Date: Thu, 8 May 2014 13:50:58 +0400 From: Eygene Ryabinkin <rea@freebsd.org> To: "Bjoern A. Zeeb" <bzeeb-lists@lists.zabbadoz.net> Cc: net@freebsd.org Subject: Re: Allowing CARP to use arbitrary OUI prefix and allocating block from FreeBSD's OUI space assignment for that Message-ID: <ht5ioSVMkYjTNaadctX3apO0nWQ@dHhGgwofm7uNfL6/X5%2BbGIkDUYs> In-Reply-To: <97B3C7CB-3E64-4FE0-81C8-F1FE6FB456A2@lists.zabbadoz.net> References: <Nlrr646yWAdJ7EfNj9z5ymwq5ZQ@dHhGgwofm7uNfL6/X5%2BbGIkDUYs> <97B3C7CB-3E64-4FE0-81C8-F1FE6FB456A2@lists.zabbadoz.net>
next in thread | previous in thread | raw e-mail | index | archive | help
--s/l3CgOIzMHHjg/5 Content-Type: text/plain; charset=utf-8 Content-Disposition: inline Content-Transfer-Encoding: quoted-printable Bjoern, good day. Thu, May 08, 2014 at 09:37:37AM +0000, Bjoern A. Zeeb wrote: > On 08 May 2014, at 08:10 , Eygene Ryabinkin <rea@freebsd.org> wrote: > > As current CARP implementation somewhat hijacks OUI MAC space for VRRP > > (00:00:5e:00:01:$VRID) and this sometimes create problems, because > > routing people tend to be different from the ones that run CARP > > clusters, so their VRID/VHID can clash inside single L2 domain, and > > this often leads to breakages (because of same MAC values for the > > different clustered instances), >=20 > It often leads to a bit of logging about =E2=80=9Chey I don=E2=80=99t kno= w this > =E2=80=98version' of VRRP=E2=80=9D (well yeah) on some $vendor devices wh= o should > know better by now. Here you're talking about protocol (112) and version number that CARP packets use (version number is one higher than VRRP's one). This was fixed by most vendors, most notably Cisco. > Apart from that I thought the different version number was sufficient The thing is that both VRRP and CARP packets use MAC address (on Ethernet at least) that equals to 00:00:5e:00:01:$VRID. So in case that $VRID is the same and VRRP and CARP admins aren't aware of each other, there will be MAC conflict, so L2 packets will be switched in a "funny" manner. So, it isn't about the "control plane" messages that carry CARP/VRRP protocol type and numbers, but rather than the "data plane" messages =66rom CARP/VRRP nodes saying "hi, I am here" in replies for ARP requests and switches caching MACs in their FDB on the ports into which replies ingress. > (as it is for other protocols, and so have others who actually > started to write a draft for an independent submission early last > year and stalled on it). I am actually not in the loop on what we > ended up with in 10 but I guess given the new CARP started to > understand the old stuff glebius did not end up bumping it finally > in FreeBSD? So the problem might remain that we are on a > conflicting =E2=80=9CVRRP/CARP version=E2=80=9D? No, we're conflicting with VRRP on the MAC address space. And, as I understand, CARP in 10 hadn't changed protocol in any way, it just refurbished now CARP instances are configured and attached to the interfaces. Could be wrong here, though. --=20 Eygene Ryabinkin ,,,^..^,,, [ Life's unfair - but root password helps! | codelabs.ru ] [ 82FE 06BC D497 C0DE 49EC 4FF0 16AF 9EAE 8152 ECFB | freebsd.org ] Please, CC me: I am not subscribed to this list. --s/l3CgOIzMHHjg/5 Content-Type: application/pgp-signature -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.22 (FreeBSD) iL4EABEKAGYFAlNrU4JfFIAAAAAALgAoaXNzdWVyLWZwckBub3RhdGlvbnMub3Bl bnBncC5maWZ0aGhvcnNlbWFuLm5ldDgyRkUwNkJDRDQ5N0MwREU0OUVDNEZGMDE2 QUY5RUFFODE1MkVDRkIACgkQFq+eroFS7PvK8AD8CX6NSJ31vc2dqm1ox+PUFakQ 5uJhezXz134Vp1BHUYABAJLoH3gLYMULjgDIEYEIB9xxBXJcIXjuVgspIJnan3Dv =EK2n -----END PGP SIGNATURE----- --s/l3CgOIzMHHjg/5--
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?ht5ioSVMkYjTNaadctX3apO0nWQ>