Date: Fri, 12 Mar 2004 17:05:43 +0000 From: Matthew Seaman <m.seaman@infracaninophile.co.uk> To: Chuck Swiger <cswiger@mac.com> Cc: questions@freebsd.org Subject: Re: ip binding Message-ID: <20040312170543.GA89612@happy-idiot-talk.infracaninophile.co.uk> In-Reply-To: <4051DA0B.60107@mac.com> References: <E1B1lLE-0008R0-Q1@rslrs2-server.com> <4051DA0B.60107@mac.com>
next in thread | previous in thread | raw e-mail | index | archive | help
--OXfL5xGRrasGEqWY
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable
On Fri, Mar 12, 2004 at 10:40:59AM -0500, Chuck Swiger wrote:
> You can use "ifconfig ... alias" to bind more than one IP address to a NI=
C;=20
> you must you a different subnet mask, however (usually 0xfffffff). You=
=20
> also can use netgraph (ng_many2one?) to trunk several NICs together for=
=20
> fault-tolerance.
ng_one2many(4). This you can use for channel bonding -- to make a
virtual network interface out of several physical NICs, with the
implied extra bandwidth available.
However, one thing it doesn't actually do is provide failure
tolerance. As the man page says:
LINK FAILURE DETECTION
At this time, the only algorithm for determining when a link has faile=
d,
other than the hook being disconnected, is the ``manual'' algorithm: t=
he
node is explicitly told which of the links are up via the
NGM_ONE2MANY_SET_CONFIG control message (see below). Newly connected
links are down until configured otherwise.
That is, you have to manually reconfigure the interface group if one
of it's components should happen to fail. There's no means of
automatically testing that all of the components are still working
properly, and if not, of reconfiguring the interface group to work
around the problem.
ng_one2many is clearly the basis upon which such failure tolerance
could be built, but so far no one has committed the necessary patches
to ng_one2many to do that. If you need failover, what you can
apparently use is the ng_fec module by Bill Paul which implements the
Cisco Fast EtherChannel mechanism, but which apparently has no man
pages or other docs available. See Bill's announcement message at
http://docs.freebsd.org/cgi/getmsg.cgi?fetch=3D700715+0+archive/2001/fr=
eebsd-hackers/20010211.freebsd-hackers
However, you either need to be using a point-to-point link or via a
Cisco switch that supports FEC.
Cheers,
Matthew
--=20
Dr Matthew J Seaman MA, D.Phil. 26 The Paddocks
Savill Way
PGP: http://www.infracaninophile.co.uk/pgpkey Marlow
Tel: +44 1628 476614 Bucks., SL7 1TH UK
--OXfL5xGRrasGEqWY
Content-Type: application/pgp-signature
Content-Disposition: inline
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.4 (FreeBSD)
iD8DBQFAUe3ndtESqEQa7a0RAqeBAJ4zDdCIz/RfA5Ng098NWeh0ghofEQCeOX5p
tOd2nPBu4K/5Xraxy4jISHE=
=Ku1K
-----END PGP SIGNATURE-----
--OXfL5xGRrasGEqWY--
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20040312170543.GA89612>