15:00:49 <krtaylor> #startmeeting third-party
15:00:50 <openstack> Meeting started Wed Jun 10 15:00:49 2015 UTC and is due to finish in 60 minutes.  The chair is krtaylor. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:00:51 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
15:00:54 <openstack> The meeting name has been set to 'third_party'
15:01:11 <sweston> o/
15:01:12 <krtaylor> who's here for third party ci working group?
15:01:13 <patrickeast> hi
15:01:16 <mmedvede> o/
15:01:19 <asselin_> hi
15:01:30 <marcusvrn> Hi
15:01:44 <BobBall> o_
15:01:46 <BobBall> That's a mostly.
15:02:02 <ja> 3pci
15:02:23 <krtaylor> hi everyone! looks like we have a good group today
15:02:52 <krtaylor> here's the agenda
15:02:55 <krtaylor> #link https://wiki.openstack.org/wiki/Meetings/ThirdParty#6.2F10.2F15_1500_UTC
15:03:36 <krtaylor> #topic Announcements
15:03:52 <asselin_> BobBall, are you double-booked?
15:03:58 <BobBall> Yup!
15:04:11 <BobBall> Always am... But moving the other meeting (as it happens :) )
15:04:19 * BobBall is also in #openstack-meeting :)
15:04:22 <krtaylor> The cinder volume driver deadline is June 19th, coming up
15:04:52 <krtaylor> re: double booking, I am hoping to fix that, we'll discuss in a bit
15:05:13 <krtaylor> any other quick announcements or reminders?
15:05:22 <asselin_> cinder deadling is actually June 12th for CI to start posting 1 week before June 19th to be merged
15:05:47 <krtaylor> ah, thanks for that clarification asselin_
15:06:52 <krtaylor> if nothing else, on to active work items then
15:07:01 <krtaylor> #topic Monitoring dashboard status
15:07:30 <krtaylor> just to level set, we had a good initial discussion in the Monday office hours meeting
15:07:51 <krtaylor> the spec for the dashboard has stalled
15:08:27 <krtaylor> there are several instances of the scoreboard being seen around
15:08:51 <BobBall> Any examples worth sharing?
15:08:56 <krtaylor> so the question is, either we need to get support behind the existing spec
15:09:03 <BobBall> Do these scoreboards look at all CIs or just their own?
15:09:07 <krtaylor> yes, links coming
15:09:23 <BobBall> Also spec link would be good :)
15:09:30 <krtaylor> spec:
15:09:35 <krtaylor> #link  https://review.openstack.org/#/c/135170/
15:10:12 <krtaylor> I just had the scoreboard link, one sec
15:10:58 <asselin_> patrickeast commented to use the scoreboard he setup
15:11:07 <sweston> this spec has been difficult to move forward ... every time I think we are about to get a +2, somebody posts another review that stalls it.  Reminds me of the charlie brown football.
15:11:12 <sweston> http://ec2-54-67-102-119.us-west-1.compute.amazonaws.com:5000/?project=openstack%2Fcinder&user=&timeframe=24
15:11:17 <krtaylor> asselin_, do you have that link handy?
15:11:27 <sweston> ^ link for scoreboard
15:11:40 <krtaylor> thats it, thanks sweston
15:11:47 <sweston> krtaylor: yup
15:11:55 <BobBall> "This web page is not available"
15:11:59 <BobBall> does it work for others?
15:12:07 <asselin_> works for me
15:12:14 <mmedvede> wfm as well
15:12:27 <BobBall> Stupid corporate firewall!
15:12:27 <asselin_> do we need a spec to do the scoreboard?
15:12:38 <asselin_> patrickeast made one w/o a spec
15:12:54 <asselin_> so I'm questioning the purpose of the spec really....
15:13:00 <patrickeast> sry, i keep dropping off the meeting (on my phone)
15:13:02 <krtaylor> asselin_, good question, but the initial radar was also made w/o spec
15:13:06 <patrickeast> asselin_: i was wondering about that too
15:13:27 <asselin_> I was looking at this unrelated spec: https://review.openstack.org/#/c/188574/
15:13:28 <krtaylor> I can see where sweston would think that was somewhat unfair
15:13:45 <asselin_> and basically, someone set something up, and wrote a spec afterwards to host it by openstack
15:13:47 <krtaylor> since sweston had to go through the painful spec process
15:14:25 <patrickeast> i agree 100%, its kind of backwards
15:14:29 <asselin_> I think we should do something similar for the dashboard
15:14:41 <patrickeast> but on the other hand… we need something
15:14:58 <patrickeast> whether its my scoreboard, or sweston’s dashboard, or radar or whatever
15:15:02 <sweston> yup, this is the circle we've been in for months
15:15:12 <asselin_> get something working, iterate, and once we're happy, copy that spec outline to have it hosted by infra
15:15:21 <patrickeast> and i think i’m not alone in saying that waitin for another round of spec hmm-hawing isn’t going to get us anywhere
15:15:33 <asselin_> patrickeast, +1
15:15:34 <krtaylor> agreed, so the question is, do we stay with the new radar?
15:15:36 <patrickeast> unless our goal is to revisit it again in M
15:16:00 <sweston> radar should be the permanent solution ... the scoreboard is good for adhoc reporting, but does not help the understanding of history and trends
15:16:01 <asselin_> M is for Maybe
15:16:15 <patrickeast> sweston: +1
15:17:14 <patrickeast> one thing im wondering is if maybe we are having this conversation in the wrong venue
15:17:25 <patrickeast> since clearly even if we all agree on something it wont make it happen
15:17:31 <patrickeast> we kind of need the infra cores to back it
15:17:38 <asselin_> patrickeast, no we don't
15:17:50 <patrickeast> aren’t they the +2’s we need on that spec?
15:17:55 <asselin_> patrickeast, you didn't to create the scorecard
15:18:03 <patrickeast> true
15:18:04 <krtaylor> so, I'd propose, abandoning the spec, having infra host scoreboard for a tactical solution, then radar for the strategic way forward
15:18:09 <asselin_> patrickeast, goes back to what I was saying...why do we need the spec at all?
15:18:19 <patrickeast> asselin_: haha yea, fair point
15:18:45 <krtaylor> exactly, stand up one, patch to have it hosted, that would be the spec
15:18:50 <asselin_> so, we do need a spec to have infra host scoreoard
15:19:08 <asselin_> or host radar
15:19:34 <krtaylor> I'd say no, else we would have had an approved spec long ago
15:19:49 <asselin_> and if we want a spec, it should be in third-party-specs where we can +2 it ourselves
15:20:03 <krtaylor> hm, good point
15:20:04 <sweston> krtaylor: I like your plan.  I am reluctant to abandon the spec, as in my opinion I don't think the spec should be related to infra hosting the solution
15:20:37 <sweston> asselin_: yes, that would be a good solution, +1
15:21:14 <krtaylor> sweston, it can remain actually, or we can have a third-party-specs copy
15:21:18 <wznoinsk> sweston: sorry to inject out of nowhere here, would nagios check and bothering alert email be just enough? you can pull report from nagios on how many times the check failed/what status of the check was (like CI name that was suspected) ?
15:21:51 <krtaylor> wznoinsk, lets have that discussion in the open discussion at the end of the meeting
15:22:01 <asselin_> wznoinsk, I think we need something more visual to compare all 3rd party cis
15:22:24 <krtaylor> oh, as an alternative, understand now
15:22:33 <sweston> wznoinsk: no problem ... nagios needs data to run against, the dashboard is for the collection of the data
15:22:47 <krtaylor> yes, we know email will be ignored
15:22:54 <sweston> krtaylor: hehe
15:23:36 <krtaylor> so, do we have an agreement?
15:23:58 <asselin_> krtaylor, can you summarize just to be clear?
15:24:00 <sweston> krtaylor: can you clarify what is on the table now
15:24:10 <krtaylor> I think we all agree on scoreboard immediately, radar long term
15:24:30 <asselin_> +1
15:24:32 <sweston> can we vote? makes me feel important ...
15:24:42 <sweston> ;-)
15:24:44 <krtaylor> hehheh sure
15:24:52 <asselin_> let's see if it works this time
15:25:07 <krtaylor> #startvote
15:25:08 <openstack> Unable to parse vote topic and options.
15:25:13 <krtaylor> lol
15:25:15 <krtaylor> nope
15:25:17 <asselin_> #help
15:25:28 <sweston> ah, well, most fun I've had all week.
15:25:41 <sweston> +1 for scoreboard short term, dashboard long term
15:25:51 <patrickeast> +1
15:25:52 <krtaylor> #startvote scoreboard immediately, radar long term
15:25:52 <openstack> Unable to parse vote topic and options.
15:25:53 <ja> +1
15:26:04 <marcusvrn> +1
15:26:07 <mmedvede> +1
15:26:08 <asselin_> krtaylor, use #help
15:26:14 <asselin_> +1
15:26:18 <krtaylor> #help startvote
15:26:22 <krtaylor> #help
15:26:25 <BobBall> +1
15:26:27 <krtaylor> lol
15:26:30 <asselin_> or not :)
15:26:30 <krtaylor> +1
15:26:43 <BobBall> I don't see any -1s though!
15:26:49 <krtaylor> I can show it this way
15:27:04 <krtaylor> #topic Vote on  scoreboard immediately, radar long term
15:27:57 <krtaylor> #topic Vote results: 7 voting +1, 0 voting -1
15:28:00 <asselin_> Example: #startvote Should bshum be the release manager from now on? Yes, No
15:28:51 <krtaylor> #startvote  scoreboard immediately, radar long term? yes,  no
15:28:52 <openstack> Begin voting on: scoreboard immediately, radar long term? Valid vote options are yes, no.
15:28:54 <openstack> Vote using '#vote OPTION'. Only your last vote counts.
15:28:59 <BobBall> #vote yes
15:29:03 <krtaylor> #vote yes
15:29:04 <asselin_> #vote yes
15:29:05 <mmedvede> #vote yes
15:29:05 <sweston> #vote yes
15:29:15 <marcusvrn> #vote yes
15:29:21 <BobBall> You're right sweston - it does make me feel important.
15:29:28 <krtaylor> yea, we figured out the vote
15:29:34 <sweston> BobBall: right? ;-)
15:29:48 <patrickeast> #vote yest
15:29:49 <openstack> patrickeast: yest is not a valid option. Valid options are yes, no.
15:29:51 <patrickeast> #vote yes
15:29:53 <patrickeast> whew
15:29:56 <krtaylor> hahhahaaaa
15:29:59 <patrickeast> almost missed it and paniced
15:30:26 <krtaylor> anyone else?
15:30:29 <patrickeast> panicked even
15:31:20 <krtaylor> #topic Monitoring dashboard
15:31:29 <sweston> I hate to get stuck on this, but I do not feel that we've solved the second issue of what to do about the current spec
15:31:29 <asselin_> krtaylor, need to #endvote
15:31:31 <krtaylor> last call
15:31:37 <BobBall> Was going to say, shouldn't it be finished ;)
15:31:42 <krtaylor> #endvote
15:31:44 <openstack> Voted on "scoreboard immediately, radar long term?" Results are
15:31:45 <openstack> yes (7): asselin_, krtaylor, sweston, BobBall, marcusvrn, mmedvede, patrickeast
15:32:12 <krtaylor> cool, so the question is: what to do with the spec?
15:33:24 <asselin_> I think we need a new spec to get scoreboard hosted by infra
15:33:26 <krtaylor> do we want a third-party-specs
15:33:50 <krtaylor> do we need a spec for that? I guess infra should answer that
15:34:01 <asselin_> do we need  a new repo? or just add a folder to out exiting?
15:34:13 <asselin_> krtaylor, yes we do
15:34:58 * krtaylor is having trouble parsing this
15:35:10 <krtaylor> folder in thrid-party-ci-tools?
15:35:14 <krtaylor> third
15:35:24 <asselin_> krtaylor, yes
15:35:46 <asselin_> krtaylor, or a new third-party-ci-tools-spec repo?
15:36:03 <krtaylor> that would give us a place for the existing spec, sweston 's
15:36:33 <sweston> I think the other factor is, that this spec should have been closed a long time ago, but there has not been a +2 quick enough from a core, and then additional questions get posted to it.  We had agreement on this spec a month ago
15:37:00 <sweston> so I am inclined to move it to a place where we can approve it
15:37:04 <krtaylor> yes, and we have 2 topics here
15:37:16 <krtaylor> 1) spec 2) hosting
15:37:42 <sweston> yes, may we vote, again on 1, and then we can vote on 2
15:38:00 <krtaylor> how about we just make a dir under the tools repo for now, we can always pull it out into a new repo
15:38:19 <asselin_> krtaylor, +1
15:38:32 <sweston> krtaylor: +1
15:38:53 <krtaylor> so the proposal is: move the existing radar spec to a /spec dir under the third-party-ci-tools repo
15:39:13 <asselin_> #agree  move the existing radar spec to a /spec dir under the third-party-ci-tools repo
15:39:22 <krtaylor> #startvote move the existing radar spec to a /spec dir under the third-party-ci-tools repo? yes, no
15:39:23 <openstack> Begin voting on: move the existing radar spec to a /spec dir under the third-party-ci-tools repo? Valid vote options are yes, no.
15:39:24 <openstack> Vote using '#vote OPTION'. Only your last vote counts.
15:39:32 <asselin_> #vote yes
15:39:36 <sweston> #yes
15:39:39 <krtaylor> now I am abusing the start vote tool  :)
15:39:39 <mmedvede> #vode yes
15:39:41 <sweston> ah ..
15:39:47 <sweston> #vote yes
15:39:47 <mmedvede> #vote yes
15:39:48 <krtaylor> #vote yes
15:39:59 <marcusvrn> #vote yes
15:40:13 <ja> #vote yes
15:40:21 <sweston> krtaylor: feels good, though ... right?
15:40:25 <krtaylor> lol
15:40:42 <krtaylor> last call
15:41:00 <krtaylor> #endvote
15:41:00 <openstack> Voted on "move the existing radar spec to a /spec dir under the third-party-ci-tools repo?" Results are
15:41:01 <openstack> yes (6): asselin_, krtaylor, sweston, marcusvrn, mmedvede, ja
15:41:26 <krtaylor> cool, so, last is, spec for hosting or just patchset
15:42:16 <krtaylor> let's see if a patch will do it, pointing to a running copy of scoreboard, that might be spec enough
15:42:28 <krtaylor> comments?
15:42:41 <asselin_> krtaylor, not sure we have an option....or I'm misunderstanding
15:43:11 <sweston> I am not following either, a patch to what, exactly
15:44:48 <krtaylor> not sure either, the process of hosting vs spec
15:45:00 <krtaylor> do vs. talk about it
15:45:58 <mmedvede> getting it hosted by infra vs hosting ourselves?
15:46:08 <asselin_> krtaylor, we can't just 'do' if we want infra to host it
15:46:10 <krtaylor> yes, thats the question
15:46:20 <asselin_> krtaylor, but if we host it ourselves, no spec needed
15:46:29 <mmedvede> I think it is better for infra to host it
15:46:33 <krtaylor> ok, now I understand your comment
15:46:41 <krtaylor> yes, we want infra to host
15:46:58 <asselin_> if scoreboard is shorterm, we can consider hosting ourselves
15:47:09 <asselin_> patrickeast already has one hosted
15:47:31 <krtaylor> but where is that hosted?
15:47:39 <krtaylor> ourselves is infra :)
15:47:40 <asselin_> amazon lol :)
15:47:44 <krtaylor> hehheh
15:47:56 <sweston> yeah, but infra hosting it has the benefit of standardizing on the url, one supported place for everyone to go
15:48:03 <marcusvrn> and how can we contribute to the patrickeas's scoreboard?
15:48:09 <mmedvede> But patrickeast is paying for it, and it would not be enough once it would get a lot of hits
15:48:26 <patrickeast> haha yea its a very small vm on ec2
15:48:34 <patrickeast> it couldn’t handle much more load than it gets now
15:48:35 <krtaylor> agreed, infra needs to host so that all projects can use it as a reference for CI system reliability
15:48:38 <marcusvrn> I cloned the scoreboard and added more CIs, for example
15:48:39 <sweston> patrickeast should have some ads at the top for revenue generation
15:48:43 <patrickeast> it is technically sponsored by Pure Storage
15:48:46 <sweston> hehe
15:48:47 <patrickeast> i’m using my work account
15:49:09 <asselin_> +1 for ad support :)
15:49:13 <patrickeast> lol
15:49:16 <mmedvede> haha
15:49:22 <marcusvrn> lol
15:49:35 <asselin_> we'd get a lot of hits for sure
15:49:45 <krtaylor> marcusvrn, you can patch the third-party-ci-tools/scoreboard for improvements
15:50:06 <krtaylor> #link https://github.com/stackforge/third-party-ci-tools/tree/master/monitoring/scoreboard
15:50:30 <asselin_> I can help write a spec to get it hosted by infra
15:51:10 <krtaylor> ok, so the proposal is to have infra host the scoreboard for an immediate solution
15:51:20 <krtaylor> another vote  ?  :)
15:51:27 <sweston> whohoo, one more abuse of the voting?
15:51:37 <asselin_> let's vote to vote
15:51:40 <sweston> I say go
15:51:44 <marcusvrn> krtaylor: ok
15:51:52 <krtaylor> any no's?
15:52:01 <krtaylor> (running out of time)
15:52:08 <krtaylor> last call
15:52:37 <krtaylor> ok, asselin_, you have the pen on that spec?
15:52:58 <asselin_> sure
15:53:40 <krtaylor> asselin_, I'll be happy to help, I'm sure others will review as well
15:54:00 <krtaylor> sweston, any other thoughts?
15:54:20 <sweston> krtaylor: nope, you've got them all for the moment
15:54:46 <krtaylor> sweston, again, I want to thank you for all the hard work on the spec and radar code, it was a great effort
15:54:56 <krtaylor> we'll keep this moving
15:55:10 <krtaylor> #topic Meeting frequency and time
15:55:15 <sweston> krtaylor: you're welcome! and thanks for the appreciation, feels good
15:55:17 <marcusvrn> krtaylor: short term means how much time for scoreboard?
15:55:39 <krtaylor> quickly, I want to mention that I have proposed moving this meeting time and day
15:55:41 <asselin_> marcusvrn, until radar is ready
15:55:58 <krtaylor> here is the proposal:
15:56:07 <krtaylor> #link https://review.openstack.org/#/c/190221/
15:56:23 <krtaylor> please review
15:56:33 <krtaylor> oops it failed, gotta fix that
15:56:44 <krtaylor> anyway, I'll also follow up with email
15:57:13 <marcusvrn> krtaylor: agreed! This meeting is at my lunch time :P
15:57:20 <krtaylor> the proposal is to move for less neutron  and cinder conflicts, eliminate poorly attended 0400UTC time
15:57:50 <sweston> jenkins checks ical files now? but my coffee maker is still broken ...
15:58:04 <krtaylor> we are also discussing TPCIWG items in the Monday meetings now, so I see less need for this time every week, move to every other
15:58:37 * patrickeast should start going to monday meetings again
15:58:55 <mmedvede> me too
15:59:14 <krtaylor> we can continue to have deeper dives on topics here
15:59:19 <krtaylor> almost out of time
15:59:29 <krtaylor> thanks everyone, really good meeting today
15:59:41 <sweston> yes, thanks all
16:00:08 <krtaylor> #endmeeting