[musicbrainz-server] 14reosarevok opened pull request #2562 (03master…MBS-12395): MBS-12395: Report for videos in mediums that shouldn't support video https://github.com/metabrainz/musicbrainz-serve...
ephemer0l has quit
odnes has quit
PrathameshG
Hi @alastairp Just a little update:
Almost done with exams, and I'll be home on 13th June, and would be able to give my undivided attention to this project :))
I couldn't check in with this channel daily, so I hope I didn't miss anything major!
alastairp
hi PrathameshG, thanks for the update. I already have 13th June in my calendar! how are your exams going?
lucifer: hi, I'm here but have been working on other stuff today. around for questions, otherwise back to it tomorrow
lucifer
alastairp: ah cool, wanted to release the LB flask update. tomorrow then.
PrathameshG
@alastairp The exams are going fine! Even better than I imagined actually :)
I also had a course for data warehousing and mining this semester. Can't wait to apply what I learned there in this project 👌
alastairp
btw, IRC doesn't use @ to address a person :) I get a notification if the text of my name is anywhere in the message
lucifer: as you wish, what is remaining? just testing on beta? I'm happy to be around to help with this if you want to do it
lucifer
alastairp: PR is pending review. i already deployed on test yesterday
alastairp
let me take a look
great to see that other deps are updated too!
PrathameshG
noice, IRC is pretty convenient in that sense ig
BTW was there any GSOC global summit around this time?
Another GSOC candidate posted a blog about it, and apparently there was no email sent for invitation!
it doesn't help to get the initial list of dependencies (your one looks good for that!) but it would allow us to have a top level of deps and then compile into a list of exact deps
lucifer: added some comments. all superficial stuff, nothing very important
given that there are dumps changes, did you try and make a dump and reimport it?
is floyd ok? didn't we have a disconnect issue about this time yesteday too?
lucifer
alastairp: makes sense, will add docs and utility function for assertRedirects. i didn't try dumps yet but because the tests seems to catch the possible issues but probably makes sense to test those manually as well so will do today.
there seems to be heavy load there
alastairp
yeah, I hope that the tests do actually work for what we want
I need to do the same with CB, to check that the modified tests work too
bitmap, yvanzo, zas: about the cpu surge on floyd, it may have been related to sir. there seem to be increased number of messages around the same time in rmq (although its overall and i am not sure if the individual queue stats can be retrieved at this point). and also solr cloud incoming packets were heightened during the same time.
bitmap
yes, I was just looking at the PG logs and noticed the same thing
all of the slow queries during this timeframe are from sir
lucifer
also, the high eth0 on floyd sent around 100mbps at that time much more than received.
the player emits at least 3 state changes on track change. (as you note in the comment there is no onTrackEnd event so we need to resort to a workaround).
maybe the player started emitting another extra state at some point or they modified the contents of those such that now multiple state changes match the condition.
will add some more logging for more visibility into the issue.
btw, spotify playback is entirely broken on FF. i saw some other people report it on spotify forums so probably not an issue on our end though.
v6lur joined the channel
v6lur has quit
monkey
lucifer: thanks, I had that suspicion but lacked time to investigate. Nice sleuthing! Will have a closer look, probably in about 10 days
