[Nfd-dev] NDN traffic generator

[NAREN N]

Hello,
Yes I believe that is the case. So what would be a good way to get a sense
of NFD performance?? Basically, we are trying to stress test the NFD and
see when it will start dropping interests. So should I set the freshness
period high and the interest lifetime to be low and increase the interest
rates?

-Naren

On Tue, Dec 9, 2014 at 11:18 AM, Lan Wang (lanwang) <lanwang at memphis.edu>
wrote:

>
>
>
>  On Dec 9, 2014, at 12:57 PM, NAREN N <nnaren1902 at gmail.com>
>  wrote:
>
>  Hello everyone,
> I was running some experiments for a course project and was using NDN
> traffic generator to measure of the performance and how each factor
> contributes to the performance of the NFD.
> However, I noticed something weird during the experiments.
>
>  When having the interestlifetime=100ms and the freshness period=100ms
> same between the runs, and also the contents of the data packet being the
> same, I ran the traffic generator for different burst rates. Here are the
> rates
>
>  >1 interest / second -----> RTT=71ms and InterstLoss = 20%
> >10 interests/second------> RTT=27ms and interestLoss = 1%
> > 100 interests/second -------> RTT=13ms and interestLoss = 0%
> >1000 interests/second ------>RTT=22ms and interestLoss=0%
>
>  So I was wondering why would the roundtrip time decrease with increase
> in Interest rate?? I have a feeling this has something do with the traffic
> generator logger code, which isn't reporting the RTT properly...
>
>  Also I see a interest loss when the Interest rate is lower, which is
> completely weird. I was expecting the interest loss to be higher when there
> was a burst of interests but this is opposite to what I was expecting.
>
>  Any help would be appreciated with this!
>
>
>  Have you set the option MustBeFresh to be true in the client's
> configuration file?  If so, your observation is due to your freshness
> period setting and interval between Interests.  If the sending rate is 1/s,
> then all of the Interests will have to fetch new data from the producer
> (because the data is expired when the new Interest comes).  The higher the
> sending rate, the more likely that the data in the local cache or an
> intermediate cache is fresh and the Interest does not have to go all the
> way to the producer.
>
>  Lan
>
>
>  regards,
> Naren
>  _______________________________________________
> Nfd-dev mailing list
> Nfd-dev at lists.cs.ucla.edu
> http://www.lists.cs.ucla.edu/mailman/listinfo/nfd-dev
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.lists.cs.ucla.edu/pipermail/nfd-dev/attachments/20141209/b576ba03/attachment.html>