07-17-2008 07:54 AM - edited 03-06-2019 12:15 AM
I have a setup with 2 routers at each site on individual T1 links that I am running an IGP (eigrp) across and then using GLBP on the inside interfaces for load-balancing / fail over. When I am advertising the dynamic routing protocol into the core-switch at each location, the switches are peering with the physical address of the inside interface, and not the GLBP virtual address. I need to force this, or I guess I could just have the core switch default-route to the GLBP address.
One site is eigrp into the core, the other site I am redistributing eigrip into rip (hp switch as the core).
Any suggestions?
07-23-2008 11:02 AM
Bill
GLBP/HSRP etc. are primarily used for end hosts to have redundant gateways and not by routing protocols which will always peer on the physical addresses, at least the IGP's anyway.
If you want load-balancing/failover then you don't need GLBP. The core switch should see 2 equal cost paths to both routers for all destinations on the WAN side assuming the core switch has a link to each router ?
Jon
07-23-2008 12:44 PM
Jon,
Thanks for the advice. I think I was coming to this conclusion. Let me think about this for another day, and i'll get back with either points or more questions.
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