1

I have a cisco core switch that is connected directly to our mpls router. It learns about the 172.16.0.0/12 network via iBGP from the mpls router.

We have a lab subnet 172.16.100.0/22 behind a firewall which directly connects to the core switch.

I configured a static route on the core switch for the 172.16.100.0/22 network with the next hop of the firewall.

I cannot ping anything on the 172.16.100.0/22 network from the core switch. The firewall is open so this is not the reason.

However, a traceroute to 172.16.100.253 shows the route being taken via the mpls router.

Is there any reason why the core switch would prefer the iBGP route to 172.16.0.0/12 over the static route to 172.16.100.0/22 ???

Many thanks,

Paul

||improve this question||||
  • 1
    Can you please provide the relevant output of your routing table (show ip route)? – one.time Aug 12 '15 at 2:29
  • Do'h it was the next hop that was the issue...I had 10.26.x.x instead of 10.27.x.x thanks very much for the help :) – Tullio_IRL Aug 12 '15 at 10:16
6

Based on the limited information you provided, I'd bet that the next hop in your static route isn't reachable (isn't in the routing table). Is the static route to the lab subnet seen in a "show ip route static"?

Would need at least partial configs to troubleshoot more.

||improve this answer||||
-4

static routes default administrative distance is 254. ibgp is lower by default. but this is not the issue i misread your question. The issue is that you probably have auto summarization on. so if you look at your routing table they are probably being summarized into the same route.

also make sure the interface is configured correctly on the firewall and the subnets are the same and your port priveledges are good correct too.

check the arp table on your firewall and see if it see's the switch. check it the other way too. see if your arp table on the switch sees the firewall. if they dont check that all your interfaces are configured correctly and that auto summarization is off. your probably summarizing the 172.16.100.x into the 172.16.x.x route.

||improve this answer||||
  • 1
    Administrative distance for static routes is usually the 2nd lowest (most preferred) next to connected routes. – John Jensen Aug 14 '15 at 20:28
  • i mis read the question re read my solution. and neither of the routes the path is taking is a connected route. his traceroute says its going out the mpls route with ibgp. connected routes would be to layer 2 devices on the same subnet. none of his devices are on the same subnet he is routing to different subnets so connected routes are irrelivant to his problem – jradetsk Aug 14 '15 at 20:33
  • 1
    You're still wrong about the AD of static routes for Cisco routers. Also stating "connected routes would be to layer 2 devices on the same subnet" is also wrong. You simply can't configure more than one link within the same subnet on a router. I only mentioned connected routes because they're the only type of route that has an AD which is lower than static routes. – John Jensen Aug 14 '15 at 20:36
  • connected routes are to end devices with the same subnet that your interface has. router1 port 1 10.10.10.1 <--->router2 port1 10.10.10.2 connected route – jradetsk Aug 14 '15 at 20:45
  • 2
    If the next-hop of a static route isn't reachable, it's not going to make it into the RIB, period. The router isn't going to "try" the incorrectly configured static route first, because it isn't going to exist, hence why the router thought the best path was the available longest prefix match, 172.16.0.0/12, learned via iBGP. – John Jensen Aug 14 '15 at 21:04

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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