| *** itlinux has quit IRC | 00:11 | |
| *** MelindaBunny has quit IRC | 00:11 | |
| *** NM has joined #openstack-swift | 00:15 | |
| *** NM has quit IRC | 00:17 | |
| *** clarkb has joined #openstack-swift | 00:24 | |
| clarkb | clayg: hey, so I've made a bit of progress in other areas that were tricky for enabling tls by default in devstack. Looks like the swift functional tests still fail with that enabled. Anything I can do to help out around that? | 00:25 |
|---|---|---|
| clarkb | clayg: I'm happy to write the chagnes if you can point me in the right direction | 00:25 |
| *** alenavolk has quit IRC | 00:29 | |
| *** alenavolk has joined #openstack-swift | 00:31 | |
| *** alenavolk has quit IRC | 00:38 | |
| *** zhengyin has joined #openstack-swift | 00:47 | |
| timburke | well that's...interesting http://logs.openstack.org/69/469269/1/check/gate-swift-docs-ubuntu-xenial/9566728/console.html#_2017-05-30_21_55_27_938896 | 01:03 |
| timburke | looks like https://github.com/openstack/swift/commit/2ca3035 was a bit premature -- so when sphinx was updated in the gate following https://github.com/openstack/requirements/commit/6b4753c our docs broke | 01:05 |
| timburke | meanwhile, the failure in https://8b86aea46fb38e6450f2-0e5f4c086da474abc1df58826577db2f.ssl.cf1.rackcdn.com/469269/6698/probetests/console.txt has a traceback that simply doesn't make any sense | 01:08 |
| *** lucasxu has joined #openstack-swift | 01:17 | |
| *** lucasxu has quit IRC | 01:17 | |
| *** lucasxu has joined #openstack-swift | 01:20 | |
| *** lucasxu has quit IRC | 01:22 | |
| *** lucasxu has joined #openstack-swift | 01:23 | |
| *** NM has joined #openstack-swift | 01:31 | |
| *** ianychoi_ is now known as ianychoi | 01:35 | |
| *** tovin07_ has joined #openstack-swift | 01:39 | |
| kota_ | good morning | 01:44 |
| *** aselius has quit IRC | 01:45 | |
| *** lan1 has quit IRC | 01:49 | |
| *** NM has quit IRC | 01:55 | |
| *** lucasxu has quit IRC | 02:01 | |
| *** klrmn has quit IRC | 02:02 | |
| *** alenavolk has joined #openstack-swift | 02:10 | |
| *** lucasxu has joined #openstack-swift | 02:11 | |
| *** lucasxu has quit IRC | 02:13 | |
| *** alenavolk has quit IRC | 02:14 | |
| *** lucasxu has joined #openstack-swift | 02:14 | |
| *** alenavolk has joined #openstack-swift | 02:25 | |
| *** mwheckmann has joined #openstack-swift | 02:30 | |
| *** lucasxu has quit IRC | 02:40 | |
| *** lucasxu has joined #openstack-swift | 02:45 | |
| *** lucasxu has quit IRC | 02:47 | |
| notmyname | reminder, team meeting in about 4 hours | 02:49 |
| *** klrmn has joined #openstack-swift | 02:51 | |
| openstackgerrit | Tim Burke proposed openstack/swift master: Revert "Make Sphinx treat warnings as errors" https://review.openstack.org/469314 | 02:53 |
| *** lucasxu has joined #openstack-swift | 02:56 | |
| openstackgerrit | Tim Burke proposed openstack/swift master: Make Sphinx treat warnings as errors https://review.openstack.org/469315 | 02:57 |
| timburke | land the first one to un-stick the gate, work with infra to get green balls on the second one | 02:58 |
| *** lucasxu has quit IRC | 03:04 | |
| *** mwheckmann has quit IRC | 03:05 | |
| *** psachin has quit IRC | 03:29 | |
| *** sanchitmalhotra has joined #openstack-swift | 03:30 | |
| mattoliverau | timburke: nice work on the gate issues. | 03:33 |
| *** chsc has joined #openstack-swift | 03:33 | |
| *** chsc has joined #openstack-swift | 03:33 | |
| *** chsc has quit IRC | 03:39 | |
| *** cshastri has joined #openstack-swift | 03:41 | |
| *** links has joined #openstack-swift | 03:45 | |
| *** lucasxu has joined #openstack-swift | 03:45 | |
| *** kei_yama has quit IRC | 03:46 | |
| *** psachin has joined #openstack-swift | 03:47 | |
| *** kestrels has joined #openstack-swift | 03:55 | |
| *** kestrels2 has quit IRC | 03:55 | |
| *** kei_yama has joined #openstack-swift | 04:03 | |
| *** wasmum has quit IRC | 04:05 | |
| *** jaosorior has quit IRC | 04:10 | |
| *** lucasxu has quit IRC | 04:11 | |
| *** jaosorior has joined #openstack-swift | 04:13 | |
| *** zhurong has joined #openstack-swift | 04:15 | |
| *** zhurong has quit IRC | 04:19 | |
| *** alenavolk has quit IRC | 04:22 | |
| *** alenavolk has joined #openstack-swift | 04:25 | |
| *** zhengyin has quit IRC | 04:32 | |
| *** chsc has joined #openstack-swift | 04:51 | |
| *** chsc has quit IRC | 04:58 | |
| *** alenavolk has quit IRC | 05:09 | |
| *** alenavolk has joined #openstack-swift | 05:11 | |
| *** zhengyin has joined #openstack-swift | 05:11 | |
| *** alenavolk has quit IRC | 05:25 | |
| *** alenavolk has joined #openstack-swift | 05:26 | |
| *** alenavolk has quit IRC | 05:26 | |
| *** skudlik has joined #openstack-swift | 05:30 | |
| *** gyee has quit IRC | 05:34 | |
| mahatic | rledisez: sure, no problem. Thanks for updating the agenda. | 05:43 |
| *** ChubYann has quit IRC | 05:58 | |
| notmyname | hello, world | 06:08 |
| *** jaosorior has quit IRC | 06:08 | |
| *** mvk has quit IRC | 06:09 | |
| *** rcernin has joined #openstack-swift | 06:10 | |
| mattoliverau | notmyname: o/ | 06:12 |
| notmyname | hey mattoliverau. how are you? | 06:13 |
| mattoliverau | Ok, got some interviews lined up, so that's exciting. Been enjoying time with the family. Lucy is as a cool interactive stage.. so so far unemployment has been kinda nice :) | 06:14 |
| *** jaosorior has joined #openstack-swift | 06:17 | |
| notmyname | that's fun. I hope you do find something soon | 06:23 |
| *** m_kazuhiro has joined #openstack-swift | 06:29 | |
| *** zhengyin has quit IRC | 06:33 | |
| *** zhengyin has joined #openstack-swift | 06:34 | |
| tovin07_ | hi notmyname | 06:41 |
| tovin07_ | recently, we talked about osprofiler in swift | 06:41 |
| tovin07_ | last week, i publish a patch for this https://review.openstack.org/#/c/468316/ | 06:41 |
| patchbot | patch 468316 - swift - [WiP] OSprofiler in OpenStack Swift | 06:41 |
| tovin07_ | can you take a look at it and drop some comments? | 06:41 |
| notmyname | yep. I glanced over it this morning, actually | 06:41 |
| tovin07_ | oh, thanks :D | 06:42 |
| notmyname | I definitely appreciate all the detail you put in the commit message | 06:43 |
| notmyname | and I'm glad to see timburke working with you on it | 06:43 |
| notmyname | I'm still nervous about the extra dependencies | 06:44 |
| *** zhengyin has quit IRC | 06:44 | |
| tovin07_ | yes, about extra dependencies, I tried not to use webob, and make osprofiler turned of by default | 06:47 |
| tovin07_ | *off | 06:47 |
| *** tesseract has joined #openstack-swift | 06:53 | |
| *** klrmn has quit IRC | 06:58 | |
| mahatic | fyi, meeting time in about 3 mins in #openstack-meeting | 06:58 |
| *** zhengyin has joined #openstack-swift | 07:00 | |
| notmyname | meeting time in #openstack-swift | 07:00 |
| kota_ | notmyname: thx for reminder | 07:01 |
| acoles | good morning | 07:02 |
| *** mvk has joined #openstack-swift | 07:08 | |
| *** geaaru has joined #openstack-swift | 07:08 | |
| * kota_ is feeling i need availabe commands list when i was the chair on the irc meeting | 07:19 | |
| openstackgerrit | Christian Schwede proposed openstack/swift master: Add support to increase object ring partition power https://review.openstack.org/337297 | 07:19 |
| notmyname | kota_: #link #info #agreed #topic and #startmeeting are the ones I know | 07:22 |
| notmyname | the vote one are tricky, so I just don't mess with them | 07:22 |
| kota_ | notmyname: cool, thanks | 07:22 |
| kota_ | I'll be add them into meeting wiki for other chairs | 07:23 |
| kota_ | s/be// | 07:23 |
| *** pcaruana has joined #openstack-swift | 07:26 | |
| acoles | kota_: the commands are listed by meetbot whent he meeting starts, but having them on the wiki would be great too | 07:31 |
| *** cbartz has joined #openstack-swift | 07:33 | |
| *** skudlik has quit IRC | 07:34 | |
| *** qwerty has joined #openstack-swift | 07:40 | |
| *** qwerty has quit IRC | 07:41 | |
| *** aselius has joined #openstack-swift | 07:45 | |
| *** adriant has quit IRC | 07:52 | |
| *** jeffli has joined #openstack-swift | 07:57 | |
| acoles | rledisez: which is highest priority for you - https://review.openstack.org/#/c/435771/ or https://review.openstack.org/#/c/456921/ ? | 08:01 |
| patchbot | patch 435771 - swift - Rewrite redirection in cname_lookup & domain_remap | 08:01 |
| patchbot | patch 456921 - swift - Fix SSYNC failing to replicate unexpired object | 08:01 |
| acoles | rledisez: or both? :) | 08:01 |
| rledisez | SSYNC one, it's critical for, data availability | 08:01 |
| rledisez | cname_lookup, it's just broken :) | 08:01 |
| rledisez | so not a big deal :D | 08:02 |
| notmyname | mahatic: I'll review the meeting log my tomorrow and work on updating the meeting agenda page. also feel free to update it yourself as well :-) | 08:02 |
| hugokuo | anyone knows how's the correct string should be for delete object name like this | 08:02 |
| acoles | mahatic: ^^ note rledisez's response , if you have time to look at a patch | 08:02 |
| * notmyname is off to bed | 08:02 | |
| acoles | notmyname: good night! | 08:02 |
| hugokuo | abc/abc/you're fine!abc | 08:03 |
| mahatic | notmyname: sure, thanks. good night! | 08:03 |
| hugokuo | If I put backslash prior to examination mark, Swift CLI decoded it as %5C%21 , but the actual file name should be decode as %21 | 08:04 |
| hugokuo | not sure if this should be a bug in Swift CLI | 08:04 |
| mahatic | acoles: thanks, will do. SSYNC already seems to have a couple of people around. I'll try to get around the broken fix :) | 08:04 |
| mahatic | will try to keep a tab on how SSYNC patch is progressing as well | 08:04 |
| kota_ | notmyname: good night | 08:05 |
| *** oshritf has joined #openstack-swift | 08:10 | |
| *** m_kazuhiro has quit IRC | 08:10 | |
| *** oshritf has quit IRC | 08:10 | |
| *** oshritf has joined #openstack-swift | 08:20 | |
| *** oshritf has quit IRC | 08:33 | |
| *** kei_yama has quit IRC | 08:44 | |
| *** vint_bra1 has joined #openstack-swift | 08:53 | |
| *** ediardo_ has joined #openstack-swift | 08:54 | |
| *** remixtj has joined #openstack-swift | 08:55 | |
| *** jeffl has joined #openstack-swift | 08:56 | |
| *** baffle__ has joined #openstack-swift | 08:56 | |
| *** briancli1e has joined #openstack-swift | 08:56 | |
| *** portante_ has joined #openstack-swift | 09:00 | |
| *** dmellado_ has joined #openstack-swift | 09:00 | |
| *** jeffli has quit IRC | 09:01 | |
| *** links has quit IRC | 09:01 | |
| *** remix_tj has quit IRC | 09:01 | |
| *** vint_bra has quit IRC | 09:01 | |
| *** ediardo has quit IRC | 09:01 | |
| *** baffle has quit IRC | 09:01 | |
| *** dmellado has quit IRC | 09:01 | |
| *** portante has quit IRC | 09:01 | |
| *** jistr has quit IRC | 09:01 | |
| *** briancline has quit IRC | 09:01 | |
| *** portante_ is now known as portante | 09:01 | |
| *** ediardo_ is now known as ediardo | 09:01 | |
| *** jistr has joined #openstack-swift | 09:02 | |
| *** charz has quit IRC | 09:03 | |
| *** mingyu has joined #openstack-swift | 09:03 | |
| *** charz has joined #openstack-swift | 09:06 | |
| *** links has joined #openstack-swift | 09:06 | |
| *** skudlik has joined #openstack-swift | 09:10 | |
| *** mingyu has quit IRC | 09:17 | |
| *** jaosorior is now known as jaosorior_lunch | 09:19 | |
| *** NM has joined #openstack-swift | 09:36 | |
| *** NM has quit IRC | 09:45 | |
| *** dmellado_ is now known as dmellado | 09:46 | |
| *** aselius has quit IRC | 09:54 | |
| *** mvk has quit IRC | 09:55 | |
| *** tovin07_ has quit IRC | 09:55 | |
| *** mvk has joined #openstack-swift | 10:10 | |
| openstackgerrit | Alistair Coles proposed openstack/swift master: Add structure to storage policy configuration guide https://review.openstack.org/468360 | 10:11 |
| *** links has quit IRC | 10:16 | |
| *** links has joined #openstack-swift | 10:36 | |
| *** jaosorior_lunch is now known as jaosorior | 10:41 | |
| *** zhengyin has quit IRC | 11:08 | |
| *** mingyu has joined #openstack-swift | 11:23 | |
| *** mingyu has quit IRC | 11:33 | |
| *** psachin has quit IRC | 11:49 | |
| *** psachin has joined #openstack-swift | 11:49 | |
| *** mingyu has joined #openstack-swift | 11:52 | |
| *** mingyu has quit IRC | 11:57 | |
| *** MVenesio has joined #openstack-swift | 12:01 | |
| *** chlong has quit IRC | 12:03 | |
| *** psachin has quit IRC | 12:07 | |
| *** cshastri has quit IRC | 12:09 | |
| *** jeffl has quit IRC | 12:10 | |
| *** links has quit IRC | 12:11 | |
| tdasilva | good morning | 12:18 |
| tdasilva | just got caught up with the meeting | 12:18 |
| tdasilva | mahatic: terrific job chairing the meeting! congrats! | 12:18 |
| tdasilva | just coincidentally yesterday I was trying to re-load patch 371150 into my brain, so I will continue to work on it | 12:19 |
| patchbot | https://review.openstack.org/#/c/371150/ - swift - Return 404 on a GET if tombstone is newer | 12:19 |
| tdasilva | driving to the office, will be back in a bit... | 12:20 |
| *** NM has joined #openstack-swift | 12:28 | |
| *** cshastri has joined #openstack-swift | 12:44 | |
| *** mat128 has joined #openstack-swift | 13:04 | |
| *** lan has joined #openstack-swift | 13:05 | |
| *** oshritf has joined #openstack-swift | 13:08 | |
| *** mat128 has quit IRC | 13:08 | |
| *** psachin has joined #openstack-swift | 13:09 | |
| *** lan has quit IRC | 13:09 | |
| *** lucasxu has joined #openstack-swift | 13:11 | |
| *** oshritf has quit IRC | 13:11 | |
| acoles | mattoliverau: thanks for speedy review! | 13:14 |
| *** lifeless has quit IRC | 13:15 | |
| *** chlong has joined #openstack-swift | 13:29 | |
| *** mat128 has joined #openstack-swift | 13:30 | |
| *** klamath has joined #openstack-swift | 13:33 | |
| *** lifeless has joined #openstack-swift | 13:33 | |
| *** klamath has quit IRC | 13:33 | |
| *** klamath has joined #openstack-swift | 13:34 | |
| *** oshritf has joined #openstack-swift | 13:40 | |
| *** zhurong has joined #openstack-swift | 13:42 | |
| *** chlong has quit IRC | 13:43 | |
| *** chlong has joined #openstack-swift | 13:46 | |
| *** cshastri has quit IRC | 13:49 | |
| *** psachin has quit IRC | 13:53 | |
| *** mwheckmann has joined #openstack-swift | 13:55 | |
| *** kestrels has quit IRC | 14:02 | |
| *** cebreidian has quit IRC | 14:02 | |
| *** kestrels has joined #openstack-swift | 14:02 | |
| *** hseipp has joined #openstack-swift | 14:08 | |
| *** kestrels has quit IRC | 14:09 | |
| *** aselius has joined #openstack-swift | 14:13 | |
| *** oshritf has quit IRC | 14:23 | |
| *** zhurong has quit IRC | 14:35 | |
| *** hoonetorg has quit IRC | 15:04 | |
| *** gyee has joined #openstack-swift | 15:09 | |
| *** mat128 has quit IRC | 15:14 | |
| *** hoonetorg has joined #openstack-swift | 15:16 | |
| *** mvk has quit IRC | 15:19 | |
| *** alenavolk has joined #openstack-swift | 15:24 | |
| *** cebruns has joined #openstack-swift | 15:57 | |
| *** cbartz has left #openstack-swift | 15:58 | |
| *** chsc has joined #openstack-swift | 15:58 | |
| *** chsc has quit IRC | 16:03 | |
| *** rcernin has quit IRC | 16:07 | |
| *** tesseract has quit IRC | 16:08 | |
| notmyname | good morning | 16:12 |
| *** hseipp has quit IRC | 16:29 | |
| notmyname | zhongjun: great to see you in the meeting yesterday. I hope the time was more convenient for you | 16:33 |
| clayg | clarkb: I'll spin that up and try to capture some notes - I think I had a working(ish) swift test.conf | 16:35 |
| clayg | clarkb: then there's this change needed in devstack -> https://gist.github.com/clayg/0134511526537410f9348e81d625ed95 | 16:35 |
| clayg | and swift needs to fix this lp bug #1667833 | 16:36 |
| openstack | Launchpad bug 1667833 in OpenStack Object Storage (swift) "func tests must not assert accept-ranges on 304" [Undecided,New] https://launchpad.net/bugs/1667833 | 16:36 |
| clarkb | clayg: do you want to push the devstack change or is it easier if I do it? | 16:36 |
| clayg | ... which I don't think should be complicated - it really should mostly just be the typing and then explaining the reason *why* | 16:36 |
| clayg | clarkb: for the devstack change it would certainly be easier for you to do it - that may not be the best thing for my personal growth - but I think you've been more than patient and helpful - so... enough with that let's get shit done | 16:38 |
| timburke | good morning | 16:39 |
| clayg | there is another change needed to devstack... something with how it writes out the test.conf when TLS is enabled | 16:39 |
| clayg | IIRC I had gotten as far as what was "wrong" with the test.conf generated by devstack's perspective - and created a test.conf that would "work" - but not so far as where/what change needed to be made to swift/devstack so that the two could compromise | 16:40 |
| clayg | I think perhaps - i have some vague memory (timburke help me out) of perhaps - sdauge working out a change to swift-*client*'s test.conf so that it could use a full auth_uri instead of `ssl = true`and other sillyness | 16:41 |
| notmyname | yeah, that landed | 16:41 |
| timburke | no it didn't | 16:42 |
| notmyname | oh. well it was *proposed*! | 16:42 |
| clayg | notmyname: stand back and let timburke drop some knowledge | 16:42 |
| timburke | https://review.openstack.org/#/c/456791/ | 16:42 |
| patchbot | patch 456791 - python-swiftclient - Allow swift functional tests to use keystone auth uri | 16:42 |
| notmyname | that's the one | 16:42 |
| timburke | (one aspect of) the trouble is that one of the swiftclient gates doesn't actually test the in-review change | 16:43 |
| clayg | timburke: that's cool | 16:44 |
| *** rcernin has joined #openstack-swift | 16:46 | |
| *** chsc has joined #openstack-swift | 16:49 | |
| *** chsc has joined #openstack-swift | 16:49 | |
| *** mvk has joined #openstack-swift | 16:49 | |
| *** pcaruana has quit IRC | 16:50 | |
| *** klrmn has joined #openstack-swift | 17:02 | |
| *** klrmn has quit IRC | 17:05 | |
| *** klrmn has joined #openstack-swift | 17:09 | |
| *** ChubYann has joined #openstack-swift | 17:11 | |
| notmyname | timburke: https://review.openstack.org/#/c/469314/ needs to land. any reason not to click +A on it right now? | 17:17 |
| patchbot | patch 469314 - swift - Stop having Sphinx treat warnings as errors | 17:17 |
| notmyname | clarkb: ^^ are there cross-project implications? | 17:17 |
| notmyname | do you know? | 17:17 |
| timburke | notmyname: not that i can think of. there's a chance that something *else* has gone wrong, but we can address that if it comes up | 17:17 |
| clarkb | there shouldn't be | 17:18 |
| notmyname | clarkb: thanks | 17:18 |
| clarkb | since the docs are fairly self contained and where they reference between them its all via urls | 17:18 |
| notmyname | timburke: ok, I'll land it | 17:18 |
| clarkb | curious why you wouldn't just fix the warnings instead? | 17:18 |
| notmyname | then we can reject https://review.openstack.org/#/c/456921/ | 17:18 |
| patchbot | patch 456921 - swift - Fix SSYNC failing to replicate unexpired object | 17:18 |
| timburke | it weirds me out that (1) the releasenotes job ran on the first patchset and (2) it failed because reno was treating warnings as errors (!?) | 17:18 |
| notmyname | clarkb: it's that the gate is using an old version of liberasurecode that emits a warning that sphinx catches and barfs on | 17:19 |
| clarkb | oh interesting | 17:19 |
| notmyname | not reject. recheck | 17:19 |
| notmyname | clarkb: http://logs.openstack.org/21/456921/3/gate/gate-swift-docs-ubuntu-xenial/efa34c2/console.html | 17:19 |
| timburke | notmyname: yeah, "recheck" -- that's the one :-) | 17:19 |
| *** mat128 has joined #openstack-swift | 17:22 | |
| notmyname | timburke: so this doesn't break all of the gate | 17:25 |
| notmyname | I mean, we've had stuff land | 17:25 |
| notmyname | the sphinx change happened nearly 2 months ago | 17:26 |
| timburke | notmyname: but new sphinx didn't come into upper-constraints until yesterday -- https://github.com/openstack/requirements/commit/6b4753c | 17:26 |
| notmyname | oh, so it's the new sphinx that is doing the warn/error thing? | 17:27 |
| notmyname | or that is catching the dependency warning for the first time? | 17:27 |
| *** tonanhngo has joined #openstack-swift | 17:31 | |
| *** ujjain has quit IRC | 17:33 | |
| *** oshritf has joined #openstack-swift | 17:38 | |
| *** jaosorior is now known as jaosorior_away | 17:44 | |
| *** cebruns has quit IRC | 17:56 | |
| *** ujjain has joined #openstack-swift | 18:01 | |
| *** ujjain has quit IRC | 18:01 | |
| *** ujjain has joined #openstack-swift | 18:01 | |
| *** chlong has quit IRC | 18:05 | |
| *** cebruns has joined #openstack-swift | 18:10 | |
| *** cebruns has quit IRC | 18:21 | |
| *** chlong has joined #openstack-swift | 18:22 | |
| *** cebruns has joined #openstack-swift | 18:36 | |
| *** cebruns has quit IRC | 18:36 | |
| *** cebruns has joined #openstack-swift | 18:38 | |
| *** cebruns has quit IRC | 18:39 | |
| *** cebruns has joined #openstack-swift | 18:41 | |
| *** cebruns has quit IRC | 18:41 | |
| *** cebruns has joined #openstack-swift | 18:42 | |
| *** cebruns has quit IRC | 18:42 | |
| portante | notmyname, clayg, torgomatic: have you folks run into behavior differences in the error output between ubuntu and fedora, where on uses `leading back-tick quotes' and the other users 'just single quotes'? | 18:42 |
| portante | if so, do you know if there is a way to control that behavior? | 18:43 |
| *** klrmn has quit IRC | 18:43 | |
| *** cebruns has joined #openstack-swift | 18:46 | |
| *** alenavolk has quit IRC | 18:59 | |
| *** alenavolk has joined #openstack-swift | 19:00 | |
| *** alenavolk has quit IRC | 19:05 | |
| *** alenavolk has joined #openstack-swift | 19:06 | |
| *** alenavolk has quit IRC | 19:08 | |
| *** mdrabe has joined #openstack-swift | 19:13 | |
| *** lucasxu has quit IRC | 19:13 | |
| *** geaaru has quit IRC | 19:13 | |
| *** oshritf has quit IRC | 19:31 | |
| *** oshritf has joined #openstack-swift | 19:36 | |
| portante | https://unix.stackexchange.com/questions/364004/cp-error-message-uses-different-quoting-on-ubuntu-12-04-and-fedora-24 | 19:41 |
| notmyname | portante: I have not seen that | 19:44 |
| notmyname | it's not a LOCALE thing? | 19:44 |
| *** pcaruana has joined #openstack-swift | 19:45 | |
| portante | notmyname: not that we have seen, which is what I would have thought | 19:45 |
| portante | kills our hacky unit tests | 19:46 |
| portante | :( | 19:46 |
| portante | I suspect when we find the answer we'll say, "Doh!" | 19:47 |
| notmyname | portante: also, ubuntu 12.04? I know you work at redhat and are used to supporting an OS for like 20 years or something, but canonical shot 12.04 in the head a while back I think ;-) | 19:54 |
| portante | yeah, travis runs that version of ubuntu for us. :( | 19:55 |
| clarkb | portante: iirc you can pick a different version on travis | 19:59 |
| clarkb | its some directive in your yaml file | 19:59 |
| timburke | interesting... i've got a 14.04 vm handy that defaults to cannot stat ‘foo’ (i.e., it uses http://www.fileformat.info/info/unicode/char/2018/index.htm and http://www.fileformat.info/info/unicode/char/2019/index.htm) but it responds well to LC_ALL=C | 19:59 |
| clayg | LC_ALL=C is *key* | 20:00 |
| *** pcaruana has quit IRC | 20:02 | |
| openstackgerrit | Merged openstack/swift master: Stop having Sphinx treat warnings as errors https://review.openstack.org/469314 | 20:12 |
| portante | clarkb: thanks, we'll try that | 20:14 |
| *** klrmn has joined #openstack-swift | 20:14 | |
| portante | clayg: thanks, we tried that and it does not seem to work | 20:14 |
| clarkb | clayg: re increasing the allowed header size, I am writing the commit message now. Is theer a short version of why this is required? is it just that swift is able to use so many bytes for metadata? | 20:18 |
| clarkb | and since the non tls'd jobs go straight to eventlet its a non issue there? | 20:18 |
| clayg | yeah | 20:19 |
| clayg | I'm realizing now it may have been possible to update the swift constraints to have a smaller limit on metadata... it's possible the tests already exercise the advertised limit and the problem was just the artificial limit in the middle rather than any specific requirement | 20:20 |
| *** mat128 has quit IRC | 20:20 | |
| clarkb | clayg: https://review.openstack.org/469642 | 20:20 |
| patchbot | patch 469642 - openstack-dev/devstack - Increase allowed header size with tls proxy | 20:20 |
| *** lucasxu has joined #openstack-swift | 20:20 | |
| *** rcernin has quit IRC | 20:21 | |
| *** lucasxu has quit IRC | 20:21 | |
| *** oshritf has quit IRC | 20:22 | |
| *** rcernin has joined #openstack-swift | 20:22 | |
| clayg | clarkb: the other thing that gets us much closer to working on lp bug #1667833 is something with /etc/swift/test/conf | 20:23 |
| openstack | Launchpad bug 1667833 in OpenStack Object Storage (swift) "func tests must not assert accept-ranges on 304" [Undecided,New] https://launchpad.net/bugs/1667833 | 20:23 |
| clayg | OMM functests with tls enabled fail spectacularly until I change `auth_ssl = no` / `auth_ssl = yes` | 20:24 |
| clayg | I see a `auth_uri = https://192.168.8.8:35357` option already in play - so maybe the best thing is to patch swift to use it? | 20:25 |
| clayg | notmyname: timburke: is there already a patch for that in progress? | 20:25 |
| *** joeljwright has joined #openstack-swift | 20:26 | |
| *** ChanServ sets mode: +v joeljwright | 20:26 | |
| clarkb | ya devstack should know what the auth_uri is for the entire cloud and you should be able to write that in as appropriate to the swift config | 20:31 |
| *** rcernin has quit IRC | 20:42 | |
| notmyname | swift team meeting in 15 minutes | 20:45 |
| *** skudlik has quit IRC | 20:46 | |
| *** m_kazuhiro has joined #openstack-swift | 20:52 | |
| openstackgerrit | Tim Burke proposed openstack/python-swiftclient master: Allow swift functional tests to use keystone auth uri https://review.openstack.org/456791 | 20:52 |
| *** chlong has quit IRC | 20:53 | |
| timburke | i think that^^ is what we want, though i'm fairly certain it'll require a swiftclient release to actually be useable in swift's gate. if we want the gate on tls *before* that, we should plumb auth_ssl into lib/swift alongside the auth_port plumbing around https://github.com/openstack-dev/devstack/blob/master/lib/swift#L533-L537 | 20:55 |
| timburke | i suppose i should have just gone ahead and done that when i did https://github.com/openstack-dev/devstack/commit/f43ea47 | 20:56 |
| kota_ | good morning | 20:58 |
| mattoliverau | Morning | 20:59 |
| timburke | things would be *even better* if we land something like https://review.openstack.org/#/c/443104/ -- then we don't need to mangle auth_url based on auth_version (auth_version could almost even *go away* at that point, except we need a way to know whether to use keystone or v1 auth) | 20:59 |
| patchbot | patch 443104 - python-swiftclient - Use generic keystone client instead of versioned one (ABANDONED) | 20:59 |
| notmyname | meeting time in #openstack-meeting | 21:00 |
| joeljwright | timburke: somehow we need to work towards better keystone handling | 21:01 |
| timburke | joeljwright: yup :-( | 21:01 |
| *** chlong has joined #openstack-swift | 21:05 | |
| clayg | mattoliverau: I met someone from Oz the other day (yesterday?)... shit I where were they from... Brisbane I think - I couldn't remember where *you* were living now - I said you used to live in Melbourne - am I making this up?! | 21:11 |
| *** pcaruana has joined #openstack-swift | 21:12 | |
| mattoliverau | Close enough, I was in Melbourne, I'm now at the beach in a place called Surf Beach on the south coast of NSW.. but I was in Melbourne so yeah that works ;) | 21:12 |
| *** skudlik has joined #openstack-swift | 21:12 | |
| tdasilva | mattoliverau: there will be a couple of awesome soccer matches in Melbourne in a couple of weeks ;) | 21:13 |
| mattoliverau | There is an OpenStack day Melbourne on today.. but couldn't convince work to pay for me to go :p | 21:14 |
| cschwede_ | timburke: looks like we have a gate blocker (the sphinx issue?) i saw your patch 469315 - but do we need an "override" for it? | 21:26 |
| patchbot | https://review.openstack.org/#/c/469315/ - swift - Make Sphinx treat warnings as errors (again) | 21:26 |
| timburke | cschwede_: nope; that's the follow-up to put it back. gate should be unblocked now, and i know there a re a few rechecks working their way through currently | 21:27 |
| cschwede_ | timburke: oh my, i'm too tired. "warning-is-error = 1" - yeah, read first, talk second. nevermind, thx :) | 21:28 |
| timburke | no worries :-) | 21:28 |
| *** pcaruana has quit IRC | 21:29 | |
| tdasilva | timburke, clayg: hi, timburke left a comment asking in patch 371150 should also apply for containers and accounts, and to be honest I don't remember why I put the check there, it might be just because the bug called out objects, but i'm not sure | 21:30 |
| patchbot | https://review.openstack.org/#/c/371150/ - swift - Return 404 on a GET if tombstone is newer | 21:30 |
| tdasilva | thoughts? | 21:31 |
| *** NM has quit IRC | 21:31 | |
| tdasilva | any concerns about doing the same for account/containers? | 21:31 |
| *** MVenesio has quit IRC | 21:31 | |
| *** MVenesio has joined #openstack-swift | 21:31 | |
| *** MVenesio has quit IRC | 21:32 | |
| openstackgerrit | Merged openstack/swift master: Add structure to storage policy configuration guide https://review.openstack.org/468360 | 21:36 |
| acoles | timburke: thanks for rechecking the ssync patch | 21:43 |
| *** jarbod_ has quit IRC | 21:47 | |
| *** jarbod_ has joined #openstack-swift | 21:47 | |
| tdasilva | timburke: fwiw, i'm leaning more towards keeping it as is for the reasos you just mentioned, I think operators should definetely get in there and fix things, otherwise they might lose data | 22:01 |
| timburke | speaking of deprecation and warnings, there's also patch 468065, patch 468077, and patch 468084 | 22:03 |
| patchbot | https://review.openstack.org/#/c/468065/ - swift - Log deprecation warnings for run_pause | 22:03 |
| patchbot | https://review.openstack.org/#/c/468077/ - swift - Log deprecation warnings for account_white/blacklist | 22:03 |
| patchbot | https://review.openstack.org/#/c/468084/ - swift - Log deprecation warning for allow_versions | 22:03 |
| clayg | notmyname: re-reading my comment on patch 468105 - I was hoping to trick someone into posting something to the os-ops ML regarding the coming change? And maybe also to solicit feedback from people that have in the past few releases maybe started to see this warning? Or are on an older version of swift and have a bad config - but with *no* warning? | 22:03 |
| patchbot | https://review.openstack.org/#/c/468105/ - swift - Require that known-bad EC schemes be deprecated | 22:03 |
| mattoliverau | K breakfast time o/ | 22:04 |
| timburke | 'cause i totally want to rip out some config options a la patch 468099 -- but i want to make sure we've been complaining for a while first | 22:05 |
| patchbot | https://review.openstack.org/#/c/468099/ - swift - Remove deprecated vm_test_mode option | 22:05 |
| clayg | timburke: joeljwright: can you think of *any* other use-case for wrapping segments in inline strings besides tar-archives? | 22:15 |
| clayg | I don't want to add a generic feature that's entirely use-case specific | 22:15 |
| *** klamath has quit IRC | 22:18 | |
| timburke | you could build large xml documents from component ones | 22:19 |
| clayg | timburke: ok nice - that's something! | 22:20 |
| clayg | {"preamble": "<div class='some outer'>", path="some/inner.html", "postamble": "</div>"} | 22:21 |
| timburke | could probably use it for various container formats -- though i don't know a whole lot about them. some of them probably demand more interleaving than you'd want to do with an slo -- better to just save the transcode as a separate object | 22:21 |
| timburke | yup | 22:22 |
| joeljwright | clayg: I can think of examples | 22:22 |
| joeljwright | but they're a bit more out there | 22:22 |
| joeljwright | we've considered building video files from audio/video segments with pre/postamble to build the container | 22:23 |
| joeljwright | but we've got no further than the idea | 22:23 |
| clayg | strictly speaking tho - it seems like support for inline segments *has* to support a superset of what you can do with pre/post-amble per segment. But perhaps significantly more complex for a client that wants to use inline segments to wrap individual segments with buffers? | 22:24 |
| clayg | i.e. [{'base64ineline': "preamble"}, {'path': "c/o2"}, {'base64inline': "postamble-preamble"}, {'path': 'c/o'}, {'base64inline': "postamble-footer"}] annoying if the who thing is to "wrap" each segment | 22:26 |
| clayg | and I guess if you have a big thing made of a bunch of little things - it's pretty common that the big thing is going to have little bits of data "around" the little things | 22:27 |
| clayg | so this implementation might have lowest impedance mismatch for known use-cases | 22:28 |
| * clayg shrugs | 22:28 | |
| clayg | do any other object storage api's have similar use-case support? s3 has http://docs.aws.amazon.com/AmazonS3/latest/dev/S3TorrentRetrieve.html | 22:29 |
| timburke | yeah, i keep going back and forth on which way i like more | 22:29 |
| clayg | I sorta feel like "can I download this SLO as a tar archive" is not such a weird feature to implement - or even "here's a list of segments - make me a link I can download as a tar archive" | 22:30 |
| timburke | i think torgomatic also had a use-case... something something proxyfs? ish? | 22:31 |
| clayg | why does the client have to tell us how to make it a tar archive... I guess it gives them more control - but do they *want* that much control? If we just do it - and then maybe a few feature requests for strings to pull out (think content-deposition header support for tempurls or whatever) - maybe that's good enough better? | 22:31 |
| *** mdrabe has quit IRC | 22:32 | |
| timburke | i totally expect that this is infrastructure so that we can have the auto-tar come next | 22:32 |
| clayg | ok, sorry I keep struggling to embrace the idea with first class API support - seems pretty esoteric - but the use-case seems *entirely* sensible - which makes me think "oh it should be *simpler* to turn a list of segments into a tar!?" | 22:33 |
| clayg | oic.... but if the goal is auto-tar... is first class pre/post-amble really a pre-req or just a half feature? | 22:34 |
| timburke | i thought we were all about pushing complexity to the client ;-) | 22:34 |
| joeljwright | :) I do want to explore more than tar using this too | 22:35 |
| joeljwright | but I see your point | 22:35 |
| clayg | timur: have you ever heard of this pre/post-amble SLO stuff? Can you think of any analog for turning some soft of manifest or collection of objects into another format? Or some sort of dynamic inline/padding bytes adorned to a response who's body is mostly data(s) from the object store? | 22:36 |
| *** m_kazuhiro has quit IRC | 22:36 | |
| timburke | joeljwright: how far has that gotten? in your explorations, has it felt natural to always attach the extra bytes to a segment, or were there cases where it felt forced, like the data didn't really belong "with" one segment or the other, but just needed to be *between* them? | 22:37 |
| clayg | timburke: I think you misunderstood - I want API's with clear affordances and don't mind clients that have to be explicit in their requests - if clients *want* to upload tars with pre/post-amble wrapping the objects - by all means - give them their cake! If they want to turn list of segments into tar archives - don't make them dance like monkies. | 22:37 |
| clayg | timburke: that's a good question | 22:38 |
| joeljwright | timburke: honestly I haven't had the chance to dig into that | 22:39 |
| joeljwright | I'll see if I can spend some time thinking about that before next week | 22:41 |
| joeljwright | but I have to go now | 22:41 |
| joeljwright | thanks for the help | 22:41 |
| joeljwright | :) | 22:41 |
| *** joeljwright has quit IRC | 22:42 | |
| clayg | timburke: what about instead of a json manfiest we just had a template of some kind? | 22:42 |
| timburke | clayg: yes! let's dep on jinja2! | 22:43 |
| clayg | read manifest, send bytes till {% SOME_OBJECT %}, send bytes in SOME_OBJECT, keep sending bytes | 22:43 |
| clayg | timburke: i mean maybe... I don't know if jinja2 supports a streaming semantic - i wouldn't want to have to render the whole template - and I don't know if jinja would be able to do sufficiently arbitrary byte strings to render a tar... | 22:44 |
| timburke | i was being difficult :P | 22:45 |
| clayg | i'm just recognizing that in the xml case for example the pre-post idea seems like a difficult needle to thread | 22:45 |
| clayg | then there's all this encoding to make the bytes we want work in json | 22:45 |
| notmyname | why do I switch back to my IRC client and see a serious(?) conversation about adding a dependency on templating engines? | 22:46 |
| timburke | we need *some* kind of encoding to differentiate between data and control channels | 22:46 |
| clayg | i just wish someone else already had an API we could draw inspiration from... redis has pre-comipled LUA scripts - you define your own aritity and shit, make up your own new functions "download_as_tar obj1, obj2, ... objN" | 22:46 |
| openstackgerrit | Merged openstack/swift master: Fix SSYNC failing to replicate unexpired object https://review.openstack.org/456921 | 22:49 |
| notmyname | whoo! | 22:50 |
| *** Sukhdev has joined #openstack-swift | 22:51 | |
| clayg | oh... that does sound nice? | 22:52 |
| openstackgerrit | Tim Burke proposed openstack/python-swiftclient master: Allow swift functional tests to use keystone auth uri https://review.openstack.org/456791 | 22:53 |
| clayg | clarkb: (just rubber ducking, but ...) my apache is failing to start with `file '/opt/stack/data/devstack-cert.pem' does not exist or is empty` i'm wondering if `tls-proxy` in my enable_service local.conf is no longer entirely sufficient to get all the ssl goodness... | 22:56 |
| clayg | :q | 22:58 |
| clarkb | clayg: it should be sufficient, there was a recent change to that code to handle python35 hopefully I didn't break something (it passed tests >_>) | 23:00 |
| *** chlong has quit IRC | 23:00 | |
| clarkb | no it passed tests which would exercise the non python35 and python35 paths so it should be fine | 23:01 |
| clarkb | clayg: was this on a fresh run? | 23:02 |
| clayg | it was not on a "fresh run in the gate" - no it was just in my environment - but i did just recreate the vm and run stack.sh with what I felt like was a reasonable configuration | 23:04 |
| clarkb | do you have that localrc? | 23:05 |
| clarkb | double checking gate job runs tls-proxy is all we set in ENABLED_SERVICES | 23:07 |
| clarkb | clayg: the code that runs make_cert only runs if the file does not exist, so maybe you had the file exist and its not valid for some reason? | 23:09 |
| clayg | yeah, it's strange code trace seems to indicate that's the only thing needed to create the file? I didn't see the error where it tried to create it an failed tho? http://paste.openstack.org/show/611146/ | 23:09 |
| clarkb | eg it is empty? | 23:09 |
| clayg | it doesn't exist now - perhaps it did? re-running stack.sh did not create the file. | 23:09 |
| openstackgerrit | Tim Burke proposed openstack/swift master: Add support to increase object ring partition power https://review.openstack.org/337297 | 23:11 |
| clarkb | clayg: https://git.openstack.org/cgit/openstack-dev/devstack/tree/stack.sh#n814 init_cert there is what writes that file | 23:12 |
| clarkb | and its just checking if tls-proxy is enabled | 23:12 |
| clarkb | you can likely look in the devstack log and see if is_service_enabled tls-proxy returned true or false as devstack should run udner set -x | 23:12 |
| clayg | my code tracing lead me to a similar conclusion - i'm trying to verify I'm maybe not getting that far? | 23:15 |
| clarkb | clayg: http://logs.openstack.org/89/443189/24/gate/gate-tempest-dsvm-neutron-full-ubuntu-xenial/15b8873/logs/devstacklog.txt.gz#_2017-05-31_20_20_43_159 is an example of where it does that in a gate job | 23:15 |
| clayg | oic, and here is where i blow up? http://logs.openstack.org/89/443189/24/gate/gate-tempest-dsvm-neutron-full-ubuntu-xenial/15b8873/logs/devstacklog.txt.gz#_2017-05-31_20_20_40_798 | 23:17 |
| clayg | so close! | 23:17 |
| clarkb | oh interesting | 23:18 |
| clayg | wonder how my tls-ness got in the config files ahead of when the apache's wanted to start in this case... maybe I didn't blow away the vm!? | 23:18 |
| clarkb | ya I'm guessing you have preexisting apache vhost configs floating around | 23:18 |
| clarkb | clayg: that is probably a good reason to move up the ca and cert stuff in the stack.sh script though | 23:19 |
| clayg | sure, yeah probably | 23:19 |
| clarkb | https://git.openstack.org/cgit/openstack-dev/devstack/tree/stack.sh#n784 to there I think | 23:19 |
| clarkb | maybe try move the block locally and see if that fixes it for you | 23:20 |
| *** chsc has quit IRC | 23:20 | |
| *** chsc has joined #openstack-swift | 23:22 | |
| *** chsc has joined #openstack-swift | 23:22 | |
| *** kei_yama has joined #openstack-swift | 23:30 | |
| *** chsc has quit IRC | 23:40 | |
| torgomatic | timburke: yeah, I wanted the pre/postamble for my own purposes, but if it doesn't happen, I'll deal | 23:53 |
| timburke | torgomatic: no no! clayg was just looking for more use-cases | 23:53 |
| torgomatic | timburke: fair enough. Having the pre/postambles (in whatever form it takes) in SegmentedIterable would make my life easier at some future point. | 23:54 |
| clayg | clarkb: ok, http://paste.openstack.org/show/611148/ womm fwiw | 23:59 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!