RE: [dhcwg] Computation of T1, T2, and lease expiration time

As usual, I missed that section of text.

Actually, I was referring to this sentence in paragraph two:

	The client records its own local time for later use in 
	computing the lease expiration.  The client then broadcasts 
	the DHCPDISCOVER...

If you consider both of these paragraphs, it appears that the
timers should be calculated from the DHCPDISCOVER.

I would be in favor of including this for clarification in 
a future version of draft-ietf-dhc-implementation-01.txt.

--kan--

> -----Original Message-----
> From: Bernie Volz [mailto:volz@cisco.com]
> Sent: Monday, 05 April, 2004 4:49 PM
> To: 'Kevin A. Noll'; dhcwg@ietf.org
> Subject: RE: [dhcwg] Computation of T1, T2, and lease expiration time
> 
> 
> I think the text you refer to is:
> 
>    The client records the lease expiration time as the sum of
>    the time at which the original request was sent and the duration of
>    the lease from the DHCPACK message.
> 
> And, the question is what does "original request" refer to - the
> DISCOVER or REQUEST?
> 
> To be completely safe, using the DISCOVER would be best (since that will
> result in the shortest lease / T1 / T2 times for the client). But, I
> think practially speaking that REQUEST is OK.
> 
> But, any client and server need to be prepared for some grace period for
> leases as there is no requirement that the client and server have
> synchronized clocks and their individual clocks may run fast or slow.
> 
> Perhaps this would be item to include in a future revision of
> draft-ietf-dhc-implementation-01.txt?
> 
> - Bernie
> 
> -----Original Message-----
> From: dhcwg-admin@ietf.org [mailto:dhcwg-admin@ietf.org] On Behalf Of
> Kevin A. Noll
> Sent: Monday, April 05, 2004 3:01 PM
> To: dhcwg@ietf.org
> Subject: [dhcwg] Computation of T1, T2, and lease expiration time
> 
> 
> 
> Paragraph 2 of Section 4.4.1 of RFC2131 implies that T1, 
> T2, and the lease expiration time are all computed relative 
> to the time that the client sends the DHCPDISCOVER.
> 
> Is this correct, or should these values be computed relative 
> to the time that the client acquires the lease (i.e. when it receives
> the DHCPACK)?
> 
> In a perfect world, the former would make sense because the time to get
> from DISCOVER to ACK should be just a couple of 
> seconds at most. 
> 
> In a not so perfect world, the latter would make sense. For example, if
> the network or DHCP server is not performing very well, it may take many
> seconds and several retries to actually get from DISCOVER to ACK. If
> this happens and the DISCOVER time 
> is the reference, then there could be a considerable discrepency 
> between the client's and server's notion of when the lease 
> should expire.
> 
> Am I correct, or did I miss another section of the document that
> clarifies this?
> 
> --kan--
> --
> Kevin A. Noll, KD4WOZ
> CCIE 10948, CCDP
> Perfect Order, Inc.		
> Kevin.Noll@perfectorder.com
> 717-796-1936
> 
> 
> _______________________________________________
> 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