At 8:48 AM -0500 3/6/01, Ralph Droms wrote:
>In the case where L1 and L2 are site local addresses, I think B must
>be configured so that it doesn't forward site local addresses between
>its two interfaces (the behavior of B w.r.t. site local addresses wasn't
>specified in the original message). That is, A and the left-hand
>interface of B are part of one site while C and the right-hand
>interface of B are part of a different site.
That's correct. This is all covered in the aforementioned Scoped
Address Architecture draft.
>I'll ask for confirmation of this hypothesis, as thinking about site
>local addresses generally makes my head hurt.
If it helps your head at all, you can analogize this to what an IPv4
router has to do if it is attached to more than one "private" address
space (e.g., more than one hunk of real or virtual [as in VPN] topology
that reuses the same "net 10" address space). If that just makes your
head hurt worse, then forget I suggested it.
Steve
--------------------------------------------------------------------
IETF IPng Working Group Mailing List
IPng Home Page: http://playground.sun.com/ipng
FTP archive: ftp://playground.sun.com/pub/ipng
Direct all administrative requests to [EMAIL PROTECTED]
--------------------------------------------------------------------
