On Wed, 2006-04-26 at 09:29 +0000, zarrar yousaf wrote:
> what is the reason for having a wireless link between the router R and
>  the HA as proposed in the MIPv6 HowTo test bed scenario.

There is no reason.  HA can be connected via wired link to R, and R
doesn't need the wireless link at all, just 2 wired interfaces.

> What if R and HA are connected through a cross-over ethernet cable? 
>  
> How significant is this and how will it effect the tests and operation
> of MIP?

Doesn't matter at all.  To simplify the test bed, you can have R and HA
in the same box.  Just make sure the routes match this too.

A realistic scenario might be something like this:

ISP--|R|--corporate network--|HA|--|AP|...|MN|

You have some heavy duty wired router connected to internet (meaning
ISP's network) and corporate network (I leave out firewalls and
everything else for simplicity).  If there are a lot of potential MNs,
HA might be a dedicated server in the machine room with two wired
interfaces, one connected to the corporate network and the other
connected to a wireless access point (bridge).  HA would advertise
itself on the wireless link provided by AP, which would be the home link
for the corporation's MNs.

In a home environment you don't need all that and your network could
look something like this:

ISP--|HA|...|MN|

Here the HA might be a combined ADSL router, wireless AP, and Home agent
(and firewall etc.).  You could run everything on e.g. OpenWRT supported
box.  Since in a home environment, there wouldn't be that many MNs, this
would be sufficient.

The point of all this is, that Mobile IPv6 may be configured the same in
both cases.  You would only have to change routes to match the topology.

Regards,

Antti



_______________________________________________
mipl mailing list
[email protected]
http://www.mobile-ipv6.org/cgi-bin/mailman/listinfo/mipl

Reply via email to