when adding a non-default static route with not yet reachable IP address as next-hop, the route is added to the routing table with the next-hop taken from the default route (and possibly another less specific route). This creates problems if the configured static route is intended to become active if and only if the directly connected next-hop subnet becomes active, and otherwise a less specific, but non-default blackhole route shall drop the traffic.
Added static route:
Intended routing table:
Actual routing table with Vlan 47 down
In the example case, traffic to 192.0.2.128/25 is sent to the default next-hop, instead of dropping it until the correct next-hop 198.51.100.111 is routed via the connected subnet 198.51.100.110/31 (Vlan 47).
Is there a way to prevent this from happening? I did not find anything in the manual or CLI help.
I have seen this on an older Arista 7050 switch running EOS 4.14.x, perhaps newer EOS versions or different hardware work differently?
Hm, the table formatting did not make it to the post, making it hard to read. I did not see that yesterday, as the question was held for moderation…
Anyway, using fully specified static routes (if I remember the terminology correctly) using both the interface and the next-hop results in interesting effects if interface and IP do not match (typo, configuration change, etc.).
Post your Answer
You must be logged in to post an answer.