Just an FYI that I'm interested in collaborating on getting an kamailio-as-sbc instance up and running in Kubernetes, and specifically in OpenShift (it's running k8s). I'm an author of this project https://github.com/dougbtv/docker-asterisk and I've been using that as a basis for running kamailio / asterisk in CoreOS using fleet as a scheduler. While it's a rather feasible way to run kamailio & asterisk with HA -- fleet as a scheduler is rather low-level compared to k8s and takes a lot more care. I'd like to leverage some of the higher level features of k8s.
I'll come back at some point with some suggestions on what you've got going, but, after more digging on my part. Last I really looked, I started to get the feeling there's a need for a router for k8s to be written to accommodate routing the wide ranges of UDP and potentially for other UDP behavior one would want with a SIP setup.
P.S. imo, NAT is just a pain, it's not "bad" per se.
Just an FYI that I'm interested in collaborating on getting an kamailio-as-sbc instance up and running in Kubernetes, and specifically in OpenShift (it's running k8s). I'm an author of this project https://github.com/dougbtv/docker-asterisk and I've been using that as a basis for running kamailio / asterisk in CoreOS using fleet as a scheduler. While it's a rather feasible way to run kamailio & asterisk with HA -- fleet as a scheduler is rather low-level compared to k8s and takes a lot more care. I'd like to leverage some of the higher level features of k8s.
I'll come back at some point with some suggestions on what you've got going, but, after more digging on my part. Last I really looked, I started to get the feeling there's a need for a router for k8s to be written to accommodate routing the wide ranges of UDP and potentially for other UDP behavior one would want with a SIP setup.
P.S. imo, NAT is just a pain, it's not "bad" per se.