mozilla :: #mdndev

14 Mar 2017
00:43rjohnsongood night mdndev
15:16shobsonI have two bug triages on my calendar today. What time is the real one?
15:16shobsonfscholz: ^?
15:24jpettorjohnson: i reviewed and it looks good, but i'd love a second eye on the python bits of PR 4151
15:24jpettojust to be safe
15:24jpettonext, we'll need to create the experiment pages on prod after the push
15:24jpettoand i *think* that's it
15:25rjohnsonjpetto: ok, i'll start with the review
15:26jpettoi'd say a good once over is sufficient :)
15:27rjohnsonok, sounds good
15:27fscholzshobson: first one works for me. Does it for everyone else?
15:28shobsonfscholz: works for me
16:00rjohnsonfscholz: sorry i missed the triage meeting, got wrapped up in something else
16:01fscholzrjohnson: it is happening in 30 minutes
16:02* fscholz thinks so at least
16:02* fscholz hates time zones
16:28sheppyYeah, triage in 2 minutes
16:30rjohnsonfscholz: SRE meeting finishing soon, will be in triage meeting after
16:30fscholzrjohnson: ok! :)
18:00rjohnsonjpetto: just finished meetings, back to your content pr
18:18sheppyThe MDN demo server functionality recently added (https://kuma.readthedocs.io/en/latest/development.html#deis-workflow-demo-instances) -- is that something that any contributor can use or is that just for you guys? I have something I'd love others to play with without having to do their own custom build.
18:22shobsonOkay, that is very cool! Does it populate with the sample database?
18:22jpettorjohnson: let me know when you're ready :)
18:23shobsonsheppy: I would assume it is only available to people who have permissions to create branches on the kuma repo.
18:23sheppyshobson: I'd think so too. :)
18:23sheppyJust wishful thinking
18:26rjohnsonshobson, sheppy: it's really nice, but not quite ready for prime time yet (it doesn't render as currently deployed)...metadave is close to having the sample db auto-deployed as well
18:26metadavehey, I just got it working and I'm writing docs and will have a PR within the fortnight
18:26sheppyI need to find a way to let people try out changes I've made without having to branch and build and set up an environment themselves.
18:26metadaveer, well, hopefully a few hours :-)
18:27rjohnsonmetadave: lol
18:27metadaveI'm pretty sure the dynamic k8s demo instances are for gh org members only
18:27* sheppy nods
18:29metadaveallowing folks outside the org to provision Kubernetes namespaces + resources could be ... tricky
18:29sheppyOh yeah
18:30sheppyI'm Moz staff but not sure if I have that access.
18:30rjohnsoni started working on getting an api service into the deployment yesterday (when i realized it was missing and causing my render issues), and made some progress. jgmize and i are going to work on that this afternoon
18:30sheppyDon't think so
18:31metadavesheppy: you don't need Kubernetes CLI access to use https://kuma.readthedocs.io/en/latest/development.html#deis-workflow-demo-instances
18:31metadaveI think the only requirement is that the branch is actually IN the kuma repo
18:31metadavenot a remote
18:31metadaveI could be totally wrong there though (as I am about most things computer related)
18:31sheppyMakes sense.
18:32metadaveif you have some time, give it a shot and if it doesn't work, I can assist
18:32sheppyk
18:32metadaveare you in the mozilla org?
18:32rjohnsonjpetto: i did a limited review of pr 4151 (just focused on the backend python code) and it looks great
18:32metadaveer, do you have push access to kuma?
18:32sheppyI don't believe so.
18:32jpettorjohnson: awesome
18:33rjohnsonjpetto: let me make a formal comment on the pr and then do you want to merge?
18:33sheppymetadave: Only org I'm a member of is mdn.
18:33jpettoaside from creating the pages in prod, do you see any other post-push steps needed?
18:33metadavesheppy: looking... one sec
18:35metadavesheppy: I don't see you as part of https://github.com/orgs/mozilla/teams/mdn
18:35metadaveif you were in that group, you'd have admin access
18:35sheppyhuh
18:36sheppymetadave: Yeah, I'm just in https://github.com/mdn
18:37sheppyI'm the owner of that, in fact.
18:37metadaveoh, my bad, there's an org AND a team
18:37jpettorjohnson: yep, happy to merge after your comment
18:37sheppymetadave: yeah
18:37rjohnsonjpetto: hmm, just noticed, yes maybe so, and it worries me a bit. since we're modifying the django settings, i think we have to do that manually on staging and prod
18:38jpettohmmm
18:38metadavesheppy: I'm not an admin of the MDN team
18:39rjohnsonjpetto: i don't think the staging and prod servers use the common.py settings, so we'd have to add those manually to the appropriate file
18:40jpettorjohnson: it looks like stage.py and prod.py import * from common.py?
18:40metadavesheppy: we'll have to defer to John when he's back/in Toronto
18:43rjohnsonjpetto: i'm almost certain those files aren't used...i'll have to check
18:43jpettorjohnson: we should be able to merge and push without error though, right?
18:44jpettothe experiments just may not work until we get settings figured out?
18:45rjohnsonjpetto: i'll have to check that, but i suspect it could cause errors on all doc requests until the correct settings are installed
18:46jpettoah, if CONTENT_EXPERIMENTS isn't defined
18:46* rjohnson is checking...
18:46rjohnsonexactly
18:54jpettorjohnson: gonna grab some lunch. be back in ~30 min
18:54rjohnsonjpetto: is this a high priority to go out today? it's going to take some time for me to look into this (the settings situation on prod) and i have lunch and then a long working meeting at 1pm with the SRE's
18:55jpetto|lunchi don't think today is critical - Topal was asking when it could go out
18:55jpetto|lunchi'm off tomorrow and thursday to move, but i don't think i'm necessary for the merge/push
18:57rjohnsonjpetto: i feel like it would be better to wait for jwhitlock next week, as i have lots to do before the end of the week as well, but if it's necessary this week i can focus on it tomorrow
19:01jpetto|lunchrjohnson: hm, Topal, what do you think?
19:01jpetto|lunchcould be something we hash out/launch in toronto
19:03* jpetto|lunch lunches for real now
19:06shobsonWe need one of the Johns on hand for the push I think.
19:09rjohnson|lunchi can at least look into the settings situation to see if that's a problem or not, but given that jwhitlock did the backend work on this i'd feel more comfortable if he was part of the launch
19:10Topalrjohnson|lunch: jpetto|lunch When I talked to John on Friday he mentioned that pushing the experiment should be fine without him this week.
19:11TopalIf thats problematic still I wouldnt insist on it. Its just that wed like to start the actual A/B testing in Toronto, and this would give us a few days of data and confidence that the mechanism is valid.
19:14rjohnson|lunchTopal: thanks, good to know that john thought it'd be fine, that makes me feel better...let me check into the settings issue and perhaps we can push tomorrow if everything looks fine
19:17Topalrjohnson|lunch: thanks! dont stress yourself out though, if we cant launch it, we cant launch it.
20:53rjohnsonjgmize: fyi, we need to scale the api service from its zero start, so i'll add that to the pr
22:21rjohnsonmetadave: forgot to say thanks so much for your https://github.com/mozilla/kuma/pull/4153 ! nice!
23:18rjohnsonjgmize: on second thought, do you mind if we keep the "mdn-demo-maint-mode" around for a few more days?
23:35jgmizerjohnson: no problem, I'll just add a cleanup task to the backlog
15 Mar 2017
No messages
   
Last message: 72 days and 19 hours ago