Helix API timeout when trying to get my followers

Now please, what kind of statement is this? We are not internet 3rd world here. If there are consistent problems with connections from Germany as claimed by you, it is 100% something Twitch/Amazon or the service providers used by you (e.g. Fastly is hosting the API host) can and has to fix. We have extensive and well built networks here and there are usually no problems with any service or target whatsoever. I’m sorry, but this really sounds like a cop out reply.

However your claim of general “really bad issues” does not hold true in my opinion anyway. There are tons of Twitch streamers located in Germany and we all stream without any issues 99,999% of the time. When we experience problems they are almost always due to our ISPs. The only cases where it was not like that I remember where when Twitch had bigger/global outages.

Furthermore, I have these issues both from my server in the datacenter of one of the biggest Hosters in Germany as from my home internet connection. Both have totally different connection paths to api.twitch.tv. So it really makes no sense to claim that the problem is anywhere else than in the realms of fastly or on that server.

Forcing a different IP through a override in /etc/hosts for example does not help, as the routing for the netblock still ends up at the Fastly peering in Frankfurt.

I am convinced that there is a issue with that specific peering of Fastly in Frankfurt or that specific server and that Fastly should have a look into that instead of making statements like “Germany has generally problems”. Again, we have world class internet here so claiming that this is a issue of Germany and not of Fastly or the API setup quite frankly does not make any sense at all.