This document describes common design patterns when Aviatrix Transit Firewall Network (Transit FireNet) is deployed.
FireNet supports Insane Mode.
You can deploy two Firewall Networks, one dedicated for East-West traffic inspection and another for egress inspection.
Note you must follow the configuration sequence below:
- Disable the Traffic Inspection of the FireNet Gateway intended for egress control.
- Enable Egress Control for FireNet Gateway intended for egress control.
- Build connection policies.
When Aviatrix FQDN gateway is deployed in a VPC/VNet, it uses a public IP address to perform both whitelisting and NAT function for Internet-bound traffic. Sometimes these Internet bound traffic are partner API calls and these partners require to limit the number of IP addresses for each customer of theirs. In such situation, you can deploy FQDN in a centralized manner as shown in the diagram below.
Since the default routes are propagated over the Aviatrix Transit Gateway peering, you can consolidate the Internet-bound egress traffic to the firewalls in one region, as shown in the diagram below.
If you need to have a distributed egress for each region, make sure you filter out the default route 0.0.0.0/0 when you build the Aviatrix Transit Gateway peering, as shown in the diagram below.
This Ingress Protection design pattern is to have the traffic forward to firewall instances directly in Aviatrix Transit FireNet VPC/VNet as shown in the diagram below. In this design pattern, each firewall instance must configure (1) SNAT on its LAN interface that connects to the Aviatrix FireNet Gateway and (2) DNAT to the IP of application server/load balancer. The drawback of this design is source IP address is not preserved when traffic reaches the application.
For an example configuration workflow, check out Ingress Protection via Aviatrix Transit FireNet with Fortigate.