I’ve been using the FH to TGW connection via GRE/BGP for some time now with no issues.
I’m now looking into how we might run another FH in another AWS region and utilize the DR feature in the PepVPN connection. There are no issues connecting PepVPNs this way, but the problem comes when you look at the AWS side routing.
Say your on-prem networks need access to AWS VPC resources, the routes for this can be easily made from the FH propagated BGP tables>AWS TGW> VPC and back. As I said, this has been working great for me.
Now lets say I’d like a FH in US-West-2 as a DR for my original in US-East-1. For traffic to pass from US-West-2 to US-East-1, I have to make static routes to pass that traffic through a peering connection between those TGWs. Problem is the BGP tables from one FH are the same as the other. There would be conflicts in the routing needed to make this happen.
Even if this was done in one region with one TGW and separate availability zones, (NOT TRUE DR) there would still be conflicts with the two FHs connecting to the same TGW.
Wondering if AWS new Cloud Wan feature can help with this at all or if any testing has been done on DR Fusion Hubs in AWS?
Thanks