Subject: Re: CARP (OpenBSD)
To: None <current-users@NetBSD.org>
From: Michael Hertrick <email@example.com>
Date: 11/05/2003 23:12:49
Emre Yildirim wrote:
>This is kinda cool. Theo always objected to implementing "VRRP" into OpenBSD,
>because supposedly, it was patented by Cisco and that was a bad thing or something.
>I recently had to infect some of my machines with Linux to try out load balancing
>(LVS project), but I guess now I have a good excuse to try out this CARP feature.
>Hopefully it will make it into NetBSD some day.
Yes, it is patented by Cisco in the US and they require one to license
the "technology" in order to implement HSRP/VRRP like things. The
patent is pretty generic and was probably granted before the US patent
office even knew what a router was. Nortel apparently has a VRRP patent
in Europe as well. I also expect to see the freevrrpd project die once
Cisco catches wind of it. http://www.bsdshell.net/hut_fvrrpd.html Even
so, a friend of mine recently got freevrrpd to work on NetBSD with some
minor changes... e-mail me if you're interested.
I saw today that someone has already patched FreeBSD-current to support
I'm trying my best to force-feed it to NetBSD, but I'm still learning
here.... so it may be done some time in 2010. :-(
A list of the files modified in OpenBSD for CARP can be found here:
Once upon a time David Maxwell proclaimed:
>>On Wed, Nov 05, 2003 at 03:51:33PM -0500, Michael Hertrick wrote:
>>>In case you hadn't heard, OpenBSD has a new protocol in -current called
>>>CARP (Common Address Redundancy Protocol) to facilitate automatic
>>>failover and load-balancing.
>>>This is great news to me, a networker who has wanted these features for
>>>my BSD routers, firewalls, and servers for years. It would be even
>>>greater news if CARP was incorporated into NetBSD.
>>>Will NetBSD implement CARP in the future, or some other VRRP
>>>alternative? Or, is CARP still too young and volatile to consider for
>>Hopefully, things don't get refused for inclusion in NetBSD because of
>>Quality/Stability, and Design Goals might not be compatible - those are
>>the actual reasons. People could misinterpret them as 'ageism'.
>>If the design makes sense, and someone is interested in working on it,
>>it could well be pulled into NetBSD sooner rather than later. The
>>simpler explanation is that, since it seems to have been committed
>>around Oct 16, no NetBSD developers have noticed, or had time to
>>investigate it yet.
>>David Maxwell, firstname.lastname@example.orgemail@example.com -->
>>(About an Amiga rendering landscapes) It's not thinking, it's being artistic!
>> - Jamie Woods