17:01:28 <krtaylor> #startmeeting ironic-qa 17:01:29 <openstack> Meeting started Wed Dec 2 17:01:28 2015 UTC and is due to finish in 60 minutes. The chair is krtaylor. Information about MeetBot at http://wiki.debian.org/MeetBot. 17:01:31 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 17:01:33 <openstack> The meeting name has been set to 'ironic_qa' 17:01:34 <jroll> ohai 17:01:44 <krtaylor> anyone here for ironic qa meeting? 17:01:48 <thingee> o/ 17:01:49 <sambetts> o/ yup 17:01:51 <sinval> \o 17:02:02 <[1]cdearborn> \o/ 17:02:07 <liliars> o/ 17:02:08 <jroll> /o\ 17:02:15 <krtaylor> ah, ok, whew :) 17:02:31 <krtaylor> hi everyone 17:03:00 <krtaylor> I am running the meeting this week, I am assuming jlvillal is still in class 17:03:13 <krtaylor> here is the agenda, which is very light 17:03:20 <mjturek1> o/ 17:03:25 <krtaylor> #link https://wiki.openstack.org/wiki/Meetings/Ironic-QA 17:03:56 <krtaylor> so this may be a short one 17:04:16 <krtaylor> anyone have anything to add to grenade or functional status? 17:04:53 <sinval> about functional, I started a simple change (WIP) for publishing driver interfaces functional tests 17:05:25 <krtaylor> sinval, great, link? 17:06:02 <jroll> sinval: yeah, looking at those, any reason not to make them part of tempest runs? AIUI functional tests aren't meant to deal with hardware and such 17:06:02 <sinval> there you go: https://review.openstack.org/#/c/249515/ 17:06:10 <jroll> (I feel like we talked about this once already though) 17:06:33 <krtaylor> #link https://review.openstack.org/#/c/249515/ 17:07:18 <sinval> jroll, I have some thoughts about this, would be nice to have you suggestions on the patch, I'll update it, since I'm more aware how the functional tests thing is working in Ironic and python-ironicclient, so, yes 17:07:22 <rpioso> o/ 17:07:26 <jroll> sinval: sure thing 17:07:27 * krtaylor looks 17:08:25 <rajinir_> o/ 17:08:27 <sinval> the code is still not finished, it is just a sketch for ideas and discussion... 17:08:52 <krtaylor> #topic Grenade and Functional testing 17:09:01 <krtaylor> forgot to do that... 17:10:35 <krtaylor> ok, anything else on this topic? 17:11:07 <krtaylor> sinval, I'll review that also 17:11:17 <sinval> krtaylor, thanks 17:11:44 <krtaylor> ok, well, onward then 17:11:54 <krtaylor> #topic third party CI 17:12:07 <krtaylor> so, the spec got a bunch of great reviews 17:12:51 <krtaylor> #link https://review.openstack.org/#/c/241294/ 17:13:16 <krtaylor> I had assumed that we were going to merge and deal with the specifics in the docs 17:13:39 <krtaylor> but, I can address these in the spec and push a new version if needed 17:13:42 <krtaylor> comments? 17:13:56 <jroll> krtaylor: yeah, can you push a new one and we can land it? 17:14:01 <jroll> I'm sad that isn't merged yet :( 17:14:36 <thingee> seems like there is enough consensus without showstoppers 17:15:07 <thingee> also would like to note communication went out last monday to the list as well to driver maintainers I've identified via git http://lists.openstack.org/pipermail/openstack-dev/2015-December/080867.html 17:15:34 <krtaylor> yes, that was on my list to thingee 17:15:36 <jroll> thingee: \o/ thank you for that 17:15:38 <krtaylor> thanks for that 17:15:39 <liliars> yeah, I saw that, thxx thingee 17:15:51 <thingee> krtaylor: sorry for jumping ahead :) 17:15:57 <krtaylor> also thingee did this -> #link https://wiki.openstack.org/wiki/Ironic/Testing#Third_Party_CI 17:16:13 <thingee> thanks all for the help in getting this out. 17:16:13 <krtaylor> since we are at it, thanks for that too! 17:17:52 <krtaylor> jroll, I'll be happy to address the comments and push a new version of the spec, and to seek approval quickly 17:18:15 <jroll> krtaylor: yeah, I'll be around for a bit tomorrow morning, otherwise can review monday if you just ping me 17:19:02 <jroll> (ping me when it's up, that is) 17:19:21 <krtaylor> will do 17:19:56 <krtaylor> thingee, any questions or comments off the maillist, I haven't seen any traffic on it 17:20:24 <krtaylor> re: the driver communication email 17:21:00 <thingee> nope 17:21:16 <thingee> they will come though 17:21:28 <thingee> unfortunately as we near the deadline 17:21:31 <krtaylor> without doubt :) 17:21:56 <krtaylor> ok, well, anything else on this? 17:22:22 <liliars> I'm happy to say our team managed to overcome network issues and our CI is finally able to test a deploy workflow \o/ just finishing some automatizing scripts for that and we'll put it up 17:22:52 <liliars> if anyone has got issues and want some tips, feel free to ping me and (I guess hah) sinval :) 17:22:56 <sambetts> \o/ 17:22:58 <krtaylor> liliars, amazing how close we are, my team has just done the same 17:22:59 <sinval> We also saw new statements on CI spec regarding the tests that every CI should run (dsvm-pxe_ipa), we need to work on this 17:23:07 <krtaylor> maurosr, mjturek1 ^^ 17:23:17 <liliars> krtaylor, cool!! 17:23:40 <mjturek1> yep, hoping to be commenting soon :) 17:23:49 <krtaylor> we are expecting to start soon 17:23:54 <krtaylor> yes :) 17:24:14 <sambetts> my progress has been slowed re: the cisco ci because of a lack of space for equipment, more progress should be happening in the new year :( 17:24:47 <krtaylor> excellent, any other systems coming online soon? 17:25:44 * krtaylor is interested in comparing notes for how the CI systems were set up 17:26:08 <krtaylor> ok, anything else? 17:26:16 <krtaylor> next then 17:26:21 <krtaylor> #topic General QA and Open Discussion 17:26:39 <sambetts> do we need to reiterate Anita's message? 17:26:46 <krtaylor> which one? 17:26:55 <jroll> sambetts: do iiiiit 17:27:11 <krtaylor> not to announce on -dev? 17:27:22 <sambetts> yeah and to use the third party wiki 17:27:51 <mjturek1> sorry, might be out of the loop. Not announce what exactly? 17:27:55 <krtaylor> yeah, the history there is that when we started seeing a lot of systems come up, there was a lot of noise, so we started the wiki 17:28:33 <krtaylor> mjturek1, there was an ironic CI system announce they were reporting 17:28:44 <jroll> #link http://lists.openstack.org/pipermail/openstack-dev/2015-November/080808.html 17:28:58 <jroll> ^ that 17:29:14 <mjturek1> ahhhhh got it, thanks jroll and krtaylor 17:29:14 <krtaylor> I tried for some time to get active status displayed on theta wiki page, but it is difficult because of the wiki formatting 17:29:45 <krtaylor> that wiki page is meant to list all active CI systems and their status 17:30:22 <krtaylor> there are instructions there that show hoe to update the status working around some of the wiki strangeness 17:30:39 <sambetts> we've also got this https://etherpad.openstack.org/p/IronicCI for CI development updates 17:31:17 <krtaylor> yes, and cinder and other projects also have sub-driver testing pages, we can as well 17:31:57 <sambetts> How does this get updated? http://stackalytics.com/report/driverlog?project_id=openstack%2Fironic 17:32:09 <sinval> krtaylor, +1 17:32:24 <jroll> sambetts: https://wiki.openstack.org/wiki/DriverLog#How_To:_Add_a_new_driver_to_DriverLog 17:32:31 <krtaylor> sambetts, there is a patch to add your system 17:32:34 <krtaylor> yes that ^^^ 17:33:03 <jroll> that's pretty out of date, does anyone want to take an action to update that (besides CI updates)? 17:33:26 <sambetts> haha I was about to say should we update it with all the drivers we know about now 17:33:34 <krtaylor> the list or the instructions? 17:34:18 <sambetts> It should be updated with all the drivers listed https://etherpad.openstack.org/p/IronicCI 17:34:40 <sambetts> and then we should have instructions to tell those implementing CI to update their status on there 17:35:17 <krtaylor> the "official" list is the wiki at: #link https://wiki.openstack.org/wiki/ThirdPartySystems 17:36:25 <krtaylor> I would think that stackalytics would mine that...somehow 17:36:44 <jroll> well, that's the official CI list, not driver list, yeah? 17:36:47 <krtaylor> but, alas, multiple places to keep the info synced 17:37:17 <krtaylor> jroll, good point, but there are few systems that have more than one driver 17:38:06 <sambetts> driverlog looks like a nice way to store the info because its a repo full of data, and anything could use that data 17:38:10 <jroll> krtaylor: but there are many drivers without CI systems :P 17:38:25 <krtaylor> hehheh how true 17:39:26 <krtaylor> for us, let's start with our etherpad/wiki and see what we need to add to stackalytics 17:39:56 <krtaylor> as far as I know, stackalytics isn't used for any official purpose for infra or anything 17:40:00 <sambetts> ok :) 17:41:03 <jroll> sure 17:41:42 <krtaylor> sinval, re: pxe_ipa tests 17:42:00 <krtaylor> we were not planning on having ipa testing initially 17:42:46 <sinval> krtaylor, ok thanks 17:45:15 <krtaylor> sinval, so what should be our required base set of tests 17:45:32 <krtaylor> feel free to add that as a comment in the spec 17:45:52 <krtaylor> but the requirement is for N remember :) 17:45:58 <jroll> whoa whoa wait 17:46:06 <jroll> "we were not planning on having ipa testing initially" ??? 17:46:23 <jroll> the bash ramdisk is deprecated, and the recommended deploy ramdisk is IPA 17:46:29 <jroll> there's no reason not to use that for testing 17:46:52 <krtaylor> jroll, we (as in PowerKVM), we don't (yet) support IPA 17:46:59 <jroll> the *agent driver* is different, but dsvm-pxe_ipa is basically our baseline tempest job right now 17:47:02 <jroll> oh. 17:47:45 <krtaylor> I guess what I was poking at was - is that a problem for any other system to have that by N 17:47:46 <sinval> krtaylor: ok, will do, I thought that we were going to discuss about that during the docs development, but talk during the spec sounds reasonable 17:48:07 <krtaylor> sinval, looks like we have another round of comments on the spec :) 17:48:24 <jroll> krtaylor: it's the only supported deployment mechanism by the end of N, so... 17:48:52 <jroll> there shouldn't be any reason why a system doesn't support it, curious why powerkvm can't 17:49:20 <jroll> or why a driver couldn't support it 17:49:56 <krtaylor> jroll, it is probably close for us (powerkvm), we have been focused on IPMI initially to get our test system set up 17:50:16 <jroll> krtaylor: dsvm-pxe_ipa is just pxe_ipmitool with an IPA ramdisk 17:51:10 <krtaylor> yes, we haven't got around to trying that yet 17:51:13 <jroll> which is kind of the "default" or "recommended" deployment right now; which is why we chose it as the baseline test for CI 17:51:14 <jroll> ok 17:51:18 <jroll> it's just linux :D 17:51:32 <krtaylor> yes, but different platform :) 17:52:25 <jroll> indeed 17:52:27 <jroll> :P 17:52:52 <krtaylor> I'll roll up/answer the latest comments on the CI spec, any other actions? 17:53:12 <jroll> +1, thanks for doing that 17:53:30 <krtaylor> it seems we are winding down, anything else? 17:53:36 <sinval> 5 minutes 17:53:41 <sambetts> nothing from me 17:54:12 <krtaylor> thanks everyone, it was another good meeting 17:54:34 <liliars> thanks krtaylor! bye everyone 17:54:39 <sambetts> o/ 17:54:50 <krtaylor> #endmeeting