| *** ianw has joined #openstack-infra | 00:09 | |
| *** ericw has joined #openstack-infra | 00:10 | |
| *** annegentle has quit IRC | 00:10 | |
| *** annegentle has joined #openstack-infra | 00:12 | |
| *** ericw has quit IRC | 00:13 | |
| Alex_Gaynor | jeblair: do you ahve a screen shot of what the NNFI visualizer looks like? | 00:15 |
|---|---|---|
| *** dkehn has quit IRC | 00:16 | |
| *** ericw has joined #openstack-infra | 00:19 | |
| *** notmyname_ has joined #openstack-infra | 00:19 | |
| *** jhesketh has quit IRC | 00:23 | |
| *** kpepple has quit IRC | 00:23 | |
| *** fbo_away has quit IRC | 00:23 | |
| *** enikanorov-w has quit IRC | 00:23 | |
| *** ekarlso has quit IRC | 00:23 | |
| *** bearhands has quit IRC | 00:23 | |
| *** NobodyCam has quit IRC | 00:23 | |
| *** lyz has quit IRC | 00:23 | |
| *** vishy has quit IRC | 00:23 | |
| *** notmyname has quit IRC | 00:23 | |
| *** jmeridth has quit IRC | 00:23 | |
| *** jhesketh has joined #openstack-infra | 00:24 | |
| *** notmyname_ is now known as notmyname | 00:24 | |
| *** kpepple has joined #openstack-infra | 00:27 | |
| *** lyz has joined #openstack-infra | 00:27 | |
| *** bearhands has joined #openstack-infra | 00:27 | |
| *** NobodyCam has joined #openstack-infra | 00:28 | |
| *** ekarlso has joined #openstack-infra | 00:29 | |
| *** dkliban has joined #openstack-infra | 00:34 | |
| *** thomasm has joined #openstack-infra | 00:40 | |
| *** SlickNik_ has joined #openstack-infra | 00:40 | |
| *** dcramer_ has joined #openstack-infra | 00:40 | |
| *** annegentle has quit IRC | 00:43 | |
| *** ericw has quit IRC | 00:43 | |
| *** dprince has joined #openstack-infra | 00:43 | |
| *** thomasm has quit IRC | 00:43 | |
| *** annegentle has joined #openstack-infra | 00:44 | |
| *** SlickNik_ has quit IRC | 00:46 | |
| *** sarob has joined #openstack-infra | 00:49 | |
| *** mrda has joined #openstack-infra | 00:50 | |
| *** sarob has quit IRC | 00:53 | |
| *** yaguang has joined #openstack-infra | 00:56 | |
| *** blamar has joined #openstack-infra | 00:59 | |
| *** jhesketh__ has quit IRC | 01:02 | |
| *** michchap_ has quit IRC | 01:05 | |
| *** michchap has joined #openstack-infra | 01:07 | |
| *** dkehn has joined #openstack-infra | 01:09 | |
| *** ryanpetrello has joined #openstack-infra | 01:10 | |
| mrda | hey everyone, I'm having a little trouble signing the Individual Contributor Licence Agreement in gerrit. Anyone able to help me? | 01:18 |
| clarkb | mrda: sure | 01:21 |
| clarkb | mrda: can you be more specific about the issue(s)? | 01:21 |
| mrda | clarkb: So I'm on https://review.openstack.org/#/settings/new-agreement , have clicked the ICLA. | 01:22 |
| mrda | It's prefilled my name and email address from launchpad | 01:22 |
| lifeless | mrda: velkommen | 01:22 |
| lifeless | mrda: but so far you haven't described an issue :) | 01:22 |
| mrda | I've entered my address, country and phone number (not incl fax), entered "AGREE" in the "Compete the agreement" box | 01:22 |
| mrda | ...which allows me to click the "Submit Agreement" button. But all that does is to remove the AGREE text, and grey out the button. | 01:23 |
| mrda | Revisiting my profile shows me that it hasn't registered my agreement. It just like there's an error on form submission (without error text). | 01:23 |
| mrda | lifeless: hey, glad to be here ;) | 01:24 |
| *** dprince has quit IRC | 01:24 | |
| clarkb | mrda: weird, if there is an error it should feed that back to you. Have you signed up with the OpenStack foundation? and if so is the email you are using when signing the ICLA the same one you have provided the foundation? | 01:25 |
| mrda | yes, same email address | 01:25 |
| clarkb | mrda: also what is your Account ID according to https://review.openstack.org/#/settings/ | 01:25 |
| mrda | 8125 | 01:25 |
| *** Ryan_Lane has joined #openstack-infra | 01:26 | |
| clarkb | there are no preexisting CLA agreements according to the DB for that account | 01:27 |
| mrda | clarkb: Chrome 29 on Ubuntu 13.04, FWIW | 01:29 |
| clarkb | I wonder if the foundation side isn't accepting connections from gerrit or responding to gerrit /me pulls out laptop to check logs | 01:29 |
| mrda | clarkb: foundation profile: http://www.openstack.org/community/members/profile/12965 | 01:30 |
| *** lcestari has quit IRC | 01:31 | |
| *** Ryan_Lane has quit IRC | 01:32 | |
| clarkb | I don't see anything in the logs that scream at me as being related to the foundation side doing something funny but my browser can't get www.openstack.org | 01:35 |
| clarkb | so I think that theory is still a possibility | 01:36 |
| clarkb | and unfortunately I don't have access to things on that end. | 01:36 |
| clarkb | jeblair: fungi ^ | 01:36 |
| *** CaptTofu has joined #openstack-infra | 01:37 | |
| mrda | clarkb: just tried Firefox, doesn't work there either :-/ | 01:39 |
| clarkb | mrda: ya, I think this is an issue on the foundation side. http://www.openstack.org is dead | 01:39 |
| *** lcestari has joined #openstack-infra | 01:39 | |
| mrda | mrda: oh well, glad it's not me :) Thanks for your help. | 01:40 |
| clarkb | mrda: if jeblair and/or fungi show up I believe they may be able to look into the foundation stuff | 01:40 |
| mrda | clarkb: thanks | 01:40 |
| clarkb | mrda: the way CLA signing works is Gerrit collects the info from you, gpg encrypts it and POSTs it to a server at the foundation. The foundation does a 1:1 correlation based on name and email (or maybe just email) and if everythign lines up returns a 200 response to Gerrit | 01:41 |
| jhesketh | umm, openstack.org (the website at least) works for me | 01:42 |
| clarkb | jhesketh: interesting, it isn't up for me | 01:42 |
| jhesketh | yep, and I can log in | 01:42 |
| * clarkb tries the phone | 01:42 | |
| fungi | seems up | 01:42 |
| jhesketh | mrda: can you ping openstack.org? | 01:43 |
| clarkb | oh man derp | 01:43 |
| clarkb | I may be pebkacing /me checks a thing | 01:43 |
| clarkb | yup pebkac. proxies should diaf | 01:44 |
| clarkb | fungi: any other suggestions on what to look at? | 01:44 |
| fungi | i'll try resubmitting my contact info in gerrit to make sure the connector's not broken | 01:44 |
| fungi | but will take me a couple minutes | 01:44 |
| mrda | fungi: thanks | 01:44 |
| *** sarob has joined #openstack-infra | 01:49 | |
| lifeless | clarkb: you had a broken proxy ? | 01:49 |
| clarkb | lifeless: yes, I had neglected to remove my proxy settings after using the wire in the conference room all week | 01:50 |
| lifeless | :P | 01:50 |
| lifeless | clarkb: ah, not a broken proxy. Broken proxy config :) | 01:50 |
| *** sarob has quit IRC | 01:54 | |
| clarkb | though oddly my phone isn't happy with it either and that shouldn't be behind any proxy | 01:54 |
| clarkb | now it works I think chrome was taking a break | 01:56 |
| fungi | mrda: seems fine for me. after submitting i get... "Contact information last updated on Sep 23, 2013 at 1:54 AM." | 01:57 |
| *** ryanpetrello has quit IRC | 01:58 | |
| clarkb | fungi: do CLA signing errors end up in error_log? I see some sshd connection restsets and some tracebacks for melody but that is it | 01:58 |
| clarkb | probably worth tracking down which ssh peer is doing that | 01:59 |
| clarkb | but not tonight | 01:59 |
| mrda | fungi: Do you have any ideas on how I can proceed? Just tried again to no avail. | 01:59 |
| *** ericw has joined #openstack-infra | 02:00 | |
| fungi | well, there wouldn't be cla signing errors. possibly contact filing errors but the two are distinct and separate processes, they just happen to both be accessible from that same screen | 02:01 |
| fungi | you can still try them as separate steps | 02:01 |
| fungi | i notice that the submit button for submitting my contact information seems to have hung my browser after displaying the update notice | 02:03 |
| mrda | fungi: So I've signed up as a foundation member, now just wanting to sign the ICLA from gerrit. | 02:03 |
| fungi | not on a real computer at the moment, but i'm going to see if i can poke around in the database and confirm whether it updated | 02:04 |
| mrda | fungi: thanks | 02:04 |
| clarkb | fungi: I can check your account if that is easier | 02:04 |
| fungi | mrda: you can try agreeing without filling in the contact info on that screen, and then submit the contact info separately in your gerrit account settings page | 02:06 |
| *** w_ is now known as olaph | 02:06 | |
| mrda | fungi: cool, shall try. | 02:06 |
| fungi | clarkb: i can get to it, just typing speed here is at best 10wpm | 02:06 |
| mrda | fungi: same problem, even not filling in the contact info. The "Submit Agreement" button greys out, and it removes the "AGREE" text, but no ICLA agreement signed. | 02:08 |
| fungi | huh | 02:08 |
| clarkb | fungi: ya the failure mode seems to be different than any other related to the CLA that I have seen before | 02:11 |
| clarkb | particularly since the was no mix up in CLA selection according to the DB (account_agreements has no rows for account 8125) | 02:11 |
| fungi | well, the db does reflect my contact info update at least, so the connection to the foundation member system is intact | 02:12 |
| *** SlickNik_ has joined #openstack-infra | 02:13 | |
| *** Ryan_Lane has joined #openstack-infra | 02:13 | |
| mrda | clarkb, fungi: Just tried it via an iPhone. Same result. So I think it's not browser related. Has anyone else signed the contributor agreement today? | 02:17 |
| *** fbo_away has joined #openstack-infra | 02:17 | |
| *** enikanorov-w has joined #openstack-infra | 02:17 | |
| *** vishy has joined #openstack-infra | 02:17 | |
| *** jmeridth has joined #openstack-infra | 02:17 | |
| *** sungju has joined #openstack-infra | 02:18 | |
| *** sungju has quit IRC | 02:18 | |
| *** sarob has joined #openstack-infra | 02:19 | |
| *** sarob has quit IRC | 02:24 | |
| fungi | it's hard to say since that's not timestamped by gerrit | 02:26 |
| fungi | i'm looking to see if any lib updates happened in the past day or two | 02:27 |
| mikal | mrda: I wonder what the chrome developer tools network accesses thing says | 02:31 |
| fungi | no deb updates of note in the past week or so which could be involved | 02:34 |
| fungi | and last modified times on the gerrit warfile and its java deps are months back | 02:35 |
| mrda | fungi, clarkb, mikal: I found the problem. PEPKAC. It's "I AGREE", not "AGREE". Sorry for wasting your time. | 02:42 |
| fungi | mrda: ha! good catch | 02:42 |
| fungi | defocusing the input field with the wrong content will gray out that button | 02:43 |
| fungi | can't believe it didn't dawn on me that's what you were seeing | 02:44 |
| mrda | fungi: I'm not sure why my brain saw "enter I AGREE in the box to the left" as needing me to enter "AGREE" when it should have been "I AGREE". Thanks again! | 02:45 |
| fungi | no worries. glad it's sorted | 02:46 |
| clarkb | fungi: thanks for helping out | 02:47 |
| *** zehicle has joined #openstack-infra | 02:47 | |
| *** sarob has joined #openstack-infra | 02:50 | |
| fungi | yw | 02:52 |
| *** sarob has quit IRC | 02:55 | |
| *** amotoki has quit IRC | 02:56 | |
| *** blamar has quit IRC | 02:57 | |
| *** ericw has quit IRC | 02:58 | |
| *** SlickNik_ has quit IRC | 02:59 | |
| *** ericw has joined #openstack-infra | 02:59 | |
| *** ericw has quit IRC | 03:03 | |
| *** annegentle has quit IRC | 03:05 | |
| *** dims has quit IRC | 03:06 | |
| *** SergeyLukjanov has joined #openstack-infra | 03:06 | |
| *** annegentle has joined #openstack-infra | 03:08 | |
| *** CaptTofu has quit IRC | 03:11 | |
| *** CaptTofu has joined #openstack-infra | 03:12 | |
| *** ericw has joined #openstack-infra | 03:12 | |
| *** dcramer_ has quit IRC | 03:25 | |
| *** slong has joined #openstack-infra | 03:29 | |
| *** sarob has joined #openstack-infra | 03:50 | |
| *** sarob has quit IRC | 03:55 | |
| *** SergeyLukjanov has quit IRC | 04:17 | |
| *** michchap_ has joined #openstack-infra | 04:18 | |
| *** michchap has quit IRC | 04:21 | |
| *** michchap has joined #openstack-infra | 04:28 | |
| *** michchap_ has quit IRC | 04:31 | |
| *** michchap has quit IRC | 04:42 | |
| *** michchap has joined #openstack-infra | 04:45 | |
| *** alexpilotti has joined #openstack-infra | 04:46 | |
| *** annegentle has quit IRC | 04:49 | |
| *** sarob has joined #openstack-infra | 04:51 | |
| *** annegentle has joined #openstack-infra | 04:53 | |
| *** sarob has quit IRC | 04:55 | |
| *** afazekas has joined #openstack-infra | 05:01 | |
| *** UtahDave has quit IRC | 05:02 | |
| *** senk has joined #openstack-infra | 05:09 | |
| *** senk has quit IRC | 05:13 | |
| *** jhesketh has quit IRC | 05:20 | |
| openstackgerrit | A change was merged to openstack-infra/devstack-gate: Enable q-vpn service https://review.openstack.org/47335 | 05:24 |
| *** bearhands is now known as comstud | 05:35 | |
| *** amotoki has joined #openstack-infra | 05:47 | |
| *** sarob has joined #openstack-infra | 05:52 | |
| *** sarob has quit IRC | 05:56 | |
| *** yolanda has joined #openstack-infra | 06:07 | |
| *** senk has joined #openstack-infra | 06:10 | |
| *** Ryan_Lane has quit IRC | 06:13 | |
| *** senk has quit IRC | 06:14 | |
| *** SlickNik_ has joined #openstack-infra | 06:24 | |
| *** AJaeger has joined #openstack-infra | 06:33 | |
| *** Ryan_Lane has joined #openstack-infra | 06:43 | |
| *** jhesketh has joined #openstack-infra | 06:47 | |
| *** jhesketh__ has joined #openstack-infra | 06:47 | |
| *** xchu has joined #openstack-infra | 06:49 | |
| *** pblaho has joined #openstack-infra | 06:50 | |
| *** sarob has joined #openstack-infra | 06:52 | |
| *** sarob has quit IRC | 06:57 | |
| *** SlickNik_ has quit IRC | 06:57 | |
| *** fbo_away is now known as fbo | 07:12 | |
| *** dizquierdo has joined #openstack-infra | 07:13 | |
| *** yportnova has joined #openstack-infra | 07:19 | |
| *** jcoufal has joined #openstack-infra | 07:23 | |
| *** SergeyLukjanov has joined #openstack-infra | 07:32 | |
| *** flaper87|afk is now known as flaper87 | 07:32 | |
| *** adam_g has quit IRC | 07:42 | |
| *** adam_g has joined #openstack-infra | 07:42 | |
| *** adam_g has quit IRC | 07:42 | |
| *** adam_g has joined #openstack-infra | 07:42 | |
| *** sarob has joined #openstack-infra | 07:53 | |
| *** pabelanger has quit IRC | 07:53 | |
| *** jhesketh__ has quit IRC | 07:53 | |
| *** jhesketh__ has joined #openstack-infra | 07:55 | |
| *** pabelanger has joined #openstack-infra | 07:55 | |
| *** pabelanger has quit IRC | 07:55 | |
| *** pabelanger has joined #openstack-infra | 07:55 | |
| *** Ryan_Lane has quit IRC | 07:58 | |
| *** sarob has quit IRC | 07:58 | |
| *** kpavel has joined #openstack-infra | 08:04 | |
| *** kpavel_ has joined #openstack-infra | 08:05 | |
| *** dizquierdo has left #openstack-infra | 08:08 | |
| *** kpavel has quit IRC | 08:09 | |
| *** kpavel_ is now known as kpavel | 08:09 | |
| *** shardy_afk is now known as shardy | 08:10 | |
| *** Ryan_Lane has joined #openstack-infra | 08:28 | |
| *** Ryan_Lane has quit IRC | 08:32 | |
| *** dizquierdo has joined #openstack-infra | 08:39 | |
| *** boris-42 has joined #openstack-infra | 08:41 | |
| *** jhesketh has quit IRC | 08:52 | |
| *** sarob has joined #openstack-infra | 08:53 | |
| *** sarob has quit IRC | 08:57 | |
| *** Ryan_Lane has joined #openstack-infra | 08:58 | |
| *** mrda has quit IRC | 08:59 | |
| *** winston-d has quit IRC | 09:04 | |
| *** Ryan_Lane has quit IRC | 09:05 | |
| *** markmc has joined #openstack-infra | 09:11 | |
| *** alexpilotti has quit IRC | 09:14 | |
| *** ruhe has joined #openstack-infra | 09:29 | |
| *** sarob has joined #openstack-infra | 09:54 | |
| *** sarob has quit IRC | 09:58 | |
| *** SlickNik_ has joined #openstack-infra | 10:01 | |
| *** jcoufal has quit IRC | 10:04 | |
| *** ruhe has quit IRC | 10:05 | |
| *** ruhe has joined #openstack-infra | 10:11 | |
| *** mrda has joined #openstack-infra | 10:19 | |
| *** yaguang has quit IRC | 10:21 | |
| *** xchu has quit IRC | 10:22 | |
| *** pcm_ has joined #openstack-infra | 10:26 | |
| *** pcm__ has joined #openstack-infra | 10:30 | |
| *** pcm_ has quit IRC | 10:32 | |
| *** SergeyLukjanov has quit IRC | 10:36 | |
| *** ruhe has joined #openstack-infra | 10:50 | |
| *** hashar has joined #openstack-infra | 10:50 | |
| *** boris-42 has quit IRC | 10:51 | |
| *** sarob has joined #openstack-infra | 10:54 | |
| *** sarob has quit IRC | 10:59 | |
| *** SlickNik_ has quit IRC | 11:03 | |
| *** SergeyLukjanov has joined #openstack-infra | 11:05 | |
| *** hashar_ has joined #openstack-infra | 11:09 | |
| *** hashar has quit IRC | 11:10 | |
| *** hashar_ is now known as hashar | 11:10 | |
| *** dims has joined #openstack-infra | 11:12 | |
| *** w_ has joined #openstack-infra | 11:23 | |
| *** olaph has quit IRC | 11:25 | |
| *** ruhe has quit IRC | 11:27 | |
| *** pcm__ has quit IRC | 11:29 | |
| *** CaptTofu has quit IRC | 11:33 | |
| *** mkerrin has quit IRC | 11:37 | |
| *** CaptTofu has joined #openstack-infra | 11:42 | |
| *** CaptTofu has quit IRC | 11:45 | |
| *** CaptTofu has joined #openstack-infra | 11:46 | |
| *** ruhe has joined #openstack-infra | 11:46 | |
| *** ArxCruz has joined #openstack-infra | 11:46 | |
| sdague | so apparently we never released the gerrit lib with the bulk query defined? | 11:46 |
| sdague | any chance we could do that today, I'd actually like to use it in a tool for tempest | 11:47 |
| *** CaptTofu has quit IRC | 11:49 | |
| *** CaptTofu has joined #openstack-infra | 11:49 | |
| *** weshay has joined #openstack-infra | 11:49 | |
| *** CaptTofu has quit IRC | 11:51 | |
| *** CaptTofu has joined #openstack-infra | 11:51 | |
| *** sarob has joined #openstack-infra | 11:55 | |
| openstackgerrit | Sean Dague proposed a change to openstack-infra/gerritlib: fix param ordering to fix review function https://review.openstack.org/47821 | 11:59 |
| *** sarob has quit IRC | 11:59 | |
| *** cppcabrera has joined #openstack-infra | 12:04 | |
| *** mrda has quit IRC | 12:04 | |
| *** lcestari_ has joined #openstack-infra | 12:09 | |
| *** CaptTofu has quit IRC | 12:11 | |
| *** CaptTofu has joined #openstack-infra | 12:11 | |
| *** boris-42 has joined #openstack-infra | 12:13 | |
| *** thomasm has joined #openstack-infra | 12:24 | |
| *** rfolco has joined #openstack-infra | 12:27 | |
| *** thomasbiege has joined #openstack-infra | 12:31 | |
| *** adalbas has joined #openstack-infra | 12:34 | |
| *** thomasbiege has quit IRC | 12:35 | |
| *** dkliban has quit IRC | 12:36 | |
| fungi | i'm sure it's doable. we'll likely have enough people around for consensus once the earth spins a little more | 12:41 |
| *** zehicle has quit IRC | 12:42 | |
| sdague | silly earth roundness | 12:45 |
| sdague | ok, time to migrate to office anyway | 12:45 |
| sdague | I did manage to write a nice little script to run through old reviews and get jenkins to recheck on them, to expose merge conflicts | 12:45 |
| fungi | sdague: neat! | 12:47 |
| *** ruhe has quit IRC | 12:50 | |
| *** dims has quit IRC | 12:53 | |
| *** zehicle has joined #openstack-infra | 12:53 | |
| *** dims has joined #openstack-infra | 12:54 | |
| *** alexeyo has joined #openstack-infra | 12:57 | |
| *** sandywalsh has quit IRC | 12:57 | |
| *** alexeyo has quit IRC | 13:00 | |
| *** pcm_ has joined #openstack-infra | 13:00 | |
| *** alexeyo has joined #openstack-infra | 13:01 | |
| *** prad has joined #openstack-infra | 13:01 | |
| *** dprince has joined #openstack-infra | 13:03 | |
| *** prad has quit IRC | 13:05 | |
| *** dcramer_ has joined #openstack-infra | 13:05 | |
| alexeyo | hi all | 13:06 |
| alexeyo | I have been working on a scheduler redesign and stumbled upon the following problem: http://logs.openstack.org/67/45867/4/check/gate-nova-python26/087aba0/console.html | 13:06 |
| *** rustlebee is now known as russellb | 13:08 | |
| *** sandywalsh has joined #openstack-infra | 13:10 | |
| *** amotoki has quit IRC | 13:10 | |
| *** dkliban has joined #openstack-infra | 13:12 | |
| *** julim has joined #openstack-infra | 13:12 | |
| *** julim has quit IRC | 13:13 | |
| *** CaptTofu has quit IRC | 13:15 | |
| alexeyo | fungi hi | 13:15 |
| *** CaptTofu has joined #openstack-infra | 13:16 | |
| *** jhesketh__ has quit IRC | 13:16 | |
| alexeyo | apparently I can't add requirements anymore (link to log above) | 13:16 |
| *** mriedem has joined #openstack-infra | 13:17 | |
| alexeyo | is it an ok behavior? | 13:17 |
| *** CaptTofu has quit IRC | 13:17 | |
| *** julim has joined #openstack-infra | 13:17 | |
| *** mriedem has quit IRC | 13:17 | |
| *** CaptTofu has joined #openstack-infra | 13:17 | |
| *** mriedem has joined #openstack-infra | 13:18 | |
| sdague | alexeyo: you have to change global requirements first, you can't make requirements change in a project unless their are already matching lines in global requirements | 13:18 |
| *** dprince has quit IRC | 13:19 | |
| alexeyo | I see, so I have to make a regular commit to openstack/requirements with my would-be requirement first, right? | 13:21 |
| fungi | alexeyo: that's been the case throughout the entire havana development cycle as well as at least part of the grizzly cycle. also, getting additions to the global requirements list approved is fairly unlikely to happen between now and when we start work on the icehouse release | 13:21 |
| *** mriedem has quit IRC | 13:22 | |
| *** ryanpetrello has joined #openstack-infra | 13:22 | |
| alexeyo | okay, so it has to be approved to be a global requirement first and only after that I could add it to project's reqs, do i get this rigth? | 13:23 |
| *** mriedem has joined #openstack-infra | 13:23 | |
| mordred | morning | 13:23 |
| fungi | the hope is that we give downstream packagers and integrators a period of stability in our dependency lists throughout the release candidate period, and also free up developers to fix release-critical bugs rather than working on new features | 13:23 |
| fungi | alexeyo: yes, the requirements reviewers, in addition to making sure it's a safe time in the development cycle to add/change the requirements list (early after the last release is generally better), the additions are also checked for licensing conformance, release versioning stability and other sorts of concerns | 13:24 |
| *** jcoufal has joined #openstack-infra | 13:24 | |
| mordred | alexeyo: zigo has expressed interest in getting the sqlalchemy req changed, and zul has confirmed that ubuntu is ok with 0.8 | 13:25 |
| mordred | I have no idea if that will actually get changed this cycle | 13:25 |
| *** sarob has joined #openstack-infra | 13:25 | |
| boris-42 | mordred imho I very worried about switching to sqla0.8 | 13:25 |
| zul | i think we need to have a session at ODS for the requirements project | 13:25 |
| alexeyo | I see | 13:26 |
| boris-42 | mordred I wrote a lot of sqla hacks so there could be a lot of hidden bugs that won't be tested by our current tests.. | 13:26 |
| fungi | zul: that'll make at least three summits in a row we've had a session about the requirements project. almost as if it's taken on a life of its on | 13:26 |
| boris-42 | mordred we should test all this stuff by hands before.. | 13:27 |
| zul | fungi: yeah i think we need something written down for it | 13:27 |
| alexeyo | so in case one needs a new requirement for a new feature where should he/she start? | 13:27 |
| zul | yeah | 13:27 |
| zul | alexeyo: like processes stuff | 13:28 |
| fungi | boris-42: well, at the moment my crude test of sqla 0.8.x is definitely not happy in devstack-tempest jobs, no idea about unit tests even... https://review.openstack.org/47745 | 13:28 |
| alexeyo | zul: sorry, didn't get this one | 13:28 |
| *** changbl has quit IRC | 13:29 | |
| boris-42 | fungi we planned migration to sqla 0.8 after all projects switch to oslo-db code | 13:29 |
| *** tvb has joined #openstack-infra | 13:29 | |
| boris-42 | fungi end when we test all by hands and fix all bugs | 13:29 |
| *** SergeyLukjanov has quit IRC | 13:29 | |
| *** tvb has quit IRC | 13:29 | |
| *** tvb has joined #openstack-infra | 13:29 | |
| zul | alexeyo: things like when would it not be ok to change requirements | 13:29 |
| *** dkranz has joined #openstack-infra | 13:29 | |
| *** sarob has quit IRC | 13:29 | |
| boris-42 | fungi so we should avoid bumping version of sqla at this moment imho =) | 13:30 |
| fungi | boris-42: oh, i entirely agree--the change i uploaded was purely a test to see where things stand with regard to integration tests. i have no vested interest there | 13:30 |
| alexeyo | what I am trying to say is that if one decides to build a new feature and needs an extra package to make it work then it is not clear how to start with integrating this feature | 13:31 |
| alexeyo | especially if it is the only one to use this package | 13:31 |
| alexeyo | (at least for now) | 13:31 |
| fungi | alexeyo: general recommendations are, 1. follow the openstack-announce and openstack-dev mailing lists so you know what's going on (release freezes for example), 2. submit a change to the project you're working on which includes that new requirement and leave a comment on that change that requirements tests won't pass until the requirement gets added to the global list, 3. if core reviewers on that | 13:32 |
| fungi | project are in favor of your change submit a corresponding change to openstack/requirements noting which change needs it | 13:32 |
| *** tvb has quit IRC | 13:33 | |
| alexeyo | while there is no feature for which a requirement is needed it is pointless to approve it, while requrements are not satisfied the feature won't pass the gate | 13:33 |
| *** bnemec has joined #openstack-infra | 13:33 | |
| *** matty_dubs|gone is now known as matty_dubs | 13:33 | |
| sdague | alexeyo: global requirements are there to ensure that anything that openstack might depend on is something the distros are able to package, and passes sanity checking around licensing and maintainability | 13:34 |
| *** SergeyLukjanov has joined #openstack-infra | 13:34 | |
| mordred | boris-42: I agree with you - but apparently ubuntu and debian have both upgraded themselves already, so I'm also a little worried about what happens for people trying to run on ubuntu from packages | 13:34 |
| boris-42 | mordred fuuuu* | 13:34 |
| boris-42 | =) | 13:34 |
| mordred | :) | 13:34 |
| alexeyo | fungi: thanks, it is becoming much clearer now | 13:34 |
| fungi | alexeyo: also if the requirement change is a new dependency, then it's worth discussing on the openstack-dev ml well ahead of time to see whether there are any likely concerns, particularly if there's something unusual about it | 13:34 |
| zul | mordred: we tested it and reported stuff that we see and even helped fixed them | 13:35 |
| sdague | zul: then why can't keystone start on 0.8? | 13:35 |
| sdague | you guys have a patch for that? | 13:35 |
| zul | sdague: it does for us | 13:35 |
| sdague | it seems like a big disconnect | 13:35 |
| zul | it does im not exactly sure debian' | 13:36 |
| zul | debian's problem | 13:36 |
| zul | (new keyboard) | 13:36 |
| sdague | zul: well it doesn't with pure upstream, so I'm curious what's been patched to make it work | 13:36 |
| *** rpodolyaka has joined #openstack-infra | 13:36 | |
| zul | sdague: im not sure | 13:36 |
| *** blamar has joined #openstack-infra | 13:37 | |
| sdague | that's the logger head that we're add right now, the fact that we can't start keystone on 0.8, but the distros can, means there is a difference that needs to get resolved | 13:37 |
| fungi | zul: it doesn't work on our devstack-tempest tests (see http://logs.openstack.org/45/47745/1/check/gate-tempest-devstack-vm-full/7aa67c7/console.html for example) | 13:37 |
| sdague | it's an import issue | 13:37 |
| sdague | from what I can tell | 13:37 |
| zul | sdague: maybe we have a patch in the sqlachemy package | 13:37 |
| mordred | gah | 13:37 |
| sdague | zul: that's entirely possible | 13:37 |
| * mordred punches someone | 13:37 | |
| mordred | the distro-patches-upstream thing is vastly overused, btw | 13:38 |
| zul | mordred: so sue me ;) | 13:38 |
| mordred | zul: lawsuit in mail | 13:38 |
| sdague | zul: anyone from canonical able to help unwind this? us doing archaeology on debs without a guide isn't the fastest way to resolve | 13:38 |
| sdague | mordred: it could have been a debian patch as well that was inherited | 13:38 |
| mordred | sdague: yup. I believe all the distros are more or less equally to blame in the general state of things there | 13:39 |
| *** DennyZhang has joined #openstack-infra | 13:39 | |
| fungi | zul: http://logs.openstack.org/45/47745/1/check/gate-tempest-devstack-vm-full/7aa67c7/logs/screen-key.txt.gz "CRITICAL log logging_excepthook cannot import name exceptions" | 13:39 |
| zul | sdague: fastest way is to install ubuntu precise in a vm with the cloud archive and run your tempest tests | 13:39 |
| fungi | that may be a red herring tho | 13:39 |
| zul | fungi: yeah thats an api change with sqlalchemy | 13:39 |
| rpodolyaka | ...looks like I'm late for the party... are you guys looking for a particular patch which made sqlalchemy-migrate work with sqlalchemy 0.8.x? | 13:40 |
| fungi | zul: on that note! we just tried to upgrade to using cloud archive over the weekend | 13:40 |
| zul | fungi: oh? | 13:40 |
| zul | did it blow up? | 13:40 |
| fungi | zul: libvirt 1.x was a bit of a problem for us | 13:40 |
| zul | fungi: how so? | 13:40 |
| fungi | splodey | 13:40 |
| fungi | we had to do an emergency rollback to get devstack working again | 13:40 |
| mordred | rpodolyaka: we're trying to figure out why sqla 0.8 works if you install everything from distro packages, and doesn't if you install from pip | 13:41 |
| mordred | rpodolyaka: we're surmising that there is a patch in the distro packages somewhere | 13:41 |
| *** w_ is now known as olaph | 13:41 | |
| sdague | fungi: actually, on that front, you got a copy of the bum image somewhere? I'd like to exactly replicate it if possible | 13:41 |
| rpodolyaka | mordred: Ubuntu (or Fedora) guys patched it. it's really small. I'll post a link to it in a few minutes | 13:41 |
| fungi | sdague: we deleted the image, but even if we did have, our "openstack" providers don't know how to make a public glance yet (well, it's apparently in progress in both of them but not officially ready yet) | 13:42 |
| zul | mordred: this is compat patch for sqlachemy migrate fql sqla 8.0 http://summit.openstack.org/cfp/details/109 | 13:42 |
| rpodolyaka | mordred: https://github.com/stackforge/sqlalchemy-migrate/commit/74ccf7a39762020671bf73820bae030d4b28bc96 | 13:42 |
| zul | oops no it isnt | 13:42 |
| zul | http://paste.ubuntu.com/6145576/ | 13:42 |
| rpodolyaka | mordred: so we already have it master of sqlalchemy-migrate | 13:42 |
| rpodolyaka | mordred: and this one too https://github.com/stackforge/sqlalchemy-migrate/commit/2ff11c176af15cae3a5ea0223ee3fc236f1eab13 | 13:43 |
| fungi | sdague: btw, do you have links to any of the changes/logs for the failures we saw? i didn't think to make a note of any for posterity | 13:43 |
| mordred | ah! wow, I love it when there's folks who actually know what's going on! | 13:44 |
| rpodolyaka | :) | 13:44 |
| mordred | so - step one is "release a new sqlalchemy-migrate" - step two is "tests sqla 0.8 again until boris-42 is happy" - step three is "bump requirements" - step four is "profit" ? | 13:45 |
| boris-42 | step 4 ..... | 13:45 |
| boris-42 | step 5 PROFITTT | 13:45 |
| boris-42 | =) | 13:45 |
| mordred | boris-42: :) | 13:46 |
| fungi | sdague: if not, no worry, i can probably dig one up | 13:46 |
| sdague | fungi: I made a bug, but launchpad's not letting me log in right now | 13:46 |
| sdague | so finding it... more difficult | 13:46 |
| zul | boris-42: https://bugs.launchpad.net/cinder/+bug/1206561 (cinder sqla8 bug) | 13:47 |
| uvirtbot | Launchpad bug 1206561 in cinder "Migration tests fail with sqlalchemy 0.8" [Undecided,New] | 13:47 |
| boris-42 | zul ouch=) | 13:47 |
| fungi | sdague: no worries, i just realized my sqla req bump triggered it too | 13:48 |
| fungi | zul: http://logs.openstack.org/45/47745/1/check/gate-tempest-devstack-vm-full/f926000/console.html "Server bac84e47-cc28-46ea-ba71-3f740bdee508 failed to reach ACTIVE status within the required time (400 s)." | 13:49 |
| *** alexeyo has quit IRC | 13:49 | |
| fungi | zul: apparently libvirt was dying behind the scenes | 13:49 |
| zul | fungi: you have the libvirt log for it? :) | 13:50 |
| zul | /var/log/libvirtd | 13:50 |
| sdague | zul: I don't think we capture that | 13:50 |
| fungi | zul: i'm looking but i don't theink | 13:50 |
| fungi | what sdague said | 13:50 |
| zul | ok | 13:50 |
| sdague | n-cpu just stops | 13:50 |
| *** rcleere has quit IRC | 13:50 | |
| *** DennyZhang has quit IRC | 13:50 | |
| *** DennyZhang has joined #openstack-infra | 13:51 | |
| *** mrmartin has joined #openstack-infra | 13:51 | |
| sdague | ok, is launchpad not letting anyone else in? | 13:51 |
| fungi | sdague: working for me | 13:51 |
| fungi | sdague: what project did you open the bug against? | 13:52 |
| sdague | nova | 13:52 |
| sdague | https://bugs.launchpad.net/nova/+bug/1228977 | 13:53 |
| uvirtbot | Launchpad bug 1228977 in nova "n-cpu seems to crash when running with libvirt 1.0.6 from ubuntu cloud archive" [High,New] | 13:53 |
| sdague | it's in the recheck list | 13:53 |
| fungi | aha, bug 1228977 | 13:53 |
| uvirtbot | Launchpad bug 1228977 in nova "n-cpu seems to crash when running with libvirt 1.0.6 from ubuntu cloud archive" [High,New] https://launchpad.net/bugs/1228977 | 13:53 |
| fungi | oh, you just linked it | 13:53 |
| zul | wait why are you guys using 1.0.6? | 13:53 |
| fungi | zul: that's what installed as soon as we added ubuntu cloud archive to our sources | 13:54 |
| zul | havana-proposed has 1.1+ though? | 13:54 |
| openstackgerrit | Marton Kiss proposed a change to openstack-infra/config: Groups-dev drupal drush update to 6.0.0 https://review.openstack.org/47838 | 13:55 |
| sdague | zul: it's whatever was in havana/updates | 13:55 |
| fungi | zul: perhaps https://review.openstack.org/45896 was wrong? | 13:55 |
| fungi | yeah, precise-updates/havana | 13:55 |
| *** michchap has quit IRC | 13:56 | |
| zul | i dunno maybe you guys might want to used havana-proposed | 13:56 |
| *** michchap has joined #openstack-infra | 13:56 | |
| zul | or precise precise-proposed | 13:57 |
| fungi | zul: i think part of the goal is to test our changes against the libs ubuntu has already tested and approved | 13:57 |
| zul | ah ok | 13:57 |
| mordred | yes. that is correct | 13:57 |
| fungi | so it might make more sense to wait until what's in precise-updates/havana works with our tests, and then start using it going forward | 13:57 |
| mordred | (I really love this particular chicken and egg) | 13:58 |
| sdague | zul: there is also the issue of asymetry, because if packages are dropping into there without having proven themselves in our gate first, then we can have what happened this weekend | 13:59 |
| sdague | where an ubuntu package update blocks all code merge for 16 hours | 13:59 |
| *** michchap has quit IRC | 13:59 | |
| *** michchap has joined #openstack-infra | 14:00 | |
| sdague | ok... so lp login isn't working today, and lp is being extra slow in sending reset email | 14:00 |
| sdague | that makes for a bad way to start a monday | 14:01 |
| fungi | sdague: granted, we already have that (precise security updates could in theory cause it too) | 14:01 |
| *** CaptTofu has quit IRC | 14:01 | |
| sdague | fungi: yes, though the risk surface on that is less | 14:01 |
| *** CaptTofu has joined #openstack-infra | 14:02 | |
| fungi | and even with some dependency caps in place, it doesn't entirely stop new/broken packages on pypi from causing the same | 14:02 |
| sdague | it would be good to make sure the image can run devstack gate before it promotes | 14:02 |
| sdague | fungi: sure | 14:02 |
| fungi | sdague: we do already perform a sanity check on the candidate image before it gets used, so that might just be a question of extending what constitutes sanity | 14:03 |
| *** tvb has joined #openstack-infra | 14:05 | |
| *** tvb is now known as Guest72636 | 14:05 | |
| sdague | fungi: ok, cool, where is that in code base? | 14:05 |
| fungi | sdague: currently looking for it in nodepool | 14:06 |
| *** Guest72636 has quit IRC | 14:06 | |
| mordred | sdague: we've talked about doing that in the future when we're dib based for images | 14:07 |
| sdague | mordred: cool | 14:07 |
| mordred | sdague: I'm mildly concerned about making it too strong though | 14:07 |
| mordred | sdague: because we _want_ to know when upstream breaks us | 14:08 |
| mordred | and if we just stop promoting images | 14:08 |
| mordred | no one will notice | 14:08 |
| sdague | mordred: well we need signalling on it | 14:08 |
| sdague | so we know there was a fail | 14:08 |
| mordred | so far, our track record on watching signallying not related to code reviews is bad :( | 14:08 |
| fungi | sdague: right now here's what goes on for an image update... https://git.openstack.org/cgit/openstack-infra/nodepool/tree/nodepool/nodepool.py#n340 | 14:08 |
| openstackgerrit | Marton Kiss proposed a change to openstack-infra/config: Groups-dev distbuild checked wrong Git repository commit hash https://review.openstack.org/47844 | 14:10 |
| fungi | sdague: while a full tempest run is probably something which would need more work to integrate, some devstack-specific sanity checks in https://git.openstack.org/cgit/openstack-infra/config/tree/modules/openstack_project/files/nodepool/scripts/prepare_devstack.sh might help? | 14:13 |
| fungi | or https://git.openstack.org/cgit/openstack-infra/config/tree/modules/openstack_project/files/nodepool/scripts/devstack-cache.py | 14:13 |
| fungi | anything which didn't make the resultant image unsuitable for reuse is potentially a candidate | 14:14 |
| *** rnirmal has joined #openstack-infra | 14:15 | |
| *** ruhe has joined #openstack-infra | 14:19 | |
| *** rcleere has joined #openstack-infra | 14:22 | |
| *** CaptTofu has quit IRC | 14:22 | |
| *** CaptTofu has joined #openstack-infra | 14:22 | |
| mordred | fungi: https://review.openstack.org/47848 | 14:24 |
| mordred | fungi: there's the basic/simple version of the tox update we want for everyhone | 14:24 |
| *** CaptTofu has quit IRC | 14:25 | |
| fungi | excellent | 14:25 |
| *** rcleere has quit IRC | 14:25 | |
| *** rcleere has joined #openstack-infra | 14:26 | |
| *** sarob has joined #openstack-infra | 14:26 | |
| fungi | that seems to jive with the couple of reviews of yours i already went through | 14:26 |
| *** CaptTofu has joined #openstack-infra | 14:26 | |
| *** rcleere has quit IRC | 14:27 | |
| *** markmcclain has joined #openstack-infra | 14:28 | |
| *** dprince has joined #openstack-infra | 14:29 | |
| *** sarob has quit IRC | 14:30 | |
| *** CaptTofu has quit IRC | 14:31 | |
| *** dkranz has quit IRC | 14:31 | |
| *** CaptTofu has joined #openstack-infra | 14:32 | |
| jd__ | mordred: around? | 14:32 |
| mordred | jd__: always | 14:34 |
| sdague | mordred: yeh, so that's probably a good summit session. A better way to get a handle on status of some non code review initiated changes | 14:34 |
| mordred | sdague: ++ | 14:34 |
| jd__ | mordred: we just realized that the MANIFEST.in from Horizon is bugged | 14:35 |
| jd__ | mordred: I think nobody noticed because pbr use git ls-files rather than MANIFEST.in if sdist is run from a git clone | 14:35 |
| jd__ | mordred: I wonder if it shouldn't be better to always pick MANIFEST.in if it's there | 14:35 |
| mordred | uhm. one sec - let me llok so I have better context | 14:35 |
| jd__ | mordred: pbr/packaging.py around line 545 | 14:36 |
| mordred | oh! | 14:36 |
| mordred | no, I know that | 14:36 |
| mordred | I wanted to see horizon's manifest.in | 14:36 |
| mordred | that's a crazy MANIFEST.in file | 14:36 |
| jd__ | somehow | 14:36 |
| mordred | why don't we just remove it? | 14:36 |
| *** zehicle has quit IRC | 14:37 | |
| mordred | or, replace it just with one that has exclude openstack_dashboard/local/local_settings.py | 14:37 |
| mordred | since that seems to be the only thing locally that would need to be added in addition to what pbr does | 14:37 |
| jd__ | mordred: I think we could remove it | 14:37 |
| jd__ | that's the fix for Horizon | 14:37 |
| jd__ | the other question that remains to me, is pbr having a good behaviour prefering git ls-files over MANIFEST.in, in general | 14:38 |
| mordred | in any case - pbr _should_ use both | 14:38 |
| jd__ | ah :) | 14:38 |
| jd__ | that's a valid answer to that remaining question | 14:38 |
| mordred | but it's going to process them in a sequence | 14:38 |
| mordred | and it's possible that we're doing them in a bad order | 14:38 |
| mordred | in terms of processing things like exclude + include lines | 14:38 |
| mordred | jd__: what was the bug you uncovered in horizon? like, which thing happened or didn't happen? | 14:39 |
| jd__ | mordred: horizon/conf/__init__.py file isn't included for example | 14:39 |
| mordred | jd__: wow, really? | 14:40 |
| jd__ | yup | 14:40 |
| jd__ | you can try easily by copying your git clone, rm -rf .git and run python setup.y sdist | 14:40 |
| mordred | oh! no. | 14:41 |
| mordred | that's not supported | 14:41 |
| mordred | at | 14:41 |
| mordred | all | 14:41 |
| mordred | never has been | 14:41 |
| mordred | don't intend for it to be | 14:41 |
| mordred | if that's a thing that people want to support, it's a complete design change and would require a decently large amount of work | 14:41 |
| jd__ | what? | 14:41 |
| *** dafter has joined #openstack-infra | 14:42 | |
| jd__ | I don't follow you | 14:42 |
| mordred | we do not expect python setup.py sdist to work outside of a git repo | 14:42 |
| jd__ | running sdist while there's a MANIFEST.in is supported AFAIU | 14:42 |
| mordred | not by OpenStack it's not | 14:42 |
| *** dafter is now known as tvb | 14:42 | |
| mordred | (I understand it's supported by python itself) | 14:42 |
| jd__ | mordred: this isn't nice to packages :) | 14:42 |
| jd__ | packagers | 14:42 |
| *** tvb has quit IRC | 14:42 | |
| *** tvb has joined #openstack-infra | 14:42 | |
| mordred | sure it is | 14:42 |
| mordred | in fact | 14:43 |
| mordred | it's exactly in line with how packaging is supposed to work | 14:43 |
| mordred | packagers are supposed to consume the tarball that upstream produces | 14:43 |
| mordred | they are NOT supposed to produce their own | 14:43 |
| jd__ | fair enough | 14:43 |
| *** dkranz has joined #openstack-infra | 14:43 | |
| *** datsun180b has joined #openstack-infra | 14:43 | |
| mordred | and - if they want to use something like git-buildpackage - I strongly recommend the --pristine-tarball optiion and git import-tarball | 14:44 |
| mordred | scuse me - import-orig | 14:44 |
| * mordred needs to write a specific write up on this - it keeps coming up | 14:44 | |
| jd__ | mordred: agreed, talking with zigo | 14:44 |
| jd__ | about this | 14:44 |
| mordred | jd__: if there are ways in which I can better help packaging workflow though, I definitely want to do them - there is a 100% chance there are things I have not thought about :) | 14:46 |
| *** pcrews has joined #openstack-infra | 14:46 | |
| jd__ | mordred: well having the ability to build the SOURCES.txt correctly from scratch would be nice I think | 14:47 |
| jd__ | and it's not that far away | 14:47 |
| mordred | jd__: is SOURCES.txt not doing the right thing? | 14:47 |
| jd__ | mordred: well it depends heavily on the context you are building it | 14:47 |
| fungi | part of the problem here, particularly with some of the gbp howtos, is that they're written to help package maintainers cope with upstreams who don't produce releases and/or don't generate tarballs, and so they assume workflows which are counter to what we're doing in openstack | 14:47 |
| jd__ | like I said, in Horizon depending on you being in a git clone or not, MANIFEST.in is taking into account or not and does not generate the same | 14:48 |
| mordred | jd__: I have horizon/conf/__init__.py in my SOURCES.txt when I make a tarball | 14:48 |
| mordred | oh - you still mean outside of a git clone | 14:48 |
| jd__ | yes :) | 14:48 |
| mordred | you should never be running commands in packaging that regenerate that | 14:48 |
| jd__ | for example that works fine with Ceilometer, but not Horizon | 14:48 |
| mordred | really | 14:48 |
| jd__ | mordred: yeah, YMMV I guess, though using upstream tarballs is fine with me | 14:49 |
| mordred | ah - ceilometer seems to be maintaining a MANIFEST.in file | 14:49 |
| jd__ | I just need to convince zigo :) | 14:49 |
| mordred | we gave up on maintaining those in favor of git about 1.5 years ago because the tarballs would consistently have things missing | 14:49 |
| mordred | jd__: I agree with what fungi said above - the 'right' way to do packaging involves consuming upstream tarballs if they exist | 14:50 |
| fungi | gbp has plenty of support for consuming upstream tarballs, so i still can't figure out why there's been such resistance to using that in the debian openstack packaging workflow | 14:50 |
| mordred | that's the reason that debian has a .orig.tar.gz in the set of files associated with it | 14:50 |
| mordred | the only reason gbp has support for the other thing is for hipster upstreams who do not release anything | 14:50 |
| mordred | if the packager produces his own orig.tar.gz, then there is no way to verify that the contents are what they are claimed to be | 14:51 |
| mordred | debian's .orig.tar.gz shoudl be binary == to openstack's released .tar.gz | 14:51 |
| mordred | it becomes even more important when we start signing our tarballs | 14:52 |
| fungi | i'm particularly concerned that by not using our tarballs, debian is carrying source packages with regenerated orig tarballs which do not match checksums in our release announcements or on our tarballs server/pypi/lp/et cetera | 14:52 |
| *** rcleere has joined #openstack-infra | 14:52 | |
| *** fbo is now known as fbo_away | 14:52 | |
| mordred | fungi: exactly | 14:52 |
| mordred | it's a giant bug | 14:52 |
| fungi | what mordred said | 14:52 |
| mordred | and should be fixed | 14:52 |
| *** zehicle has joined #openstack-infra | 14:52 | |
| mordred | with haste | 14:52 |
| mordred | because otherwise the entire design of debian packaging is being subverted | 14:52 |
| fungi | and i really suspect the culprit is blind following of some howto which told someone how to use gbp with $random_project which doesn't publish tarballs | 14:53 |
| * mordred needs to write a howto doc | 14:53 | |
| *** dizquierdo has quit IRC | 14:54 | |
| * fungi goes to check what's landed in sid/jessie recently | 14:55 | |
| *** mrmartin has quit IRC | 14:56 | |
| zul | fyi we use upstream tarballs for our releases ;) | 14:57 |
| *** zehicle has quit IRC | 14:59 | |
| fungi | oh, for starters, they're repacking our tarballs with xzi | 15:00 |
| fungi | p | 15:00 |
| mordred | why? | 15:00 |
| fungi | http://ftp.de.debian.org/debian/pool/main/n/nova/nova_2013.1.3.orig.tar.xz | 15:00 |
| mordred | sigh | 15:00 |
| fungi | because $shiny_new_thing | 15:00 |
| mordred | fungi: sorry. I forgot. gzip is terrible | 15:01 |
| mordred | fungi: should we release .tar.bz2 and .tar.xz to maybe prevent people from doig that? | 15:01 |
| *** sarob has joined #openstack-infra | 15:01 | |
| jeblair | don't forget rar | 15:01 |
| fungi | admittedly it's about 50% the size of our tarball for nova, but still that's tiny to start with | 15:01 |
| *** tvb has quit IRC | 15:02 | |
| fungi | i would not be opposed to having our tarball jobs start publishing .tar.xz versions and including their checksums in future release announcements if ttx doesn't object | 15:02 |
| *** tvb has joined #openstack-infra | 15:03 | |
| *** sarob_ has joined #openstack-infra | 15:03 | |
| *** tvb is now known as Guest82175 | 15:03 | |
| *** thedodd has joined #openstack-infra | 15:04 | |
| *** sarob has quit IRC | 15:04 | |
| fungi | current consensus is there's little point to releasing .tar.bz2 if you also do .tar.xz and .tar.gz since xzip performs at least as well as bzip2 under most circumstances and for everyone else who doesn't have $shiny_new_tool there's still your tried-and-true gzip version | 15:04 |
| ttx | fungi: they should at least keep the same .tar inside the recompressed thing | 15:04 |
| *** Guest82175 is now known as tvb | 15:04 | |
| ttx | i.e. if you take the signed .gz and their thing and uncompress it, you should end up with the same .tar | 15:05 |
| *** tvb is now known as Guest31682 | 15:05 | |
| ttx | That's been the rule for recompressing in debian packaging | 15:05 |
| *** changbl has joined #openstack-infra | 15:05 | |
| ttx | so that they don't lose the verification chain | 15:06 |
| fungi | ttx: yeah, they don't. (just compared checksums for unxz nova_2013.1.3.orig.tar.xz and gunzip nova-2013.1.3.tar.gz) | 15:06 |
| ttx | now *that* would be bad debian packaging. | 15:07 |
| * ttx with his debian packager hat on | 15:07 | |
| fungi | their nova tarball, uncompressed, is also much smaller than ours | 15:07 |
| fungi | diffing the contents now to see what's left out | 15:07 |
| pabelanger | fungi, I though dfsg was to be append if modifying the upstream tarball? | 15:09 |
| mordred | fungi: that's probably because they're actually not using our tarball at all, but are re-generating it from source incorrectly, afaict | 15:09 |
| *** sarob_ has quit IRC | 15:09 | |
| mordred | pabelanger: I believe you are correct | 15:09 |
| mordred | because the only valid reason in debian for modifying the upstream tarball is to fix dfsg violations | 15:09 |
| pabelanger | mordred, yes, that is my understanding also | 15:09 |
| fungi | pabelanger: that's convention when the tarball is repacked specifically for compliance with debian free software guidelines | 15:10 |
| mordred | jd__: you may have just awakened the beast which is a channel full of old-school debian packagers | 15:10 |
| pabelanger | fungi, Oh, I see | 15:10 |
| fungi | there are actually other circumstances where the package maintainer might need to repack a tarball, but generally it should be avoided unless those are very, very good reasons | 15:10 |
| pabelanger | fungi, agreed | 15:11 |
| jd__ | mordred: :-) | 15:11 |
| fungi | and it's still convention to at least modify the name of the orig tarball when doing that, yes | 15:11 |
| Shrews | mordred: this change (http://lists.openstack.org/pipermail/openstack-dev/2013-September/015495.html) should be made to stackforge project too, i assume? | 15:11 |
| mordred | Shrews: yes. it will make everyone happier | 15:11 |
| fungi | pabelanger: and also a get-orig-source target is good in the debian/rules to make it easy for others to repeat the repacking process deterministically | 15:11 |
| mordred | Shrews: if it doesn't make you happier, you're wrong | 15:12 |
| fungi | pabelanger: and a readme.source file to describe precisely why it was repacked | 15:12 |
| *** zigo_ has joined #openstack-infra | 15:13 | |
| fungi | okay, so it looks like the debian regenerated nova 2013.1.3 tarball omits our AUTHORS, ChangeLog, PKG-INFO and nova.egg-info while including .gitignore and .gitreview, and also strips a blank line from the end of setup.cfg | 15:18 |
| *** reed has joined #openstack-infra | 15:18 | |
| fungi | basically, didn't run setup.py sdist | 15:18 |
| mordred | yup | 15:19 |
| pabelanger | Do we know why they are regenerating the tarball or is it just a static part of there process? | 15:19 |
| mordred | I consider that a bug | 15:19 |
| mordred | pabelanger: I believe they are following the "how to consume an upstream git repo" process, rather than the "how to consume an upstream tarball" process | 15:19 |
| jd__ | you can ask zigo_ directly :) | 15:20 |
| mordred | hey zigo_ | 15:20 |
| mordred | we're talking smack about you | 15:20 |
| pabelanger | mordred, Ah, ya, that would do it | 15:20 |
| mordred | but it's really because we'd like to help | 15:20 |
| zigo_ | mordred: Adding *.py in the "recursive-include horizon" line of the MANIFEST.in fixed our problems. | 15:20 |
| zigo_ | mordred: I believe that's a bug in the original MANIFEST.in of horizon, which nobody really uses. | 15:21 |
| *** mgagne has joined #openstack-infra | 15:22 | |
| mordred | zigo_: it is not a bug in the original MANIFEST.in, it's a bug in the debian packaging | 15:22 |
| mordred | zigo_: what can I do to move you away from not using our tarballs? | 15:22 |
| fungi | zigo_: the bigger question is why you're distributing tarballs in your source packages which claim to be original upstream tarballs, but aren't | 15:22 |
| mordred | right. what fungi said | 15:22 |
| zigo_ | fungi: I haven't claim such thing. | 15:23 |
| mordred | zigo_: nova_2013.1.3.orig.tar.xz | 15:23 |
| zigo_ | mordred: And? | 15:23 |
| mordred | zigo_: is not, in fact, related to nova_2013.1.3.tar.gz | 15:23 |
| fungi | zigo_: the tarball does not include a readme.source explaining why you're not using upstream tarballs, nor a get_orig_source target in debian/rules | 15:23 |
| mordred | it's contents are wrong, for one | 15:23 |
| zigo_ | Probably switching to .xz compression would be a good start. Removing the ChangeLog which is huge, and which the FTP masters *refused* would also help. | 15:24 |
| fungi | and yes, most importantly, it lacks files generated by our sdist and the checksums are obviously different from what we sign and attest to in release announcements | 15:24 |
| fungi | ftpmaster said our ChangeLog file was unsuitable for inclusion in the source tarball? | 15:24 |
| fungi | why, specifically? | 15:24 |
| flaper87 | dhellmann: ping | 15:25 |
| zigo_ | fungi: For a binary of let's say 30KB, it adds about 1.5 MB of changelog. | 15:25 |
| dhellmann | flaper87: pong | 15:25 |
| zigo_ | Multiply this by the number of .deb, and you get a quite bad result. | 15:25 |
| zigo_ | Also, dealing with pristine-tar branch and upstream branch is *very* annoying. | 15:25 |
| fungi | zigo_: just because it's in the orig tarball doesn't mean it needs to get included in the packages | 15:26 |
| zigo_ | With tags, I got everything contained in a single packaging branch. | 15:26 |
| zigo_ | That's a lot more convenient. | 15:26 |
| mordred | zigo_: but it's the right thing to do if upstream has an orig tarball | 15:26 |
| zigo_ | Well, it's what *you* think is right! :) | 15:26 |
| *** Guest31682 is now known as tvb | 15:26 | |
| fungi | as upstream, yes | 15:26 |
| zigo_ | Explain to me why it's not valid to use what's in the Git... | 15:26 |
| *** tvb has quit IRC | 15:26 | |
| *** tvb has joined #openstack-infra | 15:26 | |
| mordred | you aren't using it right | 15:26 |
| flaper87 | dhellmann: hey, I wanted to know if we could let those 3 requirement patches related to marconi land. The 3 of them are blocking other patches and that's going to slow down the work until ith development cycle begins | 15:26 |
| mordred | zigo_: if you want to use our git as a source | 15:27 |
| fungi | zigo_: because upstream says it's not what they release? | 15:27 |
| dhellmann | flaper87: let me look again | 15:27 |
| flaper87 | dhellmann: https://review.openstack.org/#/c/46757/ | 15:27 |
| mordred | you need to run python setup.py sdist in the context of our git repo so that you get the correct metadata into the package | 15:27 |
| flaper87 | dhellmann: https://review.openstack.org/#/c/47580/ | 15:27 |
| mordred | which your package is missing | 15:27 |
| zigo_ | fungi: Well, upstream is not always right. For example, upstream tries to write in /usr/share in horizon... | 15:27 |
| flaper87 | dhellmann: https://review.openstack.org/#/c/47582/ | 15:27 |
| mordred | zigo_: that's completely different | 15:27 |
| fungi | zigo_: i would consider that a bug, definitely | 15:27 |
| mordred | zigo_: that's what the upstream tarball does at install time | 15:27 |
| flaper87 | The 3 of them are *marconi specific* and don't affect any other module | 15:27 |
| mordred | which you control via the files in debian/ | 15:28 |
| flaper87 | I mean, not directly | 15:28 |
| zigo_ | mordred: I'm only replying to "because upstream says so", saying that this is not a valid argumentation... :) | 15:28 |
| mordred | zigo_: I agree. I do not think it's a valid argument either. | 15:28 |
| dhellmann | flaper87: the change to falcon bothers me because it implies there is not actually any work being done to investigate moving off of falcon, as promised | 15:28 |
| mordred | I'm saying, as a packager, regenerating the upstream tarball incorrectly | 15:28 |
| mordred | because it's 'more convenient' | 15:28 |
| mordred | is not the right thing to do | 15:28 |
| dhellmann | flaper87: I think the other changes were just for stability at this point in the cycle | 15:28 |
| mordred | upstream has nothing to do with it | 15:28 |
| sdague | zul: so your havana libvirt / qemu requires ceph? | 15:29 |
| mordred | also, you are breaking the ability to track that the software came from where you said it came from | 15:29 |
| zul | sdague: i believe so what are you seeing? | 15:29 |
| dhellmann | flaper87: I'd be ok with letting those other 2 in if mordred agrees | 15:29 |
| mordred | since, by regenerating the tarball, the fingerprints can no longer be folloed | 15:29 |
| flaper87 | dhellmann: AFAIR, the request was to study the possibility to move away from falcon which we'll do https://blueprints.launchpad.net/marconi/+spec/pecan-framework | 15:29 |
| zigo_ | mordred: Just explain to me why using the git sources is wrong then! Last time at the summit, I asked what was different, and the only thing you told me was that there was a ChangeLog and an AUTHORS file. | 15:29 |
| flaper87 | dhellmann: but AFAIR, it wasn't a strong requirement | 15:29 |
| sdague | zul: well, it's not a devstack dep, and apparently it's not a hard enough requirement to blow things up early | 15:29 |
| mordred | dhellmann: point me at a thing? | 15:29 |
| flaper87 | that doesn't mean we're not considering / evaluating that, though | 15:30 |
| sdague | ok, so apparently I'm close enough to try to replicate now | 15:30 |
| mordred | zigo_: it's wrong because you are reporting bugs in our MANIFEST.in files | 15:30 |
| dhellmann | mordred: https://review.openstack.org/#/c/47580/ and https://review.openstack.org/#/c/47582/ | 15:30 |
| mordred | zigo_: which means you're doing something which is causing things to not operate as designed | 15:30 |
| zul | sdague: ive ran libvirt with ceph though pretty much everyday so I dont know what you are seeing | 15:30 |
| zigo_ | mordred: Well, it really *is* a bug this time. | 15:30 |
| mordred | it's not | 15:30 |
| dhellmann | flaper87: I have a hard time reconciling new features going into falcon with the idea that marconi could become less dependent on it. | 15:30 |
| mordred | the tarballs I produce work fine | 15:31 |
| sdague | zul: well you have symbol linkage, so you can't even start a 64bit guest, unrelated to openstack, unless you have ceph installed | 15:31 |
| zigo_ | mordred: How come "python setup.py install" doing the wrong thing isn't a bug??? | 15:31 |
| zul | sdague: erm? | 15:31 |
| zul | sdague: can you open up a bug please | 15:31 |
| flaper87 | dhellmann: but falcon development doesn't depend on Marconi. It's its own release cycle, bugs fixes and whatnot. I understand your concern, though. | 15:31 |
| *** tvb has quit IRC | 15:31 | |
| mordred | zigo_: MANIFEST.in does not affect operation of setup.py install | 15:31 |
| sdague | zul: sure, once I get further on this tempest thing | 15:32 |
| *** tvb has joined #openstack-infra | 15:32 | |
| *** tvb has joined #openstack-infra | 15:32 | |
| mordred | zigo_: so, if you had to modify that file to get a file instaled | 15:32 |
| zul | sdague: k | 15:32 |
| mordred | then you are doing something wrong a step before | 15:32 |
| mordred | which is that you are improperly generating a source tarball | 15:32 |
| dhellmann | flaper87: right, I don't actually care about falcon development per se, but if marconi is depending on new features of falcon that doesn't say to me there is serious work being done to think about moving to the same toolset the rest of us are using | 15:32 |
| mordred | which is why the file was not there for you to install | 15:32 |
| zigo_ | mordred: Let's agree that we don't agree... :) | 15:32 |
| fungi | zigo_: also, i still find it hard to believe that ftpmaster would tell you our changelog files had to be removed from our tarballs to be redistributable. i can understand that they wouldn't want you including it in the binary packages because of multiplicity issues, but insisting that you repack the upstream tarball to remove the upstream changelog doesn't sound like something they would insist on | 15:33 |
| mordred | zigo_: ok. well, just understand that I'm not going to fix bugs in any openstack project that are due to you as a downstream interacting with my output in ways that I've EXPLICITLY said are not supported | 15:33 |
| mordred | if you're ok with that, then I am ok with agreeing to disagree :) | 15:33 |
| jd__ | dhellmann: flaper87: I wanted to say about that, that anyhow I got the feeling that's something the TC will have to take into consideration when voting for incubation ending | 15:33 |
| zigo_ | fungi: You are right. | 15:33 |
| *** lcestari_ has quit IRC | 15:33 | |
| dhellmann | jd__: good point, so perhaps we should wait for that discussion to happen | 15:34 |
| zigo_ | mordred: You wont add a -1 to a patch to the MANIFEST.in of horizon, will you? | 15:34 |
| flaper87 | dhellmann: there are serious considerations going on but no serious work just yet. As I mentioned before, that's part of our plan before going into graduation but again, although it wasn't a strong requirement during the incubation voting (AFAIR) we've marked that blueprint as essential | 15:34 |
| jd__ | dhellmann: hm I'm not sure it's going to happen soon | 15:34 |
| zigo_ | mordred: I can change my way of packaging things later on, but right now, it's a bit late for me before Havana anyway. | 15:34 |
| dhellmann | jd__: oh, right, that's savanah that's being discussed tomorrow, I was confused | 15:34 |
| jd__ | mordred: zigo_: taking a step back, I still think there's a little bug with pbr having different behaviour wrt .git/MANIFEST.in | 15:35 |
| mordred | zigo_: fair. | 15:35 |
| mordred | jd__: I still haven't seen a usecase that describes one of those | 15:35 |
| mordred | jd__: _except_ where the .git repo isn't present | 15:35 |
| fungi | zigo_: other requests which "upstream" might honor to make our released tarballs more usable downstream include carrying an xz-compressed version, and possibly finding ways to summarize the changelog to not be quite to massive | 15:35 |
| fungi | er, quite SO massive | 15:36 |
| mordred | jd__: and in those cases, it's not pbr doing it, it's trying to operate in a context where there is no git metadata and where we don't maintain support for that in MANIFEST.in | 15:36 |
| pabelanger | FWIW: There is a best practices for repacking source tarballs: http://www.debian.org/doc/manuals/developers-reference/best-pkging-practices.html#bpp-origtargz Might help close the gap | 15:36 |
| jd__ | mordred: what about "I use pbr. I don't want to include in my sdist tarball everything that is in my git, so I wrote a MANIFEST.in. pbr doesn't take it into account" | 15:36 |
| dhellmann | flaper87: well, I've only voted -1, so if mordred removes the freeze on requirements changes then my vote can be overridden. if we need to discuss it more, maybe the mailing list would be a better venue. | 15:36 |
| mordred | jd__: if you can give me a test case where pbr is ignoring a MANIFEST.in that exists, I will accept it | 15:36 |
| pabelanger | EG: You should upload packages with a pristine source tarball if possible, but there are various reasons why it might not be possible. This is the case if upstream does not distribute the source as gzipped tar at all, or if upstream's tarball contains non-DFSG-free material that you must remove before uploading. | 15:36 |
| mordred | jd__: I do not believe that is what has been descirbed thusfar today | 15:36 |
| mordred | flaper87, dhellmann: crap. too many conversations... lemme try to read scrollback on this one | 15:37 |
| dhellmann | flaper87: and I'll subscribe to that blueprint, too | 15:37 |
| dhellmann | mordred: the marconi folks need some requirements updated, but we were under a freeze so I voted -1 on 2 (and -1 for a diff. reason on a third) | 15:37 |
| flaper87 | dhellmann: cool, thanks | 15:38 |
| jd__ | mordred: hm... pbr tried to include my whole $HOME because it (or a git command) looked recursively to find a .git and it ended up finding /home/jd/.git … so it ignored MANIFEST.in and blowed up (I've a big home ya know) | 15:38 |
| mordred | flaper87: oy. I do not like the response from kgriff on that patch | 15:38 |
| mordred | jd__: ok. so, that sounds like a bug I could potentially investigate fixing. finding a .git dir outside of the actual current source tree is certainly a bug | 15:40 |
| *** bnemec has quit IRC | 15:41 | |
| *** markmc has quit IRC | 15:41 | |
| jd__ | mordred: yep -- though bypassing it with SKIP_GIT_SDIST=1 solved the problem for that time | 15:41 |
| *** bnemec has joined #openstack-infra | 15:41 | |
| sdague | zul: where do you want to me to file the bug report? | 15:41 |
| zul | sdague: at your convience | 15:42 |
| sdague | zul: no, under what project | 15:42 |
| *** olaph has quit IRC | 15:42 | |
| sdague | this is basically the crux of the issue | 15:42 |
| sdague | cloud-archive-test:~> ldd /usr/bin/qemu-system-x86_64 | grep librbd | 15:42 |
| sdague | librbd.so.1 => /usr/lib/librbd.so.1 (0x00007f09c32bf000) | 15:42 |
| zul | ubuntu-cloud-archvie | 15:42 |
| mordred | flaper87: so, I need to read these reviews a little more closely, but I really do need to close the loop with you guys on this pecan/falcon thing | 15:42 |
| zul | need to reboot | 15:43 |
| *** zul has quit IRC | 15:43 | |
| mordred | flaper87: adding new projects to openstack that are in active rebellion against overall openstack design choices is not something we can afford to do | 15:43 |
| flaper87 | mordred: I understand. What I can say is that we're seriously investigating the migration to Pecan and both implementations will be tested and benchmarked. Again, there's nothing against using Pecan instead of Falcon but my understanding from our incubation meeting is that priorities were around the storage backends, integration with other OpenStack projects and the client | 15:43 |
| *** yaguang has joined #openstack-infra | 15:44 | |
| mordred | flaper87: it's the way it's worded in that blueprint that has me concerned | 15:44 |
| mordred | because it sounds like "we'll seriously look at it because we were asked to, but we're going to test to see how good pecan is at being falcon, and if it's not as good at being falcon as falcon is, then we're going to decide to not switch" | 15:45 |
| openstackgerrit | Antoine Musso proposed a change to openstack-infra/jenkins-job-builder: doc: fix up strict errors https://review.openstack.org/47861 | 15:45 |
| flaper87 | mordred: I wouldn't say Marconi is in active rebellion against openstack design choices. We've been integrating the project since its early stages. We've been using all tools and respecting all OpenStack standards and tools | 15:45 |
| mordred | flaper87: well, except for falcon | 15:45 |
| flaper87 | mordred: when we started using falcon, the pecan discussion wasn't even started. We couldn't afford migrating the whole thing at that moment | 15:46 |
| flaper87 | that's the only reason why we kept it | 15:46 |
| mordred | grumble | 15:47 |
| flaper87 | anyway, don't get me wrong here. | 15:47 |
| *** zul has joined #openstack-infra | 15:48 | |
| *** mrodden1 has quit IRC | 15:49 | |
| mordred | flaper87: yeah, same here. I basically just want to make sure we don't dig ourselves more of a hole | 15:49 |
| sdague | zul: https://bugs.launchpad.net/cloud-archive/+bug/1229255 | 15:49 |
| uvirtbot | Launchpad bug 1229255 in cloud-archive "qemu-system-x86_64 is linked to librbd.so so will not exec unless ceph is installed" [Undecided,New] | 15:49 |
| zul | sdague: cool ill have a look | 15:49 |
| openstackgerrit | Antoine Musso proposed a change to openstack-infra/jenkins-job-builder: stop ignoring doc/source/* https://review.openstack.org/47862 | 15:50 |
| flaper87 | mordred: totally agree, whatever is better for the whole project and the community. | 15:50 |
| sdague | zul: also it looks like n-cpu didn't crash, it's deadlocked with that version of libvirt | 15:51 |
| zul | sdague: im going to try to reproduce this in the afternoon | 15:51 |
| *** kgriffs has joined #openstack-infra | 15:51 | |
| zul | how did you create the deadlock? | 15:51 |
| flaper87 | mordred: it would be really cool to let the other 2 patches land for now | 15:52 |
| sdague | running tempest on a 4 core machine | 15:52 |
| flaper87 | we can discuss the falcon one a bit further | 15:52 |
| sdague | tox -e full | 15:52 |
| kgriffs | hey guys | 15:52 |
| sdague | has 4 simultaneous processes hitting things | 15:52 |
| mordred | kgriffs: you missed all the fun! | 15:54 |
| *** danger_fo_away is now known as danger_fo | 15:54 | |
| kgriffs | heh | 15:54 |
| kgriffs | so I heard | 15:54 |
| *** olaph has joined #openstack-infra | 15:55 | |
| mordred | kgriffs: tl;dr - falcon stuff - ongoing dev in falcon, concern that will make it harder to be able to migrate off, manifesting in conversation around bumping the falcon req in openstack/requirements | 15:55 |
| kgriffs | ok | 15:56 |
| *** dprince has quit IRC | 15:56 | |
| mordred | kgriffs: with the main concern (ironically) that I don't want pecan v. falcon to become a bigger deal or even continue to be a deal | 15:56 |
| kgriffs | yes, I agree | 15:56 |
| mordred | it's possible that this conversation may indicate the opposite of that :) | 15:56 |
| *** zehicle has joined #openstack-infra | 15:59 | |
| *** rpodolyaka has left #openstack-infra | 15:59 | |
| mordred | kgriffs: but the underlying concern I've got is in revisiting everything every time we add a new project. pecan v. falcon itself on its merits I actually care very little about. what I care more about is that we made a consolidated project direction decisions last summit, and now it seems like there is a change we're going to reopen that line of thought | 15:59 |
| mordred | and long term, having the same thing over and over and over again is just not scalable | 16:00 |
| *** mrodden has joined #openstack-infra | 16:00 | |
| mordred | so as long as we're on the same page about _that_, I can shut up for a while | 16:00 |
| *** matty_dubs is now known as matty_dubs|lunch | 16:02 | |
| zigo_ | mordred: & fungi: https://bugs.launchpad.net/horizon/+bug/1229264 (reported the fact that horizon trying to write into /usr/share/openstack-dashboard/openstack_dashboard/local/) | 16:03 |
| uvirtbot | Launchpad bug 1229264 in horizon "Horizon tries to write in /usr/share/openstack-dashboard/openstack_dashboard/local" [Undecided,New] | 16:03 |
| zigo_ | Ooops! :) | 16:03 |
| zigo_ | uvirtbot: Thanks! | 16:03 |
| uvirtbot | zigo_: Error: "Thanks!" is not a valid command. | 16:03 |
| mgagne | haha | 16:03 |
| fungi | zigo_: i agree that's ugly | 16:04 |
| zigo_ | I already have a symlink to local_settings.py in /etc ... | 16:04 |
| zigo_ | :) | 16:04 |
| kgriffs | mordred: I don't think anyone is arguing that it was silly to continue maintaining the OpenStack WSGI framework. | 16:05 |
| fungi | zigo_: probably some examples from other existing django-based webapps packaged for debian might help | 16:05 |
| sdague | fungi: https://bugs.launchpad.net/cloud-archive/+bug/1228977 | 16:05 |
| uvirtbot | Launchpad bug 1228977 in nova "n-cpu seems to crash when running with libvirt 1.0.6 from ubuntu cloud archive" [High,New] | 16:05 |
| fungi | zigo_: though the horizon peeps may have this worked out already and just slipped up in that one situation | 16:05 |
| sdague | looks like a hard lockup actually | 16:05 |
| fungi | sdague: awesome! | 16:05 |
| sdague | zul: that's for you as well, I put it on cloud-archive as well | 16:06 |
| *** hashar is now known as hasharCal | 16:06 | |
| *** hasharCal is now known as hasharCall | 16:06 | |
| zul | ack | 16:06 |
| *** zehicle has quit IRC | 16:06 | |
| zigo_ | fungi: There's not much more django apps packaged in Debian. apt-rdepends -r python-django gives me: bcfg2-web, dicoweb, djagios, graphite-web, insanity-web, mumble-django, and it seems to be it. | 16:08 |
| fungi | zigo_: you're on the cutting edge then (try not to get cut too badly!) | 16:09 |
| zigo_ | :) | 16:09 |
| zigo_ | I actually am going to bed. | 16:09 |
| zigo_ | :) | 16:09 |
| zigo_ | (midnight over here...) | 16:09 |
| *** flaper87 is now known as flaper87|afk | 16:11 | |
| mordred | kgriffs: oh gosh no. I'm certain no one is arguing that. I'm more concerned that we're going to end up down another rathole of two different frameworks with each side claiming "it's too much work for us to switch" | 16:12 |
| *** UtahDave has joined #openstack-infra | 16:13 | |
| mordred | zigo_: heh. come on - midnight is when it's time to get up and go drinking! | 16:13 |
| kgriffs | mordred: sounds like this has happened before? | 16:14 |
| *** dims has quit IRC | 16:14 | |
| mordred | kgriffs: some times I feel like the only thing I do in openstack is mediate between diametrically opposed sets of people who want to use different libraries | 16:14 |
| kgriffs | heh | 16:15 |
| zigo_ | mordred: Probably when we'll be in HK. There's boats to Macao directly from the venue, it's a *must do* ! :) | 16:15 |
| mordred | kgriffs: if you want some fun reading - go search the mailing list archives for the initial eventlet vs. twisted conversation | 16:15 |
| mordred | zigo_: woot! | 16:15 |
| kgriffs | I totally understand the benefits from consolidating libraries | 16:15 |
| *** cthulhup_ has joined #openstack-infra | 16:16 | |
| *** wchrisj has joined #openstack-infra | 16:16 | |
| kgriffs | that being said, in my experience it is virtually impossible to consolidate down to 1 library/thing in any organization that is much bigger than a startup. | 16:16 |
| hub_cap | mordred: thats just mean... the eventlet/twisted convo.. dont do it kgriffs | 16:16 |
| kgriffs | the best you can do is minimize the chaos | 16:16 |
| kgriffs | basically, not all projects are the same, and one-size-fits all doesn't always fit | 16:17 |
| kgriffs | that being said, the reasons for using different things or rolling your own are often more based on emotion than data | 16:17 |
| *** pcrews_ has joined #openstack-infra | 16:19 | |
| *** boris-42 has quit IRC | 16:19 | |
| *** zehicle has joined #openstack-infra | 16:20 | |
| *** pcrews has quit IRC | 16:20 | |
| mordred | indeed to the emotion bit | 16:21 |
| mordred | the thing is, we're SO big that if we let thigns proliferate at a natural rate, things would quickly become crazy | 16:21 |
| mordred | which means I get to be the person who is frustrating to people a bunch | 16:21 |
| kgriffs | yeah | 16:21 |
| kgriffs | you have to constantly fight the entropy. :p | 16:22 |
| mordred | yup | 16:22 |
| mordred | I'm fearing that this particular one is going to die a heat-death of emotion, since we have the author of both frameworks involved, which means someone is going to be sad eventually | 16:22 |
| mordred | and I don't like when people are sad | 16:23 |
| jd__ | NIH syndrom strikes again | 16:23 |
| mordred | otoh, 1200 devs | 16:23 |
| kgriffs | heh | 16:23 |
| mordred | so, sometimes people are sad | 16:24 |
| kgriffs | TBH, I'm more interested in *what* is right, not *who* is right | 16:24 |
| mordred | yup | 16:24 |
| kgriffs | Falcon was started because nothing existed that does what it does | 16:24 |
| jd__ | "When I get sad, I stop being sad and be awesome instead." | 16:24 |
| jd__ | True story. | 16:24 |
| kgriffs | FWIW, in the case of Marconi, the Falcon-based WSGI driver is already done, and we have been planning on writing a Pecan one before graduation, so at least you don't have to worry about the "it's too hard to switch" argument coming from my guys | 16:24 |
| jeblair | jd__: do you accomplish that by drinking? ;) | 16:25 |
| kgriffs | if Pecan is awesome and fast enough, we will totally use it | 16:25 |
| mordred | great | 16:25 |
| clarkb | morning | 16:25 |
| mordred | clarkb: don't read the scrollback | 16:25 |
| mordred | clarkb: you will go blind | 16:25 |
| zaro | morning | 16:25 |
| kgriffs | but if it isn't, then we need to decide to either fix Pecan or allow 3 frameworks in OpenStack | 16:25 |
| jd__ | jeblair: I've heard that can help :-) | 16:25 |
| kgriffs | (Pecan, Falcon, swob) | 16:25 |
| clarkb | mordred: but now I must out of curiousity | 16:25 |
| kgriffs | or something like that | 16:25 |
| mordred | kgriffs: indeed. well, fwiw, pecan lives in stackforge now, so fixing should be something openstack can accomplish | 16:26 |
| mordred | but, you know, that bridge, when we get there | 16:26 |
| kgriffs | yep | 16:26 |
| mordred | I just wanted to be clear the the pecan work was a real thing before approving updates to falcon things in requirements (the conversations have always been slightly vague and stuff) | 16:27 |
| kgriffs | yes, it is a real thing and is on our list of "must do" prior to graduation | 16:28 |
| kgriffs | rough list here: https://wiki.openstack.org/wiki/Marconi/Incubation/Graduation#To_Do | 16:28 |
| *** yidclare has joined #openstack-infra | 16:28 | |
| mordred | awesome | 16:28 |
| *** yidclare has left #openstack-infra | 16:29 | |
| mordred | so - for your current three reqs patches, we _are_ in requirements freeze - how urgent are there/should we look at FFE for them? or can they wait until icehouse opens? | 16:29 |
| *** julim has quit IRC | 16:29 | |
| *** afazekas has quit IRC | 16:37 | |
| *** reed has quit IRC | 16:37 | |
| kgriffs | mordred: tbh, I don't think we need pylru at all (need to benchmark one more time to be sure) | 16:38 |
| *** reed has joined #openstack-infra | 16:38 | |
| kgriffs | falcon 0.1.7 we can wait on | 16:39 |
| kgriffs | mmh3 is the tricksy one | 16:39 |
| *** pblaho has quit IRC | 16:40 | |
| *** adam_g has quit IRC | 16:40 | |
| kgriffs | we have a patch that needs some kind of stable hash that won't change between python versions or anything, as is reasonably fast. | 16:40 |
| kgriffs | let me noodle on that a bit - may be an alternative. | 16:40 |
| mordred | kgriffs: we _can_ do a freeze exception - especially since it's only for a thing that isn't part of havanna | 16:40 |
| mordred | it just requires slightly more chatting and stuff | 16:41 |
| mordred | dhellmann, Alex_Gaynor: ^^ thoughts? | 16:41 |
| Alex_Gaynor | mordred: what's the question? | 16:41 |
| *** reed has quit IRC | 16:41 | |
| mordred | possible FFE for mmh3 for marconi | 16:41 |
| *** reed has joined #openstack-infra | 16:42 | |
| *** julim has joined #openstack-infra | 16:42 | |
| Alex_Gaynor | kgriffs: was just using sha considewred? | 16:43 |
| *** yaguang has quit IRC | 16:43 | |
| *** zehicle has quit IRC | 16:43 | |
| *** kpavel has quit IRC | 16:43 | |
| kgriffs | yes, but it is too slow | 16:43 |
| kgriffs | actually, we used MD5 which is faster than SHA | 16:43 |
| *** adam_g has joined #openstack-infra | 16:43 | |
| *** adam_g has quit IRC | 16:43 | |
| *** adam_g has joined #openstack-infra | 16:43 | |
| kgriffs | but still, dropped our req/sec quite a bit in benchmarking | 16:43 |
| Alex_Gaynor | kgriffs: what about siphash :D | 16:44 |
| kgriffs | (Marconi is a data API, and we have high-throughput reqs) | 16:44 |
| * kgriffs looking | 16:44 | |
| Alex_Gaynor | kgriffs: (Sorry I'm thinking about the FFE question, just spitballing as I do) | 16:44 |
| dhellmann | mordred: I'm ok with adding mmh3, since nothing else uses that. It doesn't bring in any other dependency conflicts, does it? | 16:44 |
| Alex_Gaynor | not that siphash is in the stdlib | 16:44 |
| *** wchrisj has quit IRC | 16:45 | |
| dhellmann | mordred: although I'll defer to Alex_Gaynor's superior knowledge of alternatives. :-) | 16:45 |
| mordred | dhellmann: no. it's clean - no other depends | 16:45 |
| mordred | claims 2 and 3 support (which I suppose is the point) | 16:45 |
| mordred | this: | 16:45 |
| mordred | 'License :: Public Domain', | 16:45 |
| mordred | is a little funny, but should be fine | 16:46 |
| kgriffs | it is slow on PyPy but I plan to fix that. ;) | 16:46 |
| kgriffs | (needs some cffi love) | 16:46 |
| Alex_Gaynor | But it does work on pypy? | 16:46 |
| kgriffs | is crc32 already in the deps? | 16:46 |
| kgriffs | yes, it works | 16:46 |
| kgriffs | (mmh3 works on pypy) | 16:46 |
| *** jcoufal has quit IRC | 16:47 | |
| kgriffs | (doesn't look like crc32 is in there) | 16:47 |
| Alex_Gaynor | crc32 is in the stdlib | 16:47 |
| Alex_Gaynor | binascii module IIRC | 16:47 |
| kgriffs | oh, right. duh | 16:47 |
| dhellmann | jeblair: how hard would it be to add checkboxes to the requirements project in gerrit for things like "supports python3" and "supports pypy"? | 16:48 |
| kgriffs | i suppose we could maybe use that for our particular need (crc32) | 16:48 |
| jeblair | dhellmann: very easy if we upgrade to gerrit 2.6 | 16:48 |
| * kgriffs keeps forgetting how extensive python stdlib is | 16:48 | |
| * dhellmann hopes that's a "when" rather than an "if" ;-) | 16:48 | |
| *** dcramer_ has quit IRC | 16:49 | |
| kgriffs | guys, let me look at a couple alternatives to mmh3 and get back to you in a bit | 16:49 |
| jeblair | dhellmann: zaro is working on getting some stuff upstream so we can support WIP; otherwise we would need to ditch WIP because we don't have the manpower to maintain a gerrit fork. | 16:49 |
| dhellmann | jeblair: ah, right, that's come up before | 16:49 |
| dhellmann | kgriffs: post what you find to the review for mmh3 if you still need it, and I'll +2 | 16:50 |
| kgriffs | cool, sounds like a plan | 16:50 |
| mordred | zaro: what's the latest on that? still blocked on "nice patch, but could you also drastically expand the scope of the patch?" | 16:52 |
| mordred | hasharCall: who's your gerrit core dude again? | 16:52 |
| * cppcabrera would love to see "supports python 3", "supports pypy" checkboxes in gerrit | 16:52 | |
| hasharCall | mordred: ^d | 16:52 |
| clarkb | mordred: re sqlalchemy funness, we should reduce the restriction on the upper bound, do whatever fixing is necessary then move the lower bound (thinking back to keystoneclient problems earlier this cycle) | 16:52 |
| hasharCall | mordred: Chad "^d" Horohoe. | 16:52 |
| *** svarnau has joined #openstack-infra | 16:52 | |
| mordred | hasharCall: thanks | 16:52 |
| hasharCall | mordred: if not available you can ask for qchris who is Europe based, though he is no more working on Gerrit. | 16:53 |
| *** hemnafk is now known as hemna | 16:54 | |
| mordred | zomg. | 16:54 |
| mdenny | I'm setting up havana h3 deployment on 4 nodes Ubuntu 12.04.3 - Controller node, Network node, 2 Compute nodes | 16:54 |
| mordred | the upstream gerrit people have added "Diffy Cuckoo" | 16:54 |
| mdenny | Controller node has ntp, Mysql, keystone, RabbitMQ, Glance and Neutron working as expected. | 16:55 |
| mordred | mdenny: almost certainly the wrong channel | 16:55 |
| mdenny | apt-get install -y novnc novnc : Depends: libjs-swfobject but it is not installable | 16:55 |
| zaro | mordred, jeblair, dhellmann : patch is here.. https://gerrit-review.googlesource.com/#/c/48254/ | 16:55 |
| mordred | mdenny: this is for people running the openstack developer infrastructure - you probably want #openstack or possibly #openstack-dev | 16:55 |
| mdenny | mordred: Thanks | 16:55 |
| *** SergeyLukjanov has quit IRC | 16:56 | |
| zaro | mordred, jeblair, dhellmann : i believe i've got what they want this time. mfick says he will take a look at it today. | 16:57 |
| *** zehicle has joined #openstack-infra | 16:57 | |
| mordred | zaro: w00t! | 16:58 |
| jeblair | zaro: excellent; it seems like there was little response to your RFC thread which was one of the outcomes that mfick anticipated | 16:58 |
| mordred | zaro: that would be the best thing that's ever happened | 16:58 |
| *** pcrews_ has quit IRC | 16:58 | |
| jeblair | meaning, it seems no one feels strongly or objects to applying it to those permissions | 16:59 |
| *** pcrews_ has joined #openstack-infra | 16:59 | |
| zaro | it became a larger change than i expected. i've gone thru testing it and it seems to be working pretty well. | 17:00 |
| jeblair | ooh, i like the shaded vote box... "User cannot vote in this category". that will be nice for identifying core people | 17:00 |
| jeblair | mordred: ^ | 17:00 |
| *** matty_dubs|lunch is now known as matty_dubs | 17:02 | |
| *** zehicle has quit IRC | 17:03 | |
| mordred | jeblair: ++ | 17:04 |
| openstackgerrit | Jay Pipes proposed a change to openstack-infra/config: Enable gate on OpenStack Chef Repository https://review.openstack.org/47880 | 17:04 |
| *** KurtMartin has joined #openstack-infra | 17:04 | |
| mordred | jeblair: I'm happy for them that they have automatic verification of changes. I'm a little sad that they wrote their own thing | 17:05 |
| *** hasharCall has quit IRC | 17:05 | |
| jeblair | mordred: they are insane | 17:06 |
| mordred | jeblair: the new projects list page is nice too | 17:06 |
| mordred | jeblair: https://gerrit-review.googlesource.com/#/admin/projects/plugins/avatars/external | 17:06 |
| zaro | mordred: i don't think it includes tests though. | 17:07 |
| *** kgriffs is now known as kgriffs_afk | 17:11 | |
| *** mrmartin has joined #openstack-infra | 17:13 | |
| *** BobBall is now known as BobBall_Away | 17:14 | |
| mordred | https://gerrit.googlesource.com/plugins%2Frename-project | 17:14 |
| mordred | I wonder if that would allow us to rename on a live server? | 17:14 |
| *** jerryz has joined #openstack-infra | 17:15 | |
| mordred | well, if it existed, that would be even better | 17:15 |
| mordred | dhellmann: there is a tuskar thread on openstack-dev you might be interested in | 17:18 |
| hub_cap | mordred: hey dude im tryign to do this refactoring for ya on the client, and i got things pretty far, but im hitting a wonderful wall of reviews... so im going ot have to play people-review-tag for a while to get some new oslo / keystoneclient things done | 17:21 |
| *** gyee has joined #openstack-infra | 17:21 | |
| *** markmcclain1 has joined #openstack-infra | 17:21 | |
| *** markmcclain has quit IRC | 17:22 | |
| jerryz | jeblair: ping | 17:22 |
| jeblair | jerryz: pong | 17:22 |
| *** lcestari_ has joined #openstack-infra | 17:23 | |
| jerryz | jeblair: i am switching to nodepool from d-g, but when launching a instance, the nodepool server can not access it. | 17:23 |
| jerryz | jeblair: the authorized_keys does not have server's public key | 17:23 |
| jeblair | jerryz: in our puppet modules, we configure the ssh public key for the jenkins user | 17:24 |
| mordred | hub_cap: cool. let me know if you want me to help! | 17:24 |
| jeblair | jerryz: so for us the process is: nodepool spins up a server to build an image from, logs in with the key it creates, runs puppet which installs the jenkins ssh key, then saves the image | 17:25 |
| hub_cap | roger mordred. its more of a 3 people w/ 3 different ideas thing. i might get a ML chain started and ask you to thwack people | 17:25 |
| mordred | hub_cap: point me at the reviews? | 17:25 |
| jeblair | jerryz: then nodepool spins up new servers from that image and uses the jenkins ssh key to log into them (what key it uses for that is configured in nodepool.yaml) | 17:25 |
| *** pcrews_ has quit IRC | 17:25 | |
| jeblair | jerryz: so to summarize the missing documentation: | 17:26 |
| hub_cap | whats the general stance on stackforge projects? should we not integrate them until they go incubated, or integrated even? im looking specifically @ designate for dns support in trove | 17:26 |
| jerryz | my problem is when creating the snap_image, the key is not injected to my image | 17:26 |
| jeblair | jerryz: you should run _something_ in the image creation scripts that installs a known ssh key, and then you should configure nodepool.yaml to use that ssh key for the login check it does when it launches a server | 17:26 |
| mordred | hub_cap: you should not integrate with them in such a way that you would need to start including them in your trunk gating | 17:27 |
| mordred | hub_cap: including experimental/optional support for them is fine | 17:27 |
| jerryz | you mean i should copy and paste the public key manually while the image is being created | 17:27 |
| hub_cap | <3 as always mordred | 17:27 |
| mordred | jerryz: yes. | 17:28 |
| *** pcrews_ has joined #openstack-infra | 17:28 | |
| jeblair | jerryz: no, there's a step where it runs some scripts before creating the snap image; so you should do something there to create the key | 17:28 |
| jeblair | mordred: yes? manually? | 17:28 |
| mordred | well, not MANUALLY - I meant in the script | 17:28 |
| * mordred butts out of the conversation - jeblair is doing fine | 17:28 | |
| jeblair | jerryz: it might help to know some context -- do you want images that are just like ours (with our puppet config, etc), or are you using it for something else? | 17:29 |
| jerryz | jeblair: yes | 17:30 |
| *** sarob has joined #openstack-infra | 17:30 | |
| jerryz | jeblair: only difference is my nodepool server is not manged by puppet | 17:30 |
| jeblair | jerryz: ok, so if you are using our nodepool setup scripts (the ones at http://git.openstack.org/cgit/openstack-infra/config/tree/modules/openstack_project/files/nodepool/scripts ) | 17:31 |
| *** SergeyLukjanov has joined #openstack-infra | 17:31 | |
| jeblair | jerryz: then you can just set the "NODEPOOL_SSH_KEY" environment variable to contain the public key when you run nodepool | 17:31 |
| jerryz | i have already copied those scripts to the location where nodepool.yaml specifies | 17:31 |
| jeblair | jerryz: it does the same thing that the devstack gate ssh key variable used to do.... | 17:31 |
| jerryz | jeblair: that is what i missed | 17:32 |
| jeblair | jerryz: where it substitutes your public key for jenkins | 17:32 |
| jeblair | jerryz: then configure nodepool.yaml to use the private key, and you should be good to go | 17:32 |
| fungi | ahh, sorry i didn't think to suggest overriding it with that the other night jerryz | 17:32 |
| jeblair | jerryz: you can look at http://git.openstack.org/cgit/openstack-infra/config/tree/modules/openstack_project/files/nodepool/scripts/prepare_node.sh | 17:32 |
| jeblair | fungi, jerryz: nodepool docs are high on my list, sorry i'm a bit behind | 17:32 |
| jeblair | but the general mechanism is that nodepool will pass through any NODEPOOL_* environment variables to the scripts | 17:33 |
| jeblair | so you can pass anything else you need to the scripts that way as well | 17:33 |
| jeblair | (that's also what i use to test nodepool locally) | 17:34 |
| *** julim has quit IRC | 17:35 | |
| jerryz | thanks | 17:35 |
| jeblair | jerryz: you're welcome! | 17:36 |
| *** ruhe has quit IRC | 17:38 | |
| *** kgriffs_afk is now known as kgriffs | 17:41 | |
| *** pcrews_ has quit IRC | 17:42 | |
| *** dprince has joined #openstack-infra | 17:46 | |
| clarkb | the electrical outlets in my cube work again \o/ | 17:49 |
| mordred | w00t! | 17:50 |
| *** boris-42 has joined #openstack-infra | 17:50 | |
| mgagne | clarkb: you can now plug back that coffee maker :D | 17:50 |
| *** jerryz_ has joined #openstack-infra | 17:50 | |
| *** jerryz has quit IRC | 17:50 | |
| *** nicedice has joined #openstack-infra | 17:51 | |
| *** pcrews_ has joined #openstack-infra | 17:54 | |
| *** kgriffs is now known as kgriffs_afk | 17:55 | |
| *** jerryz_ has quit IRC | 17:55 | |
| fungi | and the space heater | 17:56 |
| hub_cap | clarkb: did u see my email to u about ovz+iscisd ? | 17:56 |
| *** ArxCruz has quit IRC | 17:56 | |
| *** Ryan_Lane has joined #openstack-infra | 17:56 | |
| *** moted has quit IRC | 17:57 | |
| *** moted has joined #openstack-infra | 17:58 | |
| clarkb | hub_cap: not yet | 17:59 |
| clarkb | hub_cap: oh actually yes | 17:59 |
| hub_cap | you will get a kick out of it | 17:59 |
| clarkb | that email I loled | 17:59 |
| *** morganfainberg has quit IRC | 17:59 | |
| clarkb | hub_cap: mordred stevebaker pleia2 https://etherpad.openstack.org/testing-heat-trove-with-dib | 18:00 |
| hub_cap | but i have afeeling, if we file a bug, it will get fixed | 18:00 |
| hub_cap | and fast | 18:00 |
| *** sarob has quit IRC | 18:00 | |
| clarkb | that is from ^ my whiteboard | 18:00 |
| pleia2 | clarkb: thanks | 18:00 |
| hub_cap | yes thx | 18:00 |
| *** mestery has quit IRC | 18:01 | |
| zaro | mordred: can i get a review from you for this one since you tried to do the same thing? https://gerrit-review.googlesource.com/#/c/48254/ | 18:01 |
| zaro | mordred: oops meant this one.. https://review.openstack.org/#/c/45329/ | 18:01 |
| *** kpavel has joined #openstack-infra | 18:01 | |
| *** jerryz has joined #openstack-infra | 18:01 | |
| *** morganfainberg has joined #openstack-infra | 18:02 | |
| *** ruhe has joined #openstack-infra | 18:03 | |
| *** ArxCruz has joined #openstack-infra | 18:03 | |
| *** ruhe has quit IRC | 18:03 | |
| openstackgerrit | A change was merged to openstack-infra/jenkins-job-builder: stop ignoring doc/source/* https://review.openstack.org/47862 | 18:05 |
| *** fbo_away is now known as fbo | 18:06 | |
| *** morganfainberg has quit IRC | 18:06 | |
| *** morganfainberg has joined #openstack-infra | 18:07 | |
| *** morganfainberg has quit IRC | 18:07 | |
| *** mestery has joined #openstack-infra | 18:08 | |
| *** morganfainberg has joined #openstack-infra | 18:08 | |
| clarkb | mordred: have you sorted out why some of your tox.ini changes for tox 1.6 seem to hate the setup.py develop stuff? | 18:14 |
| clarkb | jeblair: fungi mordred https://review.openstack.org/#/c/47393/ I think we should merge that asap to avoid conflicts | 18:16 |
| clarkb | also possibly fix the comment in tthere about global requirements | 18:16 |
| hub_cap | mordred: and co, what ever came of the "vote thru gerrit" stuff ? | 18:17 |
| clarkb | hub_cap: its up and should be useable | 18:17 |
| hub_cap | WHAT!?!? i wanna test it | 18:17 |
| hub_cap | i just did a #fakevote in my channel | 18:17 |
| clarkb | oh I don't know if it is ready for general consumption | 18:17 |
| hub_cap | and ya it didnt go over too well.. is there a rtfm tutorial on it clarkb ? | 18:17 |
| clarkb | I think it was created for the TC | 18:17 |
| hub_cap | awwwwwwww | 18:17 |
| *** melwitt has joined #openstack-infra | 18:18 | |
| hub_cap | build me up just to break me down clarkb | 18:18 |
| clarkb | hub_cap: meetbot has a vote mechanism for in meeting votes | 18:18 |
| hub_cap | yes then i have to wait for my meeting... :( | 18:18 |
| jeblair | hub_cap: you should wait for your meeting, unless you're trying to subvert democracy. :) | 18:18 |
| hub_cap | shhhhh | 18:19 |
| clarkb | hub_cap: the basic idea is you have a project in gerrit that tracks proposals as gerrit changes, this way the TC can make proposals offline that everyone can think through and not make up their minds in an hour | 18:19 |
| hub_cap | see thats what i want, more of a offline vote process | 18:19 |
| clarkb | then as positions are formed individuals can vote on the change, then if approved change is merged leaving a record of the decision | 18:19 |
| hub_cap | so u know, if you arent in a meeting you can still vote :) (not to subvert democracy) | 18:19 |
| jeblair | hub_cap: lot's of people do it, you're in good company! Nixon... Reagan... Bush... Bush... Obama... wow you're looking positively presedential. :) | 18:20 |
| * hub_cap puts a tie on | 18:20 | |
| clarkb | but I am not sure that works if you mulitplex many sets of decisions makers on a single project | 18:20 |
| hub_cap | so do yall think thats something a ptl could use for a project vote? | 18:20 |
| clarkb | I think the process is sound, it may require more projects though | 18:21 |
| clarkb | jeblair: ^ | 18:21 |
| fungi | clarkb: i'm good with merging 47393 now, as it's no longer saturday | 18:21 |
| fungi | otherwise, yes, rebase hell ensues | 18:22 |
| fungi | approved | 18:22 |
| clarkb | fungi: cool and ty | 18:22 |
| fungi | will keep an eye on the gate for a while after it merges | 18:22 |
| jeblair | hub_cap: so every change to a project is already a project vote; i think the innovation here is that the tc didn't have a code project it could do that with, now it does. | 18:22 |
| clarkb | fungi: awesome, there is another change in the queue that needs to be rechecked after that merges, I will recheck that change | 18:22 |
| jeblair | hub_cap: but i'm guessing you're talking about more long-ranging items where a change isn't ready or appropriate yet | 18:22 |
| clarkb | jeblair: good point | 18:22 |
| hub_cap | jeblair: or something thats maybe not code related, at least in the short term (which i believe is a paraphrase of what you just said, ya?) | 18:23 |
| jeblair | hub_cap: but you could still do that with documentation patches, or something similar... | 18:23 |
| fungi | clarkb: yeah, there are several stacked behind it which need rechecks after it merges | 18:23 |
| jeblair | patches to README, TODO files, etc... | 18:24 |
| hub_cap | thats a good point jeblair | 18:24 |
| *** kiranmurari has joined #openstack-infra | 18:24 | |
| jeblair | hub_cap: the main thing i'd suggest is try not to be too different than other openstack projects... if you experiment with something like that, make sure people in meetings or mailing lists know about it -- so you don't subvert democracy. :) | 18:24 |
| *** dcramer_ has joined #openstack-infra | 18:24 | |
| clarkb | fungi: merge conflict already :( | 18:24 |
| fungi | poop | 18:25 |
| fungi | fixing | 18:25 |
| clarkb | fungi: can you correct that comment about global requirements as well | 18:25 |
| jeblair | hub_cap: (and if it's awesome, it may become standard for other projects) | 18:25 |
| clarkb | fungi: jeblair left it inline | 18:25 |
| fungi | will do--missed it earlier | 18:25 |
| kiranmurari | koolhead17: heyy... | 18:26 |
| hub_cap | jeblair: and then people would like me! | 18:26 |
| hub_cap | oh wait.... | 18:26 |
| hub_cap | thats probably not true | 18:26 |
| hub_cap | ill try to think of something that requires a changeset but is minimal overhead and experiment | 18:27 |
| hub_cap | use the tool i already got so to speak | 18:27 |
| hub_cap | too bad recording the vote result will result in a rebase / re-review lol | 18:28 |
| openstackgerrit | A change was merged to openstack-infra/zuul: add forge author identity privilege to support pushing refs to gerrit https://review.openstack.org/47233 | 18:29 |
| *** tstevenson has joined #openstack-infra | 18:29 | |
| clarkb | hub_cap: ya, that is a possibility (I expect that may become an issue with the TC project depending on how things are recorded within git) | 18:30 |
| openstackgerrit | Andreas Jaeger proposed a change to openstack-infra/config: Publish atom.xml for manuals https://review.openstack.org/47897 | 18:30 |
| clarkb | I suppose a different file could be used for each proposal there | 18:30 |
| openstackgerrit | James E. Blair proposed a change to openstack-infra/zuul: Fix NNFI bug with two failing changes at head https://review.openstack.org/47898 | 18:30 |
| kiranmurari | hello folks, i need to get a couple of posts removed from planet.openstack.org... | 18:31 |
| jeblair | clarkb, fungi, mordred: ^ that is a bug that I have seen in production; it could allow incorrectly tested changes to merge | 18:31 |
| hub_cap | clarkb: thats what im going w/ right now (doc/votes/{some_proposal} | 18:31 |
| clarkb | jeblair: looking | 18:31 |
| jeblair | kiranmurari: i believe if you remove them from your feed they will be removed automatically | 18:32 |
| kiranmurari | i deleted them from our wordpress blog, but before that they got populated on planet and don't go away from planet | 18:32 |
| jeblair | kiranmurari: when did you remove? | 18:32 |
| kiranmurari | jeblair: thanks for the pointer... but its been almost 36 hrs that i deleted them from our wordpress... but they are still appearing on planet :( | 18:32 |
| *** Ajaeger1 has joined #openstack-infra | 18:33 | |
| jeblair | kiranmurari: which posts? | 18:33 |
| kiranmurari | jeblair: te title of both the posts is 'Summary: OpenStack India Day, 2013' and the author is 'kiranmurari' | 18:33 |
| kiranmurari | jeblair: therez been a slight goofup so had to delete them | 18:34 |
| mrmartin | fungi: hi | 18:37 |
| fungi | mrmartin: hi there! what's next on the groups-dev agenda? | 18:37 |
| mrmartin | could you review my two patches: https://review.openstack.org/#/c/47838/ and https://review.openstack.org/#/c/47844/ | 18:37 |
| fungi | mrmartin: sure | 18:38 |
| mrmartin | one of them is updating drush to 6.0.0 from 5.9.0 and the other one repairs a bug that affects auto-site deployment | 18:38 |
| openstackgerrit | Jeremy Stanley proposed a change to openstack-infra/config: Fix pep8 errors in modules dir https://review.openstack.org/47644 | 18:38 |
| openstackgerrit | Jeremy Stanley proposed a change to openstack-infra/config: Remove launchpad_sync module https://review.openstack.org/47643 | 18:38 |
| openstackgerrit | Jeremy Stanley proposed a change to openstack-infra/config: Remove the now vestigal pyflakes tox env https://review.openstack.org/47641 | 18:38 |
| openstackgerrit | Jeremy Stanley proposed a change to openstack-infra/config: Remove pyflakes jobs https://review.openstack.org/47393 | 18:38 |
| fungi | clarkb: ^ | 18:38 |
| clarkb | fungi: ty | 18:38 |
| clarkb | jeblair: zuul change reviewed. lgtm | 18:38 |
| fungi | (rebased and edited out that one offending comment) | 18:38 |
| mrmartin | it was checking a bad git file for recent commit hash | 18:38 |
| * fungi is reading through jeblair's nnfi fix and then will take a look at the groups-dev fixes | 18:39 | |
| *** sarob has joined #openstack-infra | 18:39 | |
| mordred | zomg. too much scrollback | 18:41 |
| jeblair | mordred: can you look at kiranmurari's question about planet? | 18:42 |
| mordred | looking | 18:43 |
| *** sarob has quit IRC | 18:44 | |
| kiranmurari | jeblair: thx | 18:44 |
| mordred | kiranmurari, jeblair: planet does some caching, so it might not see the upstream deletions | 18:44 |
| mordred | I _believe_ we can just delete the planet cache which will cause the next run of planet to run uncached | 18:44 |
| *** cody-somerville has joined #openstack-infra | 18:44 | |
| sdague | mordred: I believe you are correct on the cache front | 18:45 |
| kiranmurari | mordred: thatz what i expected... is it something like PlanetData | 18:45 |
| jeblair | mordred: ok. is there a tunable that will make it cache less aggressively? | 18:45 |
| openstackgerrit | A change was merged to openstack-infra/config: Remove pyflakes jobs https://review.openstack.org/47393 | 18:45 |
| openstackgerrit | A change was merged to openstack-infra/zuul: Fix NNFI bug with two failing changes at head https://review.openstack.org/47898 | 18:45 |
| *** kgriffs_afk is now known as kgriffs | 18:46 | |
| mordred | feed_timeout = 20 | 18:46 |
| mordred | but I don't really konw what 20 means | 18:46 |
| clarkb | jeblair: fungi: I have not approved https://review.openstack.org/#/c/47636/ because it is semi process related and figured having more reviewers look at it could only help (the code looks fine) | 18:46 |
| jeblair | http://www.intertwingly.net/code/venus/docs/config.html says Number of seconds to wait for any given feed | 18:46 |
| mordred | kiranmurari: what's the url of your blog? | 18:47 |
| sdague | jeblair: I think it's like recheckwatch, there really is no cache cleaning | 18:47 |
| jeblair | mordred, sdague: the current behavior seems broken; everything it needs to DTRT is in the http headers | 18:47 |
| kiranmurari | jeblair: http://www.openstackindia.wordpress.com | 18:47 |
| jeblair | sdague: recheckwatch cleans its cache. :) | 18:47 |
| sdague | ok, fair :) | 18:47 |
| kiranmurari | mordred: link above | 18:47 |
| mordred | jeblair: agree | 18:47 |
| mordred | jeblair: cache_keep_entries might be interesting to us - although I agre with you re: http headers | 18:48 |
| mordred | ah | 18:49 |
| *** cthulhup_ has quit IRC | 18:49 | |
| mordred | we're not running the expunge command | 18:49 |
| *** CaptTofu has quit IRC | 18:49 | |
| jeblair | oh neat | 18:49 |
| *** vipul is now known as vipul-away | 18:49 | |
| jeblair | jog0, clarkb: i'm curious about https://review.openstack.org/#/c/47636/ | 18:49 |
| *** vipul-away is now known as vipul | 18:50 | |
| jeblair | it references a bug that was marked wontfix for the ossa project, but not nova | 18:50 |
| jeblair | presumably it was rechecked in the context of nova, not ossa since ossa does not have a project in gerrit | 18:50 |
| *** ruhe has joined #openstack-infra | 18:51 | |
| fungi | that's a good question... how should recheckwatch operate in the face of a bug with multiple bugtasks | 18:51 |
| jeblair | in nova, it is in progress and fix committed on two branches... | 18:51 |
| jeblair | so i'm not convinced that's a great example for that fix | 18:51 |
| *** ruhe has quit IRC | 18:51 | |
| fungi | it's a great example of something though | 18:51 |
| jeblair | fungi: yes :) | 18:52 |
| jeblair | i'll write something in the review | 18:52 |
| fungi | for example, someone might still recheck a grizzly change against that bug because it's not committed there yet | 18:52 |
| fungi | and the ossa project doesn't have any code repositories (not that recheckwatch knows anything about lp/git matchups) | 18:53 |
| clarkb | jeblair: thank you | 18:53 |
| fungi | so basically recheckwatch is somewhat ignorant of per-project and per-branch bugtasks in the context of a change which could be affected by a bug | 18:54 |
| fungi | and that seems like it would require a rather nontrivial solution | 18:54 |
| clarkb | fungi: it should be able to do a set union of all the states and if something not in the closed list is present treat the bug as open | 18:55 |
| jeblair | fungi: i'm not convinced it would be hard. it _does_ know what projects have triggered rechecks, so it can examine bug states for all of those. | 18:55 |
| clarkb | jeblair: zaro: does https://review.openstack.org/#/c/46124/ need rebasing? | 18:56 |
| kiranmurari | mordred: jeblair: sdague: the cache can be expunged from admin interface or with 'python planet.py -x config.ini' - http://www.intertwingly.net/code/venus/docs/admin.html | 18:56 |
| *** wchrisj has joined #openstack-infra | 18:56 | |
| fungi | yeah, i guess a 100% solution would be hard but a 95% solution might not be too complicated | 18:56 |
| *** morganfainberg has quit IRC | 18:57 | |
| *** morganfainberg has joined #openstack-infra | 18:57 | |
| kiranmurari | mordred: jeblair: sdague: but i hope that wouldn't expunge the whole cache!!! | 18:58 |
| dhellmann | mordred: thanks for the pointer to the tuskar thread, I've responded | 18:59 |
| jeblair | fungi, clarkb: but that won't work because then we would start ignoring openstack-ci bugs | 19:00 |
| mordred | kiranmurari: it does not seem to do _anything_ helpful | 19:02 |
| jeblair | fungi, clarkb: I think the change is okay as-is. | 19:03 |
| openstackgerrit | Andreas Jaeger proposed a change to openstack-infra/config: Use Jenkins templates for old manual jobs https://review.openstack.org/47691 | 19:03 |
| *** hashar has joined #openstack-infra | 19:04 | |
| openstackgerrit | A change was merged to openstack-infra/config: Make "Won't Fix" a closed status in recheckwatch https://review.openstack.org/47636 | 19:05 |
| *** sarob has joined #openstack-infra | 19:10 | |
| *** dmakogon_ has joined #openstack-infra | 19:11 | |
| sdague | hey, so 26 jobs are blowing up a lot on failed to clone issues - https://jenkins02.openstack.org/job/gate-nova-python26/3893/console | 19:12 |
| sdague | any idea if rack fell over in a funny way | 19:12 |
| sdague | this is https://bugs.launchpad.net/openstack-ci/+bug/1229352 | 19:13 |
| uvirtbot | Launchpad bug 1229352 in openstack-ci "intermitent failure of git clone from git.openstack.org" [Critical,In progress] | 19:13 |
| sdague | and if you look at the gate right now, there are 2 py26 fails there that will cause resets that are that issue (I had reverified one earlier) | 19:14 |
| fungi | sdague: corrupt workspaces i think, saw one last week | 19:15 |
| sdague | ok, 3 in a row in a couple of hours seems suspicous though | 19:15 |
| fungi | agreed | 19:15 |
| fungi | sdague: which slaves besides centos6-6? | 19:16 |
| sdague | only centos 26 as far as I know | 19:16 |
| sdague | 2013-09-23 19:08:57.131 | Building remotely on centos6-2 in workspace /home/jenkins/workspace/gate-nova-python26 | 19:17 |
| fungi | -2 as well, okay | 19:17 |
| sdague | 2013-09-23 18:54:46.663 | Building remotely on centos6-2 in workspace /home/jenkins/workspace/gate-nova-python26 | 19:17 |
| sdague | ok, these are both on -2 | 19:17 |
| fungi | the one in your bug was -6 | 19:17 |
| fungi | taking them both out of service for abit | 19:18 |
| sdague | ok cool | 19:18 |
| sdague | yeh, these are still in the gate | 19:18 |
| fungi | but yes if i try to remote update in the workspace for that job on -6 the remote does hang up on me | 19:18 |
| sdague | I was just looking before I flushed a bunch of code to figure out how backed up things are | 19:19 |
| fungi | it *is* possible for a change to foul up the .git directory in a workspace such that we can't gerrit-git-prep any longer, though not sure yet if that's what's happened | 19:20 |
| *** ArxCruz has quit IRC | 19:21 | |
| *** ArxCruz has joined #openstack-infra | 19:22 | |
| *** ArxCruz has quit IRC | 19:22 | |
| *** tvb has quit IRC | 19:27 | |
| *** tvb has joined #openstack-infra | 19:28 | |
| fungi | i saved a copy of the broken workspace from -8 last week and can compare to it as well | 19:28 |
| *** tvb is now known as Guest48536 | 19:28 | |
| fungi | i'm starting to wonder if gerrit-git-prep shouldn't do a git gc before git remote update | 19:32 |
| *** sarob has quit IRC | 19:34 | |
| fungi | so far on each occurrence of these i've seen, git fsck reports one or more dangling commits which git gc takes care of and after that git remote update works fine | 19:35 |
| fungi | though it's entirely possible the dangling commits reported by fsck are unrelated and some other side effect of te gc is what's getting it working again | 19:36 |
| sdague | fungi: ok, -5 is bad now too | 19:39 |
| sdague | 2013-09-23 19:33:18.964 | Building remotely on centos6-5 in workspace /home/jenkins/workspace/gate-nova-python26 | 19:39 |
| sdague | that failed in the gate | 19:39 |
| fungi | fixing | 19:44 |
| mordred | fungi: maybe if ! `git remote update`; then git gc ; git remote upate ; fi ? | 19:46 |
| fungi | mordred: sure, that might make snese | 19:47 |
| fungi | sense too | 19:47 |
| *** yolanda has quit IRC | 19:48 | |
| sdague | yeh, -5's killed 2 gate changes atm | 19:49 |
| *** ruhe has joined #openstack-infra | 19:52 | |
| mrodden | so ... i must have missed something ... is Tuskar the new Ironic or is it the new Tripleo? | 19:52 |
| jeblair | this is really helpful for exercising nnfi | 19:52 |
| jeblair | it's like chaos monkey for zuul | 19:53 |
| *** dmakogon__ has joined #openstack-infra | 19:53 | |
| *** dmakogon_ has quit IRC | 19:53 | |
| *** dmakogon__ has quit IRC | 19:53 | |
| *** dmakogon_ has joined #openstack-infra | 19:53 | |
| SpamapS | what is the session store for gerrit? I've had to re-sign-in quite a few times today. | 19:54 |
| *** schwicht has joined #openstack-infra | 19:54 | |
| sdague | jeblair: so is nnfi actually happening? | 19:54 |
| sdague | the UI just not updating for it/ | 19:54 |
| sdague | ? | 19:54 |
| jeblair | SpamapS: any chance you had multiple tabs open? try closing them all and starting over | 19:54 |
| *** Guest48536 is now known as tvb | 19:55 | |
| *** tvb has joined #openstack-infra | 19:55 | |
| SpamapS | jeblair: yeah I always have quite a few open | 19:55 |
| jeblair | sdague: yes, nnfi is in production; the change to make it apparent on the status screen is in review. i'm fixing bugs with both | 19:55 |
| sdague | jeblair: cool | 19:55 |
| fungi | jeblair: our entire developer base is like chaos monkey for all of our infrastructure | 19:55 |
| sdague | heh | 19:55 |
| SpamapS | jeblair: is there some server affinity magic going on? | 19:55 |
| sdague | yeh, well the gate's good and busy today as well. Trying to help clear the boards on some easy stuff | 19:56 |
| jeblair | SpamapS: no -- google javascript magic | 19:56 |
| jog0 | mordred: http://docs-draft.openstack.org/26/47226/5/check/gate-hacking-docs/9610239/doc/build/html/ nice! | 19:56 |
| jeblair | SpamapS: if you have a tab with an old session around, it will invalidate a newer session | 19:56 |
| jeblair | SpamapS: so if you sign in, you need to not use any tabs that were opened during an older session | 19:56 |
| jeblair | SpamapS: (but openening new tabs with your new session is fine) | 19:57 |
| SpamapS | jeblair: weird | 19:57 |
| jeblair | SpamapS: isn't GWT great? | 19:57 |
| mordred | jog0: sexy right? | 19:57 |
| mordred | SpamapS: Java! It's for winners! | 19:58 |
| jog0 | mordred: very | 19:58 |
| mordred | jog0: I just did the same thing for tripleo-incubator | 19:58 |
| mordred | speaking of ... | 19:58 |
| jog0 | the formatting is really nice good job to whoever did that one | 19:58 |
| mordred | jeblair: for people who have been relying on github's md/rst rendering feature, I've been going through and moving the files into sphinx trees | 19:58 |
| mordred | jeblair: which, rather than adding rst/md render support to cgit, I think is actually the right-er answer | 19:59 |
| mordred | jog0: that's using dhellmann's oslo.sphinx thing | 19:59 |
| jog0 | dhellmann: ^ great job | 19:59 |
| clarkb | sorry stepped away for lunch | 19:59 |
| dhellmann | jog0: I just put our theme in a library, no big deal. :-) | 20:00 |
| *** sarob has joined #openstack-infra | 20:00 | |
| fungi | "librarification: it's a big deal" | 20:01 |
| dhellmann | fungi: I think that should be Oslo's motto :-) | 20:02 |
| *** fbo is now known as fbo_away | 20:02 | |
| clarkb | jeblair: I think https://review.openstack.org/#/c/46124/ needs a rebase then we can push it through | 20:02 |
| jog0 | dhellmann: ++ | 20:03 |
| fungi | all the projects need pithy mottoes | 20:04 |
| dhellmann | & t-shirts | 20:04 |
| clarkb | if only you could print moving gifs on tshirts | 20:05 |
| fungi | mordred: i agree--that was actually my feeling on the cgit autorendering proposal. i'm in favor of contextual syntax highlighting, but rendering is really best published somewhere | 20:05 |
| dhellmann | clarkb: the fiber optics don't do well in the washing machine | 20:05 |
| mtreinish | jog0: about https://review.openstack.org/#/c/47491/ I think to use tox to run flake8 as a gating job we need a setup.py(and other changes to tox.ini) does that mean it's time to use cookiecutter? | 20:06 |
| jeblair | mordred, sdague: the current nnfi visualizer i'm working on doesn't handle the current status because there's a change with three items directly behind it. iow, zuul is creating more complex trees than i anticipated. | 20:06 |
| jog0 | mtreinish: you read my mind | 20:06 |
| clarkb | mtreinish: correct and yes. jog0 mentioned cookiecuttering elasticRecheck for these reasons and to make it setup.py installable because that will help deployment | 20:06 |
| jog0 | was about to do that unles you want to go | 20:06 |
| mtreinish | jog0: no you get the honors | 20:06 |
| *** rockyg has joined #openstack-infra | 20:06 | |
| openstackgerrit | A change was merged to openstack-dev/hacking: Move the hacking guidelines to sphinx docs https://review.openstack.org/47226 | 20:06 |
| jog0 | will do it in a but | 20:07 |
| jeblair | mordred, sdague: (that can happen with some kinds of failing changes) | 20:07 |
| jog0 | bit* | 20:07 |
| *** julim has joined #openstack-infra | 20:08 | |
| openstackgerrit | Jeremy Stanley proposed a change to openstack-infra/config: Garbage collect Git repos in gerrit-git-prep https://review.openstack.org/47915 | 20:09 |
| fungi | sdague: mordred: ^ maybe like that | 20:09 |
| clarkb | fungi: is git having a hard time dealing with lots of unreachable commits? | 20:10 |
| fungi | clarkb: it's entirely possible | 20:10 |
| mtreinish | jog0: ok I do have the update for 47491 ready that adds the pep8 job to be gating on the project. I can push that out | 20:10 |
| jog0 | go for it | 20:10 |
| fungi | clarkb: though the places i've seen this, fsck generally only mentions a handful of dangling commits | 20:10 |
| fungi | clarkb: and seems to just be nova+centos6 | 20:11 |
| clarkb | interesting | 20:11 |
| clarkb | nova is a large repo and centos6 comes with old git | 20:11 |
| clarkb | nova may be large/special enough to tickle bugs in that old git (sounds familiar) | 20:12 |
| fungi | yep. i have four example workspaces (so far) if you want to play around with them | 20:12 |
| fungi | all exhibiting basically the same behavior with differing dangling commits reported by fsck and all fixed by gc | 20:12 |
| fungi | and i tested that specific conditional block on one as well, just for completeness, before submitting for review | 20:13 |
| *** vipul is now known as vipul-away | 20:13 | |
| *** vipul-away is now known as vipul | 20:13 | |
| clarkb | fungi: our change looks good to jeblair and I. I didn't +approve it because it hasn't been through check tests first, but g-g-p isn't check tested and the zuul queues are a bit long irght now | 20:13 |
| fungi | yeah | 20:14 |
| openstackgerrit | Matthew Treinish proposed a change to openstack-infra/config: Add elastic-recheck to openstack-infra https://review.openstack.org/47491 | 20:15 |
| fungi | since config doesn't share a gate queue with anything, i'll just punch it through | 20:15 |
| fungi | otherwise things could just get worse before they get better | 20:16 |
| clarkb | ++ | 20:16 |
| lifeless | jeblair: btw I mailed openstack-infra@lists.o.o about the baremetal testing plan | 20:18 |
| *** ruhe has quit IRC | 20:19 | |
| *** Ajaeger1 has quit IRC | 20:19 | |
| openstackgerrit | A change was merged to openstack-infra/config: Remove the now vestigal pyflakes tox env https://review.openstack.org/47641 | 20:20 |
| openstackgerrit | A change was merged to openstack-infra/config: Garbage collect Git repos in gerrit-git-prep https://review.openstack.org/47915 | 20:20 |
| *** dkliban has quit IRC | 20:22 | |
| *** kiranmurari has quit IRC | 20:22 | |
| *** dprince has quit IRC | 20:30 | |
| *** SergeyLukjanov has quit IRC | 20:30 | |
| mordred | jeblair: stellar | 20:31 |
| clarkb | fyi I just approved https://review.openstack.org/#/c/46499/ I plan on keeping an eye on that but if you see anything funny with teh bug commit message stuff let me know | 20:31 |
| Ryan_Lane | hey, are you guys using elastic search for anything? | 20:32 |
| clarkb | Ryan_Lane: we are | 20:32 |
| Ryan_Lane | do you have it installed via package? | 20:32 |
| clarkb | Ryan_Lane: I am using their debian packages | 20:32 |
| Ryan_Lane | I'd like to add https://www.mediawiki.org/wiki/Extension:CirrusSearch to the wiki | 20:33 |
| clarkb | Ryan_Lane: which allows me to install a specific version which logstash is cranky about | 20:33 |
| *** sandywalsh has quit IRC | 20:33 | |
| Ryan_Lane | but that requires adding elastic search to the instance | 20:33 |
| Ryan_Lane | how are you adding the package without a repo? does puppet allow evil things like that? :D | 20:33 |
| clarkb | Ryan_Lane: it does let me show you | 20:33 |
| Ryan_Lane | eeeeeevil. heh | 20:34 |
| Ryan_Lane | (wikimedia just keeps a custom repo for stuff like this) | 20:34 |
| clarkb | Ryan_Lane: https://git.openstack.org/cgit/openstack-infra/config/tree/modules/logstash/manifests/elasticsearch.pp#n25 lines 25 to 41 | 20:34 |
| openstackgerrit | A change was merged to openstack-infra/jeepyb: Alter comments left on 'Related' bugs https://review.openstack.org/46499 | 20:35 |
| Ryan_Lane | heh. nice | 20:35 |
| clarkb | totally insecure, but acceptable for the logstash stuff | 20:35 |
| Ryan_Lane | yeah, was about to pm you about how insecure that is ;) | 20:35 |
| Ryan_Lane | should atleast hash the package and check it against the sha1 they provide | 20:36 |
| clarkb | yeah | 20:36 |
| clarkb | Ryan_Lane: if you add that I will gladly use it :) | 20:36 |
| *** morganfainberg has quit IRC | 20:36 | |
| clarkb | Ryan_Lane: I will probably add that when I next need to upgrade elasticserach | 20:36 |
| clarkb | which will happen soonish | 20:36 |
| Ryan_Lane | heh. ok. I'll try. should be as easy as an onlyif | 20:36 |
| *** morganfainberg has joined #openstack-infra | 20:37 | |
| Ryan_Lane | hm. maybe not | 20:37 |
| clarkb | Ryan_Lane: it should be as we can provide the sha1/md5 to check against directly in the puppet | 20:37 |
| *** sarob has quit IRC | 20:38 | |
| Ryan_Lane | right, but package doesn't have it | 20:38 |
| clarkb | oh | 20:38 |
| clarkb | :( | 20:38 |
| *** sarob has joined #openstack-infra | 20:38 | |
| Ryan_Lane | so, maybe a chained exec | 20:38 |
| Ryan_Lane | get, check, and package that depends on check | 20:38 |
| clarkb | and have the check delete the .deb if the sha/md5 doesn't match to be doubly sure puppet can't do anything with it | 20:39 |
| Ryan_Lane | I think about saner ways :) | 20:39 |
| clarkb | sanity? what is that | 20:39 |
| Ryan_Lane | yeah | 20:39 |
| Ryan_Lane | it may be possible to just make the download also do a check | 20:39 |
| clarkb | oh maybe do wget/curl support such things? | 20:39 |
| Ryan_Lane | maybe. I'll see | 20:40 |
| clarkb | we can check in https://git.openstack.org/cgit/openstack-infra/config/tree/modules/logstash/manifests/init.pp#n40 too. | 20:42 |
| * Ryan_Lane nods | 20:42 | |
| *** sarob has quit IRC | 20:42 | |
| Ryan_Lane | :D | 20:44 |
| Ryan_Lane | let's switch it to https, too | 20:44 |
| clarkb | ++ do they support https? | 20:44 |
| Ryan_Lane | I guess http is fine, assuming we check the hash | 20:44 |
| Ryan_Lane | https://download.elasticsearch.org/elasticsearch/elasticsearch/elasticsearch-0.90.5.deb | 20:44 |
| clarkb | might as well use the https | 20:44 |
| * Ryan_Lane nods | 20:45 | |
| *** sandywalsh has joined #openstack-infra | 20:45 | |
| *** KurtMartin is now known as kmartin | 20:45 | |
| clarkb | lifeless: https://review.openstack.org/#/c/46557/1 does that cause problems for you? If you are using our sudoers module which will be used by the user virtual resource instantiation then you will get an admin group | 20:47 |
| clarkb | lifeless: so I don't think it will cause problems | 20:47 |
| openstackgerrit | Matthew Treinish proposed a change to openstack-infra/jeepyb: Revert "Remove tempest from list of projects that direct release." https://review.openstack.org/47921 | 20:47 |
| mtreinish | dkranz: ^^^ from the irc meeting last week | 20:48 |
| dkranz | mtreinish: Thanks. I gave it my +1 :) | 20:49 |
| *** mrmartin has quit IRC | 20:50 | |
| hub_cap | hey wheres the magic bit to start logging irc channels? | 20:50 |
| hub_cap | where bit, im assuming, is in some project like config maybe? | 20:51 |
| lifeless | clarkb: I have an admin group I think, so should be fine | 20:51 |
| clarkb | hub_cap: https://git.openstack.org/cgit/openstack-infra/config/tree/modules/openstack_project/manifests/eavesdrop.pp#n25 | 20:52 |
| lifeless | clarkb: though it's in a copy-pasted file, so needs manual porting. | 20:52 |
| hub_cap | <3 clarkb | 20:52 |
| clarkb | lifeless: ok thanks | 20:52 |
| hub_cap | ill be submitting a changeset shortly clarkb :) | 20:52 |
| *** dmakogon_ has left #openstack-infra | 20:54 | |
| DennyZhang | Does anyone know about tempest stress: I've run a tempest stree, how can I get the performance report, like the error rate, througput, latency, etc? | 20:54 |
| openstackgerrit | Michael Basnight proposed a change to openstack-infra/config: Add trove channel to eavesdrop https://review.openstack.org/47924 | 20:54 |
| clarkb | DennyZhang: you will want to ask in #openstack-qa | 20:54 |
| DennyZhang | thanks, clarkb | 20:55 |
| *** dims has joined #openstack-infra | 20:55 | |
| *** svarnau_ has joined #openstack-infra | 20:56 | |
| *** svarnau has quit IRC | 20:56 | |
| *** wchrisj has quit IRC | 20:57 | |
| openstackgerrit | A change was merged to openstack-infra/config: Make ci-launch directory readable by admin users https://review.openstack.org/46557 | 20:57 |
| clarkb | jeblair: does https://review.openstack.org/#/c/46607/1 conflict with the nnfi changes you are making? if not I will approve | 21:00 |
| *** amotoki has joined #openstack-infra | 21:01 | |
| *** markmcclain1 has quit IRC | 21:03 | |
| *** markmcclain has joined #openstack-infra | 21:03 | |
| jeblair | clarkb: all kinds of conflict, but i'll rebase. | 21:07 |
| jeblair | clarkb: (so feel free to aprv) | 21:08 |
| *** cppcabrera has left #openstack-infra | 21:08 | |
| *** sarob has joined #openstack-infra | 21:09 | |
| clarkb | jeblair: thanks | 21:11 |
| *** sarob has quit IRC | 21:13 | |
| *** pcrews_ has quit IRC | 21:15 | |
| *** lifeless has quit IRC | 21:17 | |
| *** emagana has joined #openstack-infra | 21:17 | |
| *** thomasm has quit IRC | 21:18 | |
| mordred | clarkb, fungi: this is now my current oldest outstanding patch: https://review.openstack.org/#/c/40068/9 | 21:18 |
| mordred | clarkb, fungi it's been sitting there for 10 days with no feedback | 21:18 |
| mordred | clarkb, fungi could I bother you for a review? :) | 21:18 |
| *** lifeless has joined #openstack-infra | 21:19 | |
| openstackgerrit | Ryan Lane proposed a change to openstack-infra/config: Separate logstash/elasticsearch and verify hash https://review.openstack.org/47928 | 21:19 |
| Ryan_Lane | clarkb: ^^ that hasn't been tested. | 21:19 |
| *** lcestari_ has quit IRC | 21:19 | |
| *** kpavel has quit IRC | 21:20 | |
| clarkb | Ryan_Lane: thanks | 21:21 |
| Ryan_Lane | yw | 21:21 |
| clarkb | mordred: starred | 21:21 |
| mordred | fungi: also, if you get a chance, https://review.openstack.org/#/c/41945 could use some love | 21:21 |
| clarkb | jeblair: safe to approve the corresponding zuul change as well? | 21:22 |
| openstackgerrit | A change was merged to openstack-infra/config: Don't show a revision when it's "None" https://review.openstack.org/46607 | 21:22 |
| openstackgerrit | A change was merged to openstack-infra/config: Add job to propose tag merges https://review.openstack.org/41927 | 21:23 |
| *** lifeless has quit IRC | 21:25 | |
| *** pcrews_ has joined #openstack-infra | 21:25 | |
| openstackgerrit | Monty Taylor proposed a change to openstack-infra/config: Gerrit docs improvements - user and groups. https://review.openstack.org/45001 | 21:25 |
| openstackgerrit | Monty Taylor proposed a change to openstack-infra/config: Document review.pp parameters a bit. https://review.openstack.org/44969 | 21:25 |
| openstackgerrit | Monty Taylor proposed a change to openstack-infra/config: Make gerrit DB setup match actual practice. https://review.openstack.org/44993 | 21:25 |
| openstackgerrit | Monty Taylor proposed a change to openstack-infra/config: Phase 3 infra bootstrap docs: gerrit. https://review.openstack.org/44970 | 21:27 |
| openstackgerrit | Monty Taylor proposed a change to openstack-infra/config: Improve Zuul docs. https://review.openstack.org/45163 | 21:27 |
| openstackgerrit | Monty Taylor proposed a change to openstack-infra/config: Document push key acceptance. https://review.openstack.org/45162 | 21:27 |
| openstackgerrit | Monty Taylor proposed a change to openstack-infra/config: Document basic admin hints for jeepyb. https://review.openstack.org/45043 | 21:27 |
| openstackgerrit | Monty Taylor proposed a change to openstack-infra/config: Document Jenkin slave management. https://review.openstack.org/45345 | 21:27 |
| openstackgerrit | Monty Taylor proposed a change to openstack-infra/config: Improve Jenkins documentation. https://review.openstack.org/45215 | 21:27 |
| openstackgerrit | Monty Taylor proposed a change to openstack-infra/config: Explain API projects a little. https://review.openstack.org/45111 | 21:27 |
| mordred | clarkb: I just rebased the -1'd changes out of that set | 21:27 |
| clarkb | cool, I can swing back around to those after the current review round | 21:27 |
| *** dkliban has joined #openstack-infra | 21:29 | |
| mordred | clarkb: I'm doing them - most of them have 2x +2 anyway | 21:30 |
| *** sarob has joined #openstack-infra | 21:31 | |
| *** boris-42 has quit IRC | 21:31 | |
| openstackgerrit | A change was merged to openstack-infra/config: Document review.pp parameters a bit. https://review.openstack.org/44969 | 21:33 |
| openstackgerrit | A change was merged to openstack-infra/config: Make gerrit DB setup match actual practice. https://review.openstack.org/44993 | 21:33 |
| openstackgerrit | A change was merged to openstack-infra/config: Gerrit docs improvements - user and groups. https://review.openstack.org/45001 | 21:33 |
| clarkb | mordred: did you also want jeblair to review https://review.openstack.org/#/c/40068/9 ? | 21:34 |
| clarkb | mordred: also, that script depends on stuff in openstack/requirements, what do you think about putting that script in the repo? | 21:35 |
| clarkb | mordred: instead of in openstack-infra/config? the translations jobs make sense in config because we run them for all the projects and they don't depend on eg projects.txt | 21:36 |
| mordred | clarkb: hrm. I could put it in requirements.txt | 21:36 |
| clarkb | fungi: is fifieldt near UTC or towards australia/asia? | 21:36 |
| mordred | or, blah, requirements | 21:36 |
| *** svarnau_ has quit IRC | 21:36 | |
| clarkb | mordred: I think either works, was just putting that out there | 21:36 |
| fungi | clarkb: he's in australia most of the time, not sure about today specifically | 21:36 |
| *** julim has quit IRC | 21:37 | |
| clarkb | fungi: just remembered we need to sort out nova and cinder then I need to do trove on transifex | 21:37 |
| mordred | clarkb: otoh, it has the jenkins git config stuf fin it | 21:37 |
| mordred | clarkb: let's just leave it there for now | 21:37 |
| fungi | ahh, yes | 21:37 |
| *** svarnau has joined #openstack-infra | 21:37 | |
| mordred | it seems more jenkins-y than not | 21:37 |
| clarkb | mordred: ok | 21:38 |
| clarkb | mordred: it has my +2 I haven't approved yet as I have a hunch jeblair may want to take a look at it | 21:38 |
| fungi | mordred: i agree also because the script depends on things in arbitrary branches of requirements, so does it need branch-specific versions? and if not, which branch does it live in normally? | 21:38 |
| *** boris-42 has joined #openstack-infra | 21:38 | |
| fungi | so outside the requirements repo makes more sense to me | 21:38 |
| mordred | fungi: nod | 21:38 |
| *** zehicle has joined #openstack-infra | 21:39 | |
| clarkb | mordred: oh speaking of branches does that job need to handle all the branches? | 21:39 |
| clarkb | we pin the translation stuff to master only | 21:39 |
| *** mriedem has quit IRC | 21:40 | |
| *** dkranz has quit IRC | 21:40 | |
| Ryan_Lane | -_- I'm getting warnings on unnamespaced global variables, but I'm referencing variables from the parameterized class | 21:40 |
| Ryan_Lane | is there some specific way to list a namespace for that, or should the warnings just be ignored? | 21:40 |
| openstackgerrit | A change was merged to openstack-infra/config: Groups-dev drupal drush update to 6.0.0 https://review.openstack.org/47838 | 21:42 |
| clarkb | Ryan_Lane: I wonder if $version is special in puppet land? what you have should work I think | 21:42 |
| Ryan_Lane | it probably is | 21:42 |
| clarkb | Ryan_Lane: you can try ${elasticsearch::version} | 21:42 |
| Ryan_Lane | this is something I hate about puppet | 21:42 |
| clarkb | or use something like $ver | 21:42 |
| Ryan_Lane | I'll do elasticsearch_version | 21:43 |
| fungi | clarkb: at the moment we pin requirements jobs to master, but want to start branching on havana | 21:43 |
| Ryan_Lane | stupid, stupid puppet | 21:43 |
| openstackgerrit | A change was merged to openstack-infra/config: Groups-dev distbuild checked wrong Git repository commit hash https://review.openstack.org/47844 | 21:43 |
| openstackgerrit | A change was merged to openstack-infra/config: Document basic admin hints for jeepyb. https://review.openstack.org/45043 | 21:43 |
| openstackgerrit | A change was merged to openstack-infra/config: Explain API projects a little. https://review.openstack.org/45111 | 21:43 |
| openstackgerrit | A change was merged to openstack-infra/config: Document push key acceptance. https://review.openstack.org/45162 | 21:43 |
| fungi | clarkb: to prevent projects from introducing requirements changes in stable releases, or to sync them up when needed | 21:43 |
| openstackgerrit | A change was merged to openstack-infra/config: Improve Zuul docs. https://review.openstack.org/45163 | 21:43 |
| openstackgerrit | A change was merged to openstack-infra/config: Improve Jenkins documentation. https://review.openstack.org/45215 | 21:43 |
| openstackgerrit | A change was merged to openstack-infra/config: Document Jenkin slave management. https://review.openstack.org/45345 | 21:43 |
| openstackgerrit | A change was merged to openstack-infra/config: Phase 3 infra bootstrap docs: gerrit. https://review.openstack.org/44970 | 21:44 |
| clarkb | fungi: requirements changes in stable releases should be fine unless we start pinning them when we release | 21:44 |
| clarkb | which is something we seem to go back and forth on | 21:44 |
| *** pcm_ has quit IRC | 21:44 | |
| *** pcrews_ has quit IRC | 21:44 | |
| openstackgerrit | Ryan Lane proposed a change to openstack-infra/config: Separate logstash/elasticsearch and verify hash https://review.openstack.org/47928 | 21:44 |
| fungi | clarkb: and our pre-havana answer at this point is going to be pinning supported stable-1 to the current releases of all dependencies, including transitive ones | 21:44 |
| *** pcrews_ has joined #openstack-infra | 21:44 | |
| fungi | clarkb: point being, say in 4 months a new version of some lib comes out which breaks multiple projects for havana, we would love to be able to update openstack/requirements stable/havana branch and have changes automatically generated for all projects listing that particular dependency | 21:45 |
| clarkb | right | 21:46 |
| clarkb | fungi: or we just pin them all | 21:46 |
| openstackgerrit | A change was merged to openstack-dev/cookiecutter: Add a CONTRIBUTING file with bug tracker info https://review.openstack.org/47658 | 21:46 |
| clarkb | which is the alternative we seem to decide we should do then go back on | 21:46 |
| *** dizquierdo has joined #openstack-infra | 21:46 | |
| fungi | well, pin has a couple different meanings (or at least granularities) here | 21:47 |
| clarkb | pin means foo==X.Y | 21:47 |
| *** lifeless has joined #openstack-infra | 21:47 | |
| clarkb | or >=X.Y,<X.Y+1 | 21:47 |
| *** DennyZhang has quit IRC | 21:47 | |
| fungi | right, well >=whatever_was_already_listed,<=todays_specific_version | 21:47 |
| clarkb | but basically stop testing against any thing new | 21:48 |
| fungi | and including all transitive dependencies in the list in the releases where we're hard-pinning them | 21:48 |
| clarkb | I think where I am confused is why are we doing it for stable+1 but not stable+0? | 21:49 |
| fungi | because those have a tendency to move on their own and right now we don't limit what they can be since we don't list them at all | 21:49 |
| mordred | clarkb, fungi: I'm not sure that scrit handles branches | 21:49 |
| clarkb | mordred: it doesn't. The question is should it | 21:49 |
| clarkb | mordred: if we pin all the things on stable then it doesn't need to | 21:49 |
| mordred | clarkb: we still might update the pins | 21:49 |
| clarkb | but fungi suggests it will need to for stable+0 but not stable+1 | 21:49 |
| clarkb | mordred: I think you can do that manually | 21:50 |
| fungi | mordred: the script doesn't specifically handle multiple branches right now. the reason for moving the hard-coded list of projects out of the script and into the requirements repo was so that we could branch it later as needed | 21:50 |
| *** dizquierdo has quit IRC | 21:50 | |
| mordred | nod | 21:50 |
| mordred | ok. howabout we defer deciding what to do with automatic requirements proposals for stable branches until such a time as we know what we want to do | 21:51 |
| fungi | clarkb: the why stable vs. stable-1 was mostly up to ttx. i think the rationale was that most recent stable might still want loose version caps on dependencies to support security/bugfix updates from them automatically | 21:51 |
| clarkb | mordred: wfm | 21:51 |
| fungi | vs stable-1 where we just hard-cap it at today's specific version, add in all transitive dependencies and ignore it from then on | 21:52 |
| clarkb | fungi: :/ I think I get the argument, but we don't like snowflakes around here | 21:52 |
| clarkb | and that creates 3 of them | 21:52 |
| fungi | clarkb: yes, i personally don't care about the distinction. we could just hard-cap all dependencies on release day, but we still need something to prevent individual projects from landing non-conforming dependency changes without coordination | 21:53 |
| mordred | yes. which is why I think from an infra perspective, the mechanism should be identical across branches | 21:54 |
| mordred | from a content perspective, we can do different thigns with teh files in the repos | 21:54 |
| mordred | jog0: https://review.openstack.org/#/c/47491/3 <-- you were talking about re-doing the repo on top of a cookiecutter repo, yeah? do you want to do that before we suck your stuff in? | 21:54 |
| *** dkehn is now known as dkehn_away | 21:54 | |
| *** boris-42 has quit IRC | 21:55 | |
| fungi | i also don't think we're in a great position to backport the current requirements enforcement to folsom, might be able to get it working for grizzly okay, but for havana it should be a slam dunk since they're already basically in sync | 21:55 |
| jog0 | mordred: yeah | 21:55 |
| jog0 | doing t a we speak | 21:55 |
| clarkb | fungi: I wouldn't worry about folsom too much as it goes away in a month | 21:55 |
| jog0 | that as we* | 21:55 |
| *** CaptTofu has joined #openstack-infra | 21:55 | |
| clarkb | fungi: agreed about havana | 21:56 |
| fungi | clarkb: right. however hard-capping grizzly in a month might be easier than backporting the requirements enforcement to it | 21:56 |
| clarkb | ya, so maybe we keep the snowflakes for one cycle | 21:56 |
| mordred | jog0: k. I'm going to WIP that patch then just to be sure. un-wip it when you're ready? | 21:57 |
| jog0 | sounds good | 21:57 |
| *** matty_dubs is now known as matty_dubs|gone | 21:57 | |
| mordred | clarkb, jd__: I believe we can't land this: https://review.openstack.org/#/c/45862/ because we reverted cloud-archive, right? fungi <--? | 21:58 |
| fungi | mordred: ahh, yep | 21:58 |
| mtreinish | mordred: actually I made an indentation mistake in https://review.openstack.org/#/c/47491/3/modules/openstack_project/files/jenkins_job_builder/config/projects.yaml | 21:58 |
| clarkb | mordred: good catch | 21:58 |
| mtreinish | I was about to respin it | 21:58 |
| mordred | mtreinish: well, just re-submit when you guys are ready and cookiecuttered | 21:58 |
| *** weshay has quit IRC | 21:59 | |
| mtreinish | mordred: ok | 21:59 |
| * fungi needs to semi-afk for a bit... past due to start cooking dinner | 22:00 | |
| openstackgerrit | A change was merged to openstack-infra/gerritlib: fix param ordering to fix review function https://review.openstack.org/47821 | 22:01 |
| *** sarob has quit IRC | 22:04 | |
| *** sarob has joined #openstack-infra | 22:04 | |
| *** mrda has joined #openstack-infra | 22:05 | |
| clarkb | mordred: Shrews: I have reviewed https://review.openstack.org/#/c/46741/ any feedback on that? | 22:06 |
| *** rockyg has quit IRC | 22:07 | |
| jog0 | mordred: cookiecutter failes flake8 heh | 22:07 |
| *** rockyg has joined #openstack-infra | 22:07 | |
| clarkb | ha | 22:07 |
| mtreinish | heh | 22:07 |
| *** sarob has quit IRC | 22:09 | |
| *** rockyg has quit IRC | 22:09 | |
| *** rockyg has joined #openstack-infra | 22:09 | |
| *** rockyg has quit IRC | 22:10 | |
| *** rockyg has joined #openstack-infra | 22:11 | |
| jog0 | ohh its the - in the name | 22:11 |
| *** jhesketh has joined #openstack-infra | 22:11 | |
| openstackgerrit | Ryan Lane proposed a change to openstack-infra/config: Add elasticsearch to the mediawiki node https://review.openstack.org/47933 | 22:12 |
| mordred | jog0: hrm. | 22:12 |
| jog0 | folders named 'elastic-recheck' | 22:12 |
| *** changbl has quit IRC | 22:13 | |
| clarkb | fungi: do you guys chat with fifieldt tomorrow morning? any chance you can point him #openstack-infra's way? possibly during the meeting? want to sort out the transifex things | 22:13 |
| fungi | clarkb: 2200-2300 utc tomorrow | 22:14 |
| *** amotoki has quit IRC | 22:14 | |
| mordred | clarkb, fungi, jeblair: see #openstack-dev for amusing anecdotes about just how stupid the cloudbees jenkins-as-a-service+pull-requests is | 22:15 |
| mordred | or - just how good our stuff is | 22:16 |
| Ryan_Lane | hm. ${elasticsearch_version} still gives me the same warning in the lint | 22:16 |
| Ryan_Lane | I guess I'll try scoping to ${elasticsearch::version} | 22:17 |
| *** tvb has quit IRC | 22:18 | |
| openstackgerrit | Ryan Lane proposed a change to openstack-infra/config: Separate logstash/elasticsearch and verify hash https://review.openstack.org/47928 | 22:19 |
| fungi | mordred: wow, atrocious | 22:19 |
| clarkb | Ryan_Lane: weird, I am not sure why it is doing that then | 22:19 |
| *** vipul is now known as vipul-away | 22:20 | |
| Ryan_Lane | me neither. puppet perplexes me sometimes | 22:20 |
| *** vipul-away is now known as vipul | 22:22 | |
| dhellmann | jeblair, clarkb, fungi : I thought you'd find this cloudbees error (caused by amending a commit and re-pushing to github) fun: https://netflixoss.ci.cloudbees.com/job/aminator-pull-requests/119/ | 22:22 |
| mgagne | Ryan_Lane: try ${version} ? | 22:22 |
| Ryan_Lane | mgagne: that's what I started with :) | 22:22 |
| mgagne | Ryan_Lane: please use $ in your param name | 22:24 |
| Ryan_Lane | shit, am I missing that? | 22:24 |
| mgagne | Ryan_Lane: don't know? you check? =) | 22:24 |
| Ryan_Lane | ah, the other param in that one doesn't use it either | 22:25 |
| clarkb | fungi: https://review.openstack.org/#/c/46805/ when dinner is over. Is that to accomodate projects where the project name != the project name? | 22:25 |
| *** adalbas has quit IRC | 22:25 | |
| Ryan_Lane | that's what I get for following the code that's there ;) | 22:25 |
| fungi | clarkb: yes, wsme in git vs WSME tarball | 22:25 |
| mgagne | Ryan_Lane: puppet-lint or syntax should have cried about it. :-/ | 22:25 |
| jog0 | mtreinish: pushing cookie cutter changes | 22:26 |
| jog0 | still have some code reworking but that can wait | 22:26 |
| Ryan_Lane | mgagne: it didn't because it wasn't used in the manifest, but in an erb | 22:26 |
| clarkb | fungi: thanks | 22:26 |
| openstackgerrit | Ryan Lane proposed a change to openstack-infra/config: Separate logstash/elasticsearch and verify hash https://review.openstack.org/47928 | 22:26 |
| Ryan_Lane | mgagne: thanks for catching that ;) | 22:27 |
| mgagne | Ryan_Lane: np | 22:27 |
| openstackgerrit | A change was merged to openstack-infra/config: Add trove channel to eavesdrop https://review.openstack.org/47924 | 22:27 |
| jog0 | mtreinish: pushed | 22:29 |
| jog0 | mordred: patch is ready, unless mtreinish found another bug | 22:31 |
| *** openstack has joined #openstack-infra | 22:31 | |
| openstackgerrit | Clark Boylan proposed a change to openstack-infra/zuul: Update tox.ini to current standards https://review.openstack.org/47391 | 22:36 |
| clarkb | mordred: ^ I rebased that for you | 22:36 |
| *** rcleere has quit IRC | 22:39 | |
| *** rcleere has joined #openstack-infra | 22:39 | |
| *** thedodd has quit IRC | 22:43 | |
| *** dims has quit IRC | 22:43 | |
| openstackgerrit | Ryan Lane proposed a change to openstack-infra/config: Add elasticsearch to the mediawiki node https://review.openstack.org/47933 | 22:43 |
| *** rcleere has quit IRC | 22:43 | |
| clarkb | we are a bit behind on available centos6 nodes. I think nnfi must have increased throughput such that we are not keeping up again | 22:44 |
| openstackgerrit | Khai Do proposed a change to openstack-infra/config: Setup private gerrit for security reviews https://review.openstack.org/47937 | 22:46 |
| clarkb | jeblair: re NNFI, it appears that more than one failed change can be in the gate queue at a time | 22:47 |
| clarkb | jeblair: shouldn't we always have a max of 1? | 22:47 |
| clarkb | oh wait, things that can't merge will be skipped | 22:47 |
| clarkb | but actual failures should be restarted if they are behind a different failure | 22:47 |
| *** zehicle has quit IRC | 22:49 | |
| *** cthulhup has joined #openstack-infra | 22:50 | |
| openstackgerrit | Khai Do proposed a change to openstack-infra/config: Setup private gerrit for security reviews https://review.openstack.org/47937 | 22:50 |
| *** vipul is now known as vipul-away | 22:51 | |
| mtreinish | jog0: sorry I was driving home. Yeah the indent is wrong in projects.yaml | 22:51 |
| *** datsun180b has quit IRC | 22:51 | |
| mtreinish | 4 spaces instead of 2 | 22:51 |
| clarkb | I just approved https://review.openstack.org/#/c/47644/ which makes the python we have in openstack-infra/config flake8 clean | 22:52 |
| clarkb | I will keep my eyes open for breakage | 22:52 |
| zaro | clarkb: would you like to handle adding new params to heira for this? https://review.openstack.org/#/c/47937 | 22:53 |
| sdague | clarkb: nice | 22:53 |
| clarkb | zaro: not yet, it would be good to get some feedback on whether or not we need all of them and if they should be different than normal gerrit | 22:54 |
| *** cthulhup has quit IRC | 22:54 | |
| clarkb | dhellmann: https://review.openstack.org/#/c/47702/1 doesn't the rawinstall test cover your concern? | 22:54 |
| openstackgerrit | Matthew Treinish proposed a change to openstack-infra/config: Add elastic-recheck to openstack-infra https://review.openstack.org/47491 | 22:55 |
| mtreinish | jog0, clarkb: ^^^ That should fix the tab | 22:55 |
| zaro | clarkb: what is the forum for that? | 22:55 |
| clarkb | zaro: the change you have in gerrit is perfect | 22:56 |
| zaro | i've set it to WIP because of the dependency on heira params. should i remove WIP so others will review? | 22:57 |
| clarkb | zaro: I think it should still be WIP then you can solicit reviews here or in the meeting tomorrow | 22:57 |
| jeblair | clarkb: yeah; sorry i'm bogged down in status page javascript. it's taking quite a bit of work. | 22:57 |
| clarkb | mordred: https://review.openstack.org/#/c/47702/ reviewed. Is my comment about rawinstall correct? | 22:58 |
| clarkb | jeblair: I figured | 22:58 |
| zaro | clarkb: cool, thx. | 22:58 |
| clarkb | jeblair: what is the yeah referring to? (I have pinged you a couple times) | 22:58 |
| jeblair | clarkb: nnfi | 22:58 |
| *** dims has joined #openstack-infra | 23:00 | |
| *** vipul-away is now known as vipul | 23:01 | |
| jhesketh | jeblair, clarkb: Just wondering if you have time to take a quick look at this: https://review.openstack.org/#/c/46871/ | 23:05 |
| jhesketh | (specifically my last comment) | 23:05 |
| clarkb | jhesketh: looking | 23:07 |
| jhesketh | cheers | 23:08 |
| *** dcramer_ has quit IRC | 23:08 | |
| *** kgriffs is now known as kgriffs_afk | 23:15 | |
| *** rnirmal has quit IRC | 23:15 | |
| *** danger_fo is now known as danger_fo_away | 23:20 | |
| clarkb | jhesketh: reviewed and turbo-hipster is a great name for an SNES usb controller with a turbo button | 23:21 |
| *** thomasm has joined #openstack-infra | 23:21 | |
| *** changbl has joined #openstack-infra | 23:22 | |
| jhesketh | we have this thing where we just randomly generate names until a good one comes up | 23:22 |
| jhesketh | laughing-spice formats logs for me | 23:22 |
| jhesketh | clarkb: thanks for reviewing :-) | 23:22 |
| *** hemna is now known as hemnafk | 23:24 | |
| mordred | jhesketh: turbo-hipster? | 23:28 |
| clarkb | mordred: do you have 15 minutes? I have found so many things while reviewing | 23:28 |
| clarkb | mordred: would like to just run through them now with you if possible | 23:28 |
| jhesketh | mordred: https://github.com/rcbau/turbo-hipster | 23:28 |
| *** tvb has joined #openstack-infra | 23:29 | |
| *** tvb is now known as Guest31624 | 23:29 | |
| mordred | oh wow | 23:29 |
| mordred | jeblair: did you see turbo-hipster? (I know you're in javascript land, but you might find it interesting) | 23:29 |
| clarkb | mordred: I am just going to start https://review.openstack.org/#/c/46996/ is a potentially backward incompatible change. Are we just biting the bullet as pbr hasn't been included in an integrated release yet? | 23:30 |
| * mordred HATES the name... but ... | 23:30 | |
| * mordred did release a piece of software called pbr | 23:30 | |
| mordred | so | 23:30 |
| * mordred probably should shut up | 23:30 | |
| clarkb | mordred: https://review.openstack.org/#/c/46736/ has a major bug in it (see comments) | 23:30 |
| clarkb | we sorry | 23:31 |
| jog0 | mtreinish: thanks | 23:31 |
| clarkb | wow "er sorry" https://review.openstack.org/#/c/46741/ is the one with a bug | 23:31 |
| mordred | clarkb: both aX and devX are pre-release versions | 23:31 |
| clarkb | mordred: can you confirm my comment on https://review.openstack.org/#/c/47702/ is correct? | 23:32 |
| jog0 | mordred: any comments on how to do the code layout for elastic-search, with the dash in the name and all | 23:32 |
| mordred | clarkb: and we've never made any claim about monotonic increasing numbers (and tell people to use PBR_VERSION to override) | 23:32 |
| mordred | jog0: replace the - with a _ | 23:32 |
| jog0 | mordred: and leave the name of the repo? | 23:33 |
| clarkb | mordred: cool I will approve that then | 23:33 |
| mordred | jog0: and then there is _either_ a bug in cookiecutter, or in our template, for producing a python package of that name | 23:33 |
| *** Guest31624 has quit IRC | 23:33 | |
| clarkb | mordred: https://review.openstack.org/#/c/47164/ should be abandoned or we need to discuss it with lifeless | 23:33 |
| mordred | jog0: yes. and of the pypi package | 23:33 |
| mordred | clarkb: good question. Shrews: https://review.openstack.org/#/c/46741/ | 23:34 |
| jog0 | mordred: so elastic-recheck/elastic_recheck/code.py ? | 23:34 |
| mordred | yeah | 23:34 |
| jog0 | and pip isntall elastic-recheck, import elastic_recheck | 23:34 |
| mordred | yeah | 23:34 |
| jog0 | makes sense to me | 23:34 |
| jog0 | thanks | 23:34 |
| *** gyee has quit IRC | 23:34 | |
| mordred | sure! | 23:35 |
| *** rfolco has quit IRC | 23:35 | |
| jog0 | mtreinish mordred: now that the repo is close to being under -infra, which reminds me we need to create the gerrit groups for it, I will draft the email to announce it | 23:35 |
| lifeless | clarkb: right, tie is not python. | 23:36 |
| clarkb | jog0: the groups will be greated when the change merges then we can populate them | 23:36 |
| lifeless | mordred: ^ | 23:36 |
| jog0 | clarkb: neato | 23:36 |
| * jog0 eates a really late lunch .. and dislikes jury duty | 23:38 | |
| clarkb | I was going to approve the dev pbr change but noticed http://logs.openstack.org/96/46996/2/check/gate-tempest-devstack-vm-postgres-full/b7f8302/console.html fails anyone seen that before? | 23:39 |
| clarkb | I can't find an obvious bug for it on the recheck page or in the tempet bug list | 23:39 |
| mordred | lifeless: btw - you may also want to check out turbo-hipster | 23:40 |
| mordred | lifeless: it's a zuul-gearman based non-jenkins job runner | 23:40 |
| jog0 | clarkb: I think I have seen that | 23:40 |
| clarkb | mordred: https://review.openstack.org/#/c/47164/ can that be abandoned? | 23:40 |
| jog0 | check logstash :) | 23:40 |
| jog0 | and fie a bug, we can always mark as a dup later | 23:41 |
| *** mrodden has quit IRC | 23:41 | |
| mordred | lifeless: are you saying that there is not python code in tie? | 23:41 |
| clarkb | ok will do | 23:41 |
| mordred | lifeless: (I agree that that patch was a mistake given current state of tie repo - just asking) | 23:43 |
| clarkb | mordred: I think there is a setup.py and that is it | 23:45 |
| clarkb | before I forget I will probably leave a little early tomorrow as I am going to a Mariners game and need to be there in time for the first pitch | 23:46 |
| *** mrda has quit IRC | 23:47 | |
| *** arosen has quit IRC | 23:47 | |
| *** thomasm has quit IRC | 23:48 | |
| *** dkliban has quit IRC | 23:49 | |
| *** matsuhashi has joined #openstack-infra | 23:50 | |
| *** dcramer_ has joined #openstack-infra | 23:50 | |
| *** arosen has joined #openstack-infra | 23:51 | |
| openstackgerrit | Joshua Hesketh proposed a change to openstack-infra/config: Test nova changes against a real database https://review.openstack.org/46871 | 23:51 |
| reed | a question about the concept of 'committer' on gerrit | 23:51 |
| reed | this changeset: https://review.openstack.org/#/c/41213/ | 23:51 |
| reed | owner is sukhdev, committer is Markmc | 23:51 |
| jhesketh | clarkb: when you have a moment I've updated the job checker to have a list of non jjb jobs | 23:52 |
| reed | this review: https://review.openstack.org/#/c/46796/ has sukhdev owner and committer... | 23:52 |
| *** mrodden has joined #openstack-infra | 23:52 | |
| reed | now, question is: what is 'committer' again? | 23:52 |
| clarkb | reed: owner is the person that pushed the first patchset. commit auther authored the patch, committer committed the patch | 23:52 |
| clarkb | reed: author and committer can be different if for example someome else rebases a commit for you to fix merge conflicts | 23:53 |
| clarkb | jhesketh: looking | 23:53 |
| reed | so author is not necessarily the owner? | 23:53 |
| *** mrodden1 has joined #openstack-infra | 23:54 | |
| jhesketh | reed: you can change author via git commands | 23:54 |
| reed | oh, wait.. author is what we need | 23:54 |
| jhesketh | but the owner still owns the review | 23:54 |
| reed | let me look at this again | 23:54 |
| *** dkranz has joined #openstack-infra | 23:54 | |
| jhesketh | reed: something like git commit --amend --author "New Author Name <email@address.com>" ? | 23:55 |
| reed | yeah... ok, I think I got it right | 23:55 |
| clarkb | jhesketh: reviewed, there was one small thing that snuck by after your testing :) | 23:55 |
| jhesketh | woops! | 23:56 |
| openstackgerrit | Joshua Hesketh proposed a change to openstack-infra/config: Test nova changes against a real database https://review.openstack.org/46871 | 23:56 |
| *** mrodden has quit IRC | 23:57 | |
| clarkb | Alex_Gaynor: https://review.openstack.org/#/c/47756/2 does requirements pass when run under pypy? | 23:57 |
| clarkb | Alex_Gaynor: any chance you could leave a comment with testing info? | 23:57 |
| *** sarob has joined #openstack-infra | 23:57 | |
| *** pcm_ has joined #openstack-infra | 23:58 | |
| reed | problem is that gerrit json api only exposes the uploader and the owner http://gerrit.googlecode.com/svn-history/r6507/documentation/2.2.1/json.html#patchset | 23:58 |
| *** dcramer_ has quit IRC | 23:59 | |
| *** sarob_ has joined #openstack-infra | 23:59 | |
| *** Ryan_Lane has quit IRC | 23:59 | |
| *** pcm_ has joined #openstack-infra | 23:59 | |
| *** Ryan_Lane1 has joined #openstack-infra | 23:59 | |
| *** Ryan_Lane1 is now known as Ryan_Lane | 23:59 | |
| *** Ryan_Lane has joined #openstack-infra | 23:59 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!