04-05-2023 03:02 PM
Hi there,
If understand your topology you should have at the start a 'trust' zone which holds the services your remote hosts want to access.
To prevent the hosts behind the remote VPN peers from communicating, place each VPN tunnel in its own security zone. Create explicit security polices permitting traffic between the 'trust' and VPN individual security zones. The default inter-zone policy of 'deny' will ensure the VPN hosts cannot communicate.
How is routing between the hub firewall and VPN peers being handled? Is the hub firewall advertising a default route? Do the VPN peers have a static default route with a next-hop of the hub firewall?
If you are dynamically advertising routes between the hub and VPN peers, then so long as you do not re-advertise received VPN peer subnets there will be no way for this inter-VPN traffic to be routed from the remote sites. BGP would be the best way to control this.
cheers,
Seb.
