On Jun 15, 2011, at 7:10 PM, Lorenzo Colitti wrote:
On Wed, Jun 15, 2011 at 3:58 PM, Keith Moore
<moore(_at_)network-heretics(_dot_)com> wrote:
That said, I would argue that most or all 6to4 traffic could just as well
use IPv4, since both parties to the communication obviously have access to
a public IPv4 address. What is the advantage of using 6to4 over IPv4 that
makes it worth suffering an 80% failure rate?
it can communicate with hosts that have only IPv6,
it can communicate with hosts that are stuck behind a single IPv4 address (if
the router acts as a 6to4 gateway) without a NAT being in the way,
it can be used to develop and test IPv6 applications without having to build
a special-purpose network,
it can be used to deploy applications now that already support IPv6 and so
are in some sense future-proofed,
it can be deployed on either a single host or a network
... about 80% of the time.
A single figure doesn't really describe the situation. The successes/failures
aren't independent of one another. It's not that you get only an 80%
probability of things working on every connection. It's more like, it either
works fairly well for what you need it to do, or it doesn't.
Again, don't assume that this is like the web and the connections are mostly
client-to-server, or one source to large numbers of different destinations.
I would argue that cases 1, 3, 4, 5, and 6 can be solved more reliably using
configured tunnels,
There are lots of different tunneling methods each with strengths and
weaknesses. What you probably do in practice is try one, and if that doesn't
work for your purposes, try another. 6to4 is very easy to try, and it often
works.
Keith
_______________________________________________
Ietf mailing list
Ietf(_at_)ietf(_dot_)org
https://www.ietf.org/mailman/listinfo/ietf