<suvid[m]> "on my testing environment, it..." <- Wait did this message mention you monkey? I had replied to your previous message on my pc but it doesn't show it as a reply on my phone so I'm confused
Kladky joined the channel
lucifer[m]
monkey, ansh hi! email on support about issues in connecting apple music to listenbrainz from safari on iphone. any ideas?
zas, julian45: hi! i would like to upgrade spark cluster servers to latest ubuntu lts (noble). we are on focal atm which is eol next month so maybe we should upgrade other servers too.
also, I am planning to update dockerfiles for other python projects to noble with python 3.13.
zas[m] joined the channel
zas[m]
lucifer: you can upgrade them one by one, and we can make changes to ansible so those servers are added to noble group
lucifer[m]
zas: makes sense, the spark cluster is free for the rest of the day. so if you have time, i can coordinate with you on that today. do you run ansible to upgrade the servers or just do-release-upgrade on the server itself?
cynic
hello! i'm from wuvt.vt.edu, where we're attempting to attach release-group MBIDs to hand-typed records of our stationphysical media
eep, didn't mean to hit enter there.
regardless, we've hit rate limits for the solr indexed search. the natural next step would be setting up a local copy, probably in our k8s cluster, but the whole 'outdated solr version' thing seems a bit spooky
zas[m]
lucifer: You can just run do-release-upgrade on each server, but we need to make changes on ansible config too. We can't upgrade systems from Ansible directly atm afaik.
cynic
do the folks itc have any advice for our attempted goal here? wrt setting up a mirror, or just in general. auto-assigning MBIDs is prone to error, we're aware, but it's a quick solution and hopefully not an entirely permanent one.
lucifer[m]
zas: the upgrade fails on michael due to lack of disk space on /boot.
zas[m]
Be sure to run apt autoremove before starting the upgrade, in order to remove old kernels
BrainzGit
[listenbrainz-android] 1407jasjeet merged pull request #553 (03main…bug_user_is_coming_null_if_we_click_on_you_in_feed): Fix null profile screen if we click on "You" in feed https://github.com/metabrainz/listenbrainz-andr...
[bookbrainz-utils] 14dependabot[bot] closed pull request #52 (03master…dependabot/npm_and_yarn/cleaner/multi-c90c854c9a): chore(deps): bump cross-spawn, eslint and typescript-eslint-parser in /cleaner https://github.com/metabrainz/bookbrainz-utils/...
lucifer[m]
zas: hi again, i did run apt autoremove but that doesn't remove anything atm. i think its to do with https://askubuntu.com/questions/563483/why-does.... its probably safer if you could tell me what changes to make on one server to allow removal of old files and then i can do the same for remaining servers.
zas[m]
lucifer: would it be possible to do an install from scratch for those servers?
lucifer[m]
zas: sorry not sure what you mean
zas[m]
I mean proceeding to a full reinstall (losing all data)?
lucifer[m]
yes i think it should be possible.
do you think we can also upgrade spark servers at the same time?
if its no extra cost.
zas[m]
If we want to upgrade the servers (for example to AX42) we'll have a setup fee for each and their prices monthly cost increased a lot (from 35€ to 49€ excl. VAT). But cpu are 30-40% faster (more cores and more threads), lower consumption (but it changes nothing for us price-wise). The rest (RAM/storage) are the same, but likely a bit faster.
If we keep the current servers, no extra cost. We reinstall from scratch (if possible) and get new software on same hardware.
For the same extra cost we could get 2 more AX41-nvme but Hetzner don't offer them anymore, I dunno how the Spark cluster performance scale with cpu and/or increase of nodes.
lucifer[m]
i see, let's just reinstall from scratch then, zas for now.
zas[m]
Ok, do you have data to save first?
lucifer[m]
no we can proceed with the reinstalls.
minimal has quit
zas[m]
ok, I'll start with michael then.
lucifer[m]
sounds good
vardhan has quit
vardhan_ has quit
minimal joined the channel
vardhan_ joined the channel
vardhan joined the channel
zas[m]
@lucifer: michael is now ready, I'll proceed with jackie
lucifer: randy is in the set too right?
I can upgrade jackie jermaine marlon randy tito in parallel, michael upgrade went smoothly
lucifer: ping, can you confirm randy is in the set of servers?
bitmap, zas: what do you think about moving all the base images we have into one repo? base-image, consul template base, python, node? not sure if i am missing any.
KrishnaCosmic has quit
bitmap[m] joined the channel
bitmap[m]
<lucifer[m]> "bitmap, zas: what do you think..." <- I could see merging the first two since they're closely linked. the others don't seem very related to me, e.g. how would you manage tags?
lucifer[m]
ah right, yeah good point about tags.
i was thinking we could do date based versioning for git tags and each image tag uses the date in it somehow too.
Sophist_UK has quit
Sophist-UK joined the channel
minimal has quit
minimal joined the channel
BrainzGit
[listenbrainz-server] 14fettuccinae opened pull request #3218 (03master…sanitize_username): LB-1442: Fix sending emails if username contains a special character https://github.com/metabrainz/listenbrainz-serv...
I'm a bit sick again. What the heck... I did hang out with a friend's children recently though! I blame the little plague dealers.
Just the usual comms last week, but of design discussion, but very quiet overall.
'''
derat says: "Nothing to report again, but I'll hopefully have time to start working on some miscellaneous server tickets this week."
And yvanzo says "Was off last week."
On my list for today: julian45, lucifer, kellnerd, mayhem, outsidecontext, bitmap, monkey, MonkeyPython, jasje, Gautam Shorewala, reosarevok, ansh, zas
julian45: wanna get us started?
julian45[m] joined the channel
julian45[m]
Sure!
Hi all, since last week I worked with zas to fork and fix the GitHub Action we use to manage our DNS records now (action wasn't being maintained upstream and needed to be updated), and I also made some inroads on the SSO project. Hopefully will be reaching out to the tester group on that this week!
That's all, go kellnerd ?
kellnerd[m] joined the channel
kellnerd[m]
Hey 👋
Last week I merged a couple of userscript PRs for some of ROpedebee's scripts.
I started with the Harmony integration of the cover art uploader which YoMo12 provided before I reviewed a couple of other PRs.
And since I was already looking at it, I finished the Harmony integration to seed release URLs instead of just the front cover URL and made it display the image sizes as well.
outsidecontext[m joined the channel
outsidecontext[m
Sounds awesome
kellnerd[m]
Both changes have been released today, that was one of the most wanted features which people were missing form a-tisket.
(And most never realized that it is actually that userscript which is doing the magic.)
That's it, I am currently creating release lookup tests for the Tidal v1 API provider from cached data 😓
outsidecontext: Wanna go next?
outsidecontext[m
sure, hi everyone
reosarevok[m]
Still on my list for today: lucifer, mayhem, bitmap, monkey, MonkeyPython, jasje, Gautam Shorewala, reosarevok, ansh, zas
outsidecontext[m
last week I did code reviews of zas' refactorings for Picard. Also checking some tickets and bugfixes
LupinIII
!m kellnerd[m]
BrainzBot
You're doing good work, kellnerd[m]!
suvid[m]
Hey reosarevok
I would also like to chip in if its possible 😅
outsidecontext[m
I also did experiment with and actually implement Picard browser integration using a custom protocol. Trigger for this was latest Safari releases on macOS preventing the existing browser integration links to load
It's generally working, but brings its own issues. Very platform dependent, and debugging platform specific issues is already very time consuming
