RE: Some comments on draft-ietf-vrrp-ipv6-spec-08.txt

Seems like this means changing ND somewhat so I think I prefer a
different approach.  Or draft could mention, both as alternatives
(either don't run dad or modify ND as described).  I wasn't looking for
much here, perhaps just a note that mentions the issue.   

-Steve 

-----Original Message-----
From: John Cruz (johcruz) [mailto:[email protected]] 
Sent: Wednesday, May 09, 2007 4:14 PM
To: Stephen Nadas (RL/TNT); [email protected]
Cc: [email protected]
Subject: RE: [VRRP] Some comments on draft-ietf-vrrp-ipv6-spec-08.txt 

Hi Steve,

This is regarding issue 1:

Rather than not running DAD, would the following achieve the same
effect?

The VRRP master does not respond to ND neighbor solicitation for VRRP
protected addresses if the source address is the unspecified address
(::). With this, when the address owner is back up after being down, it
will send a ND-NS for duplicate address detection process. The current
master would not respond.
Therefore, the IPv6 address will get installed and VRRP will start. This
will then make the current master to move to standy state.

Does this make sense?
John

> Issue 1:  
> We think that the draft should mention that one MUST NOT run DAD on 
> interfaces where the VRRP protected IP address == interface IP 
> address.
> It may even make sense to say one MUST NOT run DAD on any VRRP 
> protected address.  In either case, the usual result for DAD on such 
> addresses is
> a DAD failure.    
> 

_______________________________________________
vrrp mailing list
[email protected]
https://www1.ietf.org/mailman/listinfo/vrrp