horrific k8s devstream happening now https://trash.cloud/
the computer is the true vampire
as far as i can tell, for any non-http port (e.g., 1935 for rtmp), you can expose that port to one (1) service on your entire k8s cluster https://kubernetes.github.io/ingress-nginx/user-guide/exposing-tcp-udp-services/
so it's literally impossible to run more than one rtmp streaming server on the same cluster
someone asks 'is there any way to actually let multiple services listen on the same port, like two FTP servers on 21' and dude replies with this colossal comment that boils down to 'yes, here's how to set up two services, listening on 6379 and 6380 respectively' https://stackoverflow.com/questions/61430311/exposing-multiple-tcp-udp-services-using-a-single-loadbalancer-on-k8s
just beautiful brain worms
anyway at least https://trash.cloud/ is back up, there's a possum for you
the good news is OBS can stream over RTMP to ports other than the default (1935), so you can run multiple streaming servers on the same cluster if you manually assign them each a port in your ingress controller https://gist.github.com/dunn/7df5a5426b8216a86b514e6d8e9c7733#configure-the-ingress-controller
because of that i spent the last two days spelunking through the absolute nightmare that is traefik's documentation, only to find they have effectively the exact same problem because of a bug nobody seems in a rush to fix: https://github.com/traefik/traefik/issues/7144
?????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????? https://docs.nginx.com/nginx-ingress-controller/configuration/transportserver-resource/#customization-via-configmap
the nginx corporation is my last hope https://docs.nginx.com/nginx-ingress-controller/configuration/transportserver-resource/
what are the odds it won't work either