fix(create-cluster): use Traefik config of workers (#817)

Traefik API connections originate only from the localhost, access from
cluster VPN is not needed.

The cluster VPN is not automatically inserted in Traefik configuration
of worker nodes. As result there is a configuration difference between the original
cluster leader and other nodes. This commit wants to remove this
conflicting configuration.

Revert 7c90ee2

Author: DavidePrincipi

core/imageroot/var/lib/nethserver/cluster/actions/create-cluster/50update

@@ -29,7 +29,6 @@ import agent.tasks
 import cluster.grants
 import cluster.vpn
 import uuid
-import pwd
 
 request = json.load(sys.stdin)
 
@@ -64,17 +63,11 @@ rdb.hset(f'node/{NODE_ID}/vpn', mapping={
     "endpoint": endpoint
 })
 
-# Our first Traefik instance is missing the VPN network: fix it here.
-agent.run_helper('sed', '-i',
-                f'/- 127.0.0.1$/a\ \ \ \ \ \ \ \ - {network}',
-                f'{pwd.getpwnam("traefik1").pw_dir}/.config/state/configs/_api.yml').check_returncode()
-
 # Update the cluster-localnode record in /etc/hosts
 agent.run_helper('sed', '-i',
     '-e', f'/cluster-localnode$/c\{ip_address} cluster-localnode',
     '/etc/hosts').check_returncode()
 
-
 # VPN bootstrap
 agent.run_helper('systemctl', 'stop', '[email protected]')
 cluster.vpn.initialize_wgconf(ip_address)