| clarkb | jeblair: I think what I am tripping up on is what I said earlier about qemu-img-options | 00:00 |
|---|---|---|
| clarkb | jeblair: because it is entirely possible that those will be different between clouds | 00:00 |
| jeblair | clarkb: yeah, so i'm suggesting they ride with image types | 00:00 |
| clarkb | jeblair: so you end up with same format, same image, multiple times | 00:00 |
| jeblair | clarkb: only if that's necessary... | 00:01 |
| jeblair | clarkb: let's forget about the tripleo cloud for a moment so i can stop saying "except for the tripleo cloud" and we'll get back to that in a minute... | 00:01 |
| clarkb | I'm not sure how tripleo cloud got involved fwiw | 00:01 |
| jeblair | clarkb: okay, then i've been really unclear :) | 00:01 |
| jeblair | clarkb: continue to forget about it for the moment then, and i'll try splainin again | 00:02 |
| greghaynes | clarkb: we can also make dib accept multipe --image-options | 00:02 |
| greghaynes | clarkb: since its simlar to -t | 00:02 |
| jeblair | clarkb: we won't build any more images than necessary. we need to build 'devstack-trusty' for rackspace and hp, because they use different image formats | 00:02 |
| jeblair | clarkb: so regardless, we will end up with 2 devstack-trusty files on our filesystem, yeah? one in vhd and one in qcow2. | 00:02 |
| *** annegentle has quit IRC | 00:02 | |
| clarkb | jeblair: yup | 00:03 |
| jeblair | clarkb: so specifically, we will _not_ end up with 16, or even 8 devstack-trusty files like we currently have devstack-trusty snapshot images. | 00:03 |
| *** achanda_ has joined #openstack-infra | 00:03 | |
| *** patrickeast has quit IRC | 00:04 | |
| jeblair | clarkb: because i'm not saying build one for each pseudo-provider, i'm saying, found out what types we need to build and build those types automatically. but still per-diskimage. | 00:04 |
| *** yamamoto has joined #openstack-infra | 00:04 | |
| jeblair | s/found/find/ | 00:04 |
| jeblair | clarkb: so for us, we _need_ both of those files, but nothing else. and with either of my proposals, we will get exactly only those two files. | 00:05 |
| jeblair | clarkb: make sense so far? | 00:06 |
| *** baoli_ has quit IRC | 00:06 | |
| *** achanda has quit IRC | 00:07 | |
| clarkb | jeblair: mostly but only if you consider image-tpe | 00:07 |
| clarkb | jeblair: I think that breaks down when you add in the other knobs that are semi related | 00:07 |
| jeblair | clarkb: yes, that's important. i'm considering image-type and image options as a unit. | 00:07 |
| *** dboik_ has quit IRC | 00:07 | |
| jeblair | clarkb: so if provider A uses qcow 0.10 and provider B uses qcow 0.20, and provider C uses vhd, you will get 3 files for each image | 00:08 |
| *** achanda_ has quit IRC | 00:08 | |
| *** ddieterly has joined #openstack-infra | 00:08 | |
| jeblair | clarkb: because an image type is really an image-type-with-options. so qcow0.10 is not the same as qcow0.20 | 00:08 |
| *** tonytan4ever has quit IRC | 00:08 | |
| clarkb | right ok | 00:08 |
| greghaynes | Why dont we just make it a boolean use_old_qcow and if any providor wants that for a qcow we just make the qcow in 0.10 format | 00:08 |
| *** yamamoto has quit IRC | 00:09 | |
| jeblair | greghaynes: it's functionally the same i think, but allowing options makes it more future-proof | 00:09 |
| jeblair | clarkb: k, so if we add tripleo back in, we see the difference between my first and second proposals | 00:09 |
| *** sigmavirus24 is now known as sigmavirus24_awa | 00:09 | |
| greghaynes | The thinking there is that youd get two images in that case, not 3 | 00:09 |
| jeblair | greghaynes: no, we need 3 because one provider uses old and one new, and one something else | 00:09 |
| greghaynes | oh, the one who uses the new doesnt work with the old? | 00:10 |
| jeblair | greghaynes: beats me. this stuff is all insane and i'd rather just pass it through (like we do now). i agree though if we think that's feasible to support, we could enhance the proposal to cover it. | 00:11 |
| greghaynes | AIUI you could use 0.10 everywhere and be fine, but 0.20 did not | 00:11 |
| jeblair | anyway, back to tripleo | 00:11 |
| greghaynes | heh, ok | 00:11 |
| clarkb | jeblair: sorry still don't follow on tripleo | 00:11 |
| clarkb | jeblair: we shouldn't need to do anything special there right? | 00:11 |
| clarkb | jeblair: oh wait I see | 00:11 |
| jeblair | clarkb: my first was super simple, it was just "iterate over all providers, and collect image-type-plus-options into a set, and build that set of types for all images" | 00:11 |
| clarkb | we would only qcow2 that image because rax doesn't use it | 00:11 |
| jeblair | clarkb: right. so my first proposal would have us building vhd tripleo images which we don't need. | 00:12 |
| *** carl_baldwin has quit IRC | 00:12 | |
| jeblair | clarkb: the second actually says "examine each provider _and_ which images it uses, and collect the union of all file types for diskimages used by that provider" | 00:12 |
| *** weshay has joined #openstack-infra | 00:13 | |
| jeblair | clarkb: so with the first, we build a few extra image files. i don't think it will kill us because it shouldn't be that many, but yeah, it is inefficient. | 00:13 |
| openstackgerrit | Clark Boylan proposed openstack-infra/nodepool: Support per provider image type for image uploads https://review.openstack.org/126747 | 00:14 |
| jeblair | clarkb: however, it does mean that you can add an image to hp first, then add an image to rax later and already have the file built | 00:14 |
| jeblair | clarkb: if we went with the second thing, nodepool would have to say "oh, i need to rebuild or convert this qcow image to vhd because it's now being used by a provider that requires vhd" | 00:14 |
| nibalizer | asselin_: wow okay, thats not at all what I expected | 00:15 |
| *** baoli has joined #openstack-infra | 00:15 | |
| jeblair | clarkb: i don't know off the top of my head how complicated that logic would be. if it's not terrible, then that's the better way to go. | 00:15 |
| nibalizer | but ill open my mind a bit :) | 00:15 |
| clarkb | jeblair we dont want the first thing because disk already hurts us | 00:15 |
| clarkb | jeblair and we need something like 16gb ti build a vhd image | 00:16 |
| asselin_ | nibalizer, good or bad? anyway....looking for better ways to test.... | 00:16 |
| clarkb | but new patchset is incomplete start at doing the second thing you describe | 00:16 |
| jeblair | nibalizer, asselin_: what's the topic of conversation? | 00:16 |
| jeblair | clarkb: to build or store? | 00:16 |
| clarkb | I have to go now but will pick that back up later | 00:16 |
| asselin_ | jeblair, https://review.openstack.org/169117 | 00:16 |
| jeblair | clarkb: ok, np. anyway, no argument there, the second way is better. just a tradeoff we can evaluate. :) | 00:17 |
| clarkb | jeblair to build because of the vhd utils multi step thing you basically end up with 4 copies of the image during build | 00:17 |
| nibalizer | I thouht we were gonna write a puppet module. I didn't expect to write shell scripts :P | 00:17 |
| clarkb | so 4*4GB == 16GB | 00:17 |
| asselin_ | nibalizer, I'm not a puppet expert by any means....learning as I go | 00:17 |
| nibalizer | as for the testing... its gonna be in shell or serversspec or beaker or envassert, and they're all pretty similar so :shrug: | 00:18 |
| jeblair | clarkb: yeah, i think storage space is the main thing here since diskimage builds are single threaded. | 00:18 |
| clarkb | jeblair its raw, qcow2, intermediate, vhd | 00:18 |
| *** sputnik13 has quit IRC | 00:18 | |
| greghaynes | clarkb: I dont think you end up with 4 copies in vhd | 00:18 |
| *** ibiris_away is now known as ibiris | 00:18 | |
| greghaynes | I think you get 2 | 00:18 |
| clarkb | greghaynes no its 4 total see above | 00:19 |
| *** unicell has quit IRC | 00:19 | |
| *** unicell1 has joined #openstack-infra | 00:19 | |
| greghaynes | yes, IIRC the vhd-util does an in-place modify | 00:19 |
| greghaynes | and then an mv | 00:19 |
| *** hdd has quit IRC | 00:19 | |
| greghaynes | I should re-test though | 00:19 |
| clarkb | yup I think our maths line up | 00:19 |
| *** Somay has quit IRC | 00:19 | |
| jeblair | clarkb: anyway, you need that space to build regardless; the advantage of option 2 over option 1 is storage space, not in build workspace | 00:20 |
| clarkb | ya I am just cranky we dont have 16gb currently | 00:20 |
| *** rossella_s has quit IRC | 00:20 | |
| jeblair | clarkb: we should add a cinder volume | 00:20 |
| clarkb | so will need new disk volime sooner than later | 00:20 |
| clarkb | yup | 00:20 |
| *** tsg has quit IRC | 00:20 | |
| *** rossella_s has joined #openstack-infra | 00:20 | |
| greghaynes | clarkb: ah, so I think the raw shouldnt be included | 00:22 |
| jesusaurus | jeblair: can i pick your brain for a minute about gertty? | 00:22 |
| greghaynes | clarkb: its destructive to the raw image which is why I have the sorting logic to run vhd last | 00:22 |
| jeblair | jesusaurus: yep | 00:22 |
| jeblair | jesusaurus: (also, i left a comment on 168495) | 00:22 |
| greghaynes | hrm, did that get removed | 00:22 |
| clarkb | greghaynes uh it shouldnt do that since we need raw | 00:22 |
| clarkb | and yes its not in dib | 00:22 |
| greghaynes | clarkb: yea, disreguard that, the sorting logic was for the raw patch | 00:23 |
| jesusaurus | jeblair: yeah, i saw the comment, thanks | 00:23 |
| jesusaurus | jeblair: im confused about how parser.py works and what p[0] represents | 00:23 |
| jesusaurus | jeblair: how is p_is_term() getting called? | 00:23 |
| jeblair | jesusaurus: so the first thing to know is that i didn't make that up. :) that's more or less the standard lex/yacc way of doing things; see: http://www.dabeaz.com/ply/ | 00:24 |
| jeblair | jesusaurus: but the way i think about it is that when the parser matches grammar described at the top of the function... | 00:25 |
| *** asettle has joined #openstack-infra | 00:25 | |
| jeblair | jesusaurus: (the _comments_ at the top of the function are actually the grammar -- they aren't just comments, the thing actually works by parsing the comments and building the grammar directly from them!) | 00:25 |
| jesusaurus | jeblair: oh wow, that's a little bit of magic there | 00:25 |
| jeblair | jesusaurus: it invokes that function. the arguments p[2]...p[n] are the parts after the ':' | 00:26 |
| jeblair | jesusaurus: p[1] is the part before the ':' (so the literal "is:" in this case) | 00:26 |
| *** dims has joined #openstack-infra | 00:27 | |
| jeblair | jesusaurus: and p[0] is the result | 00:27 |
| jeblair | jesusaurus: (or the 'production') | 00:27 |
| *** mtanino_ has joined #openstack-infra | 00:27 | |
| *** tiswanso has joined #openstack-infra | 00:28 | |
| *** ibiris is now known as ibiris_away | 00:28 | |
| jeblair | jesusaurus: so if you have "status:open AND is:watched" -- the "is:watched" produces a sql query for whether something is watched | 00:28 |
| jesusaurus | jeblair: ah, so putting the return value in p[0] is just part of how ply works? | 00:28 |
| jeblair | jesusaurus: same thing for status:open | 00:28 |
| jeblair | jesusaurus: and then it runs the function for "AND", which does a sql query operation that combines its two arguments and returns _that_ as p[0] | 00:29 |
| jeblair | jesusaurus: so that's how they get built up into one big sql query at the end | 00:29 |
| *** mtanino has quit IRC | 00:29 | |
| jeblair | jesusaurus: and yep, the whole p[0] thing is just the way ply works | 00:29 |
| *** mwagner_lap has joined #openstack-infra | 00:29 | |
| *** signed8b_ has joined #openstack-infra | 00:30 | |
| jesusaurus | jeblair: ok, im starting to see how the pieces fit together | 00:30 |
| jeblair | jesusaurus: yeah, take a look at some of the boolean things (and note that there's special handling in there since 'and' isn't necessary -- any two expressions next to each other are 'anded'). i think what they do with arguments are illustrative. | 00:32 |
| *** signed8bit has quit IRC | 00:32 | |
| jeblair | jesusaurus: and also, note the 'main' funciton in __init__.py, which is good for playing with things and testing out the result | 00:33 |
| jeblair | jesusaurus: i have to run now, but let me know how you get on | 00:33 |
| jesusaurus | jeblair: awesome, thanks | 00:33 |
| clarkb | jeblair oh we also need another dib change to support this fwiw | 00:36 |
| clarkb | but should be fine we already have a stack of those | 00:36 |
| greghaynes | clarkb: oh? | 00:37 |
| greghaynes | I am dib changes as a service lately | 00:37 |
| clarkb | the qemu ing options to avoid rebuilding inages for that | 00:37 |
| clarkb | not sure how to model that yet | 00:38 |
| greghaynes | ah, well that gets at my question of why not just make all images in 0.10 format | 00:38 |
| greghaynes | since IIRC the issue was that some things didnt recognize 0.20 | 00:38 |
| *** tiswanso has quit IRC | 00:38 | |
| clarkb | not 1.0 byt there are other options too | 00:38 |
| greghaynes | ah | 00:38 |
| *** tiswanso has joined #openstack-infra | 00:39 | |
| clarkb | I mean I am mostly trying to expect the unexpected here | 00:39 |
| greghaynes | So weve been talking about making raw output format caching | 00:39 |
| clarkb | since thats been the tldr on this stuff | 00:39 |
| greghaynes | so if you want another format, and theres a raw image cached, wed just do a convert from that | 00:39 |
| *** yamamoto_ has joined #openstack-infra | 00:43 | |
| *** otter768 has joined #openstack-infra | 00:43 | |
| SpamapS | Hm, we should check that we don't rely on anything that pulls from code.google.com | 00:52 |
| SpamapS | http://google-opensource.blogspot.com/2015/03/farewell-to-google-code.html | 00:52 |
| greghaynes | yeep | 00:53 |
| clarkb | SpamapS we have a year and yes we have thibgs, gerrit sort of, yappi just off the top of my head | 00:53 |
| ianw | clarkb / fungi : see also https://github.com/eliben/pycparser/issues/72 ; that seems to isolate the memory usage to pycparser ... if there's a bug there is to be determined | 00:54 |
| *** ghostpl_ has joined #openstack-infra | 00:54 | |
| *** Longgeek has joined #openstack-infra | 00:54 | |
| *** YorikSar has quit IRC | 00:55 | |
| fungi | ianw: that's an excellent reduction | 00:56 |
| *** annegentle has joined #openstack-infra | 00:58 | |
| *** emagana has quit IRC | 01:01 | |
| *** rfolco has joined #openstack-infra | 01:02 | |
| *** ghostpl_ has quit IRC | 01:04 | |
| openstackgerrit | K Jonathan Harker proposed stackforge/gertty: Add is:watched to p_is_term() https://review.openstack.org/168495 | 01:06 |
| *** ChuckC has quit IRC | 01:08 | |
| *** imcsk8 has quit IRC | 01:12 | |
| *** baoli has quit IRC | 01:12 | |
| *** amotoki has joined #openstack-infra | 01:12 | |
| *** imcsk8 has joined #openstack-infra | 01:12 | |
| *** annegentle has quit IRC | 01:12 | |
| *** annegentle has joined #openstack-infra | 01:13 | |
| *** amotoki has quit IRC | 01:13 | |
| * ianw takes a lunch break and starts reducing some leeks for dinner ... much tastier type of reduction :) | 01:14 | |
| *** Longgeek has quit IRC | 01:17 | |
| *** Longgeek has joined #openstack-infra | 01:17 | |
| *** yamamoto_ has quit IRC | 01:17 | |
| anteaya | ianw: sounds yummy | 01:19 |
| *** asselin has joined #openstack-infra | 01:28 | |
| *** thingee has quit IRC | 01:30 | |
| *** aduarte has quit IRC | 01:30 | |
| *** unicell1 has quit IRC | 01:31 | |
| *** unicell has joined #openstack-infra | 01:31 | |
| *** kaisers has joined #openstack-infra | 01:33 | |
| *** ChuckC has joined #openstack-infra | 01:34 | |
| *** xyang1 has quit IRC | 01:35 | |
| *** kaisers1 has quit IRC | 01:35 | |
| openstackgerrit | Thai Tran proposed openstack/requirements: Updated incorrect name for angular lrdragndrop https://review.openstack.org/169132 | 01:38 |
| *** weshay has quit IRC | 01:38 | |
| *** stevemar has joined #openstack-infra | 01:39 | |
| *** jerryz_ has joined #openstack-infra | 01:39 | |
| *** baoli has joined #openstack-infra | 01:40 | |
| *** amotoki has joined #openstack-infra | 01:42 | |
| *** annegentle has quit IRC | 01:42 | |
| *** tqtran has quit IRC | 01:45 | |
| *** baoli has quit IRC | 01:47 | |
| *** amotoki has quit IRC | 01:47 | |
| *** yamahata has joined #openstack-infra | 01:47 | |
| *** amotoki has joined #openstack-infra | 01:47 | |
| *** carl_baldwin has joined #openstack-infra | 01:47 | |
| *** yamamoto_ has joined #openstack-infra | 01:49 | |
| *** asettle has quit IRC | 01:49 | |
| *** tsg has joined #openstack-infra | 01:49 | |
| *** unicell has quit IRC | 01:49 | |
| *** dboik has joined #openstack-infra | 01:50 | |
| *** amotoki has quit IRC | 01:52 | |
| *** sputnik13 has joined #openstack-infra | 02:04 | |
| *** sputnik13 has quit IRC | 02:06 | |
| *** sputnik13 has joined #openstack-infra | 02:06 | |
| *** dims has quit IRC | 02:07 | |
| *** dims has joined #openstack-infra | 02:08 | |
| *** Longgeek has quit IRC | 02:09 | |
| *** Longgeek_ has joined #openstack-infra | 02:09 | |
| *** sputnik13 has quit IRC | 02:11 | |
| *** sputnik13 has joined #openstack-infra | 02:12 | |
| *** mfink has left #openstack-infra | 02:12 | |
| *** dims has quit IRC | 02:14 | |
| *** carl_baldwin has quit IRC | 02:14 | |
| *** SumitNaiksatam has quit IRC | 02:14 | |
| *** annegentle has joined #openstack-infra | 02:15 | |
| *** marun has quit IRC | 02:16 | |
| *** baoli has joined #openstack-infra | 02:18 | |
| *** Longgeek_ has quit IRC | 02:20 | |
| *** ghostpl_ has joined #openstack-infra | 02:21 | |
| *** carl_baldwin has joined #openstack-infra | 02:21 | |
| *** Longgeek_ has joined #openstack-infra | 02:23 | |
| *** signed8b_ has quit IRC | 02:26 | |
| *** amotoki has joined #openstack-infra | 02:28 | |
| *** ghostpl_ has quit IRC | 02:28 | |
| *** otter768 has quit IRC | 02:30 | |
| *** carl_baldwin has quit IRC | 02:31 | |
| *** achanda has joined #openstack-infra | 02:34 | |
| *** sputnik13 has quit IRC | 02:35 | |
| *** garyh has quit IRC | 02:38 | |
| *** otter768 has joined #openstack-infra | 02:39 | |
| *** carl_baldwin has joined #openstack-infra | 02:41 | |
| *** amotoki has quit IRC | 02:43 | |
| *** crc32 has quit IRC | 02:44 | |
| *** achanda has quit IRC | 02:44 | |
| *** yamahata has quit IRC | 02:44 | |
| jgriffith | anybody looking at teh XStatic-angular-irdragndrop issue? | 02:44 |
| jgriffith | ie it appeasr the version referenced in setup has been removed :( | 02:45 |
| *** achanda has joined #openstack-infra | 02:45 | |
| jgriffith | Oh.. yep, it's here http://goo.gl/NZaSzc | 02:46 |
| *** miqui has joined #openstack-infra | 02:47 | |
| *** mtanino_ has quit IRC | 02:50 | |
| *** carl_baldwin has quit IRC | 02:51 | |
| openstackgerrit | John Griffith proposed openstack/requirements: Updated incorrect name for angular lrdragndrop https://review.openstack.org/169132 | 02:51 |
| jgriffith | WTH.. is that an L or an I LOL | 02:53 |
| *** dannywilson has joined #openstack-infra | 02:54 | |
| clarkb | l | 02:57 |
| *** Somay has joined #openstack-infra | 02:58 | |
| jgriffith | clarkb: so it's correct on pypi... but not our mirror | 02:58 |
| clarkb | may just need an update? I cant cgeck it now | 02:58 |
| *** hichihara has quit IRC | 02:59 | |
| *** dannywilson has quit IRC | 02:59 | |
| *** hichihara has joined #openstack-infra | 02:59 | |
| *** hichihara has quit IRC | 02:59 | |
| jgriffith | Cool, if that's it devstack should work again, and the global update will go in (although that's not so critical) | 03:00 |
| *** erlon has quit IRC | 03:01 | |
| *** Somay has quit IRC | 03:02 | |
| jgriffith | clarkb: well now I'm confused... Thai is the author of the package on PyPi but he submitted a patch with the wrong letter | 03:03 |
| *** marcusvrn1 has quit IRC | 03:06 | |
| *** marcusvrn has joined #openstack-infra | 03:06 | |
| *** ddieterly has quit IRC | 03:07 | |
| *** harlowja_ is now known as harlowja_away | 03:07 | |
| *** bhunter71 has quit IRC | 03:08 | |
| *** amotoki has joined #openstack-infra | 03:10 | |
| *** tobiash_ has joined #openstack-infra | 03:13 | |
| *** tobiash has quit IRC | 03:13 | |
| jgriffith | clarkb: well.. it's fixed now | 03:14 |
| jgriffith | clarkb: if you fixed it with super powers... thanks :) | 03:14 |
| jgriffith | clarkb: if it fixed itself..... thanks just the same :) | 03:14 |
| *** otter768 has quit IRC | 03:15 | |
| *** bknudson has quit IRC | 03:16 | |
| *** amotoki has quit IRC | 03:16 | |
| *** tsg has quit IRC | 03:19 | |
| *** unicell has joined #openstack-infra | 03:20 | |
| *** spzala has quit IRC | 03:22 | |
| *** pal has joined #openstack-infra | 03:25 | |
| *** esker has joined #openstack-infra | 03:25 | |
| *** unicell has quit IRC | 03:27 | |
| *** unicell has joined #openstack-infra | 03:28 | |
| *** hdd has joined #openstack-infra | 03:32 | |
| *** carl_baldwin has joined #openstack-infra | 03:32 | |
| *** sputnik13 has joined #openstack-infra | 03:33 | |
| *** ajmiller__ has joined #openstack-infra | 03:34 | |
| openstackgerrit | Sean M. Collins proposed openstack-infra/elastic-recheck: Refine the query for #1326813 to pick up more hits https://review.openstack.org/169155 | 03:35 |
| *** baoli has quit IRC | 03:36 | |
| *** ajmiller_ has quit IRC | 03:37 | |
| *** SumitNaiksatam has joined #openstack-infra | 03:38 | |
| *** esker has quit IRC | 03:42 | |
| *** hdd has quit IRC | 03:42 | |
| *** amotoki has joined #openstack-infra | 03:43 | |
| *** rfolco has quit IRC | 03:46 | |
| *** ayoung has quit IRC | 03:48 | |
| openstackgerrit | Sean M. Collins proposed openstack-infra/elastic-recheck: Add fingerprint for XStatic-Angular-Irdragndrop dependency failure https://review.openstack.org/169156 | 03:49 |
| openstackgerrit | Sean M. Collins proposed openstack-infra/elastic-recheck: Add fingerprint for XStatic-Angular-Irdragndrop dependency failure https://review.openstack.org/169156 | 03:51 |
| openstackgerrit | Sean M. Collins proposed openstack-infra/elastic-recheck: Add fingerprint for XStatic-Angular-Irdragndrop dependency failure https://review.openstack.org/169156 | 03:52 |
| *** asettle has joined #openstack-infra | 03:56 | |
| *** rfolco has joined #openstack-infra | 03:58 | |
| *** armax has quit IRC | 04:00 | |
| *** sputnik13 has quit IRC | 04:02 | |
| *** sputnik13 has joined #openstack-infra | 04:03 | |
| *** carl_baldwin has quit IRC | 04:05 | |
| *** ddieterly has joined #openstack-infra | 04:08 | |
| openstackgerrit | Spencer Krum proposed openstack-infra/system-config: Fully qualify puppetdb classes https://review.openstack.org/169157 | 04:09 |
| *** krtaylor has quit IRC | 04:11 | |
| *** ddieterly has quit IRC | 04:12 | |
| *** achanda has quit IRC | 04:12 | |
| *** ghostpl_ has joined #openstack-infra | 04:13 | |
| *** xnox has quit IRC | 04:13 | |
| *** krtaylor has joined #openstack-infra | 04:13 | |
| *** ghostpl_ has quit IRC | 04:17 | |
| *** achanda has joined #openstack-infra | 04:19 | |
| *** hdd has joined #openstack-infra | 04:19 | |
| *** sputnik13 has quit IRC | 04:20 | |
| *** sputnik13 has joined #openstack-infra | 04:21 | |
| *** achanda has quit IRC | 04:27 | |
| *** sputnik13 has quit IRC | 04:30 | |
| *** achanda has joined #openstack-infra | 04:30 | |
| *** aduarte has joined #openstack-infra | 04:31 | |
| *** sputnik13 has joined #openstack-infra | 04:31 | |
| *** oomichi_ has joined #openstack-infra | 04:33 | |
| *** Longgeek_ has quit IRC | 04:35 | |
| *** achanda has quit IRC | 04:40 | |
| *** annegentle has quit IRC | 04:40 | |
| *** achanda has joined #openstack-infra | 04:40 | |
| *** btully has quit IRC | 04:41 | |
| *** carl_baldwin has joined #openstack-infra | 04:41 | |
| *** Somay has joined #openstack-infra | 04:42 | |
| *** Longgeek has joined #openstack-infra | 04:42 | |
| *** annegentle has joined #openstack-infra | 04:46 | |
| *** annegentle has quit IRC | 04:49 | |
| *** achanda has quit IRC | 04:51 | |
| *** thingee has joined #openstack-infra | 04:54 | |
| *** ildikov has quit IRC | 04:55 | |
| *** sputnik13 has quit IRC | 04:55 | |
| *** mrmartin has joined #openstack-infra | 04:56 | |
| *** ibiris_away is now known as ibiris | 04:56 | |
| *** yfried has quit IRC | 05:07 | |
| *** ddieterly has joined #openstack-infra | 05:09 | |
| stevemar | something broke everything :( | 05:09 |
| stevemar | angular something or other | 05:09 |
| *** btully has joined #openstack-infra | 05:10 | |
| greghaynes | stevemar: XStatic-Angular-Irdragndrop? | 05:11 |
| stevemar | greghaynes, thats the one, is everything resolved now? | 05:11 |
| greghaynes | I dont believe so, I just had a failing test | 05:12 |
| greghaynes | but theres some patches up to fix - seems like the pacakge might have been yanked from pypi for some reason | 05:12 |
| greghaynes | or possibly a mirroring issue | 05:12 |
| stevemar | great timing, just before RC time :) | 05:13 |
| *** ddieterly has quit IRC | 05:13 | |
| greghaynes | yea, the package seems to be missing on pypi | 05:14 |
| *** carl_baldwin has quit IRC | 05:15 | |
| *** hdd has quit IRC | 05:15 | |
| *** otter768 has joined #openstack-infra | 05:16 | |
| *** ociuhandu has quit IRC | 05:16 | |
| dougwig | you guys aware of the 4 hour horizon requirements breakage? | 05:18 |
| dougwig | ahh yes, i see you are. | 05:19 |
| *** harlowja_at_home has joined #openstack-infra | 05:19 | |
| *** otter768 has quit IRC | 05:20 | |
| greghaynes | hrm, I notice the package name doesnt match the name we have in requirements | 05:22 |
| greghaynes | http://git.openstack.org/cgit/stackforge/xstatic-angular-lrdragndrop/tree/xstatic/pkg/angular_irdragndrop/__init__.py | 05:22 |
| *** achanda has joined #openstack-infra | 05:22 | |
| greghaynes | XStatic-IrDragNDrop vs XStatic-Angular-Irdragndrop | 05:22 |
| *** harlowja_at_home has quit IRC | 05:25 | |
| *** prmtl has joined #openstack-infra | 05:25 | |
| *** dulek has joined #openstack-infra | 05:29 | |
| *** oomichi_ has quit IRC | 05:29 | |
| greghaynes | oh my, it is actually uploaded as lrdragndrop | 05:35 |
| *** pal has quit IRC | 05:37 | |
| StevenK | Haha | 05:37 |
| greghaynes | ah, it appears that is the real name and we just added a req with a totally wrong name? | 05:40 |
| openstackgerrit | greghaynes proposed openstack/requirements: Updated incorrect name for angular lrdragndrop https://review.openstack.org/169132 | 05:42 |
| *** aduarte has quit IRC | 05:42 | |
| *** vigneshvar has joined #openstack-infra | 05:43 | |
| *** vigneshvar_ has joined #openstack-infra | 05:43 | |
| *** ajo has joined #openstack-infra | 05:44 | |
| *** yfried has joined #openstack-infra | 05:45 | |
| greghaynes | http://logs.openstack.org/66/147366/3/gate/gate-tempest-dsvm-full/08d97e7/logs/devstacklog.txt.gz#_2015-02-24_23_33_56_658 | 05:45 |
| greghaynes | so looks like that package existed once upon a time | 05:45 |
| *** reed has quit IRC | 05:47 | |
| *** ildikov has joined #openstack-infra | 05:47 | |
| *** markvoelker has quit IRC | 05:48 | |
| *** yfried_ has joined #openstack-infra | 05:49 | |
| *** yfried has quit IRC | 05:49 | |
| greghaynes | and https://pypi.python.org/pypi/XStatic-Angular-lrdragndrop/1.0.2.2 was just uploded today | 05:51 |
| *** annegentle has joined #openstack-infra | 05:58 | |
| *** pcrews_ has joined #openstack-infra | 05:59 | |
| *** pcrews has quit IRC | 05:59 | |
| greghaynes | and openstackci is not an owner on pypi | 05:59 |
| *** pcrews_ has quit IRC | 05:59 | |
| *** ghostpl_ has joined #openstack-infra | 06:01 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack-infra/project-config: Normalize projects.yaml https://review.openstack.org/169174 | 06:01 |
| *** annegentle has quit IRC | 06:02 | |
| *** mmedvede has quit IRC | 06:04 | |
| *** yfried_ is now known as yfried|afk | 06:06 | |
| *** pcrews has joined #openstack-infra | 06:06 | |
| *** achanda has quit IRC | 06:08 | |
| *** achanda has joined #openstack-infra | 06:08 | |
| *** thingee has quit IRC | 06:08 | |
| *** ghostpl_ has quit IRC | 06:09 | |
| *** ddieterly has joined #openstack-infra | 06:10 | |
| AJaeger | greghaynes, it got renamed last week | 06:11 |
| AJaeger | seems we missed the wrong pypi setup during review ;( | 06:11 |
| greghaynes | ah! | 06:12 |
| AJaeger | flaper87, could you merge the translations for glance, please? https://review.openstack.org/#/c/158570/ | 06:12 |
| *** pal has joined #openstack-infra | 06:12 | |
| AJaeger | greghaynes, change I38385017423de00fc91ec4fe4293c29c746dda07 | 06:12 |
| AJaeger | let me email the author... | 06:12 |
| *** ddieterly has quit IRC | 06:14 | |
| greghaynes | AJaeger: So, I wonder how that works with the actual module - its named -irdragndrop | 06:15 |
| greghaynes | are we going to uplaod a pacakge lrdragndrop with a modulre irdragndrop | 06:15 |
| greghaynes | actually, I guess thats whats already uploaded | 06:15 |
| AJaeger | greghaynes, that looks wrong and needs to be changed IMHO. | 06:15 |
| greghaynes | yep, its just going to break horizon | 06:15 |
| greghaynes | we might have to do an import guard in horizon, rename the module, remove import guard? | 06:16 |
| AJaeger | greghaynes, can you fix it? | 06:16 |
| AJaeger | greghaynes, sorry, I can't give good advise here... | 06:16 |
| greghaynes | heh, its np, just thinking out loud :) | 06:16 |
| *** markvoelker has joined #openstack-infra | 06:19 | |
| flaper87 | AJaeger: yup | 06:22 |
| AJaeger | thanks, flaper87 ! | 06:23 |
| *** yfried|afk is now known as yfried_ | 06:24 | |
| *** BharatK has joined #openstack-infra | 06:24 | |
| *** markvoelker has quit IRC | 06:26 | |
| *** tnovacik has joined #openstack-infra | 06:27 | |
| *** amotoki has quit IRC | 06:27 | |
| *** mrda is now known as mrda-away | 06:27 | |
| *** Hal has joined #openstack-infra | 06:29 | |
| *** scheuran has joined #openstack-infra | 06:29 | |
| *** Hal is now known as Guest81334 | 06:29 | |
| mikal | Hey, is there a simple way to see all the recent votes for a given gerrit user? | 06:31 |
| mikal | I'm trying to determine if the hyper-v ci for nova is as broken as I think it is? | 06:31 |
| *** nmagnezi has joined #openstack-infra | 06:31 | |
| AJaeger | mikal, let me dig out something... | 06:34 |
| mikal | Sure, just writing an email now and it would be nice to say if its passed at all in say the last 24 hours | 06:34 |
| AJaeger | https://review.openstack.org/#/q/reviewer:%22Oracle+ZFSSA+CI%22,n,z | 06:34 |
| mikal | AJaeger: can I search by the vote for that user? | 06:35 |
| mikal | https://review.openstack.org/#/q/project:openstack/nova+status:open+reviewer:%22Microsoft+Hyper-V+CI%22,n,z is what I have so far | 06:35 |
| AJaeger | http://stackalytics.com/report/driverlog contains hyper v | 06:36 |
| mikal | Ahhh, nice. Thanks. | 06:37 |
| AJaeger | mika, your URL should works and is the best I know for nova | 06:38 |
| mikal | AJaeger: yeah, it just doesn't filter for positive votes | 06:39 |
| mikal | But the driverlog thing is close enough, so that keeps me rolling | 06:39 |
| openstackgerrit | Merged openstack-infra/project-config: Normalize projects.yaml https://review.openstack.org/169174 | 06:39 |
| AJaeger | Ah, you look for positive - should be possible... | 06:39 |
| *** dboik has quit IRC | 06:40 | |
| AJaeger | mikal, btw. could you import translations for nova, please? https://review.openstack.org/#/c/162030/ | 06:40 |
| * mikal looks | 06:40 | |
| mikal | Dine | 06:40 |
| mikal | Done even | 06:40 |
| AJaeger | thanks! | 06:41 |
| mikal | NP | 06:41 |
| *** Guest81334 has quit IRC | 06:43 | |
| AJaeger | mikal, do you know the login name of the hyper v ci? | 06:43 |
| AJaeger | You should be able to add "label:Code-Review=1,aname" | 06:43 |
| mikal | I have the username, its "Microsoft Hyper-V CI" IIRC | 06:44 |
| AJaeger | but I can't get that one working with name | 06:44 |
| mikal | Oh, I see | 06:44 |
| mikal | We need the username, not the long name | 06:44 |
| AJaeger | mikal,exactly | 06:44 |
| mikal | That should be on the wiki IIRC | 06:44 |
| AJaeger | That'S my guess after the long name does not work. | 06:45 |
| mikal | hyper-v-ci it seems | 06:45 |
| mikal | https://wiki.openstack.org/wiki/ThirdPartySystems/Hyper-V_CI | 06:45 |
| mikal | Oh, that wiki page says "offline for maintenance" | 06:45 |
| *** deepakcs has joined #openstack-infra | 06:45 | |
| mikal | Oh, ignore me | 06:45 |
| mikal | That was at the start of the month | 06:45 |
| mikal | Its nearly April now apparenlty | 06:45 |
| AJaeger | Use: reviewer:"Microsoft Hyper-V CI" label:Code-Review=1,hyper-v_ci@microsoft.com | 06:46 |
| AJaeger | but that looks wrong -1 does not get anything either ;( | 06:46 |
| mikal | Its ok, the driverlog thing is enough I think | 06:47 |
| *** YorikSar has joined #openstack-infra | 06:47 | |
| *** Somay has quit IRC | 06:47 | |
| *** jamielennox is now known as jamielennox|away | 06:49 | |
| AJaeger | mikal, ok - otherwise, you need to ask for more help during US time, hope there'll be somebody around with better gerrit search skills | 06:49 |
| *** Somay has joined #openstack-infra | 06:50 | |
| AJaeger | mikal, this works: label:Code-Review=0,user:hyper-v_ci@microsoft.com | 06:50 |
| AJaeger | Change 0 to 1.. | 06:50 |
| *** marun has joined #openstack-infra | 06:51 | |
| AJaeger | Note this matches only most recent review | 06:51 |
| AJaeger | mikal, ignore me - it should give a result but does not - I'm giving up. | 06:51 |
| mikal | Good plan | 06:52 |
| *** amotoki has joined #openstack-infra | 06:53 | |
| *** sergsh has joined #openstack-infra | 06:53 | |
| AJaeger | mikal, I saw your email - asking is a good idea ;) | 06:54 |
| *** marun has quit IRC | 06:55 | |
| *** sdake has quit IRC | 06:56 | |
| *** tkelsey has joined #openstack-infra | 06:58 | |
| *** annegentle has joined #openstack-infra | 06:58 | |
| openstackgerrit | Jan Hruban proposed openstack-infra/jenkins-job-builder: Add support for Display Upstream Changes Plugin https://review.openstack.org/169193 | 06:59 |
| *** asettle has quit IRC | 07:01 | |
| flaper87 | https://review.openstack.org/#/c/169132/ :( | 07:01 |
| *** amotoki has quit IRC | 07:01 | |
| *** amotoki has joined #openstack-infra | 07:02 | |
| *** markus_z has joined #openstack-infra | 07:02 | |
| *** jyuso has quit IRC | 07:03 | |
| AJaeger | flaper87, see also change I38385017423de00fc91ec4fe4293c29c746dda07 | 07:04 |
| *** annegentle has quit IRC | 07:04 | |
| *** jyuso1 has joined #openstack-infra | 07:04 | |
| *** yamamoto_ has quit IRC | 07:04 | |
| *** yfried__ has joined #openstack-infra | 07:04 | |
| flaper87 | AJaeger: thank you! | 07:04 |
| *** liusheng has joined #openstack-infra | 07:04 | |
| *** achanda has quit IRC | 07:05 | |
| *** tkelsey has quit IRC | 07:05 | |
| flaper87 | AJaeger: what's the next step there? Do we just have to wait until the new version is uploaded? | 07:05 |
| AJaeger | flaper87, I don't know ;(, let's tell the author. I reached out and told him already about the missing openstackci but not more. greghaynes was bringing this up one hour ago as well... | 07:06 |
| *** yfried has joined #openstack-infra | 07:06 | |
| AJaeger | flaper87, sorry, can't help further with this | 07:07 |
| *** ajo_ has joined #openstack-infra | 07:07 | |
| flaper87 | AJaeger: :( ok, I guess I'll just have to wait | 07:07 |
| flaper87 | AJaeger: no worries | 07:07 |
| *** yfried_ has quit IRC | 07:08 | |
| *** ZZelle has quit IRC | 07:08 | |
| *** yfried__ has quit IRC | 07:09 | |
| *** ajo has quit IRC | 07:10 | |
| *** ddieterly has joined #openstack-infra | 07:10 | |
| AJaeger | flaper87, now I see also the failures ;( | 07:13 |
| AJaeger | flaper87, greghaynes: Might be best to send an email to openstack-dev and copy Thai Tran | 07:14 |
| *** teran has joined #openstack-infra | 07:14 | |
| *** ddieterly has quit IRC | 07:15 | |
| *** stevemar has quit IRC | 07:16 | |
| *** otter768 has joined #openstack-infra | 07:16 | |
| *** vigneshvar has quit IRC | 07:17 | |
| *** vigneshvar_ has quit IRC | 07:17 | |
| *** ala_ has joined #openstack-infra | 07:19 | |
| *** mpavone has joined #openstack-infra | 07:21 | |
| *** otter768 has quit IRC | 07:21 | |
| *** markvoelker has joined #openstack-infra | 07:22 | |
| *** rfolco has quit IRC | 07:24 | |
| *** MaxV has joined #openstack-infra | 07:25 | |
| *** arxcruz has joined #openstack-infra | 07:26 | |
| *** markvoelker has quit IRC | 07:27 | |
| AJaeger | flaper87, greghaynes: For all the xstatic packages: They are not uploaded via CI ;( | 07:28 |
| AJaeger | That's why reviewing did not notice it - there were no publish-to-pypi jobs | 07:29 |
| *** nilasae has joined #openstack-infra | 07:29 | |
| AJaeger | Still, the commit message for the rename should have raised some questions but didn't for me ;( | 07:29 |
| *** MaxV has quit IRC | 07:30 | |
| *** MaxV has joined #openstack-infra | 07:30 | |
| greghaynes | AJaeger: The old named package was uploaded, see http://logs.openstack.org/66/147366/3/gate/gate-tempest-dsvm-full/08d97e7/logs/devstacklog.txt.gz#_2015-02-24_23_33_56_658 | 07:30 |
| greghaynes | I think the author must have deleted it? | 07:30 |
| AJaeger | greghaynes, might have been deleted - yes. It was uploaded manually, not via OpenStack CI. | 07:31 |
| greghaynes | yep | 07:31 |
| greghaynes | and same with the new one | 07:31 |
| *** mpaolino has joined #openstack-infra | 07:32 | |
| *** liusheng has quit IRC | 07:33 | |
| * greghaynes needs to sleep, way past EOD | 07:33 | |
| *** liusheng has joined #openstack-infra | 07:34 | |
| AJaeger | greghaynes, good night! Here the new day is already over 9hours old ;) | 07:34 |
| *** MaxV has quit IRC | 07:35 | |
| StevenK | AJaeger: Heh, not far enough in the future. The day is over 18 hours old here. :-) | 07:35 |
| AJaeger | StevenK :) | 07:36 |
| *** rfolco has joined #openstack-infra | 07:36 | |
| *** jlanoux has joined #openstack-infra | 07:37 | |
| *** woodster_ has quit IRC | 07:40 | |
| *** jistr has joined #openstack-infra | 07:43 | |
| *** Somay has quit IRC | 07:52 | |
| *** ghostpl_ has joined #openstack-infra | 07:53 | |
| *** Hal has joined #openstack-infra | 07:55 | |
| *** btully has quit IRC | 07:55 | |
| *** Hal is now known as Guest22614 | 07:55 | |
| *** YorikSar has quit IRC | 07:56 | |
| *** yamamoto_ has joined #openstack-infra | 07:57 | |
| *** AJaeger has quit IRC | 07:57 | |
| *** ghostpl_ has quit IRC | 07:58 | |
| *** yamamoto_ has quit IRC | 07:59 | |
| *** annegentle has joined #openstack-infra | 07:59 | |
| *** Guest22614 has quit IRC | 08:00 | |
| *** MaxV has joined #openstack-infra | 08:00 | |
| *** Murad has quit IRC | 08:01 | |
| *** notnownikki has joined #openstack-infra | 08:01 | |
| *** hashar has joined #openstack-infra | 08:03 | |
| *** liusheng has quit IRC | 08:03 | |
| *** rdopiera has joined #openstack-infra | 08:04 | |
| rdopiera | hi guys, we are having some problem with the global-requirements pip cache here, it seems: http://logs.openstack.org/23/167323/10/check/gate-horizon-pep8/97e8b53/console.html | 08:04 |
| *** zz_johnthetubagu is now known as johnthetubaguy | 08:04 | |
| *** liusheng has joined #openstack-infra | 08:04 | |
| rdopiera | Could not install requirement XStatic-Angular-Irdragndrop>=1.0.2.1 (from -r /home/jenkins/workspace/gate-horizon-pep8/requirements.txt (line 45)) because of error 404 Client Error: Not Found | 08:04 |
| *** annegentle has quit IRC | 08:05 | |
| *** AJaeger has joined #openstack-infra | 08:09 | |
| *** AJaeger has quit IRC | 08:09 | |
| *** AJaeger has joined #openstack-infra | 08:09 | |
| *** ddieterly has joined #openstack-infra | 08:11 | |
| *** dizquierdo has joined #openstack-infra | 08:11 | |
| *** yfried_ has joined #openstack-infra | 08:11 | |
| *** prmtl has quit IRC | 08:13 | |
| *** yfried has quit IRC | 08:15 | |
| *** ddieterly has quit IRC | 08:15 | |
| *** prmtl has joined #openstack-infra | 08:15 | |
| *** yfried__ has joined #openstack-infra | 08:16 | |
| *** yfried_ has quit IRC | 08:16 | |
| *** markvoelker has joined #openstack-infra | 08:23 | |
| openstackgerrit | Julien Danjou proposed openstack-infra/project-config: Gnocchi: do not check requirements https://review.openstack.org/169210 | 08:25 |
| *** markvoelker has quit IRC | 08:27 | |
| cschwede | rdopiera: i saw the same error on two jobs, added a comment to https://bugs.launchpad.net/openstack-ci/+bug/1334550 | 08:29 |
| openstack | Launchpad bug 1334550 in OpenStack-Gate "Could not find any downloads that satisfy the requirement X" [Low,Fix released] | 08:29 |
| *** derekh has joined #openstack-infra | 08:29 | |
| *** e0ne has joined #openstack-infra | 08:29 | |
| *** YorikSar has joined #openstack-infra | 08:30 | |
| *** ssam2 has joined #openstack-infra | 08:33 | |
| ttx | AJaeger: should we push an alert ? | 08:37 |
| AJaeger | ttx, might be a nice idea. flaper87, what do you think? | 08:38 |
| ttx | Looks like it's completely stuck | 08:38 |
| flaper87 | ttx: AJaeger +2A for the alert | 08:38 |
| AJaeger | ttx, is really everything failing? Then let's raise an alert and sent an email to openstack-dev | 08:38 |
| flaper87 | it's *completely* stuck | 08:38 |
| AJaeger | flaper87 and ttx, will either of you sent an email to openstack-dev, please? | 08:39 |
| AJaeger | flaper87, we can still merge manuals and infra ;) | 08:39 |
| ttx | flaper87: if you have a good understanding of the fail, I'll let you write the email | 08:39 |
| ttx | I can send the alert based on the language you put in the email | 08:40 |
| flaper87 | ok, I can send it | 08:40 |
| flaper87 | lemme gather the required info | 08:41 |
| AJaeger | flaper87, ttx, there was some discussion on #openstack-horizon as well, see http://eavesdrop.openstack.org/irclogs/%23openstack-horizon/%23openstack-horizon.2015-03-31.log | 08:42 |
| ttx | Ifear that fixing it will require some direct monkeying that is not available for us mortals | 08:43 |
| *** Hal has joined #openstack-infra | 08:43 | |
| AJaeger | ttx, yeah | 08:43 |
| *** Hal is now known as Guest61703 | 08:44 | |
| ttx | SergeyLukjanov could help but not sure he knows what lever to play with | 08:44 |
| *** dtantsur|afk is now known as dtantsur | 08:44 | |
| *** e0ne has quit IRC | 08:45 | |
| *** ibiris is now known as ibiris_away | 08:45 | |
| AJaeger | Agreed, we need a plan on how to fix this and I fear that needs some expert knowledge from our US fire fighter team ;) | 08:45 |
| *** pelix has joined #openstack-infra | 08:46 | |
| flaper87 | ttx: AJaeger http://paste.openstack.org/show/bnrQTSuJ4H82e8AgCMq4/ | 08:48 |
| ttx | flaper87: looks good | 08:50 |
| flaper87 | sending | 08:50 |
| ttx | I propose alert "CI Check/Gate pipelines currently stuck due to a bad dependency creeping in the system. No need to recheck your patches at the moment." | 08:51 |
| ttx | does that sound good? | 08:51 |
| flaper87 | ttx: it does | 08:51 |
| * flaper87 will use it as subject | 08:51 | |
| flaper87 | :P | 08:51 |
| ttx | #status alert CI Check/Gate pipelines currently stuck due to a bad dependency creeping in the system. No need to recheck your patches at the moment. | 08:51 |
| openstackstatus | ttx: sending alert | 08:51 |
| -openstackstatus- NOTICE: CI Check/Gate pipelines currently stuck due to a bad dependency creeping in the system. No need to recheck your patches at the moment. | 08:53 | |
| *** ChanServ changes topic to "CI Check/Gate pipelines currently stuck due to a bad dependency creeping in the system. No need to recheck your patches at the moment." | 08:53 | |
| *** ghostpl_ has joined #openstack-infra | 08:54 | |
| *** ibiris_away is now known as ibiris | 08:54 | |
| anteaya | ttx thanks | 08:54 |
| ttx | now we can redirect questions to /topic :) | 08:55 |
| openstackstatus | ttx: finished sending alert | 08:56 |
| openstackgerrit | Fabien Boucher proposed openstack-infra/system-config: Move some hardcoded values as parameters for zuul manifests https://review.openstack.org/168330 | 08:56 |
| *** ociuhandu has joined #openstack-infra | 08:59 | |
| *** ghostpl_ has quit IRC | 08:59 | |
| anteaya | ttx: ha ha ha | 08:59 |
| * anteaya goes back to bed | 09:00 | |
| *** tkelsey has joined #openstack-infra | 09:00 | |
| *** annegentle has joined #openstack-infra | 09:00 | |
| *** ZZelle has joined #openstack-infra | 09:04 | |
| *** ghostpl_ has joined #openstack-infra | 09:04 | |
| *** yfried__ is now known as yfried|lunch | 09:05 | |
| *** annegentle has quit IRC | 09:06 | |
| *** ildikov has quit IRC | 09:06 | |
| *** ghostpl_ has quit IRC | 09:09 | |
| *** afazekas has joined #openstack-infra | 09:11 | |
| *** yamamoto has joined #openstack-infra | 09:12 | |
| *** ddieterly has joined #openstack-infra | 09:12 | |
| *** yfried|lunch is now known as yfried|afk | 09:15 | |
| *** Longgeek has quit IRC | 09:16 | |
| *** ddieterly has quit IRC | 09:16 | |
| *** nmagnezi_ has joined #openstack-infra | 09:17 | |
| *** nmagnezi has quit IRC | 09:17 | |
| *** otter768 has joined #openstack-infra | 09:17 | |
| *** mpaolino has quit IRC | 09:18 | |
| *** ildikov has joined #openstack-infra | 09:19 | |
| *** amotoki_ has joined #openstack-infra | 09:20 | |
| *** otter768 has quit IRC | 09:22 | |
| *** ajmiller_ has joined #openstack-infra | 09:23 | |
| *** markvoelker has joined #openstack-infra | 09:23 | |
| *** ajmiller__ has quit IRC | 09:26 | |
| openstackgerrit | yolanda.robla proposed openstack-infra/nodepool: Independent provider manager for polling tasks https://review.openstack.org/169235 | 09:27 |
| *** markvoelker has quit IRC | 09:28 | |
| openstackgerrit | Rob Cresswell proposed openstack/requirements: Updated incorrect name for angular lrdragndrop https://review.openstack.org/169132 | 09:28 |
| *** robcresswell has joined #openstack-infra | 09:29 | |
| *** yfried|afk is now known as yfried|lunch | 09:36 | |
| odyssey4me | Any chance we can get https://review.openstack.org/#/c/168531/ reviewed so that we can fix the stackforge packaging configuration for angular lrdragndrop ? | 09:36 |
| odyssey4me | I have the patch ready, but it's being rejected. | 09:36 |
| *** punith has joined #openstack-infra | 09:38 | |
| odyssey4me | AJaeger ^ | 09:38 |
| *** e0ne has joined #openstack-infra | 09:39 | |
| *** ghostpl_ has joined #openstack-infra | 09:41 | |
| *** jamielennox|away is now known as jamielennox | 09:41 | |
| robcresswell | Just been chasing this up on the Horizon side, but I'm not terribly familiar with the packaging workflow: https://bugs.launchpad.net/horizon/+bug/1438543 | 09:41 |
| openstack | Launchpad bug 1438543 in OpenStack Dashboard (Horizon) "wrong package name 'XStatic-Angular-Irdragndrop' in horizon/requirements.txt" [Critical,Confirmed] - Assigned to Rob Cresswell (robcresswell) | 09:41 |
| *** ajo_ is now known as ajo | 09:42 | |
| *** jlibosva has joined #openstack-infra | 09:46 | |
| *** yfried|lunch is now known as yfried|afk | 09:46 | |
| *** jamielennox is now known as jamielennox|away | 09:47 | |
| odyssey4me | robcresswell yeah, I see so - it would seem that the packaging was originally setup in https://github.com/stackforge/xstatic-angular-lrdragndrop | 09:47 |
| odyssey4me | that repo was recently renamed (implied in https://review.openstack.org/#/c/168531/ ) | 09:47 |
| odyssey4me | so I have a patch for that repo ready, but gerrit is rejecting the patch - very possibly because there is no .gitreview present | 09:48 |
| robcresswell | odyssey4me: Rename is here: https://review.openstack.org/#/c/167798/ | 09:50 |
| robcresswell | odyssey4me: There are patches out to rename the Horizon imports and update global-reqs. Do they not solve the issues? | 09:51 |
| odyssey4me | robcresswell You mean https://review.openstack.org/169132 and https://review.openstack.org/169133 ? | 09:51 |
| robcresswell | odyssey4me: Indeed. Just trying to follow as best I can. | 09:51 |
| *** cdent has joined #openstack-infra | 09:52 | |
| *** ildikov has quit IRC | 09:54 | |
| *** rlandy has joined #openstack-infra | 09:54 | |
| rdopiera | I don't think those patches will merge with dvsm failing... | 09:55 |
| rdopiera | odyssey4me: you don't need the .gitreview file | 09:56 |
| *** liusheng has quit IRC | 10:00 | |
| *** yamamoto has quit IRC | 10:01 | |
| *** liusheng has joined #openstack-infra | 10:01 | |
| *** Guest61703 has quit IRC | 10:01 | |
| *** annegentle has joined #openstack-infra | 10:01 | |
| AJaeger | odyssey4me, 168531 is a patch for lrdgragndrop, nothing I can do | 10:01 |
| AJaeger | rdopiera, which patch is rejected? | 10:02 |
| odyssey4me | AJaeger ah, of course - let me track down the project cores | 10:03 |
| *** dims has joined #openstack-infra | 10:03 | |
| AJaeger | odyssey4me, https://review.openstack.org/#/admin/groups/385,members | 10:03 |
| openstackgerrit | Rob Cresswell proposed openstack/requirements: Updated incorrect name for angular lrdragndrop https://review.openstack.org/169132 | 10:03 |
| *** achanda has joined #openstack-infra | 10:06 | |
| *** annegentle has quit IRC | 10:06 | |
| *** ildikov has joined #openstack-infra | 10:09 | |
| *** jcoufal has joined #openstack-infra | 10:09 | |
| *** achanda has quit IRC | 10:10 | |
| *** ominakov has joined #openstack-infra | 10:13 | |
| *** ddieterly has joined #openstack-infra | 10:13 | |
| ekarlso | 4hmmms | 10:13 |
| ekarlso | doesn't devstack work with pgsql ? | 10:13 |
| *** ominakov has quit IRC | 10:13 | |
| ekarlso | 2015-03-31 10:11:10.249 | FATAL: no pg_hba.conf entry for host "192.168.27.100", user "root", database "template1", SSL off | 10:13 |
| *** dhritishikhar_ has joined #openstack-infra | 10:15 | |
| *** yfried|afk is now known as yfried|lunch | 10:16 | |
| *** Longgeek has joined #openstack-infra | 10:17 | |
| *** ddieterly has quit IRC | 10:17 | |
| *** yamamoto has joined #openstack-infra | 10:17 | |
| *** yfried|lunch is now known as yfried | 10:17 | |
| *** pc_m has joined #openstack-infra | 10:18 | |
| *** e0ne is now known as e0ne_ | 10:19 | |
| *** e0ne_ is now known as e0ne | 10:23 | |
| *** markvoelker has joined #openstack-infra | 10:24 | |
| *** btully has joined #openstack-infra | 10:25 | |
| rdopiera | AJaeger: the global-requirements.txt one: https://review.openstack.org/#/c/169132/ | 10:26 |
| *** dhritishikhar_ has quit IRC | 10:28 | |
| *** jlanoux_ has joined #openstack-infra | 10:28 | |
| *** markvoelker has quit IRC | 10:29 | |
| *** btully has quit IRC | 10:30 | |
| *** Longgeek has quit IRC | 10:30 | |
| ekarlso | so devstack is not tested with pgsql ? | 10:31 |
| *** jlanoux has quit IRC | 10:31 | |
| *** teran has quit IRC | 10:31 | |
| *** samueldmq has joined #openstack-infra | 10:36 | |
| samueldmq | morning | 10:36 |
| samueldmq | so we already have something to solve early in the morning? :-) | 10:36 |
| *** e0ne is now known as e0ne_ | 10:37 | |
| *** yamamoto has quit IRC | 10:37 | |
| *** ghostpl_ has quit IRC | 10:38 | |
| *** e0ne_ is now known as e0ne | 10:40 | |
| yfried | AJaeger: any news on gate status? | 10:41 |
| *** ghostpl_ has joined #openstack-infra | 10:42 | |
| *** jlanoux has joined #openstack-infra | 10:43 | |
| yfried | boris-42: see topic | 10:45 |
| yfried | AJaeger: is there a way to disable "recheck" temporarily? | 10:46 |
| *** jlanoux_ has quit IRC | 10:46 | |
| openstackgerrit | Marton Kiss proposed openstack-infra/system-config: Add backup scripts to ask.o.o site https://review.openstack.org/169260 | 10:49 |
| *** claudiub has joined #openstack-infra | 10:52 | |
| AJaeger | yfried, not sure about disabling ;( | 10:52 |
| AJaeger | rdopiera, do you know which changes we need? I've started creating an etherpad where we can collect all information: https://etherpad.openstack.org/p/lrdragndrop-stuck | 10:53 |
| *** xnox has joined #openstack-infra | 10:55 | |
| robcresswell | AJaeger: Those patches already listed seem key. I'm not sure if tests will have to be disabled to allow the merges though. | 10:56 |
| *** gmann_ has joined #openstack-infra | 10:57 | |
| *** gmann_ has left #openstack-infra | 10:57 | |
| AJaeger | robcresswell, thanks | 10:57 |
| robcresswell | tbh, the old package should not have been taken down before the update | 10:57 |
| AJaeger | ttx, flaper87 fungi, clarkb, jeblair: https://etherpad.openstack.org/p/lrdragndrop-stuck | 10:57 |
| flaper87 | AJaeger: looking | 10:59 |
| *** mmedvede has joined #openstack-infra | 11:00 | |
| * AJaeger is just trying to get facts together so that the infra admin team can quickly help... | 11:00 | |
| robcresswell | AJaeger: Just adding any useful info I can think of. | 11:01 |
| AJaeger | robcresswell, Thanks! | 11:01 |
| robcresswell | no problem :) | 11:01 |
| AJaeger | robcresswell, that's what we need right now - all information easily accessible... | 11:02 |
| *** pal has quit IRC | 11:02 | |
| *** annegentle has joined #openstack-infra | 11:02 | |
| robcresswell | AJaeger: Yeah, understood. I've updated both patches to link the relevant bugs/ dependents. But they are failing the tests; I wonder if the tests may need disabling for the global-req change to pass, then reenabled for the horizon change. | 11:03 |
| AJaeger | robcresswell, infra admins can approve without gating if really necessary. | 11:04 |
| AJaeger | I'm not an infra admin - so just help to collect data so that they can easier access the situation, they have far more knowledge here than I have... | 11:05 |
| robcresswell | AJaeger: Gotcha. I'm in the same boat. A little out of my depth, just trying to help out. | 11:06 |
| *** yfried is now known as yfried|afk | 11:06 | |
| *** ildikov has quit IRC | 11:06 | |
| *** BharatK has quit IRC | 11:07 | |
| *** annegentle has quit IRC | 11:07 | |
| AJaeger | robcresswell, thanks for your help | 11:07 |
| *** jcoufal has quit IRC | 11:08 | |
| *** ujuc has joined #openstack-infra | 11:08 | |
| *** jcoufal has joined #openstack-infra | 11:08 | |
| *** dmellado has quit IRC | 11:08 | |
| openstackgerrit | Flavio Percoco proposed openstack-infra/project-config: Install required services for Zaqar https://review.openstack.org/169269 | 11:10 |
| *** jcoufal has quit IRC | 11:10 | |
| *** flaper87 has quit IRC | 11:10 | |
| *** yfried|afk is now known as yfried | 11:11 | |
| *** ddieterly has joined #openstack-infra | 11:13 | |
| *** teran has joined #openstack-infra | 11:15 | |
| *** tsufiev is now known as tsufiev_ | 11:15 | |
| *** Longgeek has joined #openstack-infra | 11:17 | |
| *** ddieterly has quit IRC | 11:18 | |
| *** otter768 has joined #openstack-infra | 11:18 | |
| *** _nadya_ has joined #openstack-infra | 11:19 | |
| *** teran has quit IRC | 11:19 | |
| *** Hal has joined #openstack-infra | 11:20 | |
| *** Hal is now known as Guest47177 | 11:20 | |
| rdopiera | robcresswell: I wonder if I should just upload a package under the old name | 11:21 |
| rdopiera | robcresswell: to pypi | 11:21 |
| *** ssam2 has quit IRC | 11:21 | |
| *** ildikov has joined #openstack-infra | 11:22 | |
| *** flaper87 has joined #openstack-infra | 11:22 | |
| *** otter768 has quit IRC | 11:23 | |
| *** mpaolino has joined #openstack-infra | 11:24 | |
| *** pblaho has joined #openstack-infra | 11:24 | |
| *** markvoelker has joined #openstack-infra | 11:25 | |
| *** teran has joined #openstack-infra | 11:25 | |
| *** flaper87 has quit IRC | 11:26 | |
| *** flaper87 has joined #openstack-infra | 11:26 | |
| *** salv-orlando has quit IRC | 11:26 | |
| AJaeger | rdopiera, that'S one option out of the mess - the question is why was renaming needed then? | 11:26 |
| AJaeger | rdopiera, do you have permissions? | 11:26 |
| *** hogepodge has quit IRC | 11:27 | |
| robcresswell | rdopiera: That is probably the safest and most sensible solution. | 11:27 |
| *** sc68cal has quit IRC | 11:27 | |
| openstackgerrit | Joshua Hesketh proposed openstack-infra/zuul: Add support for 'connection' concept https://review.openstack.org/121528 | 11:27 |
| openstackgerrit | Joshua Hesketh proposed openstack-infra/zuul: Document the new connections in zuul https://review.openstack.org/169270 | 11:27 |
| *** sc68cal has joined #openstack-infra | 11:28 | |
| openstackgerrit | Merged openstack-infra/project-config: Move aiopcpu to nova check queue https://review.openstack.org/169043 | 11:29 |
| *** markvoelker has quit IRC | 11:29 | |
| *** changbl has quit IRC | 11:30 | |
| rdopiera | robcresswell: I just did that, and discovered another problem | 11:31 |
| rdopiera | robcresswell: the code on stackforge is two versions behind the code on pypi | 11:31 |
| AJaeger | rdopiera, please update the etherpad with what you did and learnt | 11:31 |
| AJaeger | rdopiera, this one? https://pypi.python.org/pypi/XStatic-IrDragNDrop/1.0.2.2 That misses the Angular in the name | 11:32 |
| *** jistr is now known as jistr|english | 11:32 | |
| *** jistr|english is now known as jistr|class | 11:33 | |
| rdopiera | AJaeger: I just did | 11:33 |
| rdopiera | argh | 11:33 |
| *** deepakcs has quit IRC | 11:33 | |
| *** ssam2 has joined #openstack-infra | 11:34 | |
| rdopiera | AJaeger: ok, fixed | 11:34 |
| AJaeger | rdopiera, do you know why the package was renamed? | 11:34 |
| robcresswell | AJaeger: from 'I' to 'l'? | 11:34 |
| AJaeger | robcresswell, yes | 11:35 |
| robcresswell | AJaeger: The original github repo is called lrdragndrop after the author | 11:35 |
| rdopiera | AJaeger: because the original project is named with l | 11:35 |
| robcresswell | at some point, someone saw the 'l' and though it was 'I' | 11:35 |
| rdopiera | AJaeger: and it was packager's mistake to use I | 11:35 |
| AJaeger | rdopiera, AH! Understood now. | 11:35 |
| rdopiera | and he forgot about the fundamental rule about releases | 11:36 |
| rdopiera | once you release something, it stays releases | 11:36 |
| rdopiera | once you release something, it stays released | 11:36 |
| rdopiera | you cannot unrealease anything | 11:36 |
| AJaeger | rdopiera, please tell him once he shows up ;) | 11:37 |
| openstackgerrit | Derek Higgins proposed openstack-infra/zuul: Document ZUUL_CHANGES https://review.openstack.org/129243 | 11:37 |
| robcresswell | The horizon patch has passed jenkins, so once the global-reqs passes it should be fine.The bugs are marked so should close themselves. | 11:37 |
| rdopiera | AJaeger: I will do better, I will add it to the docs about creating xstatic packages | 11:37 |
| robcresswell | Just need reviews once Jenkins has passed now :) | 11:37 |
| AJaeger | rdopiera, can we upload the xstatic packages via gerrit in the future? | 11:38 |
| AJaeger | robcresswell, that quickly? | 11:38 |
| rdopiera | AJaeger: yes, but I couldn't get that to work, I would appreciate any help with it | 11:38 |
| rdopiera | AJaeger: whenever you or anyone else has time | 11:39 |
| AJaeger | rdopiera, could you join the infra meeting at 19:00 UTC today? Then let's add it as a topic there... | 11:39 |
| robcresswell | AJaeger: For some reason, the horizon patch had already passed, I have no idea why. The global reqs patch has a recheck running from rdopiera. | 11:40 |
| *** ssam2 has quit IRC | 11:40 | |
| rdopiera | certainly | 11:40 |
| AJaeger | rdopiera, are xstatics some how specail from packaging point of view? | 11:40 |
| rdopiera | AJaeger: no, they are regular python packages | 11:40 |
| *** salv-orlando has joined #openstack-infra | 11:40 | |
| rdopiera | AJaeger: as mundane, as possible | 11:40 |
| AJaeger | rdopiera, then it should be trivial to upload them - just add publish-to-pypi to them... | 11:40 |
| AJaeger | rdopiera, let me do a quick patch... | 11:41 |
| rdopiera | AJaeger: if you can make a patch on one, I can do the rest | 11:41 |
| rdopiera | AJaeger: thanks | 11:41 |
| *** tsufiev_ is now known as tsufiev | 11:42 | |
| rdopiera | AJaeger: by the way, if you see any other checks or anything that would be nice to have there | 11:42 |
| AJaeger | rdopiera, let's ask the other infra cores for suggestions as well... I need to do some other stuff ;) | 11:43 |
| AJaeger | but will soon have a patch... | 11:43 |
| openstackgerrit | Andreas Jaeger proposed openstack-infra/project-config: Setup publishing to pypi for xstatic-angular https://review.openstack.org/169274 | 11:44 |
| AJaeger | rdopiera, feel free to take my patch and finish it - I just took the first project to show what needs to be done | 11:44 |
| AJaeger | and if you do, add yourself as co-author | 11:44 |
| AJaeger | ttx, you can send an allclear again and ask for rechecks... | 11:45 |
| ttx | AJaeger: just a sec | 11:46 |
| AJaeger | ttx: rdopiera, gave us some breathing time ;) | 11:46 |
| rdopiera | AJaeger: how is the upload triggered? | 11:46 |
| rdopiera | AJaeger: by tagging a release? | 11:47 |
| rdopiera | AJaeger: or on every commit? | 11:47 |
| *** dmellado has joined #openstack-infra | 11:47 | |
| ttx | #status ok Check/Gate unstuck, feel free to recheck your abusively-failed changes. | 11:47 |
| openstackstatus | ttx: sending ok | 11:47 |
| AJaeger | rdopiera, http://docs.openstack.org/infra/manual/drivers.html#tagging-a-release - tagging a release | 11:47 |
| AJaeger | ttx, thanks! | 11:47 |
| rdopiera | AJaeger: ok | 11:48 |
| ttx | rdopiera: thx for saving the day! | 11:48 |
| ttx | AJaeger: and thanks for seeing this through. | 11:48 |
| rdopiera | sorry for the kerfuffle, it's all caused by inadequate documentation and processes on our side | 11:48 |
| AJaeger | ttx, you're welcome | 11:48 |
| *** ChanServ changes topic to "Discussion of OpenStack Developer and Community Infrastructure | docs http://docs.openstack.org/infra/manual/ http://ci.openstack.org/ | bugs https://storyboard.openstack.org/ | source https://git.openstack.org/cgit/openstack-infra/" | 11:49 | |
| AJaeger | rdopiera, let's follow the usual pypi release process as everybody else and we can all share and improve the same documentation ;) | 11:49 |
| -openstackstatus- NOTICE: Check/Gate unstuck, feel free to recheck your abusively-failed changes. | 11:49 | |
| robcresswell | rdopiera, AJaeger: Thanks for the fix. | 11:50 |
| AJaeger | I'm retriggering https://review.openstack.org/#/c/169132/ - it fails right now, the recheck was too early | 11:50 |
| openstackgerrit | Andreas Jaeger proposed openstack/requirements: Updated incorrect name for angular lrdragndrop https://review.openstack.org/169132 | 11:51 |
| openstackstatus | ttx: finished sending ok | 11:51 |
| robcresswell | AJaeger: I think a recheck would've worked fine, rather than a new patch, no? | 11:51 |
| robcresswell | either way, thanks | 11:52 |
| AJaeger | robcresswell, no, it wouldn't - a recheck would be ignored right now. | 11:52 |
| robcresswell | AJaeger: Ah, okay | 11:52 |
| flaper87 | w000t | 11:52 |
| *** ihrachyshka has joined #openstack-infra | 11:52 | |
| flaper87 | thanks folks | 11:52 |
| AJaeger | The change was failing and the fastest way to queue it in again is not waiting until it'S finished and then recheck but do a new patch... | 11:52 |
| *** punith has quit IRC | 11:52 | |
| AJaeger | flaper87, care to followup on your email with the all-clear? | 11:53 |
| robcresswell | AJaeger: Yep, I wasn;t sure whether recheck would reset a patch that was currently being checked. | 11:53 |
| *** ssam2 has joined #openstack-infra | 11:53 | |
| *** weshay has joined #openstack-infra | 11:53 | |
| *** MaxV has quit IRC | 11:53 | |
| flaper87 | AJaeger: will do, right away | 11:54 |
| flaper87 | AJaeger: mind to point me to the summary of the fix ? | 11:54 |
| flaper87 | if there's any | 11:54 |
| *** pal_ has joined #openstack-infra | 11:54 | |
| AJaeger | flaper87, rdopiera uploaded the deleted package again ;) | 11:54 |
| flaper87 | ooook | 11:54 |
| AJaeger | NOw we can do the change properly... | 11:54 |
| *** ldnunes has joined #openstack-infra | 11:55 | |
| *** tiswanso has quit IRC | 11:55 | |
| flaper87 | sent | 11:55 |
| openstackgerrit | Andreas Jaeger proposed openstack-infra/project-config: Setup publishing to pypi for xstatic-angular https://review.openstack.org/169274 | 11:56 |
| *** mpaolino has quit IRC | 11:57 | |
| *** mpaolino has joined #openstack-infra | 11:59 | |
| *** mpavone has quit IRC | 12:00 | |
| *** woodster_ has joined #openstack-infra | 12:01 | |
| *** hdd has joined #openstack-infra | 12:02 | |
| *** annegentle has joined #openstack-infra | 12:03 | |
| *** ildikov has quit IRC | 12:03 | |
| openstackgerrit | Flavio Percoco proposed openstack-infra/project-config: Install required services for Zaqar https://review.openstack.org/169269 | 12:05 |
| *** annegentle has quit IRC | 12:08 | |
| *** dprince has joined #openstack-infra | 12:09 | |
| *** erlon has joined #openstack-infra | 12:09 | |
| *** MaxV has joined #openstack-infra | 12:10 | |
| *** _nadya_ has quit IRC | 12:12 | |
| *** dkranz has joined #openstack-infra | 12:12 | |
| *** aysyd has joined #openstack-infra | 12:13 | |
| *** bswartz has quit IRC | 12:13 | |
| *** ddieterly has joined #openstack-infra | 12:14 | |
| *** mpavone has joined #openstack-infra | 12:14 | |
| *** mpaolino has quit IRC | 12:16 | |
| *** yamamoto has joined #openstack-infra | 12:16 | |
| *** hdd has quit IRC | 12:16 | |
| *** markvoelker has joined #openstack-infra | 12:17 | |
| *** yfried is now known as yfried|afk | 12:17 | |
| *** bhunter71 has joined #openstack-infra | 12:18 | |
| *** kgiusti has joined #openstack-infra | 12:18 | |
| *** ildikov has joined #openstack-infra | 12:18 | |
| *** ddieterly has quit IRC | 12:19 | |
| *** amotoki_ has quit IRC | 12:20 | |
| openstackgerrit | Radomir Dopieralski proposed openstack-infra/project-config: Setup publishing to pypi for xstatic-* https://review.openstack.org/169274 | 12:23 |
| *** yfried|afk is now known as yfried | 12:24 | |
| *** dims has quit IRC | 12:25 | |
| *** adalbas has joined #openstack-infra | 12:25 | |
| *** dims has joined #openstack-infra | 12:25 | |
| *** claudiub has quit IRC | 12:27 | |
| *** claudiub has joined #openstack-infra | 12:27 | |
| *** mpaolino has joined #openstack-infra | 12:28 | |
| *** dboik has joined #openstack-infra | 12:30 | |
| *** dkliban_afk is now known as dkliban | 12:31 | |
| *** Guest47177 has quit IRC | 12:31 | |
| *** robcresswell has left #openstack-infra | 12:33 | |
| *** amotoki_ has joined #openstack-infra | 12:34 | |
| fungi | does anyone have a summary on how the package rename (which happened on friday) didn't cause any problems for integration tests until today? | 12:34 |
| *** amotoki_ has quit IRC | 12:34 | |
| fungi | manual actions taken on pypi.python.org is my take from briefly skimming the scrollback here | 12:36 |
| *** yfried is now known as yfried|afk | 12:37 | |
| *** Hal has joined #openstack-infra | 12:37 | |
| *** Hal is now known as Guest1663 | 12:38 | |
| *** tiswanso has joined #openstack-infra | 12:38 | |
| *** yamamoto has quit IRC | 12:39 | |
| *** yfried|afk is now known as yfried | 12:41 | |
| *** e0ne is now known as e0ne_ | 12:42 | |
| *** bknudson has joined #openstack-infra | 12:42 | |
| *** radez_g0n3 is now known as radez | 12:43 | |
| *** e0ne_ is now known as e0ne | 12:43 | |
| *** smccully has joined #openstack-infra | 12:47 | |
| fungi | aha. mailing list thread says the original package name, which horizon was still relying on, got _deleted_ from pypi (seriously, who deletes packages from pypi like that?!?) | 12:50 |
| *** gordc has joined #openstack-infra | 12:50 | |
| *** dboik has quit IRC | 12:50 | |
| mordred | fungi: wow. that's special | 12:51 |
| fungi | indeed | 12:51 |
| *** dboik has joined #openstack-infra | 12:52 | |
| *** yfried has quit IRC | 12:53 | |
| *** kerpan has joined #openstack-infra | 12:53 | |
| *** tnovacik has quit IRC | 12:53 | |
| *** chris613 has joined #openstack-infra | 12:53 | |
| *** bswartz has joined #openstack-infra | 12:54 | |
| *** ildikov_ has joined #openstack-infra | 12:58 | |
| *** jistr|class is now known as jistr | 12:58 | |
| *** ddieterly has joined #openstack-infra | 12:58 | |
| fungi | rdopiera: aha, it was you. i guess we each get to learn that lesson at some point! deleting releases from pypi isn't really ever a good idea | 12:59 |
| fungi | thanks for straightening it out | 12:59 |
| openstackgerrit | Claudiu Belu proposed openstack-infra/project-config: Adds cloud-init to StackForge https://review.openstack.org/169293 | 13:00 |
| openstackgerrit | Radomir Dopieralski proposed openstack/requirements: Updated incorrect name for angular lrdragndrop https://review.openstack.org/169132 | 13:00 |
| *** ildikov has quit IRC | 13:00 | |
| *** nilasae has quit IRC | 13:01 | |
| *** ajo has quit IRC | 13:01 | |
| *** ajo has joined #openstack-infra | 13:01 | |
| rdopiera | fungi: it's actually more complicated, we need to do the change in several steps... | 13:01 |
| fungi | rdopiera: i expect so. figured renaming the git repo was only one of many | 13:02 |
| *** annegentle has joined #openstack-infra | 13:02 | |
| *** smccully_ has joined #openstack-infra | 13:05 | |
| *** hogepodge has joined #openstack-infra | 13:05 | |
| *** baoli has joined #openstack-infra | 13:05 | |
| *** prmtl has quit IRC | 13:06 | |
| *** tonytan4ever has joined #openstack-infra | 13:06 | |
| *** subscope has joined #openstack-infra | 13:07 | |
| *** nmagnezi_ is now known as nmagnezi | 13:08 | |
| *** e0ne is now known as e0ne_ | 13:08 | |
| AJaeger | fungi, https://etherpad.openstack.org/p/lrdragndrop-stuck has all information | 13:08 |
| fungi | thanks! | 13:09 |
| *** eharney has joined #openstack-infra | 13:10 | |
| mrmartin | fungi: I've added the last piece of ask.o.o backup https://review.openstack.org/#/c/169260/ but it is failing to build until the two other pgsql-backup puppet not gets an approval | 13:10 |
| *** dboik has quit IRC | 13:11 | |
| AJaeger | rdopiera, do you want to fix the xstatic patch or shall I? | 13:11 |
| *** radez is now known as radez_g0n3 | 13:11 | |
| *** macjack has joined #openstack-infra | 13:11 | |
| rdopiera | AJaeger: what do you mean? | 13:12 |
| rdopiera | aah | 13:12 |
| rdopiera | I see, I will fix it | 13:12 |
| AJaeger | rdopiera, thanks! | 13:12 |
| mrmartin | fungi: I cannot participate today in the infra meeting, but the ask.o.o seems to be ok, I cannot reproduce the Chinese lang solr search issue on the new instance and dev environment, so if the backup is passing, then I don't know about more open issues. | 13:12 |
| AJaeger | Remember adding yourself as "Co-Authored-By" | 13:12 |
| *** e0ne_ is now known as e0ne | 13:13 | |
| openstackgerrit | Radomir Dopieralski proposed openstack-infra/project-config: Setup publishing to pypi for xstatic-* https://review.openstack.org/169274 | 13:13 |
| *** peristeri has joined #openstack-infra | 13:13 | |
| ctlaugh | krtaylor: I just saw your PM and responded. | 13:14 |
| openstackgerrit | Radomir Dopieralski proposed openstack-infra/project-config: Setup publishing to pypi for xstatic-* https://review.openstack.org/169274 | 13:14 |
| rdopiera | AJaeger: roger | 13:14 |
| *** btully has joined #openstack-infra | 13:14 | |
| AJaeger | rdopiera, looks fine, thanks! | 13:15 |
| *** ddieterly has quit IRC | 13:16 | |
| *** ddieterly has joined #openstack-infra | 13:17 | |
| *** btully has quit IRC | 13:19 | |
| *** otter768 has joined #openstack-infra | 13:19 | |
| *** mriedem_away is now known as mriedem | 13:20 | |
| *** ddieterly has quit IRC | 13:21 | |
| *** ayoung has joined #openstack-infra | 13:23 | |
| openstackgerrit | Joshua Hesketh proposed openstack-infra/zuul: Document the new connections in zuul https://review.openstack.org/169270 | 13:23 |
| openstackgerrit | Joshua Hesketh proposed openstack-infra/zuul: Add support for 'connection' concept https://review.openstack.org/121528 | 13:23 |
| openstackgerrit | Joshua Hesketh proposed openstack-infra/zuul: Add in extra connections tests https://review.openstack.org/169300 | 13:23 |
| *** dustins has joined #openstack-infra | 13:24 | |
| *** otter768 has quit IRC | 13:24 | |
| *** wayneeseguin has quit IRC | 13:24 | |
| *** tnovacik has joined #openstack-infra | 13:25 | |
| *** unicell1 has joined #openstack-infra | 13:26 | |
| *** unicell has quit IRC | 13:26 | |
| *** mpavone has quit IRC | 13:27 | |
| *** mfink has joined #openstack-infra | 13:28 | |
| *** jaypipes has quit IRC | 13:29 | |
| *** xyang1 has joined #openstack-infra | 13:30 | |
| *** ssam2 has quit IRC | 13:30 | |
| *** dboik has joined #openstack-infra | 13:31 | |
| *** spzala has joined #openstack-infra | 13:31 | |
| *** garyh has joined #openstack-infra | 13:41 | |
| *** dhritishikhar_ has joined #openstack-infra | 13:42 | |
| *** jaypipes has joined #openstack-infra | 13:42 | |
| openstackgerrit | Claudiu Belu proposed openstack-infra/project-config: Adds cloud-init to StackForge https://review.openstack.org/169293 | 13:42 |
| *** hdd has joined #openstack-infra | 13:43 | |
| *** macjack has quit IRC | 13:43 | |
| *** ssam2 has joined #openstack-infra | 13:43 | |
| *** ddieterly has joined #openstack-infra | 13:45 | |
| *** masoz has quit IRC | 13:46 | |
| *** tsg_ has joined #openstack-infra | 13:46 | |
| *** amitgandhinz has joined #openstack-infra | 13:46 | |
| openstackgerrit | Jordan Pittier proposed openstack-infra/system-config: Third party CI: document how to properly format message with Gerrit Trigger https://review.openstack.org/167716 | 13:48 |
| fungi | mrmartin: thanks for the update. sounds like we're close to being able to do a cut-over maintenance for it | 13:50 |
| *** mwagner_lap has quit IRC | 13:51 | |
| *** amotoki_ has joined #openstack-infra | 13:51 | |
| rcarrillocruz | hey folks, may I please get another core review on https://review.openstack.org/#/c/167553/ ? just need another one to bring the project to stackforge | 13:52 |
| rcarrillocruz | thanks | 13:52 |
| *** ddieterly has quit IRC | 13:54 | |
| *** armax has joined #openstack-infra | 13:55 | |
| *** ildikov_ has quit IRC | 13:57 | |
| mordred | fungi: you want an easy review? https://review.openstack.org/#/c/168671/1 :) | 13:57 |
| fungi | mordred: any idea why the parent change wasn't approved already? | 13:59 |
| *** sigmavirus24_awa is now known as sigmavirus24 | 13:59 | |
| mordred | fungi: nope. but I was going to approve it in a sec | 13:59 |
| fungi | too late | 14:00 |
| mordred | woot! | 14:00 |
| mordred | less things remaining in the queue == good | 14:00 |
| *** wenlock has joined #openstack-infra | 14:01 | |
| mordred | fungi: also, while I'm bugging you - can I get an opinion on https://review.openstack.org/#/c/168734/ ? | 14:02 |
| mordred | fungi: it's more urllib warnings | 14:02 |
| *** andreykurilin has joined #openstack-infra | 14:03 | |
| *** andreykurilin has left #openstack-infra | 14:03 | |
| mordred | fungi: in this case, the rackspace deprecated cert thing which makes using this stuff very verbose/chatty on a topic we can't fix | 14:03 |
| sigmavirus24 | mordred: that's still not fixed? | 14:03 |
| * sigmavirus24 goes to bug people | 14:03 | |
| fungi | sigmavirus24: you should totally bug people | 14:04 |
| *** zz_dimtruck is now known as dimtruck | 14:04 | |
| *** sabeen has joined #openstack-infra | 14:04 | |
| mordred | sigmavirus24: nope | 14:04 |
| sigmavirus24 | It was supposed to have been fixed by last week | 14:04 |
| * sigmavirus24 is tracking down the people who promised that | 14:04 | |
| *** mrmartin has quit IRC | 14:05 | |
| mordred | sigmavirus24: oh - wait a sec .. | 14:05 |
| jroll | sigmavirus24: wow, my nags went nowhere :| | 14:05 |
| *** ayoung has quit IRC | 14:05 | |
| fungi | /usr/local/lib/python2.7/dist-packages/requests/packages/urllib3/connection.py:251: SecurityWarning: Certificate has no `subjectAltName`, falling back to check for a `commonName` for now. This feature is being removed by major browsers and deprecated by RFC 2818. (See https://github.com/shazow/urllib3/issues/497 for details.) | 14:05 |
| mordred | sigmavirus24: I just retested without my silencing of it | 14:05 |
| mordred | fungi: did you get that recently? | 14:06 |
| sigmavirus24 | fungi: I know the error | 14:06 |
| * fungi is still getting that when testing with novaclient on rax just now | 14:06 | |
| sigmavirus24 | fungi: I added that warning myself | 14:06 |
| *** gokrokve has joined #openstack-infra | 14:06 | |
| mordred | :) | 14:06 |
| fungi | yep, so i gathered | 14:06 |
| * sigmavirus24 is an evil person | 14:06 | |
| fungi | just confirming it hasn't changed on us | 14:06 |
| *** btully has joined #openstack-infra | 14:06 | |
| AJaeger | mordred, do you want an easy review? ;) https://review.openstack.org/168135 or https://review.openstack.org/168148 or https://review.openstack.org/168159 is my spring cleanup | 14:07 |
| fungi | it's sandwiched between a couple of the "you're running a python 2.x more than three months old, you should upgrade" warnings, so it's not as easy to spot as it could be | 14:07 |
| openstackgerrit | Ricardo Carrillo Cruz proposed openstack-infra/system-config: Expose puppetmaster_server param on openstack_project::puppetmaster https://review.openstack.org/168572 | 14:08 |
| *** yfried has joined #openstack-infra | 14:08 | |
| *** dkranz has quit IRC | 14:08 | |
| mordred | fungi: you know, when I go to test that I'm getting it, I really should test against rax | 14:08 |
| fungi | i wondered if you stopped seeing it because you were connecting to hpcloud ;) | 14:09 |
| *** mpavone has joined #openstack-infra | 14:09 | |
| *** ChuckC has quit IRC | 14:09 | |
| *** mpaolino has quit IRC | 14:10 | |
| *** nelsnelson has joined #openstack-infra | 14:10 | |
| sigmavirus24 | fungi: that python warning isn't mine =P | 14:11 |
| *** _nadya_ has joined #openstack-infra | 14:11 | |
| openstackgerrit | Monty Taylor proposed openstack-infra/shade: Disable warnings about old Rackspace certificates https://review.openstack.org/168734 | 14:11 |
| fungi | sigmavirus24: yeah, that's bubbling up from urllib3 directly, i gather | 14:11 |
| mordred | ok. I have a better way of filtering just that one | 14:11 |
| mordred | Shrews: ^^ you'll like that better | 14:11 |
| Shrews | mordred: i hate it. | 14:12 |
| Shrews | though i haven't looked yet | 14:12 |
| * mordred throws a cow at Shrews | 14:12 | |
| *** baoli has quit IRC | 14:12 | |
| jroll | mordred: "there is nothing a user can do" they can call up rackspace and complain loudly :) | 14:12 |
| mordred | jroll: we're done that already :) | 14:13 |
| *** carl_baldwin has joined #openstack-infra | 14:13 | |
| *** dboik_ has joined #openstack-infra | 14:13 | |
| Shrews | let's all complain to jroll on a daily basis until it is fixed | 14:13 |
| *** ildikov_ has joined #openstack-infra | 14:13 | |
| jroll | mordred: not loudly enough, apparently :| | 14:13 |
| sigmavirus24 | fungi: they both come from urllib3, but I only added one of them =P | 14:13 |
| sdague | jroll: we should have the warning include the rax support line number then | 14:14 |
| jroll | heh | 14:14 |
| openstackgerrit | Monty Taylor proposed openstack-infra/nodepool: Disable warnings about old Rackspace certificates https://review.openstack.org/169327 | 14:14 |
| mordred | fungi: why not add it there too ^ ? | 14:14 |
| *** dboik_ has quit IRC | 14:14 | |
| *** dboik_ has joined #openstack-infra | 14:15 | |
| *** dboik has quit IRC | 14:16 | |
| sigmavirus24 | so best estimate I can get now is early April (is when that certificate will work) | 14:16 |
| sigmavirus24 | mordred: to be fair, the certificate isn't old | 14:16 |
| openstackgerrit | Monty Taylor proposed openstack-infra/shade: Create swift container if it does not exist https://review.openstack.org/167823 | 14:17 |
| *** dhritishikhar_ has quit IRC | 14:18 | |
| openstackgerrit | Monty Taylor proposed openstack-infra/shade: Fix a use of in where it should be equality https://review.openstack.org/169105 | 14:18 |
| *** ayoung has joined #openstack-infra | 14:18 | |
| mordred | sigmavirus24: fair. "deprecated" makes me think "old" | 14:19 |
| openstackgerrit | Monty Taylor proposed openstack-infra/shade: Namespace caching per cloud https://review.openstack.org/168615 | 14:19 |
| openstackgerrit | Monty Taylor proposed openstack-infra/shade: Allow for passing cache class in as a parameter https://review.openstack.org/162148 | 14:19 |
| openstackgerrit | Monty Taylor proposed openstack-infra/shade: Add 'rebuild' to shade https://review.openstack.org/161067 | 14:19 |
| openstackgerrit | Monty Taylor proposed openstack-infra/shade: Migrate API calls to task management https://review.openstack.org/168665 | 14:19 |
| openstackgerrit | Monty Taylor proposed openstack-infra/shade: Add task management framework https://review.openstack.org/168664 | 14:19 |
| sigmavirus24 | mordred: yeah I get that | 14:19 |
| *** ZZelle has quit IRC | 14:19 | |
| mordred | fungi, Shrews: that's all basically just rebasing that stack - I pulled three patches out and made them standalone and not in the dependency chain | 14:19 |
| sigmavirus24 | Certificates without SANs are "okay" but every major browser is removing support for them because it's a good hardening measure | 14:19 |
| sigmavirus24 | We're planning to do the same, but we're also not sure /when/ we'll do that | 14:20 |
| *** ZZelle has joined #openstack-infra | 14:20 | |
| sigmavirus24 | Things like removing support for RC4 ciphers, however, are easy to decide to do on the other hand =P | 14:20 |
| *** timcline has joined #openstack-infra | 14:21 | |
| Shrews | mordred: that container change... there was a change i made in Feb to remove the call to create the container within create_object(). It shouldn't be necessary. You're finding it is? | 14:21 |
| *** dkranz has joined #openstack-infra | 14:21 | |
| Shrews | mordred: commit 91bc40 | 14:22 |
| *** carl_baldwin has quit IRC | 14:22 | |
| *** enikanorov has quit IRC | 14:23 | |
| *** mpaolino has joined #openstack-infra | 14:23 | |
| *** timcline has quit IRC | 14:24 | |
| *** sergsh has quit IRC | 14:24 | |
| *** timcline has joined #openstack-infra | 14:24 | |
| openstackgerrit | Merged openstack-infra/shade: Add api-level timeout parameter https://review.openstack.org/168596 | 14:26 |
| *** timcline_ has joined #openstack-infra | 14:26 | |
| openstackgerrit | Merged openstack-infra/shade: Pass socket timeout to all of the Client objects https://review.openstack.org/168671 | 14:26 |
| *** enikanorov has joined #openstack-infra | 14:27 | |
| *** tsg_ has quit IRC | 14:27 | |
| openstackgerrit | Davide Michelino proposed openstack-infra/project-config: Add Networking-portforwarding project https://review.openstack.org/169332 | 14:28 |
| *** sdake has joined #openstack-infra | 14:29 | |
| mordred | Shrews: yes. it is necessary | 14:29 |
| mordred | Shrews: on at least one of the clouds | 14:29 |
| *** timcline has quit IRC | 14:30 | |
| *** dboik_ has quit IRC | 14:30 | |
| Shrews | mordred: poo. then let's put it back in create_object() instead | 14:30 |
| mordred | Shrews: but - I agree it should not be in create_object in general - that was the wrong level ... it _should_ be in upload_image because the user shouldn't need to know they wanted a container | 14:30 |
| mordred | Shrews: oh - ok | 14:30 |
| mordred | Shrews: see, you'd convinced me on taht one :) | 14:30 |
| *** zz_jgrimm is now known as jgrimm | 14:31 | |
| Shrews | i think your own argument aligns with my suggestion | 14:31 |
| mordred | hrm | 14:31 |
| mordred | you're probably right | 14:31 |
| Shrews | mordred: i wonder if it is an error to create a container that already exists | 14:32 |
| *** baoli has joined #openstack-infra | 14:33 | |
| *** mattfarina has joined #openstack-infra | 14:36 | |
| Shrews | oh, create_container() takes care of that | 14:37 |
| rcarrillocruz | fungi: hiya, if you have +2 powers on project-config (i assume you do), could you have a look at https://review.openstack.org/#/c/167553/ pls ? | 14:38 |
| openstackgerrit | Sean M. Collins proposed openstack-infra/elastic-recheck: Add fingerprint for XStatic-Angular-Irdragndrop dependency failure https://review.openstack.org/169156 | 14:39 |
| *** ChuckC has joined #openstack-infra | 14:39 | |
| openstackgerrit | Monty Taylor proposed openstack-infra/nodepool: Add tests for shade cloud config parameter https://review.openstack.org/169339 | 14:41 |
| fungi | rcarrillocruz: out of curiosity, how is gozer leveraging autossh? | 14:41 |
| mordred | jeblair: I added tests for teh config validation as per ianw's request, but I did it in a follow on patch to reduce churn | 14:41 |
| mordred | fungi: I would counsel rcarrillocruz to not tell you that | 14:42 |
| mordred | fungi: only for the purposes of good though, I promise | 14:42 |
| fungi | mordred: you and your corporate security ;) | 14:42 |
| fungi | s/security/secrecy/ | 14:42 |
| rcarrillocruz | mordred has spoken! | 14:43 |
| mordred | I will say that, in general, autossh does a great job at maintaining long-lived ssh connections | 14:43 |
| openstackgerrit | Radomir Dopieralski proposed openstack-infra/project-config: Publish tuskar-ui-extras to pypi https://review.openstack.org/169341 | 14:43 |
| *** pal has joined #openstack-infra | 14:43 | |
| *** pal_ has quit IRC | 14:43 | |
| rcarrillocruz | it sure does | 14:44 |
| fungi | i've never tried it, but i do know people who use it for personal purposes and are pleased with the result | 14:44 |
| mordred | ianw, yolanda: speaking of - I adressed your concerns in https://review.openstack.org/#/c/168603/ but in a follow on patch | 14:44 |
| claudiub | fungi: hello | 14:44 |
| fungi | claudiub: what have i broken now? | 14:44 |
| claudiub | fungi: I am having a little bit og trouble with a commit on project-config | 14:45 |
| fungi | claudiub: which commit? | 14:45 |
| claudiub | fungi: I have a Jenkins -1 that I do not know how to solve: https://review.openstack.org/#/c/169293/2 | 14:45 |
| openstackgerrit | Monty Taylor proposed openstack-infra/shade: Create swift container if it does not exist https://review.openstack.org/167823 | 14:45 |
| mordred | Shrews: there you go ^^ | 14:45 |
| claudiub | fungi: basically, it says: #cloud-init is not registered with ChanServ. | 14:45 |
| * fungi checks | 14:46 | |
| Shrews | mordred: +1'd | 14:46 |
| fungi | claudiub: yep, if you `/msg chanserv access #cloud-init list` you'll be told the same thing by chanserv | 14:47 |
| fungi | claudiub: freenode's instructions point to http://blog.freenode.net/2008/04/registering-a-channel-on-freenode/ | 14:47 |
| *** amotoki_ has quit IRC | 14:49 | |
| claudiub | fungi: aha, so basically, the channel exists, but it wasn't registered, correct? | 14:49 |
| *** hashar is now known as hasharShopping | 14:49 | |
| fungi | claudiub: yes, in irc, a channel comes into existence the moment at least one user tries to join it | 14:50 |
| TheJulia | win 5 | 14:50 |
| fungi | claudiub: however, to register a channel with the network's services bots, you need to follow the additional steps documented there | 14:50 |
| claudiub | fungi: I see. Thanks for the help! :) | 14:51 |
| *** ayoung has quit IRC | 14:51 | |
| openstackgerrit | Claudiu Belu proposed openstack-infra/project-config: Adds cloud-init to StackForge https://review.openstack.org/169293 | 14:51 |
| fungi | hrm, minor hitch in my plan to start running jobs on devstack-.* workers instead of bare-.* | 14:53 |
| anteaya | oh? | 14:53 |
| fungi | we don't actually _have_ any devstack-centos6 | 14:53 |
| mordred | oh | 14:53 |
| anteaya | well that would be a hitch yes | 14:53 |
| *** jaypipes has quit IRC | 14:53 | |
| mordred | fungi: we should make some | 14:53 |
| fungi | yep, shouldn't be hard as long as we don't actually expect devstack itself to work on them | 14:54 |
| fungi | just making a mental note that i'll need to add that | 14:54 |
| anteaya | yup | 14:54 |
| mordred | fungi: maybe we should rename them | 14:54 |
| *** carl_baldwin has joined #openstack-infra | 14:54 | |
| fungi | that's the eventual plan. when bare-trusty for example goes away because we're running all its jobs on devstack-trusty, we can just rename that to trusty | 14:54 |
| mordred | yah. it'll be less confusing then | 14:55 |
| rcarrillocruz | thx fungi | 14:55 |
| *** sputnik13 has joined #openstack-infra | 14:55 | |
| mordred | fungi: fwiw, I recommend calling them "ubuntu-trusty" and not just trusty | 14:56 |
| fungi | mordred: cool. and centos-6? | 14:56 |
| mordred | because that way we can call the centos and fedora nodes "centos-6" and "fedora-21" | 14:56 |
| mordred | yah | 14:56 |
| *** marun has joined #openstack-infra | 14:56 | |
| mordred | that way it's "$DISTRO-$RELEASE" everywhere | 14:56 |
| * rcarrillocruz finds disturbing that some manifests have a mix of '' and 'UNDEF' on class params... | 14:56 | |
| mordred | fungi: I suppoe we could skip it and just have an image named "6" and one named "21" :) | 14:57 |
| openstackgerrit | Aleksey Ripinen proposed openstack-infra/storyboard-webclient: Support for tags https://review.openstack.org/166847 | 14:57 |
| *** devvesa has joined #openstack-infra | 14:58 | |
| openstackgerrit | Merged openstack-infra/project-config: Add stackforge/puppet-autossh https://review.openstack.org/167553 | 14:58 |
| *** dangers_away is now known as dangers | 14:58 | |
| *** dustins_ has joined #openstack-infra | 14:58 | |
| *** dulek has quit IRC | 14:58 | |
| fungi | mordred: in that case, i may want to change how the version and release strings are emitted by bindep since i sort of chose them to align with the worker labels we're using (in absence of any existing format defined to express them) | 14:58 |
| fungi | i guess we can try a quick bikeshed on standardizing our worker labels during today's meeting | 15:00 |
| mordred | fungi: each of the dib builds I've got now take each of those two things - largely because there is a base dib element you need for a given thing | 15:00 |
| mordred | but then it takes a parameter for version | 15:00 |
| *** bswartz has quit IRC | 15:00 | |
| *** devvesa has quit IRC | 15:00 | |
| mordred | there is a centos6 base element in the dib elements which kinda makes everything horky | 15:00 |
| mordred | so with the minimal elements, I kept it ${DISTRO}-minimal and each of those take a param for release so you can build what you want with a decent pattern | 15:01 |
| *** baoli has quit IRC | 15:01 | |
| *** claudiub has quit IRC | 15:01 | |
| *** dustins has quit IRC | 15:01 | |
| fungi | eventual alignment between dib, bindep and our nodepool configuration might be nice, yeah | 15:01 |
| *** dboik has joined #openstack-infra | 15:02 | |
| *** dboik has quit IRC | 15:02 | |
| *** cdent has quit IRC | 15:02 | |
| *** dboik has joined #openstack-infra | 15:03 | |
| *** rdopiera has quit IRC | 15:03 | |
| *** bswartz has joined #openstack-infra | 15:03 | |
| *** ayoung has joined #openstack-infra | 15:04 | |
| *** dims has quit IRC | 15:05 | |
| *** baoli has joined #openstack-infra | 15:05 | |
| openstackgerrit | Aleksey Ripinen proposed openstack-infra/storyboard-webclient: Added query case to buildSignature https://review.openstack.org/169360 | 15:06 |
| *** annegentle has quit IRC | 15:08 | |
| *** dustins_ has quit IRC | 15:08 | |
| *** dims_ has joined #openstack-infra | 15:08 | |
| *** pal has quit IRC | 15:08 | |
| *** mrmartin has joined #openstack-infra | 15:08 | |
| anteaya | fungi: do you think we need any discussion of the ldragndrop issue in today's meeting, in terms of messaging around renaming projects | 15:09 |
| anteaya | as you say it appears someone deletes from pypi just once in their life | 15:09 |
| *** sputnik13 has quit IRC | 15:10 | |
| fungi | anteaya: i doubt it would be worth covering "don't delete releases from pypi" in the infra team meeting. i think we would be mostly preaching to the choir | 15:11 |
| *** dustins has joined #openstack-infra | 15:11 | |
| *** MaxV has quit IRC | 15:11 | |
| fungi | it wasn't any regular participant in the infra meetings who did it | 15:11 |
| anteaya | yes, but in our messaging around renaming projects | 15:13 |
| *** ajmiller has joined #openstack-infra | 15:13 | |
| anteaya | should we do anything differently there | 15:13 |
| anteaya | in terms of what we offer to others going through the process | 15:13 |
| *** annegentle has joined #openstack-infra | 15:13 | |
| *** reed has joined #openstack-infra | 15:13 | |
| anteaya | I'm fine if you disagree, just sharing a thought | 15:13 |
| openstackgerrit | James E. Blair proposed stackforge/gertty: Support searching by URL https://review.openstack.org/169371 | 15:15 |
| *** zz_gondoi has joined #openstack-infra | 15:15 | |
| *** ajmiller_ has quit IRC | 15:15 | |
| *** MaxV has joined #openstack-infra | 15:15 | |
| *** zz_gondoi is now known as gondoi | 15:15 | |
| *** claudiub has joined #openstack-infra | 15:16 | |
| *** emagana has joined #openstack-infra | 15:18 | |
| *** MarkAtwood has joined #openstack-infra | 15:18 | |
| fungi | it seems at most tangentially related to the topic of repo renames. also it was a stackforge project without any release integration in our infrastructure (someone performing their release process by hand) and therefore is just as likely to happen to us in situations completely unrelated to repo renaming | 15:18 |
| fungi | even by projects which are in no way involved in our community whatsoever | 15:18 |
| *** carl_baldwin has quit IRC | 15:19 | |
| anteaya | okay | 15:19 |
| *** otter768 has joined #openstack-infra | 15:20 | |
| *** doug-fish has quit IRC | 15:20 | |
| jeblair | i kind of think if there's a stackforge project in global requirements it's in the wrong place | 15:20 |
| *** carl_baldwin_ has joined #openstack-infra | 15:20 | |
| openstackgerrit | yolanda.robla proposed openstack-infra/nodepool: Independent provider manager for polling tasks https://review.openstack.org/169235 | 15:21 |
| yolanda | mordred, jeblair, proposal for independent manager ^ | 15:21 |
| *** zigo__ is now known as zigo | 15:23 | |
| *** dannywilson has joined #openstack-infra | 15:24 | |
| *** signed8bit has joined #openstack-infra | 15:24 | |
| *** dannywilson has quit IRC | 15:24 | |
| *** otter768 has quit IRC | 15:25 | |
| *** dannywilson has joined #openstack-infra | 15:25 | |
| timrc | I started created a visualizer for our nodepool lifecycle events... it's far from complete, it may never be complete.. wondering if anyone else here is doing similar? http://15.125.65.125/ | 15:25 |
| openstackgerrit | Merged openstack-infra/elastic-recheck: Add fingerprint for XStatic-Angular-Irdragndrop dependency failure https://review.openstack.org/169156 | 15:26 |
| fungi | timrc: is that pulling from our nodepool stats counters in graphite.openstack.org? | 15:27 |
| timrc | fungi, No that is parsing debug.logs on our own instance. | 15:28 |
| fungi | ahh | 15:28 |
| fungi | timrc: well, nodepool can be configured to emit stats events to a statsd | 15:29 |
| *** wicope has joined #openstack-infra | 15:29 | |
| fungi | timrc: so no need to parse the logs to get that data | 15:29 |
| timrc | I might do as cody-somerville suggested privated to do box and whisker. | 15:29 |
| *** mrmartin has quit IRC | 15:29 | |
| *** stevemar has joined #openstack-infra | 15:29 | |
| *** jlibosva has quit IRC | 15:30 | |
| jeblair | timrc: and if a stat is missing from nodepool statsd, should be easy to add | 15:30 |
| timrc | fungi, Cool. I haven't looked much into that. It was convenient that our logs rotate daily and the events with timestamps were there. | 15:30 |
| fungi | and if there are events you find in the log which are useful to trend and don't have corresponding wrappers to emit stats, then yeah | 15:30 |
| fungi | what jeblair just said | 15:30 |
| *** maurosr has quit IRC | 15:30 | |
| clarkb | the neat thing with graphite is it can likely do much of the graphing for you via its api (but it will also give you raw data in json iirc if doing it yourself is necessary) | 15:31 |
| mordred | I will echo the suggestion to add stats counters to nodepool - honestly being able to have a live graph of those sorts of times would be fantastic | 15:32 |
| timrc | clarkb, Yeah d3 made me want to punch a wall more than once. | 15:32 |
| timrc | I thought the graphs graphite generate were relatively static though. I eventually want to make these more interactive. | 15:33 |
| mordred | in fact, you could really probably figure out a way to do it directly from TaskManager - since we already log the start/stop there | 15:33 |
| *** dizquierdo has quit IRC | 15:33 | |
| timrc | mordred, I'll look into what's there today and attempt to use it as a data source if everything I need is in place otherwise I'll submit a patch. | 15:33 |
| mordred | there are other frontends that can work with the graphite data in dynamic ways | 15:33 |
| timrc | Parsing logs and creating csv's is not my idea of a good time. | 15:34 |
| *** scheuran has quit IRC | 15:35 | |
| mordred | timrc: http://jondot.github.io/graphene/ <-- I hear good things about this | 15:35 |
| *** maurosr has joined #openstack-infra | 15:35 | |
| jeblair | they aren't dynamic in the sense of flashy javascript, but you can make any graph via url, so it's quite easy to use | 15:35 |
| jeblair | so we can still easily graph post-facto any of the data we collect | 15:36 |
| *** YorikSar has quit IRC | 15:37 | |
| *** asselin has quit IRC | 15:38 | |
| *** jlanoux has quit IRC | 15:38 | |
| fungi | timrc: yeah, parsing logs to generate stats is the sort of thing you do when you can't easily modify the application to give you the stats in a usable format | 15:38 |
| *** gondoi is now known as zz_gondoi | 15:39 | |
| *** thedodd has joined #openstack-infra | 15:40 | |
| jeblair | yolanda: what have you learned about the performance of 169235? | 15:41 |
| timrc | mordred, That does look pretty shiny. | 15:41 |
| openstackgerrit | Merged openstack-infra/project-config: Add functional test job for python-saharaclient https://review.openstack.org/167641 | 15:42 |
| *** harlowja_at_home has joined #openstack-infra | 15:42 | |
| *** sabeen3 has joined #openstack-infra | 15:42 | |
| openstackgerrit | Clark Boylan proposed openstack-infra/nodepool: Support per provider image build settings https://review.openstack.org/126747 | 15:43 |
| openstackgerrit | Clark Boylan proposed openstack-infra/nodepool: Remove qemu-img-options https://review.openstack.org/169383 | 15:43 |
| *** MaxV has quit IRC | 15:43 | |
| clarkb | jeblair: ^ that doesn't detect when a new image is necessary due to a format change yet, but is that generally what you had in mind? | 15:43 |
| *** sabeen3 has quit IRC | 15:43 | |
| *** sabeen has quit IRC | 15:43 | |
| *** MaxV has joined #openstack-infra | 15:44 | |
| *** sabeen has joined #openstack-infra | 15:44 | |
| *** harlowja_at_home has quit IRC | 15:44 | |
| *** nmagnezi has quit IRC | 15:44 | |
| fungi | heading out for a lunch walk. back soon | 15:44 |
| anteaya | enjoy | 15:45 |
| *** mestery has quit IRC | 15:45 | |
| *** dprince has quit IRC | 15:46 | |
| jogo | clarkb: nice job with the 64bit live migration! | 15:46 |
| clarkb | jogo: the libvirt stop/start vs restart really tripped me up... Would've been done much sooner if I Had sorted that out | 15:47 |
| *** skolekonov has quit IRC | 15:47 | |
| *** MaxV has quit IRC | 15:48 | |
| *** ghostpl_ has quit IRC | 15:49 | |
| *** ildikov_ has quit IRC | 15:49 | |
| *** dustins has quit IRC | 15:49 | |
| jeblair | clarkb: why is DIB_IMAGE_CACHE a per-provider env var? | 15:49 |
| *** ghostpl_ has joined #openstack-infra | 15:50 | |
| *** arxcruz has quit IRC | 15:50 | |
| clarkb | jeblair: its currently per image | 15:51 |
| clarkb | jeblair: I thought you wanted to pull that stuff up out of images | 15:51 |
| *** doug-fish has joined #openstack-infra | 15:51 | |
| clarkb | it doesn't have to be though. The current code supports it either in the image or in the provider | 15:51 |
| jeblair | clarkb: i think things that affect the content of the filesystem created in the image should be attached to the diskimage thing, and things that affect the output format used by a provider should be attached to the provider... | 15:53 |
| jeblair | clarkb: i guess the cache dir doesn't fall into either category since it shouldn't affect either of those things, right? | 15:53 |
| clarkb | correct | 15:53 |
| *** ala_ has quit IRC | 15:53 | |
| jeblair | clarkb: so if we had to adjust those definitions to accomodate it, i figure it's more related to the content than the output format, inasmuch as it is used by the elements specified in the diskimage section... | 15:54 |
| jeblair | clarkb: so i'd lean toward saying that should be in the diskimage section | 15:54 |
| jeblair | clarkb: a second, related question -- why is compat an env variable? i thought that was a qemu command line option? | 15:54 |
| mordred | jeblair: which change? | 15:54 |
| mordred | jeblair: it gets passed via env var | 15:55 |
| jeblair | mordred: 126747 | 15:55 |
| clarkb | jeblair: it is a qemu-img command line option that is set via dib's env | 15:55 |
| clarkb | jeblair: which is why my next change in that stack is to remove the special qemu-img-options nodepool config since its basically redundant (we haven't used it that way either) | 15:55 |
| jeblair | clarkb, mordred: why is that not a command line argument to dib? | 15:56 |
| jeblair | clarkb, mordred: i mean, you tell dib what _major_ format to use for the output via command line, why would you not tell it what minor format to use? | 15:57 |
| yolanda | jeblair, mordred, so that's etherpad for benchmarking , working on it right now | 15:57 |
| yolanda | https://etherpad.openstack.org/p/Nodepool_benchmarks | 15:57 |
| mordred | jeblair: I think because it only is related to one of the image output formats - so if you did -t vhd the QEMU options wouldnt' be very meaninful? | 15:58 |
| yolanda | the traditional approach is over a more general period and a real usage, and the priorityqueue is for a shorter period, and on my test account, so it's not 100% reliable now | 15:58 |
| clarkb | jeblair: it can be; however, dib has a ton of options that can only be set via env var so to me it makes sense to use env vars as the preferred option there | 15:58 |
| jeblair | mordred: that is a problem many software systems have solved without resorting to environment variables | 15:58 |
| mordred | jeblair: dib drives a substantial number of things via env vars | 15:58 |
| mordred | jeblair: so it is consistent with overall operation | 15:58 |
| greghaynes | generally disk-image-create config is done by command line option and element config is done via env var | 15:59 |
| greghaynes | because otherwise wed need to invent some other way to tunnel that config through | 15:59 |
| jeblair | yolanda: cool, so i think that running equivalent experiments would be helpful as well as overall throughput figures | 15:59 |
| mordred | greghaynes: well, isnt this a disk-image-create config? | 15:59 |
| jeblair | greghaynes, mordred: i'm really confused. i would think that... exactly what mordred said. | 15:59 |
| mordred | greghaynes: it's not really element config | 15:59 |
| yolanda | yes, i know, i need to dedicate a bit more time | 15:59 |
| mordred | greghaynes: should I make a patch? | 15:59 |
| greghaynes | yes, I think that wasnt entirely thought through here ;) | 15:59 |
| greghaynes | yes | 15:59 |
| mordred | k | 16:00 |
| * mordred makes patch | 16:00 | |
| clarkb | hold on | 16:00 |
| clarkb | ignore dib for a second | 16:00 |
| *** emagana has quit IRC | 16:00 | |
| mordred | echo " --qemu-img-options -- option flags to be passed directly to qemu-img." | 16:00 |
| clarkb | mordred: yes it supports that see above | 16:00 |
| clarkb | what I am trying to say is that building images necessarily requires you set a half dozen env things | 16:00 |
| jeblair | clarkb, mordred: and nodepool uses that | 16:01 |
| mordred | clarkb: ah - you're saying that since we're setting env vars anyway | 16:01 |
| clarkb | so rather than support a half dozen nodepool config options that will only grow, it makes sense for nodepool to pass through an env setup | 16:01 |
| clarkb | mordred: yes | 16:01 |
| mordred | we might as well set all the things that way | 16:01 |
| clarkb | yes | 16:01 |
| mordred | I can see that | 16:01 |
| jeblair | makes sense, but i don't agree in this case | 16:01 |
| annegentle | get ready for a really out there git question. Is it possible to interact with git.openstack.org so that I could extract the exact URL of a source doc file? Such as http://git.openstack.org/cgit/openstack/openstack-manuals/tree/doc/playground-user-guide/source/backup_db_incremental.rst | 16:01 |
| *** claudiub has quit IRC | 16:01 | |
| *** tonytan4ever has quit IRC | 16:01 | |
| jeblair | clarkb: i believe that the diskimage entry in the config should support env vars | 16:01 |
| jeblair | clarkb: but there should be no per-provider env vars | 16:02 |
| zaro | morning | 16:02 |
| clarkb | jeblair: ok thats fine, I can just pull the per provider envvars stuff out | 16:02 |
| jeblair | clarkb: the provider should only specify the type and any additional options (like the qemu compat thing) related to that | 16:02 |
| *** dboik_ has joined #openstack-infra | 16:02 | |
| jeblair | clarkb: cool | 16:02 |
| clarkb | I do not think we should provide any qemu compat options | 16:02 |
| jeblair | er, don't we need to? | 16:02 |
| clarkb | because next week it will be vhd compat options | 16:02 |
| clarkb | and the week after tarball flags and so on | 16:03 |
| mordred | honestly, the qemu compat thing is about teh format | 16:03 |
| *** tsg_ has joined #openstack-infra | 16:03 | |
| clarkb | instead we should just use the env to do this | 16:03 |
| *** mestery has joined #openstack-infra | 16:03 | |
| mordred | I'd go a completely different way and just have hardcoded logic that says "if format = qcow2, then qemu-img-options=compat" | 16:03 |
| *** e0ne is now known as e0ne_ | 16:03 | |
| mordred | but I'm wrongthinking many times | 16:03 |
| * greghaynes was suggesting this | 16:04 | |
| mordred | that we have to specify this compat setting is a bug in something | 16:04 |
| mordred | not a faeture | 16:04 |
| jeblair | clarkb: that's too generalized for nodepool to make decisions about when different images need to be made | 16:04 |
| jeblair | greghaynes, mordred: do we know 0.10 works everywhere? | 16:04 |
| clarkb | jeblair: we do not | 16:04 |
| mordred | jeblair: well, there is only one cloud we have access to that uses qcow2 | 16:04 |
| *** emagana has joined #openstack-infra | 16:04 | |
| mordred | oh - wait - the tripleo cloud uses qcow2 | 16:05 |
| jeblair | i think tripleo and hpcloud yeah | 16:05 |
| mordred | greghaynes: any chance you know if uploading an image to the tripleo cloud works if the image was built with 0.10? | 16:05 |
| *** dboik has quit IRC | 16:05 | |
| *** tnovacik has quit IRC | 16:05 | |
| clarkb | jeblair: its not too generalized because dib is making those decisions | 16:05 |
| greghaynes | mordred: I havent done it, no | 16:05 |
| clarkb | jeblair: so the qemu options are only valid when making qcow2, vhd stuff would only be valid for vhd images, tar flags would only be valid for tarballs | 16:05 |
| mordred | blerg | 16:05 |
| jeblair | clarkb: i'm saying that if you pass in any env variables, nodepool doesn't know which ones affect image output format | 16:05 |
| clarkb | so you can set all of them in the env and it will just work (tm) | 16:06 |
| jeblair | clarkb: only if we can set one value and have it work everywhere | 16:06 |
| jeblair | clarkb: in which case we may as well go with greghaynes idea and just set that value | 16:06 |
| * SpamapS catches up | 16:06 | |
| mordred | I say we _start_ with just setting the value until we actually have a real need to set it to different things | 16:06 |
| mordred | we can add extra configurability later | 16:07 |
| jeblair | mordred: we allow setting it now. you advocate for removing it. | 16:07 |
| mordred | I advocate for hardcoding it | 16:07 |
| jeblair | sure | 16:07 |
| anteaya | annegentle: I don't see that anyone else has replied to your question, I'm not sure what you mean by interact | 16:07 |
| annegentle | anteaya: ah, thanks for thinking through it with me. So is there an API for git.openstack? | 16:08 |
| *** unicell1 has quit IRC | 16:08 | |
| anteaya | annegentle: it is cgit so if cgit has an API then yes, but we didnt' set one up ourselves | 16:08 |
| annegentle | anteaya: I want to extract (did I say interact, sorry for imprecise language) a URL that poitns to a source file with preferrably the info from a conf.py file | 16:08 |
| *** e0ne_ has quit IRC | 16:08 | |
| anteaya | and I dont' recall that question before | 16:08 |
| mordred | anteaya: you can just do string substitution | 16:08 |
| mordred | gah | 16:08 |
| annegentle | anteaya: ok I'll look for cgit | 16:08 |
| mordred | annegentle: | 16:08 |
| mordred | there is no api | 16:09 |
| mordred | but | 16:09 |
| *** medieval1 has joined #openstack-infra | 16:09 | |
| *** jistr has quit IRC | 16:09 | |
| jeblair | well, technically, there is a fully featured read-only REST API | 16:09 |
| mordred | the base will always be http://git.openstack.org/cgit/ | 16:09 |
| mordred | jeblair: :) | 16:09 |
| annegentle | :) | 16:09 |
| mordred | annegentle: then you'll always have the project name, then the word tree, then the path | 16:09 |
| annegentle | are there endpoints? | 16:09 |
| *** carl_baldwin_ has quit IRC | 16:09 | |
| SpamapS | Maybe we need to make image formats like elements | 16:09 |
| annegentle | mordred: ah, so glue it together with what I know | 16:09 |
| mordred | annegentle: http://git.openstack.org/cgit/{projectname}/tree/{path} | 16:09 |
| mordred | annegentle: yes | 16:09 |
| annegentle | mordred: lovely, you all know you are my rubber duckies that I talk to to solve my programming challenges :) | 16:10 |
| SpamapS | I'm not sure that solves this.. | 16:10 |
| annegentle | anteaya: sometimes I just need help articulating desired outcome, thank you! | 16:10 |
| mordred | annegentle: in your example above- projectname=openstack/openstack-manuals and path=doc/playground-user-guide/source/backup_db_incremental.rst | 16:10 |
| SpamapS | but I realized just now part of our frustration with this is the interface for image making is entirely different for everything else that we put into diskimage-builder | 16:10 |
| mordred | just to be clear | 16:10 |
| *** Sukhdev has joined #openstack-infra | 16:11 | |
| annegentle | mordred: cool, and I think I can input projectname in the conf.py for doc builds | 16:11 |
| *** whoops has joined #openstack-infra | 16:11 | |
| annegentle | I think this scales across lots of doc builds | 16:11 |
| mordred | SpamapS: you'd have a vhd element and a qcow element? and if you specified both you'd get both things? that actually seems not great to me | 16:11 |
| jeblair | clarkb, mordred, greghaynes, SpamapS: let's address these questions one at a time: regardless of mechanism, do we want qemu compat to be configurable in nodepool? | 16:11 |
| anteaya | annegentle: always happy to listen to you while you solve it yourself | 16:11 |
| mordred | jeblair: ++ to starting with that question | 16:11 |
| mordred | jeblair: I vote "no, not until we have a usecase for actually using it" | 16:12 |
| mordred | largely because I want one image as much as possible | 16:12 |
| mordred | so if we had two qcow2 clouds | 16:12 |
| greghaynes | sounds like there was a question of whether there is a single value that would work, seems like the issue we were running into can be worked around with a single value but it still is probably good to verify that | 16:12 |
| mordred | I'd rather build one qcow2 cloud and upload to both | 16:13 |
| mordred | otherwise we're going to need to put compat into the filename or something | 16:13 |
| greghaynes | yep, agreed | 16:13 |
| mordred | beacuse we don't want to have file-per-provider (defeats the point) | 16:13 |
| clarkb | I am starting to come around to mordred here | 16:13 |
| jeblair | mordred: true, we would need to do that, though that's not super-hard. we already have the format in the filename, we can add the compat flag too. | 16:14 |
| clarkb | basically one qcow2 with the least common denominator version until necessary to do otherwise | 16:14 |
| mordred | jeblair: yes. we could also invent fake formats | 16:14 |
| *** dprince has joined #openstack-infra | 16:14 | |
| mordred | we coould make a qcow2.0.10 format can a qcow2 format | 16:14 |
| jeblair | and to be clear, we're still talking identical images, just multiple files of them in different formats. | 16:14 |
| mordred | but I thnk that gets lame | 16:14 |
| jeblair | yeah | 16:14 |
| mordred | yes | 16:14 |
| *** e0ne has joined #openstack-infra | 16:14 | |
| greghaynes | The only downside I can think of is that were not testing qcow 0.20 and we dont get new 0.20 features, and I think both of those dont really matter to us | 16:15 |
| mordred | i agree | 16:15 |
| jeblair | okay, so it seems like we're cool with just lowest-common-denominatoring qcow | 16:15 |
| jeblair | and we have an idea of what the future might look like if we need to change our minds on that :) | 16:15 |
| mordred | ++ | 16:15 |
| mordred | SpamapS: now - for your element thing ... let me mention a think I suggested to greghaynes | 16:15 |
| SpamapS | mordred: so IMO, if we have two versions of qcow2 that are sometimes not compatible, those are not both "qcow2". However it also sounds like we are building a bridge to nowhere by pursuing this question? | 16:15 |
| jeblair | so having said that... | 16:15 |
| SpamapS | Ah ok lag fail. What jeblair said. ;) | 16:16 |
| jeblair | i'd propose that we define env vars only in the nodepool diskimage section, define image types only in the provider section, and not specify qemu compat options anywhere | 16:16 |
| *** ivar-lazzaro has joined #openstack-infra | 16:16 | |
| *** Guest1663 has quit IRC | 16:16 | |
| jeblair | do folks agree with that ^ ? | 16:16 |
| *** ivar-lazzaro has quit IRC | 16:17 | |
| clarkb | yes | 16:17 |
| mordred | jeblair: yes | 16:17 |
| greghaynes | SGTM | 16:17 |
| jeblair | woot, i think that concludes nodepool bikeshedding. i will write that up in the review. feel free to proceed with dib bikeshedding :) | 16:17 |
| *** ivar-lazzaro has joined #openstack-infra | 16:17 | |
| clarkb | fwiw we can make the qemu version thing work today as is without any of that sillyness above, but that is because its set per diskimage and not per provider | 16:18 |
| mordred | SpamapS: I would like for dib to be able to not rebuild a new image, but instead just do a followup image conversion | 16:18 |
| *** claudiub has joined #openstack-infra | 16:19 | |
| mordred | SpamapS: as in "disk-image-create -n -t qcow2,raw -o ubuntu ubuntu-minimal vm infra ; disk-image-create --no-rebuild -n -t qcow2,raw -o ubuntu ubuntu-minimal vm infra" | 16:19 |
| mordred | SpamapS: gah | 16:19 |
| mordred | SpamapS: as in "disk-image-create -n -t qcow2,raw -o ubuntu ubuntu-minimal vm infra ; disk-image-create --no-rebuild -n -t qcow2,raw,vhd -o ubuntu ubuntu-minimal vm infra" | 16:19 |
| SpamapS | mordred: so be a thin layer in front of the vhd tools? | 16:19 |
| mordred | SpamapS: so the second command would notice that it already has an ubuntu.qcow2 and an ubuntu.raw so it does not start from scratch and would do the conversion it would have done the first time if I'd requested one | 16:20 |
| greghaynes | on the second run, yes | 16:20 |
| SpamapS | I mean, realistically you don't need the element list in the second invocation. | 16:20 |
| SpamapS | --no-rebuild would just jump right to the image output step right? | 16:20 |
| mordred | you don't - but if I don't have the images, I would expect it to build | 16:20 |
| mordred | SpamapS: "disk-image-create -n -t qcow2,raw -o ubuntu ubuntu-minimal vm infra ; rm ubuntu* ; disk-image-create --no-rebuild -n -t qcow2,raw,vhd -o ubuntu ubuntu-minimal vm infra" | 16:20 |
| mordred | would build thigns twice | 16:20 |
| jeblair | clarkb: i'm not sure i agree with that -- that still does not provide a way to map different qemu compat options to the qcow2 builds of a given image that might be used with multiple providers. | 16:21 |
| greghaynes | Yea, so those are my questions - what level of cache invalidation do we want to do and do we want to store a fully copy on disk of the image | 16:21 |
| mordred | SpamapS: basically, make the second form idempotent | 16:21 |
| *** bswartz has quit IRC | 16:21 | |
| SpamapS | mordred: so we use the argument to -o as a cache key basically. | 16:21 |
| *** derekh has quit IRC | 16:21 | |
| mordred | SpamapS: yeah, basically | 16:22 |
| mordred | SpamapS: this came to mind in my brain | 16:22 |
| mordred | SpamapS: because I'd like to at some point make our own infra base images, and I'd like to have an ansible playbook that I can just run to make images as needed | 16:22 |
| clarkb | jeblair: oh right, we would have two or more different source images | 16:22 |
| *** ildikov_ has joined #openstack-infra | 16:22 | |
| mordred | SpamapS: once an image is made, I NEVER want to remake it | 16:22 |
| mordred | SpamapS: but, it's conceivable I might grow an additional cloud that wants an additional format | 16:22 |
| mordred | and it would be nice to be able to just add that format to the playbook | 16:23 |
| SpamapS | if [ "${rebuild_image}" = "1" ] || [ ! -e ${OUTPUT_BASE}.raw ] ; then do_all_the_build_things ; fi | 16:23 |
| mordred | and have the new cloud get properly formatted image | 16:23 |
| greghaynes | I think thatll work fine as long as we make it a --update-formats or similar rather than advertise it as a smart caching layer | 16:23 |
| mordred | greghaynes: yah | 16:23 |
| SpamapS | mordred: that seems like a.. rare occurrance. | 16:23 |
| mordred | SpamapS: yes. if we never write this I will not die | 16:23 |
| SpamapS | mmk. I see all the merits, but place it in the Wishlist status unless there's some giant optimization I'm missing. | 16:24 |
| greghaynes | mordred: actually, one issue is it wont help with the img format thing | 16:25 |
| mordred | SpamapS: the main thing is that otherwise I'll have to add image conversion scripting to the ansible role which is sad because dib already knows how to go from raw to each of the formats my clouds want | 16:25 |
| greghaynes | mordred: because youll have to use a different name for the image | 16:25 |
| mordred | greghaynes: what do you mean? | 16:25 |
| greghaynes | er, qcow2 options | 16:25 |
| mordred | greghaynes: I'm not talking qcow2 options - totally agree it won't help | 16:25 |
| greghaynes | ok then | 16:25 |
| mordred | but seriously - this is super wishlist -it's not really a problem right now | 16:25 |
| mordred | I just wanted to mention it because of SpamapS idea to make image formats more element like | 16:26 |
| SpamapS | mordred: I'm confused how that would need to be automated. You add clouds, what, once very 3 years? | 16:26 |
| SpamapS | also btw, I may have accidentally shamed Cisco into wanting to provide a rack of servers for infra on Thursday night. ;) | 16:26 |
| mordred | SpamapS: I just grabbed myself a dreamhost account over the weekend | 16:26 |
| mordred | SpamapS: nice | 16:26 |
| clarkb | that reminds me I have runabove instances I was playing with (not in relation to doing infra workloads) | 16:27 |
| mordred | SpamapS: but yeah - like I said - not a problem right now - I mainly wanted to inject the usecase in your head so that didn't rework somethign to make doing that _harder_ | 16:27 |
| *** mtanino has joined #openstack-infra | 16:27 | |
| SpamapS | Also we need to update the infra slides to include the big tent. | 16:27 |
| *** ivar-laz_ has joined #openstack-infra | 16:27 | |
| SpamapS | mordred: cool. Image formats as elements would actually make it easier. | 16:28 |
| mordred | SpamapS: I believe you | 16:28 |
| *** yfried is now known as yfried|afk | 16:29 | |
| *** yamahata has joined #openstack-infra | 16:29 | |
| * anteaya heads out for a walk | 16:29 | |
| *** ivar-lazzaro has quit IRC | 16:30 | |
| SpamapS | mordred: but anyway, it's not a priority as image formats also come along about once every 3 years. ;) | 16:31 |
| cody-somerville | mordred: You should take a look at live-build from the Debian live project. It has a really nice architecture. | 16:31 |
| mordred | cody-somerville: why would I do that? | 16:31 |
| SpamapS | cody-somerville: oh no you diii'nt ;) | 16:31 |
| SpamapS | cody-somerville: actually since you're paying attention... does live-build use something like schroot to build the OS level chroots? | 16:31 |
| *** yfried|afk is now known as yfried | 16:31 | |
| SpamapS | One thing that has constantly made me go "hmmmmm" in dib is whether or not we should make it use schroot.. which would give us some really big optimizations for machines that do lots of builds (like using lvm and union filesystems) | 16:32 |
| *** dhritishikhar_ has joined #openstack-infra | 16:32 | |
| SpamapS | but then I go "optimization trap" and move on. | 16:32 |
| mordred | cody-somerville: ok. I looked. it has terrible documentation and seems to have no interesting features | 16:33 |
| *** carl_baldwin has joined #openstack-infra | 16:33 | |
| mordred | cody-somerville: it's also a system designed to produce something that i don't want to produce | 16:33 |
| mordred | cody-somerville: so yeah, that was fun to look at, I now know that live-build is not useful to me | 16:34 |
| openstackgerrit | Clark Boylan proposed openstack-infra/nodepool: Support per provider image build settings https://review.openstack.org/126747 | 16:34 |
| openstackgerrit | Clark Boylan proposed openstack-infra/nodepool: Just set qcow2 version https://review.openstack.org/169383 | 16:34 |
| clarkb | jeblair: mordred greghaynes ^ I think those two patches address all of the above stuff including detection of missing image format requiring a rebuild | 16:34 |
| *** unicell has joined #openstack-infra | 16:34 | |
| * greghaynes isnt exactly proud of the DIB docs yet | 16:34 | |
| clarkb | its likely I will need to write more tests though | 16:34 |
| cody-somerville | mordred: I was just looking at the website as well. It used to have a very accessible manual. | 16:34 |
| cody-somerville | mordred: but I was suggesting looking at it from architecture perspective | 16:35 |
| SpamapS | greghaynes: they're getting MUCH better every day. :) | 16:35 |
| *** ujuc has quit IRC | 16:35 | |
| greghaynes | SpamapS: Speaking of: https://review.openstack.org/#/c/168727 | 16:35 |
| *** baoli has quit IRC | 16:35 | |
| *** ssam2 has quit IRC | 16:36 | |
| mordred | cody-somerville: sure - I don't think we're at the stage where we're really looking at architectures for build systems though | 16:36 |
| SpamapS | greghaynes: yes, 3rd on my list ATM | 16:37 |
| *** baoli has joined #openstack-infra | 16:37 | |
| greghaynes | \O/ | 16:37 |
| *** annegentle has quit IRC | 16:40 | |
| *** mpaolino has quit IRC | 16:40 | |
| mordred | greghaynes: so - now that we have all of the unittesting - we should maybe live-test the resulting images on rax and hp again :) | 16:42 |
| *** aduarte has joined #openstack-infra | 16:42 | |
| *** YorikSar has joined #openstack-infra | 16:42 | |
| greghaynes | Sounds like a good idea | 16:42 |
| *** dangers is now known as dangers_away | 16:44 | |
| openstackgerrit | Davide Michelino proposed openstack-infra/project-config: Add Networking-portforwarding project https://review.openstack.org/169332 | 16:45 |
| *** sputnik13 has joined #openstack-infra | 16:45 | |
| *** harlowja_away is now known as harlowja_ | 16:45 | |
| *** SumitNaiksatam has quit IRC | 16:46 | |
| *** Swami has joined #openstack-infra | 16:46 | |
| *** hasharShopping has quit IRC | 16:47 | |
| *** jaypipes has joined #openstack-infra | 16:48 | |
| *** jlanoux has joined #openstack-infra | 16:48 | |
| openstackgerrit | Andreas Jaeger proposed openstack-infra/reviewstats: Update docs and security-docs teams https://review.openstack.org/168290 | 16:50 |
| *** AJaeger_ has joined #openstack-infra | 16:50 | |
| *** dims_ has quit IRC | 16:51 | |
| *** lhcheng has joined #openstack-infra | 16:51 | |
| *** dims_ has joined #openstack-infra | 16:51 | |
| *** mwagner_lap has joined #openstack-infra | 16:52 | |
| *** yfried is now known as yfried|afk | 16:52 | |
| *** MaxV has joined #openstack-infra | 16:54 | |
| *** MarkAtwood has quit IRC | 16:54 | |
| *** mpaolino has joined #openstack-infra | 16:54 | |
| *** e0ne is now known as e0ne_ | 16:56 | |
| *** bswartz has joined #openstack-infra | 16:57 | |
| *** tonytan4ever has joined #openstack-infra | 17:00 | |
| openstackgerrit | Luigi Toscano proposed openstack-infra/project-config: Make saharaclient-dsvm-functional voting https://review.openstack.org/169426 | 17:04 |
| *** andreykurilin_ has joined #openstack-infra | 17:05 | |
| *** claudiub has quit IRC | 17:05 | |
| *** yfried|afk is now known as yfried | 17:06 | |
| *** lhcheng has left #openstack-infra | 17:06 | |
| *** e0ne_ is now known as e0ne | 17:06 | |
| *** tkelsey has quit IRC | 17:06 | |
| *** baoli has quit IRC | 17:08 | |
| *** tqtran has joined #openstack-infra | 17:09 | |
| *** dimtruck is now known as zz_dimtruck | 17:09 | |
| *** radez_g0n3 is now known as radez | 17:10 | |
| *** dustins has joined #openstack-infra | 17:12 | |
| *** tkelsey has joined #openstack-infra | 17:12 | |
| *** dtantsur is now known as dtantsur|afk | 17:14 | |
| *** baoli has joined #openstack-infra | 17:14 | |
| *** patrickeast has joined #openstack-infra | 17:14 | |
| *** SumitNaiksatam has joined #openstack-infra | 17:16 | |
| *** tkelsey has quit IRC | 17:17 | |
| *** johnthetubaguy is now known as zz_johnthetubagu | 17:18 | |
| *** funzo has joined #openstack-infra | 17:18 | |
| *** baoli has quit IRC | 17:19 | |
| *** wayneeseguin has joined #openstack-infra | 17:19 | |
| openstackgerrit | Joshua Harlow proposed openstack-infra/project-config: Remove the g-r requirements check https://review.openstack.org/169434 | 17:19 |
| openstackgerrit | K Jonathan Harker proposed openstack-infra/jeepyb: Log project names as the projects are processed https://review.openstack.org/169435 | 17:19 |
| *** baoli has joined #openstack-infra | 17:20 | |
| *** baoli has quit IRC | 17:20 | |
| *** otter768 has joined #openstack-infra | 17:21 | |
| cody-somerville | http://lnav.org/ <-- this looks pretty cool. I wonder how well it would work with local Open Stack logs or something. | 17:22 |
| *** achanda has joined #openstack-infra | 17:23 | |
| *** dboik_ has quit IRC | 17:25 | |
| *** otter768 has quit IRC | 17:26 | |
| *** dboik has joined #openstack-infra | 17:26 | |
| *** juggler has joined #openstack-infra | 17:26 | |
| SpamapS | cody-somerville: looks like an improvement on ccze :) | 17:27 |
| *** markus_z has quit IRC | 17:30 | |
| jogo | clarkb: how do I get the logs from the second node in aiopcpu into logstash? | 17:30 |
| jogo | clarkb: since directory is logs/*/n-cpu etc | 17:31 |
| jogo | instead of logs/n-cpu | 17:31 |
| *** mpavone has quit IRC | 17:32 | |
| clarkb | will need to update the log gearman client to get jobs submitted for them | 17:32 |
| *** e0ne has quit IRC | 17:33 | |
| jogo | clarkb: hmm all I know about is modules/openstack_project/files/logstash/jenkins-log-client.yaml | 17:33 |
| *** Sukhdev has quit IRC | 17:35 | |
| greghaynes | clarkb: I am noticing that hpcloud vs my dev box makes a huge difference in time to do qemu image conversions, which is probably why youall wanted that short-circuit so badly | 17:35 |
| *** tkelsey has joined #openstack-infra | 17:36 | |
| *** ghostpl_ has quit IRC | 17:37 | |
| mordred | greghaynes: :) | 17:37 |
| *** sdake_ has joined #openstack-infra | 17:37 | |
| clarkb | well we have to convert | 17:38 |
| clarkb | I wanted to stop copying files unnecessarily | 17:38 |
| *** rhe00 has quit IRC | 17:39 | |
| *** jerryz_ has quit IRC | 17:39 | |
| jogo | clarkb: so not sure how to update the log gearman client | 17:40 |
| *** tkelsey has quit IRC | 17:40 | |
| *** annegentle has joined #openstack-infra | 17:41 | |
| *** doude has quit IRC | 17:41 | |
| *** sdake has quit IRC | 17:41 | |
| clarkb | jogo: https://git.openstack.org/cgit/openstack-infra/puppet-log_processor/tree/files/log-gearman-client.py is the script | 17:42 |
| clarkb | jogo: however, I think it would be great if we could have a single n-cpu log file | 17:42 |
| clarkb | though every time we try to use syslog it explodes on us so probably not worth the effort | 17:42 |
| *** ayoung is now known as hopig | 17:43 | |
| *** baoli has joined #openstack-infra | 17:43 | |
| *** hopig is now known as ayoung | 17:43 | |
| *** bswartz has quit IRC | 17:44 | |
| clarkb | jogo: I think our options are A) always put logs in known location B) grab index file and determine where subnode logs are C) use some sort of centralized log system to coalesce the logs | 17:46 |
| *** annegentle has quit IRC | 17:46 | |
| mordred | clarkb: d) give up and drink beer | 17:47 |
| clarkb | mordred: too early in the day for that | 17:48 |
| fungi | options a,b and c can also be amended with "and drink beer" | 17:48 |
| *** melwitt has joined #openstack-infra | 17:48 | |
| *** sdake has joined #openstack-infra | 17:48 | |
| fungi | i just always assumed it was an implied suffix | 17:48 |
| openstackgerrit | Valeriy Ponomaryov proposed openstack-infra/project-config: Add tempest + postgres job for manila https://review.openstack.org/165935 | 17:50 |
| *** pelix has quit IRC | 17:50 | |
| SpamapS | fungi: only 'and drink' is implied as it is an abstract method. You have to implement the actual method and populate with cups, location, company, and beverage of choice | 17:51 |
| clarkb | Shrews: for 167371 what do you think about having shade do the logical "make networking work" which will create subnet, network, and router in one go | 17:52 |
| mordred | SpamapS: are you sure you don't want to implement a drink factory function that produces factories that produce drinks? | 17:52 |
| fungi | got it. i pretty much just used it as a class decorator | 17:52 |
| clarkb | Shrews: as a user of shade I never want to think about routers ever | 17:52 |
| sdake | hey folks, rhallisey has created a channel for #kolla developers to congregate | 17:52 |
| *** sdake_ has quit IRC | 17:52 | |
| mordred | clarkb: I think we could add a function like that | 17:52 |
| clarkb | Shrews: but not sure if this would be a good building block for what I decsribe or if we should change 167371 at all for that? | 17:52 |
| sdake | but I'm concerned the channel wasn't registered propery | 17:52 |
| mordred | clarkb: similar to auto_ip | 17:52 |
| mordred | for create_server | 17:52 |
| sdake | he said he registered it with chanserv | 17:52 |
| *** rhallisey has joined #openstack-infra | 17:52 | |
| *** doude has joined #openstack-infra | 17:52 | |
| fungi | sdake: checking now | 17:53 |
| rhallisey | sdake, hey | 17:53 |
| sdake | what is the appropriate technique to update teh chanserv settings? | 17:53 |
| sdake | hey rhallisey | 17:53 |
| fungi | sdake: rhallisey: /msg chanserv access #kolla list | 17:53 |
| fungi | sdake: rhallisey: that looks fine to me | 17:53 |
| sdake | fungi I don't have access | 17:53 |
| Shrews | clarkb: that would assume everyone wants the same kind of network, no? | 17:53 |
| mordred | Shrews: there is a basic kind of network that everyone wants | 17:54 |
| mordred | Shrews: similar to the "this is a server and it has an IP" | 17:54 |
| sdake | fungi how do we get the status bot in the channel? | 17:54 |
| Shrews | mordred: really? does everyone want the same subnet IP range??? | 17:54 |
| mordred | if you want a different kind of thing, the lower-level primitives are there | 17:54 |
| clarkb | Shrews: not the same range but thats just an option | 17:54 |
| fungi | sdake: rhallisey: see http://ci.openstack.org/irc.html | 17:54 |
| mordred | Shrews: maybe that's an argument to the function | 17:54 |
| *** doude has quit IRC | 17:54 | |
| clarkb | Shrews: and if you need super fanc use the primitives. But imo its a major bug that a user has to configure 3 different resources to make logical networking work | 17:54 |
| Shrews | mordred: what about multiple subnets??? | 17:55 |
| mordred | clarkb: ++ | 17:55 |
| clarkb | Shrews: no | 17:55 |
| mordred | Shrews: use the more complex things | 17:55 |
| clarkb | openstack needs some opinions | 17:55 |
| mordred | this is "I just want wroking network" | 17:55 |
| mordred | if you want fancy, you can always opt in | 17:55 |
| clarkb | working network is one subnet, one network, and a router | 17:55 |
| *** bswartz has joined #openstack-infra | 17:55 | |
| mordred | yup | 17:55 |
| *** pblaho has quit IRC | 17:55 | |
| mordred | and those connected to each other | 17:55 |
| clarkb | all hooked up and on the enxternal net if available | 17:55 |
| mordred | yup | 17:55 |
| Shrews | i am not convinced | 17:55 |
| clarkb | Shrews: you haven't clouded then | 17:56 |
| clarkb | this is my biggest pain poit ever using openstack | 17:56 |
| *** adalbas has quit IRC | 17:56 | |
| mordred | yah. mine too | 17:56 |
| sdake | fungi thanks fore verifyign :) | 17:56 |
| clarkb | oh also because order is important | 17:56 |
| mordred | also, the docs for it are terrible | 17:56 |
| fungi | sdake: rhallisey: oh, for the "openstackstatus" statusbot specifically the channel needs to be added to https://git.openstack.org/cgit/openstack-infra/system-config/tree/manifests/site.pp#n266 | 17:56 |
| clarkb | so not only do I have to remember to do everything I have to be very careful about how I do it | 17:56 |
| *** mtanino has quit IRC | 17:56 | |
| mordred | "my cloud doesn't work" "why" "oh, I didn't create 6 different objects" *HEADDESK* | 17:56 |
| *** zz_dimtruck is now known as dimtruck | 17:57 | |
| clarkb | so anyways I think these router primitives are fine, having primitives is good but I think oen of the reasons shade exists is to serve the opinionated case so keeping that in mind as we go is good | 17:57 |
| mordred | Shrews: also - the router api calls are all blocking? | 17:58 |
| openstackgerrit | Spencer Krum proposed openstack-infra/system-config: Add a helpful friend to gerrit reviews https://review.openstack.org/169010 | 17:58 |
| Shrews | clarkb: So, a good compromise is to allow full control by the user (what I'm creating), but we can also provide a do-it-all-for-me API method. I don't think that should be part of this review, though. | 17:58 |
| mordred | Shrews: have I mentioned that I love the consistency | 17:58 |
| *** Somay has joined #openstack-infra | 17:59 | |
| mordred | Shrews: yes. sorry - that's what IW as thinking | 17:59 |
| mordred | Shrews: we make each of the apis like you're making | 17:59 |
| clarkb | Shrews: yes thats what I am saying | 17:59 |
| *** rlucio has joined #openstack-infra | 17:59 | |
| mordred | Shrews: also - is this correct or a typo: new_router = neutron.create_router(dict(router=router)) | 17:59 |
| clarkb | also at some point we need to have one change that adds mock as a dependency | 17:59 |
| Shrews | clarkb: good, because i loathe reviews that are all "hey, add this other functionality that i just thought of" :) | 18:00 |
| clarkb | then rease all of the others on that otherwise we will end up with funny mock dep listings | 18:00 |
| Shrews | mordred: maybe? | 18:00 |
| mordred | clarkb: yah - I'm going to rebase Shrews patch into the other series because I need to rebase the "migrate to tasks" patch on top of shrews' patch | 18:00 |
| mordred | clarkb: I mean, unless Shrews feels like doing that | 18:00 |
| *** timcline_ has quit IRC | 18:01 | |
| mordred | but it seemed rude to ask him to | 18:01 |
| Shrews | mordred: go for it | 18:01 |
| mordred | Shrews: because passing in a dict as a positional parameter with a parameter of "router" which is a dict is CRAZY | 18:01 |
| *** timcline has joined #openstack-infra | 18:01 | |
| sdake | fungi thanks as usual the infrastructure team rocks :) | 18:01 |
| mordred | Shrews: but, I would not put it past things for that to be the interface | 18:02 |
| fungi | sdake: i like to think so, and you're welcome! ;) | 18:02 |
| Shrews | mordred: the neutron api is a bit weird, yeah | 18:02 |
| mordred | Shrews: OH DEAR GOD YOUR PATCH IS CORRECT | 18:02 |
| mordred | def create_router(self, body=None): | 18:02 |
| clarkb | Shrews: I am definitely not going to -1 the patch because of a missing feature. I just want shade to in general keep these things in mind as we make the primitives | 18:03 |
| mordred | clarkb: ^^ I'm going to guess that's been used by literally no humans ever | 18:03 |
| Shrews | clarkb: agreed. i am not core, though, so you guys drive the direction :-P | 18:03 |
| clarkb | mordred: create_router()? | 18:03 |
| mordred | clarkb: yeah | 18:03 |
| mordred | it takes "body" as a single arg, which is expected to be a dict | 18:04 |
| clarkb | mordred: I always use the horizon ui because I can never remember the entire list of things and what order to do them in | 18:04 |
| Shrews | mordred: subnet stuff is the same way | 18:04 |
| mordred | Shrews: it seems to be the pattern in the library | 18:04 |
| Shrews | yup | 18:04 |
| mordred | this may even win the prize for worst thing | 18:04 |
| Shrews | took me a bit of digging to figure out wth is going on | 18:04 |
| mordred | it's a tough call between this and the glance task api | 18:04 |
| *** jamielennox|away is now known as jamielennox | 18:04 | |
| mordred | Shrews: you win a cookie | 18:05 |
| mordred | Shrews: it's a little used, sorry | 18:05 |
| *** baoli has quit IRC | 18:05 | |
| clarkb | oh ya you have to enable dhcp too iirc | 18:05 |
| clarkb | and set the range (rather than just inferring it from the subnet) | 18:05 |
| asselin_ | are these commands correct? ./tools/zuul-changes.py ... http://ci.openstack.org/zuul/zuul.html#starting-zuul | 18:06 |
| openstackgerrit | Ryan Hallisey proposed openstack-infra/system-config: Add status bot to kolla channel https://review.openstack.org/169445 | 18:07 |
| *** MarkAtwood has joined #openstack-infra | 18:08 | |
| *** adalbas has joined #openstack-infra | 18:08 | |
| *** e0ne has joined #openstack-infra | 18:09 | |
| clarkb | asselin_: https://git.openstack.org/cgit/openstack-infra/zuul/tree/setup.cfg#n23 is the mapping and Ithink that help text is generated by running that command | 18:09 |
| clarkb | asselin_: so yes? | 18:09 |
| SpamapS | mordred: Heat devs begged for changes to Neutron API and got told all the reasons they were wrong. Bring your A game. :) | 18:09 |
| SpamapS | mordred: and Neutron is by far the most difficult thing to orchestrate with Heat, so likely the same with something like Ansible. | 18:10 |
| clarkb | mordred: Shrews is the get_resource() pattern of iterating through a list rather than doing a show on said resource type iwth the name/id a thing in shade because not every resource supports shows on names? | 18:10 |
| *** esker has joined #openstack-infra | 18:10 | |
| *** esker has quit IRC | 18:10 | |
| asselin_ | clarkb, seem's something not right: http://paste.openstack.org/show/197732/ | 18:11 |
| clarkb | oh zuul-changes is in the tools dir that different | 18:12 |
| *** MarkAtwood has quit IRC | 18:12 | |
| Shrews | clarkb: yeah | 18:13 |
| Shrews | convenience method | 18:13 |
| clarkb | asselin_: https://git.openstack.org/cgit/openstack-infra/zuul/tree/tools/zuul-changes.py it doesn't take the 'rehcec' comment | 18:13 |
| clarkb | asselin_: it instead prints out zuul-enqueue commands so no gerrit stuff is involved | 18:13 |
| Shrews | mordred: if you are doing mind-altering things with your tasks stuff, i'll hold off on my subnet patch until i see what you are doing | 18:14 |
| mordred | SpamapS: yes | 18:14 |
| mordred | gah | 18:14 |
| mordred | clarkb: yes | 18:14 |
| mordred | clarkb: also, many of those "show" on a name are actually doing a list in the client library | 18:15 |
| clarkb | mordred: nice | 18:15 |
| mordred | or, rather, you're talking about list(name="foo") - not show | 18:15 |
| mordred | show is different | 18:15 |
| mordred | but if we do the pattern, then we can control caching | 18:15 |
| *** yfried is now known as yfried|afk | 18:15 | |
| mordred | clarkb: there are so many things you learn in doing this that you never wanted to know | 18:15 |
| clarkb | Shrews: does update router need to also take args to change the external net port? I am trying to remember how neutron exposes that | 18:16 |
| *** rhallisey has left #openstack-infra | 18:16 | |
| *** yfried|afk is now known as yfried | 18:16 | |
| *** Longgeek has quit IRC | 18:16 | |
| *** Longgeek has joined #openstack-infra | 18:17 | |
| jogo | clarkb: A sounds like the most straight forward | 18:17 |
| jogo | clarkb: instead of $IP-subnode, we can just use subnode | 18:17 |
| clarkb | jogo: yup | 18:17 |
| jogo | or something like that | 18:17 |
| clarkb | jogo: I would say subnode-1 | 18:18 |
| *** tonytan4ever has quit IRC | 18:18 | |
| jogo | and works for me | 18:18 |
| *** e0ne is now known as e0ne_ | 18:18 | |
| *** AJaeger_ has quit IRC | 18:18 | |
| *** tonytan4ever has joined #openstack-infra | 18:19 | |
| openstackgerrit | Joe Gordon proposed openstack-infra/devstack-gate: Make aiopcpu subnode directory more deterministic https://review.openstack.org/169449 | 18:19 |
| clarkb | Shrews: ya looking at http://developer.openstack.org/api-ref-networking-v2-ext.html the external network id is a thing you can update. Is that something you think should go into this patch? | 18:19 |
| jogo | clarkb: like taht ^ ? | 18:19 |
| clarkb | jogo: no, you should do it where X is arbitrary based on the number of nodes | 18:20 |
| *** _nadya_ has quit IRC | 18:21 | |
| clarkb | also for fun the network_id for external network id appears to be a dict | 18:21 |
| Shrews | clarkb: i can add that pretty easily | 18:21 |
| *** teran_ has joined #openstack-infra | 18:21 | |
| clarkb | Shrews: I am writing comments, I will probably +2 then let you decide if it needs to go in this change or a followup | 18:22 |
| *** ihrachyshka has quit IRC | 18:22 | |
| Shrews | clarkb: i'll post a followup review then | 18:23 |
| Shrews | clarkb: thanks | 18:23 |
| openstackgerrit | Merged openstack/requirements: Updated incorrect name for angular lrdragndrop https://review.openstack.org/169132 | 18:23 |
| *** e0ne_ has quit IRC | 18:23 | |
| openstackgerrit | Joe Gordon proposed openstack-infra/devstack-gate: Make aiopcpu subnode directory more deterministic https://review.openstack.org/169449 | 18:23 |
| Shrews | jogo: how does one pronounce "aiopcpu" ???? | 18:23 |
| *** packet has joined #openstack-infra | 18:24 | |
| jogo | Shrews: mumble it | 18:24 |
| Shrews | ah, of course | 18:24 |
| jogo | all in one plus CPU | 18:24 |
| *** tonytan4ever has quit IRC | 18:24 | |
| clarkb | jogo: also I got dtroyer feedback on how to make my devstack change for the cpu_map update mergeable | 18:25 |
| clarkb | so going to do that next | 18:25 |
| *** teran has quit IRC | 18:25 | |
| jogo | nice | 18:25 |
| *** teran_ has quit IRC | 18:25 | |
| *** chris613 has quit IRC | 18:26 | |
| clarkb | Shrews: commented | 18:29 |
| *** mpaolino has quit IRC | 18:29 | |
| fungi | wow. we are somehow ending up with 167 additional packages on devstack-trusty in rax compared to hpcloud, but only 26 additional when comparing bare-trusty between the providers. this means we're somehow ending up with a ton of crap preinstalled on our devstack-trusty workers in rax | 18:31 |
| fungi | things which we presumably install when making bare-trusty workers | 18:31 |
| clarkb | fungi: rax makes the base images and probably adds a bunch of goodness for us | 18:32 |
| fungi | yeah, i'm guessing that the majority of that difference is stuff we wind up adding for bare-trusty on hpcloud but is already present for us in rax | 18:32 |
| *** dkranz has quit IRC | 18:33 | |
| *** garyh has quit IRC | 18:33 | |
| fungi | this adds an interesting extra challenge to the problem of validating that bindep is installing everything we need. it's easier for me to snag ready nodes in rax, much less so in hpcloud these days | 18:33 |
| clarkb | fungi: you could just boot the hpcloud images locally | 18:34 |
| fungi | yep, i may end up doing that | 18:34 |
| openstackgerrit | David Shrewsbury proposed openstack-infra/shade: Let router update to specify external gw net ID https://review.openstack.org/169454 | 18:34 |
| Shrews | clarkb: ^^^^ | 18:34 |
| clarkb | fungi: https://github.com/greghaynes/dib-dev-deploy makes it easy | 18:34 |
| fungi | or even just using them as chroots, not booting them | 18:34 |
| clarkb | fungi: ya I guess you aren't concerend about booting just package availability | 18:35 |
| mordred | fungi: I use them as chroots | 18:35 |
| mordred | fungi: also - this is one of the reasons I like my minimal base images | 18:35 |
| clarkb | I have typically been booting because booting has been the trouble area :) | 18:35 |
| mordred | fungi: it should help make sure that we actually _do_ list the things we need | 18:35 |
| mordred | clarkb: :) | 18:36 |
| fungi | since all i really want to compare is the package list from a devstack image after bindep runs against the package list from a bare image. so i can chroot into the devstack dib image and feed bindep results into the package manager in the chroot | 18:36 |
| mordred | clarkb: I've been booting them recently | 18:36 |
| fungi | this shouldn't theoretically have any bearing on booting | 18:36 |
| *** e0ne has joined #openstack-infra | 18:36 | |
| mordred | fungi: yup | 18:36 |
| clarkb | Shrews: the neutron docs indicate that the network_id is a dict, do we want to make user pass in dict or make a dict for them? | 18:36 |
| mordred | makea dict for them | 18:36 |
| mordred | that's a stupid interface | 18:36 |
| clarkb | Shrews: ^ ok so thats a -1 from mordred :) | 18:36 |
| *** eharney has quit IRC | 18:37 | |
| Shrews | clarkb: it's only a single value. let's not make it more difficult than it is :) | 18:37 |
| clarkb | Shrews: though I haven't confirmed that this is the interface in neutronclient | 18:37 |
| fungi | mordred: bindep totally needs to drop all its command line options and configuration in favor of passing a dict as an envvar | 18:37 |
| * clarkb goes to read neutronclient source | 18:37 | |
| mordred | fungi: yes | 18:37 |
| mordred | clarkb: it's the only way to know how anything works | 18:37 |
| fungi | it's the interface of teh future | 18:37 |
| fungi | sorta like dippin' dots | 18:38 |
| mordred | Shrews: I like the patch - as long as it doesn't require that to be a dict :) | 18:38 |
| Shrews | clarkb: good luck! neutronclient is stupidly absent of information | 18:38 |
| Shrews | clarkb: try 'neutron help router-update' or whatever the command is | 18:38 |
| Shrews | no info on how to pass the update information | 18:38 |
| clarkb | Shrews: wow yes it is | 18:38 |
| Shrews | major client fail | 18:38 |
| clarkb | so I think the api docs are what we must look at and that says its a dict | 18:38 |
| clarkb | hrm maybe I should look in the neutron api server | 18:38 |
| Shrews | clarkb: yup, i've been going by that | 18:39 |
| mordred | THAT IS CRAZY | 18:39 |
| clarkb | Shrews: rgr trying my hand at it now | 18:39 |
| *** yfried is now known as yfried|afk | 18:41 | |
| clarkb | hrm | 18:42 |
| *** afazekas has quit IRC | 18:43 | |
| greghaynes | wow non kvm qemu is slow | 18:45 |
| clarkb | Shrews: https://git.openstack.org/cgit/openstack/neutron/tree/neutron/extensions/l3.py#n105 so it is a dict | 18:46 |
| greghaynes | trying to make a dib test out of it, taking 10s of mins to boot a stock ubuntu vm | 18:46 |
| clarkb | Shrews: I will -1 with a pointer there | 18:46 |
| clarkb | greghaynes: I told you | 18:46 |
| greghaynes | :( | 18:46 |
| *** dkranz has joined #openstack-infra | 18:46 | |
| Shrews | clarkb: i'm passing a dict... not sure what you are -1'ing | 18:46 |
| clarkb | its something like 45 seconds to boot ubuntu with openstack intsalled | 18:46 |
| clarkb | Shrews: oh derp I was reading poorly | 18:47 |
| Shrews | :) | 18:47 |
| *** Somay has quit IRC | 18:49 | |
| *** achanda has quit IRC | 18:49 | |
| Shrews | clarkb: and thanks for the reviews | 18:50 |
| greghaynes | clarkb: 45secs would be awesome | 18:51 |
| clarkb | greghaynes: sorry minutes :) | 18:51 |
| clarkb | I just caught that typo myself :) | 18:51 |
| greghaynes | "what!? its faster than kvm!?" | 18:51 |
| *** Rockyg has joined #openstack-infra | 18:51 | |
| *** dkranz has quit IRC | 18:51 | |
| *** annegentle has joined #openstack-infra | 18:52 | |
| *** dhritishikhar_ has quit IRC | 18:52 | |
| *** fbo has quit IRC | 18:52 | |
| *** tonytan4ever has joined #openstack-infra | 18:52 | |
| *** stevemar has quit IRC | 18:54 | |
| *** tnovacik has joined #openstack-infra | 18:54 | |
| *** pblaho has joined #openstack-infra | 18:55 | |
| openstackgerrit | Merged openstack-infra/shade: Disable warnings about old Rackspace certificates https://review.openstack.org/168734 | 18:56 |
| *** yfried|afk is now known as yfried | 18:56 | |
| *** flaviosr has joined #openstack-infra | 18:58 | |
| tchaypo | Does DST flip this weekend in the US, or are you a few weeks later? | 18:58 |
| openstackgerrit | Monty Taylor proposed openstack-infra/shade: Namespace caching per cloud https://review.openstack.org/168615 | 18:58 |
| openstackgerrit | Monty Taylor proposed openstack-infra/shade: Allow for passing cache class in as a parameter https://review.openstack.org/162148 | 18:58 |
| openstackgerrit | Monty Taylor proposed openstack-infra/shade: Add 'rebuild' to shade https://review.openstack.org/161067 | 18:58 |
| openstackgerrit | Monty Taylor proposed openstack-infra/shade: Migrate API calls to task management https://review.openstack.org/168665 | 18:58 |
| openstackgerrit | Monty Taylor proposed openstack-infra/shade: Add task management framework https://review.openstack.org/168664 | 18:58 |
| mordred | tchaypo: we've already flipped | 18:58 |
| mordred | clarkb, Shrews: I rebased on top of Shrews patch for routers and update the last patch to put those calls into the task framework | 18:59 |
| dstufft | DST is the worst | 18:59 |
| mordred | I also +A'd both of those patches, so they shoudl land soon | 18:59 |
| Shrews | hazzah | 18:59 |
| openstackgerrit | Merged openstack-infra/shade: Add methods for logical router management https://review.openstack.org/167371 | 18:59 |
| jeblair | meeting time! | 19:00 |
| tchaypo | This year DST makes my Easter long weekend 97 hours instead of the usual 96 | 19:00 |
| *** jlanoux has quit IRC | 19:00 | |
| *** primemin1sterp is now known as primeministerp | 19:00 | |
| *** Somay has joined #openstack-infra | 19:01 | |
| *** Sukhdev has joined #openstack-infra | 19:01 | |
| openstackgerrit | Michael Krotscheck proposed openstack-infra/storyboard-webclient: Fixed reference to experimental API https://review.openstack.org/169458 | 19:02 |
| *** fbo has joined #openstack-infra | 19:04 | |
| *** dboik has quit IRC | 19:04 | |
| *** dkranz has joined #openstack-infra | 19:04 | |
| *** stevemar has joined #openstack-infra | 19:04 | |
| *** dboik has joined #openstack-infra | 19:04 | |
| ekarlso | anyone know how with sqla migrate to create a enum and then reference that enum in a table, then create the table with having it creating the enum again that's already created in pgsql ? | 19:05 |
| *** Sukhdev has quit IRC | 19:05 | |
| openstackgerrit | Monty Taylor proposed openstack-infra/shade: Namespace caching per cloud https://review.openstack.org/168615 | 19:05 |
| openstackgerrit | Monty Taylor proposed openstack-infra/shade: Allow for passing cache class in as a parameter https://review.openstack.org/162148 | 19:05 |
| openstackgerrit | Monty Taylor proposed openstack-infra/shade: Add 'rebuild' to shade https://review.openstack.org/161067 | 19:05 |
| openstackgerrit | Monty Taylor proposed openstack-infra/shade: Migrate API calls to task management https://review.openstack.org/168665 | 19:05 |
| openstackgerrit | Monty Taylor proposed openstack-infra/shade: Add task management framework https://review.openstack.org/168664 | 19:05 |
| *** tiswanso has quit IRC | 19:07 | |
| *** adalbas has quit IRC | 19:07 | |
| dansmith | are there ceph tests in experimental or something? | 19:07 |
| *** stevemar has quit IRC | 19:08 | |
| clarkb | dansmith: there is a non voting devstack + ceph job on nova and cinder's check queue | 19:08 |
| *** stevemar has joined #openstack-infra | 19:08 | |
| dansmith | okay, thought I needed to hit the experimental button for that | 19:09 |
| dansmith | guess I need to pay more attention | 19:09 |
| openstackgerrit | Merged openstack-infra/devstack-gate: Redirect subnode cleanup to devstack-gate-cleanup-host.txt https://review.openstack.org/168102 | 19:10 |
| *** salv-orl_ has joined #openstack-infra | 19:11 | |
| *** garyh has joined #openstack-infra | 19:11 | |
| *** salv-orlando has quit IRC | 19:13 | |
| *** yfried is now known as yfried|afk | 19:14 | |
| *** Somay has quit IRC | 19:14 | |
| *** ZZelle_ has joined #openstack-infra | 19:16 | |
| openstackgerrit | Stefano Maffulli 'reed' proposed openstack-infra/infra-manual: Creates a sandbox repo tutorial https://review.openstack.org/138195 | 19:19 |
| *** medieval1 has quit IRC | 19:19 | |
| *** otter768 has joined #openstack-infra | 19:22 | |
| mtreinish | clarkb: do you remember what that bug # was from the gate bug we were tracking a while ago that you hit on hp cloud? | 19:22 |
| *** sarob has joined #openstack-infra | 19:22 | |
| jlvillal | dhellmann: On https://github.com/dhellmann/ansible-znc-on-znc Would you have any interest in patches for Fedora support? | 19:23 |
| clarkb | mtreinish: the neutron port leak bug? I do not remember the numer | 19:24 |
| clarkb | mtreinish: but I can find it if you like | 19:24 |
| *** yfried|afk is now known as yfried | 19:24 | |
| *** adalbas has joined #openstack-infra | 19:24 | |
| clarkb | mtreinish: https://bugs.launchpad.net/nova/+bug/1356157 | 19:25 |
| openstack | Launchpad bug 1356157 in OpenStack Compute (nova) icehouse "make nova floating-ip-delete atomic with neutron" [High,Fix released] - Assigned to Aaron Rosen (arosen) | 19:25 |
| mtreinish | clarkb: yeah I think that was the one thanks | 19:25 |
| *** mfink has left #openstack-infra | 19:26 | |
| *** otter768 has quit IRC | 19:26 | |
| *** achanda has joined #openstack-infra | 19:27 | |
| *** ynasser has joined #openstack-infra | 19:28 | |
| *** Sukhdev has joined #openstack-infra | 19:29 | |
| *** sdake_ has joined #openstack-infra | 19:31 | |
| openstackgerrit | Christoph Gysin proposed stackforge/gertty: Add support for auth-type=form https://review.openstack.org/169467 | 19:32 |
| *** SumitNaiksatam has quit IRC | 19:33 | |
| openstackgerrit | Christoph Gysin proposed stackforge/gertty: Fix repository checking https://review.openstack.org/169468 | 19:33 |
| openstackgerrit | Christoph Gysin proposed stackforge/gertty: Add config option for git clone URL https://review.openstack.org/169469 | 19:33 |
| anteaya | jlvillal: he is on vacation, should be back maybe tomorrow or Thursday | 19:34 |
| jlvillal | anteaya: Thank you. | 19:34 |
| anteaya | jlvillal: welcome | 19:34 |
| *** sdake__ has joined #openstack-infra | 19:35 | |
| openstackgerrit | Christoph Gysin proposed stackforge/gertty: Add script to start gertty out of source dir https://review.openstack.org/169470 | 19:35 |
| *** sdake has quit IRC | 19:35 | |
| *** reed has quit IRC | 19:36 | |
| *** reed has joined #openstack-infra | 19:36 | |
| *** Sukhdev has quit IRC | 19:36 | |
| *** Sukhdev has joined #openstack-infra | 19:37 | |
| *** tkelsey has joined #openstack-infra | 19:37 | |
| reed | I fat-fingered killed X | 19:37 |
| *** sdake_ has quit IRC | 19:38 | |
| *** Rockyg has quit IRC | 19:39 | |
| *** alexpilotti has joined #openstack-infra | 19:41 | |
| *** tkelsey has quit IRC | 19:41 | |
| *** SumitNaiksatam has joined #openstack-infra | 19:44 | |
| anteaya | :( | 19:47 |
| *** chris613 has joined #openstack-infra | 19:47 | |
| *** yfried is now known as yfried|afk | 19:47 | |
| *** chris613 has left #openstack-infra | 19:47 | |
| *** Rockyg has joined #openstack-infra | 19:47 | |
| fungi | debian jessie release date set for april 25. sooner than i expected | 19:47 |
| anteaya | good to know | 19:48 |
| *** tim_o has joined #openstack-infra | 19:49 | |
| *** thingee has joined #openstack-infra | 19:50 | |
| *** chris___2 has joined #openstack-infra | 19:50 | |
| *** annegentle has quit IRC | 19:51 | |
| *** Sukhdev has quit IRC | 19:54 | |
| *** gokrokve_ has joined #openstack-infra | 19:57 | |
| *** hashar has joined #openstack-infra | 19:58 | |
| *** gokrokve_ has quit IRC | 19:59 | |
| *** gokrokve_ has joined #openstack-infra | 19:59 | |
| *** dboik_ has joined #openstack-infra | 20:00 | |
| openstackgerrit | Clark Boylan proposed openstack-infra/nodepool: Support per provider image build settings https://review.openstack.org/126747 | 20:00 |
| openstackgerrit | Clark Boylan proposed openstack-infra/nodepool: Just set qcow2 version https://review.openstack.org/169383 | 20:00 |
| *** gokrokve has quit IRC | 20:00 | |
| openstackgerrit | Clark Boylan proposed openstack-infra/nodepool: Use correct diskimage when building diskimages https://review.openstack.org/169474 | 20:00 |
| anteaya | so I was hoping we would hit open discussion but we did get through the agenda | 20:00 |
| clarkb | mordred: jeblair ^ that should pass tests now | 20:01 |
| mordred | woot | 20:01 |
| clarkb | time for foods | 20:01 |
| greghaynes | oo fun | 20:01 |
| anteaya | so do we want to talk about what to do about a summit etherpad thing? | 20:01 |
| anteaya | as many other teams are launching them | 20:01 |
| *** yfried|afk is now known as yfried | 20:01 | |
| *** ibiris is now known as ibiris_away | 20:02 | |
| yolanda | jeblair, so per benchmarking etherpad, i'll be collecting more info tomorrow so we can compare properly | 20:03 |
| *** dboik has quit IRC | 20:03 | |
| jeblair | anteaya: not yet | 20:03 |
| anteaya | jeblair: very good | 20:04 |
| *** dkranz has quit IRC | 20:05 | |
| *** andreykurilin_ has quit IRC | 20:06 | |
| *** dkranz has joined #openstack-infra | 20:09 | |
| *** annegentle has joined #openstack-infra | 20:10 | |
| *** tonytan4ever has quit IRC | 20:10 | |
| *** tonytan4ever has joined #openstack-infra | 20:11 | |
| *** tim_o has quit IRC | 20:11 | |
| *** sabeen has quit IRC | 20:11 | |
| *** sabeen has joined #openstack-infra | 20:11 | |
| *** yfried is now known as yfried|afk | 20:13 | |
| *** eharney has joined #openstack-infra | 20:14 | |
| *** yfried|afk is now known as yfried | 20:17 | |
| *** ayoung has quit IRC | 20:17 | |
| *** hashar has quit IRC | 20:17 | |
| *** tiswanso has joined #openstack-infra | 20:19 | |
| *** achanda has quit IRC | 20:20 | |
| *** mfink has joined #openstack-infra | 20:20 | |
| *** tiswanso has quit IRC | 20:20 | |
| *** tiswanso has joined #openstack-infra | 20:21 | |
| *** alexpilotti has quit IRC | 20:21 | |
| *** dboik_ has quit IRC | 20:22 | |
| *** dboik has joined #openstack-infra | 20:23 | |
| *** baoli has joined #openstack-infra | 20:26 | |
| *** ibiris_away is now known as ibiris | 20:27 | |
| harlowja_ | mordred fyi; just saw this | 20:28 |
| *** adalbas has quit IRC | 20:28 | |
| harlowja_ | [13:18:34] <Odd_Bloke> smoser: Am I right in thinking that vendor-data simply isn't a thing in the version of cloud-init in precise? | 20:28 |
| harlowja_ | [13:26:55] <smoser> right . not at all. | 20:28 |
| harlowja_ | [13:27:02] <smoser> Odd_Bloke, i'd sponsor an sru | 20:28 |
| harlowja_ | sooo mordred maybe that will help with HP cloud cloudinit? | 20:29 |
| harlowja_ | or whatever | 20:29 |
| mordred | harlowja_: nope. I'm not using it | 20:29 |
| mordred | but thanks | 20:29 |
| harlowja_ | lol | 20:29 |
| harlowja_ | mordred do u need a hug again | 20:29 |
| *** teran has joined #openstack-infra | 20:29 | |
| mordred | nope! I'm quite happy about it | 20:29 |
| harlowja_ | hugggg | 20:29 |
| harlowja_ | lol | 20:29 |
| mordred | :) | 20:29 |
| harlowja_ | don't deny the hug man | 20:29 |
| harlowja_ | jeeez | 20:29 |
| harlowja_ | jeeeeezsh | 20:29 |
| harlowja_ | lol | 20:29 |
| mordred | hugs are good | 20:29 |
| harlowja_ | ha | 20:29 |
| harlowja_ | hugs are made better by cloudinit | 20:30 |
| harlowja_ | lol | 20:30 |
| mordred | harlowja_: and enhanced ever further by simple-init :) | 20:30 |
| *** medieval1 has joined #openstack-infra | 20:30 | |
| *** annegentle has quit IRC | 20:30 | |
| reed | has anyone ever tried to abandon a change from CLI, using ssh as in https://review.openstack.org/Documentation/cmd-review.html? | 20:31 |
| *** sarob has quit IRC | 20:32 | |
| openstackgerrit | Spencer Krum proposed openstack-infra/system-config: Revert "Add a helpful friend to gerrit reviews" https://review.openstack.org/169500 | 20:33 |
| *** ayoung has joined #openstack-infra | 20:33 | |
| *** annegentle has joined #openstack-infra | 20:33 | |
| *** annegentle is now known as agentle_ | 20:33 | |
| *** annegentle has joined #openstack-infra | 20:34 | |
| openstackgerrit | Merged openstack-infra/shade: Let router update to specify external gw net ID https://review.openstack.org/169454 | 20:37 |
| sdague | reed: yes | 20:37 |
| arosen | quick questions, so this patch was just merged that makes congress an official openstack project: https://review.openstack.org/#/c/165204/ | 20:37 |
| *** ghostpl_ has joined #openstack-infra | 20:38 | |
| arosen | I guess I need to now update project-config to move congress to openstack rather than stackforge? | 20:38 |
| reed | sdague, what do you think causes it to barf on my saying "fatal: "test" is not a valid patch set" | 20:38 |
| reed | ssh -l smaffulli -p 29418 review.openstack.org 'gerrit review ' --project openstack-dev/sandbox.git -m "Abandon test patch" --abandon 169460 | 20:38 |
| *** dizquierdo has joined #openstack-infra | 20:38 | |
| *** baoli has quit IRC | 20:38 | |
| *** baoli has joined #openstack-infra | 20:39 | |
| anteaya | arosen: yes that would be a good patch to offer | 20:39 |
| sdague | you need the patch level as well iirc | 20:39 |
| sdague | or the idempotent id | 20:39 |
| sdague | it's either X/Y or Ixxxxxxxxxxxxxxxxxxxxxxx | 20:39 |
| anteaya | arosen: then add a link to the patch to upcoming project renames: https://wiki.openstack.org/wiki/Meetings/InfraTeamMeeting | 20:40 |
| anteaya | arosen: so we can schedule a rename | 20:40 |
| openstackgerrit | Spencer Krum proposed openstack-infra/system-config: Add a helpful friend to gerrit reviews https://review.openstack.org/169010 | 20:40 |
| sdague | 169460 doesn't identify specifically enough | 20:40 |
| *** david-lyle has quit IRC | 20:40 | |
| openstackgerrit | Spencer Krum proposed openstack-infra/system-config: Revert "Add a helpful friend to gerrit reviews" https://review.openstack.org/169501 | 20:41 |
| *** e0ne has quit IRC | 20:42 | |
| reed | sdague, I've tried with change-id too | 20:42 |
| *** e0ne has joined #openstack-infra | 20:42 | |
| *** ghostpl_ has quit IRC | 20:42 | |
| reed | sdague, AH! | 20:42 |
| reed | you need to specify the patchset, too | 20:42 |
| *** mrda-away is now known as mrda | 20:46 | |
| *** david-lyle has joined #openstack-infra | 20:49 | |
| *** ibiris is now known as ibiris_away | 20:50 | |
| *** claudiub has joined #openstack-infra | 20:51 | |
| *** e0ne is now known as e0ne_ | 20:52 | |
| *** e0ne_ is now known as e0ne | 20:52 | |
| fungi | which i find odd given that abandonment shouldn't need to take the patchset number into consideration, but i guess that's a non-optional part of their change manipulation api | 20:53 |
| nibalizer | jeblair: sdague can you review 169010 ? | 20:53 |
| greghaynes | one sec | 20:53 |
| greghaynes | I need to edit something :p | 20:53 |
| openstackgerrit | Stefano Maffulli 'reed' proposed openstack-infra/infra-manual: Creates a sandbox repo tutorial https://review.openstack.org/138195 | 20:53 |
| greghaynes | well, I could make my change dependent | 20:54 |
| greghaynes | go for it | 20:54 |
| *** ayoung has quit IRC | 20:54 | |
| *** yfried is now known as yfried|afk | 20:55 | |
| *** mrmartin has joined #openstack-infra | 20:56 | |
| openstackgerrit | Michael Krotscheck proposed openstack-infra/storyboard-webclient: Added commandline commands to package.json https://review.openstack.org/168488 | 20:56 |
| openstackgerrit | Michael Krotscheck proposed openstack-infra/storyboard-webclient: Removed grunt:server https://review.openstack.org/163165 | 20:56 |
| *** dimtruck is now known as zz_dimtruck | 20:57 | |
| harlowja_ | mordred ha; maybe :-P | 20:58 |
| *** zz_dimtruck is now known as dimtruck | 20:58 | |
| *** ibiris_away is now known as ibiris | 20:59 | |
| *** annegentle has quit IRC | 21:01 | |
| *** agentle_ has quit IRC | 21:01 | |
| *** annegentle has joined #openstack-infra | 21:01 | |
| anteaya | reed: did you get what you needed with that patch? | 21:02 |
| reed | anteaya, yes, I just submitted my final work. Have a look at it please, let me know what you think | 21:02 |
| anteaya | reed: okay | 21:03 |
| *** mrmartin has quit IRC | 21:03 | |
| *** dangers_away is now known as dangers | 21:04 | |
| anteaya | reed: you need to rebase https://review.openstack.org/#/c/104951/ on your new patch | 21:05 |
| anteaya | since that is the link to the tutorial | 21:05 |
| reed | ah, so complicated :) | 21:06 |
| openstackgerrit | James E. Blair proposed openstack-infra/system-config: Revert "Add a helpful friend to gerrit reviews" https://review.openstack.org/169508 | 21:06 |
| *** dkliban is now known as dkliban_afk | 21:06 | |
| *** annegentle has quit IRC | 21:07 | |
| *** ayoung has joined #openstack-infra | 21:07 | |
| anteaya | reed: yes | 21:08 |
| fungi | reed: looks like a substantial addition. i look forward to reading through | 21:08 |
| openstackgerrit | greghaynes proposed openstack-infra/system-config: Give some actions to our friend https://review.openstack.org/169509 | 21:08 |
| *** packet has quit IRC | 21:08 | |
| *** aduarte has quit IRC | 21:09 | |
| *** packet has joined #openstack-infra | 21:10 | |
| *** rfolco has quit IRC | 21:10 | |
| reed | I think I'm messing things up | 21:12 |
| *** aysyd has quit IRC | 21:13 | |
| *** tiswanso has quit IRC | 21:13 | |
| openstackgerrit | Stefano Maffulli 'reed' proposed openstack-infra/infra-manual: Adds a link for new developers to use sandbox https://review.openstack.org/104951 | 21:14 |
| openstackgerrit | Stefano Maffulli 'reed' proposed openstack-infra/infra-manual: Creates a sandbox repo tutorial https://review.openstack.org/138195 | 21:14 |
| *** achanda has joined #openstack-infra | 21:14 | |
| jogo | mordred jeblair: I thought I saw you guys say this morning that you don't think we are running out of nodes in nodepool is that correct? | 21:14 |
| anteaya | reed: that should work, let's wait for docs to come back | 21:14 |
| jeblair | jogo: well, they are a renewable resource. :) | 21:15 |
| *** ldnunes has quit IRC | 21:15 | |
| fungi | jogo: that sounds like an overly broad assertion | 21:15 |
| jeblair | we can harvest them, and they regrow at a rate which varies based on the overall health of the ecosystem | 21:16 |
| jogo | fungi: it is overly broad | 21:16 |
| fungi | we at times (for example right now) have jobs queued up waiting to eventually get an available node, and they would be done sooner in many cases if we had a larger quota | 21:16 |
| jogo | fungi: right. | 21:16 |
| ianw | my internet dropped before the open time in the meeting : can i get some eyes on the f20 removal patches https://review.openstack.org/#/q/status:open+topic:remove-f20,n,z | 21:16 |
| jeblair | (or clouds that worked) | 21:16 |
| openstackgerrit | K Jonathan Harker proposed openstack-infra/project-config: Make the unrecognized exception more verbose https://review.openstack.org/169511 | 21:16 |
| jogo | we seem to hit that state daily now | 21:16 |
| ianw | one job to switch to centos, key people have +1'd that, then we can remove the f20 nodes. one less problem to deal with | 21:16 |
| jogo | jobs queued waiting for available nodes | 21:16 |
| ianw | and workarounds | 21:16 |
| mordred | jogo: yes, but we're working through some issues and changs to work on that - in general, I do not think we need to stop testing things because we're out of nodes | 21:17 |
| fungi | jogo: yes, there are performance issues around turnover in our largest provider, and that seems to have us capped around 700-800 jobs per hour. that's still... a lot of jobs | 21:17 |
| openstackgerrit | Stefano Maffulli 'reed' proposed openstack-infra/infra-manual: Adds a link for new developers to use sandbox https://review.openstack.org/104951 | 21:17 |
| *** irdkehn has joined #openstack-infra | 21:17 | |
| *** baoli has quit IRC | 21:17 | |
| jeblair | jogo: also, if you look at the node graph, we have plenty of capacity overnight. so even with our currently reduced capacity, our test activity is sustainable. | 21:18 |
| *** baoli has joined #openstack-infra | 21:18 | |
| jogo | mordred: ahh when do you expect those issues to be resolved? | 21:18 |
| *** bswartz has quit IRC | 21:19 | |
| *** david-lyle_ has joined #openstack-infra | 21:19 | |
| jogo | jeblair: that a great metric IMHO. When what should be a 45 minute test takes 2 or 3 hours to run | 21:19 |
| jogo | it significantly slows iteration time down | 21:19 |
| mordred | jogo: it does -but that can also be the case when we have inifinite nodes | 21:20 |
| jeblair | jogo: so in summary, i appreciate and thank you for your effort to find and remove jobs that are not actually testing anything. but i don't think we need to remove jobs that are actually testing things, or avoid adding new jobs that test things. | 21:20 |
| jogo | fungi: a lot is relative, we also have a lot of git repos | 21:20 |
| jeblair | jogo: do you mean that it is not a great metric? | 21:20 |
| jeblair | jogo: i'm having trouble following you. | 21:20 |
| fungi | i agree that "jobs per hour" is not a great metric, but it's a metric we have | 21:20 |
| jogo | jeblair: err my bad, that is "not a great metric"* | 21:20 |
| jeblair | jogo: we've _never_ promised that the system will always be instantly responsive. | 21:21 |
| *** _nadya_ has joined #openstack-infra | 21:21 | |
| fungi | more accurate is to say that at peak we have ~400 nodes in constant use and another ~450 in a state of overhead (being built or deleted) | 21:21 |
| jeblair | jogo: in fact, the metric we _most_ concern ourselves with is whether we are able to merge changes faster than they are added to the merge queue. | 21:22 |
| jogo | jeblair: yeah trying to remove jobs that actually test nothing first. And not saying we *need* to drop useful tests. But we may need to in the future at this rate | 21:22 |
| *** Sukhdev has joined #openstack-infra | 21:22 | |
| sdague | jeblair: I think the concern that jogo is expressing, which I share, is whether or not the system is believed to have headroom, when it's typically taking 3 hrs to return test results, it creates a climate that makes people not want to expand test scope | 21:22 |
| jeblair | jogo: as for anything else, the system is eventually consistent. | 21:22 |
| jogo | jeblair: merge throughput is a useful metric, but that ignores the turn around time in development | 21:22 |
| jeblair | jogo: the primary goal of this system is not to outsource developer testing. | 21:23 |
| *** otter768 has joined #openstack-infra | 21:23 | |
| jogo | sdague: correct, that and it has a direct impact on iteration time of work | 21:23 |
| fungi | it also looks like at the moment changes complete check in under 2 hours, and changes are taking at worst a couple hours to make it through the gate | 21:23 |
| *** e0ne has quit IRC | 21:23 | |
| *** annegentle has joined #openstack-infra | 21:23 | |
| *** doug-fish has left #openstack-infra | 21:24 | |
| *** david-lyle has quit IRC | 21:24 | |
| jogo | fungi: yeah gate has been pretty solid recently. | 21:24 |
| openstackgerrit | Victor Seva proposed openstack-infra/jenkins-job-builder: Allow use of aliases defined previously inside included files https://review.openstack.org/157735 | 21:24 |
| jeblair | that sounds like well within what i think has traditionally been considered acceptable times for quite a while | 21:24 |
| jogo | jeblair: right. How would you sate the primary goal of the system? | 21:24 |
| fungi | though it looks like glance unit tests may have a problem at the moment | 21:25 |
| *** ibiris is now known as ibiris_away | 21:25 | |
| jogo | jeblair: right. How would you state the primary goal of the system? | 21:25 |
| jeblair | jogo: to ensure that only changes which pass tests merge | 21:25 |
| jeblair | jogo: to ensure that only changes which pass tests merge | 21:25 |
| *** _nadya_ has quit IRC | 21:25 | |
| *** peristeri has quit IRC | 21:26 | |
| jogo | jeblair: my first line had a typo, your didn't :) | 21:26 |
| fungi | a goal so nice he stated it twice! | 21:26 |
| jogo | lol | 21:26 |
| jeblair | jogo: drat, didn't see it, otherwise i would have put one in! :) | 21:26 |
| *** SumitNaiksatam has quit IRC | 21:26 | |
| jeblair | jogo: i guess i was just 'satisfied' with your first question | 21:26 |
| *** jgrimm is now known as zz_jgrimm | 21:26 | |
| fungi | it passed his tests | 21:26 |
| jeblair | (or... 'sated') | 21:26 |
| fungi | ohhh... indeed! | 21:27 |
| *** otter768 has quit IRC | 21:27 | |
| *** mjturek1 has quit IRC | 21:28 | |
| jogo | on a slightly tangential topic user survey results for postgres usage are fun http://superuser.openstack.org/articles/openstack-user-survey-insights-november-2014 | 21:28 |
| jeblair | jogo: anyway, i would like the system to be very fast and responsive, but when we're in a degraded state, i think it's reasonable for things to slow down a bit and for us to spread the load out over a larger time period | 21:28 |
| jogo | jeblair: agreed, I didn't realize we were still in a degraded state | 21:29 |
| *** pc_m has quit IRC | 21:30 | |
| jeblair | jogo: if circumstances were slightly different, that's just about the most awesome thing you could have said. :) | 21:30 |
| jogo | jeblair: hehe true | 21:30 |
| mordred | jogo: nice survey | 21:30 |
| mordred | jogo: so, basically, PG is in the margin of error | 21:31 |
| jeblair | jogo: also, it looks like load is getting very close to capacity, so we should be getting back to instantaneous response territory soon. | 21:31 |
| *** gulic has joined #openstack-infra | 21:31 | |
| *** Somay has joined #openstack-infra | 21:32 | |
| jogo | mordred: yup, more people use devstack that postgres | 21:34 |
| *** asselin has joined #openstack-infra | 21:34 | |
| *** gordc has quit IRC | 21:34 | |
| mordred | yah | 21:34 |
| *** samueldmq has quit IRC | 21:34 | |
| *** dprince has quit IRC | 21:35 | |
| *** mwagner_lap has quit IRC | 21:35 | |
| ekarlso | have ATC codes been sent yet ? | 21:35 |
| *** dkranz has quit IRC | 21:36 | |
| *** pblaho has quit IRC | 21:37 | |
| JayF | Yes, a long ways back iirc | 21:37 |
| *** asselin has quit IRC | 21:38 | |
| *** sdake has joined #openstack-infra | 21:38 | |
| fungi | ekarlso: in fact today is the last day to use them before the ticket price goes up and the discount code ceases to cover 100% of the cost | 21:39 |
| ekarlso | any idea on what the subject looks like ? :p | 21:39 |
| ekarlso | or what email their sent from | 21:39 |
| ekarlso | oh nmv ! | 21:39 |
| ekarlso | found it in the estack | 21:40 |
| ekarlso | not haystack :P | 21:40 |
| fungi | make sure to use it immediately | 21:41 |
| *** sdake__ has quit IRC | 21:42 | |
| *** david-lyle_ is now known as david-lyle | 21:43 | |
| jeblair | anyone noticed gertty failing to do its periodic sync lately? | 21:44 |
| jeblair | i just noticed mine hadn't synced for a while, and i don't see an error | 21:44 |
| jeblair | i've added more logging locally to try to catch it | 21:44 |
| jhesketh | Morning | 21:45 |
| *** Somay has quit IRC | 21:45 | |
| *** mriedem is now known as mriedem_away | 21:45 | |
| openstackgerrit | Merged openstack-infra/system-config: Add a helpful friend to gerrit reviews https://review.openstack.org/169010 | 21:46 |
| jeblair | nibalizer: ^ | 21:46 |
| dstufft | mordred: that makes me sad, postgres is a great db :D | 21:48 |
| mordred | dstufft: postgres is a great db! | 21:48 |
| mordred | dstufft: but most scalable applications are designed for _a_ db, not for _any_ db | 21:48 |
| mordred | in fact, I'd go so far as to say _all_ | 21:48 |
| dstufft | yea | 21:48 |
| mordred | scalable appliations | 21:49 |
| dstufft | I think that pluggable DBs are a bad scene | 21:49 |
| mordred | so - I dont' want us to delete postgres support because it'sa bad db | 21:49 |
| dstufft | except for frameworks and such | 21:49 |
| *** ivar-laz_ has quit IRC | 21:49 | |
| mordred | I want us to delete it because it's a stupid idea to support more than one db backend for non-frameworks | 21:49 |
| dstufft | mordred: to be clear, it makes me sad because i'd rather delete mysql support because of my own biases and the numbers don't back that up ;) | 21:49 |
| fungi | "we should also support postgres because someone may want to run openstack in a country where mysql is banned by law" | 21:49 |
| mordred | (I agree about framework pluggability - django supporting multi-databases is a great idea) | 21:49 |
| mordred | dstufft: :) | 21:49 |
| openstackgerrit | melanie witt proposed openstack-infra/project-config: Adjust regression exceptions for Nova Cells V1 job https://review.openstack.org/166396 | 21:49 |
| mordred | dstufft: that's fair - I want to delete postgres also because of my own biases | 21:50 |
| mordred | dstufft: I'm just lucky and have the numbers on my side | 21:50 |
| jeblair | mordred: that's not luck, that's skill | 21:50 |
| mordred | jeblair: :) | 21:50 |
| lifeless | mordred: I think we need to get community consensus about whether we're a product or a product toolkit | 21:50 |
| dstufft | also even beyond scaling applications, supporting multi dbs means supporting some lowest common denominator subset | 21:50 |
| nibalizer | jeblair: awesome | 21:50 |
| dstufft | so you don't even get to use any of the cool features of a particular db | 21:50 |
| lifeless | we're all over the map on that | 21:50 |
| mordred | lifeless: I believe that the foudnation has a very specific point of view on that already, actually | 21:50 |
| nibalizer | jeblair: my gertty always has syncing problems | 21:50 |
| mordred | dstufft: ++ | 21:51 |
| ekarlso | what's bad with postgres ? | 21:51 |
| nibalizer | im always quitting and forcing syncing and stuff, sometimes it gets better | 21:51 |
| mordred | ekarlso: nothing | 21:51 |
| mordred | ekarlso: we'r enot going to have a db debate in here | 21:51 |
| vishy | her does anyone here know if the sysctl for net.bridge got removed in recent kernels? | 21:51 |
| *** erlon has quit IRC | 21:51 | |
| *** stevemar has quit IRC | 21:51 | |
| mordred | vishy: I do not know | 21:51 |
| ekarlso | mordred: ;) | 21:51 |
| lifeless | mordred: that may be, but the dev community still gets up in knots on questions in that space | 21:51 |
| vishy | recently updated to kernel 3.19 and /proc/sys/net/bridge is gone | 21:51 |
| fungi | vishy: bridge support compiled as a module? loaded? | 21:52 |
| *** markvan has quit IRC | 21:52 | |
| lifeless | mordred: and the DB question is one of those | 21:53 |
| fungi | vishy: it's still in the 3.19 ip-sysctl.txt doc | 21:53 |
| mordred | lifeless: yes, I agree | 21:53 |
| *** MaxV has quit IRC | 21:54 | |
| mordred | lifeless: we still have not yet learned very well how to say no to things | 21:54 |
| vishy | fungi: yes the module is loaded | 21:54 |
| vishy | lsmod shows it | 21:54 |
| *** markvan has joined #openstack-infra | 21:54 | |
| fungi | unfortunately i don't have any 3.19 systems handy | 21:54 |
| vishy | http://paste.openstack.org/show/197741/ | 21:55 |
| dstufft | openstack dictating a DB would likely be good for the project | 21:55 |
| vishy | is there something that needs to run post module load to create the psuedofilesystem entries? | 21:55 |
| mordred | dstufft: yes, it certainly would be | 21:55 |
| mordred | dstufft: however, so many other battles to fight | 21:56 |
| fungi | vishy: not to my knowledge. /proc/sys is just a window into the kernel so it should reflect real-time what the kernel knows | 21:56 |
| lifeless | vishy: no, /proc is live from the kernel | 21:56 |
| vishy | so i am totally confused as to what might have happened then | 21:57 |
| dstufft | mordred: I am well acquainted with that feel | 21:57 |
| *** baoli has quit IRC | 21:58 | |
| *** signed8b_ has joined #openstack-infra | 21:58 | |
| fungi | vishy: oh, did you use brctl to add a bridge? if there are no bridge interfaces that might not be present | 21:58 |
| vishy | there are bridge interfaces | 21:58 |
| *** baoli has joined #openstack-infra | 21:58 | |
| *** mattfarina has quit IRC | 21:58 | |
| * fungi is out of ideas | 21:58 | |
| russellb | probably systemd's fault | 21:59 |
| * fungi hands vishy a "blame lennart" tee shirt | 21:59 | |
| *** sdake_ has joined #openstack-infra | 21:59 | |
| vishy | possibly | 21:59 |
| vishy | there are two possibilities | 21:59 |
| vishy | a) something about running this new kernel on an old ubuntu build makes them not show up | 22:00 |
| *** ociuhandu has quit IRC | 22:00 | |
| vishy | b) the /proc/sys/net/bridge/bridge-nf-* settings were removed from the bridge module | 22:00 |
| *** tonytan_brb has joined #openstack-infra | 22:00 | |
| *** tonytan_brb has quit IRC | 22:00 | |
| *** signed8bit has quit IRC | 22:00 | |
| fungi | oh... /proc/sys/net/bridge is there but the bridge-nf-* knobs are absent? | 22:01 |
| *** amitgandhinz has quit IRC | 22:01 | |
| *** tonytan_brb has joined #openstack-infra | 22:01 | |
| openstackgerrit | Ramy Asselin proposed openstack-infra/zuul: Update zuul-changes.py doc to match usage https://review.openstack.org/169527 | 22:01 |
| fungi | the netfilter bridge support module is separate from the bridge interface support module. you might not have the netfilter one loaded | 22:01 |
| *** david-lyle has quit IRC | 22:02 | |
| vishy | lololol | 22:02 |
| vishy | [ 133.282713] bridge: automatic filtering via arp/ip/ip6tables has been deprecated. Update your scripts to load br_netfilter if you need this. | 22:02 |
| vishy | in dmesg | 22:02 |
| fungi | logs ftw! | 22:02 |
| *** sdake has quit IRC | 22:03 | |
| *** rlandy has quit IRC | 22:03 | |
| *** tonytan4ever has quit IRC | 22:04 | |
| *** nelsnelson has quit IRC | 22:04 | |
| *** harlowja_ is now known as harlowja_away | 22:04 | |
| * ianw is trying very hard not to go postal at the tech support person telling me my ip address is 0.0.0.0 | 22:07 | |
| *** harlowja_away is now known as harlowja_ | 22:07 | |
| tchaypo | jeblair: I can’t find the patches I thought I read adding support for gertty using a proxy | 22:08 |
| tchaypo | did I dream about that? | 22:08 |
| tchaypo | ianw: that’s … special.. | 22:08 |
| taron | you are a far better human than I then, I'd just lose it | 22:08 |
| jeblair | tchaypo: not ringing a bell, i think it was all a dream | 22:09 |
| openstackgerrit | Aaron Rosen proposed openstack-infra/project-config: Move congress projects from stackforge to openstack https://review.openstack.org/169530 | 22:09 |
| tchaypo | it was a pleasant dream | 22:09 |
| tchaypo | I wonder if requests has magical support? | 22:09 |
| *** salv-orl_ has quit IRC | 22:09 | |
| *** salv-orlando has joined #openstack-infra | 22:13 | |
| anteaya | tchaypo: if you can get more nodes in your dream work, that is awesome | 22:13 |
| anteaya | tchaypo: bring that code over to this side | 22:13 |
| *** SumitNaiksatam has joined #openstack-infra | 22:14 | |
| fungi | tchaypo: requests has proxy support via the usual envvars afaik | 22:14 |
| * fungi heads out to get dinner, back in a little while | 22:14 | |
| tchaypo | yeah, it seems to be working | 22:15 |
| *** mtanino has joined #openstack-infra | 22:15 | |
| tchaypo | but it seems to not be a solution to my problem | 22:15 |
| fungi | proxies are far more often the source of problems than the solution to them | 22:15 |
| tchaypo | from looking at the logs, I *think* that gertty might be slow because it opens many many connections to https://review.openstack.org | 22:15 |
| tchaypo | I was hoping to be able to use squid to persist a single connection and pipeline requests | 22:16 |
| tchaypo | but I’d forgotten about the ’s’ in https | 22:16 |
| tchaypo | so I think my next bet is to look to see if requests has some kind of connection pooling, and if gertty already makes use of that | 22:17 |
| *** jlvillal_ has joined #openstack-infra | 22:19 | |
| *** harlowja_ has quit IRC | 22:21 | |
| *** tonytan_brb has quit IRC | 22:23 | |
| *** asettle has joined #openstack-infra | 22:23 | |
| *** tnovacik has quit IRC | 22:23 | |
| *** medieval1 has quit IRC | 22:24 | |
| openstackgerrit | James E. Blair proposed openstack-infra/nodepool: Add a simulator script https://review.openstack.org/169539 | 22:26 |
| openstackgerrit | James E. Blair proposed openstack-infra/nodepool: Add API sleeps to fakes https://review.openstack.org/169540 | 22:26 |
| jeblair | yolanda: ^ | 22:26 |
| *** harlowja has joined #openstack-infra | 22:26 | |
| jeblair | yolanda: the first change is a very simple approximation of the nodepool api request queue model. with it, you can make changes and run very approximate simulations quickly. | 22:27 |
| jeblair | yolanda: based on that, assuming i've modeled things close enough, i would expect the two-provider-queue change to result in a 60% increase in the average number of nodes we can produce over a 24 hour period under full load. | 22:28 |
| *** ociuhandu has joined #openstack-infra | 22:28 | |
| jeblair | (that script runs in a couple of seconds) | 22:28 |
| jeblair | yolanda: the second change will let you test out changes more accurately by using nodepool fakes instead of real quota. It's not quite complete yet -- it needs delays added for some more API calls, but it's mostly there. | 22:29 |
| jeblair | yolanda: it runs in real-time though, so it's a better second-level test. | 22:29 |
| *** ayoung has quit IRC | 22:29 | |
| jeblair | mordred, clarkb, fungi: ^ fyi | 22:30 |
| mordred | jeblair: you are a flashing unicorn of pork | 22:30 |
| tchaypo | nope, as far as I can tell it’s already pipelining everything through a single connection | 22:30 |
| *** carl_baldwin has quit IRC | 22:31 | |
| jeblair | cody-somerville: ^ | 22:31 |
| jeblair | mordred: i also noticed that the server list call is surprisingly slow... | 22:31 |
| jeblair | mordred: do you know if it suffers from the same lock issue? | 22:31 |
| mordred | jeblair: I do not know that | 22:32 |
| jeblair | it doesn't hurt us nearly as much as the delay in create/delete, because we cache it and use it for all of our servers | 22:32 |
| jeblair | but it still takes 10 seconds | 22:32 |
| *** carl_baldwin has joined #openstack-infra | 22:32 | |
| jeblair | (which, btw, i think means that the priority queue change would in fact be very dangerous) | 22:32 |
| mordred | jeblair: agree | 22:32 |
| mordred | jeblair: well, I would imagine it's going to be listing that same table | 22:33 |
| jeblair | (i did not run that one through the simulator) | 22:33 |
| mordred | jeblair: so I _betcha_ if you stopped nodepool such that there was no activity on our account | 22:33 |
| mordred | and then ran a list (waiting a minute to let stale DB things fall away) | 22:33 |
| mordred | it would run more quickl | 22:33 |
| mordred | quickly | 22:33 |
| mordred | I'm betting its' table locks behind queued events | 22:33 |
| mordred | *guess* | 22:33 |
| jeblair | *nod* | 22:33 |
| *** dizquierdo has quit IRC | 22:34 | |
| mordred | jeblair: I think we can do much smarter caching aroudn that too, btw - but I have not _EVEN_ begun to code up examples of that | 22:35 |
| *** timcline has quit IRC | 22:35 | |
| jeblair | also, it's worth noting that the simulator does not take into account the possibility that near-continuous use of the table lock by one task queue might cause even worse performance by the second (polling) task queue. that we will have to observe experimentally. | 22:35 |
| mordred | ++ | 22:35 |
| jeblair | mordred: how? | 22:35 |
| *** weshay has quit IRC | 22:35 | |
| mordred | jeblair: I believe most of the time we know when we're manipulating somethign that should cause a change in teh status of a nova server | 22:36 |
| jeblair | mordred: the polling is about waiting for something on the cloud side to change | 22:36 |
| mordred | yes | 22:36 |
| clarkb | ok I have train tickets to vancouver \o/ now if I can just manage to make the trip :) | 22:36 |
| mordred | but we can poll with a show rather than a list | 22:36 |
| anteaya | clarkb: what are your chances? | 22:36 |
| *** packet has quit IRC | 22:36 | |
| anteaya | clarkb: I'm still going with you will have babies before summit | 22:36 |
| mordred | jeblair: and that operation always needs to be direct to the cloud, because it makes no sense for it not to be | 22:37 |
| jeblair | mordred: we list so that we can perform one call and get 600 results, rather than 600 calls | 22:37 |
| mordred | jeblair: oh right | 22:37 |
| mordred | jeblair: sorry - my brain was in a different context | 22:37 |
| clarkb | anteaya: last round of Dr appointments were still full of shrugs, basically we are doing our best but distinct possibility I don't go | 22:37 |
| jeblair | mordred: we used to hit the server directly, but we spent all of our time polling | 22:37 |
| *** whoops has quit IRC | 22:37 | |
| mordred | jeblair: ignore me | 22:37 |
| anteaya | clarkb: for which I will be happy for you (yay babies) and sad for me since you won't be at summit | 22:37 |
| jeblair | mordred: no prob | 22:37 |
| *** aduarte has joined #openstack-infra | 22:37 | |
| anteaya | clarkb: shrugs are good | 22:37 |
| *** Rockyg has quit IRC | 22:37 | |
| mordred | jeblair: if only we had some sort of notification system ... | 22:37 |
| anteaya | clarkb: means the babies could sit a while longer, better for them | 22:37 |
| *** jlvillal_ has quit IRC | 22:38 | |
| *** annegentle has quit IRC | 22:38 | |
| jeblair | mordred: some way to make messages accessible across the internet | 22:38 |
| *** Krinkle has joined #openstack-infra | 22:38 | |
| anteaya | clarkb: but yay tickets and I hope you get to attend summit | 22:38 |
| mordred | jeblair: I don't know - I'm not sure mongodb is designed for that | 22:38 |
| jeblair | haha | 22:38 |
| mordred | jeblair: I mean, you could probably mongo the redis into a rabbit and then erlang something with a twisted deferred ... | 22:39 |
| jeblair | clarkb: i hope you make it but will also be happy for you if you don't :) | 22:39 |
| Krinkle | I'm looking for a way in Zuul to declare a post-merge job should doesn't queue. Similar to how Jenkins behaves by default when not allowing concurrency. Except that, due to the builds being parametized, this Jenkins behaviour doesn't happen. instead one gets a backlog of redundant post-merge jobs that run synchronously (e.g. generating documentation or doing code coverage). Basically to have it eithe | 22:40 |
| Krinkle | r ignore triggers if one is already running, or dequeue/cancel existing builds and queue a new one (like the dependenant pipeline) either would be cool. | 22:40 |
| greghaynes | clarkb: 516 train? | 22:41 |
| clarkb | greghaynes: ya | 22:41 |
| greghaynes | sweet, me too | 22:41 |
| clarkb | on saturday | 22:41 |
| *** medieval1 has joined #openstack-infra | 22:41 | |
| greghaynes | yep | 22:41 |
| *** agentle_ has joined #openstack-infra | 22:41 | |
| *** annegentle has joined #openstack-infra | 22:41 | |
| mordred | Krinkle: as in, you want "build docs for $project for commit X" to skip commit Y if Z has already come in by the time you're ready to build it again, yes? | 22:42 |
| *** sigmavirus24 is now known as sigmavirus24_awa | 22:42 | |
| Krinkle | mordred: Yeah, post-merge we trigger builds to generate documentation, but it takes 10 minutes, and if a minute later another one has been merged it should either cancel the running one or drop the event. Either would be fine. | 22:42 |
| jeblair | Krinkle: i don't think there's a way to do that out of the box; i think we'd need to add functionality to recognize that a commit is a later commit on the same ref. that shouldn't be too hard to do. | 22:43 |
| Krinkle | By target branch basically | 22:43 |
| *** marcusvrn has quit IRC | 22:43 | |
| jeblair | yeah | 22:43 |
| Krinkle | making it refer to a branch instead of a zuul/ref. Just queue a build for 'master'. It would have the benefit of picking up the latest commit if it takes a while to get a free executor slot. | 22:44 |
| jeblair | Krinkle: you would need to implement the "isUpDateOf()" method on the Ref class in zuul/model.py | 22:44 |
| jeblair | er "isUpdateOf" | 22:44 |
| jeblair | that currently always returns false | 22:44 |
| *** andreykurilin_ has joined #openstack-infra | 22:45 | |
| Krinkle | I noticed today we've got a release day (once a month) and there are lots of backports all being merged around the same time and our doc publisher is basically just spending the entire day generating documetnation for old commits. | 22:45 |
| *** bhunter71 has quit IRC | 22:45 | |
| tchaypo | jeblair: so tailing the gertty logs, it looks like every time I restart gertty it starts scanning down the list of projects alphabetically - which means it spends a lot of time grovelling through all the diskimage-builder changes it last synced a few minute ago and never gets down to the tripleo projects it hasn’t synced in ages | 22:46 |
| *** agentle_ has quit IRC | 22:47 | |
| clarkb | Krinkle: as a simple alternative you could trigger those jobs on tags instead of merges | 22:47 |
| *** marcusvrn has joined #openstack-infra | 22:47 | |
| clarkb | (which is moving the problem not solving it) | 22:47 |
| Krinkle | clarkb: Hm.. that would work for release branches which eventually get tagged. | 22:47 |
| *** dimtruck is now known as zz_dimtruck | 22:47 | |
| Krinkle | Wouldn't help for 'master' though. | 22:48 |
| *** jlvillal_ has joined #openstack-infra | 22:48 | |
| *** jlvillal_ has quit IRC | 22:49 | |
| *** devlaps has joined #openstack-infra | 22:50 | |
| *** packet has joined #openstack-infra | 22:50 | |
| *** salv-orlando has quit IRC | 22:51 | |
| *** dustins has quit IRC | 22:52 | |
| *** sabeen has quit IRC | 22:52 | |
| openstackgerrit | Aaron Rosen proposed openstack-infra/project-config: Move congress projects from stackforge to openstack https://review.openstack.org/169530 | 22:53 |
| *** devlaps has quit IRC | 22:54 | |
| *** salv-orlando has joined #openstack-infra | 22:59 | |
| arosen | arosen | 22:59 |
| *** claudiub has quit IRC | 23:04 | |
| *** baoli has quit IRC | 23:05 | |
| *** dboik_ has joined #openstack-infra | 23:05 | |
| mordred | mordred | 23:05 |
| arosen | whoops * | 23:06 |
| arosen | :P | 23:06 |
| arosen | Hi, just curious if you guys have seen these kernel issues in the upstream ci? http://logs2.aaronorosen.com/78/169278/1/check/dsvm-tempest-full-congress-nodepool/46c1873/logs/syslog.txt.gz#_Mar_31_22_53_49 | 23:06 |
| arosen | I'm using the nodepool with dib and trusty | 23:07 |
| *** thedodd has quit IRC | 23:07 | |
| *** dboik has quit IRC | 23:08 | |
| *** bswartz has joined #openstack-infra | 23:09 | |
| *** jaypipes has quit IRC | 23:09 | |
| *** dboik_ has quit IRC | 23:10 | |
| salv-orlando | I opened gerrit and clippy popped up! Is it a fool's day joke or do I have to burn my laptop? | 23:11 |
| openstackgerrit | greghaynes proposed openstack-infra/system-config: Give some actions to our friend https://review.openstack.org/169509 | 23:11 |
| anteaya | salv-orlando: you can review the patch to remove it | 23:11 |
| salv-orlando | anteaya: ok... doing that ;) | 23:12 |
| anteaya | salv-orlando: https://review.openstack.org/#/c/169501/ | 23:12 |
| openstackgerrit | greghaynes proposed openstack-infra/system-config: Give some actions to our friend https://review.openstack.org/169509 | 23:13 |
| anteaya | salv-orlando: you are also welcome to burn your laptop, send a photo if you do | 23:13 |
| anteaya | nibalizer: is that the patch folks should be reviewing? | 23:13 |
| *** annegentle has quit IRC | 23:13 | |
| nibalizer | anteaya: there are two reverts in flight | 23:14 |
| *** dims_ has quit IRC | 23:14 | |
| nibalizer | mine and jims | 23:14 |
| anteaya | jeblair nibalizer who wins in the patch | 23:14 |
| anteaya | yeah, can we pick one | 23:14 |
| nibalizer | i can abandon mine | 23:14 |
| anteaya | so salv-orlando's review hits the right one? | 23:14 |
| nibalizer | though his has a -1 saying use mine so thats confusing | 23:14 |
| anteaya | yes | 23:14 |
| anteaya | so jeblair how do you feel? | 23:14 |
| nibalizer | mine is abandoned | 23:15 |
| anteaya | jeblair: you have 508, nibalizer has 501 | 23:15 |
| anteaya | okay | 23:15 |
| *** davideagnello has quit IRC | 23:15 | |
| *** pcrews_ has joined #openstack-infra | 23:15 | |
| anteaya | salv-orlando: review jim's at 508 please | 23:15 |
| *** pcrews has quit IRC | 23:15 | |
| *** sputnik13 has quit IRC | 23:15 | |
| *** dims_ has joined #openstack-infra | 23:17 | |
| *** bhunter71 has joined #openstack-infra | 23:17 | |
| *** pcrews__ has joined #openstack-infra | 23:18 | |
| *** pcrews__ has quit IRC | 23:18 | |
| openstackgerrit | James E. Blair proposed stackforge/gertty: Sync starred changes regardless of subscription https://review.openstack.org/168508 | 23:19 |
| openstackgerrit | James E. Blair proposed stackforge/gertty: Sync own changes regardless of subscription https://review.openstack.org/168181 | 23:19 |
| openstackgerrit | James E. Blair proposed stackforge/gertty: Support searching by URL https://review.openstack.org/169371 | 23:19 |
| openstackgerrit | James E. Blair proposed stackforge/gertty: Add checkout and cherry-pick to change list https://review.openstack.org/168510 | 23:19 |
| openstackgerrit | James E. Blair proposed stackforge/gertty: Add held changes https://review.openstack.org/167838 | 23:19 |
| openstackgerrit | James E. Blair proposed stackforge/gertty: Highlight starred changes in list https://review.openstack.org/166572 | 23:19 |
| *** carl_baldwin has quit IRC | 23:20 | |
| *** pcrews_ has quit IRC | 23:20 | |
| *** mwagner_lap has joined #openstack-infra | 23:21 | |
| jeblair | anteaya: i wasn't planning on merging the revert change until thursday | 23:21 |
| *** Sukhdev has quit IRC | 23:22 | |
| *** marun has quit IRC | 23:22 | |
| *** medieval1 has quit IRC | 23:23 | |
| greghaynes | I was just able to test 169509 in prod if theres any interest in adding some more life to our friend. I can update the revert once it goes in | 23:23 |
| *** jaypipes has joined #openstack-infra | 23:23 | |
| *** bhunter71 has quit IRC | 23:23 | |
| *** otter768 has joined #openstack-infra | 23:24 | |
| *** hichihara has joined #openstack-infra | 23:24 | |
| *** asettle has quit IRC | 23:24 | |
| *** davideagnello has joined #openstack-infra | 23:24 | |
| jeblair | greghaynes: do you know if it works with old and new change screens? | 23:28 |
| anteaya | jeblair: oh okay | 23:28 |
| *** otter768 has quit IRC | 23:28 | |
| anteaya | salv-orlando: burn your laptop | 23:28 |
| reed | is there documentation for the process (TC approval and all) to move from stackforge to openstack/? | 23:28 |
| greghaynes | jeblair: easy to test, I believe I am using the old one, how do I get to the new? | 23:29 |
| openstackgerrit | Nan Liu proposed openstack-infra/puppet-elasticsearch: Use archive to fetch/checksum instead of wget. https://review.openstack.org/164908 | 23:29 |
| jeblair | greghaynes: https://review.openstack.org/#/settings/preferences | 23:30 |
| nibalizer | greghaynes: or s/c/c2/ in your url | 23:30 |
| anteaya | reed: http://docs.openstack.org/infra/manual/creators.html#add-project-to-the-governance-repository | 23:30 |
| *** melwitt has quit IRC | 23:30 | |
| anteaya | reed: is this for the new contributors work you are doing? | 23:30 |
| reed | anteaya, no, it's for this project ... https://github.com/stackforge/networking-bgpvpn | 23:31 |
| reed | and the fact that they referenced github can be considered a sign | 23:31 |
| *** andreykurilin_ has quit IRC | 23:31 | |
| reed | but the readme lists git.openstack as the source, which is good | 23:32 |
| ianw | why have we added something that auto-plays sound and can't be turned off ... it's rather obnoxious even as a joke. at least you could turn the real clippy off | 23:33 |
| jeblair | nibalizer: there is sound? | 23:33 |
| *** unicell has quit IRC | 23:34 | |
| anteaya | reed: why do they need to be in openstack? | 23:34 |
| reed | anteaya, that page on docs is not enough for my purpose. It doesn't describe or point to the requirements that the TC expects, like diversity of contributors and criteria for the voting etc | 23:34 |
| *** unicell has joined #openstack-infra | 23:34 | |
| reed | anteaya, I don't know, that was not their question :) | 23:34 |
| jeblair | reed: the tc does not require diversity of contributors | 23:34 |
| jeblair | reed: the tc requirements are in the governance repo | 23:34 |
| anteaya | reed: that is the resolution in the governance repo | 23:34 |
| jeblair | reed: you can point there | 23:34 |
| reed | jeblair, anteaya: heading to the gov repo, thanks | 23:34 |
| anteaya | is it taboo to ask people their motivation for their actions? | 23:35 |
| * anteaya still does so | 23:35 | |
| nibalizer | jeblair: yes | 23:35 |
| clarkb | jeblair: apparently there is sound. I operate with sound off as a general rule so didn't notice | 23:35 |
| nibalizer | and im not sure if we can turn that off | 23:35 |
| *** wenlock has quit IRC | 23:35 | |
| reed | anteaya, no, I just don't have time to engage at that level every time | 23:35 |
| anteaya | reed: I can understand that | 23:35 |
| jeblair | nibalizer, clarkb: i think we need to revert then :( | 23:35 |
| clarkb | or turn sound off | 23:36 |
| anteaya | reed: to be honest they should familiarize themselves with the whole governance repo | 23:36 |
| reed | http://governance.openstack.org/reference/new-projects-requirements.html | 23:36 |
| jeblair | clarkb: yeah, but nibalizer says that's an unknown | 23:36 |
| reed | that's what I was looking for | 23:36 |
| clarkb | https://github.com/smore-inc/clippy.js/blob/master/src/animator.js#L160 just comment out that line | 23:37 |
| greghaynes | jeblair: does not work on new one, and looks like theres not really an easy way to make it work there | 23:38 |
| *** changbl has joined #openstack-infra | 23:38 | |
| greghaynes | jeblair: but, it also doesnt fail there, just doesnt do anything new | 23:38 |
| *** tkelsey has joined #openstack-infra | 23:38 | |
| jeblair | nibalizer: can you fix the sound thing quickly? | 23:38 |
| nibalizer | jeblair: i think so | 23:39 |
| greghaynes | also, trying to figure out how gwt generated js works is a bit painful | 23:39 |
| *** achanda has quit IRC | 23:41 | |
| jeblair | i have to run now; unless another core is around to babysit the fix, i think we need to revert it. | 23:42 |
| *** sputnik13 has joined #openstack-infra | 23:42 | |
| clarkb | I can merge the fix if it hsows up | 23:42 |
| jeblair | kk | 23:42 |
| *** tkelsey has quit IRC | 23:42 | |
| greghaynes | nibalizer: I think I have a fix | 23:45 |
| greghaynes | nibalizer: if you want me to push | 23:45 |
| openstackgerrit | Aaron Rosen proposed openstack-infra/project-config: Move congress projects from stackforge to openstack https://review.openstack.org/169530 | 23:45 |
| nibalizer | greghaynes: doit | 23:45 |
| *** signed8b_ is now known as signed8bit_ZZZzz | 23:45 | |
| nibalizer | i tried just commenting out the line clarkb pointed out and minifying but thats failing on my testing rig | 23:45 |
| nibalizer | ofc my testing rig is kinda janky so that could be just me | 23:45 |
| *** pcrews has joined #openstack-infra | 23:47 | |
| openstackgerrit | greghaynes proposed openstack-infra/system-config: Disable our friend's noise https://review.openstack.org/169567 | 23:47 |
| clarkb | greghaynes: I assume you have tested? | 23:47 |
| greghaynes | Yes, let me doubletest though | 23:48 |
| clarkb | kk | 23:48 |
| greghaynes | yep, no sound and seems to work | 23:49 |
| greghaynes | (if anyone else wants to test just paste the js into your browser console without date restriction) | 23:49 |
| nibalizer | i've got gregs deployed here and i can't get any sound out of it | 23:51 |
| nibalizer | http://reviewtest.cloud.nibalizer.com/#/c/1 | 23:51 |
| clarkb | its approved expect it to be everywhere shortlyish | 23:51 |
| *** hdd has quit IRC | 23:54 | |
| *** dannywilson has quit IRC | 23:54 | |
| *** wicope has quit IRC | 23:54 | |
| *** hdd has joined #openstack-infra | 23:55 | |
| harlowja | is review.openstack.org slow for anyone else? | 23:55 |
| *** achanda has joined #openstack-infra | 23:55 | |
| harlowja | the images for reviews are taking forevdr to load | 23:55 |
| harlowja | sorta weird | 23:56 |
| clarkb | harlowja: it does that occasionally because memory leak | 23:56 |
| harlowja | ah | 23:56 |
| clarkb | I can check how it is doing | 23:56 |
| clarkb | hrm we don't appear to be leaking at least its not terribad like normal | 23:56 |
| *** smccully has quit IRC | 23:56 | |
| harlowja | hmmm, maybe just something else then, weird | 23:57 |
| *** smccully_ has quit IRC | 23:57 | |
| *** ayoung has joined #openstack-infra | 23:57 | |
| clarkb | looking at chromium dev tools it seems to be spending its time in gerrit api rpc calls | 23:59 |
| clarkb | also a fair bit of time looking up avatars that don't exist for our users | 23:59 |
| clarkb | zaro: ^ any way to turn that off? | 23:59 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!