| *** herndon_ has quit IRC | 00:00 | |
| fungi | i'm liking the possibility of embedding gerrit views into storyboard by making use of a stable rest api | 00:00 |
|---|---|---|
| jeblair | fungi: it's a fun idea to think about, but storyboard needs to exist first. | 00:00 |
| krotscheck | fungi: What would your expectations on that be? Diffs in storyboard? Synchronized comment threads? | 00:01 |
| fungi | krotscheck: not sure whether i have expectations at this point. just the potential to tie them together more tightly than we do with launchpad (bug and blueprint linking and updating) would be a huge improvement | 00:02 |
| mordred | ++ | 00:02 |
| mordred | I see that as a natural progression, honestly - sort of like what reviewday is doing now - except not batched | 00:02 |
| mordred | and then taking further steps | 00:02 |
| mordred | I also have no concrete examples | 00:02 |
| mordred | krotscheck: have you seen reviewday? | 00:02 |
| jeblair | let's develop that idea after storyboard tracks bugs. | 00:02 |
| mordred | jeblair: ++ | 00:03 |
| fungi | yeah, working core first, awesome new features later | 00:03 |
| mordred | ALL NEW FEATURES NOW | 00:03 |
| *** herndon has joined #openstack-infra | 00:03 | |
| *** herndon has quit IRC | 00:03 | |
| krotscheck | mordred: Indeed I have. | 00:04 |
| krotscheck | jeblair: I'm just collecting data for the roadmap that colette's putting together. | 00:04 |
| jeblair | storyboard isn't a mythical future repository where we put all of our hopes; it's an actual system that tracks bugs. | 00:05 |
| jeblair | the fact that it is not currently running is bug #1 | 00:05 |
| *** ArxCruz has quit IRC | 00:05 | |
| uvirtbot | jeblair: Error: Could not parse data returned by Launchpad: HTTP Error 503: Service Unavailable | 00:06 |
| jeblair | hehe | 00:06 |
| jeblair | once it's doing that, i think we will have room to consider a scope expansion, and i think it's very exciting. | 00:06 |
| jeblair | but i feel like it's way premature to be putting things like that on a roadmap | 00:07 |
| fungi | (for example, uvirtbot replacement which groks storyboard) | 00:07 |
| sdague | yeh, agreed, less puppies and unicorns | 00:07 |
| * mordred takes blame for storyboard not running ... | 00:07 | |
| * mordred introduced a roadblock and then became absent | 00:07 | |
| fungi | speaking of working, nova tox -r -epy27 works now with new mysql-python and hand-patched virtualenv "1.11.1" (one-liner from the pull request) by updating tox.ini to... | 00:08 |
| fungi | install_command = pip install --allow-external netaddr --allow-unverified netaddr -U {opts} {packages} | 00:08 |
| *** mriedem has quit IRC | 00:08 | |
| mordred | fungi: neat! netaddr is the only one remaining? | 00:08 |
| fungi | so looks like netaddr is the only one it needs | 00:09 |
| jeblair | fungi: nice | 00:09 |
| openstackgerrit | Sean Dague proposed a change to openstack-infra/config: Gerrit-2.8: Add secondary index support https://review.openstack.org/60080 | 00:09 |
| fungi | remember, this is just nova. no idea yet where else there be dragons | 00:09 |
| mordred | fungi: that's awesome | 00:09 |
| zaro | fungi: i added the 'CLA Accepted - ICLA' group to the gerrit db. however i can't seem to test where it works anymore since i've already accepted the agreement. is there a way to not agree anymore so i can go thru process again? | 00:09 |
| fungi | zaro: interesting. did you restart gerrit and make sure project.conf is set back to using that group name for the icla members? | 00:10 |
| mordred | we should pressure netaddr upstream | 00:10 |
| fungi | gerrit shouldn't think you've agreed to it | 00:10 |
| *** michchap has quit IRC | 00:10 | |
| mordred | dhellmann: you don't know David P. D. Moss do you? | 00:10 |
| *** michchap has joined #openstack-infra | 00:10 | |
| fungi | i thought there were discussions about replacing netaddr in nova... any guesses how far that got? | 00:11 |
| zaro | fungi: yes, double checked and the project.config is back to original. | 00:11 |
| zaro | fungi: from my point of view everything looks like it is working. | 00:12 |
| krotscheck | jeblair: Roadmaps, by their nature, are premature. And optimistic. And full of unicorns and rainbows. That doesn't make them useless, it just makes them shiny. | 00:12 |
| mordred | if it's the last thing nova is depending on, perhaps just poking the dude and saying "hi, could you run python setup.py sdist upload one time?" | 00:12 |
| fungi | zaro: okay, then the other possibility is that it was just the disjoint group name breaking gerrit, and that it actually checks both that group and the account_agreements table for backward-compatability | 00:12 |
| *** mkoderer_ has quit IRC | 00:12 | |
| jgriffith | mtreinish: wanted to confirm this is what you had in mind? http://paste.openstack.org/show/59739/ | 00:12 |
| fungi | zaro: i can poke around at the db for a few minutes and see if i can un-agree myself | 00:13 |
| jgriffith | mtreinish: so the addCleanup directive ensures that item is always called on teardown correct? | 00:13 |
| jeblair | #status ok see: https://etherpad.openstack.org/p/pip1.5Upgrade | 00:13 |
| openstackstatus | NOTICE: see: https://etherpad.openstack.org/p/pip1.5Upgrade | 00:13 |
| *** ChanServ changes topic to "Discussion of OpenStack Project Infrastructure | Docs http://ci.openstack.org/ | Bugs https://launchpad.net/openstack-ci | Code https://git.openstack.org/cgit/openstack-infra/" | 00:13 | |
| *** rongze has joined #openstack-infra | 00:14 | |
| fungi | mordred: okay, one other wrinkle... if we update the tox.ini to use that, it will work with pip 1.5 and break under 1.4.1 and earlier... how do you make pip-version-dependent branching logic in tox.ini? | 00:14 |
| *** mkoderer_ has joined #openstack-infra | 00:15 | |
| mordred | fungi: you cannot. | 00:15 |
| fungi | mordred: nevermind--as tox bundles virtualenv which carries latest pip, i guess that will never actually cause a problem? | 00:16 |
| fungi | or... hrm, i should just test | 00:16 |
| *** svarnau has quit IRC | 00:17 | |
| *** rongze has quit IRC | 00:18 | |
| zaro | fungi: actually you should not need to. you are not on the CLA accepted group. | 00:18 |
| zaro | fungi: just try to accept the agreement. | 00:19 |
| zaro | fungi: actually i can un-accept myself thru the GUI. | 00:19 |
| fungi | zaro: yeah, so even with my account listed in account_agreements as having agreed to it, gerrit 2.8 apparently only checks membership on the "CLA Accepted - ICLA" group and decides i haven't, but i can successfully agree and it adds me to the group no problem | 00:21 |
| fungi | which means we probably will have to script populating that group on upgrade, but it should be doable | 00:21 |
| *** CaptTofu has quit IRC | 00:21 | |
| zaro | fungi: yep, i just went thru the process again and it seems to work. | 00:22 |
| fungi | mordred: i tested that modified tox.ini on a system with only pip 1.4.1 globally installed and it seems to work fine, so no worries there after all | 00:22 |
| clarkb | fungi: but the same tox.ini can be used with different versions of tox so you will need to set the min tox version | 00:23 |
| *** praneshp has quit IRC | 00:24 | |
| fungi | clarkb: well, there isn't a newer version of tox which solves anything for this... people may need to update the version of virtualenv they're running, since we'll no longer be compatible with virtualenv 1.10.1, for example | 00:24 |
| fungi | i think... need to try that next | 00:25 |
| openstackgerrit | Stefano Maffulli proposed a change to openstack-infra/config: Discussion list for the personas working group https://review.openstack.org/64752 | 00:26 |
| *** jergerber has quit IRC | 00:29 | |
| openstackgerrit | Stefano Maffulli proposed a change to openstack-infra/config: Discussion list for the personas working group https://review.openstack.org/64753 | 00:30 |
| reed | damn it | 00:30 |
| fifieldt | :D | 00:30 |
| reed | git --amen is not the same thing as git --amend | 00:31 |
| fifieldt | :D | 00:33 |
| fungi | it sounds a lot more affirmative though | 00:34 |
| reed | damn it, now I'm stuck in git-review | 00:34 |
| fungi | mordred: clarkb: okay, different bad news... if we add --allow-unverified (which netaddr doesn't work without), then anyone running tox on a project using that option will need 1.11.x (just confirmed it breaks with 1.10.1 globally installed) | 00:36 |
| *** rakhmerov has joined #openstack-infra | 00:38 | |
| fungi | er, will need *virtualenv* 1.11.x | 00:38 |
| reed | what exactly is causing this error? http://logs.openstack.org/53/64753/1/check/gate-config-puppet-syntax/2c834eb/console.html | 00:41 |
| reed | is it my fault? | 00:41 |
| clarkb | reed: yes you need a comma after the description string | 00:41 |
| reed | thanks clarkb | 00:42 |
| openstackgerrit | Stefano Maffulli proposed a change to openstack-infra/config: Discussion list for the personas working group https://review.openstack.org/64753 | 00:42 |
| *** rakhmerov has quit IRC | 00:43 | |
| fungi | reed: looks like you spotted it before i could leave a review comment to that effect | 00:43 |
| reed | fungi, clarkb told me what to do | 00:43 |
| fungi | very fast ;) | 00:43 |
| reed | :) | 00:43 |
| zaro | clarkb, fungi : i've installed the core plugins to review-dev.o.o | 00:45 |
| zaro | clarkb, fungi : i'm guessing we should probably turn on review searching to test as well? | 00:47 |
| fungi | zaro: probably, yes | 00:47 |
| *** ryanpetrello has quit IRC | 00:49 | |
| *** praneshp has joined #openstack-infra | 00:53 | |
| *** praneshp_ has joined #openstack-infra | 00:56 | |
| *** praneshp has quit IRC | 00:58 | |
| *** praneshp_ is now known as praneshp | 00:58 | |
| *** jaypipes has quit IRC | 01:02 | |
| *** ^d has quit IRC | 01:11 | |
| *** rongze has joined #openstack-infra | 01:15 | |
| jeblair | reed: git commit --amen should really do something | 01:16 |
| reed | jeblair, and I thought it was just a typo :> | 01:16 |
| *** melwitt has joined #openstack-infra | 01:17 | |
| *** rongze has quit IRC | 01:20 | |
| *** AaronGr is now known as AaronGr_Zzz | 01:20 | |
| *** banix has joined #openstack-infra | 01:32 | |
| *** AaronGr_Zzz is now known as AaronGr | 01:36 | |
| *** rongze has joined #openstack-infra | 01:38 | |
| *** rakhmerov has joined #openstack-infra | 01:39 | |
| *** CaptTofu has joined #openstack-infra | 01:39 | |
| *** rakhmerov has quit IRC | 01:44 | |
| *** weshay has joined #openstack-infra | 01:44 | |
| *** StevenK has quit IRC | 01:47 | |
| *** jheiss has quit IRC | 01:48 | |
| *** rockstar has quit IRC | 01:48 | |
| *** StevenK has joined #openstack-infra | 01:48 | |
| *** jheiss has joined #openstack-infra | 01:48 | |
| *** AaronGr has quit IRC | 01:49 | |
| *** Ryan_Lane has joined #openstack-infra | 01:49 | |
| *** banix has quit IRC | 01:51 | |
| *** ryanpetrello has joined #openstack-infra | 01:53 | |
| *** sarob has joined #openstack-infra | 01:53 | |
| openstackgerrit | Michael Krotscheck proposed a change to openstack-infra/config: Added Storyboard-Webclient to gerrit https://review.openstack.org/64761 | 01:53 |
| *** praneshp has quit IRC | 01:55 | |
| *** thedodd has joined #openstack-infra | 01:55 | |
| openstackgerrit | Michael Krotscheck proposed a change to openstack-infra/config: Added Storyboard-Webclient to gerrit https://review.openstack.org/64761 | 01:55 |
| *** AaronGr has joined #openstack-infra | 01:57 | |
| *** sarob_ has quit IRC | 01:57 | |
| *** rockstar has joined #openstack-infra | 01:58 | |
| *** rockstar has joined #openstack-infra | 01:58 | |
| *** sarob has quit IRC | 01:58 | |
| *** thedodd has quit IRC | 01:58 | |
| *** AaronGr_Zzz has joined #openstack-infra | 01:59 | |
| *** krotscheck has quit IRC | 02:02 | |
| *** nati_ueno has joined #openstack-infra | 02:03 | |
| *** CaptTofu_ has joined #openstack-infra | 02:06 | |
| *** CaptTofu has quit IRC | 02:08 | |
| *** CaptTofu has joined #openstack-infra | 02:09 | |
| *** CaptTofu_ has quit IRC | 02:11 | |
| *** weshay has quit IRC | 02:13 | |
| *** dguitarbite has joined #openstack-infra | 02:16 | |
| *** dguitarbite__ has joined #openstack-infra | 02:17 | |
| *** banix has joined #openstack-infra | 02:17 | |
| *** rongze has quit IRC | 02:18 | |
| *** ryanpetrello has quit IRC | 02:19 | |
| *** rongze has joined #openstack-infra | 02:19 | |
| *** rongze has quit IRC | 02:19 | |
| *** rongze has joined #openstack-infra | 02:20 | |
| *** CaptTofu has quit IRC | 02:27 | |
| *** CaptTofu has joined #openstack-infra | 02:27 | |
| *** rongze has quit IRC | 02:32 | |
| *** rongze has joined #openstack-infra | 02:33 | |
| *** rakhmerov has joined #openstack-infra | 02:40 | |
| *** rakhmerov has quit IRC | 02:44 | |
| *** dguitarbite__ has quit IRC | 02:45 | |
| *** dguitarbite has quit IRC | 02:45 | |
| *** AaronGr has quit IRC | 02:57 | |
| *** julim has quit IRC | 02:59 | |
| *** cyeoh has quit IRC | 03:01 | |
| *** cyeoh has joined #openstack-infra | 03:02 | |
| *** zhiwei has joined #openstack-infra | 03:08 | |
| zhiwei | clarkb: hi | 03:08 |
| zhiwei | clarkb: Did you see my rename project request email in mailling list? | 03:09 |
| clarkb | zhiwei: I did, haven't had time to compose a reply | 03:11 |
| clarkb | I will do that now | 03:11 |
| zhiwei | ok, thank you very much. | 03:11 |
| *** melwitt has quit IRC | 03:13 | |
| *** AaronGr_Zzz is now known as AaronGr | 03:15 | |
| *** Daviey has quit IRC | 03:15 | |
| *** Daviey has joined #openstack-infra | 03:22 | |
| *** loq_mac has quit IRC | 03:25 | |
| *** boris-42 has quit IRC | 03:37 | |
| *** rakhmerov has joined #openstack-infra | 03:40 | |
| *** CaptTofu has quit IRC | 03:41 | |
| *** CaptTofu has joined #openstack-infra | 03:41 | |
| *** ryanpetrello has joined #openstack-infra | 03:42 | |
| *** rakhmerov has quit IRC | 03:45 | |
| *** ryanpetrello has quit IRC | 03:46 | |
| *** harlowja is now known as harlowja_away | 03:51 | |
| *** david-lyle_ has joined #openstack-infra | 03:56 | |
| openstackgerrit | A change was merged to openstack-dev/hacking: Enforce import grouping https://review.openstack.org/52221 | 04:03 |
| openstackgerrit | A change was merged to openstack-dev/hacking: Enforce grouping like imports together https://review.openstack.org/54402 | 04:03 |
| *** rongze has quit IRC | 04:04 | |
| *** coolsvap has joined #openstack-infra | 04:07 | |
| *** coolsvap has quit IRC | 04:10 | |
| *** coolsvap has joined #openstack-infra | 04:10 | |
| *** david_lyle has joined #openstack-infra | 04:18 | |
| *** david-lyle_ has quit IRC | 04:21 | |
| anteaya | just in case anyone is keeping score, change of plan. I am flying over the north pole again and am planning on being in perth sometime late Saturday night | 04:22 |
| clarkb | anteaya: weather causing problems? | 04:22 |
| pleia2 | anteaya: safe travels | 04:22 |
| anteaya | pleia2: thanks | 04:22 |
| anteaya | clarkb: well, a number of things, that meant I was going to miss my LAX to MEL flight | 04:22 |
| *** jerryz has quit IRC | 04:23 | |
| anteaya | so since I was still at YYZ, I said I should just do HK to Perth and the captian on the flight agreed with me | 04:23 |
| anteaya | and I was able to get my flight rescheduled before YYZ to LAX took off | 04:23 |
| anteaya | so here I be | 04:23 |
| clarkb | nice | 04:23 |
| anteaya | yay for airport food | 04:24 |
| pleia2 | hehe | 04:24 |
| anteaya | getting to know some good ones | 04:24 |
| anteaya | YYZ terminal 3, B and C gates have way better food options | 04:24 |
| anteaya | in case you care | 04:24 |
| anteaya | better food options than A gate (flights to US) | 04:25 |
| anteaya | oh and more fun, US passport check now requires going to a computer kiosk complete with photograph | 04:25 |
| anteaya | grimace for Uncle Sam | 04:25 |
| clarkb | fingerprints too I think | 04:26 |
| anteaya | not yet for me | 04:26 |
| jgriffith | mordred: clarkb is there a bug for the mysql-python issue that we should use for recheck? | 04:26 |
| clarkb | jgriffith: I don't think so | 04:26 |
| jgriffith | clarkb: is it worth logging one or just doing a "no bug" in this case | 04:26 |
| clarkb | no bug should be fine | 04:26 |
| jgriffith | clarkb: thanks | 04:26 |
| clarkb | anteaya: so your route is now similar ish to mine | 04:27 |
| clarkb | just in different directions :) | 04:27 |
| anteaya | clarkb: really? | 04:27 |
| anteaya | are you visiting Santa too? | 04:27 |
| anteaya | and for what reason | 04:28 |
| anteaya | oh yeah, dubai | 04:28 |
| anteaya | do you hit the pole for Seattle to dubai? | 04:28 |
| clarkb | I believe so | 04:28 |
| pleia2 | I have an 8 hour layover in aukland, that'll be fun | 04:28 |
| pleia2 | auckland too | 04:28 |
| anteaya | I'm 8 hours in hk | 04:28 |
| anteaya | more over priced shopping of stuff I would never ever wear | 04:29 |
| pleia2 | hah | 04:29 |
| clarkb | but they have dim sum in the airport :) | 04:29 |
| anteaya | I found a good hk airport restaurant | 04:29 |
| anteaya | does a great vegetarian squash soup | 04:29 |
| anteaya | looking forward to it | 04:29 |
| anteaya | plus Cathay pacific has a great touch screen | 04:30 |
| anteaya | and a good movie selection | 04:30 |
| anteaya | I hope they have some fresh ones | 04:30 |
| anteaya | so help me here, Seattle to dubai via the north pole, dubai to auckland? | 04:31 |
| clarkb | dubai to perth | 04:31 |
| anteaya | don't you have to fly over perth to get to auckland? | 04:31 |
| anteaya | so where does 8 hrs in auckland come in? | 04:31 |
| clarkb | that is pleia2 | 04:31 |
| anteaya | ah sorry | 04:31 |
| anteaya | you are both the same colour and name length | 04:31 |
| anteaya | you became one in my eyes | 04:31 |
| anteaya | I was going to say, who schedules a dubai to auckland to get to perth | 04:32 |
| pleia2 | hehe | 04:32 |
| anteaya | pleia2: so you are SF to auckland? | 04:32 |
| pleia2 | anteaya: yep | 04:33 |
| anteaya | I think notmyname has something similar | 04:33 |
| anteaya | swift | 04:33 |
| anteaya | let's find out | 04:33 |
| pleia2 | stop in hawaii on the way home though | 04:33 |
| notmyname | heh. I see my name too ;-) | 04:33 |
| anteaya | pleia2: awesome | 04:33 |
| *** gyee_ has quit IRC | 04:33 | |
| anteaya | notmyname: :D | 04:33 |
| pleia2 | anteaya: not so much, means I have to fly united hawaii to sf ;) | 04:33 |
| anteaya | notmyname: 8 hours in auckland for you too? | 04:33 |
| notmyname | I'm SFO->Auckland->MEL->Perth on the way there and Perth->Auckland->HON->SFO on the way back | 04:33 |
| anteaya | pleia2: :( | 04:34 |
| pleia2 | I don't stop in MEL on my way out | 04:34 |
| pleia2 | notmyname: flying out tomorrow? | 04:34 |
| notmyname | no, about 3 hours in AUK | 04:34 |
| notmyname | ya, I'm leaving tomorrow at 7:45 (Pacific). and I get to AUK on Sun morning. | 04:34 |
| notmyname | 21h15m in the air for tomorrow's trip | 04:35 |
| clarkb | so while I realize I am going the long way around (close to 13k miles one way iirc), I leave tomorrow evening and get there Sunday right about noon. Which isn't too bad | 04:35 |
| pleia2 | notmyname: ah, same flight then! | 04:35 |
| notmyname | pleia2: which leg? | 04:36 |
| pleia2 | notmyname: SFO to auckland | 04:36 |
| notmyname | k | 04:36 |
| pleia2 | and I'm doing Perth->Auckland->HON->SFO on saturday | 04:36 |
| notmyname | ah, cool | 04:36 |
| clarkb | I don't fly back until Tuesday, going to melt. should be fun | 04:37 |
| notmyname | I've got one thing to finish up for swiftstack tonight, then I'm planning on finally getting to check out the lca wiki and fill out the arrival page and etc | 04:37 |
| pleia2 | clarkb: hehe, yeah, it looks hot | 04:37 |
| anteaya | notmyname: there is an arrival page? | 04:37 |
| anteaya | it will be | 04:37 |
| notmyname | I saw an email about it today | 04:37 |
| pleia2 | anteaya: https://lca2014.linux.org.au/wiki/Arrivals_Departures | 04:37 |
| anteaya | Monday Jan. 6 will be 38C | 04:37 |
| anteaya | that is hot | 04:38 |
| clarkb | you should also get a 'boarding pass' which if you can print it will make life easier or something | 04:38 |
| anteaya | most other days will be 30C | 04:38 |
| anteaya | hot but manageable | 04:38 |
| pleia2 | I've started couch-to-5k, just in time for this, I might actually die :) | 04:38 |
| anteaya | clarkb: yeah, got that email sitting in the boarding lounge | 04:38 |
| *** CaptTofu has quit IRC | 04:38 | |
| pleia2 | (I did survive florida though!) | 04:38 |
| anteaya | pleia2: 5k swimming? | 04:38 |
| notmyname | https://lca2014.linux.org.au/wiki/Arrivals_Departures | 04:38 |
| pleia2 | anteaya: running | 04:39 |
| anteaya | running walking? | 04:39 |
| anteaya | eww | 04:39 |
| *** CaptTofu has joined #openstack-infra | 04:39 | |
| pleia2 | haha | 04:39 |
| anteaya | I can swim till I drop, can run at all | 04:39 |
| anteaya | can't | 04:39 |
| pleia2 | I couldn't either 3 weeks ago | 04:39 |
| anteaya | walk, I'll walk anywhere, not run | 04:39 |
| pleia2 | I ran for 3 minutes straight yesterday, twice! | 04:39 |
| anteaya | I am impressed | 04:39 |
| anteaya | good for you | 04:39 |
| anteaya | well done | 04:39 |
| pleia2 | thanks | 04:39 |
| pleia2 | san francisco is a great place for it | 04:40 |
| *** rakhmerov has joined #openstack-infra | 04:41 | |
| anteaya | awesome | 04:41 |
| *** ryanpetrello has joined #openstack-infra | 04:41 | |
| anteaya | I have no idea what my username and password is for the wiki | 04:41 |
| pleia2 | it's the same as the lca2014 website | 04:42 |
| anteaya | doesn't help me | 04:42 |
| anteaya | I have no idea what I said | 04:42 |
| anteaya | hit the door till I'm in, look I'm in, do the thing I came for | 04:42 |
| anteaya | how did you get in before? *shrug* no idea | 04:43 |
| anteaya | feel free to laugh and point | 04:43 |
| *** rakhmerov has quit IRC | 04:45 | |
| *** dcramer_ has joined #openstack-infra | 04:47 | |
| anteaya | no, I have logged into the website again and it is not accepting my name and password | 04:54 |
| anteaya | it says it should be the same, but it isn't working for me | 04:56 |
| *** melwitt1 has joined #openstack-infra | 04:58 | |
| *** zhiwei has quit IRC | 04:59 | |
| *** melwitt2 has joined #openstack-infra | 05:00 | |
| *** blamar has quit IRC | 05:00 | |
| *** melwitt1 has quit IRC | 05:03 | |
| *** Ryan_Lane has quit IRC | 05:03 | |
| *** senk has quit IRC | 05:05 | |
| *** talluri has joined #openstack-infra | 05:06 | |
| *** ^d has joined #openstack-infra | 05:07 | |
| *** ^d has joined #openstack-infra | 05:07 | |
| *** senk has joined #openstack-infra | 05:09 | |
| *** HenryG has quit IRC | 05:11 | |
| *** Ryan_Lane has joined #openstack-infra | 05:15 | |
| anteaya | it asks for username on the wiki, I have tried my first name, my first and last name, my email and my irc nick | 05:15 |
| *** zhiwei has joined #openstack-infra | 05:15 | |
| anteaya | nothing works | 05:15 |
| anteaya | in my profile for the conference I don't see a username | 05:16 |
| mikal | anteaya: the lca wiki? | 05:16 |
| anteaya | yes | 05:16 |
| mikal | Did you sign up for the conference with persona perchance? | 05:16 |
| anteaya | when I sign into the lca website it wants my email | 05:16 |
| anteaya | I have no idea what persona is, so chances are slim I did that | 05:17 |
| mikal | Ok, the wiki should be email address / password for the conf site | 05:17 |
| anteaya | I'll try that again | 05:17 |
| *** talluri has quit IRC | 05:17 | |
| notmyname | mikal: I shall get on a plane, sit down for 21 hours, and come visit you | 05:17 |
| *** talluri has joined #openstack-infra | 05:17 | |
| mikal | notmyname: good man | 05:17 |
| anteaya | mikal: okay since that narrowed down what username I was looking for I just fed it passwords until I hit the right one | 05:18 |
| anteaya | thanks | 05:18 |
| *** ^d has quit IRC | 05:19 | |
| notmyname | may I recommend a password manager? any of them will do | 05:19 |
| mikal | anteaya: lol | 05:19 |
| *** uvirtbot has quit IRC | 05:20 | |
| StevenK | anteaya: I second notmyname. Password managers are excellent. | 05:21 |
| *** Hefeweizen has quit IRC | 05:23 | |
| *** Hefeweizen has joined #openstack-infra | 05:23 | |
| anteaya | notmyname StevenK I would have to remember how to use them, would I not? | 05:24 |
| anteaya | I am designed a horizontal file system | 05:24 |
| anteaya | the size of a 4 by 8 sheet of plywood, like a big screen for a window | 05:24 |
| anteaya | able to stack them up and raise them or lower them like scenery on flies | 05:25 |
| anteaya | I would be so organized | 05:25 |
| anteaya | I am designing | 05:25 |
| StevenK | anteaya: Lastpass is awesome for that, when you add a site with a username and password, it autofills out the information | 05:25 |
| anteaya | I would remember where I put all my stuff | 05:25 |
| anteaya | StevenK: you are the man handy for the suggestions, I remember you from a prior conversation | 05:26 |
| StevenK | anteaya: :-) | 05:26 |
| anteaya | okay I'll look at lastpass while I loiter in hk airport for 8 hours | 05:26 |
| anteaya | :D | 05:27 |
| StevenK | anteaya: Since you'll be at LCA, I'd be delighted to help you | 05:27 |
| anteaya | I'll take you up on that | 05:27 |
| notmyname | I use 1Password and like it (honestly any password manager > no password manager) | 05:27 |
| anteaya | I might get a story or two out of the deal, that improves my ability to remember anything | 05:27 |
| anteaya | facts and a story work far better for me than anything else | 05:27 |
| anteaya | notmyname: yes I'm probably too oblivous to even be as ashamed as I should be | 05:28 |
| notmyname | once upon a time alice, forgot her password to the LCA wiki.... | 05:28 |
| notmyname | s/,// | 05:28 |
| anteaya | now we are talking | 05:28 |
| StevenK | notmyname: Hahaha | 05:29 |
| anteaya | what did alice do? | 05:29 |
| anteaya | and what airport was she stuck in when she did it? | 05:29 |
| anteaya | and how was the soup? | 05:29 |
| notmyname | she bought StevenK a beer and learned how to use a password manager ;-) | 05:29 |
| StevenK | It doesn't matter! All airports are the SAME! | 05:29 |
| anteaya | awesome | 05:29 |
| notmyname | StevenK: yes they are | 05:29 |
| anteaya | StevenK: not so | 05:29 |
| anteaya | no no, some have good restaurants | 05:29 |
| anteaya | they just appear the same | 05:30 |
| StevenK | That's true | 05:30 |
| anteaya | and the detroit airport has a Wendy's | 05:30 |
| notmyname | all have giant toblarone bars for sale | 05:30 |
| StevenK | The quality of resturants at international airports is hellishly variable | 05:30 |
| anteaya | gotta get a frosty in detroit | 05:30 |
| anteaya | StevenK: yes | 05:30 |
| anteaya | StevenK: what beer do you drink? | 05:30 |
| StevenK | Happily, I only have to fly domestic to Perth | 05:30 |
| anteaya | great | 05:30 |
| anteaya | you can laugh and point at me if I am jet lagged | 05:31 |
| anteaya | hk wasn't bad but coming back took about a week | 05:31 |
| StevenK | It takes me about two days to get back to feeling human after travelling to the US or Europe | 05:31 |
| StevenK | On the way back home it varies, depending on when I land in Sydney | 05:32 |
| *** david_lyle has quit IRC | 05:33 | |
| anteaya | yeah, I am finding I am doing okay back and forth to Europe | 05:34 |
| anteaya | coming back from summit, well let's just say there were a few early mornings | 05:35 |
| anteaya | I probably should pack up and change locations | 05:35 |
| StevenK | Haha | 05:35 |
| anteaya | something closer to my gate | 05:35 |
| anteaya | :D | 05:35 |
| anteaya | looking forward to meeting you StevenK | 05:35 |
| StevenK | Pack? That's so Sunday mornings problem | 05:35 |
| anteaya | and seeing you again mikal | 05:35 |
| anteaya | pack up as in vacate this restaurant | 05:36 |
| anteaya | and move to another one | 05:36 |
| StevenK | Haha | 05:36 |
| *** dstanek has quit IRC | 05:37 | |
| *** chandankumar has joined #openstack-infra | 05:37 | |
| *** AaronGr is now known as AaronGr_Zzz | 05:38 | |
| *** gilliard has quit IRC | 05:38 | |
| *** dstanek has joined #openstack-infra | 05:39 | |
| *** Ryan_Lane has quit IRC | 05:40 | |
| *** DennyZhang has joined #openstack-infra | 05:41 | |
| *** rakhmerov has joined #openstack-infra | 05:42 | |
| *** chandankumar has quit IRC | 05:44 | |
| *** chandankumar has joined #openstack-infra | 05:45 | |
| *** rakhmerov has quit IRC | 05:46 | |
| *** zhiwei has quit IRC | 05:49 | |
| *** chandankumar has quit IRC | 05:50 | |
| *** chandankumar has joined #openstack-infra | 05:51 | |
| *** mrda has quit IRC | 05:59 | |
| *** StevenK has quit IRC | 06:01 | |
| *** StevenK has joined #openstack-infra | 06:02 | |
| *** praneshp has joined #openstack-infra | 06:05 | |
| *** DennyZhang has quit IRC | 06:08 | |
| *** reed has quit IRC | 06:10 | |
| *** jpeeler has quit IRC | 06:11 | |
| *** ryanpetrello has quit IRC | 06:12 | |
| *** jishaomei has joined #openstack-infra | 06:17 | |
| *** dguitarbite has joined #openstack-infra | 06:27 | |
| *** chandankumar is now known as ciypro | 06:36 | |
| *** senk has quit IRC | 06:39 | |
| *** rakhmerov has joined #openstack-infra | 06:43 | |
| *** rakhmerov has quit IRC | 06:47 | |
| *** nicedice__ has joined #openstack-infra | 06:54 | |
| openstackgerrit | Zhiwei Chen proposed a change to openstack-infra/config: Rename a StackForge project https://review.openstack.org/64788 | 06:55 |
| *** dstanek has quit IRC | 06:55 | |
| *** dstanek has joined #openstack-infra | 06:56 | |
| *** nicedice_ has quit IRC | 06:58 | |
| mikal | Anyone with gerrit admin perms around? | 07:00 |
| *** dguitarbite has quit IRC | 07:06 | |
| *** dguitarbite has joined #openstack-infra | 07:09 | |
| *** rongze has joined #openstack-infra | 07:12 | |
| *** praneshp has quit IRC | 07:19 | |
| *** yaguang has joined #openstack-infra | 07:31 | |
| *** banix has quit IRC | 07:39 | |
| *** rakhmerov has joined #openstack-infra | 07:44 | |
| *** rcarrillocruz has quit IRC | 07:48 | |
| *** rakhmerov has quit IRC | 07:49 | |
| *** rcarrillocruz has joined #openstack-infra | 07:49 | |
| *** flaper87|afk is now known as flaper87 | 08:00 | |
| openstackgerrit | Andreas Jaeger proposed a change to openstack-infra/config: Create gate for api-site https://review.openstack.org/64795 | 08:00 |
| *** melwitt2 has quit IRC | 08:02 | |
| openstackgerrit | Andreas Jaeger proposed a change to openstack-infra/config: Create gate for api-site https://review.openstack.org/64795 | 08:06 |
| *** yolanda has joined #openstack-infra | 08:11 | |
| openstackgerrit | Andreas Jaeger proposed a change to openstack-infra/config: Create gate for api-site https://review.openstack.org/64795 | 08:13 |
| *** fifieldt has quit IRC | 08:18 | |
| openstackgerrit | Andreas Jaeger proposed a change to openstack-infra/config: Create gate for api-site https://review.openstack.org/64795 | 08:23 |
| *** vipul is now known as vipul-away | 08:27 | |
| *** vipul-away is now known as vipul | 08:29 | |
| *** vipul is now known as vipul-away | 08:44 | |
| *** cyeoh has quit IRC | 08:45 | |
| *** rakhmerov has joined #openstack-infra | 08:45 | |
| *** rakhmerov has quit IRC | 08:49 | |
| *** fbo_away is now known as fbo | 08:51 | |
| *** jpich has joined #openstack-infra | 08:59 | |
| *** rossella_s has joined #openstack-infra | 09:00 | |
| *** vipul-away is now known as vipul | 09:00 | |
| *** jishaomei has quit IRC | 09:03 | |
| *** hashar has joined #openstack-infra | 09:06 | |
| *** fifieldt has joined #openstack-infra | 09:06 | |
| *** rongze has quit IRC | 09:10 | |
| *** amotoki has joined #openstack-infra | 09:11 | |
| *** dguitarbite has quit IRC | 09:12 | |
| *** fifieldt has quit IRC | 09:14 | |
| *** derekh has joined #openstack-infra | 09:18 | |
| *** boris-42 has joined #openstack-infra | 09:21 | |
| *** hashar has quit IRC | 09:26 | |
| *** rossella_s has quit IRC | 09:36 | |
| *** rossella_s has joined #openstack-infra | 09:38 | |
| *** amotoki has quit IRC | 09:42 | |
| *** sdake has quit IRC | 09:43 | |
| *** jooools has joined #openstack-infra | 09:45 | |
| *** sdake has joined #openstack-infra | 09:46 | |
| *** rakhmerov has joined #openstack-infra | 09:46 | |
| *** rakhmerov has quit IRC | 09:50 | |
| *** johnthetubaguy has joined #openstack-infra | 09:53 | |
| *** rongze has joined #openstack-infra | 10:04 | |
| *** lcheng has quit IRC | 10:07 | |
| *** dguitarbite has joined #openstack-infra | 10:14 | |
| *** dguitarbite has quit IRC | 10:26 | |
| *** dguitarbite has joined #openstack-infra | 10:35 | |
| *** johnthetubaguy1 has joined #openstack-infra | 10:40 | |
| *** dguitarbite has quit IRC | 10:42 | |
| *** johnthetubaguy has quit IRC | 10:42 | |
| *** lcestari has joined #openstack-infra | 10:46 | |
| *** rossella_s_ has joined #openstack-infra | 10:47 | |
| *** lcheng has joined #openstack-infra | 10:47 | |
| *** rakhmerov has joined #openstack-infra | 10:47 | |
| *** rossella_s has quit IRC | 10:49 | |
| *** rossella_s_ is now known as rossella_s | 10:49 | |
| *** hashar has joined #openstack-infra | 10:50 | |
| *** rakhmerov has quit IRC | 10:52 | |
| *** yaguang has quit IRC | 10:56 | |
| *** lcheng has quit IRC | 11:13 | |
| *** yolanda has quit IRC | 11:25 | |
| *** yolanda has joined #openstack-infra | 11:26 | |
| *** hashar has quit IRC | 11:29 | |
| *** dguitarbite_ has joined #openstack-infra | 11:30 | |
| *** yolanda has quit IRC | 11:31 | |
| *** yolanda has joined #openstack-infra | 11:31 | |
| *** dguitarbite_ has quit IRC | 11:36 | |
| *** yolanda has quit IRC | 11:39 | |
| *** yolanda has joined #openstack-infra | 11:41 | |
| *** yolanda has quit IRC | 11:44 | |
| *** yolanda has joined #openstack-infra | 11:45 | |
| *** rakhmerov has joined #openstack-infra | 11:48 | |
| *** cyeoh_ has joined #openstack-infra | 11:48 | |
| *** cyeoh_ is now known as cyeoh | 11:48 | |
| *** dguitarbite has joined #openstack-infra | 11:50 | |
| *** yolanda has quit IRC | 11:51 | |
| *** yolanda has joined #openstack-infra | 11:52 | |
| *** yolanda has quit IRC | 11:52 | |
| *** rakhmerov has quit IRC | 11:53 | |
| *** yolanda has joined #openstack-infra | 11:53 | |
| *** yolanda has quit IRC | 11:56 | |
| *** yolanda has joined #openstack-infra | 11:57 | |
| *** coolsvap has quit IRC | 12:01 | |
| *** yolanda has quit IRC | 12:07 | |
| *** smarcet has joined #openstack-infra | 12:09 | |
| *** yolanda has joined #openstack-infra | 12:09 | |
| *** dguitarbite has quit IRC | 12:12 | |
| *** CaptTofu has quit IRC | 12:17 | |
| *** CaptTofu has joined #openstack-infra | 12:17 | |
| openstackgerrit | Soren Hansen proposed a change to openstack-infra/jenkins-job-builder: Allow passing credentials-id for git repos https://review.openstack.org/64817 | 12:18 |
| *** dguitarbite_ has joined #openstack-infra | 12:23 | |
| *** yolanda has quit IRC | 12:23 | |
| *** ArxCruz has joined #openstack-infra | 12:24 | |
| *** yolanda has joined #openstack-infra | 12:24 | |
| *** Ajaeger has joined #openstack-infra | 12:29 | |
| *** dguitarbite_ has quit IRC | 12:37 | |
| *** weshay has joined #openstack-infra | 12:38 | |
| *** jlibosva has joined #openstack-infra | 12:43 | |
| *** yolanda has quit IRC | 12:44 | |
| *** yolanda has joined #openstack-infra | 12:44 | |
| *** yolanda has quit IRC | 12:44 | |
| *** yolanda has joined #openstack-infra | 12:46 | |
| *** CaptTofu has quit IRC | 12:46 | |
| *** CaptTofu has joined #openstack-infra | 12:46 | |
| *** ihrachys has joined #openstack-infra | 12:47 | |
| *** CaptTofu has quit IRC | 12:47 | |
| *** CaptTofu_ has joined #openstack-infra | 12:47 | |
| *** rakhmerov has joined #openstack-infra | 12:48 | |
| ihrachys | hi all. It seems that my patch for oslo.messaging, Change-Id: I76bfa5b48bad4a70fbf06b74f4cc8234af6610c2 was not triggered for Jenkins verification, as per Zuul. Can you help me with the issue? Probably it's worth mentioning that I've created the review as a draft first, and then published it once I've made sure it's ok. | 12:49 |
| openstackgerrit | A change was merged to openstack-infra/os-loganalyze: Provide a script that starts a simple WSGI server https://review.openstack.org/60325 | 12:50 |
| *** yolanda has quit IRC | 12:51 | |
| *** yolanda has joined #openstack-infra | 12:51 | |
| *** yolanda has quit IRC | 12:52 | |
| *** rakhmerov has quit IRC | 12:53 | |
| *** CaptTofu_ has quit IRC | 12:53 | |
| *** CaptTofu has joined #openstack-infra | 12:54 | |
| *** dguitarbite has joined #openstack-infra | 12:54 | |
| *** CaptTofu has quit IRC | 12:59 | |
| *** dguitarbite has quit IRC | 13:00 | |
| ihrachys | ah, that just required adding a 'recheck' comment, and this resolved the problem. sorry for the noise above. :| | 13:05 |
| *** rongze has quit IRC | 13:10 | |
| openstackgerrit | Emilien Macchi proposed a change to openstack-infra/config: Add Puppet 3.4 gate support for OpenStack modules https://review.openstack.org/64823 | 13:17 |
| *** julim has joined #openstack-infra | 13:19 | |
| *** julim has quit IRC | 13:20 | |
| *** dguitarbite_ has joined #openstack-infra | 13:23 | |
| *** hashar has joined #openstack-infra | 13:26 | |
| *** dguitarbite_ has quit IRC | 13:27 | |
| *** b3nt_pin has joined #openstack-infra | 13:33 | |
| *** b3nt_pin is now known as beagles | 13:34 | |
| *** jpeeler has joined #openstack-infra | 13:40 | |
| *** rongze has joined #openstack-infra | 13:41 | |
| *** dprince has joined #openstack-infra | 13:46 | |
| *** rakhmerov has joined #openstack-infra | 13:49 | |
| *** dguitarbite has joined #openstack-infra | 13:49 | |
| *** ekarlso has quit IRC | 13:50 | |
| *** rakhmerov has quit IRC | 13:50 | |
| *** ciypro has quit IRC | 13:51 | |
| *** rongze has quit IRC | 13:52 | |
| *** ekarlso has joined #openstack-infra | 13:52 | |
| *** Ajaeger has quit IRC | 13:55 | |
| *** Ajaeger has joined #openstack-infra | 13:57 | |
| openstackgerrit | Sean Dague proposed a change to openstack/requirements: wip: sqla 0.8 state of the world https://review.openstack.org/64831 | 13:58 |
| *** tma996 has joined #openstack-infra | 13:58 | |
| *** HenryG has joined #openstack-infra | 14:00 | |
| *** jpeeler has quit IRC | 14:00 | |
| *** julim has joined #openstack-infra | 14:00 | |
| *** jpeeler has joined #openstack-infra | 14:01 | |
| *** dguitarbite has quit IRC | 14:01 | |
| *** dcramer_ has quit IRC | 14:03 | |
| fungi | ihrachys: did that patchset start out as a draft maybe? | 14:10 |
| fungi | if so, it's a known shortcoming of the gerrit version we're currently running (it doesn't have an event in its stream for a patchset transitioning from draft to published) | 14:11 |
| fungi | mikal: what admin perms are you needing, by the way? (if you're still around) | 14:14 |
| fungi | or, rather, what are you needing me to do with my perms? | 14:14 |
| ihrachys | fungi: yes, indeed, that was a draft [my first patch sent for review, so wanted to make sure it's ok before publishing it.] But since we can trigger tests via a recheck comment, that's fine. On a side note, another shortcoming with those drafts is that Launchpad bugs are not updated with links to corresponding reviews either. | 14:15 |
| *** mriedem has joined #openstack-infra | 14:16 | |
| * fungi ihrachys yes, we have "work in progress" as a way to mark changes which aren't ready for review. drafts are decidedly broken in several ways, and not really appropriate for our open design methodology anyway | 14:16 | |
| * fungi doesn't know why he /me'd the previous line. tabfail maybe | 14:17 | |
| *** ryanpetrello has joined #openstack-infra | 14:17 | |
| fungi | we're still looking at potential options to effectively disable drafts, but because gerrit doesn't make that configurable, the alternatives are somewhat ugly | 14:18 |
| ihrachys | fungi: ok, what's about that 'wip' thing? where can I check the proper guide on marking reviews as WiP? | 14:18 |
| fungi | there's a button in the current webui for "work in progress" which you can click on your change so that others will know it's still a work in progress | 14:19 |
| ihrachys | Ah, I see. Thanks! | 14:19 |
| fungi | the implementation is changing as we switch to gerrit 2.8 or 2.9, but once it stabilizes we'll almost certainly add a command-line option to git-review for it as well | 14:19 |
| *** jswarren has joined #openstack-infra | 14:21 | |
| *** mriedem has quit IRC | 14:25 | |
| *** mriedem has joined #openstack-infra | 14:27 | |
| openstackgerrit | Radomir Dopieralski proposed a change to openstack-dev/hacking: Add an option for ignoring h302 for some names https://review.openstack.org/64832 | 14:32 |
| openstackgerrit | A change was merged to openstack-infra/reviewday: Cleanup projects https://review.openstack.org/64474 | 14:33 |
| openstackgerrit | A change was merged to openstack-infra/reviewday: Add identity-api to projects https://review.openstack.org/64475 | 14:33 |
| mtreinish | jgriffith: actually it won't work until https://review.openstack.org/#/c/62101/ is merged | 14:40 |
| mtreinish | jgriffith: otherwise the addcleanup will be executed too late | 14:41 |
| mtreinish | but I'm not sure on the approach in that patch | 14:41 |
| *** eharney has joined #openstack-infra | 14:41 | |
| mtreinish | in the meantime it might be better to figure out a way to add the volume detach to the lifo in teardown class | 14:42 |
| *** coolsvap has joined #openstack-infra | 14:44 | |
| sdague | mtreinish: you rebasing that? | 14:45 |
| *** rongze has joined #openstack-infra | 14:45 | |
| mtreinish | sdague: I will, it's just it only works if people write scenarios properly | 14:45 |
| *** mfer has joined #openstack-infra | 14:45 | |
| mtreinish | because I can't call addCleanup from setUpClass | 14:46 |
| *** mfer has quit IRC | 14:46 | |
| sdague | can we fix that with hacking? | 14:46 |
| mtreinish | I don't think so, we have to ensure there aren't any shared resources between test cases in a class | 14:47 |
| *** dcramer_ has joined #openstack-infra | 14:49 | |
| *** rongze has quit IRC | 14:50 | |
| *** mfer has joined #openstack-infra | 14:51 | |
| *** portante is now known as portante|afk | 14:51 | |
| *** rakhmerov has joined #openstack-infra | 14:51 | |
| *** coolsvap has quit IRC | 14:52 | |
| *** yolanda has joined #openstack-infra | 14:53 | |
| *** jasondotstar has joined #openstack-infra | 14:55 | |
| *** rakhmerov has quit IRC | 14:56 | |
| *** rongze has joined #openstack-infra | 14:56 | |
| *** chmouel has quit IRC | 14:59 | |
| sdague | mtreinish: so that means really waiting to using fixtures? | 15:03 |
| mtreinish | sdague: yeah, that would probably be the ideal solution. | 15:05 |
| openstackgerrit | Radomir Dopieralski proposed a change to openstack-dev/hacking: Add an option for ignoring h302 for some names https://review.openstack.org/64832 | 15:05 |
| sdague | mtreinish: so is there a reason to do your patch before than then? if it's only going to work if other people do exactly the right things? | 15:09 |
| openstackgerrit | Cyril Roelandt proposed a change to openstack-infra/config: sphinxcontrib-docbookrestapi: enable tests for Python 2.6, 2.7 and 3.3 https://review.openstack.org/64841 | 15:09 |
| mtreinish | sdague: the only reason is so that jgriffith's patch (https://review.openstack.org/#/c/62014/ ) will work as expected | 15:11 |
| mtreinish | the fixture model will take some time to get working properly | 15:11 |
| *** ruhe has joined #openstack-infra | 15:12 | |
| openstackgerrit | Cyril Roelandt proposed a change to openstack-infra/config: sphinxcontrib-docbookrestapi: enable tests for Python 2.6, 2.7 and 3.3 https://review.openstack.org/64841 | 15:18 |
| *** coolsvap has joined #openstack-infra | 15:18 | |
| *** yolanda has quit IRC | 15:18 | |
| *** yolanda has joined #openstack-infra | 15:19 | |
| *** coolsvap_ has joined #openstack-infra | 15:22 | |
| *** thedodd has joined #openstack-infra | 15:22 | |
| *** coolsvap_ has quit IRC | 15:22 | |
| *** coolsvap_ has joined #openstack-infra | 15:22 | |
| *** russellb is now known as rustlebee | 15:23 | |
| *** coolsvap has quit IRC | 15:23 | |
| *** coolsvap_ is now known as coolsvap | 15:23 | |
| *** dcramer_ has quit IRC | 15:25 | |
| *** dansmith is now known as damnsmith | 15:26 | |
| *** dcramer_ has joined #openstack-infra | 15:27 | |
| *** rossella_s has quit IRC | 15:28 | |
| *** AaronGr_Zzz is now known as AaronGr | 15:29 | |
| *** blamar has joined #openstack-infra | 15:29 | |
| *** dims has quit IRC | 15:30 | |
| *** herndon has joined #openstack-infra | 15:31 | |
| *** yolanda has quit IRC | 15:31 | |
| *** talluri has quit IRC | 15:31 | |
| *** talluri has joined #openstack-infra | 15:32 | |
| *** yolanda has joined #openstack-infra | 15:32 | |
| *** dims has joined #openstack-infra | 15:33 | |
| *** yolanda has quit IRC | 15:33 | |
| *** yolanda has joined #openstack-infra | 15:34 | |
| *** dims is now known as dimsum | 15:34 | |
| *** yolanda has quit IRC | 15:34 | |
| *** senk has joined #openstack-infra | 15:35 | |
| *** yolanda has joined #openstack-infra | 15:36 | |
| *** talluri has quit IRC | 15:36 | |
| sdague | any chance we can get another +2 on this to get oslo.rootwrap into the gate - https://review.openstack.org/#/c/64675/1 ? | 15:40 |
| openstackgerrit | Radomir Dopieralski proposed a change to openstack-dev/hacking: Add an option for ignoring h302 for some names https://review.openstack.org/64832 | 15:43 |
| *** yolanda has quit IRC | 15:43 | |
| *** yolanda has joined #openstack-infra | 15:44 | |
| *** senk has quit IRC | 15:46 | |
| *** rnirmal has joined #openstack-infra | 15:47 | |
| *** senk has joined #openstack-infra | 15:50 | |
| *** rakhmerov has joined #openstack-infra | 15:52 | |
| *** yolanda has quit IRC | 15:55 | |
| *** rakhmerov has quit IRC | 15:56 | |
| *** ^d has joined #openstack-infra | 15:56 | |
| *** ^d has joined #openstack-infra | 15:56 | |
| *** changbl has quit IRC | 15:57 | |
| *** changbl has joined #openstack-infra | 15:57 | |
| openstackgerrit | Monty Taylor proposed a change to openstack-infra/config: Add storyboard-webclient to gerrit https://review.openstack.org/64761 | 15:59 |
| mordred | sdague: looking | 15:59 |
| *** chmouel has joined #openstack-infra | 16:00 | |
| mordred | sdague: done | 16:00 |
| *** senk has quit IRC | 16:02 | |
| mordred | zaro: ooh! I like the new commit message plugin: https://review-dev.openstack.org/plugins/commit-message-length-validator/Documentation/config.html | 16:02 |
| *** markmcclain has joined #openstack-infra | 16:02 | |
| mordred | zaro: I believe we should configure commitmessage.maxSubjectLength to 50, since that's the actual recommended length that we suggest | 16:03 |
| mordred | but I do think we should leave it just at warn | 16:03 |
| *** senk has joined #openstack-infra | 16:03 | |
| jgriffith | mtreinish: I vote you just take my original patch and we move on :) | 16:12 |
| *** yolanda has joined #openstack-infra | 16:13 | |
| jgriffith | mtreinish: I'm kidding of course | 16:13 |
| *** Ajaeger has quit IRC | 16:13 | |
| *** yolanda has quit IRC | 16:13 | |
| *** rakhmerov has joined #openstack-infra | 16:13 | |
| *** yolanda has joined #openstack-infra | 16:14 | |
| *** yolanda has quit IRC | 16:15 | |
| *** david-lyle_ has joined #openstack-infra | 16:15 | |
| *** yolanda has joined #openstack-infra | 16:16 | |
| *** yolanda has quit IRC | 16:18 | |
| *** pmathews has joined #openstack-infra | 16:19 | |
| openstackgerrit | Russell Bryant proposed a change to openstack-infra/reviewstats: Add the gantt repo to the nova (compute) stats https://review.openstack.org/64850 | 16:19 |
| *** yolanda has joined #openstack-infra | 16:19 | |
| *** DennyZhang has joined #openstack-infra | 16:20 | |
| *** yolanda has quit IRC | 16:22 | |
| *** yolanda has joined #openstack-infra | 16:23 | |
| openstackgerrit | A change was merged to openstack-infra/reviewstats: Add the gantt repo to the nova (compute) stats https://review.openstack.org/64850 | 16:24 |
| mtreinish | jgriffith: heh :) | 16:24 |
| openstackgerrit | Mate Lakat proposed a change to openstack-infra/config: Initial commit of a XenServer builder script https://review.openstack.org/60581 | 16:27 |
| mordred | dstufft: I'm always open to new pbr options - but I'm not sure I understand what that option would do for us? | 16:27 |
| dstufft | mordred: I think I'm just being grumpy | 16:28 |
| mordred | dstufft: k. grumpy is fair | 16:28 |
| dstufft | mordred: but why not declare dependencies in setup.cfg instead of a seperate file? | 16:28 |
| mordred | dstufft: hrm. I was about to argue - but in formulating the response I talked myself out of it | 16:29 |
| mordred | dstufft: I think the main thing is that then it's not clear how to make a virtualenv populated with the things the project needs not including the project itself | 16:29 |
| mordred | which is helpful when we're debugging installs | 16:30 |
| mordred | but | 16:30 |
| mordred | might not actually be super important | 16:30 |
| dstufft | mordred: btw the docs for pbr points out it requires distribute, should probably be setuptools now adays? | 16:31 |
| mordred | dstufft: + | 16:31 |
| mordred | dstufft: ++ | 16:31 |
| mordred | since what they _usually_ want to do is install the software into a venv | 16:31 |
| clarkb | until you realize that is how run_tests.sh works and breaks on neutron because pyudev | 16:31 |
| mordred | clarkb: what? | 16:31 |
| clarkb | reseparate files | 16:31 |
| mordred | clarkb: I still don't follow "how run_tests.sh works" | 16:31 |
| dstufft | mordred: fwiw https://github.com/openstack-dev/pbr/blob/master/doc/source/index.rst#usage is where the docs reco distribute, I can submit a PR thing but if you want to do it for fastering that's fine with me too! | 16:32 |
| mordred | dstufft: yeah - I can get it | 16:32 |
| clarkb | run tests just installs the reqs files and does not installthe project to the venv so pyudev does not install | 16:32 |
| mordred | ah. gotcha | 16:32 |
| mordred | so that's an argument _for_ moving to declaring directly in setup.cfg | 16:32 |
| *** yolanda has quit IRC | 16:33 | |
| dstufft | mordred: fwiw if I remember pbr automatically translates -e's in a requirements.txt into dependency links | 16:33 |
| dstufft | which won't install by default in pip 1.5 | 16:33 |
| dstufft | I could be remembering that wrong though | 16:33 |
| dstufft | (Can you tell i'm messing with pbr?) | 16:34 |
| openstackgerrit | Monty Taylor proposed a change to openstack-dev/pbr: Remove reference to distribute https://review.openstack.org/64853 | 16:34 |
| mordred | dstufft: yes. I can. and I welcome that! you know a lot of things about a lot of things | 16:34 |
| fungi | wasn't the -e handling just a short-term workaround for not being in a position yet to upload wheels of prereleases? | 16:34 |
| mordred | dstufft: fwiw, as soon as I finish my wheel uploading code, we're going to get rid of supporting links in pbr too | 16:34 |
| mordred | yes | 16:34 |
| mordred | we are not interested in people declaring that software should be gotten from not-pypi | 16:35 |
| *** ruhe has quit IRC | 16:35 | |
| dstufft | ok cool | 16:35 |
| *** yolanda has joined #openstack-infra | 16:36 | |
| *** DennyZhang has quit IRC | 16:37 | |
| *** yolanda has quit IRC | 16:38 | |
| mordred | dstufft: in general, I'm not married to the way we do any specific things we do in pbr, as long as replacements can handle all the use cases | 16:39 |
| *** yolanda has joined #openstack-infra | 16:40 | |
| *** dangers_` is now known as dangers | 16:40 | |
| mordred | and doing requirements directly in setup.cfg would certainly make adding the os selector parsing support easier | 16:42 |
| mordred | otoh- any idea on how to do what we're doing with test-requirements? | 16:42 |
| *** yolanda has quit IRC | 16:42 | |
| *** senk has quit IRC | 16:43 | |
| *** yolanda has joined #openstack-infra | 16:43 | |
| dstufft | mordred: isn't it just injecting that into tests_require? | 16:44 |
| *** talluri has joined #openstack-infra | 16:44 | |
| *** senk has joined #openstack-infra | 16:44 | |
| *** yolanda has quit IRC | 16:44 | |
| mordred | well, tests_require doesn't do much useful | 16:44 |
| mordred | so we process it directly | 16:44 |
| mordred | but the question is - how would I request of pip install . that it also include the test requirements | 16:45 |
| dstufft | yea tests_require more or less depends on ``python setup.py test`` | 16:45 |
| *** yolanda has joined #openstack-infra | 16:45 | |
| dstufft | the way I do it, which isn't perfect but it's close to the metadata 2.0 way of doing it, is just define a tests extra | 16:45 |
| dstufft | so then it'd be ``pip install .[tests]`` | 16:45 |
| mordred | dstufft: does the tests extra need to have the whole list of requirements _plus_ the tests requirements? or is it additive? | 16:46 |
| dstufft | mordred: it's additive | 16:46 |
| dstufft | it's a list of extra stuff to install | 16:47 |
| mordred | great. well then, that seems like a non-impossible thing to do | 16:47 |
| *** rossella_s has joined #openstack-infra | 16:47 | |
| dstufft | the way metadata 2.0 does it, is with an invalid extra name that maps to the tests_require | 16:47 |
| dstufft | :test: | 16:47 |
| dstufft | so in the ~future~ it'll be pip install whatever[:test:] | 16:47 |
| mordred | we'd need to re-write the global-requirements sync script too | 16:48 |
| dstufft | but since that doesn't exist yet I mimic it with ane xtral extar | 16:48 |
| dstufft | extra | 16:48 |
| dstufft | actual | 16:48 |
| *** talluri has quit IRC | 16:48 | |
| mordred | so - I could do it an alternate way in pbr ... | 16:50 |
| *** reed has joined #openstack-infra | 16:50 | |
| mordred | and put the test requirements into tests_require in setup.cfg ... but have pbr translate that into a tests extra when it does the parse | 16:50 |
| openstackgerrit | Mate Lakat proposed a change to openstack-infra/config: Initial commit of a XenServer builder script https://review.openstack.org/60581 | 16:51 |
| mordred | and then when we start emiting metadata 2.0 stuff, that could get put into :test: properly | 16:51 |
| mordred | or is that too much? | 16:51 |
| dstufft | mordred: makes sense I think | 16:52 |
| dstufft | in emtadata 2.0 it'll map tests_require to :test: | 16:52 |
| dstufft | so that would probably be fine | 16:52 |
| dstufft | mordred: also I think git has a fix for the --system-site-packages stuff in virtualenv | 16:53 |
| dstufft | there's another issue someone mentioned with pip, but i expect a X.Y.1 release sometime in the near future | 16:53 |
| *** yolanda has quit IRC | 16:53 | |
| *** UtahDave has joined #openstack-infra | 16:54 | |
| mfer | clarkb jeblair I was wondering if either of you could fill me in on the status of https://bugs.launchpad.net/openstack-ci/+bug/1242569. I see the patch set was merged but the bug is still open | 16:54 |
| mfer | I'm curious if I can pursue the new project now | 16:55 |
| *** yolanda has joined #openstack-infra | 16:57 | |
| *** lcheng has joined #openstack-infra | 16:58 | |
| clarkb | mfer: it needs proper testing. once that is done bug can be closed | 16:58 |
| *** yolanda has quit IRC | 16:58 | |
| mfer | if there is anything i can do to help please let me know | 16:59 |
| *** yolanda has joined #openstack-infra | 16:59 | |
| *** jooools has quit IRC | 17:00 | |
| *** UtahDave has quit IRC | 17:00 | |
| fungi | dstufft: yeah, the patch in git is trivial and works... we just need virtualenv to release a version which includes that now | 17:02 |
| openstackgerrit | A change was merged to openstack-infra/devstack-gate: Add oslo.rootwrap to the devstack gate https://review.openstack.org/64675 | 17:02 |
| *** gyee_ has joined #openstack-infra | 17:02 | |
| *** praneshp has joined #openstack-infra | 17:02 | |
| zaro | morning | 17:04 |
| *** yolanda has quit IRC | 17:05 | |
| dstufft | mordred: another thing, you may want to think about having pbr drop a file that encodes the version instead of using pkg_resources | 17:05 |
| dstufft | for the version stuff | 17:05 |
| *** herndon has quit IRC | 17:05 | |
| mordred | dstufft: we used to do that, actually | 17:05 |
| dstufft | pkg_resources tells you what version of a thing is installed, not what version you've imported | 17:05 |
| *** coolsvap has quit IRC | 17:05 | |
| mordred | hrm. what's the difference? | 17:06 |
| *** UtahDave has joined #openstack-infra | 17:06 | |
| *** coolsvap has joined #openstack-infra | 17:06 | |
| *** pcrews_ has quit IRC | 17:06 | |
| sdague | mordred: also, please take a look at dev list on the taskflow wedge, and make sure I'm not crazy pants | 17:06 |
| mordred | sdague: I agree with you | 17:06 |
| *** coolsvap has quit IRC | 17:06 | |
| dstufft | download a tarball and run ``python`` inside of it, you get "." on the sys.path and will be importing the code from the tarball, but since it's not installed pkg_resources won't see it, and if you already have a version of _whatever_ installed, you'll be told your importing that version instead of the actual version | 17:06 |
| dstufft | same thing with PYTHONPATH modifications and such | 17:07 |
| *** yolanda has joined #openstack-infra | 17:07 | |
| sdague | mordred: cool | 17:07 |
| dstufft | sort of a corner case, but I've seen folks get bit by using pkg_resources thinking they were debugging against 1 version when really it was another | 17:07 |
| *** coolsvap has joined #openstack-infra | 17:07 | |
| *** herndon has joined #openstack-infra | 17:08 | |
| mordred | nod | 17:09 |
| mordred | well, in general, we tel people that we don't support un-installed operation - and we're fairly heavy users of entrypoints, which forces the issue a bit | 17:09 |
| mordred | BUT | 17:09 |
| *** bnemec is now known as beekneemech | 17:09 | |
| mordred | it's a good point | 17:09 |
| jeblair | pleia2, notmyname: i'll see you tonight. :) | 17:11 |
| dstufft | mordred: yea, I rarely use entrypoints for anything so I don't really ru into that particular thing much. I just know the misinformation (probably major reason for a human to look at __version__ is debugging I think) can be a huge time waster and I figured i'd mention it! | 17:11 |
| mordred | ++ | 17:11 |
| zaro | fungi, clarkb, sdague : i turned on secondary indexing on review-dev.o.o. You can try the expanded searching capability if you like. Not sure if gerrit handles the periodic reindexing or it's something we need to setup a cron job for though. | 17:11 |
| sdague | zaro: thanks, I will do this afternoon | 17:12 |
| *** yolanda has quit IRC | 17:12 | |
| openstackgerrit | Cyril Roelandt proposed a change to openstack-infra/config: sphinxcontrib-docbookrestapi: enable tests for Python 2.6, 2.7 and 3.3 https://review.openstack.org/64841 | 17:12 |
| fungi | i think we need to rework the log uploader to ask permission rather than forgiveness. it's driving people to recheck and reverify other devstack/tempest/grenade failures against bug 1233957 | 17:13 |
| *** yolanda has joined #openstack-infra | 17:13 | |
| fungi | uvirtbot hath forsaken us... https://launchpad.net/bugs/1233957 | 17:13 |
| *** yolanda has quit IRC | 17:14 | |
| jeblair | fungi: any chance that correlates with the jenkins server that is upgraded and running the new scp plugin? | 17:15 |
| fungi | jeblair: i don't think so. this is just the benign error message where our upload script tries to grab a variety of different filename patterns and emits error messages about the ones which don't exist | 17:16 |
| jeblair | fungi: but the job failed; why? | 17:16 |
| *** rcleere has joined #openstack-infra | 17:16 | |
| *** banix has joined #openstack-infra | 17:16 | |
| fungi | and people are seeing those in other jobs (which failed for entirely other unrelated reasons), assume they're the reason for their job failure and reverify against it | 17:16 |
| jeblair | fungi: if what you are saying is happening; i agree we should deal with it, however, i'm concerned with the potentially more serious situation where that _actually is_ the reason their jobs fail... | 17:17 |
| jeblair | fungi: from what's pasted in the bug report, i can't eliminate that possibility | 17:18 |
| notmyname | jeblair: are you flying out of SFO tonight on the same flight? | 17:18 |
| jeblair | notmyname: yes | 17:19 |
| notmyname | jeblair: what time are you getting to the airport? I was going to shoot for abut 5:30 and grab something to eat | 17:19 |
| fungi | jeblair: well, i have proof people are reverifying grenade jobs against that bug incorrectly (that's how i ran across the bug in the first place), but i'll try to run down the actual bugs which they should have been rechecked against instead | 17:19 |
| jeblair | notmyname: yeah, i should be there by then, maybe a bit earlier, and was planning on doing the same thing; try to meet up? | 17:21 |
| *** Ajaeger has joined #openstack-infra | 17:21 | |
| jeblair | fungi: yeah, the link in comment #2 is a bad match, as you say | 17:22 |
| notmyname | jeblair: sure | 17:22 |
| fungi | jeblair: as is https://review.openstack.org/59686 (spotted failing in the gate just now) | 17:22 |
| jeblair | fungi: the text in the bug report, however, is concerning, assuming that it's not edited; however, the author did not link to the actual logs | 17:22 |
| *** gyee_ has quit IRC | 17:23 | |
| fungi | jeblair: the actual log for the first one is http://logs.openstack.org/17/49117/3/check/check-grenade-devstack-vm/a0080db/ (from october). i'm still digging through that one | 17:23 |
| fungi | could be it stems from a two-month-old systemic issue which i've now forgotten we dealt with, but so far i'm not spotting the actual reason it failed | 17:24 |
| *** flaper87 is now known as flaper87|afk | 17:24 | |
| Ajaeger | Hi Infra team, I'd love to see https://review.openstack.org/#/c/64567/ reviewed asap, please - to see that this really works as expected. | 17:24 |
| jeblair | fungi: ah, the original is a timeout. | 17:26 |
| *** rcarrillocruz has quit IRC | 17:26 | |
| fungi | i was just coming to that conclusion based on the timestamps, but didn't spot the actual timeout message | 17:26 |
| jeblair | fungi: so i think we sholud | 17:27 |
| jeblair | (a) close the bug report as invalid | 17:27 |
| jeblair | (b) have devstack-gate copy testr and usbunit log files into $WORKSPACE/logs; then change the devstack-logs jjb macro to stop doing that | 17:28 |
| jeblair | fungi: that should get rid of the msg from jenkins | 17:28 |
| jeblair | s/usbunit/subunit/ | 17:28 |
| fungi | seems a sane course of action. i'll take care of it | 17:28 |
| jeblair | fungi: awesome, thx | 17:28 |
| *** derekh has quit IRC | 17:29 | |
| *** moted has quit IRC | 17:29 | |
| *** moted has joined #openstack-infra | 17:30 | |
| *** praneshp has quit IRC | 17:30 | |
| *** gyee_ has joined #openstack-infra | 17:30 | |
| *** yolanda has joined #openstack-infra | 17:32 | |
| *** guest23423423 has joined #openstack-infra | 17:33 | |
| *** praneshp has joined #openstack-infra | 17:33 | |
| *** yolanda has quit IRC | 17:33 | |
| *** jpich has quit IRC | 17:34 | |
| *** krotscheck has joined #openstack-infra | 17:34 | |
| *** CaptTofu has joined #openstack-infra | 17:34 | |
| *** senk has quit IRC | 17:35 | |
| *** CaptTofu_ has joined #openstack-infra | 17:35 | |
| *** guest23423423 has left #openstack-infra | 17:37 | |
| *** harlowja_away is now known as harlowja | 17:38 | |
| *** CaptTofu has quit IRC | 17:39 | |
| *** senk has joined #openstack-infra | 17:39 | |
| *** fbo is now known as fbo_away | 17:42 | |
| openstackgerrit | Andreas Jaeger proposed a change to openstack-infra/config: Create gate for api-site https://review.openstack.org/64795 | 17:44 |
| *** ruhe has joined #openstack-infra | 17:45 | |
| fungi | yeah, so the failure i spotted that bug being used to reverify against today seems to indicate that what's actually going on at the moment is havana is untestable because grenade's failing to run the aggregates exercises on grizzly. new bitrot | 17:46 |
| harlowja | sdague yt | 17:48 |
| *** pcrews has joined #openstack-infra | 17:53 | |
| *** ruhe is now known as _ruhe | 17:57 | |
| *** tma996 has quit IRC | 17:58 | |
| zaro | fungi: i see review-dev is setup to replicate to github. where can i find the review-dev project on github? | 17:59 |
| *** markmcclain has quit IRC | 17:59 | |
| jeblair | zaro: it has non-overlapping orgs, such as 'gtest-org' | 17:59 |
| jeblair | zaro: https://github.com/gtest-org | 18:00 |
| fungi | mikal: we're getting turbo-hipster bugs reported against openstack-ci. where do you want those forwarded? | 18:00 |
| clarkb | harlowja: fwiw oslo project don't pose the same problem as taskflow because they are installed from source and not pypi when doing integration testing. Because source installs are used we can (and do) update the requirements lists to match the global requirements before install | 18:01 |
| harlowja | clarkb ok, that doesn't seem so bad, install from source vs not | 18:02 |
| fungi | mikal: nevermind... i just noticed turbo-hipster has an lp project, so i moved the bug in question to it | 18:02 |
| harlowja | clarkb why is it limited to installing oslo projects from source? stackforge ones can also? | 18:05 |
| *** bknudson has quit IRC | 18:05 | |
| clarkb | harlowja: because we don't control stackforge projects | 18:05 |
| clarkb | we can't enforce requirements on them | 18:05 |
| clarkb | or assume that doing so would work and so on | 18:06 |
| clarkb | (it really boils down to which projects are gated) | 18:06 |
| harlowja | hmmm, but u could let some stackforge projects say 'i'm ok with being enforced' | 18:06 |
| dhellmann | clarkb, harlowja : sdague added an option to devstack to install some stackforge projects from source (we use it for pecan and wsme right now) | 18:06 |
| *** bknudson has joined #openstack-infra | 18:06 | |
| zaro | jeblair: thanks. just verified that replication continues to work. | 18:06 |
| clarkb | harlowja: the inverse is not true though | 18:06 |
| clarkb | openstack doesn't want to be gated on stackforge | 18:06 |
| *** yolanda has joined #openstack-infra | 18:06 | |
| dhellmann | http://git.openstack.org/cgit/openstack-dev/devstack/tree/lib/stackforge | 18:07 |
| harlowja | clarkb hmmm, blurry lines between stackforge and oslo and gating and what is really a dependent | 18:07 |
| dhellmann | to use it, have "enable_service stackforge_libs" in your localrc | 18:07 |
| *** yolanda has quit IRC | 18:08 | |
| dhellmann | we have a few libraries in stackforge now that are maintained by mostly OS devs and are strong dependencies, but that aren't part of an OS program | 18:08 |
| clarkb | dhellmann: interesting. I don't think that solves the problem of openstack being gated on stackforge which would be necessary to solve this proble while keeping taskflow in stackforge, but it does solve part of the technical problem | 18:08 |
| fungi | right, we want to check that changes to openstack work with other non-openstack library dependencies as-released, since that's how they'll be run in the wild. if we test openstack changes against unreleased dependency sources, we risk releasing an openstack nobody can run with released deps | 18:08 |
| dhellmann | clarkb: yeah, I think sdague was working on some symmetric gating stuff in december, but I'm not sure where it stands | 18:09 |
| *** yolanda has joined #openstack-infra | 18:09 | |
| dhellmann | fungi: true, I think the devstack changes were mostly so we could gate pecan and wsme changes against openstack | 18:09 |
| *** yolanda has quit IRC | 18:10 | |
| *** jlibosva has quit IRC | 18:10 | |
| fungi | we actually also have this problem now with openstack components which are part of gating but do not follow the integrated release cadence, and so still need to add some more jobs to confirm that we're also testing against released libs we do control | 18:10 |
| clarkb | dhellmann: ya gating stackforge on openstack is fine | 18:10 |
| jeblair | yeah, gating stackforge against openstack for libraries that openstack uses _as libraries_ is fine (and pretty cool) | 18:10 |
| *** yolanda has joined #openstack-infra | 18:10 | |
| fungi | for instance, it would not be good if we released icehouse with a nova which only worked with an unreleased version of oslo.config... right now we don't have jobs which confirm we won't make that mistake | 18:11 |
| pleia2 | it's behind a pay wall (unless you're a usenix member) but my code review for sysadmins talk turned into an article here: https://www.usenix.org/publications/login-logout/january-2014-login-logout | 18:11 |
| jeblair | openstack dependencies either need to be released libraries, or other openstack projects | 18:11 |
| *** yolanda has quit IRC | 18:11 | |
| jeblair | so are we discussing the possibility of taskflow being an unreleased dependency of openstack? | 18:11 |
| clarkb | jeblair: there is a thraed on the ML, sdague is suggesting it get slurped into oslo | 18:12 |
| *** yolanda has joined #openstack-infra | 18:12 | |
| harlowja | jeblair i'm not sure yet, still trying to figure it out :-P | 18:12 |
| *** yolanda has quit IRC | 18:12 | |
| fungi | jeblair: this came up because taskflow pins sqlalchemy to the same thing as specified in requirements, thus because it's also a dependency we can't test un-pinning sqlalchemy easily | 18:12 |
| harlowja | i don't mind getting slurped into oslo, although i guess that just changes really not that much | 18:12 |
| dhellmann | it doesn't seem to change much to me, either, but I haven't read sdague's argument yet | 18:13 |
| dhellmann | I'll chime in on that after lunch | 18:13 |
| harlowja | thx dhellmann | 18:13 |
| *** _ruhe is now known as ruhe | 18:13 | |
| *** yolanda has joined #openstack-infra | 18:13 | |
| *** yolanda has quit IRC | 18:13 | |
| harlowja | ivan also just tried https://review.openstack.org/#/c/64869/ and seems ok | 18:13 |
| *** prad has joined #openstack-infra | 18:14 | |
| jeblair | caught up (i only skimmed that thread because sdague said "you should be in oslo" and harlowja said "okay"; so i moved on. :) | 18:14 |
| *** yolanda has joined #openstack-infra | 18:15 | |
| harlowja | :) | 18:15 |
| jeblair | harlowja: dhellmann can say for sure, but my understanding of the incubator is that it's only for code that's being refactored out of openstack... | 18:15 |
| harlowja | jeblair ya thats my understanding as well | 18:15 |
| dhellmann | yeah, it's not appropriate to put taskflow in the incubator | 18:15 |
| jeblair | harlowja: so since taskflow is already an extant library, it could go straight to being its own oslo lib | 18:15 |
| *** senk has quit IRC | 18:15 | |
| dhellmann | it's already a lib, what benefit do we get from attaching "oslo" to it? | 18:15 |
| *** yolanda has quit IRC | 18:16 | |
| jeblair | dhellmann: making it part of openstack | 18:16 |
| fungi | it could go straight into any existing program where it's appropriate to be associated | 18:16 |
| jeblair | fungi: ++ | 18:16 |
| dhellmann | does that solve a technical problem somehow? | 18:16 |
| clarkb | dhellmann: mutual gating | 18:16 |
| harlowja | btw, just so i can address a user of taskflow who is using it in a non-openstack cloud, at least one person is doing this (refactoring parts of his own cloud using taskflow) | 18:16 |
| harlowja | i just want to make sure that his usage doesn't become hard | 18:16 |
| krotscheck | Do we have the cobertura plugin on our jenkins? | 18:17 |
| dhellmann | clarkb: why can't we do that anyway? | 18:17 |
| *** yolanda has joined #openstack-infra | 18:17 | |
| *** yolanda has quit IRC | 18:17 | |
| clarkb | dhellmann: because we cannot gate openstack on stackforge | 18:17 |
| dhellmann | clarkb: is that a technical issue or a policy we've set? | 18:17 |
| clarkb | krotscheck: we might, but shouldn't use it | 18:17 |
| fungi | krotscheck: what are you wanting to do with it? | 18:17 |
| *** SEJeff_work has joined #openstack-infra | 18:18 | |
| clarkb | krotscheck: its one of those plugins that is slow. better to generates coverage reports in a job and stash them somewhere | 18:18 |
| clarkb | also you don't want to talk to jenkins directly | 18:18 |
| dhellmann | clarkb: I'm just trying to understand the situation, because if there's a technical issue then it may change some of my own plans for other libs like cliff and stevedore | 18:18 |
| fungi | krotscheck: right, we usually run coverage as a job with a testenv:cover section in tox.ini | 18:18 |
| jeblair | dhellmann: it's a policy we've set because when we start putting projects into the integrated gate, we're getting involved in project governance, and the only way we want to do that is using the existing project governance we have: openstack | 18:18 |
| krotscheck | clarkb: fungi: Well, I have coverage and unit test reports being generated, and I'd like to know how best to break my build if it fails. | 18:19 |
| clarkb | dhellmann: so it is clear. Gating stackforge on openstack is fine. openstack on stackforge is not | 18:19 |
| dhellmann | jeblair: ok, that makes sense | 18:19 |
| krotscheck | sorry- break my build if the coverage doesn't pass | 18:19 |
| dhellmann | clarkb: sure | 18:19 |
| clarkb | krotscheck: exit 1 :) | 18:19 |
| fungi | or sys.exit(1) | 18:19 |
| clarkb | krotscheck: also coverage 'passing' is really tricky | 18:20 |
| jeblair | krotscheck: we have all kinds of ideas on how to break your build! :) | 18:20 |
| krotscheck | clarkb: That works. So I should just dump console output if it fails? | 18:20 |
| *** thedodd has quit IRC | 18:20 | |
| fungi | krotscheck: i think you should dump console output regardless of whether it passes or fails, because it's useful for debugging why it incorrectly passed sometimes | 18:20 |
| krotscheck | fungi: Well, yes. I just want to know if there's a big fancy UI type thing that gives us pretty "Hey this needs coverage" graphs. | 18:21 |
| krotscheck | ....though I suspect the answer to that is "What does UI stand for?": | 18:21 |
| clarkb | krotscheck: most coverage tools will output them themselves | 18:21 |
| clarkb | krotscheck: the way we do it with the python projects is the coverage job runs, after completion it generates html files from the results, those files are copied to the log server and served statically | 18:21 |
| krotscheck | clarkb: Perfect. Point me at the publisher that does that? | 18:22 |
| krotscheck | clarkb: I can definitely generate HTML | 18:22 |
| harlowja | clarkb jeblair cliff and stevedore would probably also need to get pulled in olso i guess? | 18:23 |
| fungi | i was going to pull up an example nova coverage report, but apparently their coverage jobs are crashing with tracebacks | 18:23 |
| clarkb | krotscheck: http://git.openstack.org/cgit/openstack-infra/config/tree/modules/openstack_project/files/jenkins_job_builder/config/macros.yaml#n289 | 18:23 |
| clarkb | krotscheck: jobs that use that put the html in a dir called cover and that gets copied to the logs server | 18:23 |
| krotscheck | clarkb: Easy enough. | 18:24 |
| fungi | krotscheck: http://logs.openstack.org/b9/b9dd7c74055fa17fe6fe4e6632e6df219f315c08/post/swift-coverage/8924e85/ | 18:24 |
| fungi | (pretty example) | 18:24 |
| jeblair | harlowja: only if dhellmann wants tighter integration, or to gate changes to openstack on cliff and stevedore; right now the gate is only the other way so it's not a problem. | 18:25 |
| jeblair | harlowja: it seems like taskflow is hitting a need for tighter integration | 18:25 |
| harlowja | ya, or just the first to bring up transitive dependencies in this way | 18:26 |
| zaro | jeblair: do we use any features, besides replication, in gerrit that are now plugins? | 18:27 |
| jeblair | harlowja: yeah. there are other ways to solve it; knowing sqlalchemy, they are probably all bad and involve lots of things like "if sqlalchemy version < x" | 18:27 |
| *** rwsu has quit IRC | 18:28 | |
| harlowja | ya, puke | 18:28 |
| *** ruhe is now known as _ruhe | 18:28 | |
| fungi | zaro: where's the official list of things which became plugins? | 18:28 |
| clarkb | zaro: pretty sure we use the list installed on review-dev | 18:28 |
| *** dcramer_ has quit IRC | 18:28 | |
| jeblair | zaro: just looking at the list at the top of https://etherpad.openstack.org/p/gerrit-2.8-upgrade | 18:29 |
| jeblair | zaro: the only thing there that i think we should look at early is monitoring plugin | 18:29 |
| jeblair | zaro: the others can be deferred | 18:29 |
| *** johnthetubaguy1 has quit IRC | 18:30 | |
| zaro | is that something we want to test out on review-dev? | 18:30 |
| zaro | i mean now or later time? | 18:30 |
| openstackgerrit | David Lyle proposed a change to openstack-infra/reviewstats: Updating horizon core https://review.openstack.org/64877 | 18:31 |
| *** hashar has quit IRC | 18:31 | |
| jeblair | zaro: we should test monitoring now; if we can go into production with it available to help us deal with any problems, that would be good. | 18:31 |
| harlowja | jeblair cool, after i guess dhellmann thinks it over i guess we can figure out the next steps here | 18:32 |
| *** _ruhe is now known as ruhe | 18:33 | |
| zaro | i'll get going on installing and configuring that. everything seems like its working pretty good so far. | 18:33 |
| zaro | fungi: i can't seem to find an official list. | 18:36 |
| mfer | folks, is there a guide for creating a new project that includes getting setup on launchpad as a project and all the little details? | 18:37 |
| mfer | i keep searching around and i'm not finding anything that goes through the whole process | 18:37 |
| *** yolanda has joined #openstack-infra | 18:37 | |
| *** yolanda has quit IRC | 18:40 | |
| *** yolanda has joined #openstack-infra | 18:40 | |
| zaro | mfer: have you read this? https://wiki.openstack.org/wiki/How_To_Contribute | 18:41 |
| mfer | zaro yeah. i'm looking at the details of spinning up a new project starting on stackforge. things like adding the project to launchpad and all of that | 18:41 |
| mfer | i took on the golang-client and i'd like to get that setup completed | 18:42 |
| jeblair | mfer: http://ci.openstack.org/stackforge.html is all we have; launchpad isn't really covered (it is optional) | 18:42 |
| jeblair | mfer: the main thing to do is to give the 'openstackci' account bug triage access if you do decide to use it. | 18:42 |
| *** yolanda has quit IRC | 18:42 | |
| mfer | jeblair that part is done. it's the launchpad part i'm interested in now | 18:43 |
| jeblair | mfer: anything else is up to you | 18:43 |
| *** yolanda has joined #openstack-infra | 18:43 | |
| mfer | jeblair what's the right way to get a launchpad project setup as part of the openstack project? | 18:43 |
| *** markmcclain has joined #openstack-infra | 18:43 | |
| *** yolanda has quit IRC | 18:43 | |
| clarkb | I need to run and grab lunch before heading to the airport, but a quick glance at the zuul status page indicates neutron is having trouble | 18:44 |
| *** yolanda has joined #openstack-infra | 18:44 | |
| zaro | mfer all you need to do is create a launchpad project then configure one of the yaml files to point to that project. | 18:44 |
| *** rossella_s has quit IRC | 18:44 | |
| *** garyk has quit IRC | 18:45 | |
| zaro | mfer: yaml file is in openstack/config project | 18:45 |
| mfer | zaro and that will put it under the openstack (https://launchpad.net/openstack) project on launchpad? | 18:45 |
| fungi | mfer: for that, you just configure your project on lp and set the "part of" option to the openstack project group | 18:46 |
| mfer | k, thanks | 18:46 |
| fungi | there's no central control over that. individual projects seem to add themselves if they feel like it. i think the only thing it probably gets you is blueprint integration from commit messages | 18:47 |
| fungi | (since the blueprint hyperlinks in gerrit send you to a lp blueprint search against "openstack") | 18:47 |
| mfer | ok | 18:48 |
| mfer | it's also a matter of branding and direction if someone is purusing launchpad | 18:48 |
| openstackgerrit | A change was merged to openstack-infra/reviewstats: Updating horizon core https://review.openstack.org/64877 | 18:48 |
| *** coolsvap has quit IRC | 18:49 | |
| *** yolanda has quit IRC | 18:51 | |
| *** yolanda has joined #openstack-infra | 18:51 | |
| *** yolanda has quit IRC | 18:52 | |
| openstackgerrit | Michael Krotscheck proposed a change to openstack-infra/config: Add storyboard-webclient to gerrit https://review.openstack.org/64761 | 18:52 |
| *** banix has quit IRC | 18:53 | |
| *** thedodd has joined #openstack-infra | 18:55 | |
| *** dstanek has quit IRC | 18:58 | |
| *** hashar has joined #openstack-infra | 18:58 | |
| *** herndon has quit IRC | 19:00 | |
| *** coolsvap has joined #openstack-infra | 19:00 | |
| *** gyee_ has quit IRC | 19:02 | |
| *** dstanek has joined #openstack-infra | 19:02 | |
| *** herndon has joined #openstack-infra | 19:03 | |
| *** hashar has quit IRC | 19:05 | |
| morganfainberg | bknudson, ping | 19:07 |
| bknudson | morganfainberg: what's up? | 19:07 |
| morganfainberg | https://review.openstack.org/#/c/63671/13/keystone/assignment/core.py i don't think we should delete the tokens if the update fails | 19:07 |
| bknudson | morganfainberg: just because the update raised an exception doesn't mean that the project wasn't actually updated. | 19:07 |
| bknudson | the exception might happen after the commit. | 19:07 |
| bknudson | (in the case of SQL) | 19:08 |
| bknudson | so do you want the project disabled and the tokens not revoked? | 19:08 |
| morganfainberg | eh. sure i'll move it | 19:08 |
| bknudson | I believe the way the code worked before is that the tokens were revoked before the project was updated. | 19:08 |
| jeblair | clarkb, fungi, mordred: http://lists.openstack.org/pipermail/openstack-dev/2014-January/023399.html | 19:09 |
| morganfainberg | likely we should have some mechanism checking the project is actually disabled before revoking. but sure. will move it | 19:09 |
| openstackgerrit | James E. Blair proposed a change to openstack-infra/config: Add a devstack-gate core group https://review.openstack.org/64882 | 19:10 |
| fungi | jeblair: thanks! | 19:11 |
| krotscheck | Can I get someone to look at https://github.com/krotscheck/storyboard-webclient/blob/master/bin/bootstrap.sh and tell me that I'm not accidentally using a package that's not installed on the jenkins box? | 19:15 |
| *** rongze has quit IRC | 19:15 | |
| krotscheck | wget, autoconf, etc etc... | 19:15 |
| openstackgerrit | James E. Blair proposed a change to openstack-infra/config: Zuul status: don't toggle on link click https://review.openstack.org/64716 | 19:15 |
| *** rongze has joined #openstack-infra | 19:16 | |
| jeblair | krotscheck: they should all be there | 19:16 |
| krotscheck | Awesome, thanks. | 19:16 |
| openstackgerrit | Brant Knudson proposed a change to openstack-infra/elastic-recheck: Add query for bug 1265672 https://review.openstack.org/64885 | 19:17 |
| *** rongze has quit IRC | 19:20 | |
| bknudson | jenkins really had his way with that one. | 19:20 |
| bknudson | "No distributions at all found for lazr.authentication" again. | 19:21 |
| bknudson | I don't know what happened to fix it last time. | 19:22 |
| Ajaeger | jeblair, fungi, clarkb: I would love to see https://review.openstack.org/#/c/64567/ reviewed and approved if possible - to create some better docu gates. | 19:23 |
| *** praneshp_ has joined #openstack-infra | 19:25 | |
| *** jgrimm has joined #openstack-infra | 19:25 | |
| *** praneshp has quit IRC | 19:27 | |
| *** praneshp_ is now known as praneshp | 19:27 | |
| *** salv-orlando_ has joined #openstack-infra | 19:30 | |
| fungi | bknudson: pip 1.5 is refusing to download lazr.authentication because it's not hosted on pypi (this is a new behavior for pip, introduced in yesterday's 1.5 release) | 19:32 |
| fungi | bknudson: the proper fix is to get lazr.authentication to start uploading releases to pypi properly rather than merely linking to them elsewhere | 19:33 |
| *** salv-orlando has quit IRC | 19:34 | |
| *** salv-orlando_ is now known as salv-orlando | 19:34 | |
| fungi | bknudson: the short-term workaround will probably be to whitelist that package in your tox.ini so that pip 1.5 will agree to install it, even though it's potentially unsafe | 19:34 |
| fungi | bknudson: noted in item 3.2 at https://etherpad.openstack.org/p/pip1.5Upgrade | 19:35 |
| *** CaptTofu_ has quit IRC | 19:35 | |
| fungi | bknudson: you probably want to change tox.ini to have something like... | 19:36 |
| fungi | install_command = pip install --allow-external lazr.authentication --allow-unverified lazr.authentication -U {opts} {packages} | 19:36 |
| notmyname | I don't understand what's going on with https://review.openstack.org/#/c/63918/. pretty much all tests are failing (including those used by other projects), but there is just a "reverify bug 123" comment. is something else going on with the gate? | 19:36 |
| zaro | fungi: would it be possible to bring openstack/2.8 up to date with upstream/stable-2.8 ? we need a patch that's in stable-2.8 | 19:37 |
| zaro | fungi: it's for the gerrit monitoring plugin | 19:37 |
| *** dstanek has quit IRC | 19:38 | |
| clarkb | zaro: aren't we deploying upstream gerrit? | 19:38 |
| clarkb | zaro: eg it doesn't matter what we have in our gerrit fork | 19:38 |
| zaro | clarkb: we are forking so we can apply the zaro patch. | 19:39 |
| clarkb | I thought your patch was applied | 19:39 |
| clarkb | or is it 2.9 that will have it? | 19:39 |
| clarkb | (I can't keep up) | 19:40 |
| zaro | applied to master which is 2.9 | 19:40 |
| clarkb | got it | 19:40 |
| zaro | fungi: actually don't bother. i'll just build from upstream/2.8 branch for now. | 19:40 |
| zaro | fungi: i meant stable-2.8 branch | 19:40 |
| fungi | zaro: okay. well, i can certainly update the openstack/2.8 branch as well if you need | 19:40 |
| clarkb | what if we just deploy master? | 19:41 |
| clarkb | I think wikimedia does that and gerrit itself does it | 19:41 |
| *** rwsu has joined #openstack-infra | 19:41 | |
| *** dstanek has joined #openstack-infra | 19:41 | |
| jeblair | clarkb: how about we stick with 2.8 for now and consider tracking master later? | 19:42 |
| clarkb | wfm | 19:43 |
| jeblair | (so we can finish upgrading asap, and keep that delta to a known quantity) | 19:43 |
| *** elasticio has joined #openstack-infra | 19:44 | |
| zaro | fungi: ohh yeah. i think we need to move it forward so i can apply my patch on top. | 19:44 |
| *** coolsvap has quit IRC | 19:44 | |
| zaro | i forgot to do that on the current build | 19:45 |
| *** rakhmerov has quit IRC | 19:45 | |
| smarcet | fungi: could u please go ahead a rename "master" branch to "feature/openid" and "feature/openid-oauth2" to "master" on review system ? :) | 19:46 |
| *** dstanek has quit IRC | 19:47 | |
| fungi | smarcet: will do | 19:47 |
| fungi | just a moment... | 19:47 |
| smarcet | fungi: tnx sire :) | 19:48 |
| fungi | smarcet: okay, you now have a feature/openid branch at 83e5491 and a master branch at 6523a1f | 19:53 |
| smarcet | ok cool | 19:53 |
| smarcet | tnx a lot sire :) | 19:53 |
| fungi | smarcet: i'll propose a cherry-pick of the .gitreview addition onto master in just a moment | 19:54 |
| smarcet | fungi: ok cool tnx sire :) | 19:54 |
| beekneemech | notmyname: I'm seeing the same thing on some of the tests for https://review.openstack.org/#/c/61037/9 | 19:55 |
| beekneemech | The problem seems to be this line in error.txt.gz: [ERROR] /opt/stack/new/devstack/functions:378 Failure creating NET_ID for 36033ee931044cb5bc76570da0eff97b | 19:56 |
| beekneemech | No clue what's causing it though. | 19:56 |
| *** hogepodge has joined #openstack-infra | 19:56 | |
| fungi | smarcet: https://review.openstack.org/64894 | 19:58 |
| smarcet | fungi: ok cool | 19:58 |
| *** vipul is now known as vipul-away | 20:00 | |
| fungi | zaro: okay, so you want me to move openstack-infra/gerrit's openstack/2.8 branch to be at the same commit as the upstream/stable-2.8 branch (45bcb39)? | 20:05 |
| *** MarkAtwood has joined #openstack-infra | 20:05 | |
| *** vipul-away is now known as vipul | 20:07 | |
| bknudson | fungi: I was able to recreate the problem locally this time. | 20:08 |
| openstackgerrit | David Kranz proposed a change to openstack-infra/devstack-gate: comparison to stable/grizzly is not numeric https://review.openstack.org/63934 | 20:08 |
| *** flaper87|afk is now known as flaper87 | 20:08 | |
| openstackgerrit | David Kranz proposed a change to openstack-infra/devstack-gate: comparison to stable/grizzly is not numeric https://review.openstack.org/63934 | 20:10 |
| bknudson | it says "Some insecure and unverifiable files were ignored (use --allow-unverified lazr.authentication to allow)." | 20:13 |
| sdague | dhellmann: sorry I missed the earlier thread | 20:13 |
| bknudson | but when I do that it says "no such option: --allow-unverified" | 20:13 |
| *** dcramer_ has joined #openstack-infra | 20:13 | |
| bknudson | maybe it's different versions of pip | 20:14 |
| *** ArxCruz has quit IRC | 20:14 | |
| dhellmann | sdague: here? I don't know that we discussed much, just getting some context for the email thread | 20:15 |
| sdague | this all comes back to requirements wedging | 20:15 |
| *** ^d is now known as ^demon|away | 20:15 | |
| *** rongze has joined #openstack-infra | 20:16 | |
| *** rakhmerov has joined #openstack-infra | 20:16 | |
| sdague | I spent a month this summer unwinding this for all the projects, when we had to get 12 projects to uncap libs in a specific order to get them past the gate | 20:16 |
| sdague | and do client releases in a specific order | 20:16 |
| bknudson | pip 1.4.1 from /home/bknudson/dev/elastic-recheck/.tox/pep8/lib/python2.7/site-packages (python 2.7) | 20:16 |
| dhellmann | sdague: we need a way to fix that besides bringing every library we use into openstack, though | 20:17 |
| sdague | dhellmann: well, if you come up with a way, tell me :) | 20:17 |
| bknudson | I upgraded virtualenv... sudo pip install --upgrade virtualenv | 20:18 |
| dhellmann | sdague: well, some projects just aren't going to let us take over, right? | 20:18 |
| sdague | because global requirements forced update is the best idea I've had here | 20:18 |
| sdague | dhellmann: sure, but there are ones that are so embedded, that we can | 20:18 |
| sdague | and taskflow is that | 20:18 |
| dhellmann | sdague: I have considered changing stevedore so it uses pkg_resources to find plugins, but then does the import itself without enforcing any of the requirements stuff that pkg_resources does on an import | 20:18 |
| dhellmann | sdague: nature abhors the special case | 20:19 |
| sdague | dhellmann: so all that would do is take us back to the old problem here right, where when we installed taskflow, it would uninstall sqla 0.8 and install the version it liked | 20:19 |
| sdague | then depending on who ran last, we'd have a different version | 20:20 |
| dhellmann | sdague: because if we test libraries' trunks instead of releases, we will have to release the libraries at the same time that we release the apps | 20:20 |
| mriedem | beekneemech: i'm seeing the same here: http://logs.openstack.org/19/60719/4/check/gate-tempest-dsvm-neutron-large-ops/1f73559/logs/error.txt.gz | 20:20 |
| *** rakhmerov has quit IRC | 20:20 | |
| mriedem | devstack setup fails around the time of some q-metering stuff | 20:20 |
| dhellmann | sdague: well, yes, but we can ask the library author to fix the requirements, right? | 20:20 |
| fungi | bknudson: that option is new to (and required by, in your case) pip 1.5. if you're using pip embedded in a virtualenv, such as via tox, you also want virtlauenv 1.11 which bundles pip 1.5 as well | 20:20 |
| sdague | dhellmann: sure, you signing up for that? because it was a 50% job all summer long | 20:21 |
| sdague | until we got g-r | 20:21 |
| sdague | we have over 100 requirements :) | 20:21 |
| dstufft | bknudson: pip 1.4 has --allow-insecure | 20:21 |
| dstufft | so does pip 1.5 | 20:21 |
| dstufft | the option name was changed in 1.5 to be more descriptive | 20:21 |
| dhellmann | I mean, I'm aware of the amount of time we spend on chasing down these requirements conflicts and I've done my share of pushing new versions of dependencies, but I don't think making the libraries move in lockstep with the apps is the right solution | 20:21 |
| dstufft | but the old option is still there, it's just hidden from help output | 20:21 |
| openstackgerrit | Brant Knudson proposed a change to openstack-infra/elastic-recheck: Tox allow lazr.authenticate install https://review.openstack.org/64898 | 20:21 |
| openstackgerrit | Brant Knudson proposed a change to openstack-infra/elastic-recheck: Add query for bug 1265672 https://review.openstack.org/64885 | 20:21 |
| zaro | fungi: yes, that is correct. | 20:21 |
| sdague | dhellmann: so I provided the other option, no library is ever allowed to specify a max verison | 20:22 |
| fungi | dstufft: so if we use --allow-insecure instead of --allow-unverified that will work with both pip 1.5 and older (1.4.1 at least)? | 20:22 |
| sdague | ever | 20:22 |
| dstufft | fungi: yea | 20:22 |
| dhellmann | sdague: i missed that, that would work for me | 20:22 |
| sdague | it was in the email | 20:22 |
| dhellmann | sdague: I try to do that myself | 20:22 |
| sdague | dhellmann: this would be an absolute | 20:22 |
| fungi | dstufft: you are a life saver ;) | 20:22 |
| sdague | because if it's not, we're back into wedge land | 20:23 |
| *** vipul is now known as vipul-away | 20:23 | |
| beekneemech | mriedem: Yeah, it seems like there is a legitimate problem here. Think I'll open a bug and try a recheck to see if it reproduces. | 20:23 |
| dstufft | it won't work in 1.3 | 20:23 |
| dstufft | FWIW | 20:23 |
| fungi | dstufft: is there a deprecation on --allow-insecure? (will it be going away in 1.6?) | 20:23 |
| dstufft | fungi: uh | 20:23 |
| mriedem | beekneemech: was just going to open one, but go ahead - link me and i'll post my logs | 20:23 |
| dhellmann | sdague: ah, your most recent email just landed in my inbox | 20:23 |
| *** rongze has quit IRC | 20:23 | |
| dhellmann | I'm OK with that being a rule for stackforge libraries | 20:23 |
| beekneemech | mriedem: Will do. | 20:23 |
| openstackgerrit | Brant Knudson proposed a change to openstack-infra/elastic-recheck: Add query for bug 1265672 https://review.openstack.org/64885 | 20:23 |
| openstackgerrit | Brant Knudson proposed a change to openstack-infra/elastic-recheck: Tox allow install of lazr.authentication https://review.openstack.org/64898 | 20:23 |
| sdague | harlowja: you ok with removing all your caps in taskflow and releasing a new library today? | 20:24 |
| dhellmann | sdague: I like it better than testing unreleased versions of stuff | 20:24 |
| harlowja | all caps? | 20:24 |
| dstufft | fungi: the changelog says "a period of time", but tbh I doubt there will ever be a compelling reason to remove it other than as part of some sort of down the road cleanup kind of thing | 20:24 |
| harlowja | i think all caps shouldn't be to bad actually, only pbr it seems | 20:25 |
| sdague | harlowja: all specifications of maximum versions | 20:25 |
| harlowja | and SQLAlchemy | 20:25 |
| fungi | dstufft: though even if it does, that's tractable i guess. right now the sticking point i that i can't patch nova's tox.ini to work with pip 1.5 until virtualenv releases a newer version with the sitepackages fix included, because we're pinning to virtualenv 1.10.1 on our systems to get around that for now, so adding --allow-unverified would cause the embedded pip 1.4.1 to break | 20:25 |
| bknudson | looks like now jenkins is running with old virtualenv / pip: http://logs.openstack.org/98/64898/1/check/gate-elastic-recheck-pep8/47db1b2/console.html | 20:25 |
| bknudson | http://logs.openstack.org/98/64898/1/check/gate-elastic-recheck-pep8/47db1b2/console.html#_2014-01-03_20_21_47_534 | 20:26 |
| dstufft | fungi: the code has a comment for # Remove --allow-insecure after 1.7 | 20:26 |
| bknudson | "no such option: --allow-unverified" | 20:26 |
| sdague | harlowja: ok, cool, as soon as you can do that, we can probably officially support sqla 0.8 | 20:26 |
| harlowja | cool | 20:26 |
| fungi | dstufft: but sounds like --allow-insecure will get us around it for now, and we can just make a point of transitioning once virtualenv 1.11.1 or 1.12 or something comes along | 20:26 |
| harlowja | let me see if ivan is running any additional tests, but don't see it as to bad | 20:26 |
| *** ryanpetrello has quit IRC | 20:26 | |
| fungi | dstufft: thanks again | 20:26 |
| *** ryanpetrello has joined #openstack-infra | 20:27 | |
| sdague | dhellmann: so the other issue about apps vs. libs needs someone to rework the matrix then. Because it's different than what we do, and we do what we do because we'd get broken by python client releases | 20:27 |
| dstufft | fungi: np | 20:27 |
| mriedem | we don't index on logs/error.txt.gz do we :( | 20:28 |
| dhellmann | sdague: our client libraries may be the special case | 20:28 |
| *** lcestari has quit IRC | 20:28 | |
| sdague | well, I don't think they are more special than oslo? | 20:28 |
| beekneemech | mriedem: https://bugs.launchpad.net/openstack-ci/+bug/1265904 | 20:28 |
| sdague | oslo releases the same time as the servers, the clients don't | 20:28 |
| dhellmann | sdague: because I see the client libs as bound much more tightly to the apps than other libraries | 20:28 |
| mriedem | beekneemech: thanks | 20:29 |
| fungi | zaro: okay, openstack/2.8 is now at the same commit as upstream/stable-2.8 | 20:29 |
| dstufft | fungi: FWIW I think all/most of pip's options can be declared as env vars too, though i'm not 100% on that | 20:29 |
| dstufft | fungi: incase this ever comes up in a case where there isn't a simple fix like --allow-insecure | 20:30 |
| sdague | so I think there is an interesting matrix here, and maybe a way to do this better. But I'll admit I don't have time to rejigger in this cycle. | 20:30 |
| fungi | dstufft: oh, awesome i'll try that too (would be a potentially better workaround) | 20:30 |
| dhellmann | sdague: we've been conservative with oslo library releases because (a) we don't expect anyone else to be using them and (b) the code has gone through the incubator and had a chance to "stabilize" there | 20:31 |
| dhellmann | for libraries where one or neither of those things is not true, we should just be relying on releases | 20:31 |
| sdague | sure, though taskflow is pretty openstack specific at this point as well | 20:31 |
| dhellmann | my understanding from harlowja was that someone other than openstack is already using it | 20:31 |
| harlowja | ya | 20:32 |
| harlowja | SEJeff_work is | 20:32 |
| dhellmann | and in fact adopted it first? | 20:32 |
| harlowja | dhellmann not sure if first | 20:32 |
| dhellmann | ok, I may have imagined that part :-) | 20:32 |
| harlowja | and i've heard at&t is using it for other stuff internally as well (although nothing publicly said) | 20:32 |
| sdague | ok, from my pov it feels like part of oslo, as it's a common library | 20:33 |
| sdague | built for openstack, even if others are using it | 20:33 |
| zaro | fungi: thanks. | 20:33 |
| sdague | but I'll let others make that call | 20:33 |
| dhellmann | sdague: I would like to avoid having OpenStack become a black hole for libraries (oslo or otherwise) that no one else uses because they are so tightly bound up in our processes, so I am tending to push back on the idea of every library being an oslo library. | 20:34 |
| dstufft | dhellmann: +1 :D | 20:34 |
| sdague | dhellmann: that's fair | 20:35 |
| sdague | we just need to get a better way to premptively test these things en-mass | 20:35 |
| bknudson | dhellmann: how do we handle common code? | 20:36 |
| dhellmann | sdague: +1000 | 20:36 |
| dhellmann | bknudson: common? | 20:36 |
| sdague | because come the last week of March... we're going to have every other day gate wedge as people prep for pycon | 20:36 |
| bknudson | dhellmann: the stuff that's in oslo-incubator... it stays there? | 20:36 |
| sdague | I mean we couldn't even survive a virtualenv release :) | 20:36 |
| dhellmann | sdague: well, it may be appropriate for us to tighten our requirements specs during times like that | 20:36 |
| sdague | and dstufft hangs with us | 20:36 |
| dhellmann | bknudson: no, the incubation-to-graduation process markmc laid out is still the plan for that | 20:36 |
| dhellmann | sdague: what about running a periodic test against "newer" versions of our dependencies? | 20:37 |
| jeblair | it's not intractable for a library to have version caps. it's only intractable if they are intentionally keeping them at a level that doesn't allow openstack to advance | 20:37 |
| jeblair | it would work fine for taskflow to bump its sqlalchemy requirement, release, and then for openstack to depend on that newer release | 20:38 |
| harlowja | jeblair https://review.openstack.org/#/c/64905/ | 20:38 |
| jeblair | (taskflow having no caps is fine with me; not trying to argue against that) | 20:38 |
| sdague | jeblair: so.... that only works in practice when the number of libs that you are dealing with is small | 20:38 |
| jeblair | just pointing out that i don't think it's the only option. libraries specifying versions they work with is fine, but only if they _correctly_ specify the versions they work with. | 20:39 |
| sdague | wc -l global-requirements.txt | 20:39 |
| sdague | 150 global-requirements.txt | 20:39 |
| sdague | it really took me *3* weeks to get just 12 projects sorted out this summer | 20:39 |
| jeblair | sdague: yeah, as a big project our dealing with lots of libraries is hard; but i'm not sure we can expect to "control" all of them, and defining our processes based on the assumption we can ask any dependent library to remove all caps is suspect in my mind. | 20:40 |
| dhellmann | sdague: I spent a whole release cycle updating webob, so I feel your pain | 20:40 |
| jeblair | (especially when our own policy is that we must have caps because to have uncapped library depends is _crazy_) | 20:40 |
| sdague | jeblair: so the alternative is integrated gate | 20:40 |
| *** MarkAtwood has quit IRC | 20:40 | |
| jeblair | sdague: alternative to what? | 20:40 |
| sdague | to no caps | 20:41 |
| jeblair | sdague: alternative to using library version dependencies the way they are intended? we can not take that off the table. | 20:41 |
| sdague | well, someone else can manage the requirements matrix then. I think I've given all the blood I can on that one | 20:42 |
| jeblair | sdague: we have 3 choices: (a) taskflow _correctly_ describes versioun requirements (b) taskflow uncaps version requirements (c) taskflow joins openstack and the integrated gate | 20:42 |
| *** vipul-away is now known as vipul | 20:42 | |
| harlowja | or (d) taskflow bumps its version cap for sqlalchemy | 20:42 |
| jeblair | sdague: as i said, i'm not trying to talk you and harlowja out of your agreement to uncap versions | 20:42 |
| jeblair | harlowja: (that's what i meant by (a)) | 20:42 |
| harlowja | ah | 20:43 |
| jeblair | sdague: however, i am trying to talk you out of going on a crusade to convince 150 libraries to remove all their max caps. :) | 20:43 |
| sdague | :) | 20:43 |
| sdague | well, I'm also trying to be pragmatic | 20:43 |
| sdague | so I'm not actually going on that crusade | 20:44 |
| beekneemech | https://bugs.launchpad.net/openstack-ci/+bug/1265904 appears to have completely wedged the gate. | 20:44 |
| beekneemech | Everything on the zull status page is red right now. | 20:44 |
| beekneemech | *zuul | 20:44 |
| dstufft | FWIW the ~future~ will be to have dependency stuff default to something similar to >=1.1,<2 style thing | 20:44 |
| *** banix has joined #openstack-infra | 20:45 | |
| SEJeff_work | dhellmann, Yes my issue with some of the openstack bits is that it seems to be all or nothing | 20:46 |
| dhellmann | SEJeff_work: right, I'm trying to avoid that where practical in new libraries | 20:46 |
| SEJeff_work | I've love for taskflow to not hard depend on the whole of openstack. Unfortunately, openstack doesn't fit the bill of what I'm building for server automation. Taskflow however, does. I'd prefer not to require openstack where possible | 20:46 |
| sdague | jeblair: I guess the real issue is libraries that depend on things we know are unstable | 20:46 |
| dstufft | sdague: FWIW a lot of the virtualenv problem is that virtualenv has all of like 12 tests on it's own, and our largest test of virtualenv is the fact that the pip test uses the hell out of venv :( | 20:46 |
| sdague | like sqla | 20:46 |
| SEJeff_work | but am sympathetic to your side if it absolutely has to depend on openstack | 20:47 |
| SEJeff_work | dhellmann, Thanks | 20:47 |
| sdague | so we need need to have a global lever for unstable libs | 20:47 |
| sdague | otherwise upgrading is a paris peace treaty | 20:47 |
| dhellmann | sdague: it's too late for this year, but we should put together a pycon talk for next year | 20:48 |
| jeblair | sdague: yeah, that's definitely something we need to consider; and negotiating/understanding if a lib is going to mantain wide sqla compatibility (or upgrade asap on new sqla releases), or uncap sqla requirements is in-scope for requirements repo reviews. | 20:48 |
| dhellmann | sdague: my guess is most projects don't hit the sorts of conflicts we do | 20:49 |
| sdague | dhellmann: agreed | 20:49 |
| jeblair | sdague: i'm totally cool with a policy of "if there is a transitive dep on sqla, the middle lib must be responsive to upgrades and version bumps, or have no sqla version cap" | 20:49 |
| *** sarob has joined #openstack-infra | 20:49 | |
| jeblair | sdague: i'm not sure i'd want to apply that universally to all our libs and all possible transitive deps. | 20:49 |
| jeblair | but to known problems, it seems very reasonable | 20:50 |
| *** senk has joined #openstack-infra | 20:50 | |
| sdague | jeblair: so I actually think "be responsive" isn't good enough. Because it means we can't actually ever test uncaps | 20:50 |
| sdague | without lots of people negotiation | 20:50 |
| sdague | it stops being CI when it involves asking people for new releases | 20:51 |
| jeblair | sdague: sure we can; once taskflow determines it is compatible with new sqla and releases, we can. | 20:51 |
| jeblair | sdague: we shouldn't ask them. that's what "be responsive" means. :) | 20:51 |
| jeblair | sdague: be responsive to sqla having made a new release. | 20:51 |
| jeblair | sdague: that having been done, taskflow should validate their compatibility with it, and then release with a new version cap. | 20:52 |
| mriedem | beekneemech: looking through the devstack recent commits i see a few red flags related to networking and neutron | 20:52 |
| sdague | you know.... there would be another way. If pip let us do requirements overrides on packages.... | 20:53 |
| sdague | basically a similar lever flip like what g-r gives us | 20:53 |
| *** yolanda has joined #openstack-infra | 20:53 | |
| sdague | letting us speculate in a world where the underlying packages are wrong in their requirements | 20:53 |
| * fungi needs to pop out and get some food... back in a bit | 20:53 | |
| jeblair | yep | 20:53 |
| sdague | dstufft: how crazy pants does that sound to you? | 20:54 |
| dstufft | jeblair: well ideally you wouldn't be bumping with each new release, you'd set the version spec to whatever the libraries declared backwards comapt policy allows (lol yea right :V) | 20:54 |
| dstufft | sdague: not sure what you're looking for exactly, you want to be able to say "my library A depends on B, and B depends on C, but I want to pretend B actually depends on C<1.0" ? | 20:55 |
| jeblair | dstufft: yes, ideally -- but we keep using sqla as an example because like every point release is backwards incompat or something crazy like that. | 20:55 |
| sdague | dstufft: yeh, basically | 20:55 |
| sdague | dstufft: pip --overide-version "C>1.5" | 20:56 |
| sdague | then any time it tries to install C from anywhere, it ignores the requirements.txt version, and uses the override instead | 20:56 |
| dstufft | sdague: so additive is simple because in a world where pip has an actual dependency resolver (not yet :() you'd just add your own dependency on C, if you want to full out replace that's a bit different | 20:56 |
| sdague | dstufft: we typically need a full out replace | 20:57 |
| dstufft | sdague: pip's dependency resolver is really really really dumb at the moment | 20:57 |
| dstufft | but it's dumb in a way that probably allows this | 20:57 |
| jeblair | sdague: hypothetically if that existed; it would be useful for those experiments, but it's worth considering whether we want to have software that is not actually installable without using that... | 20:57 |
| sdague | dstufft: and we need the override because we actually need to do about 20 different pip installs to setup openstack | 20:57 |
| dstufft | basically pip's dependency resolver just accepts whatever the first version spec is for a package | 20:58 |
| dstufft | so if you have in a chain of dependencies, C>=1.0 and C==1.1, it'll see it as just C>=1.0 if it saw that one first | 20:58 |
| *** melwitt has joined #openstack-infra | 20:58 | |
| sdague | jeblair: right, but it would give us feedback to get upstream | 20:58 |
| dstufft | so I think all you'd need to do right now, is do ``pip install C>1.5 -r requirements.txt`` | 20:58 |
| *** vipul is now known as vipul-away | 20:59 | |
| jeblair | sdague: i would imagine that we would not actually want to land a change that required a version override; but yes, could provide info. | 20:59 |
| sdague | so we'd know "taskflow, your requirements are wrong" | 20:59 |
| sdague | and be able to put the paris peace treaty on the other end of the process, as we'd already have proven results | 20:59 |
| dstufft | as far as actually supporting that as a first class feature | 20:59 |
| SEJeff_work | Feels more like a Mexican Standoff at the moment | 21:00 |
| dstufft | it's probably doable, I'd have to think about it and it'd need discussed and such, but I don't think it's an unreasonable request | 21:00 |
| dstufft | although for testing you can get around it by just using --no-deps | 21:00 |
| *** melwitt1 has joined #openstack-infra | 21:00 | |
| sdague | dstufft: so in this particular break, I think we actually got the right thing out of pip | 21:00 |
| sdague | but then the entry point loading had the version spec encoded, and blew us up | 21:01 |
| dstufft | yea | 21:01 |
| dstufft | that's a setuptools feature | 21:01 |
| sdague | but if pip gets smarter.... having that override functionality would be really nice | 21:01 |
| dstufft | and pip doesn't really have any control over that | 21:01 |
| sdague | dstufft: agreed | 21:01 |
| sdague | but dhellmann said he was going to reconsider the way stevedore loads to get around that | 21:02 |
| sdague | about 1/3 of our requirements breaks have actually been that | 21:02 |
| sdague | it all actually would work fine, but entry points blew us up | 21:02 |
| *** vipul-away is now known as vipul | 21:02 | |
| dstufft | yea, that's mostly a side effect of the way setuptools tried to do virtual environments | 21:02 |
| *** melwitt has quit IRC | 21:03 | |
| dstufft | (install everything into sub directories of site-packages, like site-packages/C-1.0.egg/FILES and dynamically add the C 1.0 version to the sys.path) | 21:03 |
| dstufft | so it has checks like that to make sure the proper version was loaded | 21:04 |
| dstufft | sdague: FWIW console_sciprts when installed by pip from a Wheel don't use pkg_resources anymore :D | 21:05 |
| sdague | yay! | 21:05 |
| dstufft | in pip 1.5 anyways | 21:05 |
| beekneemech | mriedem: Yeah, I'm trying to reproduce locally so I can git bisect it. | 21:06 |
| dstufft | sdague: this is the new autogenerated script https://gist.github.com/dstufft/61001f4d4465e94e5bc3 , but only when installing from wheels with pip 1.5 | 21:07 |
| *** elasticio has quit IRC | 21:07 | |
| *** vipul is now known as vipul-away | 21:08 | |
| *** vipul-away is now known as vipul | 21:08 | |
| sdague | dstufft: cool, though probably getting over my head in the subtlties ;) | 21:09 |
| sdague | realistically I'm just a noob with way too many battle scars after the summer of global requirements :) | 21:09 |
| *** yolanda has quit IRC | 21:09 | |
| *** jasondotstar has quit IRC | 21:09 | |
| harlowja | k, so shall i unbind for 0.8 sdague , that seems simplest, although trunk is being worked on currently (expected to have 0.2 release sometime this month) so will have to get a couple patches in to make 0.1.2 presentable, does that seem fair? | 21:10 |
| *** yolanda has joined #openstack-infra | 21:10 | |
| *** yolanda has quit IRC | 21:10 | |
| dstufft | sdague: FWIW I'm pretty sure that global requirements is basically a requirement because of pip's brain dead dependency resolution scheme | 21:10 |
| dstufft | I *think* that fixing that in pip will remove the need for some of that orchecstration in openstack | 21:10 |
| sdague | dstufft: it's actually a requirement because we have to pip install each repo | 21:11 |
| sdague | so we don't get to do this in one go | 21:11 |
| sdague | harlowja: sure | 21:11 |
| *** yolanda has joined #openstack-infra | 21:11 | |
| sdague | I think as soon as you get a 0.1.2 out there we can probably put a nail in this sqla 0.8 thing | 21:11 |
| harlowja | k | 21:11 |
| *** senk has quit IRC | 21:11 | |
| sdague | then figure out how far off 0.9 is | 21:11 |
| dstufft | sdague: you can pip install more than one repo at a time, and realistically pip should not install something that is incompatible with the already installed stuff either | 21:12 |
| *** praneshp is now known as praneshp_lunch | 21:12 | |
| sdague | dstufft: well it just uninstalls the incompatible thing :) | 21:12 |
| sdague | because I did this we found some libraries were getting uninstalled and reinstalled up to 6 times during an openstack setup | 21:12 |
| * dhellmann envisions sdague wearing a "I survived an OpenStack requirements update" badge in Atlanta this spring | 21:12 | |
| sdague | it was awesome | 21:12 |
| harlowja | sdague as some of the other taskflow core is asleep, if this happens by say like mondayish is that ok? | 21:12 |
| sdague | harlowja: sure | 21:13 |
| sdague | harlowja: thanks for jumping on this | 21:13 |
| harlowja | np | 21:13 |
| *** senk has joined #openstack-infra | 21:13 | |
| harlowja | the never ending requirements battle :( | 21:13 |
| dstufft | sdague: what I mean, is when the dependency graph is built in any one ``pip install <foo>`` it should also include everything already installed as part of that graph | 21:13 |
| harlowja | anvil has taught me the same lessons | 21:13 |
| harlowja | *why i also made it completly avoid pip, lol | 21:13 |
| *** krotscheck_ has joined #openstack-infra | 21:14 | |
| harlowja | *well avoid installating via pip, pip good at downloading tarballs of the right version usually | 21:14 |
| *** krotscheck_ has quit IRC | 21:14 | |
| harlowja | so i feel the pain sdague :( | 21:15 |
| *** yolanda has quit IRC | 21:16 | |
| *** rakhmerov has joined #openstack-infra | 21:16 | |
| *** yolanda has joined #openstack-infra | 21:17 | |
| *** rongze has joined #openstack-infra | 21:19 | |
| *** rakhmerov has quit IRC | 21:21 | |
| *** Ajaeger has quit IRC | 21:22 | |
| *** rongze has quit IRC | 21:23 | |
| *** rfolco has quit IRC | 21:24 | |
| *** yolanda has quit IRC | 21:25 | |
| *** herndon has quit IRC | 21:28 | |
| beekneemech | mriedem: I went ahead and pushed a revert of my hostname change. Will see if it passes Jenkins now. | 21:30 |
| mriedem | beekneemech: ok, that was the only one that was really suspicious to me given the timing | 21:31 |
| *** smarcet has quit IRC | 21:32 | |
| beekneemech | mriedem: Yeah, although it merged something like 16 hours before this started happening. Could be an interaction with another change though. | 21:33 |
| *** herndon has joined #openstack-infra | 21:33 | |
| *** senk has quit IRC | 21:41 | |
| *** dprince has quit IRC | 21:41 | |
| beekneemech | mriedem: Looks like the revert is going to fail Jenkins too, so that's probably not it. :-( | 21:45 |
| *** chrisstreeter_ has joined #openstack-infra | 21:48 | |
| zaro | fungi: i think there's something wrong with the openstack/gerrit git project. | 21:53 |
| *** vipul is now known as vipul-away | 21:56 | |
| *** vipul-away is now known as vipul | 21:56 | |
| zaro | fungi: git submodule update doesn't work anymore. | 21:58 |
| *** beagles has quit IRC | 22:04 | |
| *** nati_ueno has quit IRC | 22:05 | |
| *** nati_ueno has joined #openstack-infra | 22:08 | |
| *** SergeyLukjanov has joined #openstack-infra | 22:10 | |
| *** sarob has quit IRC | 22:15 | |
| *** sarob has joined #openstack-infra | 22:16 | |
| *** rakhmerov has joined #openstack-infra | 22:17 | |
| *** prad has quit IRC | 22:17 | |
| *** rongze has joined #openstack-infra | 22:20 | |
| *** sarob has quit IRC | 22:20 | |
| *** jergerber has joined #openstack-infra | 22:20 | |
| *** rakhmerov has quit IRC | 22:22 | |
| *** nati_ueno has quit IRC | 22:22 | |
| *** rongze has quit IRC | 22:25 | |
| beekneemech | mriedem: Reproduced the problem locally. My neutron commands are failing with "'ExtensionManager' object does not support item assignment" | 22:26 |
| beekneemech | Thinking that's the problem. | 22:26 |
| mriedem | beekneemech: maybe a recent ML2 change in neutron? | 22:27 |
| *** banix has quit IRC | 22:27 | |
| beekneemech | mriedem: Not sure. I'm not even clear on whether the error is coming from neutron or the client. | 22:29 |
| *** MarkAtwood has joined #openstack-infra | 22:32 | |
| markmcclain | beekneemech we are currently experiencing problems with the client | 22:32 |
| beekneemech | mriedem: ^ | 22:33 |
| beekneemech | markmcclain: Okay, thanks. | 22:34 |
| *** jergerber has quit IRC | 22:34 | |
| mriedem | beekneemech: markmcclain: thanks | 22:34 |
| *** eharney has quit IRC | 22:35 | |
| beekneemech | Looks like probably this bug: https://bugs.launchpad.net/python-neutronclient/+bug/1265926 | 22:35 |
| *** mriedem has quit IRC | 22:35 | |
| *** vipul is now known as vipul-away | 22:35 | |
| markmcclain | just waiting on the changes to merge | 22:36 |
| *** vipul-away is now known as vipul | 22:36 | |
| openstackgerrit | Brant Knudson proposed a change to openstack-infra/reviewday: Generate JSON https://review.openstack.org/64471 | 22:39 |
| openstackgerrit | Matt Farina proposed a change to openstack-infra/config: Adding the launchpad project for the golang-client https://review.openstack.org/64919 | 22:41 |
| openstackgerrit | Brant Knudson proposed a change to openstack-infra/reviewday: Generate JSON https://review.openstack.org/64471 | 22:42 |
| *** ryanpetrello has quit IRC | 22:43 | |
| *** senk has joined #openstack-infra | 22:45 | |
| *** senk1 has joined #openstack-infra | 22:46 | |
| *** sarob has joined #openstack-infra | 22:46 | |
| *** burt has quit IRC | 22:46 | |
| *** mfer has quit IRC | 22:47 | |
| *** dcramer_ has quit IRC | 22:48 | |
| *** senk has quit IRC | 22:49 | |
| ruhe | [storyboard] can we use '{name}-branch-tarball' template with non-python project (assuming that tarball is built in dist/ too)? | 22:56 |
| clarkb | if python setup.p sdist builds the tarball then yes | 22:57 |
| *** sarob has quit IRC | 22:57 | |
| *** banix has joined #openstack-infra | 22:57 | |
| *** weshay has quit IRC | 22:58 | |
| *** hogepodge has quit IRC | 22:59 | |
| ruhe | clarkb: that project will not be built this way. should we create macros similar to '{name}-branch-tarball' under misc.yaml ? | 22:59 |
| reed | Delete is successfully failing | 23:00 |
| * reed parsing error | 23:00 | |
| reed | reading questions on Ask can be amusing | 23:01 |
| *** banix has quit IRC | 23:01 | |
| clarkb | ruhe: or just write out a job from scratch and start there | 23:02 |
| *** banix has joined #openstack-infra | 23:02 | |
| *** hogepodge has joined #openstack-infra | 23:03 | |
| *** nati_ueno has joined #openstack-infra | 23:07 | |
| ruhe | got it, thanks | 23:08 |
| *** banix has quit IRC | 23:10 | |
| *** salv-orlando has quit IRC | 23:14 | |
| *** salv-orlando has joined #openstack-infra | 23:14 | |
| *** pcrews has quit IRC | 23:15 | |
| *** flaper87 is now known as flaper87|afk | 23:15 | |
| *** thedodd has quit IRC | 23:16 | |
| *** rakhmerov has joined #openstack-infra | 23:18 | |
| *** rongze has joined #openstack-infra | 23:20 | |
| *** thedodd has joined #openstack-infra | 23:21 | |
| *** rakhmerov has quit IRC | 23:23 | |
| morganfainberg | it looks like gate is happier now the neutronclient change went in. just checking before starting to issue some rechecks. | 23:23 |
| morganfainberg | don't need to spam check/gate queues if it's going to just fail. | 23:23 |
| krotscheck | Question- given a specific pipeline trigger (say, post), are the individual jobs run in order without cleanup ( i.e. can I tarball the results from a previous job)? | 23:24 |
| openstackgerrit | Sergey Lukjanov proposed a change to openstack-infra/zuul: Collect and report last reconfigured timestamp https://review.openstack.org/63849 | 23:24 |
| jeblair | krotscheck: they are run simultaneously on different hosts, at random; however... | 23:24 |
| jeblair | krotscheck: you can indicate that a job should run _after_ another job, though the first job will need to stash its results somewhere | 23:24 |
| jeblair | krotscheck: see the -tarball and -pypi-upload jobs for an example of this | 23:25 |
| jeblair | krotscheck: but that should really only be used if the intermediate artifact is publishable on its own. | 23:25 |
| *** rongze has quit IRC | 23:25 | |
| jeblair | krotscheck: temporary storage of transient artifacts isn't expected | 23:25 |
| krotscheck | jeblair: Thanks for the clarification. ruhe's helping me get the tarball onto a server, but first we need to create a tarball. Ergo, questions blah blah. | 23:26 |
| jeblair | krotscheck: a tarball is exactly the kind of thing we would expect to be created and pushed to tarballs.o.o. | 23:26 |
| *** SergeyLukjanov has quit IRC | 23:26 | |
| *** senk1 has quit IRC | 23:26 | |
| krotscheck | jeblair: Really? Is that why it's called tarballs.o.o? | 23:29 |
| * krotscheck couldn't help himself. | 23:30 | |
| *** vipul is now known as vipul-away | 23:30 | |
| *** vipul-away is now known as vipul | 23:30 | |
| *** rcleere has quit IRC | 23:33 | |
| reed | jeblair, are you having problems today with Sonic.net DNS servers? | 23:33 |
| sdague | harlowja: so as I dive into the taskflow thing more in detail, I'm coming back towards the original integrated gate bits. Mostly from pragmatic pov. | 23:34 |
| *** thedodd has quit IRC | 23:35 | |
| harlowja | sdague sure, i'd be ok with that to from said pragmatic pov, it doesn't seem to change that much really | 23:35 |
| jeblair | reed: i don't use them; i run my own recursive resolver. | 23:35 |
| sdague | because we really can't have a library that prevents us from working on sqla 0.9 compatibility because it blocks us | 23:35 |
| sdague | anyway, put a few more thoughts back on the list | 23:36 |
| harlowja | k | 23:36 |
| jeblair | sdague, harlowja: we can only do two-way gating if it's part of openstack. | 23:36 |
| sdague | lets get over the 0.8 hump first | 23:36 |
| sdague | jeblair: agreed | 23:36 |
| jeblair | cool | 23:37 |
| * jeblair -> SFO | 23:37 | |
| sdague | which means thinking about the fact that if you require SQLA you have to be part of oslo | 23:37 |
| *** dstanek has joined #openstack-infra | 23:37 | |
| harlowja | sdague ya,, it just feels weird though | 23:37 |
| sdague | because we can't have things managing tables in an openstack environment that aren't | 23:38 |
| sdague | harlowja: I agree | 23:38 |
| reed | jeblair, should have known better :) | 23:38 |
| sdague | but, the other direction is taskflow effectively dictates the sqla versions in all of openstack | 23:39 |
| harlowja | understandable, when oslo.db shows up on pypi then i expect taskflow to use it, but that doesn't exist | 23:39 |
| sdague | the alternative would be taskflow not doing the actual storage itself, and having some kind of storage callback mechanism | 23:39 |
| harlowja | then oslo.db can dictate that version | 23:39 |
| *** jerryz has joined #openstack-infra | 23:39 | |
| harlowja | sdague so the storage is already plugabble (via these entrypoints), just one of the defaults is sqlalchemy | 23:40 |
| sdague | harlowja: so if there was taskflow and taskflow-sqla, and we didn't need taskflow-sqla in the gate, then we'd be fine as well | 23:40 |
| sdague | as with all complex systems, we often figure out what works over time by figuring out what doesn't work :) | 23:41 |
| harlowja | agreed | 23:41 |
| harlowja | entrypoints make this more hardier to figure out what the right way should be | 23:41 |
| sdague | agreed | 23:42 |
| beekneemech | morganfainberg: My recheck seems to be proceeding normally now, so I think you're good to go ahead and mass recheck. | 23:42 |
| morganfainberg | beekneemech, thank you much for the confirmation | 23:42 |
| harlowja | since people using taskflow can provide there own, or not use the sqlalchemy storage thing, but then people that do use it want to know which version to use... | 23:42 |
| *** chrisstreeter_ has quit IRC | 23:43 | |
| harlowja | like we could move it to test-requirements.txt (similar what happened with eventlet in taskflow) and then those that use it have to know to include sqlalchemy in there requirements | 23:43 |
| harlowja | dependencies sux, lol | 23:43 |
| *** ruhe is now known as _ruhe | 23:44 | |
| *** vipul is now known as vipul-away | 23:46 | |
| sdague | harlowja: well what if you split the package and made the sqla part a second install? | 23:46 |
| harlowja | possible | 23:46 |
| sdague | I guess the question is, does cinder need the sqla part | 23:47 |
| sdague | ? | 23:47 |
| harlowja | not yet | 23:47 |
| sdague | but we expect it to? | 23:47 |
| harlowja | ya | 23:47 |
| harlowja | responding to ML | 23:47 |
| harlowja | :-P | 23:47 |
| sdague | cool | 23:47 |
| *** julim has quit IRC | 23:47 | |
| sdague | yeh ML is good, as I expect we'll need to come back to this thread later | 23:47 |
| harlowja | def | 23:48 |
| harlowja | doubt this will be the first time ;) | 23:48 |
| sdague | or the last :) | 23:49 |
| openstackgerrit | Michael Krotscheck proposed a change to openstack-infra/config: Add storyboard-webclient to gerrit https://review.openstack.org/64761 | 23:51 |
| markmcclain | how long is the wait before VMs are built with the most current version of a client lib appear? | 23:51 |
| fungi | wow, i go to dinner and... *bam* scrollback | 23:52 |
| harlowja | to much chit-chat, ha | 23:54 |
| *** rnirmal has quit IRC | 23:54 | |
| sdague | markmcclain: pypi lib, or client? | 23:54 |
| sdague | we do integrated gate with the git trees of the client libraries | 23:55 |
| markmcclain | hmmm | 23:55 |
| markmcclain | https://jenkins01.openstack.org/job/gate-neutron-python27/5197/console | 23:55 |
| markmcclain | show 2.3.2 version of neutronclient lib | 23:55 |
| markmcclain | 2.3.3 is most current | 23:55 |
| sdague | ok, so unit tests are different | 23:56 |
| sdague | fungi: would know | 23:56 |
| markmcclain | or is this a case where I need to bump the min in the global reqs too for the cliff bug? | 23:57 |
| *** dimsum has quit IRC | 23:57 | |
| sdague | let's let fungi weigh in first | 23:59 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!