Re: [ippm] FW: http://tools.ietf.org/wg/ippm/draft-morton-ippm-delay-var-as-04.txt

["Romascanu, Dan (Dan)" <[e-mail]>]

Benoit,
 
Let me say that I am finding this discussion to happen a little bit too
early, but I was answering to the challenge coming from Henk. Usually
such discussions happen when the document became a work item, or even at
WGLC. 
 
Section 5 includes A LOT of history which seems to me not relevant and
usually is not included in IETF documents. Maybe a short summary should
be part of Section 6 but not all the details. In section 6 I do not see
the point of 6.5 and 6.6, and have doubts about 6.8, but maybe I should
read this one again to make sure that I understand the issues. 
 
The issue with section 8 is that it goes into some details about
measurement methods and devices, rather than metrics. It would be
interested to hear the opinion of other IPPM folks. 
 
Dan
 
 
 


________________________________

	From: Benoit Claise [mailto:[e-mail]] 
	Sent: Tuesday, February 12, 2008 2:05 AM
	To: Romascanu, Dan (Dan)
	Cc: Henk Uijterwaal; IETF IPPM WG
	Subject: Re: [ippm] FW:
http://tools.ietf.org/wg/ippm/draft-morton-ippm-delay-var-as-04.txt
	
	
	Dan,
	
	While I "could" understand your point about section 5, I don't
understand the arguments about removing section 6.
	I thought it gave some useful comparison between IPDV and PDV
when looking at parameters packet loss, path changes, clocks,
composition.
	- "In conclusion, the PDV results are affected by the packet
loss ratio.
	   The IPDV results are affected by both the packet loss ratio
and the
	   packet loss distribution.  In the extreme case of loss of
every other
	   packet, IPDV doesn't provide any results." 
	   is an important conclusion (just to reference one) for the
selection of the DV method
	- For example, do we NTP for one-way delay and delay variation
is a typical question I receive from customers.
	- Same thing for "how do I combine DV?"
	
	Again for section 8, I have to explain to customers/partners
	- "You want Poisson because it's random. But it's sometimes
better to get a random start with fixed intervals..."
	- "With PDV, it is sufficient to specify the upper percentile
(e.g.,  99.9%), while it's different with IPDV"
	- How long should I measure etc...
	
	Considering that this draft is there as an entry point for delay
variation explanations + a series of guidelines (at least this was my
intention) I spent quite some time on these sections. Specifically, if
section 5 is removed, we need to list somewhere the conclusions of the
published papers referenced in section 5. The place would then be the
current section 6. Unless you have a different view?
	
	Regards, Benoit.
	
	

		IMO section 5 and most of section 6 should probably go
away. I am not
		sure that section 8 is within the 'traditional' scope of
IPPM. 
		
		Dan
		
		
		 
		 
		
		  

			-----Original Message-----
			From: Henk Uijterwaal [mailto:[e-mail]] 
			Sent: Monday, February 11, 2008 9:23 PM
			To: Romascanu, Dan (Dan)
			Cc: IETF IPPM WG
			Subject: Re: [ippm] FW: Draft 
	
http://tools.ietf.org/wg/ippm/draft-morton-ippm-delay-var-as-04.txt
			
			Romascanu, Dan (Dan) wrote:
			    

				 If the question that is being asked is
'do you believe that the 
				document' is in good shape to become an
IPPM WG work 
				      

			item?', my answer 
			    

				is 'yes'.
				      

			We already asked that one, the next question is:
do you agree 
			with the contents, what should be
added/changed/removed?
			
			Henk
			
			
			    

				Dan
				
				
				
				 
				
				-----Original Message-----
				From: Henk Uijterwaal
[mailto:[e-mail]]
				Sent: Monday, February 11, 2008 12:26 PM
				Subject: Draft
	
http://tools.ietf.org/wg/ippm/draft-morton-ippm-delay-var-as-04.txt
				
				Hi
				
				Last year, you expressed interest in
				
				   
				      

	
http://tools.ietf.org/wg/ippm/draft-morton-ippm-delay-var-as-04.txt
			    

				and supported it as an IPPM WG document.
Did you read the latest 
				version of the document and if so, can
you please post 
				      

			comments to the 
			    

				[e-mail] list.  Even a "yes, I've
read it and it is fine" is 
				already very helpful.
				
				Thanks!
				
				Henk
	
_______________________________________________
				ippm mailing list
				[e-mail]
	
http://www.ietf.org/mailman/listinfo/ippm
				
				      

			--
	
--------------------------------------------------------------
			----------------
			Henk Uijterwaal                           Email:

			henk.uijterwaal(at)ripe.net
			RIPE Network Coordination Centre          
			http://www.amsterdamned.org/~henk
			P.O.Box 10096          Singel 258         Phone:
+31.20.5354414
			1001 EB Amsterdam      1016 AB Amsterdam  Fax:
+31.20.5354445
			The Netherlands        The Netherlands
Mobile: [phone]
	
--------------------------------------------------------------
			----------------
			
			Is one of the choices leaving the office open?
			                                       Alan
Greenspan on the 
			next elections
			
			This email was protected during delivery to
Avaya with TLS encryption
			
			
			    

		_______________________________________________
		ippm mailing list
		[e-mail]
		http://www.ietf.org/mailman/listinfo/ippm
		  

_______________________________________________
ippm mailing list
[e-mail]
http://www.ietf.org/mailman/listinfo/ippm