#metabrainz

alastairp: not from memory. We trim artist names so we wouldn't accept a whitespace only name iirc

_lucifer: replace the last line with just bash to drop you into a bash shell . then keep that up and then we can poke it.

👍

something weird is happening, the container immediately exits without any logs

docker run -v /spark:/spark -v /hadoop:/hadoop -v "/usr/lib/jvm/adoptopenjdk-11-hotspot-amd64":"/usr/lib/jvm/adoptopenjdk-11-hotspot-amd64" -v `pwd`:/rec --network host --name "spark-request-consumer" metabrainz/listenbrainz-spark-new-cluster:latest bash

ah! let me paste it

[musicbrainz-server] mwiencek merged pull request #2047 (master…mbs-11563): MBS-11563: Invalidate entity_link_type_counts on relationship inserts https://github.com/metabrainz/musicbrainz-serve...

i wonder if docker is setup correctly.

do you want to try without docker once?

hang on. still doing 15 things at the same time.

[musicbrainz-android] akshaaatt opened pull request #67 (master…starting_with_tests): Starting with Tests and DI Clean Up https://github.com/metabrainz/musicbrainz-andro...

sure!

testing docker now. I've got a busybox container up and it can ping and reach google.

works in host mode too.

so, it seems that docker is setup ok.

what are you trying to reach right now? do you have an example ip:port I can try?

_lucifer: OT: it is uncustomary to install anything in root on a host OS. ( /spark ) better in /usr/local

192.168.100.6:9866

oh! i'll move that. i had looked up a couple of guides and both of installed at /spark so i got the opposite impression.

[musicbrainz-server] reosarevok merged pull request #2058 (master…MBS-11581): MBS-11581: Don't try to display / offer darkened art as RG cover art https://github.com/metabrainz/musicbrainz-serve...

hmm, ok.

then leave it there. there might be something with that assumption.

i made a modification to the spark-submit script and now the conatiner is up

fine in this case since they are dedicated machines.

docker seems installed fine, what I can tell.

though 192.168.100.6:9866 gives me a 'Resource temporarily unavailable'

works from inside the request consumer container.

[musicbrainz-server] reosarevok merged pull request #2035 (master…MBS-11500): MBS-11500: Do not pass GID of removed release to avoid linking https://github.com/metabrainz/musicbrainz-serve...

are you unstuck now?

yeah seems so.

[musicbrainz-server] reosarevok merged pull request #2044 (master…MBS-11564): MBS-11564: Don't block Niconi Commons links with Nicovideo cleanup https://github.com/metabrainz/musicbrainz-serve...

woo!

should i modify rabbitmq config in this container so that we can test it with lemmy?

sure.

[musicbrainz-server] reosarevok merged pull request #2060 (master…no-parent-format-jsonld-warning): Fix Catalyst warning: ensure parent format id is not undef https://github.com/metabrainz/musicbrainz-serve...

[musicbrainz-server] reosarevok merged pull request #2059 (master…edit-preview-relationship-warning): Fix Catalyst warning: don't try using undef to access hash https://github.com/metabrainz/musicbrainz-serve...

reosarevok: thanks. just related to LB-510

LB-510: Minor text issue in error response for submit-listens https://tickets.metabrainz.org/browse/LB-510

so we can specifically say that if the field is there but it's blank then it's an error

Hi _lucifer 😀 I have been working on writing tests and following up on DI lately. I am still new to the areas and hence will take a while. I have just made the tests-pull request rn so that it will be convenient for you to catch up on my progress. I want to have a dedicated session with you in future where we could chat and talk about any pointers for me since I'm sure there are a few and I could use some advice.

just completed courses on Dagger-Hilt and Android Testing 🤓 I'll make a different PR and work on some other stuff as well soon. Just let me know 1-2 days before when you would review my tests PR. I am highly flexible on when we can do this, probably like April end/May beginning. Thanks a lot! 😀

ruaok: done.

let's stop the container on leader and send a request from lemmy.

so the consumer is up and running?

puurfect.

yes.

I guess start by importing a data dump?

yup let's do that

I'll let you do that, I'm still buried in biz dev nonsense.

👍

akshaaatt[m]: sure, let's do it sometime later this week. i'll ping you once I have reviewed your PRs.

<_lucifer "akshaaatt: sure, let's do it som"> Great!

huh! leader is named `newleader`. i guess it was always that way and i forgot :p.

which newleader? lol

bitmap, yvanzo: so, does it seem sensible to change https://github.com/metabrainz/musicbrainz-serve... to just drop rand sort

?

Is there any way to see stats of how many times it has been requested? Guess not?

yeah lol. leader.listenbrainz.org -> newleader and newleader.listenbrainz.org -> michael 😆

we could check recent nginx logs, I guess

remind me to clean up once we get the old new cluster ditched

sure. i sent the request but does not seem the container received it.

can we check the rabbitmq queue?

1 request in the queu.

on prince you can `docker exec -it rabbitmq-prince rabbitmqadmin -V /listenbrainz -u guest -p guest list queues`

bitmap: could I convince you to do that? :p

what rmq port are you trying to connect to? we're on a non-standard port.

cool thanks. i copied the configuration from the old request consumer container

should i PM you the config file?

reosarevok: in the past ~10 hours, I see 19 requests with &order=rand and 474 with &order=desc. so some, but not a ton

naw, hang on

Wonder how many of those are us testing it :p

62673 is the port.

at least one from me :)

aaaand, that doesn't connect.

firewall?

likely, but unlikely to be on on the (new)leader side

zas: remember that firewall hole you made for us? it seems to not work anymore.

until zas comes to rescue us, we're dead in the water. :(

reosarevok: what do you thinkabout renaming it to something like "unspecified (faster)"?

Is it *significantly* faster?

no worries, i'll pick up the a couple other pending PRs in the meantime then

Like, enough to make some timeouty queries not time out

I haven't measured it. do you have a link to a slow-ish edit search that you use handy?

Possibly, I’m fine with either option.

chaban: ^ do you have a link like that we could test with?

Some of the searches I have for closed edits are awfully slow: https://musicbrainz.org/collection/ece5ad67-e46...

Especially those I limited to past 30 days

Well, that made them faster :P

Thanks, +1 to rename it then :)

MBS-10846: BookBrainz edit searches don't work

Another kind of search I had to limit was personal label rel: https://musicbrainz.org/collection/6f7767ba-9d6...

At first 360 days was enough but eventually I had to reduce to 180 days

[musicbrainz-server] reosarevok merged pull request #2063 (master…EditReleaseEvents-map-keys): Add keys to EditReleaseEvents to stop warning https://github.com/metabrainz/musicbrainz-serve...

Weird. Personal label rel search seems to works fine now.

A while ago it timed out

alastairp: around?

[musicbrainz-server] reosarevok merged pull request #2023 (master…MBS-11533): MBS-11533: Block smart links: snd.click https://github.com/metabrainz/musicbrainz-serve...

[musicbrainz-server] reosarevok merged pull request #2016 (master…MBS-11527): MBS-11527: Block smart links: music.indiefy.net https://github.com/metabrainz/musicbrainz-serve...

[musicbrainz-server] reosarevok opened pull request #2065 (master…MBS-11582): MBS-11582: Add a sortname method for instrument guess case https://github.com/metabrainz/musicbrainz-serve...

bitmap, yvanzo ^ another tiny bug that can't hurt to fix if you have the time to check

CatQuest: thanks for finding that one - I guess the guess button for instrument sortnames was broken forever, wonder if before React it just copied it or something...

[musicbrainz-server] reosarevok merged pull request #1850 (master…MBS-11303): MBS-11514 / MBS-11515 / MBS-11516 / MBS-11517: Use React.useContext in more places rather than passing $c around https://github.com/metabrainz/musicbrainz-serve...

_lucifer: if it's a small question

[musicbrainz-server] reosarevok merged pull request #2062 (master…missing-link-type-id-warning): Fix Catalyst warning: don't try undef linktype id to access $loaded https://github.com/metabrainz/musicbrainz-serve...

alastairp: i actually updated the https://github.com/metabrainz/listenbrainz-serv... and wanted to talk about the script reuse thing

it'll now continue on error but still report it as overall failed

[musicbrainz-server] reosarevok merged pull request #2065 (master…MBS-11582): MBS-11582: Add a sortname method for instrument guess case https://github.com/metabrainz/musicbrainz-serve...

_lucifer: cool, so if any one fails it'll mark as failed. if all pass it'll mark as ok?

script reuse - you mean test.sh?

yup!

I'd prefer to talk about that tomorrow if you don't mind

sure!

ruaok: what's the problem?

the king of pop can reach the recently deceased king of r&b. :(

*can't

we had it working, but that was pre-firewall setup. I didn't think that the firewall would block outbound connections, so I'm at a loss.

did something revert on prince?

[brainzutils-python] amCap1712 opened pull request #65 (master…junit): Publish Unit Test Results https://github.com/metabrainz/brainzutils-pytho...

[musicbrainz-server] reosarevok merged pull request #2054 (master…MBS-9631): MBS-9631: Add 1200px to available image sizes https://github.com/metabrainz/musicbrainz-serve...

I don't think so, but the IP of the incoming connection perhaps changed?

136.243.155.251 is external IP

and it times out.

why is the IP always changing?

that's the third IP

I allowed this one on port 62673

but you may want to fix the core issue: origin IP isn't constant it seems

I had 94.130.35.211

is this IP still valid?

i do not know for sure but that IP with 94 might be of the old cluster setup. the IP starting with 136 is of michael where the new setup will be located.

ruaok can verify that.

[musicbrainz-server] reosarevok merged pull request #2048 (master…MBS-11569): MBS-11569: Limit entity types when creating collection from sidebar https://github.com/metabrainz/musicbrainz-serve...

Updating beta

zas: 94.130.35.211 is bono. still valid, yes.

its not that the IP keeps changing, we're adding and removing services.

and third IP is right. soon we will get rid of the old cluster and thus get rid of one.

_lucifer: works now. thanks zas.