RE: RE: [dhcwg] dhc WG last call on draft-ietf-dhc-rapid-commit-opt-05

I agree with Bernie. It's true that sending the extra 'REQUEST' is 
inexpensive, but it's just another unreliable UDP packet. I don't think 
that the suggested extra step provides enough benefit to be worth including.

-- Mark

At 08:17 AM 7/14/2004 -0400, Bernie Volz wrote:
>This likely isn't as a cut and dry as you assume. Does the DHCPREQUEST get 
>retransmitted until it is ACKed? What happens if it isn't ACKed? What 
>happens if it is NAK?
>
>- Bernie
>-----Original Message-----
>From: dhcwg-bounces@ietf.org [mailto:dhcwg-bounces@ietf.org] On Behalf Of 
>Steve Gonczi
>Sent: Tuesday, July 13, 2004 7:35 PM
>To: soohong.park@samsung.com
>Cc: dhcwg@ietf.org
>Subject: RE: RE: [dhcwg] dhc WG last call on 
>draft-ietf-dhc-rapid-commit-opt-05
>
>Hi Daniel,
>
>I have to admit I am not entirely clear on what you meant,
>although I suspect you disagree with my suggestion.
>
>Let me try to make my case:
>
>For the cost of sending out one extra message by the Rapid Commit
>client (which BTW does not cause any significant delay)
>you could eliminate the entire 3.2 section from your protocol.
>
>To recap:
>
>1)  The "no multiple servers" restriction
>2)  or "must have enough addresses" restriction
>3)  or "use short lease times" restriction
>
>Why would you not want to do this?
>
>
>/sG
>-----Original Message-----
>From: PARK SOO HONG [mailto:soohong.park@samsung.com]
>Sent: Tuesday, July 13, 2004 5:54 PM
>To: Steve Gonczi; soohong.park@samsung.com
>Cc: Bernie Volz; dhcwg@ietf.org
>Subject: Re: RE: [dhcwg] dhc WG last call on 
>draft-ietf-dhc-rapid-commit-opt-05
>
>Hi Steve, thanks your comments and see my comment inline
>
>
>
> >When the Rapid commit client decides to accept an address
> >sent via a Rapid Commit ack, it should send out a DHCPREQUEST
> >(perhaps with the rapid commit option),
> >similar to the classic DHC protocol.
>
>
>
>I'd keep the original sentence since to sync the operation of DHCPv6,
>
>also make sure this option is for *RAPID COMMIT*.
>
>
>
>
>
>
>
>Regards
>
>
>
>Daniel (Soohong Daniel Park)
>
>Mobile Platform Laboratory. SAMSUNG Electronics
>
>_______________________________________________
>dhcwg mailing list
>dhcwg@ietf.org
>https://www1.ietf.org/mailman/listinfo/dhcwg

_______________________________________________
dhcwg mailing list
dhcwg@ietf.org
https://www1.ietf.org/mailman/listinfo/dhcwg