Running into an installation failure that seems to be network related but cannot find the problem issue.
Running the installation on a single node with an all-in-one config.
The host has two networks.
Docker is using standard networking: 172.17.0.1/16
The internal pod network is the default: 10.128.0.0/14
We have moved the portal net to 18.104.22.168/16.
The issue is we are finding is that the deployment pod( the temporary one from deployment config/rc set) fails to get to the
oc logs router-1-deploy
error: couldn't get deployment router-1: Get https://22.214.171.124:443/api/v1/namespaces/default/replicationcontrollers/router-1: dial tcp 126.96.36.199:443: i/o timeout
The oc command works. The node is in a Ready state.
Other pieces of the infrastructure are running. We can even launch new pods via deployment. It is just the chain deploymentconfig->replicationcontroller->deployment_pod->API Access. that is failing.
We can go into other pods and access the API without issue.