Am I the only one who finds it a bit odd that the article starts off bragging about what was an outage?
And I as soon as I hear “we were still running with half our network down”, the first thing I think of is that no one really is using IPFS at all.
"Connection was reset."
Well, that might address the mystery :-)
>The problematic configuration which was applied manually (i.e. was not based on the default values of kubo-v0.17)
So how did 60% of the network become manually mis-configured? Or does a single company (cloudflare?) make up 60% of the network?