Re: vlans and netbsd-current

[Thor Lancelot Simon <tls%panix.com@localhost>]

On Mon, Jun 09, 2014 at 03:22:43AM +1000, Darren Reed wrote:
> On 9/06/2014 3:01 AM, Thor Lancelot Simon wrote:
> > On Mon, Jun 09, 2014 at 02:45:19AM +1000, Darren Reed wrote:
> >>> In Sun, Jun 08, 2014 at 10:56:50PM +1000, Darren Reed wrote:
> >>>>
> >>> ...
> >>>>         inet6 fe80::203:baff:fe34:a1f5%cas0 prefixlen 64 scopeid 0x6
> >>>> vlan0: flags=8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> mtu 1500
> >>>>         vlan: 200 parent: cas0
> >>>>         address: 00:03:ba:34:a1:f5
> >>>> ...
> >>>> vlan1: flags=8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> mtu 1500
> >>>>         vlan: 201 parent: cas0
> >>>>         address: 00:03:ba:34:a1:f5
> >>>
> >>>> I've got three interfaces with the same MAC address!
> >>>
> >>> This is correct.  It is an error to connect two vlan interfaces on the
> >>> same underlying physical network to the same layer 2 network.  So the
> >>> MAC address being the same can't cause problems.
> >>> ...
> >>> Note that the
> >>> pathological case where you configure two vlan interfaces on the same
> >>> physical interface with the _same_ vlanif, simulating dual-attach to the
> >>> same physical LAN in the SunOS 4 case (which is where that was 
> >>> problematic)
> >>> is also insane -- it's easy enough to work out why.
> >>
> >> I think you're wrong here.
> >>
> >> For example, what if I were to create two chroot environments on my
> >> NetBSD box and I wanted to use a dedicated NIC and IP address for each?
> >> And if I want each NIC to be its own vlan interface?
> >>
> >> Or what if I want to do virtual networking inside of NetBSD and create
> >> a vwire between two vlan interfaces?
> >>
> >> Or connect both vlan interfaces to a virtual switch inside the kernel?
> > 
> > Why do any of these require stacking two instances of vlan, with the same
> > vlanif, on the same physical interface?
> 
> In the first instance, because NetBSD vlan interfaces that have a vlanid
> cannot exist without a hardware interface underneath them. So if you only
> have one NIC, say bge0, then every vlan interface must have that under it.

Not so.  We have virtual physical interfaces, too.

> To put this another way, if I can plug bge0 and bge1 into the same LAN
> and use them in whichever way, why can't I plug vlan0 and vlan1 both
> into the same VLAN?

Because if they're on the same physical interface, the semantics are unclear
at best; and implementing this on some physical hardware (we do not have the
luxury of pretending every interface in the world is a wm or bge) is difficult
and seems likely to perform poorly.  There are other ways to build equivalent
configurations; why eat the hair instead of a hamburger?

Thor