The bucket is in a really bad state, about 95% of writes are failing and about 10% of reads are failing. I basically stopped communication between the instance and the object storage because it's useless.
Since Scaleway isn't fixing anything I'm migrating to OVH atm
So the instance will be half-working for a few more days
I closed the registrations since it wouldn't make sense to keep them open
@admin Makes sense to close them. I can't imagine this instance taking very more users without serious problems. Each instance should really cap the number of registrations to prevent data overflow / DOS.
Sent you a donation via PP!
@taoeffect @admin Actually, it's quite the opposite: scaling up improves availability because with scale you can afford more redundancy and more sysadmins at a lower cost per user.
@taoeffect @admin Even when the service is free to use, a certain percentage of users are able to donate money or volunteer their skills.
@taoeffect @admin Wikipedia scaled up to billions of users just with donations (not to mention the millions of volunteer editors).
Nowadays, Wikimedia has a funny problem: having more money in the bank than good ideas to spend it on and talented project leads execute those ideas.
@codewiz @admin You bring up a valid point. It's up to @angristan if he wants to take that approach. If he doesn't, then closing registrations is the right thing to do.