15:00:30 <mattmceuen> #startmeeting openstack-helm
15:00:31 <openstack> Meeting started Tue Dec 18 15:00:30 2018 UTC and is due to finish in 60 minutes.  The chair is mattmceuen. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:00:32 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
15:00:34 <openstack> The meeting name has been set to 'openstack_helm'
15:00:34 <jayahn> o/
15:00:35 <mattmceuen> #topic Rollcall
15:00:41 <mattmceuen> GM folks and happy holidays
15:00:55 <mattmceuen> agenda du jour is: https://etherpad.openstack.org/p/openstack-helm-meeting-2018-12-18
15:01:08 <srwilkers> o/
15:01:35 <jayahn> My team survived a big organizational change! ;)
15:02:08 <mattmceuen> oh my!  any impacts to your team?  or same old but different leadership? :)
15:02:32 <jayahn> hm.. not much on our side, just one task added.
15:02:35 <jayahn> public cloud
15:02:37 <mattmceuen> tis the season for reorgs
15:02:51 <mattmceuen> just one little small task huh :)
15:02:59 <mattmceuen> you can do that in your spare time
15:03:02 <mattmceuen> good luck jay
15:03:05 <mattmceuen> that's awesome
15:03:10 <jayahn> yeap, just one little small task.. hybrid, multi cloud..
15:03:18 <mattmceuen> #topic Docs
15:03:34 <evrardjp> o/
15:03:35 <mattmceuen> jayahn do you wanna take this one?
15:03:40 <mattmceuen> gm evrardjp
15:04:12 <jayahn> doc repo is set, jaesang just uploaded the first ps. we reorganized the current docs into intro, install, dev, operation, use cases
15:04:15 <jayahn> http://logs.openstack.org/53/625853/5/check/openstack-tox-docs/89127cb/html/
15:04:28 <jayahn> https://review.openstack.org/#/c/625853/
15:05:04 <jayahn> there has been some uncertainty on how we sync up between english docs and korean docs.
15:05:31 <mattmceuen> oh awesome
15:05:41 <jayahn> however, our approach would be uploading korean contents (not trying to sync perfectly with english one at the beginning)
15:05:43 <evrardjp> There are relatively standard folder names to use, let me help you on that in the review
15:06:02 <evrardjp> so that this way the relatively standard folders can be mapped on openstack docs page
15:06:04 <jayahn> then, discuss later on how we are going to sync between two official docs (eng and kor)
15:06:19 <jayahn> evrardjp: thanks!
15:06:39 <mattmceuen> sounds like a good plan to me.
15:06:57 <jayahn> one thing though. docs core reviewers
15:07:13 <georgk> o/
15:07:17 <mattmceuen> Sure, what changes do we need there?
15:07:25 <mattmceuen> Have we not set up the docs cores yet?
15:07:32 <jayahn> we needs to be (I mean jaesang and myself) docs core reviewer to add korean contents. I am not sure what is guideline for that.
15:07:32 <mattmceuen> o/ georgk
15:07:50 <evrardjp> mattmceuen: https://review.openstack.org/#/admin/groups/1962,members
15:08:15 <mattmceuen> ty evrardjp, I always have to fish around for that link
15:08:16 <evrardjp> as member of osh-core you should be able to add those people in.
15:08:43 <evrardjp> It was discussed a series of weeks ago, with no opposition of adding more people into docs core, so I don't see a problem of doing so
15:08:51 <mattmceuen> Agree
15:08:54 <mattmceuen> I think that's a good idea
15:09:01 <evrardjp> it would allow more flexibility/fast iterations
15:09:17 <jayahn> okay, then pls someone add jaesang and myself into doc core?
15:09:19 <mattmceuen> I'll confirm w/ portdirect and get 'er done
15:09:27 <mattmceuen> will ping you jayahn
15:09:29 <jayahn> sure thing
15:09:37 <evrardjp> jayahn: is there a risk of quality change if those cores are in the same org? I guess not, but I am asking for safety
15:09:54 <mattmceuen> That's awesome, thanks to jaesang for the work on this so far
15:10:10 <jayahn> same org as osh core?
15:10:23 <evrardjp> I am asking that because there was discussion of not letting enough time to people to get to see results before things merged in some other projects
15:10:32 <evrardjp> SKT
15:10:37 <srwilkers> evrardjp: i think that's why we'll need to make sure we have a path forward toward for syncing the contents of the repository there
15:11:22 <evrardjp> don't get me wrong, this is a positive increment to diversity, and we should have it :)
15:11:25 <jayahn> evrardjp: i got what you are asking, not really for now
15:11:58 <evrardjp> great -- it's sometimes hard to say no when management is lurking :)
15:12:02 <mattmceuen> agree - as we get more korean language contributors from other orgs we can diversivy further in the future
15:12:08 * srwilkers smirks
15:12:12 <mattmceuen> for now this is a great step
15:12:18 <evrardjp> mattmceuen: agreed
15:12:50 <mattmceuen> aside from us adding the new docs PS to our holiday reading list - any other thoughts on this topic?
15:12:51 <evrardjp> mattmceuen: can you propose officially the appropriate people on the ML?
15:12:58 <jayahn> pls look at doc ps, that restructuring was purly based on our idea, happy to have others as well.
15:13:07 <evrardjp> so that it's done and public
15:13:07 <mattmceuen> yup will do evrardjp
15:13:10 <evrardjp> thanks
15:13:18 <jayahn> agreed with evrardjp
15:13:48 <evrardjp> jayahn: I think comments on this patchsets need to happen in gerrit :) Will dot!
15:13:50 <evrardjp> do*
15:14:07 <mattmceuen> #topic Meeting next week
15:14:09 <evrardjp> and thanks for the work
15:14:13 <jayahn> one thing I added was a section called "Use Cases" hoping we can add valuable experience to leverage osh.
15:14:30 <evrardjp> mattmceuen: cancelled?
15:14:32 <mattmceuen> Next Tuesday's meeting falls on the christmas holiday, and I would expect a very small crowd.  Want to cancel next week?
15:14:38 <jayahn> 25th
15:14:50 <mattmceuen> I think yes cancel :)
15:14:51 <evrardjp> I'd totally cancel that one and the next one. I won't attend either.
15:15:00 <mattmceuen> Sounds good to me
15:15:11 <jayahn> sounds perfectly good to me
15:15:14 <evrardjp> nothing will happen much, and we can still discuss in the chan
15:15:17 <mattmceuen> done
15:15:19 <jayahn> christmas and new year day
15:15:21 <mattmceuen> yup
15:15:34 <mattmceuen> we will reconvene Jan 8
15:15:41 <evrardjp> perfect
15:15:49 <mattmceuen> I'll send out a note on the ML
15:16:00 <srwilkers> sounds good to me.  i can tell you now that i'll not be the slightest bit functional those days
15:16:05 <mattmceuen> :D
15:16:06 <srwilkers> christmas eve is always a party
15:16:13 <mattmceuen> full of good cheer
15:16:18 <mattmceuen> s/good cheer/bourbon/
15:16:22 <srwilkers> cheers and beers
15:16:33 <mattmceuen> #topic Roundtable
15:16:46 <mattmceuen> What else is on y'alls minds today
15:16:47 <jayahn> roundtable already? wait, I have not finished my beer yet..
15:16:56 <mattmceuen> We are passing beers around the table
15:17:10 <mattmceuen> virtual beer as a service, you see
15:17:17 <portdirect> mattmceuen: I can handle the ml this week
15:17:27 <portdirect> Not that far from a keyboard ;)
15:17:34 <jayahn> next summit and ptg timing is aweful for us. fyi.
15:17:41 <evrardjp> portdirect: that's how you get rsi :p
15:17:43 <srwilkers> portdirect: dont make me smack your hands
15:18:29 <evrardjp> on the roundtable bit, I have to change the periodics that simply don't trigger for now
15:18:30 <mattmceuen> sorry jayahn, I hope you guys are able to participate
15:19:13 <evrardjp> I will have to do a few iterations, so please bear with me
15:19:37 <evrardjp> I will ask a few reviews to merge, so that I can results on the next day, and so on, until it really works
15:19:38 <mattmceuen> sure thing evrardjp, let me know what assistance is needed as you progress
15:19:46 <mattmceuen> +1
15:19:54 <srwilkers> oh, speaking of periodics.  the multinode ubuntu periodic job is running green again.  i'm working on resolving the issues with the ironic periodic job next
15:20:01 <srwilkers> as it's been failing the past weekish
15:20:03 <evrardjp> timing is essential, as the jobs get scheduled only once a day, if you see what i mean
15:20:23 <evrardjp> srwilkers: I see there are jobs that are consistently failing on periodics
15:20:44 <srwilkers> evrardjp: yeah, there's a few.  i'm hoping to have those all resolved before the end of the year
15:21:07 <evrardjp> srwilkers: there is also an experimental job that is now running for 89h.
15:21:15 <srwilkers> yeah, i saw that
15:21:22 <evrardjp> we should probably work that out with infra
15:21:34 <evrardjp> obviously something went horribly wrong
15:21:39 <mattmceuen> oh boy
15:21:55 <evrardjp> (I meant queued for 89h)
15:22:04 <srwilkers> yeah -- the fedora checks been queued up forever
15:22:05 <mattmceuen> oh whew :)  that is slightly better
15:22:18 <evrardjp> sorry bad english
15:23:10 <mattmceuen> only one of us has said "y'all" in the meeting today evrardjp, you are winning the english contest
15:24:20 <evrardjp> mattmceuen: I am picturing you with a texas accent now
15:24:34 <srwilkers> thats nightmare fuel
15:24:47 <evrardjp> haha
15:24:48 <mattmceuen> and that, my friends, is the definition of rountable :)
15:25:11 <mattmceuen> I will be practicing the accent for the rest of the day
15:25:18 <mattmceuen> On that note - anything else to discuss?
15:25:25 <georgk> I do have one thing
15:25:31 <mattmceuen> hey georgk go for it!
15:25:33 <georgk> a request for support / pointers to help
15:25:55 <georgk> I have a problem with the linting that helm is doing on the helm-toolkit templates
15:26:11 <mattmceuen> what error are you seeing?
15:26:16 <georgk> it doesnt like things like ¨hugepages-1Gi¨ as a resource limit
15:26:25 <georgk> specifically, it does not like the -
15:27:03 <evrardjp> srwilkers: did we discuss changing the linting? You had a patchset for it, right?
15:27:55 <srwilkers> evrardjp: only for enforcing orders of the values.yaml files for the charts.  what chart are you seeing that issue with georgk, and what overrides specifically are you supplying?
15:27:57 <evrardjp> (not that it matters for this, but the question is asked while I think of it)
15:28:36 <mattmceuen> would definitely appreciate a PS that improves the linting georgk if you feel comfortable with that corner of the code
15:28:53 <georgk> for the dpdk support, I need to add resource limits for the hugepages. I wanted to add this to the kubernetes-resources.tpl
15:29:34 <georgk> mattmceuen: not really comfortable yet, but I wanted to look into that
15:30:40 <georgk> BTW, the error is : [ERROR] templates/: parse error in "helm-toolkit/templates/snippets/_kubernetes_resources.tpl": template: helm-toolkit/templates/snippets/_kubernetes_resources.tpl:57: bad character U+002D '-'
15:30:44 <evrardjp> I thought writing a proper linting tool for helm would be a good community improvement
15:31:55 <srwilkers> georgk: we'll need to modify the helm-toolkit snippet used there, as it only currently supports setting requests and limits for memory and cpu.  thanks for bringing this up
15:32:31 <georgk> srwilkers: right, that´s what I wanted to add, but than I ran into this issue
15:33:03 <georgk> just curious if you had a solution from the top of your head
15:33:44 <srwilkers> let me dig around -- i vaguely remember seeing issues with hyphens in the yaml keys before, but cant recall the workaround we used there
15:33:50 <srwilkers> once i find something, i'll ping you in irc
15:34:04 <georgk> srwilkers: awesome, thank you
15:34:53 <mattmceuen> ty srwilkers
15:35:17 <mattmceuen> there are workarounds that might translate between - and _ iirc
15:35:47 <mattmceuen> any other topics peeps?
15:36:03 <srwilkers> mattmceuen: yeah, i vaguely remember some replacement foo happening
15:36:43 <mattmceuen> agree w/ georgk that this is something we should address one way or the other though
15:37:15 <srwilkers> georgk: you'd likely need to include some checking in the helm-toolkit template to perform something like this on the key: https://github.com/openstack/openstack-helm-infra/blob/master/helm-toolkit/templates/utils/_daemonset_overrides.tpl#L210
15:37:58 <georgk> srwilkers: eww, nice
15:37:59 <srwilkers> ie: if defining limits/requests for hugepages_foo, do a replace to make it hugepages-foo for the key
15:38:11 <mattmceuen> alright - have some :D
15:38:20 <georgk> srwilkers: ok, I´ll play with this a bit
15:38:27 <mattmceuen> oops - got halfway through a message before starting a new message
15:38:44 <mattmceuen> disclosure:  just started hosting an audio bridge, I'm filling in for multiple folks simultaneously
15:40:35 <srwilkers> any other items to discuss? or should we wrap it up?
15:41:14 <mattmceuen> I think we can wrap
15:41:19 <mattmceuen> thanks everyone for an awesome year
15:41:25 <jayahn> thanks.
15:41:25 <mattmceuen> looking forward to even better things next year
15:41:27 <srwilkers> thanks everyone :)
15:41:33 <mattmceuen> have a good one!
15:41:41 <mattmceuen> #endmeeting