From: simon hart (simon.hart@btinternet.com)
Date: Tue Apr 26 2005 - 16:45:15 GMT-3
Hi Tim,
My understanding is that because 6to4 tunnels are by their nature automatic
then you do not explicitly define a destination address, this allows the
underlying IPv4 transport network to act as a psuedo NBMA network. Each
IPv6 network you create on the edge of the domain will have the IPv4 address
of its source tunnel embedded. I shall try and illustrate, using you're
example below
Let say RA has a S0 with an ip address of 192.168.99.1/24. Now this will be
the interface to the IPv4 domain.
Now we need to create an IPv6 Prefix for RA using this network address, the
rules state that this should be /48 address starting with 2002:: So now our
IPv6 prefix is 2002:c0a8:6301::/48 This prefix can be subnetted to create
subnets on the RA router, one for the tunnel and one for the network.
Now our IPv6 domain is on E0 of RA, so I shall assign the address
2002:c0a8:6301:1::1/64 to this interface. And to the tunnel interface I
shall assign the subnet 2002:c0a8:6301::1/64
So far so good..............
Now we have to conduct a similar exercise on RB.
RB has an S0 ipv4 address of 172.16.16.1/24. This is the interface to the
IPv4 domain.
Need to now create an IPv6 /48 prefix as above. This will result in
2002:ac10:1001::/48 From this prefix I shall create some IPv6 subnets one
for the tunnel interface and one for the E0 interface to RB's IPv6 domain.
E0 = 2002:ac10:1001:1::1/64
T0 = 2002:ac10:1001::/64
One last task to conduct on each router, I have to create a static route to
the Tunnel interfaces. On both RA and RB I shall create
ipv6 route 2002::/16 tunnel 0
This has the function that, if after performing a longest match the router
does not find a local 2002:: address it will send any packet destined to
2002:: to tunnel 0.
Now here comes the elegance of the solution!!!
A host on the subnet, lets say 2002:c0a8:6301:1::4/64 wants to send to a
host on 2002:ac10:1001:1::5/64. The IPv6 packet will be sent to the RA's E0
interface, RA will now look at its routing table and send the packet to
tunnel 0.
Tunnel 0 will recognise that it is configured for ipv6ip 6to4 and will look
at the IPv6 address of E0 and determine the source to be 192.168.199.1 it
will now look at the destination IPv6 address 2002:ac10:1001:1::5/64, it
will strip of 2002 read the next 4 bytes and determine the destination of
the tunnel is 172.16.16.1. Hey presto we have the source and destination
address of the tunnel.
Sorry for the long winded explanation, but I am not sure how to explain it
easier :(
HTH
Simon
-----Original Message-----
From: nobody@groupstudy.com [mailto:nobody@groupstudy.com]On Behalf Of
ccie2be
Sent: 26 April 2005 18:51
To: Group Study
Subject: ipv6 6to4 tunnels
Hi guys,
I need some help figuring out how this works.
Let's say there are 2 ipv6 sites separated by an ipv4 domain.
ipv6 -------- r-A --- ipv4 ----- r-B ----- ipv6
I want to configure a 6to 4 tunnel between R-A and R-B.
According to the Cisco doc's, I don't need to configure a tunnel destination
on either side of the tunnel.
See this link:
http://www.cisco.com/univercd/cc/td/doc/product/software/ios123/123cgcr/ipv6
_c/sa_tunv6.htm#wp1037465
Embedded in the ipv6 address of the tunnel interface is an ipv4 address
that's local to the router on which this tunnel is configured.
That being the case, how does the one end of the tunnel find the other end?
I assume in this example, R-A and R-B don't have to share a common link,
correct?
TIA, Tim
This archive was generated by hypermail 2.1.4 : Tue May 03 2005 - 07:55:09 GMT-3