#metabrainz

That’s because SIR image tags follow git tags, contrarily to MBS image tags which match deployment environment only.

monkey: awesome. approved

Just need tofix snapshots then I'll merge (which will merge into your PR, is that OK?)

yes that is fine

yvanzo: https://github.com/metabrainz/musicbrainz-serve... still says "sudo -E -H -u musicbrainz git reset --hard $SIR_COMMIT"

But AFAICT you're no longer specifying a SIR_COMMIT?

Good catch, this line should be dropped, I probably missed it in conflicts resolution.

yvanzo: there were 3 files changed, now there are 2. Did you also drop something unimporant?

*unimportant

I didn't memorize what the small change on the third file was, so just making sure that's intentional

It is just GitHub being slow to refresh its pages.

yvanzo: wasn't it the one specifying the new image, actually?

We're back to the Starman issue

?

Check in your local git working copy, there are 3 files.

https://github.com/metabrainz/musicbrainz-serve... - selenium failed with starman again

That is a new one, I guess the failure is even more random than I imagined.

Thanks monkey!

[musicbrainz-server] 14reosarevok opened pull request #2525 (03master…MBS-12042): MBS-12042: Allow providing an edit note when editing annotation https://github.com/metabrainz/musicbrainz-serve...

yvanzo: https://github.com/metabrainz/musicbrainz-serve...

Your commit did revert that change, it seems

(unless that's github being weird, I did not check locally, but)

grrrr, it seems the other nvme drive on trille exhibits same issue the one we just replaced, I guess we'll have to replace it too. That's weird, but those were likely from the same series.

Well, this time we know in advance that we need to move stuff, at least

yup ;)

mayhem: ^ :/

also, at least we have a good drive in the raid array

reosarevok: Ok, the 3rd file was changed and reverted in the same PR, thus GH displaying only 2 files in the tab :D

reosarevok: the script found all of them except for the missing auto-generated files

reosarevok: What time for the schema change on Monday?

yvanzo: I see :) Well, if that passes, then that looks good to me, but let bitmap take a quick look since he's here :D

(I just noticed the latter after running GenerateSQLScripts.pl and having a lot of untracked files)

bitmap: neat :) Really need to document this somewhere in "how to prepare a schema change"

this was a new script I added in the last schema change so my bad for not documenting it yet

yvanzo: we did 17 UTC last time, same?

bitmap: ^

this is fine with me

(that's also what https://docs.google.com/document/d/1M3eNUeTCu26... already says)

bitmap: do I need to give you access to that or do you have it?

it loads but I don't have edit access yet

Changed now, I think

yup, thanks

bitmap, yvanzo: so, we merge https://github.com/metabrainz/musicbrainz-serve... and then it's time for manual tests?

(on test.mb)

I'll wait for https://github.com/metabrainz/musicbrainz-serve... so I can rebase on that one last time and let tests run properly

also, I added MBS-12369 to the fix version, since it only affects standalone DBs anyway

MBS-12369: Standalone databases may be missing foreign keys for the documentation schema https://tickets.metabrainz.org/browse/MBS-12369

Sure, it'd be good if yvanzo took a look at https://github.com/metabrainz/musicbrainz-serve... before we merge anyway

[musicbrainz-server] 14yvanzo merged pull request #2522 (03schema-change-2022-q2…update-sir-solr-in-ci): Update SIR and MB Solr in CI tests https://github.com/metabrainz/musicbrainz-serve...

afaict the prod database has all of these fks from previous upgrade scripts. (however, I think the prod db may be missing FKs for the sitemaps schema somehow...)

#2523 can be rebased so tests should pass again :)

thanks, they're running now

mayhem: https://www.wikidata.org/wiki/Special:Contribut... RGs now

Noice

zas: what issue is the drive exhibiting?

alastairp, zas: do we need to come up with some sort of labelling standard to tie docker containers to projects?

is that the intent?

percentage used = 135%?

yes, it doesn't cost much to add labels (for internal use), the idea it to mark those needing special care when moving around to start with

yes, worn out drive imho

yeah, 855TBW is a lot

but the crit warning only appeared after we exchanged the other drive, it's weird

so better replace it to be safe, appointment taken

Hi atj!

just found the drive spec PDF, it doesn't have a quoted TBW endurance

Was any progress made to the UI for the currently running services we discussed a while back?

I wish they used DC drives, makes me nervous

akshaaatt: yes, I made a fair bit of progress by using datatables

in addition it actually simplifies the script I wrote to pull the docker information, as it doesn't need to do any template rendering

Have you made a PR? I'm looking to jump in and continue the work alongside you

Thanks!

[listenbrainz-server] 14MonkeyDo merged pull request #1989 (03recording-discovery-spark…listencard-recs-discovered-on): Custom timestamp on Recommendations' ListenCard https://github.com/metabrainz/listenbrainz-serv...

akshaaatt: I haven't created a PR yet, it's probably best to have a separate project for the script to pull the docker info and the web page that consumes the JSON

Interesting.

did you create a GitHub actions workflow to publish your design guidelines site?

just wondering if we could do something similar with this and use some test JSON data

Yusss atj

I can handle that

the remaining tasks I had in my mind for the UI were: 1) convert to bootstrap 5 2) remove the sub-menu with the "last updated at" 3) add a reload button to trigger datatables to reload the JSON

Sounds good to me atj!

4) add some JS to check if the JSON contains an "errors" key (which would contain a list of hosts the script failed to retrieve data for) and add an info box at the top

So where can I work on this code atj?

good question :)

atj: I would like to integrate borg backup to ansible, do you think it's possible without too much hassle? See https://github.com/metabrainz/borg-backup

Gotta ask good questions for good people

I guess we'll have to redesign it partly, because it currently relies on a repo deployed in /opt. We only need keys & configs for the target node, and borg installed

yes, I think a redesign will be needed to some extent, but it shouldn't be too difficult to integrate if just worry about the clients

we're probably going to need to start using Ansible vault at some point

yes, we'll have to manage more keys soon, for example we need them to deploy node exporter too

btw, we'll move aretha to ansible during schema upgrade

akshaaatt: ok, I'm thinking we create a repo for the HTML/JS that consumes the JSON, then create a test domain for it to be deployed to

any suggestions for a repo name?

yvanzo, bitmap: so if https://github.com/metabrainz/musicbrainz-serve... passes tests again we merge and add a ticket for mirror -> all?

I think so. I'm writing a ticket now

Neat

yvanzo: can you confirm merging seems fine based on comments / answers? :)

bitmap: you named the version 2023 Q1, should it be Q2? :)

(I'd be happier if it was Q1 and not always just after my birthday, but :D )

oops, right!

or no, it should be 2022 Q4? unless we decide to skip that, but we can rename it later in that case

Well, we've always skipped it recently

But I forget if we actively decided at some point those don't happen anymore?

mayhem?

Hi bitmap!

hi!

Skipping it hasn't been decided.

Feel free to make that decision.

(and we're still making the decisions one by one)

mayhem: oh, neat. I assume it's only annoyingly hot in May, rather than deadly so? :p

Enjoy!

33 in Jerez, but should be a lot cooler in cádiz.

bitmap: if we have some other smaller things to add to Q4, I'm fine with starting with that and seeing whether we want to do it during late summer

25C

That even sounds survivable

ok, Q4 it is for now

bitmap: the PR seems ready to merge :)

[musicbrainz-server] 14mwiencek merged pull request #2523 (03schema-change-2022-q2…schema-change-2022-q2-fixes): Schema change 2022-q2 fixes https://github.com/metabrainz/musicbrainz-serve...

yvanzo: next is what, new mb-docker image with this change?

before running a full migration test, I'll do a pg_dump of the prod schema (without data) and try a local upgrade using that. sometimes it finds issues

oh, to see if it's different than a brand new standalone schema? Sure, why not

bitmap: wanna document all this as you do it? :p

I'll try :)

bitmap: Can we merge schema-change-2022-q2 into master now?

yes, let's

(feel free if you're already on it)

ok, will do

atj I think something including `services` or `status` or `live`

bitmap, reosarevok: done, #2524 is ready for review

yvanzo: so, to check if the strings look legit etc? :) Can do

atj: `docker-containers` in itself sounds like a good repo name

reosarevok: To answer about musicbrainz-docker, I have to update it for mbdb schema 27 but also for sir 3.0.0, with full upgrade steps.

yvanzo: oh ffs :D

Apparently I accidentally added https://musicbrainz.org/instrument/267f14db-4c3... to prod instead of my local DB at some point

CatQuest: ^ :D

Sorry

So we should probably *not* push testytest for translation

akshaaatt: I think that's a bit too broad, maybe "docker-containers-dashboard"?

Entered https://musicbrainz.org/edit/89602291

reosarevok: Doesn’t matter if it is in the DB already.

Works for me atj

yvanzo: as in "we can remove it by hand" or?

interesting answer from Hetzner about nvme disk replacement

reosarevok: as in we can remove it later on

though they just replaced another disk on the same machine for the exact same reason....

yvanzo: sure, it just seems weird to make people translate "testytest" :)

Would there be someone around to approve reosarevok’s edit?