This discussion has been locked.
You can no longer post new replies to this discussion. If you have a question you can start a new discussion

XG Route-based VPN clarifications...

Hello all -- 

We're looking into converting from our current policy-based IPSec VPNs (created in v17) to the newly-available Route-based VPNs (v18), but of course, I have some confusion and would prefer to avoid the horrific sophos support hold-music. 

- Is it correct to say that route-based VPN would avoid the hassle of manually adding each site's subnets to the tunnel, or is that still manual, but done in static routing or sd-wan policies instead? We have a HA-VPN to a GCP project, which uses BGP on either side; would that be the same setup here between our two sites (so instead of the static route or sd-wan policy, we'd use BGP advertisements)?

- Possibly related to the previous question, does it matter what IP is used on the xfrm interfaces?  In the GCP connection, we use link-local (169.254.x.x) IPs, and in the sophos walk-thru video, they use [seemingly random] 3.3.3.3 and 4.4.4.4.  Does it make any difference?

- We have dual ISP links into each site, with four tunnels in a failover group. Does failover work essentially the same with route-based vpn?  Looking thru, it appears that there would only be two tunnels in the failover group, but each tunnel would have a backup gateway.

Has anyone here setup a dual-ISP failover using route-based tunnels and BGP advertising?  How does the failover work, in the real-world?  (our current policy-based failover is "hands-on", at best).

Thanks
-- michael~



This thread was automatically locked due to age.