ā04-09-2017 07:39 PM - edited ā03-05-2019 08:19 AM
Sorry for newbie questions. I was reading "Cisco Intelligent WAN (iWAN)" from Cisco Press.
Page 85 talks about "NHRP Redundancy". It gives an example that multiple NHRP mapping commands can be added to a single tunnel interface. For example:
interface Tunnel100
ip nhrp nhs 192.168.100.11 nbma 172.16.11.1 multicast
ip nhrp nhs 192.168.100.12 nbma 172.16.12.1 multicast
...
Instead of using a single tunnel, I've also seen that multiple tunnels are used, each tunnel connects to a dedicated hub.
What's the pros and cons of using multiple tunnels vs. single tunnel? (as far as redundancy is concerned)?
Thank you very much!
ā05-30-2017 05:27 AM
Personally, I would always use multiple tunnels simply because it can become more complicated to attempt the use of, say, a single /24 in multiple hub locations. If the routing between these locations will allow for the Anycasting of whatever subnet you place on the tunnels then this is not a concern.
ā05-30-2017 11:10 AM
Hello.
For iWAN each DMVPN cloud is a "WAN interface" (named) or a "transport". And PFR preferences are described for the names.
So if you have single Hub device per cloud - you would lose whole WAN transport in case of a hub device failure. At the same time if you create unique names per every transport (one hub per cloud) and write proper policies - this should not affect your PFR domain too much.
I think the reason to have multiple hubs per cloud are to simplify configuration/support and meet some iWAN limitations.
The reason to have single hub per cloud is to have extra redundancy on spoke side (as in iWAN 2.x same transport may be terminated on a branch only once).
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