[Nfd-dev] On the necessity of the STRAGGLER timer

[Klaus Schneider]

On 11/16/2016 06:11 AM, Seweryn Dynerowicz wrote:
> Dear Junxiao,
>
> On 16 November 2016 at 11:28, Junxiao Shi <shijunxiao at email.arizona.edu
> <mailto:shijunxiao at email.arizona.edu>> wrote:
>
>     Straggler timer delays transferring Nonces from PIT out-records to
>     DNL, in order to keep DNL small.
>
>
> From their definitions, an Out-Record weighs 160 bits while a DNL entry
> occupies 64 bits so I assume
> you mean "small" in the sense of the number of entries. If the worry is
> about lookup time, then isn't
> it a matter of improving the data structure underlying the DNL ?

I think Junxiao refers to Section 4.2.8
Interest Finalize Pipeline: "The Dead Nonce List is a global data 
structure designed to detect looping Interests, and we want to insert as
few Nonces as possible to keep its size down."

Another reason I can think off, is that historically there was no DNL 
and thus the straggler timer was used to detect loops. So you might be 
right, that the straggler timer is less important in the current NFD 
version.

>
>     PIT entry is needed for Data measurements, because strategy needs to
>     know which Interest is being satisfied by the incoming Data (recall
>     that Interest name is a prefix of Data name and thus may be shorter
>     than the Data name), and when that Interest was forwarded (so as to
>     calculate RTT). Neither information is available in NameTree.
>
>
> Why does the Strategy need to know that ? The basic behaviour of ICN for
> forwarding back DATA is the breadcrumbs, so you only need to send to the
> IN-Faces
> for the matching PIT entries. How does this differ in NDN ?
>
> Regarding the RTT computation, when incoming Data is received, the
> IN-Records (and therefore the specific Interest packets being satisfied)
> are removed.

Are they?

> So you only compute the RTT for the first Data packet which satisfies
> the Interest, right ? If you're only interested in RTT, the OUT-Records
> do not enable you
> to do anything useful regarding that so they could be removed
> straightaway without waiting for STRAGGLER.

I don't know about the specifics, but in principle the straggler timer 
should help you to do measurements of more then the fastest interface. 
Whenever you send out Interests on multiple faces (either in parallel or 
sequential) it can be useful to wait a little longer for responses (Data 
or NACKs).

In addition to maintaining different RTT estimates, you could store 
information about a probed path being up (receiving another Data) or 
down (receiving a NACK), packet loss estimates, or other information.

I've used the functionality enabled by the straggler timer extensively 
in my previous work: 
http://conferences2.sigcomm.org/acm-icn/2015/proceedings/p137-schneider.pdf

Best regards,
Klaus


>
>
> Best regards,
>
> S.
>
> +----------------------------------------------+
> | Dynerowicz Seweryn                           |
> | PostDoc Researcher                           |
> | SITI, COPELABS, Building U                   |
> | Universidade Lusófona                        |
> | Campo Grande, 388, 1749-024 Lisboa, Portugal |
> | Mobile: +351 913 930 302 <tel:%2B351%20913%20930%20302>
>        |
> +----------------------------------------------+
>
> I hate the empty set; he's so full of himself.
>
> "Judge a man by his questions rather than his answers",
> Pierre-Marc Gaston, Duc de Lévis
>
> "Ignorance more frequently begets confidence than does knowledge.",
> C. Darwin
>
> "Seek freedom and become captive of your desires. Seek discipline
> and find your liberty.", F. Herbert
>
>
>
> _______________________________________________
> Nfd-dev mailing list
> Nfd-dev at lists.cs.ucla.edu
> http://www.lists.cs.ucla.edu/mailman/listinfo/nfd-dev
>