#metabrainz

monkey ansh ooooh new graph!! Really cool. Love the colours

P.S. I wonder if we should be careful of not dropping the artist country graph down too far? We’ve had a few people shout that out as particular useful/different to other services

[musicbrainz-server] 14reosarevok merged pull request #3496 (03master…ac-diff-warnings): Fix key warnings in diffArtistCredits https://github.com/metabrainz/musicbrainz-serve...

Happy Holi to everyone who celebrates it !

[troi-recommendation-playground] 14mayhem opened pull request #167 (03main…fix-unresolved-releases): Minor cleanup https://github.com/metabrainz/troi-recommendati...

[musicbrainz-server] 14reosarevok opened pull request #3497 (03master…MBS-10563): MBS-10563: Add autoselect + sidebar for AWA URLs https://github.com/metabrainz/musicbrainz-serve...

leftmostcat (UTC-8): hiya! Are you the author of the perl module unidecode?

zas: hi! i think the spark servers upgrade didn't go right? marlon shows only 450G of total space

all servers i checked actually.

that's the storage they have (2 SSDs of ~500gb in RAID1)

i remember it used to 1TB before the server upgrade

on each server

they were striped, not mirrored, right?

i see yeah that's possible. i think we can do RAID-0 for spark servers because HDFS does provide redundancy for the data in the spark cluster.

oh, that would explain this

that said there have been disk issues with the cluster in the past.

but 1 TB of storage is needed on almost all nodes to avoid running into disk issues. so we'd have to add new disks if we want to keep RAID-1.

ok, I'll see if we can convert without reinstalling, perhaps tricky, but I can try

I can perhaps do it from rescue, can I stop marlon in order to test?

lucifer: ?^^

zas: its currently running a job but i don't think it will succeed with the current diskspace anyway. so yup feel free to restart any/all of the spark servers.

ok, let's see

I'll try with marlon

It worked:... (full message at <https://matrix.chatbrainz.org/_matrix/media/v3/...>)

noice!

ok, that's a bit slow because it has to rebuild the array, it takes around 30 minutes per server, but I can do the 5 remaining ones in parallel

lucifer: I'll proceed with others servers

zas: :+LupinIII (IRC)

* zas: :+1:

I'll disable alerts for those servers

sure

lucifer: should we convert randy too? It has more diskspace

[@zas666:matrix.org](https://matrix.to/#/@zas666:matrix.org) yes I think it makes sense to convert randy too

ok, rebuilding arrays atm

hello?

lucifer[m]: what is it :)

Sorry had tagged by error

ok no problem! ^_^

lucifer: all servers ready :)

[@zas666:matrix.org](https://matrix.to/#/@zas666:matrix.org) thanks!

[@zas666:matrix.org](https://matrix.to/#/@zas666:matrix.org) another question, we need about 8 VMs for spark processing of MLHD dumps for probably about a week

Whwn do you think we can do that?

If mayhem is ok (money-wise), we can. Which VM specs do you need?

yup [@mayhem:chatbrainz.org](https://matrix.to/#/@mayhem:chatbrainz.org) suggested it during LB earlier this week

but, really, lets keep an eye on these servers. lets really keep it a week and not forget them, please.

[@mayhem:chatbrainz.org](https://matrix.to/#/@mayhem:chatbrainz.org) how about less servers but bigger ones? I see there are ones with 192G of ram

ram is what we need, so, go for it.

zas: okay let's coordinate this on monday. i want to make some changes to import the mlhd dumps faster.

<mayhem[m]> "leftmostcat (UTC-8): hiya! Are..." <- That's my Doppelgänger. He's written a bunch of Perl stuff and was rude enough to steal my name and interests before I even got to them.

lolwut. that's pretty cool on the whole. its an amazing module and we rely on the python version. but there is a C++ version which might be important to me soon.

He seems like a pretty cool guy, and I'm not just saying that because our interests really are weirdly aligned.

<leftmostcatUTC-8> "He seems like a pretty cool guy,..." <- have you ever contacted him?

I have! We've been in touch a few times.

My first email to him, I introduced myself by pointing out which of the people on his list I was. 😁

so awesome. :)

monkey: https://listenbrainz.org/playlist/0433cedd-da5e...

Oooh, sweet!

Great idea

lucifer: https://listenbrainz.org/playlist/13a88afa-ddf6...

I wonder how well playing the music is going to work, until all these albums come out

and only an afternoon's work too. I think we should test this for a while and then its goes into LB Radio.

monkey[m]: hit and miss so far.

Cool idea though

with improved spotify matching this could work well.

I just needed a break from the mapper work to do something I wanted to use.

yes.

user:lucifer

ah noice

* fresh:lucifer

or

fresh:lucifer::week

Ah, well if it's already released albums, shouldn't be problematic

oh yes, nothing for the future. that dont work at all.

this was originally released in 1976. but i guess its a mapping issue.

lucifer[m]: data bug. :)

or a re-release that the mapp

I'm not so sure about that, in my playlist there are some yet-to-be-released

s/or//, s/mapp/mapper doesn't get, I would guess./

yeah i think a re-released got picked up by some step but the mapping here is to the original release.

oh, maybe I didn't constrain that yet.

anyway still pretty cool

hang on.

yes, the playlists are fun so far!

generating a new one that is properly constrained.

Nice

Would be cool to be able to exclude re-releases, as you both mentioned, they feel like they're diluting the gold

yeah, might wait for the new mapper before launching into that.

but, I think I can filter out anything that was mapped to more than {range} days ago. let me do that.

https://listenbrainz.org/playlist/a840adec-4704... monkey

I'm thinking we should publish these every friday on the created for you page.

Yes absolutely

zas: , bitmap reosarevok MB is freaking out.

very slow with loads of closed connection errors

yes, we're on it (since one hour or so), blocking a lot of IPs

ah, sorry. been deep in code.

np, it should be better now btw

it is! <3

wolf is about to use all the disk space

jasje: akshaaatt i was looking into the current ExoPlayer implementation in the LB app as part of my GSoC proposal on integrating 'Listening Now' and fixing bugs in BrainzPlayer. I was considering whether we could migrate to Media3, as I have experience working with it. I was also thinking of including this in my GSoC proposal. Let me know your thoughts

GautamShorewala[: As of now we have no plans to migrate to Media3. Afaik, its undocumented for the large part. Plus that project does not deal with exoplayer that much

jasje[m]: There is some refactoring that would take place, but it would not be huge

[musicbrainz-server] 14mwiencek opened pull request #3498 (03master…mbs-12170): MBS-12170: Output containing country code for all artists in web service https://github.com/metabrainz/musicbrainz-serve...

Ooh great idea for a new playlist mayhem !