07:00:16 <skraynev> #startmeeting heat 07:00:16 <openstack> Meeting started Wed Sep 16 07:00:16 2015 UTC and is due to finish in 60 minutes. The chair is skraynev. Information about MeetBot at http://wiki.debian.org/MeetBot. 07:00:17 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 07:00:19 <openstack> The meeting name has been set to 'heat' 07:00:42 <stevebaker> skraynev: can you add me to chair? 07:00:46 <skraynev> #topic rollcall 07:00:52 <asalkeld> o/ 07:00:54 <skraynev> #chair stevebaker 07:00:55 <openstack> Current chairs: skraynev stevebaker 07:00:56 <tspatzier> hi all 07:01:02 <skraynev> stevebaker: sure ;) 07:01:07 <pas-ha> o/ 07:01:10 <tiantian> o/ 07:01:11 <elynn> o/ 07:01:13 <xek> o/ 07:01:19 <skraynev> stevebaker: btw, you can continue ;) 07:01:23 <stevebaker> ok 07:01:36 <stevebaker> skraynev: wait another 23 hours ;) 07:01:38 <shardy> o/ 07:01:42 * skraynev just started :) 07:01:48 <asalkeld> shardy: replied here after ages https://review.openstack.org/#/c/135492/4/doc/source/template_guide/hot_spec.rst (sorry for the delay) 07:01:49 <stevebaker> #topic Adding items to the agenda 07:01:52 <stevebaker> #link https://wiki.openstack.org/wiki/Meetings/HeatAgenda#Agenda_.282015-09-16_0700_UTC.29 07:02:01 <skraynev> stevebaker: :) ok 07:02:41 <skraynev> wow. so short wiki page ;) 07:02:41 <shardy> asalkeld: thanks, will check it out, would be good to un-block that work as it could be really useful 07:02:50 <stevebaker> looks like we have an agenda to merge 07:03:28 <stevebaker> fixed it https://wiki.openstack.org/wiki/Meetings/HeatAgenda#Agenda_.282015-09-16_0700_UTC.29 07:03:51 <stevebaker> oki 07:03:51 <KanagarajM__> hi 07:03:56 <stevebaker> #topic rc1 blueprint status 07:04:18 <stevebaker> #link https://launchpad.net/heat/+milestone/liberty-rc1 07:04:28 <stevebaker> shardy: can you give us an updated on nested-validation? 07:04:42 <shardy> stevebaker: I posted a revised approach to template-validate yesterday, because jdob ran into issues reworking the existing logic 07:04:54 <shardy> I made good progress and I hope to post the final recursive part of it today 07:04:55 <asalkeld> do we have the link to summit session ideas? 07:05:04 <shardy> so, if we're aiming for rc1 next monday it could still make it 07:05:11 <stevebaker> asalkeld: not yet 07:05:13 <skraynev> asalkeld: not yet. 07:05:16 <asalkeld> lol 07:05:25 <asalkeld> ok 07:05:40 <shardy> stevebaker: my approach should have the side-effect of fixing bug #1467573 too 07:05:42 <openstack> bug 1467573 in heat "template-validate doesn't validate the same as stack-create" [Medium,In progress] https://launchpad.net/bugs/1467573 - Assigned to Steven Hardy (shardy) 07:06:02 <stevebaker> shardy: ok, lets see how we go for the rest of the week. Sounds like a reasonable chance it will be ready 07:06:12 <shardy> stevebaker: +1, thanks! 07:06:18 <ramishra> hi all 07:06:28 <stevebaker> skraynev: can you give an update on rich-network-prop? 07:07:30 <stevebaker> shardy: btw I reworked this heatclient patch and it is approved https://review.openstack.org/#/c/222325/ 07:07:37 <shardy> stevebaker: thanks! 07:07:57 <skraynev> stevebaker: I updated patch and Tripleo job looks green on this new patch 07:08:06 <skraynev> it's first in series 07:08:12 <stevebaker> there will be a heatclient release as soon as https://review.openstack.org/#/c/222325/ and https://review.openstack.org/#/c/205755/ land, and this will be the fork for stable/liberty branch 07:08:30 <KanagarajM__> shardy, stevebaker: temlate-validate fails when we use NoneResource, https://review.openstack.org/223528 , it might cross with above bug 07:08:35 <skraynev> but I need couple days to re-work replace for server (due to additional "external" ports) 07:09:39 <skraynev> stevebaker: so please review this https://review.openstack.org/#/c/223169/ 07:09:53 <shardy> KanagarajM__: Yep that should be fixed by the patch I posted yesterday, I think.. 07:10:07 <shardy> KanagarajM__: thanks, I'll check it out and add it to my functional tests 07:10:12 <skraynev> Also feedback from TripleO is appreciated 07:10:38 <stevebaker> skraynev: likely they'll be happy as long as the old behaviour continues 07:11:26 <stevebaker> skraynev: it sounds like the important bits will be ready, and anything left over could be an rc bug 07:11:27 <skraynev> stevebaker: yeah. now we create port only if we have subnet property or port extra_properties ;) 07:11:30 <shardy> skraynev: thanks, I'll chat with derekh today and one of us can verify all is OK but it's good the check job is green now :) 07:11:47 <KanagarajM__> shardy: ok. sure. 07:12:08 <stevebaker> #topic rc1 bug status 07:12:09 <skraynev> stevebaker: yes 07:12:51 <skraynev> shardy: aha. thx. will wait a feedback 07:12:55 <stevebaker> I did a big prune, removing a load of bugs from liberty, and tagging some of those as liberty-rc-potential 07:13:32 <stevebaker> #link https://bugs.launchpad.net/heat/+bugs?field.tag=liberty-rc-potential 07:13:54 <stevebaker> feel free to tag any bug liberty-rc-potential if you believe it to be a blocker for rc1 07:14:19 <stevebaker> the gate is a disaster today, so more is fixed than show up in that list 07:14:52 <asalkeld> #link https://goo.gl/igwqqB 07:14:53 <shardy> stevebaker: anything we need to help through to fix the gate? 07:15:01 <asalkeld> reviews for above bugs 07:15:11 <stevebaker> btw the fix I'm suggesting for https://bugs.launchpad.net/heat/+bug/1489548 will require a schema change, which technically requires a FFE 07:15:12 <openstack> Launchpad bug 1489548 in heat "max_resources_per_stack causes scaling issues on large stacks" [High,In progress] - Assigned to Steven Hardy (shardy) 07:15:31 <asalkeld> shardy: https://review.openstack.org/223747 07:15:33 <stevebaker> shardy: cross your fingers for this to land https://review.openstack.org/223747 07:15:35 <asalkeld> needs to go in 07:16:18 <shardy> asalkeld, stevebaker: we could ping infra and get it bumped to the head of the queue if you've not already done that 07:16:51 <stevebaker> shardy: its the first heat change at least 07:17:09 <shardy_> stevebaker: Ok, cool, will keep an eye on it, thanks 07:17:47 <stevebaker> https://bugs.launchpad.net/bugs/1493906 is High and unassigned 07:17:49 <openstack> Launchpad bug 1493906 in heat "heat.tests.engine.test_engine_worker.CheckWorkflowCleanupTest.test_is_cleanup_traversal_raise_update_inprogress gate failure" [High,Confirmed] 07:18:40 <stevebaker> is that still happening? I haven't seen it 07:18:58 <stevebaker> or is it a race? 07:19:19 <asalkeld> stevebaker: i have never seen that fail locally 07:19:22 <stevebaker> ramishra: ah, you raised that one 07:19:43 <ramishra> stevebaker: yes, its happening but intermintly 07:19:54 <ramishra> I've seen it locally too 07:20:04 <skraynev> I have met similar for test_is_cleanup_traversal_raise_update_(replace or something like that) 07:20:21 <skraynev> it's some strange py34 sporadic errors... 07:20:37 <stevebaker> any takers to look at that? I could just put it in heat-ongoing, liberty-rc-potential 07:21:05 <asalkeld> maybe, i'll see if no one has taken it tomorrow 07:21:26 <stevebaker> ok, I'll leave it in rc1 for now 07:21:51 <pas-ha> it's hard to debug a thing that does not break for you 07:22:08 <stevebaker> indeed 07:22:26 <stevebaker> #topic Liberty release notes 07:22:45 <stevebaker> I've created an empty etherpad here 07:22:47 <stevebaker> #link https://etherpad.openstack.org/p/heat-liberty-release-notes 07:23:28 <stevebaker> if you have a favourite feature then feel free to write something, otherwise I'll start populating with notable blueprints 07:23:34 <pas-ha> should everyone fill in big features (s)he implemented? 07:23:35 <pas-ha> ok 07:23:53 <stevebaker> pas-ha: ideally yes, I'll be going through it all too though 07:24:34 <stevebaker> Some detailed Upgrade Notes will be needed (new config options, etc) 07:25:13 <stevebaker> wha? 07:25:18 <stevebaker> #topic Liberty release notes 07:26:01 <stevebaker> btw they will end up here, but lets iterate on the etherpad for a bit 07:26:03 <stevebaker> #link https://wiki.openstack.org/wiki/ReleaseNotes/Liberty 07:27:17 <stevebaker> #topic gate for heat-templates seems broken 07:27:23 <pas-ha> that's mine 07:27:32 <stevebaker> pas-ha: go 07:27:41 <pas-ha> due to service based resource exposure the gate is broken 07:27:50 <stevebaker> oh, yeah 07:27:54 <pas-ha> we do not install anything besides Heat and keystone there 07:28:06 <pas-ha> so validation fails on most of resources 07:28:43 <pas-ha> what I suggest is to register some fake Keystone endpoints for all the services - might that be an option? 07:29:35 <stevebaker> pas-ha: How about we fix https://bugs.launchpad.net/heat/+bug/1495914 then have an environment which swaps out these resources with None? 07:29:36 <openstack> Launchpad bug 1495914 in heat "template-validate fails when resources are swaped with None resource" [Medium,In progress] - Assigned to Kanagaraj Manickam (kanagaraj-manickam) 07:29:56 <shardy> stevebaker: +1 that sounds good 07:30:16 <stevebaker> because mostly we just want syntax validation 07:30:29 <pas-ha> that sounds good. does None resource supports arbitrary schema? 07:30:45 <KanagarajM__> we may need to swap all the resources in env? 07:30:58 <pas-ha> KanagarajM__, yes, in the global one 07:31:10 <shardy> KanagarajM__: the resource registry supports globbing 07:31:31 <shardy> pas-ha: it does, but it basically accepts any properties and provides any attribute 07:31:32 <stevebaker> we could have an env file in heat-templates which is invoked by the validate call, rather than a system global one 07:31:49 <KanagarajM__> pas-ha, shardy : sure. would be nice choice :) 07:31:52 <shardy> pas-ha: so it's not really testing the resource schema against the templates, only the structural correctness of the template 07:32:15 <pas-ha> ok, seems we have a plan 07:32:30 <stevebaker> pas-ha: is the job non-voting for now? 07:32:43 <pas-ha> voting :( 07:33:10 <pas-ha> example https://review.openstack.org/#/c/217654/ 07:33:40 <stevebaker> it may need to be made non-voting while we sort this out 07:34:22 <stevebaker> oh, look https://review.openstack.org/#/c/217654/5/hot/monasca/autoscale_env.yaml,cm 07:34:43 <stevebaker> maybe a per-template env like ^ is best 07:34:52 <stevebaker> to specify the None resources 07:35:09 <pas-ha> that would be a big commit :) 07:35:42 <KanagarajM__> currently the tools could be updated in the template repo, to consider the global env 07:36:09 * stevebaker actually sees how many errors there are 07:36:25 <KanagarajM__> and in global env, we may use globbing, 07:36:25 <stevebaker> global env ftw 07:36:36 <stevebaker> OS::* ;) 07:36:52 <KanagarajM__> yes 07:36:57 <stevebaker> #topic Open Discussion 07:36:58 <skraynev> stevebaker: AWS:: ? 07:37:02 <pas-ha> https://github.com/openstack/heat-templates/blob/master/tools/validate-templates#L32 07:37:06 <pas-ha> plug it there 07:37:22 <shardy> skraynev: *::*::* ;) 07:37:25 <KanagarajM__> pas-ha: yes 07:37:38 <skraynev> shardy: lol. exactly! 07:37:47 <tiantian> shardy, +1 07:38:55 <skraynev> shardy: what about bug with re-authentication ? 07:39:37 <shardy> skraynev: I'm running out of time for rc1 but I'm still planning to look at it, it may end up having to be a stable backport 07:40:00 <shardy> skraynev: If I can get the nested validation stuff done today I'll look at it later this week 07:40:05 * asalkeld having supper ... 07:40:18 <shardy> skraynev: you're welcome to pick it up if you have time to do it now 07:40:33 <shardy> it should be easier now we have the auth plugin stuff landed for trusts that miguelgrinberg posted 07:40:44 <skraynev> shardy: do you have time for it ? I told with ochuprykov - He may help with it (but he have not experience in this part of code yet ) 07:41:20 <skraynev> shardy: I wanted to help with it, but now looks like rich-network chews all my free time :) 07:41:34 <shardy> skraynev: evidently I don't have time for it, but I'm trying to make some ;) 07:41:42 <pas-ha> shardy: time for possible stupid questions that is ;) 07:41:47 <stevebaker> skraynev: and it will be awesome :) 07:42:01 <shardy> pas-ha: haha ;) 07:42:17 <shardy> skraynev: I will try to look at it this week, I just want to get nested-validation working first 07:43:39 <shardy> skraynev: the quickest fix will be to have a config option which says always use the trust plugin when the stack timeout exceeds the token expiry 07:43:44 <skraynev> shardy; ok. thank you. I will say ochuprykov to ping you ;) if he meet some blockers. 07:44:03 <stevebaker> looks like all the heatclient changes have landed, no more features for liberty heatclient 07:44:06 <skraynev> shardy: aha. got it 07:44:27 <skraynev> stevebaker: +1. 07:44:39 <shardy> skraynev: ideally, we'd do something a bit more clever and switch tokens just before expiry, but that may be harder to get right 07:45:00 <tiantian> stevebaker, https://review.openstack.org/#/c/220921/ hope to land 07:45:06 <tiantian> :) 07:45:24 <shardy> skraynev: anyway, I can chat to ochuprykov about it 07:46:11 <skraynev> shardy: In mind the most part of approaches look awesome - reality puts us on the ground :) 07:46:30 <skraynev> shardy: agreed 07:47:04 <stevebaker> tiantian: its a bug fix, so we can backport it to stable/liberty 07:47:37 <shardy> stevebaker: how do pypi releases work for multiple client branches? 07:47:43 <tiantian> stevebaker, ok 07:47:48 <stevebaker> tiantian: although this stable branch thing is just for gate testing. Downstream are free to distribute later client releases however they choose 07:47:50 * shardy still doesn't get the stable client branches thing at all 07:48:04 <shardy> aha 07:48:23 <stevebaker> shardy: its really just a tool to reduce stable gate breakage 07:48:35 <shardy> So the gate always works and everyone downstream gets to discover the breakage 07:48:39 <shardy> awesome ;) 07:48:57 <pas-ha> I would think on bumping a Y or X in X.Y.Z after stable cut 07:49:11 <pas-ha> so that stable backports/fixes goes to Z 07:49:35 <skraynev> shardy: anyway, sorry if I ask you about it persistently... :) my apologize. I just wanted to be sure, that we do not do the same in parallel. 07:49:38 <stevebaker> pas-ha: yep, only Z changes can go to stable. I increment Y whenever any new feature lands 07:50:11 <shardy> skraynev: sure, I probably won't get to it until Friday, and I'll ping you/ochuprykov before I start to ensure we don't duplicate effort 07:50:17 <shardy> skraynev: thanks for your patience! ;) 07:50:46 <stevebaker> shall we finish up? 07:51:14 <pas-ha> +1, nothing from me 07:51:18 <shardy> +1 07:51:22 <skraynev> +1 07:51:46 <stevebaker> #endmeeting