1

Im trying understand hot-potato routing better using this diagram that I have made. It shows R1 and R2 in automous system 1 (AS1) are receiving full routes from their parent providers P1 and P2 respectively. Both R1 and R2 are also injecting a 0/0 default route inside AS1. The T3 link at R2 is configured as the primary link and the T1 link at R1 is used as a backup so all outbound traffic will exit the AS through R2. Interior routers R3, R4, and R5 don't see external routes and simply follow Hot Potato Routing by forwarding all packets to the nearest exit point. enter image description here

It this topology effective/prone to errors if it uses hot potato routing?

Improve this question
5
  • 2
    What is your understanding of what hot potato routing is? It is simply normal routing. Business routers do not buffer and try to forward packets as quickly as possible.
    – Ron Maupin ♦
    Sep 20, 2022 at 16:44
  • Yes that is my understanding aswel, internal AS try to route information to the closest exit point without caring too much about which exit/entry point the AS prefers.
    – john genjis
    Sep 20, 2022 at 16:51
  • Are you advertising any routes via BGP to your ISP routers? Otherwise, this is a simple case of putting the right metrics on the WAN interfaces.
    – Zac67 ♦
    Sep 20, 2022 at 16:59
  • Yes both R1 and R2 are advertising routers the ISP router. Also would having 2 default routes cause any issues when routing?
    – john genjis
    Sep 20, 2022 at 17:05
  • You haven't shared any information about how you've implemented T3 link at R2 is configured as the primary link and the T1 link at R1 is used as a backup but a typical configuration would use local-preference and would not be error-prone.
    – Jeff Wheeler
    Sep 20, 2022 at 20:47

1 Answer 1

Reset to default
3

If R2's WAN is to be preferred and R1's WAN acts as backup, that's not hot-potato routing - it's redundant routing with active-passive uplinks.

Hot-potato routing would instead always use the shortest path to WAN, ie. traffic coming from R3 and R1 would route out of R1's WAN. This could cause asymmetric routing when external traffic enters through R2 (where your ASN is advertised normally) but exits through R1 (for a host attached to R3). Whether that presents a problem depends on your firewalling - if no connection states are tracked (or if those are sync'ed between R1 and R2) then it would work. Of course, many forms of NAPT imply connection tracking and won't work with asymmetric routing.

With an active-passive setup, R2's default route would be advertised with a better metric than R1's one, so all WAN traffic would exit via R2 as long as its link is up.

Improve this answer

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.