Date: Mon, 12 Sep 2005 08:52:01 -0400 From: Chuck Swiger <cswiger@mac.com> To: =?ISO-8859-1?Q?Sten_Daniel_S=F8rsdal?= <lists@wm-access.no> Cc: freebsd-isp@freebsd.org, aaron.glenn@gmail.com Subject: Re: VLAN interfaces on FreeBSD; performance issues Message-ID: <432579F1.4010807@mac.com> In-Reply-To: <43254F76.4000505@wm-access.no> References: <ED8E7F5B-7E3F-40D8-8993-76E9AB8226F9@yfug.yumaed.org> <4322FDC4.8010609@mac.com> <18f601940509110230242e8bfc@mail.gmail.com> <43243677.6020707@mac.com> <43254F76.4000505@wm-access.no>
next in thread | previous in thread | raw e-mail | index | archive | help
Sten Daniel Sørsdal wrote: > Chuck Swiger wrote: [ ... ] >> Because you cannot put one NIC into two genuinely distinct layer-2 >> collision domains. Spanning Tree Protocol won't recognize a single NIC >> as a potential connection or loop, depending. > > A vlan should be a seen as a single nic. > On other platforms, STP considers vlans as independant nics. > But would it be multihoming if you are just bridging the vlans? > I thought the essence of multihoming was multiple ip networks to which > it was a member. A VLAN is an abstraction, a way of logically grouping or seperating ports and tagging network traffic with a VLAN header, much as an IP subnet is an abstraction. A NIC is a network interface. It's a physical object. The essence of multihoming is having two (or more) distinct NICs. The most common application for multihoming is where a device performs layer-3 routing between the two or more IP networks, but you could be using SPX/IPX, DECnet, or some other non-IP protocol. You can also do bridging at layer-2, perhaps because the two sides use a different physical layer (Cat-5 ethernet cabling and wireless? Cat5 and thinnet? Cat5 and a dialup PPP link over POTS line, ...etc...) -- -Chuck
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?432579F1.4010807>