Solved: Default Route Advertising

wesweber1 · ‎08-06-2010

We have several routers around the world connected to a VLPS cloud. We want to backhaul the traffic, Inet and private, for some of the smaller sites through other larger sites. These smaller sites usually have an internal routing switch connected to the VPLS router and to a router we're running GRE tunnels back to the corporate HQ. The primary routers are the VPLS and the GRE routers are for redundancy. We're using EIGRP for routing. Some of the routers are IOS 15.0 and some are on 12.4.

We want the routing to automatically switch to the GRE routers when there is a VPLS circuit outage.

We have a site that has that working, the VPLS router at the site, Stanford Connecitcut, gets a default route pointing to the HQ VPLS router, Mason Ohio, and if that site's circuit goes down the EIGRP reroutes through the GRE router.

The problem is, we have a site, Sydney AU, that needs to backhaul through a large site, Melbourne AU. The Sydney router currently gets a default route back to the Mason VPLS router, so we have a static default route pointing to the Melborne VPLS router. We need to be able to get a default route pointing to the Melbourne VPLS router instead of the static route.

Is there a way to do that, like filtering the incoming router protocol? and am I right in thinking the default route being advertised is the HQ VPLS router because it has the smallest EIGRP router ID? Can you force a default route advertisement from a particular router to a particular router?

Giuseppe Larosa · ‎08-07-2010

Hello Wes,

ausyd-coresw can run EIGRP ?

if possible I would run EIGRP over it

it it cannot be done for license problems or because ausyd-coresw is non cisco you can:

run an OSPF process on ausyd-coresw and R2, R3 with R2 advertising a default route only if it is receiving it from R1 on the VPLS cloud, R3 will generate a worse default route that is not used until R2 generates its own

OR you can use IP SLA to track reachability or R1's ip address on the VPLS cloud: until this IP address is reachable primary default route towards R2 is considered valid if it fails the primary static default route is removed from IP routing table on ausyd-coresw and the secondary floating static route is used, the one pointing to R3.

This is more near to your current setup but you need to check if ausyd-coresw operating system allows to do this

for reliable static routes with object tracking see the following document:

http://www.cisco.com/en/US/docs/ios/12_3/12_3x/12_3xe/feature/guide/dbackupx.html

OR you can add a link between R2 and R3 so that R2 can use R3 to reach everything when VPLS is down, you can use a vlan for which the switch ausyd-coresw acts only as a L2 switch ( non SVI on it)

Hope to help

Giuseppe

View solution in original post

Giuseppe Larosa · ‎08-06-2010

Hello Wes,

>> Can you force a default route advertisement from a particular router to a particular router?

You may want to consider using a separate routing protocol ( it can be a different EIGRP process with a different EIGRP AS number) between the two routers of interest or moving to BGP as routing protocol of use.

having a VPLS scenario this is possible.

Hope to help

Giuseppe

wesweber1 · ‎08-06-2010

I attached the jpg in case you can't read this.

All the routers are in the same EIGRP domain. All the VPLS routers are seeing R1 as the default route. The ausyd-coresw currently routes everything ( including Inet traffic) to the Melbourne VPLS router via a static route. If the circuit between R2 and the VPLS cloud goes down I have to manually change the default route on ausyd-coresw to point to R3. What I want to do is have that happen without human intervention. I was thinking of using "ip default-network" to define a couple of default networks so if the primary goes down it would reroute to the next default network. I'm not sure how that would work if at all.

Giuseppe Larosa · ‎08-07-2010

Hello Wes,

ausyd-coresw can run EIGRP ?

if possible I would run EIGRP over it

it it cannot be done for license problems or because ausyd-coresw is non cisco you can:

run an OSPF process on ausyd-coresw and R2, R3 with R2 advertising a default route only if it is receiving it from R1 on the VPLS cloud, R3 will generate a worse default route that is not used until R2 generates its own

OR you can use IP SLA to track reachability or R1's ip address on the VPLS cloud: until this IP address is reachable primary default route towards R2 is considered valid if it fails the primary static default route is removed from IP routing table on ausyd-coresw and the secondary floating static route is used, the one pointing to R3.

This is more near to your current setup but you need to check if ausyd-coresw operating system allows to do this

for reliable static routes with object tracking see the following document:

http://www.cisco.com/en/US/docs/ios/12_3/12_3x/12_3xe/feature/guide/dbackupx.html

OR you can add a link between R2 and R3 so that R2 can use R3 to reach everything when VPLS is down, you can use a vlan for which the switch ausyd-coresw acts only as a L2 switch ( non SVI on it)

Hope to help

Giuseppe

wesweber1 · ‎08-13-2010

I looked into the reliable static routing backup using object tracking and that worked perfectly. Problem solved, thanks for the tip.