| *** dstanek has joined #openstack-infra | 00:01 | |
| *** mgagne has quit IRC | 00:06 | |
| *** jcooley_ has joined #openstack-infra | 00:13 | |
| *** dstanek has quit IRC | 00:14 | |
| *** senk has quit IRC | 00:17 | |
| *** jcooley_ has quit IRC | 00:19 | |
| *** senk has joined #openstack-infra | 00:21 | |
| *** rfolco has quit IRC | 00:23 | |
| *** rakhmerov has joined #openstack-infra | 00:23 | |
| *** adam_g has quit IRC | 00:24 | |
| *** adam_g has joined #openstack-infra | 00:24 | |
| *** adam_g has quit IRC | 00:24 | |
| *** adam_g has joined #openstack-infra | 00:24 | |
| *** rakhmerov has quit IRC | 00:27 | |
| *** zehicle_at_dell has quit IRC | 00:28 | |
| *** zehicle_at_dell has joined #openstack-infra | 00:29 | |
| *** zehicle has quit IRC | 00:29 | |
| *** zehicle has joined #openstack-infra | 00:30 | |
| *** sdake has quit IRC | 00:30 | |
| *** adam_g has quit IRC | 00:35 | |
| *** adam_g has joined #openstack-infra | 00:35 | |
| *** adam_g has quit IRC | 00:35 | |
| *** adam_g has joined #openstack-infra | 00:35 | |
| *** rongze has joined #openstack-infra | 00:42 | |
| *** rongze has quit IRC | 00:46 | |
| *** yaguang has joined #openstack-infra | 00:52 | |
| *** senk has quit IRC | 00:53 | |
| *** senk has joined #openstack-infra | 00:54 | |
| *** jcooley_ has joined #openstack-infra | 01:00 | |
| *** boris-42 has quit IRC | 01:05 | |
| *** yongli has joined #openstack-infra | 01:05 | |
| *** jcooley_ has quit IRC | 01:06 | |
| *** jcooley_ has joined #openstack-infra | 01:07 | |
| *** banix has joined #openstack-infra | 01:10 | |
| *** rongze has joined #openstack-infra | 01:15 | |
| *** senk has quit IRC | 01:16 | |
| *** banix has quit IRC | 01:18 | |
| *** banix has joined #openstack-infra | 01:20 | |
| *** senk has joined #openstack-infra | 01:22 | |
| *** praneshp has joined #openstack-infra | 01:22 | |
| *** senk has quit IRC | 01:23 | |
| *** rakhmerov has joined #openstack-infra | 01:24 | |
| *** rakhmerov has quit IRC | 01:28 | |
| *** jcooley_ has quit IRC | 01:31 | |
| *** rongze has quit IRC | 01:34 | |
| *** banix has quit IRC | 01:36 | |
| *** zehicle_at_dell has quit IRC | 01:37 | |
| *** zehicle has quit IRC | 01:37 | |
| *** rongze has joined #openstack-infra | 01:49 | |
| *** fifieldt has joined #openstack-infra | 01:54 | |
| *** dcramer_ has joined #openstack-infra | 02:01 | |
| *** senk has joined #openstack-infra | 02:11 | |
| *** fifieldt has quit IRC | 02:15 | |
| *** rakhmerov has joined #openstack-infra | 02:24 | |
| *** rakhmerov has quit IRC | 02:29 | |
| *** larrycai_ has joined #openstack-infra | 02:38 | |
| *** senk has quit IRC | 02:40 | |
| *** adam_g has quit IRC | 02:48 | |
| *** adam_g has joined #openstack-infra | 02:48 | |
| *** adam_g has quit IRC | 02:48 | |
| *** adam_g has joined #openstack-infra | 02:48 | |
| *** dguitarbite has joined #openstack-infra | 03:22 | |
| *** rakhmerov has joined #openstack-infra | 03:25 | |
| *** rakhmerov has quit IRC | 03:30 | |
| *** dguitarbite has quit IRC | 03:34 | |
| *** banix has joined #openstack-infra | 03:35 | |
| *** nati_ueno has joined #openstack-infra | 04:04 | |
| *** changbl has quit IRC | 04:05 | |
| *** changbl has joined #openstack-infra | 04:10 | |
| *** nati_ueno has quit IRC | 04:16 | |
| *** rakhmerov has joined #openstack-infra | 04:26 | |
| *** rakhmerov has quit IRC | 04:30 | |
| *** loq_mac has joined #openstack-infra | 04:39 | |
| *** julim has joined #openstack-infra | 04:46 | |
| *** rongze has quit IRC | 04:47 | |
| *** rongze has joined #openstack-infra | 04:48 | |
| *** rongze has quit IRC | 04:52 | |
| *** banix has quit IRC | 04:57 | |
| *** ArxCruz has quit IRC | 05:11 | |
| *** rongze has joined #openstack-infra | 05:18 | |
| *** rongze has quit IRC | 05:20 | |
| *** rongze_ has joined #openstack-infra | 05:20 | |
| *** adam_g has quit IRC | 05:21 | |
| *** adam_g has joined #openstack-infra | 05:21 | |
| *** adam_g has quit IRC | 05:21 | |
| *** adam_g has joined #openstack-infra | 05:21 | |
| *** adam_g has quit IRC | 05:24 | |
| *** adam_g has joined #openstack-infra | 05:24 | |
| *** adam_g has quit IRC | 05:24 | |
| *** adam_g has joined #openstack-infra | 05:24 | |
| *** fifieldt has joined #openstack-infra | 05:24 | |
| *** rongze_ has quit IRC | 05:25 | |
| *** rakhmerov has joined #openstack-infra | 05:27 | |
| *** rakhmerov has quit IRC | 05:29 | |
| *** rakhmerov has joined #openstack-infra | 05:29 | |
| *** chandankumar has joined #openstack-infra | 05:48 | |
| *** loq_mac has quit IRC | 05:55 | |
| clarkb | what if we deny push access to all users on refs/draft/* | 06:18 |
|---|---|---|
| *** rongze has joined #openstack-infra | 06:18 | |
| *** rakhmerov has quit IRC | 06:22 | |
| *** AaronGr_Zzz is now known as AaronGr | 06:22 | |
| *** rongze has quit IRC | 06:23 | |
| *** rongze has joined #openstack-infra | 06:28 | |
| *** rakhmerov has joined #openstack-infra | 06:35 | |
| *** jhesketh_ has joined #openstack-infra | 06:47 | |
| *** rakhmerov has quit IRC | 06:48 | |
| *** fifieldt has quit IRC | 06:49 | |
| *** coolsvap has joined #openstack-infra | 06:50 | |
| *** fifieldt has joined #openstack-infra | 06:50 | |
| *** zigo has joined #openstack-infra | 06:56 | |
| *** fifieldt has quit IRC | 06:58 | |
| *** garyk has joined #openstack-infra | 07:03 | |
| *** yfried has joined #openstack-infra | 07:04 | |
| *** yfried has left #openstack-infra | 07:04 | |
| *** fifieldt has joined #openstack-infra | 07:14 | |
| *** rcarrillocruz has joined #openstack-infra | 07:32 | |
| *** coolsvap has quit IRC | 07:37 | |
| *** coolsvap has joined #openstack-infra | 07:39 | |
| *** rongze has quit IRC | 07:45 | |
| *** rongze has joined #openstack-infra | 07:51 | |
| *** adam_g has quit IRC | 08:07 | |
| *** adam_g has joined #openstack-infra | 08:08 | |
| *** adam_g has quit IRC | 08:08 | |
| *** adam_g has joined #openstack-infra | 08:08 | |
| *** fbo_away is now known as fbo | 08:12 | |
| *** dguitarbite_ has joined #openstack-infra | 08:12 | |
| *** hashar has joined #openstack-infra | 08:13 | |
| *** praneshp_ has joined #openstack-infra | 08:16 | |
| *** praneshp has quit IRC | 08:19 | |
| *** praneshp_ is now known as praneshp | 08:19 | |
| *** zhiyan has joined #openstack-infra | 08:24 | |
| *** fifieldt has quit IRC | 08:29 | |
| *** rakhmerov has joined #openstack-infra | 08:29 | |
| *** coolsvap has quit IRC | 08:44 | |
| *** coolsvap has joined #openstack-infra | 08:53 | |
| *** zhiyan has quit IRC | 08:55 | |
| *** jpich has joined #openstack-infra | 09:00 | |
| *** adam_g has quit IRC | 09:02 | |
| *** adam_g has joined #openstack-infra | 09:03 | |
| *** adam_g has quit IRC | 09:03 | |
| *** adam_g has joined #openstack-infra | 09:03 | |
| *** rakhmerov has quit IRC | 09:03 | |
| *** talluri has joined #openstack-infra | 09:06 | |
| *** derekh has joined #openstack-infra | 09:10 | |
| *** larrycai_ has quit IRC | 09:30 | |
| *** lcestari has joined #openstack-infra | 09:42 | |
| *** praneshp has quit IRC | 09:45 | |
| *** coolsvap has quit IRC | 09:47 | |
| *** zhiyan has joined #openstack-infra | 09:58 | |
| *** rakhmerov has joined #openstack-infra | 10:00 | |
| *** coolsvap has joined #openstack-infra | 10:03 | |
| *** rakhmerov has quit IRC | 10:04 | |
| *** zhiyan has quit IRC | 10:09 | |
| *** jooools has joined #openstack-infra | 10:15 | |
| *** yaguang has quit IRC | 10:23 | |
| *** fifieldt has joined #openstack-infra | 10:27 | |
| *** dguitarbite_ has quit IRC | 10:43 | |
| *** fifieldt has quit IRC | 10:43 | |
| *** rongze_ has joined #openstack-infra | 10:43 | |
| *** rongze has quit IRC | 10:46 | |
| *** rongze_ has quit IRC | 10:48 | |
| *** boris-42 has joined #openstack-infra | 10:50 | |
| *** ArxCruz has joined #openstack-infra | 10:58 | |
| *** rakhmerov has joined #openstack-infra | 11:01 | |
| *** rakhmerov has quit IRC | 11:05 | |
| *** rongze has joined #openstack-infra | 11:14 | |
| *** boris-42 has quit IRC | 11:16 | |
| *** boris-42_ has joined #openstack-infra | 11:16 | |
| *** rongze_ has joined #openstack-infra | 11:18 | |
| *** rongze has quit IRC | 11:19 | |
| *** rongze_ has quit IRC | 11:22 | |
| *** dstanek has joined #openstack-infra | 11:28 | |
| *** jroovers has joined #openstack-infra | 11:30 | |
| *** boris-42_ is now known as boris-42 | 11:36 | |
| *** jorisroovers has joined #openstack-infra | 11:39 | |
| *** jroovers has quit IRC | 11:41 | |
| *** hashar has quit IRC | 11:44 | |
| *** rakhmerov has joined #openstack-infra | 12:02 | |
| *** rongze has joined #openstack-infra | 12:02 | |
| *** rakhmerov has quit IRC | 12:05 | |
| *** talluri has quit IRC | 12:17 | |
| *** CaptTofu has quit IRC | 12:31 | |
| *** CaptTofu has joined #openstack-infra | 12:31 | |
| *** rongze has quit IRC | 12:36 | |
| *** rongze has joined #openstack-infra | 12:37 | |
| *** rongze has quit IRC | 12:42 | |
| *** gfa has left #openstack-infra | 12:48 | |
| *** jhesketh_ has quit IRC | 12:48 | |
| *** jorisroovers has quit IRC | 12:49 | |
| *** fifieldt has joined #openstack-infra | 12:58 | |
| *** b3nt_pin` is now known as beagles | 12:59 | |
| *** beagles is now known as Guest7975 | 13:00 | |
| *** weshay has joined #openstack-infra | 13:01 | |
| sdague | morning folks, happy new year | 13:01 |
| *** CaptTofu has quit IRC | 13:02 | |
| *** coolsvap has quit IRC | 13:02 | |
| *** CaptTofu has joined #openstack-infra | 13:02 | |
| *** rakhmerov has joined #openstack-infra | 13:03 | |
| *** rongze has joined #openstack-infra | 13:03 | |
| *** smarcet has joined #openstack-infra | 13:04 | |
| *** rakhmerov has quit IRC | 13:07 | |
| *** Guest7975 has quit IRC | 13:08 | |
| fungi | sdague: what's so happy about it? means i have to remember to type "2014" everywhere now | 13:09 |
| fungi | major inconvenience | 13:09 |
| sdague | who types years? | 13:11 |
| *** b3nt_pin has joined #openstack-infra | 13:12 | |
| *** rfolco has joined #openstack-infra | 13:13 | |
| sdague | anyone know when jog0 is flying? | 13:17 |
| sdague | I was going through his grenade patches last night, and wanted to get together with him on them | 13:18 |
| *** xchu has joined #openstack-infra | 13:25 | |
| *** dcramer_ has quit IRC | 13:26 | |
| *** johnthetubaguy has joined #openstack-infra | 13:27 | |
| fungi | i don't recall whether he said when he was departing | 13:30 |
| *** fifieldt has quit IRC | 13:37 | |
| *** b3nt_pin is now known as beagles | 13:44 | |
| *** mriedem has joined #openstack-infra | 13:44 | |
| *** MIDENN_ has quit IRC | 13:46 | |
| *** hashar has joined #openstack-infra | 13:46 | |
| *** mdenny has joined #openstack-infra | 13:47 | |
| openstackgerrit | Thierry Carrez proposed a change to openstack-infra/devstack-gate: Add oslo.rootwrap to the devstack gate https://review.openstack.org/64675 | 13:48 |
| * fungi will be back shortly... | 13:56 | |
| *** CaptTofu has quit IRC | 14:00 | |
| *** CaptTofu has joined #openstack-infra | 14:00 | |
| openstackgerrit | Russell Bryant proposed a change to openstack-infra/config: zuul status: add a visual hint for change headers https://review.openstack.org/63809 | 14:01 |
| ArxCruz | ALL: Does anyone working on PPC cirros image ? | 14:01 |
| openstackgerrit | Thierry Carrez proposed a change to openstack-infra/config: Add oslo.rootwrap to the integrated gate https://review.openstack.org/64677 | 14:01 |
| *** rakhmerov has joined #openstack-infra | 14:04 | |
| sdague | ArxCruz: I don't think so | 14:05 |
| *** rakhmerov has quit IRC | 14:08 | |
| ttx | sdague: I pushed the various changes that you mentioned for oslo.rootwrap devstack-gate integration at: https://review.openstack.org/#/q/status:open+topic:rootwrap-gate,n,z | 14:10 |
| ttx | let me know if I missed one | 14:10 |
| *** dripton has joined #openstack-infra | 14:10 | |
| *** mfer has joined #openstack-infra | 14:11 | |
| *** tma996 has joined #openstack-infra | 14:12 | |
| sdague | cool, will check it out in a few | 14:13 |
| *** dprince has joined #openstack-infra | 14:16 | |
| *** ttx has quit IRC | 14:24 | |
| *** ttx has joined #openstack-infra | 14:25 | |
| *** ttx has quit IRC | 14:25 | |
| *** ttx has joined #openstack-infra | 14:25 | |
| *** xchu has quit IRC | 14:32 | |
| fungi | ArxCruz: you might be able to make a very stripped-down debian/powerpc image which wouldn't be much bigger than cirros | 14:32 |
| ArxCruz | fungi: the problem is when I ran tempest it ask for cirros image, and some tests fails because the image is x86 :/ | 14:33 |
| openstackgerrit | Thierry Carrez proposed a change to openstack-infra/config: Add logging to #openstack-relmgr-office https://review.openstack.org/64683 | 14:37 |
| *** eharney has joined #openstack-infra | 14:40 | |
| *** ruhe has joined #openstack-infra | 14:41 | |
| *** ryanpetrello has joined #openstack-infra | 14:43 | |
| *** rakhmerov has joined #openstack-infra | 14:43 | |
| maurosr | ArxCruz: you can specify which image you want to use in devstack | 14:44 |
| ArxCruz | maurosr: yeah, that's what I'm working right now | 14:44 |
| ArxCruz | :) | 14:44 |
| maurosr | ArxCruz: then you just need to set the url of the debian-ppc image that fungi mentioned | 14:44 |
| maurosr | ok | 14:44 |
| *** markmcclain has quit IRC | 14:54 | |
| fungi | ArxCruz: right, i suspect for a given processor architecture, you will either need an emulator hypervisor (qemu can emulate x86 on powerpc, right?) or an image of the correct arch | 14:55 |
| *** ruhe has quit IRC | 14:55 | |
| ArxCruz | fungi: yeah, qemu can't emulate x86 on ppc | 14:56 |
| fungi | ArxCruz: oh, too bad. i wonder if bochs would work | 14:59 |
| ArxCruz | bochs? | 15:00 |
| *** dkliban has joined #openstack-infra | 15:01 | |
| fungi | it's an x86 emulator i used to use "back in the day" (before qemu and friends existed) and still seems to be around | 15:01 |
| *** dcramer_ has joined #openstack-infra | 15:04 | |
| sdague | ArxCruz: or build your own cirros | 15:04 |
| ArxCruz | sdague: i know a few guys at IBM is trying to build one, but no success so far :/ | 15:06 |
| fungi | the benefit to bochs is that it's a full system emulator, not just a paravirt implementation | 15:07 |
| fungi | so no need to support syscall passthrough or anything like that, which means it has more of a chance of being usable on entirely separate platforms | 15:07 |
| *** Ajaeger has joined #openstack-infra | 15:08 | |
| fungi | though i've not tried to use it for network-accessible virtual machines (that was a very new concept at the time i was using it), so no idea whether it's usable for that | 15:08 |
| sdague | fungi: yeh, but then it would be *super* slow | 15:09 |
| sdague | 2nd level kvm guests executing full emulation | 15:10 |
| sdague | not so good for running tempest tests | 15:10 |
| ArxCruz | and I'm not sure if it has ppc64 support | 15:11 |
| ArxCruz | i mean, if there's binaries for ppc64 :) | 15:12 |
| annegentle_ | fungi: have a minute to assist in real-time with the signed tagging? | 15:12 |
| *** rakhmerov has quit IRC | 15:15 | |
| *** jpeeler has joined #openstack-infra | 15:16 | |
| *** jpeeler has quit IRC | 15:16 | |
| *** jpeeler has joined #openstack-infra | 15:16 | |
| *** senk has joined #openstack-infra | 15:16 | |
| *** dkehn_ is now known as dkehn | 15:17 | |
| *** croberts has joined #openstack-infra | 15:17 | |
| mfer | folks, my review for the config repo is failing jenkins now. the recommendation is to rebase. Do you rebase something that's already out there or should it actually be a merge? Once I share code I never rebase. This is why I ask. | 15:18 |
| Ajaeger | mfer: which review is it? | 15:19 |
| mfer | https://review.openstack.org/#/c/62069/ | 15:19 |
| mfer | this originally passed but infa isn't making new projects right so it's on hold | 15:19 |
| mfer | that's why it was code reviewed with X | 15:20 |
| Ajaeger | mfer: Download it and rebase it manually. The repository has been moved forward and your patch does not apply anymore - so, before it can go in, it might need some tweaking... | 15:21 |
| mfer | rebase or merge or doesn't matter? | 15:21 |
| Ajaeger | mfer: git checkout master;git pull;git review -d 62069; git rebase origin/master;git review - unless you get a conflict, then you need to resolve the conflicts | 15:21 |
| Ajaeger | mfer: git rebase, you want your patch forward ported to the tip of the tree | 15:22 |
| mfer | Ajaeger thanks. i don't usually rebase once the code is shared publically. that's why i ask. | 15:22 |
| *** coolsvap has joined #openstack-infra | 15:22 | |
| anteaya | mfer: good asking | 15:23 |
| anteaya | we don't merge, only rebase | 15:23 |
| anteaya | gerrit is the only system that merges | 15:23 |
| Ajaeger | mfer: There seems to have been changes that make it impossible for the patch to apply cleanly, therefore you need to rebase. | 15:23 |
| mfer | good to know. this was my first request so I'm just learning | 15:23 |
| anteaya | mfer: welcome aboard | 15:23 |
| anteaya | and thanks for asking, it saves time in the long run | 15:24 |
| Ajaeger | mfer: Welcome - and if there's more you need to know, just ask. | 15:24 |
| mfer | oh I will. Thanks :) | 15:24 |
| *** senk has quit IRC | 15:24 | |
| Ajaeger | anteaya: why does it say "rebase or merge" at https://wiki.openstack.org/wiki/GerritJenkinsGit#Gerrit_Merge_Problems ? | 15:24 |
| Ajaeger | I guess that should be only rebase... | 15:25 |
| *** markmcclain has joined #openstack-infra | 15:26 | |
| *** senk has joined #openstack-infra | 15:27 | |
| anteaya | good question, Ajaeger | 15:27 |
| anteaya | I think it should be only rebase | 15:27 |
| Ajaeger | Ok, I'll change it... | 15:28 |
| anteaya | let's hear from fungi and jeblair before we change | 15:28 |
| * Ajaeger waits ;) | 15:28 | |
| anteaya | ah okay, that works too | 15:28 |
| anteaya | if there is objection we can change it back | 15:28 |
| anteaya | I say this and I am going to be awol in about an hour | 15:28 |
| anteaya | so I'll leave changes to you, but I do think it should be only rebase | 15:29 |
| anteaya | fair enough? | 15:29 |
| Ajaeger | Ok, blame me for beiing to quick - reverts are cheap in the wiki ;) | 15:29 |
| anteaya | Ajaeger: my kind of contributor | 15:30 |
| anteaya | :D | 15:30 |
| fungi | annegentle_: sorry, back. i think my irc client has stopped raising a bell in my terminal multiplexer, so i keep responding late to people. should i check my e-mail, or are you just wanting to work through it in here? | 15:31 |
| Ajaeger | anteaya: Fixed and reworked the paragraph slightly. mfer could you check the description at https://wiki.openstack.org/wiki/GerritJenkinsGit#Gerrit_Merge_Problems to see whether it answers your problem? | 15:32 |
| *** rakhmerov has joined #openstack-infra | 15:33 | |
| *** rongze has quit IRC | 15:34 | |
| fungi | mfer: rebasing published branches in a normal git workflow is indeed often frowned upon, but for code review it works out better (though you still should avoid gratuitously rebasing when there are no merge conflicts) | 15:34 |
| mfer | Ajaeger it does. what would be useful is an example of how to go about the rebase here. I say that for new folks who aren't familiar with rebasing or all of git review. I went a year working with git before I rebased the first time. | 15:34 |
| mfer | fungi makes sense. thanks | 15:34 |
| anteaya | mfer: http://anteaya.info/blog/2013/03/25/please-rebase-your-change-and-upload-a-new-patchset/ | 15:34 |
| anteaya | feel free to use anything helpful | 15:35 |
| Ajaeger | mfer, search for rebase at https://wiki.openstack.org/wiki/GerritWorkflow | 15:35 |
| Ajaeger | mfer and if you could enhance the wiki with an example, it would be great. I guess anteaya and myself can review it as needed ;) | 15:35 |
| mfer | Ajaeger thanks. I was more thinking for future folks who are not familiar than for myself. | 15:36 |
| mfer | kk | 15:36 |
| fungi | Ajaeger: anteaya: the article is correct in stating that gerrit will ff *or* merge (it isn't suggesting that the contributor should do that manually, just stating what will potentially happen to the change when it comes time for it to get added officially to the target branch after gating) | 15:36 |
| anteaya | fungi: yes | 15:36 |
| anteaya | Ajaeger just removed the part in the wiki that looked to me like it was recomending contributors merge locally | 15:37 |
| anteaya | which I think is contra-indicated with our workflow | 15:37 |
| fungi | ahh, yes, i was still catching up on scrollback. i'll take a look at the dif | 15:37 |
| fungi | f | 15:37 |
| anteaya | k | 15:37 |
| *** prad has joined #openstack-infra | 15:37 | |
| fungi | Ajaeger: your new wording helps, thanks! | 15:38 |
| Ajaeger | fungi: thanks for the review! | 15:38 |
| fungi | though we really should mostly just point people to https://wiki.openstack.org/wiki/Gerrit_Workflow#Long-lived_Topic_Branches | 15:39 |
| fungi | that article is generally more visible in the onboarding documents, so that new contributors will hopefully see it first | 15:40 |
| fungi | so that's where we try to keep the development workflow recommendations (and the GerritJenkinsGit article is intended to be more behind-the-scenes explanations of stuff) | 15:40 |
| Ajaeger | fungi: The page I changed is mentioned in the gerrit message in case of conflicts. But let's add a link to the Workflow page. | 15:40 |
| fungi | ahh, good point. i had forgotten we linked to GerritJenkinsGit from the merge conflict message | 15:41 |
| fungi | we might ought to put up a change to gerrit's config to adjust that | 15:42 |
| Ajaeger | For now I put the link into the page. | 15:43 |
| fungi | awesome--thanks again | 15:43 |
| *** prad has quit IRC | 15:44 | |
| *** banix has joined #openstack-infra | 15:44 | |
| *** rongze has joined #openstack-infra | 15:44 | |
| *** yidclare has joined #openstack-infra | 15:45 | |
| jeblair | clarkb, fungi: let's merge https://review.openstack.org/#/c/63809/3 to help with the discoverability of the new zuul status page | 15:45 |
| fungi | oh, i see, the link in the review comment is just to the test failures section, not to the Gerrit_Merge_Problems section | 15:45 |
| jeblair | sdague: happy new year to you! :) | 15:45 |
| sdague | hey, so I kicked the hornet nest again on the gerrit upgrade on infra list | 15:46 |
| sdague | when you get a chance | 15:46 |
| jeblair | sdague: people were working on it through the holidays | 15:47 |
| sdague | cool | 15:47 |
| jeblair | sdague: please don't rush it, or if it's really important, please help out. i think it's very important that the upgrade be well tested and configured. it's a huge change, and with >1k users, we need to be careful about how we make changes like that. | 15:48 |
| fungi | sdague: yeah, at the moment we've run into some substantial changes in the underlying cla feature implementation which will require some special handling on upgrade | 15:49 |
| jeblair | fungi: perhaps we should stop dealing with the cla in gerrit. | 15:49 |
| Ajaeger | fungi: yeah, it points to the wrong place indeed ;( | 15:49 |
| openstackgerrit | A change was merged to openstack-infra/config: zuul status: add a visual hint for change headers https://review.openstack.org/63809 | 15:49 |
| jeblair | sdague: https://blueprints.launchpad.net/openstack-ci/+spec/gerrit-2.8-upgrade | 15:50 |
| jeblair | sdague: that links to an etherpad where we've been working through just some of the issues. | 15:50 |
| fungi | jeblair: well, we'd still need some means of denying contributions from non-cla-signers for certain projects, or agreement that we'll stop bothering with automated enforcement | 15:50 |
| annegentle_ | fungi: oh sorry now I'm back :) yes, I'd like help on irc if you can | 15:51 |
| fungi | annegentle_: sure, how far did you get? | 15:51 |
| jeblair | fungi: can you summarize the cla situation in the etherpad? | 15:51 |
| annegentle_ | fungi: I copied the .gnupg directory onto the cloud server | 15:51 |
| sdague | jeblair: ok, cool. so it's a little hard for me to see from that etherpad what's decided and done, and what's stilll a blocker | 15:51 |
| *** sdake has joined #openstack-infra | 15:52 | |
| *** sdake has quit IRC | 15:52 | |
| *** sdake has joined #openstack-infra | 15:52 | |
| fungi | jeblair: yeah, i'll add the details there in a moment | 15:52 |
| jeblair | sdague: almost nothing is done; so far the work has focused on actually getting it running and discovering what the problems are (eg, cla) | 15:52 |
| jeblair | s/done/complete/ | 15:53 |
| fungi | annegentle_: does gpg2 --list-keys return anything? | 15:53 |
| annegentle_ | fungi: yep, including what I think is the private key | 15:53 |
| annegentle_ | fungi: but then the git tag command gives back this: https://gist.github.com/anonymous/502126886b5f473c8cd0 | 15:54 |
| annegentle_ | fungi: did I need to do anything to tell git about the key? | 15:55 |
| fungi | annegentle_: does 'gpg2 --list-keys anne@openstack.org' return anything? | 15:57 |
| fungi | it might be that you didn't have a uid for that address on your key | 15:57 |
| *** prad has joined #openstack-infra | 15:58 | |
| sdague | jeblair: ok, fair. Which is why I threw the thread out there. | 15:58 |
| annegentle_ | fungi: yep, it's listing the one I expect it to | 16:02 |
| jeblair | sdague: responding | 16:03 |
| *** CaptTofu_ has joined #openstack-infra | 16:06 | |
| *** CaptTofu has quit IRC | 16:07 | |
| *** UtahDave has joined #openstack-infra | 16:08 | |
| *** thedodd has joined #openstack-infra | 16:08 | |
| fungi | annegentle_: interesting... well, you can tell git to use a specific key by running 'git config --global --add user.signingkey <your_key_id>' | 16:09 |
| fungi | where <your_key_id> is the last 16 or so digits of the fingerprint displayed by 'gpg2 --fingerprint anne@openstack.org' (omitting spaces) | 16:10 |
| *** jgrimm has joined #openstack-infra | 16:12 | |
| annegentle_ | fungi: ok sounds like a good thing to try | 16:12 |
| *** lcheng has joined #openstack-infra | 16:13 | |
| bknudson | I've got a change that failed pep8, but can't recreate it here... | 16:14 |
| bknudson | http://logs.openstack.org/74/64474/1/gate/gate-reviewday-pep8/db995bd/console.html#_2014-01-02_15_42_20_746 | 16:14 |
| bknudson | No distributions at all found for lazr.authentication (from lazr.restfulclient>=0.9.11->launchpadlib->reviewday==0.2.0) | 16:14 |
| NobodyCam | Good morning infra.. and Happy New Year | 16:15 |
| annegentle_ | fungi: ok that did it! | 16:16 |
| *** thedodd has quit IRC | 16:16 | |
| NobodyCam | i just got a change that failed to build python-mysql? | 16:16 |
| *** ^d has joined #openstack-infra | 16:17 | |
| *** ^d has joined #openstack-infra | 16:17 | |
| fungi | annegentle_: great! | 16:17 |
| annegentle_ | fungi: ok so one more issue, I can't seem to push to gerrit | 16:18 |
| annegentle_ | fungi: here's what I see https://gist.github.com/anonymous/365dfe0c41465a65fbde | 16:18 |
| Ajaeger | fungi: Is the acl setup correctly? | 16:19 |
| annegentle_ | NobodyCam: happy new year! | 16:19 |
| *** thedodd has joined #openstack-infra | 16:21 | |
| *** markmcclain has quit IRC | 16:21 | |
| fungi | bknudson: looks like maybe that run had trouble reading the links from https://pypi.python.org/simple/lazr.authentication/ (seems it's not hosted on pypi and instead links to lp for tarball downloads) | 16:21 |
| fungi | annegentle_: i suspect we haven't added you to that group yet (so Ajaeger was close on the cause). fixing | 16:22 |
| sdague | so 3 out of 4 changes in the gate are failing docs jobs | 16:23 |
| annegentle_ | fungi: thanks | 16:23 |
| *** blamar has joined #openstack-infra | 16:23 | |
| annegentle_ | sdague: let's ditch sphinx ;) | 16:23 |
| sdague | though the pattern seems hard to figure out. 2 failed on distribute | 16:23 |
| sdague | and 1 on pbr | 16:23 |
| bknudson | fungi: strangely, the pep8 job had worked just a couple of days ago. | 16:23 |
| *** jergerber has joined #openstack-infra | 16:23 | |
| fungi | annegentle_: looks like that was it. try again and see if it's any better. also, you can add/remove additional members of openstack-docs-ptl at https://review.openstack.org/#/admin/groups/260,members now as needed | 16:24 |
| sdague | annegentle_: well, they are also failing pep8 and unit tests | 16:24 |
| annegentle_ | ok whew fungi thanks! I was mostly just glad I still remember the passphrase ha ha | 16:24 |
| annegentle_ | sdague: yeah let's ditch pep8 next | 16:24 |
| fungi | annegentle_: ;) | 16:24 |
| sdague | docs is just the job I assume is least likely to be random fails | 16:24 |
| annegentle_ | sdague: yeah it's usually solid, and everyone switched their requirements, right? | 16:25 |
| *** coolsvap has quit IRC | 16:25 | |
| *** prad_ has joined #openstack-infra | 16:26 | |
| *** prad has quit IRC | 16:26 | |
| *** prad_ is now known as prad | 16:26 | |
| *** prad has quit IRC | 16:26 | |
| sdague | annegentle_: as far as I know | 16:27 |
| annegentle_ | fungi: Ajaeger: ok, tag 0.1 for openstack-doc-tools | 16:27 |
| fungi | bknudson: aha, i should read logs more carefully... "Some externally hosted files were ignored (use --allow-external lazr.authentication to allow)." | 16:28 |
| annegentle_ | sdague: I have a search now on Gerrit for all rst changes for all repos, and it seems to be working, and only one has even been proposed over break, and hasn't merged. So yeah random. | 16:28 |
| annegentle_ | (over break I mean last 2 weeks or so) | 16:28 |
| *** andreaf has joined #openstack-infra | 16:28 | |
| Ajaeger | congrats annegentle_ for the tag! | 16:29 |
| annegentle_ | Ajaeger: whoo hoo | 16:29 |
| Ajaeger | fungi: I don't see it yet at https://pypi.python.org/pypi/openstack-doc-tools/ - how long will it take? | 16:29 |
| * Ajaeger sees the tag in git | 16:29 | |
| sdague | well, I think this is something fundamental going on | 16:29 |
| Ajaeger | Oh, there it is! GREAT! | 16:29 |
| sdague | about distribute | 16:29 |
| fungi | bknudson: i'm thinking reviewday may need a tox.ini patch to add --allow-external on the pip command | 16:30 |
| Ajaeger | whoo hoo, annegentle_ ! | 16:30 |
| bknudson | fungi: when I ran pep8 locally, it worked... pip freeze shows "lazr.authentication==0.1.2" | 16:30 |
| annegentle_ | Ajaeger: wow that's fast, fungi and all y'all that's amazing | 16:30 |
| fungi | bknudson: your version of pip is probably old | 16:30 |
| annegentle_ | Some kinda software factory you're running here. | 16:30 |
| fungi | annegentle_: it's great when it finally works! | 16:30 |
| Ajaeger | annegentle_: if you now approve https://review.openstack.org/#/c/64561/, infra can approve https://review.openstack.org/#/c/64567/ | 16:31 |
| Ajaeger | (review and approve I mean) | 16:31 |
| Ajaeger | I'm removing the WIP... | 16:31 |
| fungi | bknudson: latest pip no longer follows external links by default (requirements listed in pypi but not hosted on pypi) | 16:31 |
| bknudson | fungi: maybe it works locally because of a local pip cache and previous tox runs with different config? | 16:32 |
| fungi | NobodyCam: i want to say there was a new release and we ninja-pinned it in openstack/requirements last week | 16:32 |
| bknudson | fungi: did pip change just in the last couple of days? | 16:32 |
| fungi | bknudson: quite possibly | 16:32 |
| * fungi checks | 16:33 | |
| *** senk has quit IRC | 16:33 | |
| sdague | fungi: I think we have some bad unit test nodes | 16:33 |
| fungi | sdague: have a log? | 16:33 |
| *** senk has joined #openstack-infra | 16:34 | |
| annegentle_ | Ajaeger: testing 64561 now | 16:34 |
| sdague | https://jenkins01.openstack.org/job/gate-nova-python27/14588/console | 16:34 |
| sdague | fungi: go look at http://status.openstack.org/zuul/ | 16:34 |
| sdague | there is a really odd pattern there | 16:34 |
| fungi | NobodyCam: nevermind--i thought we pinned something, but looking back at global-requirements.txt i guess not | 16:34 |
| sdague | which is something very systematic | 16:34 |
| Ajaeger | tox did not install the package properly, any ideas? http://paste.openstack.org/show/59642/ | 16:35 |
| NobodyCam | fungi: https://review.openstack.org/#/q/status:open+project:openstack/ironic,n,z note the first three reviews | 16:35 |
| *** hashar has quit IRC | 16:35 | |
| sdague | maybe not bad nodes | 16:35 |
| sdague | nova and cinder aren't passing any of docs, pep8, or unit tests | 16:36 |
| fungi | sdague: yeah, looks like something's trying to globally install pbr as part of the job (rather than into a virtualenv) seeing if i can narrow it down | 16:36 |
| devananda | ironic's failing all tests | 16:36 |
| devananda | looks like MySQL-python? | 16:37 |
| devananda | fungi: http://paste.openstack.org/show/59643/ | 16:38 |
| sdague | well there was a new MySQL-python release today | 16:38 |
| fungi | heh... /usr/lib/python2.7/distutils/dist.py:267: UserWarning: Unknown distribution option: 'entry_poimts' | 16:38 |
| sdague | so that's a good chance that it's the break | 16:38 |
| Ajaeger | fungi: any idea what could be wrong with the package http://paste.openstack.org/show/59642/ - or is there a better place to ask? | 16:39 |
| jeblair | fungi: where do you see poimts? | 16:39 |
| HenryG | git.openstack.org is not responding to my requests ? | 16:40 |
| fungi | jeblair: jsonpatch seems to have a broken setup.py, but i think it's unrelated to the issue | 16:41 |
| *** pcrews_ has joined #openstack-infra | 16:41 | |
| anteaya | HenryG: I can open it in a browser, what is the request you are trying? | 16:41 |
| jeblair | sdague, fungi: i don't see a new mysql-python on the mirror... | 16:42 |
| jeblair | sdague: also, 1.2.4 looks like it was released last yer | 16:42 |
| jeblair | year | 16:42 |
| HenryG | git remote update (on neutron). But it finally came through after about 3 or 4 minutes | 16:42 |
| fungi | jeblair: pip 1.5 was released today. i think it may not dtrt for us under tox? | 16:43 |
| sdague | jeblair: oh, right, wrong year :) | 16:43 |
| jeblair | heh | 16:43 |
| jeblair | HenryG: the git farm doesn't seem unusually loaded | 16:43 |
| fungi | dstufft: mordred: seems we're quite possibly suffering some major breakage with today's pip 1.5 release under tox (seems to be trying to globally install things rather than using a virtualenv?) | 16:44 |
| sdague | nice... | 16:44 |
| sdague | so why is it only breaking on nova derived projects? | 16:44 |
| fungi | sdague: perhaps system packages | 16:45 |
| jeblair | the errors i'm seeing are due to mysql-python | 16:45 |
| * anteaya tries git remote update on neutron | 16:45 | |
| fungi | sdague: er, sitepackages | 16:45 |
| jeblair | s/due to/triggered by/ | 16:45 |
| *** gyee_ has joined #openstack-infra | 16:45 | |
| anteaya | HenryG: there were some pauses, but finished by a count of 10 | 16:45 |
| jeblair | which is used by nova and ironic | 16:45 |
| anteaya | HenryG: might it have something to do with your proxies? | 16:46 |
| fungi | jeblair: the log sdague linked me to was pip trying to write to /usr/local/lib from the requirements installation phase of a unit test job | 16:46 |
| jeblair | fungi: which one is that? | 16:46 |
| fungi | jeblair: https://jenkins01.openstack.org/job/gate-nova-python27/14588/consoleFull | 16:46 |
| jeblair | fungi: https://jenkins01.openstack.org/job/gate-nova-pep8/17283/consoleFull is typical of the widespread failures seen currently on zuul | 16:46 |
| HenryG | anteaya: it seems back to normal now | 16:46 |
| * fungi looks | 16:46 | |
| anteaya | HenryG: very good | 16:46 |
| jeblair | fungi: i think the different jobs may be failing in different ways | 16:47 |
| jeblair | fungi: i think my link is typical of a pep8 failure, and sdague's is typical of a unit test failure | 16:47 |
| jeblair | possibly just due to installation order? | 16:48 |
| *** rongze has quit IRC | 16:48 | |
| sdague | jeblair: probably | 16:48 |
| sdague | that would definitely be a nasty change of default interaction between pip and tox though | 16:49 |
| anteaya | here's my cab | 16:49 |
| *** rongze has joined #openstack-infra | 16:49 | |
| sdague | anteaya: enjoy the flight | 16:49 |
| fungi | jeblair: the recursion there makes me wonder if pip 1.5 is tripping over mysql-python's direct declaration of distribute as a requirement in ways which 1.4.x did not | 16:49 |
| anteaya | sdague: thanks | 16:49 |
| jeblair | fungi: yeah; maybe we have multiple problems with pip 1.5 | 16:50 |
| fungi | jeblair: well, another issue with 1.5 was also just pointed out by bknudson (no longer following external links without adding an option to enable that behavior) | 16:50 |
| dstufft | fungi: 1.5 no longer hits external urls | 16:50 |
| fungi | not so much an issue as something we need to take into account for non-mirror-enforced projects | 16:51 |
| jeblair | *nod* | 16:51 |
| fungi | dstufft: the other issues we're seeing so far are pbr under tox is trying to write to /usr/local/lib when sitepackages=True, and hitting max recursion depth on mysql-python's attempt to install distribute as a requirement | 16:51 |
| dstufft | fungi: got a log by any chance? | 16:52 |
| fungi | dstufft: in fact i do | 16:52 |
| fungi | dstufft: https://jenkins01.openstack.org/job/gate-nova-pep8/17283/consoleFull is the mysql-python/distribute recursion | 16:53 |
| fungi | dstufft: https://jenkins01.openstack.org/job/gate-nova-python27/14588/consoleFull is the pbr trying to write to system directories behavior | 16:53 |
| *** tma996 has quit IRC | 16:55 | |
| fungi | jeblair: sdague: my concern is that our modus operandi for pinning misbehaving dependencies while we work through the issue won't be possible for pip, since we install is separately from normal dependencies | 16:55 |
| fungi | er, install it separately | 16:55 |
| sdague | fungi: agreed | 16:56 |
| sdague | this one we're going to need to leave broken until we have a real fix | 16:56 |
| bknudson | How do I get tox to use the new pip? I did "sudo pip install pip --upgrade" but "tox -r -e pep8" still shows "pip 1.4.1 from /home/bknudson/dev/reviewday/.tox/pep8/lib/python2.7/site-packages (python 2.7)" | 16:56 |
| *** zzelle has joined #openstack-infra | 16:56 | |
| fungi | bknudson: you need to upgrade it with './.tox/pep8/bin/pip install pip --upgrade' instead | 16:57 |
| russellb | fungi: may want to broadcast a status message with openstackstatus | 16:57 |
| fungi | russellb: thanks for the reminder--will do | 16:57 |
| fungi | #status alert gating is severely hampered while we attempt to sort out the impact of the pip 1.5 release... no ETA for solution yet | 16:58 |
| openstackstatus | NOTICE: gating is severely hampered while we attempt to sort out the impact of the pip 1.5 release... no ETA for solution yet | 16:58 |
| *** ChanServ changes topic to "gating is severely hampered while we attempt to sort out the impact of the pip 1.5 release... no ETA for solution yet" | 16:58 | |
| russellb | great thanks :) | 16:58 |
| fungi | bknudson: and then try running tox without -r (since that may cause it to recreate it with older pip) | 16:59 |
| bknudson | fungi: right, but then it doesn't try to reinstall the deps which is where it looks like the problem is. | 16:59 |
| dstufft | fungi: did you by any chance happen to upgrade virtualenv? | 17:00 |
| *** branen has joined #openstack-infra | 17:02 | |
| fungi | dstufft: i'll check | 17:02 |
| fungi | bknudson: oh, good point | 17:02 |
| jeblair | fungi: pip installs virtualenv via puppet, right? | 17:02 |
| fungi | jeblair: yes, in modules/openstack_project/manifests/base.pp | 17:04 |
| zaro | morning | 17:04 |
| fungi | dstufft: yes, pip freeze on our slaves reflect the same version as the latest virtualenv on pypi.python.org | 17:05 |
| dstufft | fungi: I'm pretty sure that the MySQL-python problem is a setuptools problem | 17:05 |
| fungi | dstufft: jeblair: also released toay | 17:05 |
| fungi | er, today | 17:05 |
| dstufft | and you're getting the latest setuptools because we updated setuptools inside of virtualenv | 17:06 |
| *** garyk has quit IRC | 17:06 | |
| dstufft | or rather it's a problem with MySQL-python trying to force distribute | 17:06 |
| dstufft | and the latest setuptools changed something that is breaking that | 17:06 |
| fungi | dstufft: makes sense. i think we'd already determined that mysql-python is broken for declaring a distribute requirement? | 17:06 |
| fungi | yeah, that | 17:06 |
| fungi | so maybe this will force their hand in ways which our pestering did not | 17:07 |
| fungi | #status alert gating is severely hampered while we attempt to sort out the impact of today's pip 1.5/virtualenv 1.11 releases... no ETA for solution yet | 17:07 |
| openstackstatus | NOTICE: gating is severely hampered while we attempt to sort out the impact of today's pip 1.5/virtualenv 1.11 releases... no ETA for solution yet | 17:07 |
| *** ChanServ changes topic to "gating is severely hampered while we attempt to sort out the impact of today's pip 1.5/virtualenv 1.11 releases... no ETA for solution yet" | 17:07 | |
| dstufft | fungi: the pbr error happens during the execution of pbr's setup.py | 17:11 |
| dstufft | fungi: so that's either a setuptools error, or due to the nature of pbr, it's an error within pbr itself | 17:11 |
| dstufft | possibly an error the way pbr calls pip | 17:11 |
| jeblair | dstufft: which error is this? (can you give a log link?) | 17:12 |
| dstufft | https://jenkins01.openstack.org/job/gate-nova-python27/14588/consoleFull | 17:12 |
| fungi | dstufft: i didn't mean to imply that pip is broken, just that the behavior changes will need to be sorted out | 17:12 |
| dstufft | fungi: oh I know :) | 17:12 |
| dstufft | Just trying to narrow things down | 17:12 |
| fungi | dstufft: appreciated | 17:13 |
| mordred | can someone point me at the pbr error? | 17:13 |
| *** ruhe has joined #openstack-infra | 17:13 | |
| jeblair | https://jenkins01.openstack.org/job/gate-nova-python27/14588/consoleFull | 17:13 |
| fungi | mordred: ^ | 17:13 |
| jeblair | so i was chatting with mordred on the phone; he thinks it's unlikely that we will convince the python-mysqldb maintainer to make the distribute change (he has tried) | 17:14 |
| fungi | mordred: so far i've only noticed that when pbr is being installed under tox with sitepackages=True, if that helps narrow it down | 17:14 |
| mordred | fungi: stellar | 17:14 |
| jeblair | and making a small fork of that to fix setup.py might not be a crazy idea | 17:14 |
| fungi | jeblair: mordred: well, at this point it very well may be that 'pip install mysql-python' has ceased working entirely with pip 1.5, so maybe they'll reconsider their position | 17:15 |
| *** lcheng has quit IRC | 17:15 | |
| mordred | that is correct - it flat does not work - just tested | 17:16 |
| mordred | 534 virtualenv testmysql | 17:17 |
| mordred | 535 source testmysql/bin/activate | 17:17 |
| mordred | 536 pip install mysql-python | 17:17 |
| mordred | fails with latest pip and latest virtualenv | 17:17 |
| *** rongze has quit IRC | 17:17 | |
| fungi | they are now down to either revisiting their insistence on that choice, or deciding to no longer support installation via latest pip | 17:17 |
| *** rongze has joined #openstack-infra | 17:18 | |
| *** rongze has quit IRC | 17:18 | |
| *** jaypipes has joined #openstack-infra | 17:18 | |
| dstufft | mordred: fungi locally it doesn't work with pip 1.4.1 && setuptools 2.0.2 either | 17:19 |
| dstufft | nor does it work with easy_install | 17:19 |
| *** jaypipes has quit IRC | 17:19 | |
| *** rongze has joined #openstack-infra | 17:19 | |
| zzelle | hello, i am not sure but it seems pbr use --index-url or --extra-index-url ? | 17:19 |
| fungi | wow, so basically they've simply alowed mysql-python to completely bitrot into uninstallability | 17:20 |
| dstufft | nor does it work with ``python setup.py install`` | 17:20 |
| zzelle | which is no more supported by pip in version 1.5 | 17:20 |
| dstufft | zzelle: hrm? | 17:20 |
| dstufft | --index-url and --extra-index-url are still supported in 1.5 | 17:21 |
| mordred | zzelle: and pbr only uses those in its integration tests | 17:21 |
| mordred | nothing in execution | 17:21 |
| zzelle | oups, need to sleep | 17:21 |
| zzelle | i read half words | 17:21 |
| fungi | zzelle: i think you read the release notes backwards... the entry which mentions those is saying --use-mirrors, -M, and --mirrors are no longer supported | 17:21 |
| mordred | dstufft, fungi: pbr installs into a clean venv with no problem with 1.5 - I have pbr installed globally, so I'm having a hard time testing with system-site-packages enabled :) | 17:22 |
| mordred | dstufft: "/usr/bin/python -c "import setuptools;__file__='/home/jenkins/workspace/gate-nova-python27/build/pbr/setup.py';exec(compile(open(__file__).read().replace('\r\n', '\n'), __file__, 'exec'))" install --single-version-externally-managed --record /tmp/tmp.FeFChvCkt3/pip-v6F8TN-record/install-record.txt" | 17:23 |
| mordred | dstufft: why is it using /usr/bin/python there? | 17:23 |
| mordred | this should be in a virtualenv | 17:23 |
| dstufft | mordred: oh hrm | 17:23 |
| dstufft | I missed that | 17:23 |
| *** krtaylor has quit IRC | 17:24 | |
| dstufft | mordred: it just uses sys.executable | 17:24 |
| *** markmcclain has joined #openstack-infra | 17:24 | |
| dstufft | trying to think why that'd be /usr/bin/python | 17:24 |
| mordred | dstufft: shouldn't that be $venv/bin/python within a virtualenv? | 17:24 |
| mordred | otherwise, it's going to be doing entirely the wrong thing | 17:24 |
| dstufft | mordred: yes | 17:25 |
| dstufft | and it is in my test venv | 17:25 |
| mordred | same here | 17:25 |
| mordred | fwiw, https://github.com/emonty/MySQLdb1 has two changes I made to the repo to 'fix' the distribute issues | 17:27 |
| *** rwsu has joined #openstack-infra | 17:27 | |
| clarkb | maybe tox and new virtualenv dont do the right thing re sys.executable above | 17:28 |
| *** ruhe has quit IRC | 17:28 | |
| dstufft | fungi: mordred are you able to repro the pbr issues locally at all? | 17:28 |
| mordred | dstufft: I have not reproed the pbr issues - however, my patches on MyQLdb1 do allow it to be installed with new pip | 17:28 |
| mordred | which I have just verified | 17:28 |
| sdague | fungi: if you have a moment - https://review.openstack.org/#/c/63889/ would be nice to land, to help the neutron team move forward on getting isolated tests working | 17:29 |
| fungi | dstufft: well, i can reproduce it on one of our test systems, so should be able to duplicate the same scenario on a similar virtual machine | 17:29 |
| fungi | dstufft: i'll see what i can work out | 17:30 |
| clarkb | sdague: fungi: I have approved | 17:31 |
| sdague | clarkb: great | 17:31 |
| sdague | thanks | 17:31 |
| clarkb | (don't know if the puppet stuff is sufficiently decoupled from tox to pass) | 17:31 |
| fungi | sdague: 63889 needs a rebase | 17:31 |
| fungi | sdague: layout.yaml was refactored/reorganized since that change was drafted | 17:32 |
| sdague | ok, let me do that | 17:32 |
| *** senk has quit IRC | 17:32 | |
| *** rakhmerov has quit IRC | 17:33 | |
| openstackgerrit | Sean Dague proposed a change to openstack-infra/config: Adds neutron-isolated to Tempest gate as experimental https://review.openstack.org/63889 | 17:33 |
| *** jpich has quit IRC | 17:33 | |
| mordred | https://github.com/farcepest/MySQLdb1/pull/44 | 17:34 |
| mordred | there is a pull request for it | 17:34 |
| mordred | however, if we get no traction with that, I still contend that, since it's an almost unmoving project, we could do a shallow fork, release it as python-mysqldb and change our stuff to consume that instead if we can't get andy to merge that and cut a new release | 17:35 |
| dstufft | lol | 17:35 |
| dstufft | Mysql-python and python-mysql | 17:35 |
| mordred | right? | 17:35 |
| mordred | well, python-mysqldb, since that's what the module is actually called | 17:36 |
| fungi | i guess the two would be runtime-compatible so distros could substitute the not-fork when running our software | 17:36 |
| mordred | yup | 17:37 |
| *** lcheng has joined #openstack-infra | 17:37 | |
| fungi | assuming all which gets altered is setup.* | 17:37 |
| mordred | I have NO INTEREST in touching the code | 17:37 |
| fungi | mordred: though SpamapS would apparently love some help maintaining the debian packages for mysql if you get time ;) | 17:38 |
| openstackgerrit | A change was merged to openstack-infra/config: Adds neutron-isolated to Tempest gate as experimental https://review.openstack.org/63889 | 17:38 |
| *** krotscheck has joined #openstack-infra | 17:39 | |
| *** DennyZhang has joined #openstack-infra | 17:39 | |
| SpamapS | yes please :) | 17:39 |
| SpamapS | A few people have stepped up of late. | 17:39 |
| SpamapS | Though I think we have a new set of secret oracle CVE's for 5.5 coming soon. :-/ | 17:40 |
| *** jaypipes has joined #openstack-infra | 17:40 | |
| fungi | aren't all of oracle's cves effectively secret (or at least so opaque as to be useless)? | 17:41 |
| mordred | https://github.com/emonty/MySQLdb1/tree/python-MySQLdb | 17:42 |
| mordred | there is a branch that we can cut a fork from, should we decide to go that route | 17:42 |
| *** rongze has quit IRC | 17:45 | |
| dstufft | mordred: hrm | 17:46 |
| mordred | I'd rather get upstream fixed though | 17:47 |
| dstufft | mordred: | 17:47 |
| dstufft | 2014-01-02 15:58:13.628 | py27 installdeps: -r/home/jenkins/workspace/gate-nova-python27/requirements.txt, -r/home/jenkins/workspace/gate-nova-python27/test-requirements.txt | 17:47 |
| dstufft | 2014-01-02 15:58:13.629 | WARNING:test command found but not installed in testenv | 17:47 |
| dstufft | 2014-01-02 15:58:13.629 | cmd: /usr/bin/pip | 17:47 |
| dstufft | 2014-01-02 15:58:13.629 | env: /home/jenkins/workspace/gate-nova-python27/.tox/py27 | 17:47 |
| dstufft | from the pbr log | 17:47 |
| dstufft | it looks like tox was executing the global pip | 17:47 |
| dstufft | because it didn't find a pip installed into the venv? | 17:47 |
| mordred | oh. because of --system-site-packages | 17:47 |
| zaro | fungi: would you be able to add me to gerrit admin group on review-dev? | 17:47 |
| *** rongze has joined #openstack-infra | 17:47 | |
| dstufft | mordred: this should be outside of that, this is pure $PATH I think | 17:48 |
| fungi | zaro: sure thing | 17:48 |
| mordred | dstufft: nah - I just had that problem locally - because of system-site-packages, and because my global env has latest pip | 17:48 |
| mordred | I can't install pip into the virtualenv | 17:48 |
| mordred | like, it used to automatically install one - but as best I can tell, it's letting system-site-packages allow it to not do that now | 17:49 |
| dstufft | mordred: ugh | 17:49 |
| chmouel | when I send a commit for review is the commit get pushed to github before it gets approved? | 17:49 |
| openstackgerrit | Matt Farina proposed a change to openstack-infra/config: New project request: PHP-Client https://review.openstack.org/62069 | 17:49 |
| mordred | which honestly breaks everything | 17:49 |
| mordred | so I think we can consider taht a but in virtualenv perhaps | 17:49 |
| dstufft | mordred: this was probably a regression because we switched virtualenv from sdist to wheel | 17:49 |
| fungi | zaro: i've added you just now | 17:49 |
| mordred | chmouel: the commit can be found on github by its sha | 17:49 |
| mordred | dstufft: yah. that makes sense | 17:49 |
| *** harlowja has joined #openstack-infra | 17:49 | |
| chmouel | mordred: ah yeah i just figured that, how's that work? | 17:50 |
| mordred | the wheel code paths are a little different | 17:50 |
| zaro | fungi: thanks. | 17:50 |
| chmouel | mordred: if it's not approved it will be orphas there? | 17:50 |
| chmouel | mordred: orphans | 17:50 |
| mordred | chmouel: yes. | 17:50 |
| mordred | chmouel: https://github.com/openstack/swift/commit/$sha | 17:50 |
| chmouel | mordred: yep figured from my example to https://github.com/openstack-dev/devstack/commit/00b434182e3c04976e03b94490359fa26e71ef69 | 17:51 |
| mordred | should show you any commit that's been pushed up to gerrit for swift, even if it's not merged | 17:51 |
| fungi | chmouel: or https://git.openstack.org/cgit/openstack/swift/commit/?id=$sha | 17:51 |
| chmouel | i just wasn't expecting that | 17:51 |
| mordred | fungi: I believe with 2.8 we can actually change the gitweb urls to cgit perhaps | 17:51 |
| mfer | folks, i have a half dev half infra question... in Go (and i'm looking at golang-client on stackforge) you can have remote projects. When doing that you reference the full uri for the import path and Go can fetch and manage it for you. For example, github.com/stackforge/golang-client. Is it ok to point the full import path (which will need to be in the codebase itself) to github? | 17:52 |
| chmouel | fungi: nice, it makes nicer for code review on large patch i think | 17:52 |
| clarkb | mordred: I think we may need to set the slash escaping settings on apache for cgit too (maybe) | 17:52 |
| mordred | mfer: if it's something that you ever expect to use inside of infra systems | 17:52 |
| dstufft | mordred: can I bother you to raise an issue for virtualenv? | 17:52 |
| mordred | mfer: you should reference git.o.o instead, because github goes down | 17:52 |
| mordred | dstufft: I will do that right now | 17:52 |
| mordred | mfer: however, it's a stackforge project, so honestly you can do github if you want to - just know that you'll hit intermittent testing issues because of network | 17:53 |
| *** derekh has quit IRC | 17:53 | |
| mordred | dstufft: https://github.com/pypa/virtualenv/issues/521 | 17:54 |
| mfer | mordred can't really use openstack testing infra right now it seems. I've already investigated that :/ | 17:55 |
| chmouel | jeblair: you are mentioning here https://review.openstack.org/#/c/62377/ that bug #124256 is a blocker for new projects on stackforge but this seems to go to a gnome bug | 17:55 |
| uvirtbot | Launchpad bug 124256 in gnome-app-install "gnome-app-install crashed with TypeError in show_description() (dup-of: 122914)" [Undecided,New] https://launchpad.net/bugs/124256 | 17:55 |
| uvirtbot | Launchpad bug 122914 in gnome-app-install "gnome-app-install crashed with TypeError in show_description()" [Undecided,Fix released] https://launchpad.net/bugs/122914 | 17:55 |
| mordred | jeblair: based on the above ^^ your suggestion on the phone of attempting to pin virtualenv at an earlier version until that gets fixed may be the right short-term choice | 17:55 |
| *** garyk has joined #openstack-infra | 17:57 | |
| *** DennyZha` has joined #openstack-infra | 17:57 | |
| jeblair | chmouel: sorry, https://bugs.launchpad.net/openstack-ci/+bug/1242569 | 17:57 |
| uvirtbot | Launchpad bug 1242569 in openstack-ci "manage-projects error on new project creation" [Critical,In progress] | 17:57 |
| fungi | mordred: yeah, so i can confirm that trying to tox -epy27 in a fresh nova clone is ending up with no pip installed at all in the virtualenv | 17:57 |
| clarkb | fungi: jeblair: if we can add a project or two we should be able to confirm 1242569 is actually fixed now | 17:58 |
| chmouel | jeblair: ok thanks, I updated the review | 17:58 |
| clarkb | fungi: mordred: the other option (not really easy) is to use cloud archive libvirt and libvirt-python | 17:59 |
| *** DennyZhang has quit IRC | 17:59 | |
| clarkb | and stop using system packages completely | 17:59 |
| mfer | mordred what i really needed was a holiday break. i just had a realization around my problem with it on the git infra! yayz | 17:59 |
| mordred | clarkb: well, hrm. that is an option, and it is a place we want to be eventually | 17:59 |
| mordred | I think all three need to happen | 18:00 |
| fungi | mordred: also, as for --allow-external, adding that to the install_command in tox.ini doesn't seem to be causing it to get picked up, which i think we're going to need to get working if we expect people to be able to run tests under tox on their own systems without our mirror | 18:00 |
| mordred | fungi: wow, really? | 18:00 |
| mordred | install_command should be used verbatim | 18:01 |
| fungi | i'm trying again with tox -r now just to make sure | 18:01 |
| *** DennyZha` has quit IRC | 18:01 | |
| zul | hmmm...cloud archive? | 18:01 |
| fungi | mordred: still seems to not work | 18:01 |
| fungi | install_command = pip install --allow-external -U {opts} {packages} | 18:01 |
| mordred | dstufft: that's the right incantation right ^^ ? | 18:01 |
| fungi | [...] | 18:02 |
| fungi | Some externally hosted files were ignored (use --allow-external to allow). | 18:02 |
| dstufft | --allow-extenral PACKAGENAME | 18:02 |
| mordred | oh wow | 18:02 |
| dstufft | there's also an --allow-all-external | 18:02 |
| dstufft | I think it's called | 18:02 |
| mordred | ah - we shoudl start with that for now, unless we know all of the external packages we need to whitelist | 18:02 |
| fungi | aha! | 18:02 |
| dstufft | mordred: you'll probably need --allow-unverified too | 18:03 |
| fungi | yeah, trying | 18:03 |
| dstufft | which doesn't have an -all- variant | 18:03 |
| dstufft | because it's a security sensitive switch | 18:03 |
| mordred | I am in support of these switches in general - it looks like we might need to just list each package both ways explicitly rather than with -all for ease of tracking | 18:04 |
| mordred | do we get a list of which external packages were skipped? | 18:04 |
| fungi | mordred: not without iteration. it bails on the first failure | 18:04 |
| fungi | i'll work up a list for nova | 18:04 |
| mordred | fungi: sweet. thanks. I believe there should only be four, based on last time we looked in to this | 18:05 |
| mordred | dstufft: with 1.5, wheels are on by default, right? | 18:05 |
| Alex_Gaynor | We added a new one in at some point recently | 18:05 |
| Alex_Gaynor | mordred: yes | 18:05 |
| clarkb | we shouldn't need any external right? the mirror protects us | 18:06 |
| clarkb | at least for projects that only use the mirror | 18:06 |
| clarkb | oh but then everyone would have to use the mirror... | 18:06 |
| mordred | right | 18:06 |
| mordred | we don't need the list for the gate, but we'll need it for devs | 18:07 |
| mordred | OH | 18:07 |
| mordred | we'll need to update the mirror script with a whitelist | 18:07 |
| mordred | so that it doesn't bomb out trying to fetch externally hosted deps | 18:07 |
| fungi | yep | 18:07 |
| *** svarnau has joined #openstack-infra | 18:07 | |
| zaro | fungi: how is contrib agreement setup in gerrit 2.4? | 18:08 |
| mordred | https://etherpad.openstack.org/p/pip1.5Upgrade | 18:08 |
| mordred | I'm starting a list | 18:08 |
| zaro | fungi: just noticed that it's now in the project.config in ver 2.8 | 18:08 |
| dstufft | mordred: sorry for that :) | 18:09 |
| *** rongze has quit IRC | 18:09 | |
| fungi | zaro: see the etherpad--i've already updated it with the issues i identified for the cla a couple days ago | 18:09 |
| mordred | dstufft: no worries! I'm just glad the pbr error wasn't a pbr error :) | 18:09 |
| mordred | jeblair, fungi: I think I've captured each of the things we've talked about there, yeah? have I missed one? | 18:12 |
| mordred | zul: we were discussing cloud archive as a way to potentially accelarate getting a new enough libvirt that we can start using the new python-libvirt from pypi and stop doing system-site-packages in our tox.ini configs | 18:13 |
| jeblair | mordred: i missed something; what did you determine the cause of the "pbr error that wasn't a pbr error" to be? | 18:13 |
| fungi | mordred: what did the workaround for "error: could not create '/usr/local/lib/python2.7/dist-packages/pbr': Permission denied" end up being? | 18:13 |
| mordred | jeblair: bug in virtualenv | 18:13 |
| mordred | jeblair: (bullet 1) | 18:14 |
| zaro | fungi: ok. that looks good. i think one that additional thing that needs to be done is to add the a 'contributor agreement' section to project.config because the upgrade does not do that for you. | 18:14 |
| zaro | will add my notes as well. | 18:14 |
| jeblair | mordred: got it, thx; fungi i think that's the answer to your question too, right? | 18:14 |
| fungi | zaro: the upgrade did in fact add that section based on the db contents. i checked | 18:14 |
| zul | mordred: ah...well trusty has 1.2 and will have libvirt-python packaged (just maing its way to the archive) | 18:15 |
| zul | that reminds me devstack will be needed to be updated as well | 18:15 |
| fungi | jeblair: possibly. though downgrading virtualenv may not allow me to effectively iterate through the whitelist i need to build | 18:15 |
| mordred | jeblair: re: pinning virtualenv - I think that'll require manual effort on every slave | 18:15 |
| jeblair | mordred, fungi: virtualenv is installed by puppet via pip; how do we downgrade/pin that? | 18:16 |
| fungi | so i may resort to hand-patching the system installed virtualenv | 18:16 |
| fungi | (on my local test system | 18:16 |
| fungi | ) | 18:16 |
| zaro | fungi: in all-projects ? i don't see it. | 18:16 |
| jeblair | mordred: does the pip provider in puppet accept a version argument? | 18:16 |
| fungi | zaro: git clone /home/gerrit2/review_site/git/All-Projects.git && cd All-Projects && git fetch origin refs/meta/config && git checkout FETCH_HEAD | 18:16 |
| mordred | jeblair: I'm not sure there is a way to tell puppet to downgrade an installed pip package. I believe that we can express it in the puppet as "installed" rather than "latest" to keep puppet from borking us | 18:16 |
| mordred | jeblair: I _think_ it does - but I'm not sure I've ever tested whether it is smart enough to downgrade | 18:17 |
| AaronGr | mordred: yes, put a version in place of "latest" | 18:17 |
| zaro | fungi: ohh i'm just looking at the gui. so i guess it doesn't appear on the gui. | 18:17 |
| AaronGr | mordred: (provided you're using the system's package manager to install it) | 18:17 |
| mordred | AaronGr: we're not - we're using pip | 18:18 |
| *** zzelle has left #openstack-infra | 18:18 | |
| jeblair | mordred: okay, so manually downgrade on >50 machines, and either (a) set to installed, or (b) set to version [preferred if it works because it allows us to spin up new slaves with the correct config] | 18:18 |
| jeblair | fungi: what's the state of salt; can we execute the downgrade command on all the slaves using it? | 18:19 |
| mordred | jeblair: ++ | 18:19 |
| fungi | jeblair: i believe it should work, yes | 18:19 |
| * fungi checks to see whether it's remained sane since last look | 18:19 | |
| jeblair | mordred: what's the version we need? | 18:20 |
| clarkb | ugh run_tests.sh | 18:20 |
| *** johnthetubaguy has quit IRC | 18:20 | |
| clarkb | mordred: jeblair AaronGr putting a version in with the pip provider should work | 18:21 |
| mordred | jeblair: < 1.11 | 18:21 |
| mordred | jeblair: 1.10.1 I believe is what we were using before this upgrade | 18:21 |
| *** markmcclain has quit IRC | 18:21 | |
| clarkb | mordred: why wouldn't it? | 18:21 |
| clarkb | this is actually a fun side effect that pip --upgrade will also downgrade | 18:21 |
| mordred | oh - will it? | 18:22 |
| clarkb | pretty sure it will | 18:22 |
| jeblair | patching puppet now | 18:22 |
| mordred | yes. I verify that -U downgrades | 18:22 |
| clarkb | if pip --upgrade virtualenv==1.10.1 did not do the right thing it wouldn't work but it does do the correct thing | 18:22 |
| openstackgerrit | James E. Blair proposed a change to openstack-infra/config: Pin virtalenv to 1.10.1 https://review.openstack.org/64709 | 18:22 |
| *** hogepodge has joined #openstack-infra | 18:23 | |
| jeblair | clarkb, mordred, AaronGr: ^ | 18:23 |
| mordred | ok. that gets us past that problem | 18:23 |
| mordred | what do we want to do about mysql? | 18:23 |
| * mordred needs to get on the road, will have internet while in car - but would love to do higher bandwidth on this particular point | 18:24 | |
| clarkb | mordred: we could just switch to supporting only postgres | 18:24 |
| clarkb | (sorry couldn't resist) | 18:24 |
| mordred | heh | 18:24 |
| mordred | fungi is working on 3 - 4 is longer term | 18:25 |
| mordred | dhellmann: any chance you know/have pull with Andy? | 18:25 |
| fungi | this seems to work... sudo salt precise10.slave.openstack.org cmd.run 'pip install -U virtualenv\<1.11' | 18:26 |
| mordred | fungi: great. I _think_ the puppet change will just handle it - but good to know that's a good backup if it doesn't | 18:26 |
| fungi | virtualenv==1.10.1 | 18:26 |
| fungi | mordred: all the slaves seem to be responding as functional salt minions still, so s/precise10/*/ there ought to blast it everywhere | 18:27 |
| mordred | cool | 18:27 |
| dhellmann | mordred: sorry, I don't think so | 18:28 |
| *** mdenny has quit IRC | 18:28 | |
| *** david-lyle has quit IRC | 18:28 | |
| openstackgerrit | A change was merged to openstack-infra/config: Pin virtalenv to 1.10.1 https://review.openstack.org/64709 | 18:28 |
| dhellmann | mordred: oh, wait, I do think I know him | 18:29 |
| *** mdenny has joined #openstack-infra | 18:29 | |
| *** branen_ has joined #openstack-infra | 18:29 | |
| *** david-lyle has joined #openstack-infra | 18:29 | |
| dhellmann | mordred: I'll email him | 18:29 |
| mordred | jeblair: ok. I've done the python setup.py register for python-MySQLdb and have given openstackci full access to the account. I have uploaded the 'fixed' release. I went ahead and did that in case you guys decide to move forward with that plan and I'm unreachable | 18:31 |
| mordred | my preference is that we delete that and fix things - but I'm about to be driving through the desert | 18:32 |
| fungi | mordred: yeah, so trying to work around the virtualenv issue, because pip 1.4.1 doesn't grok --allow-unverified i can't effectively work out a whitelist purely by running tests under tox (it seems to refuse to use the virtualenv's pip even if i install one into it) | 18:32 |
| clarkb | mordred: when do you get on a plane for lca? | 18:32 |
| *** branen has quit IRC | 18:32 | |
| mordred | clarkb: not until Monday - I had a thing come up | 18:32 |
| jeblair | mordred: yeah, let's not bother importing it into gerrit; this shouldn't be maintained in the long run. | 18:33 |
| mordred | fungi: how about make a virtualenv without system-site-packages, and then just do pip install -U -r requirements.txt by hand into it? | 18:33 |
| mordred | jeblair: kk | 18:33 |
| *** rakhmerov has joined #openstack-infra | 18:33 | |
| mordred | jeblair: I did also tag the commit I used as 1.2.4 | 18:33 |
| jaypipes | clarkb: all of that makes sense, only the failed import of pyudev is also present when using tox on a fresh checkout of Neutron.... | 18:33 |
| jeblair | mordred: so now we propose a change to requirements to switch to it, yeah? | 18:33 |
| clarkb | jaypipes: really? it isn't for me and that is the only difference other than possibly a distro thing | 18:33 |
| fungi | mordred: well, trying to iterate on nova which will need sitepackages=True i think, but maybe i don't care since i just want to see deps install | 18:34 |
| jaypipes | clarkb: so I don't think the setup_hook is working as you state... I'm not sure what is different b/w the gate and my machine :( | 18:34 |
| clarkb | jaypipes: that is a possibility | 18:34 |
| jaypipes | clarkb: I will try and do some more digging. | 18:34 |
| jeblair | mordred: do you want to go aheand and do that, or are you signing off now? | 18:34 |
| clarkb | it does work on my test machine and in the gate. both sets of machines are ubuntu 12.04 | 18:34 |
| fungi | mordred: also, i've confirmed that virtualenv 1.10.1 actually *downgrades* the virtualenv's pip to 1.4.1 even if i install 1.5 in it by hand | 18:35 |
| mordred | jeblair: yeah. which is going to be tricky, because we're going to need to fix the mirror script, probably, due to the verified things | 18:35 |
| jaypipes | clarkb: yes, I'm on 13.04 as you know... not sure why/if that has anything to do with it.. | 18:35 |
| mordred | fungi: yah. you'llk need to do $venv/bin/pip install -U pip | 18:35 |
| *** praneshp has joined #openstack-infra | 18:35 | |
| *** sarob has joined #openstack-infra | 18:35 | |
| fungi | mordred: i did. running tox after that re-downgrades it | 18:35 |
| mordred | fungi: oh, right. I'm saying skip tox | 18:35 |
| jeblair | mordred, fungi, clarkb, dstufft, Alex_Gaynor: one of us should probably summarize the python-mysqldb situation to the mailing list; do any of you want to do that, or shall i? | 18:36 |
| fungi | ahh, yeah, i mainly wanted to confirm i had the tox.ini correct, but maybe i just wing that part | 18:36 |
| mordred | jeblair: I can take a stab at it real quick | 18:36 |
| jeblair | mordred: ack, thx. | 18:36 |
| *** markmcclain has joined #openstack-infra | 18:37 | |
| *** rakhmerov has quit IRC | 18:38 | |
| fungi | i'm being hounded for the past two hours to take a break to go out to lunch, so i'm going to need to pick the --allow-unverified testing back up in an hour or so. i think it's unlikely to matter for real until we have a fixed virtualenv anyway | 18:38 |
| *** ryanpetrello has quit IRC | 18:38 | |
| *** chandankumar has quit IRC | 18:38 | |
| fungi | (a fixed new virtualenv with pip 1.5 embedded i mean) | 18:38 |
| *** nati_ueno has joined #openstack-infra | 18:39 | |
| clarkb | sdague: https://review.openstack.org/#/c/64321/ for when you get back from lunch | 18:40 |
| *** rongze has joined #openstack-infra | 18:40 | |
| *** hogepodge has quit IRC | 18:40 | |
| * fungi heads out for a bit | 18:40 | |
| *** jroovers has joined #openstack-infra | 18:41 | |
| *** rongze has quit IRC | 18:45 | |
| *** hogepodge_ has joined #openstack-infra | 18:45 | |
| mordred | jeblair: https://etherpad.openstack.org/p/WMxbH5GK6Z | 18:46 |
| mordred | jeblair: check me on that - also, we should probably send out an email about the virtualenv issue | 18:46 |
| mordred | just for informational purposes | 18:46 |
| openstackgerrit | A change was merged to openstack-infra/zuul: Add zuul version to status.json https://review.openstack.org/64209 | 18:47 |
| *** nati_ueno has quit IRC | 18:47 | |
| jeblair | mordred: lgtm | 18:47 |
| mordred | jeblair: sent | 18:48 |
| *** jroovers has quit IRC | 18:48 | |
| jaypipes | guh, clarkb and now I can't reproduce the issue locally after uninstalling python-pyudev and rm -rf'ing my .venv and .tox dirs... :( | 18:49 |
| mordred | ok. I need to run - I'll get back online as soon as we're moving in the car | 18:49 |
| *** nati_ueno has joined #openstack-infra | 18:49 | |
| *** crank has joined #openstack-infra | 18:49 | |
| *** jerryz has joined #openstack-infra | 18:50 | |
| *** nati_ueno has quit IRC | 18:54 | |
| *** praneshp has quit IRC | 18:55 | |
| *** vipul is now known as vipul-away | 18:55 | |
| *** vipul-away is now known as vipul | 18:55 | |
| *** reed has joined #openstack-infra | 18:56 | |
| clarkb | jaypipes: so it works now? weird | 18:57 |
| *** thedodd has quit IRC | 18:58 | |
| *** hogepodge_ has quit IRC | 18:59 | |
| *** sileht has quit IRC | 18:59 | |
| *** praneshp has joined #openstack-infra | 19:00 | |
| jaypipes | clarkb: well, tests are still running, which is a good sign. :) | 19:02 |
| jaypipes | clarkb: I just have no idea why it works now and didn't earlier. | 19:02 |
| jaypipes | clarkb: I'll run a non-tox test after this and see if it's isolated to the venv tester. | 19:02 |
| *** rongze has joined #openstack-infra | 19:03 | |
| *** markmcclain has quit IRC | 19:04 | |
| clarkb | jaypipes: ok let me know how that goes | 19:04 |
| jaypipes | will do! | 19:04 |
| *** jooools has quit IRC | 19:04 | |
| *** hogepodge has joined #openstack-infra | 19:04 | |
| *** markmcclain has joined #openstack-infra | 19:05 | |
| *** sileht has joined #openstack-infra | 19:05 | |
| sdague | clarkb: so that seems fine to me, though have we actually used the crm114 data to create any finger prints yet? | 19:06 |
| sdague | I'd like to know it was useful before overselling the idea | 19:06 |
| clarkb | sdague: I am not sure we have. It does pick up on stuff though. search error_pr:[-1000.0 TO -10.0] | 19:07 |
| vishy | Is this a known issue? The required version of distribute (>=0.6.28) is not available, | 19:08 |
| vishy | trying to install mySQL-python | 19:08 |
| clarkb | vishy: yup, mordred sent mail to the list about it | 19:08 |
| clarkb | and see topic | 19:08 |
| vishy | ok cool thx | 19:08 |
| vishy | looks like it got merged | 19:10 |
| jeblair | vishy: neat. now we just need a release. :) | 19:11 |
| vishy | :) | 19:11 |
| *** rongze has quit IRC | 19:13 | |
| sdague | clarkb: yeh, so a browse of what's coming back looks like we're going to need some sort of tool to sift those results before they become useful. | 19:14 |
| sdague | so I would couch that a bit more | 19:14 |
| clarkb | sdague: really? those results are pretty reliable from a this is different perspective | 19:15 |
| clarkb | sdague: the fact the logs are noisy is a different problem :) | 19:15 |
| *** markmcclain has quit IRC | 19:15 | |
| jeblair | clarkb: i like your idea of doing something like 's/[0-9a-f]+/TOKEN/', btw; did you say you planned on writing that? | 19:15 |
| clarkb | jeblair: I didn't but I am happy to take a stab at it | 19:16 |
| jeblair | clarkb: i think it will make it more 'meaningful'; and after all, that's pretty much how we read the logs anyway. | 19:17 |
| *** markmcclain has joined #openstack-infra | 19:17 | |
| clarkb | yup | 19:17 |
| *** CaptTofu_ has quit IRC | 19:18 | |
| *** krtaylor has joined #openstack-infra | 19:18 | |
| clarkb | sdague: I can see how it doesn't give you the magical fingerprint but it does point you in the right direction imo. For example the current situation, log lines for isntalls pop up with low error_pr indicating a problem with package installs | 19:18 |
| *** CaptTofu has joined #openstack-infra | 19:18 | |
| sdague | clarkb: so I'm not convinced I've seen a win yet in that list in the first 4 pages of looking | 19:18 |
| clarkb | sdague: the first hit I have is a win | 19:19 |
| clarkb | sdague: gate-nova-python26 Installing collected packages blah blah blah blah | 19:19 |
| sdague | ok, I thought you said we didn't have any queries based on this yet? | 19:19 |
| clarkb | sdague: we don't but ^ would point you straight to the current problems | 19:19 |
| jeblair | it's still early days with basically no tuning done to it; i think the question isn't "does throwing this thing at the problem right out of the box magically work?" versus, "is this a potentially useful with some more research and tuning". | 19:20 |
| sdague | right, which is why I just suggested downplaying it | 19:21 |
| clarkb | sdague: I can do that, but isn't that one of the goals of the thing (I think it is fine to point it out as a goal, I can make it clear we haven't actually done that particular thing yet) | 19:27 |
| clarkb | before I forget, did anyone catch my idea of preventing people from pushing to refs/drafts/* | 19:27 |
| clarkb | we can't disable drafts but we can prevent people from creating them | 19:27 |
| jeblair | clarkb: i did not, but i think it is an idea whose time has come. | 19:27 |
| clarkb | it was late last night | 19:27 |
| clarkb | https://review.openstack.org/#/c/63567/ was updated but without a comment answering jog0 and my questions | 19:28 |
| clarkb | which led me to thinking about drafts and that was my idea, prevent them from ebing created | 19:28 |
| sdague | clarkb: the problem with the installing collected packages line, is then you take the general part of that string and stick it back in as a fingerprint and it's 99% success. We'll probably want some sort of reverse transform here to figure out if there is a generic relevancy here. | 19:28 |
| clarkb | sdague: I don't think you will be able to just copy pasta | 19:29 |
| *** thedodd has joined #openstack-infra | 19:29 | |
| clarkb | sdague: but that line claerly indicates there is a package problem of some sort | 19:29 |
| clarkb | because successful jobs don't have anything like it | 19:29 |
| clarkb | it would be even better if it picked up on the error pbr couldn't install line | 19:31 |
| *** CaptTofu has quit IRC | 19:33 | |
| *** sarob has quit IRC | 19:33 | |
| sdague | clarkb: success jobs don't have that exact package list, but they have "Installing collected packages:" lists | 19:33 |
| *** sarob has joined #openstack-infra | 19:33 | |
| *** CaptTofu has joined #openstack-infra | 19:33 | |
| sdague | anyway, it's something we'll explore over time here. I just want us to be careful on expectations. | 19:34 |
| clarkb | sure | 19:34 |
| *** rakhmerov has joined #openstack-infra | 19:34 | |
| clarkb | sdague: my expectation isn't that you can copy paste those lines from elasticsearch into elasticrecheck | 19:34 |
| clarkb | instead they are very good hints | 19:34 |
| clarkb | jeblair: re the draft thing, do you think we need to take a temperature on that or just do it for the greater good | 19:37 |
| sdague | clarkb: re: draft, greater good :) | 19:37 |
| sdague | drafts are terrible | 19:37 |
| jeblair | clarkb: i think we should (a) test it so we don't eat crow if it doesn't work, then (b) announce it to the ml with an implementation date... | 19:38 |
| *** rakhmerov has quit IRC | 19:38 | |
| *** sarob has quit IRC | 19:38 | |
| jeblair | clarkb: i agree with sdague that i strongly think we should do it; announcing it in advance will give people notice and an opportunity for objection | 19:39 |
| jeblair | (some people may actually be using it successfully; if such a person exists, they'll need to change their workflow) | 19:40 |
| openstackgerrit | James E. Blair proposed a change to openstack-infra/config: Zuul status: don't toggle on link click https://review.openstack.org/64716 | 19:41 |
| jeblair | Alex_Gaynor: ^ | 19:41 |
| openstackgerrit | James E. Blair proposed a change to openstack-infra/config: Zuul status: don't toggle on link click https://review.openstack.org/64716 | 19:42 |
| jeblair | now without tabs | 19:42 |
| clarkb | jeblair: sounds good | 19:42 |
| clarkb | jeblair: forgive my js fail, but line 314 is event in scope? | 19:43 |
| *** senk has joined #openstack-infra | 19:46 | |
| *** sarob has joined #openstack-infra | 19:47 | |
| *** CaptTofu has quit IRC | 19:49 | |
| *** markmcclain has quit IRC | 19:49 | |
| *** CaptTofu has joined #openstack-infra | 19:49 | |
| krotscheck | clarkb: You mentioned at one point that you managed to get xvfb running on some of the builds - I'm digging through the configs and can't find a reference anywhere. Is that a config thing? Where should I be lookign? | 19:50 |
| *** burt has joined #openstack-infra | 19:50 | |
| clarkb | krotscheck: https://git.openstack.org/cgit/openstack-infra/config/tree/modules/jenkins/files/slave_scripts/run-selenium.sh | 19:51 |
| krotscheck | Awesome, thanks | 19:51 |
| *** markmcclain has joined #openstack-infra | 19:53 | |
| openstackgerrit | Devananda van der Veen proposed a change to openstack-infra/config: Add ironic log files to logstash indexing https://review.openstack.org/64717 | 19:55 |
| jeblair | mikal: may i suggest that you could make your job names smaller by not including 'gate-' in them? the prefix is a convention we use to indicate whether a job is intended to run in the check+gate queues, or only the check queue | 19:56 |
| jeblair | mikal: therefore it's not really applicable to turbo-hipster | 19:56 |
| *** michchap has quit IRC | 19:58 | |
| clarkb | devananda: to double check, ironic uses oslo logging? | 19:58 |
| Ajaeger | anybody here that can help with a python packaging issue - or where should I ask? Looks pbr related | 19:59 |
| Ajaeger | run tox I get during the install phase "UnicodeEncodeError: 'ascii' codec can't encode character u'\u2013' in position 6938: ordinal not in range(128)" | 20:00 |
| Ajaeger | details at http://paste.openstack.org/show/59642/ | 20:00 |
| *** jgrimm has quit IRC | 20:00 | |
| *** lcestari has quit IRC | 20:05 | |
| clarkb | Ajaeger: it is creating a manifest file and somewhere in that list is an ndash | 20:06 |
| clarkb | Ajaeger: pbr can probably fix this by encoding the results of find_sources or you can rename the file with the ndash | 20:07 |
| jeblair | Ajaeger: what are the contents of /home/aj/Software/vcs/OpenStack/openstack-manuals/test-requirements.txt ? | 20:07 |
| jeblair | (i don't understand why pbr is running in git context instead of downloading the release) | 20:07 |
| Ajaeger | openstack-doc-tools | 20:08 |
| Ajaeger | is the single line | 20:08 |
| jeblair | Ajaeger: i can't reproduce that behavior locally | 20:08 |
| Ajaeger | I can run openstack-doc-tools' setup.py just fine. | 20:08 |
| Ajaeger | jeblair: both annegentle_ and myself could ;( | 20:08 |
| *** markmcclain has quit IRC | 20:09 | |
| devananda | clarkb: yes -- http://git.openstack.org/cgit/openstack/ironic/tree/openstack-common.conf#n17 | 20:10 |
| Ajaeger | jeblair: do you have any idea how to debug this? | 20:10 |
| *** markmcclain has joined #openstack-infra | 20:11 | |
| devananda | clarkb: also, those screenlogs won't exist outside of "check experimental" right now. pending work on tempest, etc... | 20:11 |
| clarkb | devananda: thats fine, if the files don't exist they are ignored | 20:12 |
| Ajaeger | There are files with an ndash in openstack-manuals - could that be a problem? | 20:15 |
| clarkb | Ajaeger: possibly, I wonder if it is running in git context because of the openstack-manuals git repo? | 20:15 |
| *** michchap has joined #openstack-infra | 20:15 | |
| clarkb | also boo for unicode file names | 20:15 |
| clarkb | and unfortunately mordred is afk right now so can't chime in about the git context vs package | 20:16 |
| jeblair | Ajaeger: do you have a commit you're working on? can you push it up for review and mark it wip so i can fetch it? | 20:17 |
| Ajaeger | jeblair, clarkb: I think I got it thanks to your help. | 20:18 |
| Ajaeger | to reproduce: checkout openstack-manuals and run "tox" | 20:18 |
| Ajaeger | I just deleted the files with an ndash in it and everything works fine... | 20:19 |
| Ajaeger | So, that's what I will do now - rename these files... | 20:19 |
| clarkb | Ajaeger: well I think that may fix a symptom of a bigger problem | 20:20 |
| Ajaeger | clarkb: what do you mean? | 20:21 |
| clarkb | Ajaeger: pbr shouldn't be adding the openstack-manuals files to the openstack-doc-tools manifest | 20:21 |
| clarkb | it shouldn't be running in a git context at all | 20:21 |
| Alex_Gaynor | jeblair: thanks | 20:22 |
| Ajaeger | clarkb: no, it shouldn't - so, what shall I do? file a bug report against pbr? | 20:22 |
| clarkb | Ajaeger: I think that is a good start | 20:22 |
| clarkb | include the paste that you had earlier showing the exception | 20:23 |
| jeblair | Ajaeger: ah, yes, i can reproduce that | 20:23 |
| fungi | Ajaeger: (but also, rename s/–/-/g on the filenames might be appreciated) | 20:23 |
| fungi | sticking to strict ascii for filenames is probably still a good idea given the way macos hfs+ uses different utf-8 normalization than typical linux/bsd filesystems | 20:25 |
| Ajaeger | fungi: Yeah, I rename now, won't wait for pbr to get fixed... | 20:25 |
| *** vipul is now known as vipul-away | 20:26 | |
| fungi | heck, a hacking check for non-ascii filenames might even be a good idea (not that i expect docs repos to necessarily use hacking jobs, but just in general) | 20:26 |
| *** vipul-away is now known as vipul | 20:26 | |
| Ajaeger | jeblair, clarkb, fungi : Thanks for your help! I filled this bug now: https://bugs.launchpad.net/pbr/+bug/1265622 | 20:28 |
| uvirtbot | Launchpad bug 1265622 in pbr "pbr running in git context when it shouldn't" [Undecided,New] | 20:28 |
| Ajaeger | Could you add some more comments to it as appropriate, please? | 20:29 |
| *** rakhmerov has joined #openstack-infra | 20:34 | |
| *** ruhe has joined #openstack-infra | 20:38 | |
| Ajaeger | what an irony: None of these four files was used at all ;( | 20:38 |
| *** ryanpetrello has joined #openstack-infra | 20:39 | |
| *** rakhmerov has quit IRC | 20:39 | |
| *** mattf has joined #openstack-infra | 20:43 | |
| mattf | folks, i'm seeing "OSError: [Errno 13] Permission denied: '/usr/lib/python2.6/site-packages/Babel-0.9.4-py2.6.egg-info/top_level.txt'" from the jenkins gate at http://logs.openstack.org/40/64540/2/gate/gate-python-savannaclient-python26/2e617d0/console.html#_2014-01-02_20_38_30_440 | 20:45 |
| mattf | anyone seen something similar / have a suggestion for me? | 20:45 |
| *** milki has quit IRC | 20:45 | |
| *** milki has joined #openstack-infra | 20:46 | |
| *** fbo is now known as fbo_away | 20:47 | |
| *** CaptTofu has quit IRC | 20:47 | |
| Ajaeger | Now https://review.openstack.org/#/c/64567/ can be reviewed ;) | 20:47 |
| *** CaptTofu has joined #openstack-infra | 20:47 | |
| dhellmann | did someone already notice that https://pypi.python.org/pypi/MySQL-python/1.2.5 was released with mordred's fix? | 20:50 |
| jeblair | dhellmann: nope that's news, thanks! | 20:51 |
| *** Ryan_Lane has joined #openstack-infra | 20:51 | |
| dhellmann | jeblair: I don't know if it fixes the problem, but it should help :-) | 20:51 |
| *** loq_mac has joined #openstack-infra | 20:52 | |
| jeblair | mattf: https://review.openstack.org/#/c/64709/ should have fixed that, but doesn't appear to have taken effect on the centos slaves | 20:55 |
| *** vipul is now known as vipul-away | 20:55 | |
| jeblair | fungi: ^ puppet is running but pip doesn't seem to be downgrading virtualenv | 20:55 |
| *** johnthetubaguy has joined #openstack-infra | 20:56 | |
| mattf | jeblair, you recommend i just try again tomorrow? | 20:56 |
| jeblair | mattf: yes, or in a few hours | 20:57 |
| jeblair | fungi: http://paste.openstack.org/show/59692/ | 20:57 |
| mattf | will do | 20:57 |
| *** loq_mac has quit IRC | 20:57 | |
| jeblair | Requested virtualenv==1.10.1, but installing version 1.11 | 20:58 |
| jeblair | a reason would be a really cool thing to include on that line. :( | 20:58 |
| *** loq_mac has joined #openstack-infra | 20:58 | |
| *** Ryan_Lane has quit IRC | 20:59 | |
| dstufft | jeblair: what version of pip is it using to do the install | 20:59 |
| jeblair | [root@centos6-5 ~]# pip --version | 20:59 |
| jeblair | pip 1.3.1 from /usr/lib/python2.6/site-packages (python 2.6) | 20:59 |
| dstufft | It's hitting the existing build dir bug | 20:59 |
| ruhe | jeblair, fungi: btw, is this how nodes get scripts updated http://git.openstack.org/cgit/openstack-infra/nodepool/tree/nodepool/nodepool.py#n614 ? | 20:59 |
| *** croberts has quit IRC | 20:59 | |
| dstufft | there's a 1.11 build dir laying around and it is installing that instead of deleting it and downloading 1.10.1 and building/installing it | 20:59 |
| dstufft | that was fixed in 1.4 or 1.5 I think | 20:59 |
| * dstufft heads out | 21:00 | |
| jeblair | ruhe: it's how cached images get updated, jenkins slave scripts included. if you're interested in a particular script+job combo, i can help confirm that's the right place. | 21:01 |
| jeblair | dstufft: thanks | 21:01 |
| *** smarcet has quit IRC | 21:01 | |
| *** UtahDave has left #openstack-infra | 21:02 | |
| *** ArxCruz has quit IRC | 21:02 | |
| ruhe | jeblair: i just want to understand the flow. once a change in the slave script is merged, how do active slave nodes get this update? | 21:02 |
| jeblair | ruhe: which slaves, and which slave scripts? | 21:03 |
| *** UtahDave has joined #openstack-infra | 21:03 | |
| *** thedodd has quit IRC | 21:05 | |
| jeblair | sudo salt centos*.slave.openstack.org cmd.run 'rm -fr /tmp/pip-build-root/' | 21:05 |
| jeblair | sudo salt centos*.slave.openstack.org cmd.run 'pip install -U virtualenv==1.10.1' | 21:05 |
| jeblair | fungi, clarkb: ^ i ran that | 21:06 |
| ruhe | jeblair: for instance any file under jenkins/files/slave_scripts get modified; i see that new images get these scripts installed with slave.pp; but i can't see how existing instances get this scripts updated | 21:07 |
| *** Ryan_Lane has joined #openstack-infra | 21:07 | |
| jeblair | ruhe: those files are updated by puppet run via cron on long-running nodes, and on single-use nodes, new versions will go into effect when an image update runs and popuplates a new image with the current version | 21:09 |
| jeblair | ruhe: so yes, for nodepool based nodes, that's how they get updated | 21:09 |
| jeblair | (once per day, currently) | 21:10 |
| ruhe | jeblair: thank you :) that's what i wanted to confirm | 21:10 |
| *** loq_mac has quit IRC | 21:11 | |
| *** croberts has joined #openstack-infra | 21:11 | |
| *** rongze has joined #openstack-infra | 21:12 | |
| fungi | jeblair: cool | 21:12 |
| *** freyes_ has quit IRC | 21:12 | |
| fungi | jeblair: salt calls seemed to have worked so far? | 21:12 |
| jeblair | fungi: yeah, looks like they all reported back with sane output | 21:13 |
| jeblair | fungi: http://paste.openstack.org/show/59701/ | 21:13 |
| *** freyes_ has joined #openstack-infra | 21:14 | |
| *** CaptTofu has quit IRC | 21:15 | |
| fungi | jeblair: awesome. also glad to know that a stale /tmp/pip-build-root will cause it to reinstall old packages (though mildly worrisome that centos6 slaves had pip 1.3.1) | 21:16 |
| *** rongze has quit IRC | 21:17 | |
| fungi | that looks very similar to the caching issues we saw with the -b option | 21:17 |
| Ajaeger | Yeah, openstack-doc-tools is now checking for annegentle_ as well ;) Thanks again, fungi, clarkb, jeblair! | 21:17 |
| annegentle_ | Thanks y'all! | 21:17 |
| fungi | annegentle_: Ajaeger: of course, glad we could help! | 21:17 |
| jeblair | fungi: yeah, i didn't want to open any new cans which may or may not have worms. | 21:18 |
| *** vipul-away is now known as vipul | 21:18 | |
| *** johnthetubaguy has quit IRC | 21:18 | |
| *** syerrapragada1 has quit IRC | 21:19 | |
| *** johnthetubaguy has joined #openstack-infra | 21:19 | |
| fungi | all our cans are full of worms | 21:19 |
| *** loq_mac has joined #openstack-infra | 21:23 | |
| openstackgerrit | Peter Balland proposed a change to openstack-infra/config: Remove gates from Congress https://review.openstack.org/64730 | 21:27 |
| *** croberts has quit IRC | 21:27 | |
| *** johnthetubaguy has quit IRC | 21:29 | |
| *** banix has quit IRC | 21:31 | |
| *** dprince has quit IRC | 21:33 | |
| *** ryanpetrello has quit IRC | 21:35 | |
| *** rakhmerov has joined #openstack-infra | 21:35 | |
| *** rakhmerov has quit IRC | 21:40 | |
| *** ruhe has quit IRC | 21:40 | |
| *** Ajaeger has quit IRC | 21:41 | |
| krotscheck | Can I get a sanity check on a diff from anyone interested in the JS toolchain? http://paste.openstack.org/show/59703/ | 21:42 |
| krotscheck | It's a followup to "Hey we want to use node for builds but don't want to install it on our boxes." | 21:42 |
| jeblair | krotscheck: sure; first thing is you could just push a change up for review; gerrit has a better diff interface :). you can mark it 'work in progress' and ping people individually if you want to make it clear you aren't ready for it to be merged.... | 21:45 |
| *** croberts has joined #openstack-infra | 21:45 | |
| jeblair | krotscheck: i think the mechanics are good, but since this is a single-project solution, it might make more sense to live as a script in the storyboard repo | 21:46 |
| krotscheck | jeblair: Those were my thoughts at first, but when presented with the option I usually default to generic solutions. | 21:47 |
| krotscheck | I'll get it into gerritt | 21:47 |
| krotscheck | (What with rumors of horizon wanting to use this toolchain etc etc) | 21:47 |
| *** vipul has quit IRC | 21:48 | |
| jeblair | krotscheck: interesting rumours; i'm not sure that fits with the project support policy, and so i would not expect this to be accepted in the openstack context. | 21:48 |
| *** ryanpetrello has joined #openstack-infra | 21:48 | |
| *** ryanpetrello has quit IRC | 21:48 | |
| krotscheck | jeblair: Project support policy? | 21:48 |
| *** vipul has joined #openstack-infra | 21:49 | |
| jeblair | krotscheck: yes, we don't do things that break the current stable releases of ubuntu or centos | 21:49 |
| *** ryanpetrello has joined #openstack-infra | 21:49 | |
| david-lyle | krotscheck: is this change just for testing? | 21:49 |
| david-lyle | if it compiles anything, that likely won't fly | 21:49 |
| *** ryanpetrello has quit IRC | 21:50 | |
| david-lyle | sorry haven't gone through the diff yet | 21:50 |
| *** beagles has quit IRC | 21:50 | |
| jeblair | krotscheck: in practice, that means we do testing on plain installations of those operating systems (or with os-vendor-provided repos like ubuntu cloud archive) | 21:50 |
| krotscheck | david-lyle: It adds the javascript toolchain (npm/grunt/bower) to WORKSPACE/.local | 21:50 |
| jeblair | krotscheck: so basically, the reason why we asked you to do it this way is _because_ we can't do it for openstack (eg, horizon) | 21:51 |
| krotscheck | david-lyle: It's intended for build tools & testing only, not as an actual server. | 21:51 |
| *** ryanpetrello has joined #openstack-infra | 21:52 | |
| krotscheck | jeblair: Wait a sec, if storyboard isn't actually openstack, then why are we using infra? | 21:52 |
| jeblair | krotscheck: so that's why it's more appropriate to just be in storyboard's repo. | 21:52 |
| jeblair | krotscheck: storyboard isn't part of openstack, it's part of openstack-infra | 21:53 |
| krotscheck | Argh. Semantics. Right | 21:53 |
| jeblair | krotscheck: openstack-infra exists to support the development of openstack | 21:53 |
| david-lyle | different rules apply | 21:53 |
| jeblair | krotscheck: not really; openstack is a cloud computing system that includes things like compute and storage management systems | 21:54 |
| jeblair | krotscheck: at the moment, no one has suggested that it also include a bug tracker | 21:54 |
| jeblair | however, that's something that the project needs for its development efforts | 21:54 |
| jeblair | krotscheck: so _openstack_, as a piece (or collection) of software has some requirements around what systems it targets and supports | 21:55 |
| jeblair | that doesn't necessarily affect openstack-infra projects (some of which are even in java, if you can imagine such a thing), but we try to stay as close as possible | 21:56 |
| krotscheck | Right, I get all that (though I might not remember most of the details). | 21:56 |
| jeblair | krotscheck: for instance, we only run servers on the latest stable releases of centos and ubuntu. so running storyboard in production will eventually bring up similar issues... | 21:56 |
| jeblair | krotscheck: it's easy if it depends on packages that are in the os repo. if not, then we have to figure out how to install them on servers, and make sure they are automatically maintained and receive security updates. | 21:57 |
| *** Ryan_Lane has quit IRC | 21:57 | |
| krotscheck | So, is there a distinction between "here's the tools needed to build the software" and "here's the operating system that the software needs to run on"? | 21:58 |
| krotscheck | Or do they all need to be the same for the sake of openstack (as opposed to openstack-infra) | 21:59 |
| krotscheck | (Just clarifying because the package output of storyboard is static HTML, JS, CSS, and images) | 21:59 |
| krotscheck | (Which only need to be hosted by insert-web-server-of-choice-here) | 22:00 |
| jeblair | krotscheck: building and installing static files like that sounds fairly easy; builds and installs can definitely be distinct. | 22:01 |
| *** mfer has quit IRC | 22:01 | |
| *** CaptTofu has joined #openstack-infra | 22:02 | |
| krotscheck | jeblair: Right - I see a lot of post-build SCP/FTP in the config files. | 22:02 |
| jeblair | krotscheck: yep, building and uploading the static files to tarballs.o.o and installing releases from there sounds like a good solution. | 22:04 |
| *** vipul is now known as vipul-away | 22:04 | |
| *** vipul-away is now known as vipul | 22:04 | |
| krotscheck | jeblair: So given that the intent of my changes is that it should _never_ run on openstack, even though it outputs assets that will... should I still pull that script back into the storyboard repo? | 22:04 |
| jeblair | krotscheck: it outputs assets that will run on a server, not "assets that will run on openstack" (we run our servers on a provider that's running openstack, but that's beside the point) | 22:06 |
| jeblair | krotscheck: but yes, i think it should be in storyboard | 22:06 |
| krotscheck | jeblair: Ok, so horizon's web interface runs on a server? | 22:07 |
| krotscheck | (It's all the same to me, I just feel that other projects might want to make the toolchain) | 22:08 |
| krotscheck | *make use of | 22:08 |
| jeblair | krotscheck: i'm not sure what you're getting at. my point was that storyboard is not a component of openstack, and it would be good to not be confused on that point. | 22:08 |
| jeblair | krotscheck: openstack projects _may not_ make use of this toolchain | 22:08 |
| krotscheck | So the horizon guys are dead in the water? | 22:08 |
| jeblair | krotscheck: they aren't dead in the water, they have a working product. | 22:08 |
| krotscheck | Huhn. Interesting | 22:08 |
| clarkb | jeblair: worth pointing out they have active blueprints to work on this stuff iirc | 22:08 |
| jeblair | clarkb: neat. they will have to address the support issue | 22:09 |
| clarkb | jeblair: I agree, just pointing out that this is a thing for horizon. https://blueprints.launchpad.net/horizon/+spec/jasmine-integration | 22:09 |
| clarkb | I am assuming tests that don't require a DOM want to run in nodejs or similar | 22:10 |
| *** thedodd has joined #openstack-infra | 22:10 | |
| krotscheck | Well, in this context nodejs is just the task runner (ivy equivalent). The tests themselves are run in [insert configurable browser here] | 22:12 |
| *** rongze has joined #openstack-infra | 22:12 | |
| vipul | so with pip 1.5, should we be removing d2to1 from the setup.py? seeing some wierd errors with our Trove packages | 22:13 |
| clarkb | vipul: d2to1 should've been removed months ago | 22:13 |
| vipul | clarkb: you're right, maybe it still lives in our internal packaging | 22:14 |
| vipul | thanks | 22:14 |
| krotscheck | jeblair, clarkb: General apologies for being obtuse about some of these topics btw. Y'all have a lot of institutional knowledge that I'm still absorbing. | 22:15 |
| *** milki has quit IRC | 22:17 | |
| *** rongze has quit IRC | 22:17 | |
| jeblair | krotscheck: no apology necessary; it takes most people several months to come up to speed. | 22:17 |
| *** milki has joined #openstack-infra | 22:17 | |
| *** Alex_Gaynor has quit IRC | 22:20 | |
| openstackgerrit | Zane Bitter proposed a change to openstack-infra/config: Separate handling of ChangeIds and Git SHAs in Gerrit https://review.openstack.org/64738 | 22:20 |
| *** Alex_Gaynor has joined #openstack-infra | 22:25 | |
| *** hogepodge has quit IRC | 22:26 | |
| *** UtahDave has quit IRC | 22:26 | |
| *** UtahDave has joined #openstack-infra | 22:27 | |
| *** croberts has quit IRC | 22:27 | |
| *** zehicle_at_dell has joined #openstack-infra | 22:29 | |
| *** zehicle has joined #openstack-infra | 22:29 | |
| devananda | clarkb: for this morning's everything-fail, what bug # should I be reverifying against? | 22:31 |
| *** mrda has joined #openstack-infra | 22:32 | |
| *** markmcclain has quit IRC | 22:32 | |
| *** rwsu has quit IRC | 22:33 | |
| *** mkoderer_ has quit IRC | 22:33 | |
| *** Ryan_Lane has joined #openstack-infra | 22:34 | |
| clarkb | devananda: I am not sure that there is one. I arrived late to the party | 22:34 |
| *** Ryan_Lane has quit IRC | 22:34 | |
| devananda | clarkb: i couldn't find one looking through status.oo/rechecks | 22:35 |
| devananda | clarkb: but apparently reverify doesn't do anything without a bug, any more :( | 22:35 |
| *** ryanpetrello has quit IRC | 22:35 | |
| clarkb | devananda: but you have approval powers just reapprove | 22:35 |
| clarkb | (admittedly a work around but an intentional one) | 22:36 |
| devananda | ah. wasn't sure that would kick the job off. thanks | 22:36 |
| *** mkoderer_ has joined #openstack-infra | 22:36 | |
| *** Alex_Gaynor has quit IRC | 22:36 | |
| *** rakhmerov has joined #openstack-infra | 22:36 | |
| mriedem | devananda: it's bug 1265641 | 22:37 |
| uvirtbot | Launchpad bug 1265641 in nova "Can not create pbr directory due to permission denied" [Undecided,New] https://launchpad.net/bugs/1265641 | 22:37 |
| mattf | jeblair, jenkins looks back to normal for me, thanks! | 22:37 |
| devananda | mriedem: hm. i think we saw the python-MySQL one in our pipe, for what ever reason. | 22:38 |
| devananda | but thanks :) | 22:39 |
| *** mattf has left #openstack-infra | 22:39 | |
| *** rwsu has joined #openstack-infra | 22:40 | |
| *** Alex_Gaynor has joined #openstack-infra | 22:41 | |
| *** rakhmerov has quit IRC | 22:41 | |
| openstackgerrit | Sean Dague proposed a change to openstack/requirements: wip: totally speculative exploration of sqla 0.9 https://review.openstack.org/64742 | 22:41 |
| clarkb | sdague: fingers are crossed here | 22:44 |
| *** mkoderer_ has quit IRC | 22:48 | |
| *** weshay has quit IRC | 22:49 | |
| *** mkoderer_ has joined #openstack-infra | 22:52 | |
| Alex_Gaynor | jeblair: new collapsed UI is awesome, thaknks | 22:55 |
| jeblair | russellb: ^ | 22:55 |
| zaro | clarkb, fungi: i fixed the contributor agreement problem on review-dev.o.o | 22:57 |
| fungi | devananda: mriedem: there are several problems all surfacing at once in the face of new pip and virtualenv releases | 22:58 |
| fungi | zaro: did you simply have to create that group and populate it? | 22:58 |
| *** rwsu has quit IRC | 22:59 | |
| zaro | clarkb: i'm not seeing the contact info error you reported. | 22:59 |
| clarkb | zaro: let me try again | 23:00 |
| *** mkoderer_ has quit IRC | 23:00 | |
| clarkb | oh it says I am verified now | 23:01 |
| clarkb | zaro: what was the fix? | 23:01 |
| sdague | clarkb: yeh, don't cross your fingers on that one. I blows up really early with keystone db setup | 23:01 |
| zaro | fungi: no, we already have a openstack-cla group. i just changed the ACLs to allow that group access. | 23:01 |
| zaro | clarkb: ^ | 23:01 |
| clarkb | zaro: ideally we would use whatever group gerrit wants to use | 23:01 |
| clarkb | zaro: I can submit contact info now too | 23:01 |
| *** mkoderer_ has joined #openstack-infra | 23:03 | |
| *** sarob has quit IRC | 23:03 | |
| zaro | I might be wrong, but i assumed that openstack-cla is the group that contains members who accepted the cla. | 23:03 |
| fungi | zaro: openstack-cla is a different group, actually. that's a cruft hold-over from when we used to sync groups from lp and approve cla signatures there | 23:04 |
| *** sarob has joined #openstack-infra | 23:04 | |
| fungi | zaro: it looks like gerrit wanted to create a different group (added a uuid for it in the groups file but hadn't successfully inserted it into the db) | 23:04 |
| zaro | fungi: ok. i'll revert and try the db route. | 23:04 |
| jeblair | yeah, if we need to provide our own group we can use openstack-cla (it's unused), but if gerrit wants to default to a different group name, let's use that | 23:05 |
| *** eharney has quit IRC | 23:06 | |
| zaro | clarkb: FWIW, i never had problems submitting contacts even before my latest changes to ACLs. | 23:06 |
| clarkb | zaro: did you do it through the CLA signing page? | 23:07 |
| clarkb | that is where I had the error | 23:07 |
| zaro | clarkb: yes. | 23:07 |
| fungi | zaro: jeblair: the project.confg and groups files refer to a "CLA Accepted - ICLA" which is simply missing from the account_groups and related db tables | 23:07 |
| zaro | clarkb: and the plugins are not installed on site creation. need to manually add those. | 23:08 |
| *** sarob has quit IRC | 23:08 | |
| jeblair | fungi: got it; creating that group and using it makes sense | 23:08 |
| fungi | i think we simply need to create it and add all accounts to to which were "V" status for cla id #2 in the account_agreements table | 23:09 |
| fungi | er, to it | 23:09 |
| clarkb | zaro: is there any way to automate that? eg can puppet put some wars or jars in special dirs? | 23:09 |
| clarkb | zaro: manual is bad | 23:09 |
| zaro | clarkb: ohh, yeah. it's just a simple copy to plugins folder then restart gerrit. should be easily puppitable. | 23:10 |
| fungi | i'm going to guess for plugins we just need to duplicate the gerrit.war copying pattern but for the plugin files | 23:10 |
| zaro | clarkb: i'm just saying that gerrit does not do it for us. | 23:10 |
| *** thedodd has quit IRC | 23:11 | |
| *** fifieldt has joined #openstack-infra | 23:11 | |
| *** rwsu has joined #openstack-infra | 23:12 | |
| *** rongze has joined #openstack-infra | 23:13 | |
| *** rongze has quit IRC | 23:18 | |
| sdague | I'm poking a review-dev a little and putting a couple of things over in the etherpad that are hopefully helpful in discovery | 23:18 |
| jeblair | sdague: cool, thanks | 23:19 |
| clarkb | was the etherpad linked in here recently? my grepping doesn't show it | 23:20 |
| jeblair | clarkb: i made a blueprint that links to the etherpad so we don't lose it, but it's here: https://etherpad.openstack.org/p/gerrit-2.8-upgrade | 23:21 |
| jeblair | clarkb: (if you need to find it again in the future, you can start with the blueprint) | 23:21 |
| pleia2 | clarkb: https://etherpad.openstack.org/p/gerrit-2.8-upgrade | 23:21 |
| pleia2 | ah, jeblair beat me to it :) | 23:21 |
| clarkb | ty] | 23:21 |
| *** dcramer_ has quit IRC | 23:22 | |
| jeblair | sdague: note https://review.openstack.org/#/c/60080/1 | 23:23 |
| sdague | jeblair: yep, found it as you were typing about it | 23:26 |
| mordred | o/ | 23:35 |
| mordred | jeblair: it seems that we have a new mysql-python in the wild | 23:36 |
| mordred | dhellmann: ^^ thank you! | 23:36 |
| jeblair | mordred: yeah! i think the virtualenv pinning is masking the problem with 1.2.4 at the moment | 23:36 |
| mordred | jeblair: awesome | 23:36 |
| mordred | jeblair: I'm going to delete my python-mysqldb from pypi | 23:37 |
| *** rakhmerov has joined #openstack-infra | 23:37 | |
| jeblair | mordred: so it looks like the periodic mirror update job for 2.7 times out because it takes > 1 hour | 23:39 |
| jeblair | mordred: https://jenkins.openstack.org/job/periodic-mirror-python27/ | 23:39 |
| mordred | jeblair: yes. I believe this is known - I thought we'd expanded the timeout for it | 23:39 |
| mordred | clarkb: ^^ didn't we extend the timeout? | 23:39 |
| jeblair | apparently not | 23:39 |
| mordred | jeblair: part of the issue is that we're now double-building everything due to also building wheels | 23:39 |
| mordred | jeblair: so the script has to do twice as much work | 23:40 |
| openstackgerrit | A change was merged to openstack-infra/zuul: Update the url for the zuul git repo https://review.openstack.org/62676 | 23:40 |
| jeblair | mordred: py26 is 90m, py27 is 60m | 23:40 |
| jeblair | py26 _takes_ 80m | 23:40 |
| *** ArxCruz has joined #openstack-infra | 23:41 | |
| *** ryanpetrello has joined #openstack-infra | 23:41 | |
| *** rakhmerov has quit IRC | 23:41 | |
| clarkb | jeblair: re new change screen I believe that is a non issue, it is user configurable whether you want old or new | 23:42 |
| clarkb | jeblair: under your user settings preferences | 23:43 |
| openstackgerrit | Sean Dague proposed a change to openstack-infra/config: Gerrit-2.8: Add secondary index support https://review.openstack.org/60080 | 23:43 |
| sdague | so, that's a stab | 23:44 |
| clarkb | now I see 3.0 will remove the old screen... | 23:44 |
| jeblair | clarkb: what's a non-issue? radically changing the ui for the keystone of the developer workflow? i care. | 23:44 |
| clarkb | jeblair: well I didn't realize we had to change it | 23:45 |
| clarkb | I thought both screens would live together happily ever after | 23:45 |
| clarkb | (it is currently toggleable, you have a choice today, but that choice will be removed apparently) | 23:45 |
| jeblair | clarkb: ah, yeah. it'd be nice if we had more time, but we don't. | 23:45 |
| *** sarob has joined #openstack-infra | 23:46 | |
| jeblair | clarkb: so given that, i'd like to try to start seeing what we can do to accomodate it, but i'm starting to lean toward old-as-default for now so we have a smoother transition | 23:46 |
| clarkb | ++ | 23:47 |
| *** sarob has quit IRC | 23:47 | |
| *** sarob_ has joined #openstack-infra | 23:47 | |
| sdague | could we light the new screen on review-dev to see the difference? | 23:47 |
| clarkb | sdague: its lit, you can select it under your preferences | 23:47 |
| sdague | ok, cool | 23:47 |
| mordred | what's the change screen difference? ah! I will go look | 23:47 |
| clarkb | gerrit needs someone like krotscheck to fix the UI for them >_> | 23:49 |
| jeblair | clarkb: _we_ fixed the ui for them. | 23:49 |
| sdague | well, there are parts of this I like better | 23:49 |
| clarkb | jeblair: ya | 23:49 |
| jeblair | then they broke it again | 23:49 |
| *** UtahDave has quit IRC | 23:49 | |
| mordred | yeah - I don't hate the new screen - but I haven't used it in anger much, so I don't have good opinions really | 23:49 |
| mordred | it is radically different | 23:50 |
| clarkb | sdague: you need to load up a change with a bunch of dependencies and revisions then try looking at it on a non widescreen monitor | 23:50 |
| clarkb | mordred: the biggest problem is it grows horizontally | 23:50 |
| sdague | would it be possible to import something like the tempest change history in? I feel like some of the UI only will show up with bigger data | 23:50 |
| mordred | clarkb: in what way? | 23:50 |
| sdague | clarkb: you have an example? | 23:50 |
| sdague | clarkb: also, define "non widescreen" | 23:50 |
| clarkb | sdague: https://gerrit-review.googlesource.com/#/c/53392/ | 23:51 |
| clarkb | sdague: basically changes for gerrit itself | 23:51 |
| clarkb | sdague: I have a chromium instance fullscreened on a 16:10 monitor and if I get anywhere near 4:3 I end up with a horizontal scroll bar | 23:51 |
| jeblair | sdague: anything less than 1440 pixels | 23:52 |
| mordred | oh. that does horizontally scroll | 23:52 |
| mordred | gross | 23:53 |
| mordred | god. there's also SO MUCH INFO on that screen that I don't care about | 23:53 |
| jeblair | sdague: current gerrit requires 1116 pixels (and that's only in side-by-side diff view | 23:53 |
| jeblair | mordred: yeah, that's one of the things i was getting at with "crowded" | 23:54 |
| mordred | ++ | 23:54 |
| mordred | well... now that it's got a REST api and the rest api is getting comprehensive | 23:54 |
| sdague | so how stable are those gwt table tags? | 23:54 |
| mordred | we could just roll with old screen by default for 2.8 and by 2.9 have our own web ui :) | 23:54 |
| clarkb | mordred: because then upgrading to 3.0 is potentially a giant process again... | 23:54 |
| krotscheck | clarkb: Oh man, I have so many amazing ideas for gerrit's UI... | 23:55 |
| mordred | clarkb: yeah. I know. I'm offering bad suggestions. grasping at straws | 23:55 |
| jeblair | sdague: my understanding is most of the gwt stuff is not stable. | 23:55 |
| mordred | clarkb: I say that- are we expecting the REST API to diverge _wildly_ for 3.0? | 23:55 |
| mordred | I mean, other stuff would break pretty terribly if the api just up and broke | 23:56 |
| jeblair | mordred: i plan on writing a text-mode interface. i can use that to implement the inbox style workflow i've always wanted. | 23:56 |
| clarkb | mordred: they don't have any real indication of stability for anything as far as I can tell | 23:56 |
| mordred | jeblair: ++ | 23:56 |
| sdague | clarkb: honestly, I'd say get to 2.8 / 2.9 as fast as possible stay on the old UI, and then we can at least speak a close language as them | 23:56 |
| jeblair | mordred: (possible bonus for you: i think it could work offline too; you could review while on a plane) | 23:57 |
| sdague | and I agree, the UX on that page is pretty bad | 23:57 |
| jeblair | mordred: (a plane over the ocean) | 23:57 |
| *** herndon_ has joined #openstack-infra | 23:57 | |
| mordred | sdague: we can also explore alternate UIs, like jeblair's text mode, or krotscheck doing a thing or something | 23:57 |
| sdague | yep, sure | 23:57 |
| mordred | becuase trying to tie those together is a pretty big bite | 23:57 |
| sdague | so yeh, I'd default old UI. It's sort of a shame we weren't closer to them already, because we have a more active gerrit then they do, and could provide them some interesting user feedback | 23:58 |
| jeblair | ok, i changed my vote on https://review.openstack.org/#/c/60094/ | 23:58 |
| jeblair | sdague: yeah, you'd think they'd care about our install. | 23:59 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!