Problem with DHCPLEASEQUERY

[Niall O'Reilly]

On 08/06/10 23:51, David W. Hankins wrote:
> On Wed, Jun 02, 2010 at 02:32:28PM +0100, Niall O'Reilly wrote:
>> Is there something I've overlooked, or is this a bug?
>
> I don't actually know offhand.
>
> I guess the reason it would be unusual to direct the reply to the
> source port number, is because we use the 'giaddr' to direct the
> reply, and there's no port information next to giaddr.  It would be
> more understandable to use the source IP and port if we were using
> the source IP to direct the reply.
>
> So I suspect the reply directed to the 'bootps' port is the expected/
> standard behaviour.

	Thanks for this reply, David.

	In the meantime, I've done some background reading, with the
	following results.

	This is not a bug. I was missing something.

	RFC2131 requires the response to be sent to giaddr#bootps.
	This wasn't changed by RFC4388, as it made no difference to
	the use case(s) considered when the latter RFC was written.
	This is unfortunate for the use case I now have in mind.

	I'm thinking of how best to work around this.  Kim Kinnear
	and Bernie Volz have made helpful suggestions.


	Best regards,
	Niall O'Reilly