IPV6 cluster support

[brummetj]

Hey there! im taking a look at the current 0.14.0 version on a ipv6 k8s cluster. it looks like there are several networking issues that are occurring and depends on which service I update the bind address for. here is what I currently have configured on the helm chart.

    - name: ARROYO__API__BIND_ADDRESS
      value: '::'
    - name: ARROYO__COMPILER__BIND_ADDRESS
      value: '::'
    - name: ARROYO__ADMIN__BIND_ADDRESS
      value: '::'
    - name: ARROYO__CONTROLLER__BIND_ADDRESS
      value: '::'
    - name: ARROYO__NODE__BIND_ADDRESS
      value: '::'
    - name: ARROYO__WORKER__BIND_ADDRESS
      value: '::'

which allows the api, compiler, and everything to startup. but once I start a job and worker to starts up but it will error out with

{"timestamp":"2025-08-05T15:32:56.899322Z","level":"WARN","fields":{"message":"Failed to connect to controller: transport error, retrying..."},"target":"arroyo_worker"}{"timestamp":"2025-08-05T15:32:58.142975Z","level":"WARN","fields":{"message":"Failed to connect to controller: transport error, retrying..."},"target":"arroyo_worker"}

any thoughts on a easy fix? believe it might be coming from

        let hostname = local_address(config.worker.bind_address);
        let rpc_address = format!("http://{}:{}", hostname, local_addr.port());

in the arroyo-worker crate that is creating the address. I can see in some logs it will create a ipv4 rpc address.