#metabrainz

/

      • daksh has quit
      • 2017-12-11 34556, 2017

      • zas
        AND fixing remaining bits of our apps which aren't docker-friendly
      • 2017-12-11 34522, 2017

      • zas
        this is something we can plan, as well the migration to full cloud
      • 2017-12-11 34513, 2017

      • zas
        it is an effort of the whole team, a bit like the one we did to move from dwni to hetzner (with the dockerization of our apps), but it is definitively the way to go.
      • 2017-12-11 34544, 2017

      • daksh joined the channel
      • 2017-12-11 34550, 2017

      • zas
        but being familiar with docker is a pre-requisite, so, imho, it is too soon. My idea is to setup services we currently have on gcloud vms (grafana/influx/nbagios/wiki/etc..) on kubernetes at gcloud, we need to study pricing etc... though
      • 2017-12-11 34521, 2017

      • zas
        it could be a starting point, before trying to migrate critical stuff
      • 2017-12-11 34554, 2017

      • ruaok
        great summary, thanks.
      • 2017-12-11 34517, 2017

      • ruaok
        in the new year, lets make a list of things that need to be done in order to make our containers docker friendly.
      • 2017-12-11 34556, 2017

      • ruaok
        then, once we get there, we should as a team all work to nuke consul and to go to pod.
      • 2017-12-11 34506, 2017

      • ruaok
        then we can do a gcloud test deploy
      • 2017-12-11 34511, 2017

      • ruaok
        how does that sound like a plan?
      • 2017-12-11 34535, 2017

      • ruaok
        iliekcomputers: as I suspected: > fatal: Not a git repository (or any of the parent directories): .git
      • 2017-12-11 34552, 2017

      • ruaok
        when the meb container starts. the commit hash needs to be captured at container build time.
      • 2017-12-11 34530, 2017

      • samj1912
        we generally do it by patching a __version__ variable in the __init__ or setup file
      • 2017-12-11 34511, 2017

      • ruaok
        got link, samj1912 ?
      • 2017-12-11 34519, 2017

      • samj1912
      • 2017-12-11 34531, 2017

      • samj1912
      • 2017-12-11 34554, 2017

      • samj1912
        is this what you are looking for?
      • 2017-12-11 34515, 2017

      • ruaok
        yeah, but for a webapp, not a module.
      • 2017-12-11 34502, 2017

      • samj1912
        there shouldn't be much of a difference if its in python?
      • 2017-12-11 34528, 2017

      • ruaok
        not much no, but this seems overly complicated. it can be done a lot simpler, methinks.
      • 2017-12-11 34520, 2017

      • samj1912
        yeah, should be
      • 2017-12-11 34527, 2017

      • ruaok
        but, thanks for the links.
      • 2017-12-11 34534, 2017

      • ZarkBit has quit
      • 2017-12-11 34516, 2017

      • ZarkBit joined the channel
      • 2017-12-11 34558, 2017

      • culinko joined the channel
      • 2017-12-11 34559, 2017

      • culinko has quit
      • 2017-12-11 34559, 2017

      • culinko joined the channel
      • 2017-12-11 34517, 2017

      • Slurpee joined the channel
      • 2017-12-11 34517, 2017

      • Slurpee has quit
      • 2017-12-11 34517, 2017

      • Slurpee joined the channel
      • 2017-12-11 34543, 2017

      • ruaok
        zas: see email for the example startup-log for the meb container now.
      • 2017-12-11 34552, 2017

      • ruaok
        I think we should standardize on something like that.
      • 2017-12-11 34528, 2017

      • ruaok
      • 2017-12-11 34542, 2017

      • ruaok
        also, samj1912 and iliekcomputers
      • 2017-12-11 34507, 2017

      • samj1912
        looks good :)
      • 2017-12-11 34553, 2017

      • ruaok
        lets see what zas says.
      • 2017-12-11 34513, 2017

      • ruaok
        but this level of startup logging should prevent the troubles I've had.
      • 2017-12-11 34522, 2017

      • cats has quit
      • 2017-12-11 34541, 2017

      • cats joined the channel
      • 2017-12-11 34526, 2017

      • UmkaDK has quit
      • 2017-12-11 34507, 2017

      • UmkaDK joined the channel
      • 2017-12-11 34558, 2017

      • UmkaDK_ joined the channel
      • 2017-12-11 34505, 2017

      • UmkaDK has quit
      • 2017-12-11 34539, 2017

      • bitmap
        yvanzo: how about we merge stuff to beta today and do the release next monday
      • 2017-12-11 34500, 2017

      • iliekcomputers
        ruaok: tests seem to be failing for some reason on that PR
      • 2017-12-11 34508, 2017

      • ruaok
        yeah, testing locally.
      • 2017-12-11 34556, 2017

      • ruaok
        heh, I see. :)
      • 2017-12-11 34508, 2017

      • bitmap
        ruaok: so, I guess that response is from AWS, and it says our AWS_PUBLIC code is not registered https://github.com/metabrainz/docker-server-confi…
      • 2017-12-11 34542, 2017

      • bitmap
        not sure what amazon account those came from...I tried logging into the one in syswiki and the password was incorrect:)
      • 2017-12-11 34517, 2017

      • ruaok
        yeah, I found it to be incorrect today too. I updated to what I had on record, let me make sure that matches what have in syswiki
      • 2017-12-11 34557, 2017

      • ruaok
        iliekcomputers: tests passing now.
      • 2017-12-11 34518, 2017

      • ruaok
        oh I see part of the problem
      • 2017-12-11 34523, 2017

      • ruaok
        AWS != amazon associates.
      • 2017-12-11 34537, 2017

      • ruaok
        let me make sure both are up to date.
      • 2017-12-11 34508, 2017

      • bitmap
        that's why I was a bit confused, since it was saying "Your $AWS_PUBLIC is not registered as an Amazon Associate." so I thought they were somehow linked
      • 2017-12-11 34523, 2017

      • ruaok
        associates updated.
      • 2017-12-11 34556, 2017

      • ruaok
        oh, I see where the confusion is.
      • 2017-12-11 34509, 2017

      • ruaok
        they are two different concepts but have the same login credentials.
      • 2017-12-11 34528, 2017

      • ruaok
        bitmap: AmazonAffiliateProgram.md
      • 2017-12-11 34543, 2017

      • ruaok
        use that to log into AWS and associates. I've made a note in the AWS file as well.
      • 2017-12-11 34553, 2017

      • ruaok
        let me know if that works.
      • 2017-12-11 34513, 2017

      • bitmap
        thanks
      • 2017-12-11 34518, 2017

      • UmkaDK joined the channel
      • 2017-12-11 34519, 2017

      • UmkaDK_ has quit
      • 2017-12-11 34504, 2017

      • bitmap
        it sent a login code to the meb email, don't think I have access
      • 2017-12-11 34520, 2017

      • ruaok
        539453
      • 2017-12-11 34525, 2017

      • bitmap
        thx
      • 2017-12-11 34540, 2017

      • bitmap
        working
      • 2017-12-11 34545, 2017

      • ruaok
        great
      • 2017-12-11 34536, 2017

      • bitmap
        I guess I should just have MB be using the keys in AmazonAWSAccessKeys.md
      • 2017-12-11 34535, 2017

      • bitmap
        I don't know where these other ones came from, for security reasons they don't show you your active keys... but apparently you can have up to two at a time
      • 2017-12-11 34552, 2017

      • ruaok
        or just generate new ones, just in case?
      • 2017-12-11 34542, 2017

      • bitmap
        sure, I'll try that instead
      • 2017-12-11 34515, 2017

      • bitmap
        ah, they show you the public codes for your active keys. I guess there weren't any generated before. so the current MB ones weren't on this account
      • 2017-12-11 34501, 2017

      • bitmap
        I'll switch out the codes and see if it works
      • 2017-12-11 34514, 2017

      • ruaok
        I also thought those were two separate things, so I guess I broke the old keys when I cleaned up the associates stuff
      • 2017-12-11 34543, 2017

      • bitmap
        not sure, it was working until very recently
      • 2017-12-11 34509, 2017

      • ruaok
        huh, yeah not sure. it was a few months ago that I cleaned up
      • 2017-12-11 34529, 2017

      • samj1912
        bitmap: the docker file for solr was always there in the mb-solrquerywriter repo, it just wasn't updated, gentlecat started using docker-solr, but git submodules are a massive PITA, especially when there are 2 layers of them so I simply updated the solrquerywriter repo with the latest docker file
      • 2017-12-11 34550, 2017

      • bitmap
        ah it stopped working on june 8, so maybe XD
      • 2017-12-11 34531, 2017

      • ruaok
        that may very well have been it
      • 2017-12-11 34543, 2017

      • ruaok
      • 2017-12-11 34548, 2017

      • ruaok
        has your exec branch running
      • 2017-12-11 34557, 2017

      • ruaok
        (with my logging and config fixes branch also)
      • 2017-12-11 34523, 2017

      • ruaok
        but it seems to work fine. I have no real way of verifying anything beyond that, though.
      • 2017-12-11 34555, 2017

      • bitmap
        awesome
      • 2017-12-11 34508, 2017

      • bitmap
        perhaps trying to update something in https://github.com/metabrainz/docker-server-confi…
      • 2017-12-11 34525, 2017

      • bitmap
        and seeing if the service reloads gracefully
      • 2017-12-11 34526, 2017

      • ruaok
        perfect, can do. hang on.
      • 2017-12-11 34517, 2017

      • ZarkBit has quit
      • 2017-12-11 34519, 2017

      • ruaok
        I pushed the config, but got nothing in the logs indicating that it was restarted.
      • 2017-12-11 34533, 2017

      • ruaok
        would it fully start the service from scratch? if so, it doesn't look like it worked.
      • 2017-12-11 34552, 2017

      • bitmap
        hmm, I think it would send a HUP signal
      • 2017-12-11 34502, 2017

      • ruaok
        oh, it took longer than expected.
      • 2017-12-11 34504, 2017

      • bitmap
        I don't know what uwsgi does in that case
      • 2017-12-11 34509, 2017

      • ruaok
        30 seconds.
      • 2017-12-11 34516, 2017

      • bitmap
        ah
      • 2017-12-11 34556, 2017

      • ruaok
        yep, seems to work fine.
      • 2017-12-11 34503, 2017

      • ruaok
        anything we should test?
      • 2017-12-11 34520, 2017

      • bitmap
        making sure it handles config updates is the only thing I can think of
      • 2017-12-11 34554, 2017

      • bitmap
        I think it shouldn't change anything else
      • 2017-12-11 34525, 2017

      • ruaok
        ok, I'll merge both PRs and deploy the updated stuff to production.
      • 2017-12-11 34554, 2017

      • ruaok
        phew. finally deployed cleanly into production.
      • 2017-12-11 34531, 2017

      • bitmap
        nice
      • 2017-12-11 34520, 2017

      • bitmap
        samj1912: ok, thanks. I think it's confusing, but maybe the mb-solrquerywriter repo just needs to be renamed
      • 2017-12-11 34513, 2017

      • bitmap
        the name of the repo and the README suggest it's for a very specific part/module to our Solr setup, so I just wouldn't expect it to be the "main" Solr repo
      • 2017-12-11 34556, 2017

      • samj1912
        yes, I was thinking the same
      • 2017-12-11 34503, 2017

      • samj1912
        I was going to ask to name suggestions :P
      • 2017-12-11 34508, 2017

      • samj1912
        I am very bad at it
      • 2017-12-11 34528, 2017

      • daksh has quit
      • 2017-12-11 34516, 2017

      • ZarkBit joined the channel
      • 2017-12-11 34525, 2017

      • culinko has left the channel
      • 2017-12-11 34513, 2017

      • anshuman73
      • 2017-12-11 34534, 2017

      • samj1912
        Freso: meeting?
      • 2017-12-11 34509, 2017

      • ruaok
        in an hour.
      • 2017-12-11 34516, 2017

      • anshuman73
        samj1912: isn't it in another hour
      • 2017-12-11 34519, 2017

      • anshuman73
        yeah
      • 2017-12-11 34537, 2017

      • samj1912 totally forgot we are back to normal times
      • 2017-12-11 34535, 2017

      • Freso
        samj1912: :p
      • 2017-12-11 34500, 2017

      • chhavi_
        totally forgot too
      • 2017-12-11 34501, 2017

      • TOPIC: MetaBrainz Community and Development channel | MusicBrainz non-development: #musicbrainz | Google Code-In https://goo.gl/8xSZem | Meeting agenda: Reviews, Consolidated contributing guidelines video (naiveai)
      • 2017-12-11 34558, 2017

      • yvanzo
        bitmap: Yes. Can we deploy gh:MBS#580 on test.mb.o?
      • 2017-12-11 34558, 2017

      • BrainzBot
        MBS-9577: Update Google custom search engine to API v2.0: https://github.com/metabrainz/musicbrainz-server/…
      • 2017-12-11 34515, 2017

      • yvanzo
        Hmm... will deploy to beta instead.
      • 2017-12-11 34523, 2017

      • zas
        ruaok: about https://github.com/metabrainz/metabrainz.org/pull… i'm not sure to understand: you write current git version to .gitversion in the top dir of the repo but where is it copied to the image ?
      • 2017-12-11 34552, 2017

      • ruaok
      • 2017-12-11 34533, 2017

      • zas
        hmmm, does it exclude .git ??
      • 2017-12-11 34545, 2017

      • ruaok
        yes
      • 2017-12-11 34505, 2017

      • ruaok
        otherwise the previous PR would've worked
      • 2017-12-11 34558, 2017

      • zas
        still i think you shouldn't copy all to /code, i wonder why .git isn't copied, while .gitattributes is
      • 2017-12-11 34509, 2017

      • zas
        do i miss something, i see no .dockerignore
      • 2017-12-11 34530, 2017

      • zas
        .vagrant is copied
      • 2017-12-11 34542, 2017

      • zas
        (at least it is in the running container on hip)
      • 2017-12-11 34506, 2017

      • ruaok
        .git is a dir and .gitattributes a file?
      • 2017-12-11 34512, 2017

      • ruaok
        but, it works ok.
      • 2017-12-11 34552, 2017

      • zas
        .vagrant is a dir too
      • 2017-12-11 34529, 2017

      • zas
        current prod container /code lists as:
      • 2017-12-11 34533, 2017

      • zas
      • 2017-12-11 34523, 2017

      • bitmap
        yvanzo: yep let's test it on beta
      • 2017-12-11 34503, 2017

      • bitmap
        thanks for merging the react pr 👍
      • 2017-12-11 34535, 2017

      • zas
        hmm
      • 2017-12-11 34531, 2017

      • Freso
        CatQuest: Ping? Around?
      • 2017-12-11 34527, 2017

      • Wizzup has quit
      • 2017-12-11 34500, 2017

      • Freso
        <BANG>
      • 2017-12-11 34505, 2017

      • CatQuest
        hey!
      • 2017-12-11 34511, 2017

      • Freso
        samj1912: Now is the time. ;)
      • 2017-12-11 34514, 2017

      • CatQuest
        I made ot home safely :D
      • 2017-12-11 34524, 2017

      • Freso
        People up: Freso, reosarevok, bitmap, zas, ruaok, yvanzo, samj1912, iliekcomputers, CatQuest – maybe up (please confirm!): LordSputnik, Leftmost, Leo_Verto, alastairp, CatCat, chhavi_, ListMyCDs, ferbncode; anyone else?