15:01:19 <anteaya> #startmeeting third-party
15:01:20 <openstack> Meeting started Mon Jun 29 15:01:19 2015 UTC and is due to finish in 60 minutes.  The chair is anteaya. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:01:22 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
15:01:25 <openstack> The meeting name has been set to 'third_party'
15:01:41 <anteaya> anyone here for the third party meeting?
15:01:50 <patrickeast> o/
15:01:52 <lennyb> Hi
15:02:00 <krtaylor> o/
15:02:25 <ctlaugh_> o/
15:02:32 <eantyshev> o/
15:02:49 <anteaya> hello
15:03:03 <anteaya> thanks to asselin for charing last weeks meeting
15:03:55 <anteaya> reminder that asselin's common-ci sprint is coming up July 8 & 9
15:03:59 <anteaya> #link https://wiki.openstack.org/wiki/VirtualSprints#OpenStack_Common-CI_Solution
15:04:23 <anteaya> does anyone have anything they wish to discuss at today's third party meeting?
15:04:50 <taturn> 0o0
15:05:06 <anteaya> taturn: you are in a meeting channel
15:05:29 <anteaya> taturn: if you wish to participate in the third party meeting you are welcome, but please contribute something of use
15:05:42 <anteaya> taturn: the general discussion channel is #openstack-dev
15:06:12 <taturn> sorry i am new here
15:06:33 <taturn> let me use translater to know what u said  i am a chinese
15:06:36 <taturn> lol
15:06:36 <anteaya> taturn: yeah I saw that in the rally meeting
15:06:47 <anteaya> taturn: I gave you some guidance
15:07:07 <anteaya> so does anyone have anything pertient to share at today's meeting
15:08:08 <anteaya> so if noone has anything of any importance to share at today's third party meeting
15:08:21 <lennyb> not me
15:08:23 <anteaya> how about we find more productive uses of our time?
15:08:30 <anteaya> hi lennyb, thanks
15:08:41 <eantyshev> our Zuul occasionally falls into busyloops of a kind: http://paste.openstack.org/show/325044/
15:09:03 <eantyshev> which happens after reconfiguration
15:09:24 <anteaya> does it ever break out of the loop?
15:09:41 <eantyshev> anteaya: Not until zuul-server restarted
15:10:03 <anteaya> so after you reconfigure zuul you need to restart it?
15:10:29 <lennyb> eantyshev: are you using the latest upstream?
15:10:30 <eantyshev> sometimes, yes
15:10:38 <patrickeast> i saw that happen to mine on friday night, filled up like 70GB of log file with that error :(
15:10:51 <anteaya> patrickeast: was a zuul restart the solution?
15:10:53 <patrickeast> i did the same thing, killed zuul, pulled all the latest changes and restarted
15:11:09 <patrickeast> unsure if there was an update that fixed it, but it went away
15:11:11 <eantyshev> lennyb: sure
15:11:31 <anteaya> eantyshev: when did you first notice this behaviour?
15:12:23 <lennyb> yantyshev: may it be related to the late patches that were added. there was some fix related to the loop ( I even think it was yours )
15:13:05 <eantyshev> anteaya: that happened last Thursday at 3pm
15:13:26 <anteaya> eantyshev: so this is recent behaviour then
15:13:43 <anteaya> eantyshev: next time you restart zuul can you be sure to pull the most recent patches
15:14:04 <anteaya> eantyshev: and if the looping behaviour persists are you willing to file a bug against zuul?
15:14:34 <anteaya> #link http://paste.openstack.org/show/325044/
15:14:35 <eantyshev> anteaya: it dissapears every time after restart
15:15:04 <anteaya> eantyshev: right, I'm asking if you are willing to pull the latest patches and evaluate if a patchset fixes the situation
15:15:20 <anteaya> eantyshev: and if it doesn't, if you are willing to file a bug against zuul
15:16:34 <anteaya> eantyshev: are you willing to file a bug against zuul if you keep seeing this?
15:16:36 <eantyshev> anteaya: Zuul is up-to-date, AFAIU, head last updated on Jun12 (Support external cross-project dependencies in ui).  And I'd rather investigate it myself, though
15:16:47 <anteaya> eantyshev: oh okay
15:17:01 <anteaya> eantyshev: can you let us know if you find anything?
15:17:16 <patrickeast> why not file a bug and assign it to yourself?
15:17:16 <anteaya> a bug report allows others to track progress
15:17:23 <anteaya> patrickeast: good point
15:17:38 <anteaya> and you can update teh bug status as you work on it
15:17:57 <eantyshev> then I'll make this a bug
15:18:02 <anteaya> it is a workflow which is considered friendlier in an open source spac
15:18:06 <anteaya> thanks eantyshev
15:18:36 <anteaya> #link https://storyboard.openstack.org/#!/project/679
15:18:48 <anteaya> is the storyboard link for zuul
15:18:56 <anteaya> anything more on this topic?
15:19:29 <eantyshev> patrickeast: did you experienced busyloops too? can you share the piece of your debug logs?
15:20:35 <patrickeast> eantyshev: yea i’ll go dig up the links
15:21:22 <anteaya> does anyone have any other topic to discuss?
15:21:49 <patrickeast> eantyshev: http://paste.openstack.org/show/321511/
15:22:01 <patrickeast> same callstack, different patchset
15:22:09 <ociuhandu> hi all, sorry for being late
15:22:14 <anteaya> #link http://paste.openstack.org/show/321511/
15:22:20 <anteaya> ociuhandu: np
15:22:30 <ociuhandu> anteaya: I’ve seen a quite long thread discussing about third-party CI voting rights
15:22:31 <anteaya> ociuhandu: do you have anything you wish to discuss today?
15:22:53 <anteaya> ociuhandu: yes the discussion is taking place on the mailing list
15:23:04 <anteaya> and voting isn't a right, it is at best a permission
15:23:04 <eantyshev> patrickeast: thanks!
15:23:18 <ociuhandu> and it seems that the topic heads towards non-voting CIs
15:23:33 <anteaya> the decision to allow any ci voting permissions lies with the ptl of the project in question
15:23:53 <anteaya> individual projects can make the decision that works best for their project
15:24:03 <anteaya> I don't expect openstack wide consensous on this
15:24:24 <anteaya> the ability to grant permissions on cis lies with the ptl of the project in question
15:25:47 <ociuhandu> anteaya: from what I have seen it seems that though this started from one project it’s now involving multiple projects and they seem to lean towards the same direction
15:26:17 <anteaya> #link http://lists.openstack.org/pipermail/openstack-dev/2015-June/067992.html
15:26:24 <anteaya> ociuhandu: okay
15:26:28 <ociuhandu> but i see your point, so we’ll have to discuss them with the individual project ptls
15:26:37 <anteaya> did you have any point you wanted to make about that?
15:26:45 <anteaya> yes
15:26:53 <anteaya> I have no say in what the decision is
15:27:34 <anteaya> I encourage you to participate in the thread if you have a perspective you wish to ensure others know about
15:27:48 <ociuhandu> well, as a general idea i feel that telling now any third-party that we’ll revoke any kind of voting rights kind of defeats the purpose of the CI requirements in the first place
15:28:10 <anteaya> you are welcome to have that opinion
15:28:14 <ociuhandu> as in, you have to have a reliable CI but we don’t let you say if a patch is good or not
15:28:17 <krtaylor> ociuhandu, I disagree, comments are useful, if they are reliable
15:28:28 <anteaya> and I do agree that it is frustrating from the perspective of teh ci operators
15:28:35 <anteaya> as the message seems to keep changing
15:28:36 <ociuhandu> (since anyway everyone is complaining about the noise the CIs add to the commenting section)
15:29:00 <anteaya> however the infra message is clear, if you have a gerrit ci account, you are expected to adheare to infra requirements
15:29:11 <anteaya> all else is up to the individual projects
15:29:28 <wznoinsk> hi all
15:29:32 <ociuhandu> ok, thanks
15:29:33 <anteaya> wznoinsk: hello
15:29:43 <anteaya> ociuhandu: welcome, thanks for bringing it up
15:30:06 <anteaya> ociuhandu: I encourage you to comment on the mailing list thread, and discuss it with the project you are testing
15:30:42 <anteaya> anything more on this topic?
15:31:10 <wznoinsk> ociuhandu: I think it came down with recent changes to not to ahve integrated releases
15:31:11 <ociuhandu> anteaya: will do, just wanted to get a deeper understanding on this from the infra perspective
15:31:45 <anteaya> ociuhandu: sure, the infra perspective is about gerrit accounts, voting permissions and many other decisions are up to the projects
15:32:47 <anteaya> does anyone have anything else they wish to discuss today?
15:32:53 <wznoinsk> in short, cores will be looking at CIs results less by the looks of things
15:33:17 <anteaya> wznoinsk: I don't understand either of your last two statements
15:33:27 <anteaya> wznoinsk: can you try again to express your point?
15:35:18 <anteaya> or shall we move on?
15:35:33 * anteaya is uncertain if waiting is a worthy use of time
15:35:55 <anteaya> I'm not trying to dismiss what you said wznoinsk, I'm trying to understand it
15:36:08 <wznoinsk> anteaya: sure, when there was any enforcing of the features in releases, cores/ptls were interested in the tests CI provide as a wider testing ground for features, now when we report rather than enforce and a lot of neutron code went out of tree, cores/ptl are not directly looking at whether they work or not
15:36:26 <anteaya> oh
15:36:47 <anteaya> well to be honest it never was the job of the core reviewers to review the ci jobs
15:36:58 <wznoinsk> to me CIs are becoming the help to features devs only
15:37:07 <anteaya> it was and is the responbility of the ci/driver maintainer
15:37:20 <anteaya> well becoming/is yes
15:37:30 <wznoinsk> anteaya: not saying it was, but as for features that were promised in a release it was  a good testing ground
15:37:33 <anteaya> it just is taking some folks a while to realize that
15:37:37 <anteaya> no it was
15:37:41 <anteaya> I am saying it was and is
15:38:32 <anteaya> the ci results are meant for the consumption of the driver maintainer
15:38:57 <krtaylor> except in the case of infra ci
15:39:01 <anteaya> and others just look to see if the driver maintainer is paying attention or not
15:39:01 <wznoinsk> the help to devs yes, I never said cores job was to look at the CI results, it was more of an additional info
15:39:16 <anteaya> krtaylor: no, especially in the case of infra ci
15:39:29 <anteaya> wznoinsk: right
15:40:01 <anteaya> the addtional info being wether the driver maintainer was paying attention or not
15:40:22 <anteaya> anything more here?
15:40:54 <wznoinsk> I sort of don't like the fact the more loose rights we have now (big tent, 4 OPENs etc) the less we pay attention to proper testing of all of that... I'm not saying cores/ptls should be looking after it tho
15:41:11 <anteaya> wznoinsk: I don't like it either
15:41:25 <anteaya> wznoinsk: I am out of suggestions for what to do about it though
15:41:41 <anteaya> wznoinsk: if you have any contructive thoughts I am looking for suggestions
15:41:43 <wznoinsk> ideally some sort of certified CI badge would be good to have
15:42:21 <anteaya> I'm not certifying anything
15:42:33 <wznoinsk> some CI ops are interest-less, some are looking after well, but may stop doing so if there's no real value it in anymore - that would hurt openstack testing
15:43:03 <anteaya> I have opinions on the use of the word certified
15:43:16 <anteaya> here is a thread from last year on the topic in case you missed it
15:43:20 <anteaya> #link http://lists.openstack.org/pipermail/openstack-dev/2014-June/036933.html
15:43:34 <anteaya> wznoinsk: I agree
15:43:40 <anteaya> values are important
15:43:58 <ttx> not necessarily "certified" but at least recognized as good. Some tag that says "CI actually means something here" could be interesting
15:44:04 <ttx> difficult to define though
15:44:11 <anteaya> and experiencing a lack of values is hard to give or recieve guidance
15:44:12 <wznoinsk> anteaya: well, maybe "powered by Openstack CI suite" could be for CIs using downstream-puppet when you have it done with proper self-diagnosis
15:44:19 <wznoinsk> anteaya: only throwing thoughts here...
15:44:28 <anteaya> wznoinsk: sure
15:44:33 <anteaya> brainstorming is good
15:44:51 <anteaya> I have nothing to offer, unfortunately, though I wish I did
15:45:00 <anteaya> but I definitely feel a lack of values
15:45:08 <anteaya> there used to be values
15:45:15 <anteaya> and I can't find them today
15:45:22 <anteaya> not sure what to do about that
15:45:28 <wznoinsk> ttx: it is difficult, once we agree on who values from CIs (and these are devs and components vendors imho only atm) we would make the certification/recognition right
15:46:03 * anteaya reiterates she has opinions on the use of the word certified
15:46:40 <anteaya> but identifying what we value
15:46:45 <anteaya> or what is valued
15:46:57 <anteaya> and recognizing when that value is respecting
15:47:01 <anteaya> I support that
15:47:07 <anteaya> as well as all those words
15:47:17 <anteaya> values, recogniztion and respect
15:47:53 <wznoinsk> on the other hand, if infra plans to extend their testing, why not to do it using community (3rd party ci)?
15:48:14 <anteaya> if infra plans to extend their testing
15:48:22 <anteaya> does infra plan to extend their testing?
15:48:27 <anteaya> if so this is news to me
15:48:43 <anteaya> what plans does infra have
15:48:51 <anteaya> and what tests are being extended
15:48:53 <anteaya> and how?
15:49:48 <anteaya> wznoinsk: what are you reading or referencing that leads you to these conclusions?
15:49:52 <wznoinsk> there were couple of talks on the summit about testing, some ppl wanted to have more testing, I don't know whether that's in projects plan (to extend by additional distro testing ie.) which would be eventually thrown at infra
15:50:10 <anteaya> do the summit sessions have names?
15:50:16 <anteaya> distro testing?
15:50:20 <anteaya> that is a new one for me
15:50:24 <asselin_> o/
15:50:53 <anteaya> we are discussing packaging repos, that were hoped to be shared by different distros and are looking less likely now
15:50:57 <anteaya> hi asselin_
15:51:13 <anteaya> we have talks every summit about testing
15:51:29 <anteaya> since we have a whole project dedicated to the topic, the qa project
15:51:52 <anteaya> I'm light on details on how this is affecting infra
15:52:13 <anteaya> I'm not saying its not, I'm saying I don't have the same perspective you have
15:52:24 <anteaya> and need some detail or references to understand
15:52:32 <anteaya> and right now I have none
15:52:34 <wznoinsk> anteaya: ok, my 'if infra' was 'in case infra' rather than 'because infra wants to increase testing'
15:52:54 <anteaya> well infra is consuming the common-ci solution
15:53:03 <anteaya> as fast as new modules are merged
15:53:16 <anteaya> so in terms of infra should use the common-ci solution
15:53:18 <anteaya> we are
15:53:27 <asselin_> yes! :)
15:53:32 <anteaya> as soon as there is anything to consume
15:53:41 <anteaya> asselin_: thanks for all your hard work!
15:54:10 <anteaya> so we seem to be in a close to wrap up place
15:54:21 <anteaya> I'll just remind folks that when discussing topics
15:54:40 <anteaya> it is really helpful to not assume anyone else knows the details of what you are talking about
15:54:48 <anteaya> so bringing links with you is really helpful
15:55:03 <anteaya> etherpads, mailing list threads, summit session topics, anything
15:55:27 <anteaya> please bring the link so that others, who may be interested in understanding what you are trying to say, can follow along
15:55:39 <anteaya> anything more for today before we wrap up?
15:56:18 <asselin_> re common-ci solution. I'll request zuul patches to merge today
15:56:25 <anteaya> asselin_: wonderful
15:56:38 <asselin_> so next, requesting reviews on the nodepool patches
15:56:41 <anteaya> have you links so taht those present can review quickly if they wish
15:57:37 <asselin_> looking for link now
15:58:00 <anteaya> thanks
15:58:12 <asselin_> #link https://review.openstack.org/#/c/188325/
15:58:35 <asselin_> didn't find the others....but ping we later if you're interested....
15:58:39 <anteaya> sure
15:58:42 <anteaya> thank you asselin_
15:58:56 <anteaya> thanks all for your attendance and participation today
15:59:00 <anteaya> see you next week
15:59:04 <anteaya> #endmeeting