Subject: Re: PR misc/23431 (CARP)
To: None <tech-net@NetBSD.org>
From: Michael Hertrick <m.hertrick@neovera.com>
List: tech-net
Date: 11/15/2003 12:51:02
Martin Husemann wrote:

>So, a few probably stupid questions from someone that has not looked at code
>or details at all:
>
> - Is CARP going to be standardizied(sp?) in any way, or is it just an ad-hoc
>   solution local to OpenBSD?
>
I've been informed that OpenBSD may change the CARP packet format soon.  
Perhaps to make it less vrrp-ish.  They do actually plan to include CARP 
in the 3.5 release.

> - Did someone verify the patent claims for VRRP do not affect CARP?
>   They probably are (see above, I have no clue, this is all FUD) intentionaly
>   written very unspecific.
>
-----

See the PS at the bottom of Ryan's original CARP post:
http://marc.theaimsgroup.com/?l=openbsd-misc&m=106642790513590&w=2

The following page answers some questions, but raises others.  It does 
provide a link to an abstract of the Cisco SRP patent.
http://www.foo.be/vrrp/

-----

>
>I just want to make sure we do not adapt a very non-standard solution instead
>of a real standard and later get the same problems we would have had with
>"the real thing".
>
>And obvious last question:
>
> - Did someone verify the VRRP license conditions are incompatible with
>   *our* project goals? I think it's pretty clear that OpenBSD concluded
>   that for their goals, but that means pretty much nothing for us.
>  
>
Cisco would be the only one to have the answer.  I have not inquired 
about it, though.  One question would be whether or not they would allow 
an open-source implementation of VRRP.  Other companies hold patents for 
router rendundancy protocols outside the US, as well... further 
complicating things for a globally distributed system like NetBSD.

>Michael, this is not to stop your work: as you said it is a practical
>solution and you had a problem to solve. Your effort is very much apreciated.
>
>Martin
>
>
>  
>

As I become more familiar with CARP, I see it has more in common with 
VRRP than the authors might want to admit.  The code is original and the 
methods are original, but the end result of the program is very similar 
to VRRP/HSRP.  Without access to the patent itself, I'm not sure if it 
infringes.

Maybe someday I'll actually get it to work.  :-/

~Mike.