Specify IP addresses the Ingress controller is listening on

[bowei]

From @cluk33 on January 17, 2017 13:47

We are running k8s on bare metal. It would be great to specify the IP addresses the nginx ingress controller is listening on.

This would enable us

1. to route traffic for specific IPV4 IPs to k8s
2. also route (http) traffic for IPV6 addresses to k8s.

AFAIK 1) can be achieved by using a service in front of the ingress ctrl with external (IPV4) addresses. But currently we do not see any possibility to achieve 2).

Might be related to #131 .

Thanks a lot!

Copied from original issue: kubernetes/ingress-nginx#137

[bowei]

From @euank on January 18, 2017 18:14

2 is related to #34; I use a daemonset with net=host to manage ipv6 ingress

[bowei]

From @bprashanth on January 25, 2017 21:8

you probably want to specify the name of the interface instead of the actual ip? eg if you just want to listen on cbr0 or docker0 with net=host, so the system is free to manage ipam. The way the nodecontroller assigns cidrs to nodes today, you don't know the range a given node will get.

[bowei]

From @cluk33 on January 26, 2017 20:9

@euank Thanks for the pointer. But isn't nginx listening on all IPV4 and IPV6 addresses which are available on the host in that setup? We are trying to migrate one IP at a time to k8s and keep the other IPs and services working outside of k8s.

[bowei]

From @cluk33 on January 26, 2017 20:13

@bprashanth Thanks for the suggestion. We are using a fixed /28 subnet for our incoming traffic. Therefore it would be great to specify exact listen addresses for nginx. What would also work is using the hostnames from the ingress object for the nginx listen directives. Though my attempts using a custom nginx template doing that failed utterly. ;)

[bowei]

From @euank on January 26, 2017 23:31

@cluk33 indeed, it is for me. It's not ideal, but it's worked for me so far. Specifying only one ipv6 address would be saner.

[bowei]

From @euank on January 26, 2017 23:32

@bprashanth on the other hand, you don't necessarily want to listen on an interface for the case of eth0 or whatever since you've potentially got multiple different ips on it.

[bowei]

From @TattiQ on October 11, 2017 15:45

I also use bare metal and would be great to be able to assign an external IP for ingress
and thus use the hostnames from the ingress resource for the nginx listen directives.
Currently I am not sure how I can get this working - private docker registry (2 rc replica's, talking to the same redis deployment + the same filesystem (nfs mount to all kube nodes) ) behind a nginx controller + ingress resource points to a hostname docker-registry.bla.bla . My understanding is docker-registry.bla.bla should be pointing to the nginx external IP in DNS which in its turn redirects traffic to the endpoints in round robin. But we have no way to point it to nginx, have we ? Please correct me if I am wrong somewhere there, I am a newbie to kubernetes

[fejta-bot]

Issues go stale after 90d of inactivity.
Mark the issue as fresh with /remove-lifecycle stale.
Stale issues rot after an additional 30d of inactivity and eventually close.

Prevent issues from auto-closing with an /lifecycle frozen comment.

If this issue is safe to close now please do so with /close.

Send feedback to sig-testing, kubernetes/test-infra and/or @fejta.
/lifecycle stale

[hdave]

How can one use the gce ingress controller on an "on-premises" kubernetes installation? I think this issue is in the wrong project....yes?

[nicksardo]

Yes, this isn't relevant for GCE.