Kbin.social upgrade
Currently, the kbin.social instance has planned maintenance. They are busy upgrading the infrastructure to Kubernetes, using Docker containers. They said it shouldn't take much longer than 1 hour downtime in total (if everything goes fine).
The migration will help to dynamically scale the instance, to remove the growing pains. And hopefully increase the availability (uptime) as well.
Soon additional setup configurations, best server practices and ansible playbooks will be shared with the community.
In the meanwhile you can of course use other instances (I bet you are reading this message right now from another instance).
142
Comments25
Hell yeah, reading from a Lemmy instance :D
Fediverse sure is great
Same. It’s awesome.
I'm very curious how this goes for you folks! Good luck from us over at lemmy.world ❤️
Migration seems successful.
Or not.. It's down again for a second time.. :o
Nice!
Kbin is a different code base from lemmy, right?
Yes it is. Kbin is also a reddit-alike alternative. Just like Lemmy. Also open-source. See code base: https://codeberg.org/Kbin/kbin-core
Is there a TLDR of what the difference between the two is?
mastodon = twitter lemmy = reddit kbin = both
Sure, it's also written by different developers, which some people might find important. Also Kbin is written in PHP + JS while Lemmy is written in Rust + TS mainly.
Rust > php any day
Both kbin and Lemmy are reddit-like content/link aggregators. Kbin combines content aggregation with twitter-like microblogging.
Ah, Kubernetes...
Actually excited for this (currently rolling out a k3s environment at work). For instances like this, the in demand scaling will be clutch.
Helm chart when?
Hmmm, wonder if that’s why my upvotes are off today, it was 300 yesterday. 🤔
Dam.. kbin still down
** it's down again*
From what I've seen, when it went back up for a few minutes, they changed the interface, the thumbnails were on the left side and it now shows your profile picture next to your username in the upper right. Those were the more obvious changes, of course.
I don't know if those changes stay permanent but it looked really nice! (That's just as an FYI for everyone who didn't see the site go back up)
These are only the cosmetic changes you see. Kbin.social was running a quite old version indeed. I try to keep my instance up-to-date every day, running the latest changes from the
developbranch.That being said, there are a lot of other changes you are not aware of. These can be database changes, performance improvements, backend features. Or any other code change, bug fix or improvement, which isn't always obvious to the end-user.
It's up again?
EDIT: But I notice some image cache issues.
Yeah I just tried to log in at Kbin (hooray for accounts on multiple instances!) and saw that it is expected to be back up by 20:00 UTC? But...it's currently 20:11 UTC so I'm guessing it's taking a little longer than expected.
I was afraid of this... I already told Ernest to list multiple instances on the page, so people can visit other instances. But it seems to be down for maintenance a second time.