish Cobertura Coverage Report' changed build result to FAILURE" | 03:19 |
| *** mikal has quit IRC | 03:26 | |
| *** mikal has joined #openstack-dev | 03:27 | |
| *** eglynn_ has joined #openstack-dev | 03:32 | |
| *** eglynn has quit IRC | 03:36 | |
| *** e1mer has joined #openstack-dev | 03:36 | |
| *** e1mer has joined #openstack-dev | 03:36 | |
| mtaylor | notmyname: bother. will look in to it. | 03:49 |
| mtaylor | LinuxJedi: ^^^^ | 03:49 |
| mtaylor | LinuxJedi: when you wake up | 03:49 |
| *** shang has quit IRC | 03:49 | |
| mtaylor | notmyname: I'm going to blame LinuxJedi ... but it's probably still my fault | 03:49 |
| *** e1mer has quit IRC | 03:50 | |
| *** shang has joined #openstack-dev | 04:02 | |
| *** yamahata_ has quit IRC | 04:06 | |
| *** shang has quit IRC | 04:07 | |
| *** dtroyer_zzz is now known as dtroyer | 04:09 | |
| *** shang has joined #openstack-dev | 04:20 | |
| *** aa__ has quit IRC | 04:30 | |
| *** aa__ has joined #openstack-dev | 04:30 | |
| *** Adri2000 has quit IRC | 04:50 | |
| *** Adri2000 has joined #openstack-dev | 04:50 | |
| *** Adri2000 is now known as Guest74775 | 04:50 | |
| *** hattwick has quit IRC | 05:01 | |
| *** garyk has quit IRC | 05:02 | |
| *** ywu has quit IRC | 05:04 | |
| *** e1mer has joined #openstack-dev | 05:10 | |
| *** aa__ has quit IRC | 05:28 | |
| *** aa__ has joined #openstack-dev | 05:32 | |
| *** e1mer has quit IRC | 05:34 | |
| LinuxJedi | notmyname: a retrigger seems to have reported back, I am not sure what is wrong yet though | 05:53 |
| *** garyk has joined #openstack-dev | 05:55 | |
| *** dtroyer is now known as dtroyer_zzz | 05:59 | |
| *** yamahata has joined #openstack-dev | 06:04 | |
| *** littleidea has quit IRC | 06:14 | |
| *** vanchester12416 has joined #openstack-dev | 06:41 | |
| *** pmezard has joined #openstack-dev | 06:57 | |
| *** fc__ has joined #openstack-dev | 07:11 | |
| *** reidrac has joined #openstack-dev | 07:13 | |
| *** pasm has joined #openstack-dev | 07:25 | |
| *** shang has quit IRC | 07:28 | |
| *** darraghb has joined #openstack-dev | 07:29 | |
| *** mindpixel has joined #openstack-dev | 07:30 | |
| *** gael has joined #openstack-dev | 07:38 | |
| *** shang has joined #openstack-dev | 07:42 | |
| *** pixelbeat has joined #openstack-dev | 07:49 | |
| *** pmezard has quit IRC | 07:54 | |
| *** pmezard has joined #openstack-dev | 07:57 | |
| *** hattwick has joined #openstack-dev | 07:58 | |
| *** derekh has joined #openstack-dev | 08:07 | |
| *** apevec has joined #openstack-dev | 08:08 | |
| *** GheRivero has joined #openstack-dev | 08:08 | |
| *** matwood has quit IRC | 08:08 | |
| *** maploin has joined #openstack-dev | 08:14 | |
| *** maploin has quit IRC | 08:15 | |
| *** maploin has joined #openstack-dev | 08:15 | |
| *** aa has joined #openstack-dev | 08:46 | |
| *** aa__ has quit IRC | 08:46 | |
| *** danpb has joined #openstack-dev | 09:14 | |
| davidha | Hi, I noticed that openstack includes its own memcached daemon | 09:21 |
| davidha | sorry - I meant client :) | 09:21 |
| davidha | Can anyone elaborate what is special about the openstack memcached client - and why a standard client was not used? | 09:22 |
| *** longhudou has joined #openstack-dev | 09:35 | |
| *** pasm has quit IRC | 09:35 | |
| *** Mandell has quit IRC | 09:40 | |
| LinuxJedi | guys: there have been a few cases where Jenkins results don't appear to be feeding back to Gerrit. If you notice this happening can you let us know? I'm trying to figure out why this is happening (and have restarted the Gerrit trigger plugin for now) | 09:42 |
| *** pasm has joined #openstack-dev | 09:47 | |
| *** b1rkh0ff has joined #openstack-dev | 09:52 | |
| *** rafaduran has quit IRC | 09:53 | |
| *** rbasak has joined #openstack-dev | 10:02 | |
| *** bourke_ has quit IRC | 10:13 | |
| *** bourke has joined #openstack-dev | 10:13 | |
| *** bencherian has quit IRC | 10:16 | |
| *** flaviamissi has quit IRC | 10:34 | |
| *** rods has joined #openstack-dev | 10:44 | |
| *** danwent has quit IRC | 10:51 | |
| *** semyazz has joined #openstack-dev | 10:51 | |
| *** milner has joined #openstack-dev | 11:08 | |
| *** wiliam has joined #openstack-dev | 11:55 | |
| *** salgado has joined #openstack-dev | 11:57 | |
| *** salgado has joined #openstack-dev | 11:57 | |
| *** flaviamissi has joined #openstack-dev | 12:05 | |
| *** dachary has joined #openstack-dev | 12:08 | |
| *** semyazz has quit IRC | 12:14 | |
| *** anand_happs has joined #openstack-dev | 12:14 | |
| *** melmoth has joined #openstack-dev | 12:15 | |
| melmoth | Hi there ! I experienced a problem i was able to reproduce twice. when running euca-associate-address things works... | 12:16 |
| melmoth | but if i remove the static network already define, then create the same..then euca-associate address do not work anymore | 12:17 |
| melmoth | I did not find an open bug for this. What should i do when creating a bug.. reproducing it on the package shipped with ubuntu 12.04, reproduce it with devstack on teh latest essex branch ? | 12:17 |
| *** edygarcia has joined #openstack-dev | 12:32 | |
| *** Ruetobas has quit IRC | 12:33 | |
| *** dachary has quit IRC | 12:35 | |
| *** dachary has joined #openstack-dev | 12:37 | |
| *** Ruetobas has joined #openstack-dev | 12:37 | |
| *** melmoth has quit IRC | 12:44 | |
| *** anand_happs has quit IRC | 12:48 | |
| *** anand_happs has joined #openstack-dev | 12:48 | |
| *** dachary has quit IRC | 12:49 | |
| *** dachary has joined #openstack-dev | 12:51 | |
| *** rafaduran has joined #openstack-dev | 12:54 | |
| *** melmoth has joined #openstack-dev | 12:56 | |
| davidha | Hi, anyone here familiar with the details of the MemcacheRing | 13:02 |
| *** deshantm has joined #openstack-dev | 13:02 | |
| *** aa has quit IRC | 13:07 | |
| *** deshantm has quit IRC | 13:09 | |
| *** iryoung has joined #openstack-dev | 13:10 | |
| davidha | GM. Openstack implements its own memcached client - apperently there are standard opensource clients to choose from - why? | 13:20 |
| mtaylor | davidha: that makes me want to cry | 13:23 |
| davidha | +mtaylor: dont cry - there may be good reasons to do that - but if there isnt, openstack should reduce scope to areas where it contributes | 13:25 |
| mtaylor | davidha: well, as someone who has worked on both libmemcached and memcached, I'll say that no, I do not believe there are good reasons to do that | 13:27 |
| *** deshantm has joined #openstack-dev | 13:27 | |
| mtaylor | davidha: but then I tend to be opinionated :) | 13:27 |
| *** adalbas has joined #openstack-dev | 13:28 | |
| adalbas | ping | 13:28 |
| davidha | mtaylor, so what triggered this work? Why 'standard' libs were not used instead? | 13:28 |
| mtaylor | davidha: no idea | 13:28 |
| mtaylor | davidha: that's why I was crying earlier | 13:29 |
| davidha | :) | 13:29 |
| davidha | mtaylor, who in the team was working on the client? I want my shot at trying to understand :) | 13:33 |
| *** salgado is now known as salgado-brb | 13:38 | |
| *** roge has joined #openstack-dev | 13:38 | |
| *** dachary has quit IRC | 13:43 | |
| *** dachary has joined #openstack-dev | 13:44 | |
| *** salgado-brb has quit IRC | 13:45 | |
| *** sstent has quit IRC | 13:52 | |
| *** sstent has joined #openstack-dev | 13:52 | |
| *** anand_happs has quit IRC | 14:01 | |
| mtaylor | davidha: on what client? do you have a link? | 14:01 |
| davidha | class MemcacheRing inside /common/memcached.py | 14:03 |
| davidha | mtaylor: (i was referring to the memcached client ^^) | 14:06 |
| mtaylor | davidha: ah, you are talking about swift | 14:09 |
| mtaylor | notmyname: how come you guys have your own memcached protocol impl? | 14:09 |
| *** dtroyer_zzz is now known as dtroyer | 14:10 | |
| davidha | mtaylor, notmyname: Does swift common files (e.g. wsgi.py, memcached.py etc.) not shared with other openstack projects? | 14:12 |
| davidha | Does=Do | 14:13 |
| mtaylor | davidha: openstack.common is an ongoing project, nobody is fully using it/sharing through it yet | 14:14 |
| mtaylor | davidha: but part of the plan is to get projects sharing common code through it - and potentially to make it in to a proper library at some pooint | 14:14 |
| *** davidha has quit IRC | 14:15 | |
| *** davidha has joined #openstack-dev | 14:16 | |
| *** littleidea has joined #openstack-dev | 14:30 | |
| *** mnaser has joined #openstack-dev | 14:33 | |
| *** openstackjenkins has quit IRC | 14:36 | |
| *** sandywalsh has joined #openstack-dev | 14:36 | |
| *** openstackjenkins has joined #openstack-dev | 14:36 | |
| *** ncode has joined #openstack-dev | 14:42 | |
| LinuxJedi | notmyname: the gerrit trigger appears to be working fine now | 14:43 |
| zykes- | any piston cloud guys here ? | 14:44 |
| *** openstackjenkins has quit IRC | 14:45 | |
| *** openstackjenkins has joined #openstack-dev | 14:46 | |
| *** vanchester12416 has quit IRC | 14:47 | |
| notmyname | LinuxJedi: thanks | 14:49 |
| *** sandywalsh_ has joined #openstack-dev | 14:49 | |
| notmyname | mtaylor: I'd have to ask around on that (re memcache impl) | 14:49 |
| *** sandywalsh has quit IRC | 14:50 | |
| gholt | davidha: mtaylor: The memcached client thing is more a question for redbo, but like many things like that, there was some shortcoming with the standard one back over three years ago when he made the Swift one. | 14:50 |
| mtaylor | notmyname: k. not super important ... davidha was just asking and then my old libmemcached developer personality kicked in and was curious | 14:50 |
| mtaylor | gholt: makes sense | 14:50 |
| mtaylor | gholt: can chalk up to hysterical raisins | 14:51 |
| *** utlemming has joined #openstack-dev | 14:51 | |
| gholt | Yeah, I do know he gave python-memcache (the packaged version on Lucid) a go first though. | 14:51 |
| gholt | And it is a dead simple protocol. Choosing webob however, was a whole different little nightmare, so... you lose some, you lose some. :) | 14:52 |
| *** salgado has joined #openstack-dev | 14:53 | |
| *** salgado has joined #openstack-dev | 14:53 | |
| *** edygarcia has quit IRC | 14:54 | |
| mtaylor | hehehe | 14:55 |
| mtaylor | oh yeah, also - lucid. I forgot | 14:55 |
| mtaylor | python-memcache is a bit of a turdpile | 14:55 |
| *** aa has joined #openstack-dev | 14:57 | |
| *** maploin has quit IRC | 14:59 | |
| *** littleidea has quit IRC | 15:02 | |
| *** openstackjenkins has quit IRC | 15:02 | |
| *** openstackjenkins has joined #openstack-dev | 15:02 | |
| gholt | Hey- Lucid was cutting edge then! | 15:03 |
| gholt | :) | 15:03 |
| mtaylor | I remember those days... | 15:03 |
| mtaylor | "hey! we can stop supporting dapper now!" | 15:04 |
| gholt | Heh, yeah, kind of a different feel to it on the support side | 15:04 |
| mtaylor | yup | 15:04 |
| mtaylor | hell, be glad you weren't doing an open source C++ project that wanted to be cutting edge but also needed to build on hardy | 15:05 |
| * mtaylor still has flashback nightmares | 15:05 | |
| LinuxJedi | mtaylor: boost? | 15:05 |
| LinuxJedi | :) | 15:05 |
| mtaylor | LinuxJedi: yup. remember the version fun we had with that? | 15:05 |
| mtaylor | at least autoconf doesn't suck-ass as much as setuptools | 15:06 |
| LinuxJedi | mtaylor: it is still a PITA whenever I do work on Drizzle :) | 15:06 |
| * mtaylor can't believe he just said those words | 15:06 | |
| *** reidrac has quit IRC | 15:06 | |
| LinuxJedi | mtaylor: it could have been worse, we could have needed Windows support too | 15:06 |
| mtaylor | solaris and freebsd were bad enough | 15:07 |
| gholt | Hmm, I guess Lucid went out 2010.04 though right?, so my timespan memory is faulty. We had our first production Swift here right about when Lucid went out. | 15:07 |
| LinuxJedi | yep, I swear if we needed AIX support I would have shot someone | 15:07 |
| LinuxJedi | mtaylor: you see the latest on Windows 8. Visual Studio Express won't compile non-metro apps. So death to Windows Open Source I guess | 15:08 |
| *** edygarcia has joined #openstack-dev | 15:09 | |
| mtaylor | LinuxJedi: wow. that's a strong choice on their part... | 15:09 |
| LinuxJedi | yep, I can kinda see why, but bad news for a lot of people | 15:09 |
| mtaylor | gholt: hrm. wow - it seems older than that :) | 15:09 |
| gholt | Maybe it was just Rackspace Cloud Files that was out for 3+ years, and the memcached part probably stems from back then. | 15:09 |
| gholt | And then, we were on the "latest" debian, if you can call it that, hehe. | 15:10 |
| mtaylor | hahahahaha | 15:10 |
| LinuxJedi | :) | 15:10 |
| gholt | Poor debian, gotta love 'em gotta, but... | 15:10 |
| davidha | gholt, mtaylor: thanks - I now get the historical context... :) I guess it would be nice to replace this code with using a standard lib next time someone needed to touch it. | 15:15 |
| *** timello has joined #openstack-dev | 15:19 | |
| *** openstackjenkins has quit IRC | 15:23 | |
| *** openstackjenkins has joined #openstack-dev | 15:24 | |
| *** littleidea has joined #openstack-dev | 15:27 | |
| *** katell has joined #openstack-dev | 15:28 | |
| *** apevec has quit IRC | 15:32 | |
| *** eglynn_ has quit IRC | 15:48 | |
| *** eglynn__ has joined #openstack-dev | 15:48 | |
| *** openstackjenkins has quit IRC | 15:51 | |
| *** openstackjenkins has joined #openstack-dev | 15:52 | |
| darraghb | tks | 16:06 |
| darraghb | damn, wrong window | 16:06 |
| *** garyk has quit IRC | 16:17 | |
| *** davidha has quit IRC | 16:20 | |
| *** davidha has joined #openstack-dev | 16:22 | |
| *** anand_happs has joined #openstack-dev | 16:32 | |
| *** gael has quit IRC | 16:36 | |
| *** mindpixel has quit IRC | 16:39 | |
| *** anand_happs has quit IRC | 16:43 | |
| *** anand_happs has joined #openstack-dev | 16:43 | |
| *** melmoth has quit IRC | 16:44 | |
| *** deshantm has quit IRC | 16:48 | |
| *** pixelbeat has quit IRC | 16:49 | |
| *** anand_happs has quit IRC | 16:50 | |
| *** bencherian has joined #openstack-dev | 16:53 | |
| *** rbasak has quit IRC | 16:56 | |
| *** derekh has quit IRC | 16:57 | |
| *** dtroyer is now known as dtroyer_zzz | 16:58 | |
| *** danwent has joined #openstack-dev | 16:59 | |
| *** darraghb has quit IRC | 17:23 | |
| *** danpb has quit IRC | 17:31 | |
| *** b1rkh0ff has quit IRC | 17:32 | |
| *** armaan has joined #openstack-dev | 17:35 | |
| *** armaan has left #openstack-dev | 17:35 | |
| *** iryoung has quit IRC | 17:36 | |
| *** b1rkh0ff has joined #openstack-dev | 17:44 | |
| *** PotHix has joined #openstack-dev | 17:51 | |
| *** GheRivero has quit IRC | 17:53 | |
| *** maplebed has joined #openstack-dev | 17:53 | |
| *** dtroyer_zzz is now known as dtroyer | 17:53 | |
| *** dtroyer is now known as dtroyer_zzz | 18:00 | |
| *** dtroyer_zzz is now known as dtroyer | 18:14 | |
| *** matwood has joined #openstack-dev | 18:19 | |
| *** dtroyer is now known as dtroyer_zzz | 18:22 | |
| *** camm has quit IRC | 18:22 | |
| *** anand_happs has joined #openstack-dev | 18:27 | |
| *** pasm has quit IRC | 18:35 | |
| *** Mandell has joined #openstack-dev | 18:40 | |
| *** utlemming has quit IRC | 18:40 | |
| *** utlemming has joined #openstack-dev | 18:40 | |
| *** dtroyer_zzz is now known as dtroyer | 18:56 | |
| *** dtroyer is now known as dtroyer_zzz | 18:58 | |
| *** dtroyer_zzz is now known as dtroyer | 19:04 | |
| *** utlemming has quit IRC | 19:06 | |
| *** notmyname has quit IRC | 19:07 | |
| *** garyk has joined #openstack-dev | 19:08 | |
| *** notmyname has joined #openstack-dev | 19:08 | |
| *** ChanServ sets mode: +v notmyname | 19:08 | |
| *** mnaser_ has joined #openstack-dev | 19:13 | |
| *** mnaser has quit IRC | 19:14 | |
| *** camm has joined #openstack-dev | 19:17 | |
| *** zaitcev has joined #openstack-dev | 19:24 | |
| *** bencherian has quit IRC | 19:24 | |
| *** deshantm has joined #openstack-dev | 19:31 | |
| *** zul has quit IRC | 19:33 | |
| *** zul has joined #openstack-dev | 19:36 | |
| *** semyazz has joined #openstack-dev | 19:46 | |
| *** semyazz has left #openstack-dev | 19:47 | |
| *** matwood has quit IRC | 19:54 | |
| *** armaan has joined #openstack-dev | 19:58 | |
| *** dtroyer is now known as dtroyer_zzz | 20:07 | |
| *** dtroyer_zzz is now known as dtroyer | 20:07 | |
| *** wiliam has quit IRC | 20:15 | |
| *** armaan has left #openstack-dev | 20:16 | |
| *** aa has quit IRC | 20:19 | |
| *** aa has joined #openstack-dev | 20:27 | |
| *** matwood has joined #openstack-dev | 20:27 | |
| *** deshantm has quit IRC | 20:27 | |
| *** matwood has quit IRC | 20:38 | |
| *** matwood has joined #openstack-dev | 20:46 | |
| *** pixelbeat has joined #openstack-dev | 20:50 | |
| *** soren_ has quit IRC | 21:04 | |
| *** soren_ has joined #openstack-dev | 21:04 | |
| *** ChanServ sets mode: +v soren_ | 21:04 | |
| *** soren_ is now known as soren | 21:04 | |
| *** davidha has quit IRC | 21:12 | |
| *** davidha has joined #openstack-dev | 21:13 | |
| *** flaviamissi has quit IRC | 21:17 | |
| *** flaviamissi has joined #openstack-dev | 21:22 | |
| *** camm has quit IRC | 21:22 | |
| *** camm has joined #openstack-dev | 21:23 | |
| *** davidha has quit IRC | 21:29 | |
| *** davidha has joined #openstack-dev | 21:30 | |
| *** garyk has quit IRC | 21:39 | |
| *** ywu has joined #openstack-dev | 21:39 | |
| *** openstackjenkins has quit IRC | 21:41 | |
| *** openstackjenkins has joined #openstack-dev | 21:42 | |
| *** b1rkh0ff has quit IRC | 21:45 | |
| davidha | notmyname hi, | 21:47 |
| notmyname | howdy | 21:47 |
| davidha | Why do we have tests in swift that include relative URIs - even more so, why are thos treated as different than teh absolute URIs? | 21:48 |
| *** ncode has quit IRC | 21:49 | |
| *** utlemming has joined #openstack-dev | 21:49 | |
| davidha | test/functional/tests.py - line 640 | 21:49 |
| davidha | notmyname: comapred to line 657 | 21:50 |
| notmyname | hmm...git blame says those lines haven't been touched in nearly 2 years. suffice to say, I have no idea without spending some time looking at the code | 21:52 |
| davidha | notmyname: ok, would it be resonable to asume that URI X/Y/Z should e treated the same as /X/Y/Z? | 21:53 |
| davidha | I.e. they represnet the same object? | 21:54 |
| davidha | assume | 21:54 |
| *** garyk has joined #openstack-dev | 21:57 | |
| notmyname | davidha: looks like that list is used starting in line 692, and there is a difference in how they are treated. I don't know why they are different yet | 21:58 |
| *** dachary has quit IRC | 21:59 | |
| *** b1rkh0ff has joined #openstack-dev | 21:59 | |
| *** dachary has joined #openstack-dev | 21:59 | |
| *** Guest74775 has quit IRC | 21:59 | |
| *** Adri2000 has joined #openstack-dev | 22:01 | |
| *** Adri2000 is now known as Guest35099 | 22:01 | |
| davidha | Running nosetests -v test.functional.tests:TestContainerPaths.testContainerListing - on my system the test seem to expect to get all 39 objects sorted - meaning that /X/Y is different than X/Y | 22:02 |
| davidha | notmyname: Actually, looking on the wire with wireshark, it seems that the URIs not relative. Instead they are appended to '/v1/AUTH_test/longcontainername/' such that /X/Y becomes '/v1/AUTH_test/longcontainername//X/Y' and X/Y becomes: '/v1/AUTH_test/longcontainername/X/Y' | 22:08 |
| davidha | Still, it is odd that the test expects that these two objects will be considered as different and appear seperatly in the container listing | 22:09 |
| *** flaviamissi has quit IRC | 22:10 | |
| gholt | Everything after the container/ is the object name. No matter what it is. So /object is a different name than object, or object/, or object// etc. :) Why? Dunno, but that's the "way it's always been". | 22:14 |
| notmyname | davidha: ah ok. so yes they are different object names (but we've discussed '/'*N patterns before) and those tests are testing the path query | 22:14 |
| notmyname | davidha: the path query for listings is an older (and intentionally non-documented) feature that exists because of historical rackspace cloud files reasons | 22:16 |
| davidha | ok - so if the URI will change to consider N*/ == / than these need to change as well. | 22:16 |
| davidha | do you btw see an immidiate issue with this change ? | 22:16 |
| notmyname | davidha: I'm not yet convinced that's a change that needs to happen ;-) | 22:17 |
| gholt | There's a change proposed? | 22:17 |
| notmyname | davidha: but you are correct that these would need to change if the N*/ == / were to happen | 22:17 |
| notmyname | gholt: only discussed in here. nothing more official than that | 22:17 |
| gholt | Ah, well such a change I would consider as "too late". Folks already have objects stored. | 22:18 |
| notmyname | gholt: davidha is working on abstracting the wsgi parts of swift to run under apache. turns out that apache squashes multiple '/' into just one | 22:18 |
| gholt | Ah, well, if the cluster was always behind Apache, then I guess there isn't a problem. But Swift itself would need to continue to support opaque object names. Or at least those I deploy. :) | 22:20 |
| notmyname | gotta run... | 22:21 |
| *** ywu has quit IRC | 22:22 | |
| gholt | Apache seems to pass along extra slashes, so I'm not sure that's a problem. | 22:23 |
| davidha | gholt - where did you see that? | 22:23 |
| gholt | In my logs at least. I mean, it squashes them to resolve to on disk file names, but it seems to log them separately. | 22:24 |
| *** ywu has joined #openstack-dev | 22:25 | |
| davidha | The logs show the URI processed by httpd - when it is offered to the wsgi application it would collapse N*/ to be / | 22:25 |
| gholt | Hmm, I wonder if that's new. Didn't used to do that, back when we did use mod_wsgi. | 22:26 |
| davidha | gholt: R U sure? - if I remember right, this seem to be a apache feature, not mod_wsgi one | 22:27 |
| * gholt shrugs | 22:27 | |
| gholt | It was a long while ago. :) | 22:27 |
| *** ywu has quit IRC | 22:28 | |
| davidha | gholt: so you figure that objects stored out there may rely on slashes not being collapsed (i.e. someone keeps both a ABC/EFG object and a seperate ABC//EFG object)? | 22:32 |
| gholt | I don't know about the double thing for sure. I do know there are many stored with multi-slashes and a leading slash and if those were removed by the time they got to Swift, they'd result in 404s. | 22:40 |
| *** pmezard has quit IRC | 22:40 | |
| *** maplebed has quit IRC | 22:42 | |
| *** andrewsben is now known as andrewsben_zz | 22:42 | |
| gholt | Some folks even use urls as the object names, as in a PUT of http://swift.com/v1/account/container/http://somehost.com/blah.html | 22:43 |
| *** hugokuo has quit IRC | 22:43 | |
| davidha | gholt - ok, I will put more thought into this - basically, nothing prevents a new installation that chooses to use httpd to simply collapse slashes while a different installation of openstack choses a different web front end and do not collapse slashes. There may be a smoother way - either a httpd configuration (I doubt it, but will look for one) or a client side solution (encoding the | 22:48 |
| davidha | slashes?) | 22:48 |
| *** ncode has joined #openstack-dev | 22:50 | |
| *** openstackjenkins has quit IRC | 23:01 | |
| *** openstackjenkins has joined #openstack-dev | 23:01 | |
| gholt | davidha: Looks like you can get the original request line from mod_wsgi with env['REQUEST_URI']. You'll have to parse off the query string, but it should get you there. | 23:08 |
| *** bencherian has joined #openstack-dev | 23:22 | |
| *** Guest35099 is now known as Adri2000 | 23:24 | |
| *** Adri2000 has joined #openstack-dev | 23:24 | |
| *** flaviamissi has joined #openstack-dev | 23:34 | |
| *** dachary has quit IRC | 23:46 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!