Container Support

When attempting to podman login to an Automation Hub deployed using the platform installer, the following error appears:

(hubtest) henderson@mars => podman login https://myhubonEC2.amazonaws.com/ --username admin --password '*************' --tls-verify=False

Error: error authenticating creds for "myhubonEC2.amazonaws.com": 
Get "https://automationhub1.testing.ansible.com/token?account=admin&service=myhubonEC2.amazonaws.com": 
dial tcp: lookup automationhub1.testing.ansible.com: no such host

Inspecting the /etc/hosts on the installation, it becomes clear that that's where the `automationhub1.testing.ansible.com comes from:

[ec2-user@myhubonEC2 ~]$ cat /etc/hosts
127.0.0.1   localhost localhost.localdomain localhost4 localhost4.localdomain4
::1         localhost localhost.localdomain localhost6 localhost6.localdomain6

# BEGIN ANSIBLE MANAGED BLOCK
<my hub's external ip address>  automationhub1.testing.ansible.com
xx.xxx.xxx.xx  tower0.testing.ansible.com
xx.xxx.xxx.xx  tower1.testing.ansible.com
xx.xxx.xxx.xx  tower2.testing.ansible.com
xx.xxx.xxx.xx  db1.testing.ansible.com
# END ANSIBLE MANAGED BLOCK
[ec2-user@myhubonEC2 ~]$ 

(ip addresses above expunged)

Of course, this hostname isn't resolvable by any external system, making it impossible to login.

Potential Resolution: pulp_container should accept a setting to configure the address that's used when communicating with container clients like podman or docker.