IPv6 in the enterprise

Mohacsi Janos mohacsi at niif.hu
Mon Apr 18 19:45:13 CEST 2011




On Mon, 18 Apr 2011, Matthew Huff wrote:

> Got a few quick questions about best practices in a corporate environment.
>
> 1) Is there a the consensus about using First-Hop Redundancy protocols 
> versus RA (HSRP, VRRP, GLBP). I assume the main reason for using them is 
> faster failover and more control (priority, timers, etc..)

If you want something standardised use VRRP. If you need rudimentary 
load-balancing usee GLBP. Otherwise use what is supported in your 
environment. Sometimes lowering reachability timer (and relying on NUD) is 
the only solution

>
> 2) For first-hop interfaces on access switches, do people use eui-64 
> addresses or static? What about HSRP,VRRP,GLBP virtual interfaces. Since 
> these will show up in monitoring systems (OpenView, OpenNMS, etc..) 
> should they be static with reverse dns lookups configured?

If you have to reach as a service point  and/or put in the DNS, use 
static.


>
> 3) If using static addresses what link local addresses are people using? ::FFFF:FFFF:FFFE? Or ::0000:0000:0001 ? or something else?

Use automatically generated link-local.


 	Best Regards,
 		Janos Mohacsi



More information about the ipv6-ops mailing list