Eugen RochkoI know that there's been intermittent stability issues with mastodon.social and mastodon.online recently. We're working on improving our infrastructure to address this.
As promised, mastodon.online infrastructure upgraded and is now easier to scale, mastodon.social is next on the list. Our mod team grows as well!
deefdragon@Gargron What do you use for infra/scaling for mastodon.online/social?
@deef Kubernetes
@Gargron Are you using the helm chart or custom deployments? I'm also kindof curious to the scale of the cluster.
Renaud Chaput@deef @Gargron This is currently deployed using custom deployments as it was the quickest for the migration, but I intent to move it to the official Helm chart once it settled down and I have time to publish the needed changes.
Currently mastodon.online is running on 8 CCX instances from Hetzner. DB and Redis are still on the old dedicaced servers.
Currently mastodon.online is running on 8 CCX instances from Hetzner. DB and Redis are still on the old dedicaced servers.
Thomas Ricouard@Gargron π€©π€©π€©π€©π€©
@Gargron do you know when the 800 newer limit for the timeline will hit .social?
@dimillian Hopefully in the next few days
@Gargron Nice! I want to do some testing.
Badass Britt 
@Gargron @dimillian wait- are we getting a character limit upgrade?
@britt @Gargron no, itβs about being able to fetch more posts for the home timeline between a gap. https://github.com/mastodon/mastodon/pull/23349
@dimillian @Gargron thank you! Iβll look into this.
Nathan Jung@britt there is a character limit upgrade included with glitch-soc https://glitch-soc.github.io/docs/
@natejung I know about this but Iβm not going to use a fork for our instance. I will patiently wait for the option to (hopefully) be available within the Admin panel. *fingers crossed*
jerry@Gargron can you point me to a link that shows me how to migrate over to mastodon.social? Can't seem to find that server to migrate over to.
ASTRELION@jerryD
https://docs.joinmastodon.org/user/moving/#move
Iirc mastodon.social is closed to registrations, though
ζθΎΉθ―·@Gargron Thank you team ππ»
Katja@Gargron ππ½
rexum@Gargron why are there two instances?
Riley Testut 
@Gargron thanks to everyone working on this! Honestly I donβt mind the growing pains β reminds me of early Twitter days π
pastora j@Gargron
Can you explain that in English? π
Can you explain that in English? π
EdenDestroyer
Kyle Rose@Gargron woohoo! Thanks!
πΉπ£ππ₯πΉππ ππ πΉ
serklarvel@Gargron so good to hear that
Oblomov@Gargron hopefully somebody is also working on optimizing the code so less upgrading is needed ;-)
Matthew Clover@Gargron appreciate all the work you do!
helplessduck@Gargron Thanks!! βΊοΈ
Devon Dundee@Gargron Thank you! π
GJ Groothedde πͺπΊ@Gargron Thanks for everything!
gh0sti 
@Gargron how are you Guyβs handling csam and abuse reports? White paper?
dmynerd78@Gargron As someone hosting a mastodon instance, I'm curious to know how you're scaling things. Will there be a blog post or anything with more technical details?
Matthias@Gargron It would be very interesting to read in an article how you and the team did the migration to k8s in detail. Especially about the practical experiences, obstacles, the sizing of the worker nodes and learnings.
Obviously this is interesting for other Mastodon operations teams.
But also for the migration of other distributed systems there is a lot to learn.
Obviously this is interesting for other Mastodon operations teams.
But also for the migration of other distributed systems there is a lot to learn.
Peter Baumann - ApolloDS@Gargron Can you explain how you did the enhancement in scaling?
It would be interesting.
It would be interesting.
Pseudo Nym@Gargron Thank you.
artisanrox@Gargron Thanks πππ team
π€ 
π€
Darko VrΕ‘iΔ@Gargron What's your opinion on @cloudflare instances?
https://blog.cloudflare.com/welcome-to-wildebeest-the-fediverse-on-cloudflare/
https://blog.cloudflare.com/welcome-to-wildebeest-the-fediverse-on-cloudflare/
