[c-nsp] GLBP vs. HSRP vs. VRRP

[Kevin Graham]

On 2/9/06, Mark Lasarko <mlasarko at co.ba.md.us> wrote:

> After considering all of the above I had concluded that GLBP was the best
> bet for optimizing traffic to the server farm, based on the optimal use of bandwidth
> when configured properly.

You mentioned 3800's as well -- for software-based forwarding
platforms that have to be adjacent to the endpoints, GLBP can be a
good way to spread CPU resources. The host-dependent load balancing
method (even though I've had no luck finding documentation on what the
hash used is) keeps troubleshooting relatively painless.