Pekka Savola wrote:

> Note: it seems to me that the scope of the resulting multicast address
> must not be greater than the generating unicast address scope.  This is
> because e.g. 'subnet ID + interface ID' in e.g. global scope could clash
> from the same values from other sites.

   Yes, you are right. Properly, our site-local multicast addresses are used
   when scop <= 5.


> GLOP-based allocation seems pointless when you can use unicast
> prefix-based allocations (useful only if you want to zero-configure so
> that every node can possibly start generating global multicast traffic
> without assignments from the _site_ owner).  Site- and link -local
> allocations would only seem to be usable within the respective site, or
> link.

   Regarding GLOP-based IPv6 multicast addresses, we received some comments
   from Dave Thaler. He proposed,
   GLOP-based IPv6 multicast addresses are much less powerful and has no
   advantages over Unicast-prefix-based addresses.
   So, we decided to follow his opinion.

   Instead, in the case of link- or site-local multicast addresses,
   we and Dave agreed the
   in order to make the syntax consistent with the Unicast-Prefix-based
   multicast draft, our 00 draft would be changed as follows :
  (we will publish revised version soon.)

   3.1 Link-local multicast addresses

     |   8    | 4  | 4  |     16     |       64       |       32      |
     +--------+----+----+------------+----------------+---------------+
     |11111111|flgs|scop|  reserved  |  Interface ID  |    group ID   |
     +--------+----+----+------------+----------------+---------------+

   3.2 Site-local multicast addresses

     |   8    | 4  | 4  |     16     |       64       |       32      |
     +--------+----+----+------------+----------------+---------------+
     |11111111|flgs|scop|     SLA    |  Interface ID  |    group ID   |
     +--------+----+----+------------+----------------+---------------+

   In addition, in stead of defining define a new flag,
   we will use same flag defined in the Unicast-Prefix-based draft.
   We and dave think we might be able to update the other draft to allow to
   do this in our draft.

   I think, our proposal can be used in link- or site-local,
   harmonizing with Unicast-Prefix-based multicast scheme
   (while Unicast-Prefix-based multicast addresses are for global.).

   Thanks,
   Myung-Ki.


--------------------------------------------------------------------
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]
--------------------------------------------------------------------

Reply via email to