#metabrainz

tag v-2021-02-10.0 pushed.

will deploy to beta now.

ruaok: no changes outside beta since last Wed

So not a code change

zas: did you make any changes?

I didn't

seems ok now. just a wobble.

I wish we could work out what causes them and fix them

Probably top radio show airing time in venezuela or something…

pg issue?

^^ cause ?

_lucifer: Mr_Monkey alastairp : beta updated. -web-beta and -follow-dispatcher-beta were both restarted.

plz help sanity check?

It beta be working !

groan.

the playlists websockets feature is not working.

seems not sure if beta is configured 100% correctly for that. any idea what is amiss?

the config probably needs to be updated.

_lucifer: Works for me on beta.

Perhaps reload the page without cache

huh! i'll try in incognito then

Mr_Monkey: got playlist you can invite me to for testing?

ruaok: https://beta.listenbrainz.org/playlist/15071daa...

done https://beta.listenbrainz.org/playlist/4245ccd3...

i am getting http error on adding collaborator, i'll delete my cache and test again

`Error HTTP Error Forbidden` while trying to add a track to that playlist, Mr_Monkey

Hm. Maybe something not working with collaborating, rather than WS

it's working beta?

:D

websockets working for me now :D

ruaok: anything in browser console?

adding track working as well

As collaborator?

no as owner

shivam-kapila: the 403, nothing else.

I can't seem to add remove or move tracks as collaborator

So that's an issue confirmed.

_lucifer: can you add me as colab to your playlist you listed above

Let me test on prod, see if that's a tregressions

shivam-kapila: done

Mr_Monkey: collab adding didnt work for me on prod

I'm getting the same 403s on prod, doesn't seem to be a regression. New WS update feature works for my own playlists, but as collaborator I don't see the changes reflected.

I'll debug this collaborator issue locally, but there's no regression I can see on beta ruaok

should I wait or proceed with the release, knowing that we're likely going to do another one soon?

You can proceed I think. Ill get on debugging that today

Mr_Monkey: https://gist.github.com/mayhem/ffd5c97f9f6fad62...

this is a spurious test failure that I can't repeat right now. but maybe related?

Yeah, I was going to add a ticket for that

ok, I'll proceed with the update.

It's the list of collaborators that's not sorted, from what I gather

yea.

So not really an issue, but something we discussed

don't bother with the ticket. I'lll just fix it after the release.

[data-set-hoster] alastair merged pull request #3 (master…master): Use open-ended versions so that downstream projects can pin exact versions https://github.com/metabrainz/data-set-hoster/p...

OK

alastairp: https://travis-ci.com/github/amCap1712/brainzut...

_lucifer: data-set-hoster tag v-2021-02-10.0 pushed

seems that pytest fails at just finding the illegal character in .py file

_lucifer: in python 2 you need the https://stackoverflow.com/questions/41680533/is... header

if there are utf-8 characters

starting the release.

[musicbrainz-server] reosarevok opened pull request #1905 (master…MBS-11375): MBS-11375: Use top-align for release event tables in Edit release https://github.com/metabrainz/musicbrainz-serve...

updated.

alastairp or iliekcomputers : please publish the changelog.

on it

whoops https://travis-ci.com/github/amCap1712/brainzut...

ERROR: Service 'test' failed to build: toomanyrequests: You have reached your pull rate limit. You may increase the limit by authenticating and upgrading: https://www.docker.com/increase-rate-limit

[listenbrainz-server] release v-2021-02-10.0 has been published by github-actions[bot]: https://github.com/metabrainz/listenbrainz-serv...

done!

thx

_lucifer: oh yeah, I've run into that before!

does it make sense to host own our images?

not much we can do on travis I guess

yeah right

_lucifer: we should sort out https://github.com/metabrainz/brainzutils-pytho...

our jenkins does a docker login, so it has higher rate limits

:D, what work is needed for it?

I got some items working, I can't rememer what was next

I wanted to merge everything together into a single run file, which mostly worked. however on jenkins you can only show logs for one test run

we also need to integrate BU#44 into this, because there will be changes, and something similar to LB#1256

I think that was most of it

/!\ this channel has moved to ##hamradio /!\

/!\ this channel has moved to ##hamradio /!\

no it hasn't!

/!\ this channel has moved to #nyymit /!\

Freso, can you take a look at the spamming :)

bitmap, yvanzo: can you check what happens with mb ws & website?

/!\ this channel has moved to #nyymit /!\

it's in #musicbrainz too

/!\ this channel has moved to #nyymit /!\

ug

/!\ this channel has moved to #nyymit /!\

hmmm, last issue was only on cage, 1mn request time...

load average: 9.38, 45.78, 31.05

what is running on cage already?

jenkins

we may need to find another home for jenkins.

clearly the two do not play well with each other.

:(

** Warning: if there is any bot in #metabrainz which should be exempted from Sigyn, contact staffers before it gets caught **

So it looks like the collaborative playlists aren't really fully implemented and we just didn't realise. Go us!

lol

No mention of collaborators in the playlist item add/move/delete API endpoints

At least it's a helpful error message

lol, at least we now reminded ourselves.

If only the helpful error message made it back all the way to the front-end to be displayed to the user… :)

But that's another bug for another time !

alastairp: can you re-run jenkins container with --cpus="6" on cage?

sure

what was it before? 4, right?

we configure jenkins to have only 4 workers

but there's no docker-level resource restrictions

zas: did you remove the lb dump private files?

nope

what's the JDK its using?

before 8u192, Java didn't play with docker

*play nice

zas: oh, I found them. totally not what I expected.

_lucifer: 11, I think

that's good!

i tested locally the python 2 change, adding the header works. thanks!

zas: on the LB dumps issue: if an old dump fails for whatever reason, we stop the dump process BUT we don't clean up the dirty files. then for the next run, we sync the whole dir and that copies out the failed dump.

[listenbrainz-server] MonkeyDo opened pull request #1275 (master…websockets_server_port): Clean up: Correction of websockets port in config template https://github.com/metabrainz/listenbrainz-serv...

alastairp: can you fast-forward the your mbdb-test branch to my fork's mbdb-test branch.

after that 44 should be ready to be merged.

yes, one sec

restarting jenkins

_lucifer: pushed

what order should we merge 44 and 46?

zas: registrator on cage restarted when I started jenkins. not sure if I made it shut down or not?

alastairp: 44 first, i'll rebase 46 then to make it ready for .erging