05-02-2005 05:19 PM - edited 03-03-2019 09:27 AM
I have two main locations each with their own Internet connection. There is a Frame relay connection between the two main locations and a Frame relay connection to 6 remotes, a PVC to each main location. Right now static routes are configured in the remotes to go to main site "A" and then to the Internet. I would like to implement redundancy for the Internet. If I redistribute the all 0 route into EIGRP at both of my main sites, will this create a routing loop of any kind when some one hits the Internet? I am guessing, from the remote site, the EIGRP routing table will take the gateway of last resort across the Frame relay connection with the best cost.
05-02-2005 05:32 PM
It boils down to the metrics.. Assuming that the links to the remote routes are the minimum bandwith (which is the most likely scenario), your metric will come down to difference in delays. So remote sites connected to site "A" will go out through A's internet link and sites connected to "B" will go out through B's internet link. There is no chance of a loop here, unless you are running as a transit AS, which does not appear to be the case. If you want all traffic to go out over one Internet link and keep the other as a backup, I suggest you increase the delay through the backup link.
HTH,
Paresh.
05-03-2005 05:03 AM
Paresh, Thank you for your assistance. It is greatly appreciated.
Brad
05-02-2005 05:35 PM
Assuming that you want the remotes to see main site A as the primary way to the Internet and to see site B only as a backup, then you should redistribute into EIGRP at A with one metric and redistribute into B with a significantly different and worse metric. Then each of the remotes will receive advertisements from both A and B and will prefer to route through A and will use B as a backup.
This will not cause any kind of routing loop.
HTH
Rick
05-03-2005 05:03 AM
Rick, Thank you for your assistance. It is greatly appreciated.
Brad
05-03-2005 11:52 AM
The quickest way to efficiently switch over the default route in case of a failure is if the backup route is seen in the EIGRP table as a feasible successor.
Make sure when you are configuring the metric's you apply metrics that will allow the backup route to stay with in the feasability condition rules other wise failover will not be as fast as it would normally be(the ADVERTISED metric for the backup route must be smaller than the primary routes total metric).
Daniel
Discover and save your favorite ideas. Come back to expert answers, step-by-step guides, recent topics, and more.
New here? Get started with these tips. How to use Community New member guide