15:00:19 <therve> #startmeeting heat
15:00:20 <openstack> Meeting started Wed May 25 15:00:19 2016 UTC and is due to finish in 60 minutes.  The chair is therve. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:00:21 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
15:00:23 <therve> #topic Roll call
15:00:24 <openstack> The meeting name has been set to 'heat'
15:00:38 <therve> Hola people of Heat
15:00:41 <ramishra> hi
15:00:50 <jaimguer> o/
15:00:50 <jdob> [T]/
15:00:55 <zaneb> \o
15:01:26 <ochuprykov_> hi
15:01:28 <ananta> Hi
15:01:36 <duvarenkov> hi
15:01:52 <skraynev> hey
15:02:23 <therve> Alright
15:02:36 <_prazumovsky> Hi all!
15:02:39 <therve> #topic Adding items to agenda
15:02:51 <therve> #link https://wiki.openstack.org/wiki/Meetings/HeatAgenda#Agenda_.282016-05-25_1500_UTC.29
15:03:08 <ananta> magnum and sahara gate jobs with convergence engine
15:03:25 <therve> Can you add to the wiki? :)
15:03:29 <ananta> sorry i couldn't add the wiki was asking for a question
15:03:50 <ananta> will try
15:03:55 <shardy> o/
15:04:03 <therve> No worries we'll talk about it in next topic anyway
15:04:10 <therve> #topic Newton-1 status
15:04:17 <therve> So!
15:04:46 <therve> #link https://launchpad.net/heat/+milestone/newton-1
15:04:52 <zaneb> ananta: updated the wiki
15:04:55 <skraynev> yeah :)
15:05:03 <ananta> zaneb: thanks :)
15:05:09 <therve> We have a bunch of bugs in progress
15:05:23 <therve> Nothing that should block the release, AFAICT
15:05:33 <therve> Objections on bugs?
15:06:00 <skraynev> nope
15:06:08 <skraynev> Probably the same for BP ?
15:06:19 <therve> So BP
15:06:37 <therve> Conditionals started getting in
15:06:40 <shardy> What's the status of https://bugs.launchpad.net/heat/+bug/1570983?
15:06:41 <openstack> Launchpad bug 1570983 in heat "raw_template files duplication wastes DB space and memory" [High,In progress] - Assigned to Crag Wolfe (cwolfe)
15:06:56 <shardy> I'm fine with it slipping to n-2, but that's kind of a big deal for TripleO I think
15:07:09 <therve> It is, but tripleo uses master anyway :)
15:07:14 <zaneb> just re-reviewing as we speak, but it looks basically ready
15:07:33 <shardy> therve: we'll be cutting releases for n-1 too FWIW
15:07:44 <therve> I think the dependent patches got in, so we'll be able to move I think
15:07:58 <shardy> zaneb: ack, thanks, I'll pull it and give it a test locally too
15:08:08 <zaneb> ++
15:08:09 <cwolferh> yeah, i also think it is about ready
15:08:33 <shardy> cwolferh: thanks for all the work on it :)
15:08:37 <zaneb> I'll trigger check experimental when I review
15:08:58 <therve> I hope conditionals with be rebased, so that we can review it
15:09:02 <cwolferh> shardy, np, i've liked this one
15:09:04 <therve> yaql got in yesterday, fwiw
15:09:49 <therve> We'll talk convergence on the next topic :). Anything else? skraynev as you added the topic?
15:10:10 <skraynev> therve: we have one week before N1, so potentially Conditions can land, if guys tell, that it looks ready
15:10:27 <skraynev> therve: only this one ^ ;)
15:10:33 <therve> OK cool
15:10:41 <therve> #topic Magnum and Sahara gate jobs with convergence engine
15:10:48 <therve> ananta so where are we?
15:10:53 <ananta> https://review.openstack.org/#/c/316073/
15:11:01 <ananta> and https://review.openstack.org/#/c/316667/
15:11:10 <ananta> the gate jobs  look good
15:11:17 <skraynev> therve: p.s. I forgot to ask you: will you move all necessary bugs, BP to N2 or I need to help ?
15:11:24 <ananta> I had to push some fixes but they are running fine now
15:11:50 <zaneb> so just waiting on https://review.openstack.org/#/c/317866/
15:11:54 <therve> skraynev, I'll try, I'll ask if I need help, thanks :)
15:12:02 <therve> zaneb, Yep
15:12:15 <skraynev> therve: ok. I'll be here ;) so welcome
15:12:19 <therve> ananta, We need to recheck the sahara gate?
15:12:24 <therve> It fails but it looks unrelated
15:12:28 <ananta> yes, https://review.openstack.org/#/c/317866/ and https://review.openstack.org/#/c/319122/
15:12:40 <ananta> yes unrelated
15:12:56 <ananta> I didn't see heat jobs for trove and murano
15:13:00 <therve> It'd be good to have a green build still
15:13:06 <ananta> so I didn't add anything there
15:13:14 <therve> OK
15:13:56 <ananta> I probably need to rebase, some how it failed for last one
15:14:22 <therve> shardy, What do we do for tripleo? Can you push a config change to force convergence to false?
15:14:38 <zaneb> ananta: can you clarify what you mean by "ddidn't see heat jobs for trove and murano"?
15:14:53 <therve> zaneb, They don't run heat in any gate job
15:14:59 <shardy> therve: yes we can do that - I'll look into it
15:15:09 <ananta> zaneb: I went through their gate jobs but didn't see any heat service being configured or running
15:15:25 <therve> shardy, Ideally it'd be in tree so that we can test using it without changing project-config
15:15:25 <zaneb> really? they don't run it at all?
15:15:37 <zaneb> huh, that's surprising
15:15:41 <skraynev> therve: I had a talk with guys from Murano and they asked me about enabling convergence. I suggested to add some tests job for it during n2 and be ready to switch on convergence by default after it ;)
15:15:55 <shardy> therve: it will be a change to the puppet hiera defaults in instack-undercloud most likely
15:16:04 <therve> shardy, Cool sounds good
15:16:16 <shardy> we can then toggle that via overrides in tripleo-ci if we want to e.g have an experimental job with it enabled
15:16:45 <ananta> zaneb: I can easily add one if we see if its not the case
15:17:39 <pas-ha> 0/
15:17:40 <therve> I mean I don't think we should add jobs to other project
15:17:58 <therve> If they don't test it... They'll get the notice and hopefully test manually
15:19:01 <therve> So I feel we're on track to do it by n1, unless someone has an objection
15:19:21 <therve> Please review convergence patches :)
15:19:23 <zaneb> DO IT
15:19:29 <therve> ananta, Anything else?
15:19:36 <ananta> nope
15:19:53 <ananta> just that we need to get the patches reviewd :)
15:20:01 <therve> Yep let's do this :)
15:20:11 <ananta> it will ease me a lot :)
15:20:20 <therve> #topic Abandoning old reviews
15:20:24 <therve> OK somewhat related :)
15:20:31 <therve> We have 200+ patches in review
15:20:46 <therve> Obviously a good chunk will never land at this rate
15:20:59 <therve> Many need rebase, a bunch are just rotting
15:21:29 <therve> I'd like to do a bit of cleanups, anyone has objections about that?
15:21:37 <jdob> nope, i'm all for it
15:21:41 <skraynev> +1
15:21:44 <ananta> there are many with one +2, those are easier to review
15:21:52 <ramishra> What's the definition of old?
15:21:55 <jdob> is there going to be a warning about abandoning them or just abandon if they are too old?
15:21:58 <skraynev> I have one ;) will abandon it ;)
15:22:31 <therve> ramishra, February I'd say?
15:22:38 <jdob> feb of 2016?
15:22:49 <therve> Yes
15:22:53 <shardy> therve: we have a similar problem for bugs, with 364 open atm
15:23:04 <jdob> that feels kinda short, I was thinking at least 6 months
15:23:20 <zaneb> therve: yeah, that sound too short to me
15:23:23 <shardy> I've been trying to tackle that for TripleO by a script that marks all bugs reported against EOL branches incomplete with a comment
15:23:29 <shardy> nova are doing similar ref ML threads
15:23:31 <jdob> ya, bugs are another area we should do this for
15:23:40 <therve> I mean, if it's on page 3 of most gerrit people, it's not like it will get reviewed
15:23:48 <therve> shardy, Can we talk about that separately :)
15:23:51 <shardy> I agree ~3 months is probably too short
15:24:00 <jdob> ya, but i've ha things take me months to get people to look at that i was trying to maintain
15:24:03 <ananta> but 6 months is too old
15:24:16 <jdob> there is a bit of a lag in getting core eyes on stuff
15:24:23 <ananta> we will again be ending in same situation sooner if we keep 6 months
15:24:32 <ananta> jdob: agree
15:24:34 <therve> jdob, Sure, but you rebase them at that point
15:24:38 <skraynev> I agree with definition of "old" as more then 6 month.
15:24:43 <skraynev> therve: is it ok ^ ?
15:24:49 <jdob> i like 6 months and then making this a release task
15:24:51 <shardy> maybe abandoning patches that have negative feedback not addressed for >= 3months is OK
15:24:58 <jdob> each new release, part of that fuzzy period before summit we do this again
15:25:00 <therve> Let's say 5 and I cut 2015 :)
15:25:07 <zaneb> shardy: +1, a lot depends on the status
15:25:08 <shardy> we want to avoid abandoning anything rotting due to lack of reviews tho
15:25:15 <shardy> (I bet there are some..)
15:25:27 <jdob> i'm sure of it :)
15:25:29 <zaneb> some relatively recent patches can be abandoned, and some very old patches should probably be kept around
15:25:37 <ananta> agree with shardy
15:25:54 <ramishra> yeah not being reviewed does not make patches old;)
15:26:28 <jdob> the low hanging fruit are "old" ones with a -1 that haven't been addressed
15:26:30 <therve> ramishra, If it's in merge conflict though?
15:26:35 <jdob> thats a pretty solid indication they aren't being maintained
15:26:42 <jdob> i think you could make the same argument for merge conflict
15:26:51 <therve> merge conflict put a -1 too
15:26:59 <jdob> sorry, human -1 v. jenkins -1
15:27:16 <ananta> IMO, it all boils down to prioritizing the feature/bug
15:27:22 <ramishra> yeah, we should give them a chance to rebase, may be a comment in the review and then do whatever we want.
15:27:38 <zaneb> +1
15:27:55 <ananta> ramishra: agree
15:28:00 <therve> ramishra, You can still revive the patch when it's abandoned?
15:28:06 <therve> That's kinda the point oo
15:28:07 <therve> too
15:28:16 <therve> "Hey we won't look at that until you do something"
15:28:41 <jdob> ya, I think we just delay the inevitable in most cases
15:28:49 <jdob> if it gets abandoned and you want it back, just restore it
15:28:54 <therve> Exactly
15:29:04 <jdob> otherwise, it just makes it a longer two steps process for the heat cores
15:29:41 <pas-ha> also, if someone else wants to work on it (not the original owner) I believe cores can restore any patch
15:29:43 <ramishra> it's quite subjective. IMO, I think the owner should be given a choice to decide.
15:30:00 <therve> ramishra, Not sure what you suggest
15:30:10 <therve> You always have the choice to rebase your patch :)
15:30:17 <zaneb> in some cases though there are valuable fixes that contributors have given up on... we need people to take over those reviews and rebase them, not put them out of sight by abandoning
15:30:52 <zaneb> once reviews are abandoned, they're basically invisible
15:31:05 <ramishra> I'm suggesting, we should not abandon others patches, may be leave a comment to rebase if there is no action, then we can abandon them.
15:31:08 <therve> Well once they're on page 3 of heat reviews as well
15:31:28 <therve> Frankly my point is mostly to revive patches, not kill them
15:31:38 <therve> I don't really care that we have 200+ lying around
15:31:53 <therve> But nobody will look at them as they are
15:32:10 <zaneb> very true
15:32:29 <ramishra> I am + on taking over some importtant patches, I see some of them good ones, started but ignored.
15:33:36 <skraynev> therve: sounds like a work for next contrbutors meet-up session ;) fast review all patches ;)
15:33:44 <skraynev> by whole team
15:34:15 <therve> OK, so it doesn't sound that we have a consensus :)
15:34:57 <zaneb> I think we have a consensus that we need to go through the review backlog, and that some of the patches in that backlog should be abandoned
15:35:08 <shardy> +1
15:35:22 <skraynev> I agree with adding comment like "are you plan to work on it" for patches older then 6 month and if it will not get effect after 1 month  - abandon them
15:35:40 <skraynev> zaneb: how about this ^ ?
15:36:12 <therve> skraynev, Won't commenting on it "update" the review?
15:36:47 <skraynev> therve: give a list, I will try to write a bot  for it :)
15:36:56 <jdob> a comment won't bump it up in the list though, will it
15:36:58 <jdob> ?
15:37:02 <jdob> or is that just how I have my dash setup
15:37:23 <skraynev> doh. it's true, I forgot about it
15:37:33 <therve> I believe it changes the "Updated" field, but maybe I miss something
15:37:35 <skraynev> then it is a bad idea
15:38:13 <therve> OK, so review triage day it is?
15:38:26 <jdob> oh shit, you're right, it does change the updated
15:38:31 <zaneb> I'm looking at gertty, which lists patches in the order they were submitted... y'all would be surprised how few have not been updated since 2015 (I'd guess <50)
15:38:41 <skraynev> therve: unfortunately is will push it on the top
15:38:48 <skraynev> like her https://review.openstack.org/#/q/project:openstack/heat
15:38:56 <skraynev> *here
15:39:06 <zaneb> impressed we still have one 5-digit review number open :D
15:39:08 <therve> zaneb, Yeah I know that's why I wanted to extend the deadline :)
15:40:22 <skraynev> wow. 225 patches in line and 10 of them from 2014 :)
15:41:04 <therve> jdob, That's how I manage to keep my patches on top in other projects :D
15:41:30 <jdob> oh my god, i'm going to abuse the hell out of this knowledge
15:41:36 <therve> Heh heh
15:41:43 <skraynev> therve: I suppose, that right idea to be a good example for reviewers, let's start cleaning fro our own patches
15:41:57 <therve> Anyway if everyone could take some time to look at the end of the list and see what's there, that'd be cool
15:41:59 <skraynev> I abandoned 2 really old patches
15:42:31 <skraynev> p.s. it were my patches ;) of course
15:42:32 <therve> shardy, And yeah we need to see what to do about bugs as well... Though it's a bit easier with milestones
15:42:57 <therve> #topic Open discussion
15:43:01 <skraynev> ramishra: I see ,that you also have some old patch ;)
15:43:11 <KanagarajM> will this help: (project:openstack/heat OR project:openstack/python-heatclient OR project:openstack/heat-templates OR project:openstack/heat-specs) status:open before:2016-01-01
15:43:26 <KanagarajM> before tag always see the time of summit of patch
15:43:48 <KanagarajM> so i feel even if review comment, it won't bother the 'before' tag
15:43:52 <skraynev> therve: another known issue is - bugs targeted on heat-next or heat-future - we need to re-review them probably
15:44:02 <therve> Yeah...
15:46:03 <shardy> There's stuff like https://review.openstack.org/#/c/110714/ which refers to an AFAIK still unfixed adopt bug
15:46:26 <shardy> we probably need to either find someone to pick that up, or deprecate adopt and wontfix the bug
15:46:53 <shardy> (I know we already disabled it by default)
15:47:22 <zaneb> shardy: Drago is looking at the abandon/adopt stuff. but we agreed that abandon is a higher priority
15:48:05 <shardy> zaneb: ack, sure - it's just an example of some of the things we lose track of if we have a mechanical solution to abandon based on age
15:48:15 <zaneb> agree
15:48:17 <shardy> I guess we still have the bug in that case
15:49:05 <zaneb> true, and hopefully the bot has linked the review so it doesn't get completely lost
15:49:06 <Drago> I understood adopt to be deprecated in the future and haven't looked into that at all
15:49:14 <skraynev> zaneb: could you clarify for me: what does mean "looking" ? fix it or deprecating this staff ?
15:50:07 <therve> skraynev, Fixing it hopefully :)
15:50:15 <zaneb> skraynev: I think he just answered your question ;)
15:50:20 <therve> At least abandon *can* work
15:50:22 <ramishra> lol
15:50:31 <zaneb> AIUI abandon is going to get fixed
15:50:39 <zaneb> nobody cares a lot about adopt
15:51:03 <zaneb> and if we ever get external resources then we can abandon abandon too
15:51:10 <therve> It's almost there
15:51:14 <shardy> zaneb: Yeah perhaps we deprecate and remove adopt in favor of the external resource thing
15:52:07 <skraynev> zaneb: yeah I see :)
15:52:36 <skraynev> therve: deprecation - fixes everything, unfortunately it's not the best "fixing"
15:52:46 <therve> :)
15:52:57 <skraynev> shardy: IMO, sounds good
15:52:58 <Drago> IIRC, the only bugs left in the abandon/adopt topic are for adopt
15:53:44 <Drago> zaneb: So abandon may be fixed already :)
15:53:55 <zaneb> great :)
15:54:08 <therve> Alright
15:54:11 <therve> #endmeeting