Re: icmpv6-v3 comment

[Pekka Savola <pekkas@xxxxxxxxxx>]

On Mon, 1 Mar 2004, Jyrki Soini wrote:
>     "An Echo Reply [SHOULD NOT| MAY] be sent in response to an Echo 
> Request message
>      sent to an IPv6 multicast or anycast address. If this potentially 
> dangerous reply
>      is sent, the source address of the reply MUST be a unicast address 
> belonging to
>     the interface on which the Echo Request message was received."

First, at this point, I'd leave anycast addresses as is.

Second, I think with multicast debugging, the behaviour has to be 
consistent -- either you MUST answer or you MUST NOT.  If you receive 
an answer from some nodes, but not from others, debugging will be 
difficult.

So, whatever we choose (I don't have really strong opinions), it 
should be a MUST or MUST NOT.

>  - echo reply to multicast packet only on link-local scope addresses

Or maybe only "global" or some other scope (e.g., "up to scope B") 
could be forbidden/rate-limited.

The problem with rate-limiting is that it requires the separate 
timers, which some will argue must be configurable, causing more 
complexity than it's worth..

-- 
Pekka Savola                 "You each name yourselves king, yet the
Netcore Oy                    kingdom bleeds."
Systems. Networks. Security. -- George R.R. Martin: A Clash of Kings


--------------------------------------------------------------------
IETF IPv6 working group mailing list
ipv6@ietf.org
Administrative Requests: https://www1.ietf.org/mailman/listinfo/ipv6
--------------------------------------------------------------------