Re: [DNA] Link Identification Scheme ComparisonPowerpointslides

[Ranjitsinh Wable <wable@samsung.com>]

Hi Sathya,

See the comments inlined

----- Original Message ----- 
From: "Sathya Narayanan" <sathya@research.panasonic.com>
To: "Ranjitsinh Wable" <wable@samsung.com>
Cc: <dna@eng.monash.edu.au>
Sent: Thursday, August 04, 2005 5:39 PM
Subject: Re: [DNA] Link Identification Scheme ComparisonPowerpointslides


>> Comparison 9:    I think Landmark also depends on CPL on DNA-nonDNA
>> movement. There is no way that the Landmark can make decision on its own.
>
> As mentioned in the footnotes in the additional information section, it is 
> the CompleteRA component which helps with building of the CPL (see 
> Comparison 4), which makes it a 'Yes' for the Landmark/CompleteRA column.

Here  again I will say that the Complete RA will aid to create the Complete 
Prefix List Faster after that the
mechanism is same.

Lets say that host on the previous link have the complete prefix List
(How they get that list, mechanism may be different), On movement from DNA 
to non-DNA
they have to check the prefix list in the new RA to the complete prefix 
list.
So that way the mechanism remains same. There is no different case for both 
the
methods.

>> Comparison 10: Support for Mixed Links
>>
>> Whether you consider this case different from the Comparison 8.
> 8, 9, 10 and too some extent 4 are related - different aspects of a 
> underlying concept.
>
>> I guess its the case when the one link have both DNA and non-DNA Routers.
> Yes.
>
>> I think this the same problem with the Landmark if the new router comesup 
>> in the same link.
> I am not sure I understand this comment - new router coming up on a link 
> introduces some kinda instability to both link identification mechansim (I 
> guess it was refered to as flashing on the mailing list).
>
I think, I was too abstract to convey the point. Sorry for that I will try 
to put it in different way,

Supose in the mixed link host gets the RA. Link ID will consider the the 
Link change process
if it gets the L2 trigger bcoz it don't know whether its non-DNA node of 
same link or
RA from new Link.

Now lets consider what happens in the Landmark case. Now if the new RA 
advertise prefix which is
 not present in the Complete prefix List available with host. (Your bracket 
case from ppt).
Host don't know its new router on same link or new router on the next link.
So the requirement is agin same.
>>
>> Comparison 11: Support for the # of prefixes
>>
>> I think here again LinkId has advantage than Landmark. How does the 
>> Landmark have the advantage?
> That was a bad choice of words, what I meant was - only the Landmark 
> functionality will be available when there are large number of prefixes so 
> that they don't fit in a single RA message.
>
Yes, you may need to rephrase it.

Regards,

Wable R. U.

>> Also I think following points shall be added in the comparison list.
>>
>> 1) Synchronisation
>> 2) No of new options
>> 3) Support for neigborhood information discovery protocols like NED, CARD
>
> I don't think 1) and 2) are useful. And I don't know about 3). Lets 
> discuss it further.
>
> thanks for your comments,
> Sathya
>
>> ----- Original Message ----- 
>> From: "Sathya Narayanan" <sathya@research.panasonic.com>
>> To: "Greg Daley" <Greg.Daley@eng.monash.edu.au>
>> Cc: <dna@eng.monash.edu.au>
>> Sent: Thursday, August 04, 2005 2:35 PM
>> Subject: Re: [DNA] Link Identification Scheme Comparison
>> Powerpointslides
>>
>> > The latest comparison slides to be used during the meeting can
>> be found
>> > at:
>> >
>> > http://www.ctie.monash.edu.au/dna/DNASolnComparison_4Aug2005.ppt
>> >
>> > Please take a look at it and provide feedback.
>> >
>> > thanks,
>> > Sathya
>> >
>> > ----- Original Message -----
>> > From: Greg Daley <Greg.Daley@eng.monash.edu.au>
>> > Date: Friday, July 29, 2005 3:39 pm
>> > Subject: [DNA] Link Identification Scheme Comparison Powerpoint
>> slides>
>> >> Hi,
>> >>
>> >> Here is a comparison of the different Link Identification features
>> >> available in draft-pentland-dna-protocol and draft-jinchoi-dna-
>> >> protocol2
>> >>
>> >> It's aimed at providing a feature distinction, although there are
>> >> also a couple of outstanding issues to be considered separately.
>> >>
>> >> http://www.ctie.monash.edu.au/dna/DNASolnComparison_28July2005.ppt
>> >>
>> >> Greg
>> >>
>> >
>> >
>> >
>>
>>
>
>
>