Greetings from the ancient city of Córdoba! I'm off for today and tomorrow (but will try for the meeting tonight) so my responses will be slower than usual.
monkey: giant tortillas have been spotted, but we opted not to wait in the insane queue.
monkey[m] joined the channel
monkey[m]
"El tortillón"
Enjoy the city!
LupinIII
HI
monkey
blup
monkey[m]
Hello
LupinIII
:D
yvanzo[m] joined the channel
yvanzo[m]
Hi zas: I submitted all my changes for Ansible playbooks.
bitmap: Thanks for having updated the blog and all. I republished the above as a separate release and restored the previous release so that the links to it are not broken. Then, I replaced the previous release notes with an erratum, and updated the links for tickets.
zas[m]
reosarevok: unblocked
reosarevok[m]
Thanks!
\- joined the channel
Sintharu joined the channel
Sintharu has quit
Sintharu joined the channel
yvanzo[m]
bitmap: I restored the cron task/daily script for MB Solr backups on aretha.
Sintharu has quit
reosarevok[m]
yvanzo: I'm regenerating schema drawings for MBS-14037 during some downtime here (too hot to move)
LupinIII, monkey: I'm still mostly off this week but I'll try to find some time to look through it more properly. If I don't, ping me on the 2nd
LupinIII
👍
reosarevok[m]
Using a real example at least under examples seems sensible, even if you still use the anchovy to explain the meaning :)
For "Imagine a collection of short stories written by different authors, were only one author's name is written prominently on the book's cover." examples would also be nice
LupinIII
I agree. having a made up example to explain it, but having specific examples ans uh examples
lucifer[m]
<outsidecontext[m> "I'm confused by the lastfm API..." <- there is a higher error when accessing very old data on lfm and there is some early data for which timestamps are invalid / set to unix epoch. but other than that should be possible to get full history yeah.
yvanzo[m]
bitmap, lucifer, zas: Some help with swappiness/cgroup would be welcome. Consistently getting this message on various hosts.
yvanzo: what the exact problem? I mean I'm not sure we want containers to modify swapiness at all on those servers. And this warning just says it can't be done this way.
For it to work we need to change kernel params & reboot, but I fail to see why this is needed
lucifer[m]
zas: to disable memory swappiness on the docker containers running solr.
yvanzo[m]
zas: I’m testing the above linked PR for mirrors, not for all of our servers.
I have 4 mailed in reviews, but I guess we can have them after mayhem so he can get himself fed
Go mayhem
mayhem[m]
wow, first thing. :)
thanks.
last week was a mixed bag for me -- I was feeling a lot better, but there were a number of really pesky things that I had to deal with.
the internet at the office was changed and then our backup servers couldn't be reached.
took a few hours to get that resolved with our ISP.
there was the GSoC meeting, arranging a board meeting, reviewing PRs and the usual cadre of metabrainz tasks that needed doing.
and then on saturday morning we got on a train and took literally 3 stops and travelled nearly 1,000km in the space of 5 hours. spanish trains are awesome.
greetings from Cordoba, fin!
back to reosarevok !
reosarevok[m]
Hi!
As I said, we have 4 mailed in reviews
First is yvanzo who says:
'''
For the last two weeks, I worked on the MusicBrainz database schema change and search upgrades.
I released new versions of MusicBrainz Solr, bringing SolrCloud 9 to Docker, then search schema changes.
I released new versions of SIR, bringing both Python 3 update and search improvements, then bugfixes.
I released a new version of MusicBrainz Docker with the database schema changes only.
Another version of it should be released next to actually carry the search upgrades to mirrors.
Then Solr 7 will be officially deprecated and our Solr 7 cluster dropped one month after that.
The database schema changes of this year are fully taken into account through the search changes.
That is a premiere in years and the hope is that it will continue for the further schema changes.
The dependency upgrades (Solr 9 and Python 3) allow to consider the numerous much needed search improvements (downtime-free reindexing, dbmirror2 live indexing, SIR/Solr performances…), bugfixes and new features (related to the previous database schema changes, such as genres).
Those upgrades only happened thanks to the sustained teamwork, and the same is needed for further improvements.
Also thanks a lot to the volunteer julian45 who helped with the Ansible playbook of our Solr cluster.
'''
Thanks for all that searching and releasing, yvanzo!
bitmap says:
'''
Last week I helped deploy the schema change release and fixed one issue affecting a small number of people in the upgrade script, MBS-14032
I also deployed lucifer's PR to move MB to Sentry cloud, and fixed some of the userscript error spam there
'''
Next is julian45 who says:
'''
hi all!
this past week was spent packing my things and moving to a new apartment! a little more cramped than my last one, but i have some much-needed flexibility now, plus i've gained the experience of driving a large moving truck 🚚
i still have to wrap up the move and put some things into storage over the next couple of days, but i'm looking forward to:
