[ndnSIM] How to choose the nearest server with App Instances

Xiaoke Jiang shock.jiang at gmail.com
Wed Jul 3 17:28:13 PDT 2013


Good Idea. In fact I think there should be a centralized controller, say DNS to help consumers to select the best surrogates. but currently I just make the every producer know the topology from FIB and choose the nearest or second nearest (if the request to the nearest server is timeout).

A Question: are you interested in CDN?  

thanks

My Regards,
Xiaoke (Shock) Jiang (蒋小可)

Ph.D. Candidate,
Department of Computer Science and Technology,
Tsinghua University, Beijing, P.R.China. 100084




On Thursday, July 4, 2013 at 8:23 AM, Alex Afanasyev wrote:

> Oh. Sorry, I completely misread your question, didn't notice that you implementing IP/CDN.
>  
> For the real CDNs you can try to approximate DNS-based style selection, when clients are getting list of "close by" nodes based on geographic approximation.  Within the list, clients send request to any of them (all of them are "close").
>  
> Another approach, that is used in many overlay peer-to-peer IP networks, is to do some path evaluations (in advance?) to other peers.  
>  
> Hope it helps,
> Alex
>  
> On Jul 3, 2013, at 5:17 PM, Xiaoke Jiang <shock.jiang at gmail.com (mailto:shock.jiang at gmail.com)> wrote:
> > Hi Alex, you are right that NDN consumers do not need to select a producer. But for IP consumers, they have to.
> >  
> > thanks
> >  
> > My Regards,
> > Xiaoke (Shock) Jiang (蒋小可)
> >  
> > Ph.D. Candidate,
> > Department of Computer Science and Technology,
> > Tsinghua University, Beijing, P.R.China. 100084
> >  
> >  
> >  
> >  
> > On Thursday, July 4, 2013 at 1:22 AM, Alex Afanasyev wrote:
> >  
> > > Hi Xiaoke,
> > >  
> > > I feel that in NDN, choice of a specific producer **should not** be consumer's task.  It should NDN network job (e.g., forwarding strategy's job on a node) to figure out (based on signals from the consumers) that interest should be forwarded somewhere else.   At the same time, consumer will just request data by name, without caring what is happening behind the scenes.
> > >  
> > > ---
> > > Alex
> > >  
> > > On Jul 3, 2013, at 8:07 AM, Xiaoke Jiang <shock.jiang at gmail.com (mailto:shock.jiang at gmail.com)> wrote:
> > > > Hi Alex,
> > > >     I think I got some hints from example: triangle routes. So I will try myself to figure it out. But thank you all the same.
> > > >  
> > > >  
> > > > thanks
> > > >  
> > > > My Regards,
> > > > Xiaoke (Shock) Jiang (蒋小可)
> > > >  
> > > > Ph.D. Candidate,
> > > > Department of Computer Science and Technology,
> > > > Tsinghua University, Beijing, P.R.China. 100084
> > > >  
> > > >  
> > > >  
> > > >  
> > > > On Wednesday, July 3, 2013 at 10:09 PM, Xiaoke Jiang wrote:
> > > >  
> > > > > Hi Alex,  
> > > > > You know I am simulating CDN scenario now. One of the problem is for the consumer for choose the best surrogate after request is timeout. But now I am not confusing on how to implement it.
> > > > >  
> > > > > I created a class named CDNIPConsumer, which contains producerList and currentProducer. Timeout request triggers the rechoose another producer from producerList. Would you give me some hint? How to pick nearest/second nearest producer?
> > > > >  
> > > > > Thank you !  
> > > > >  
> > > > > thanks
> > > > >  
> > > > > My Regards,
> > > > > Xiaoke (Shock) Jiang (蒋小可)
> > > > >  
> > > > > Ph.D. Candidate,
> > > > > Department of Computer Science and Technology,
> > > > > Tsinghua University, Beijing, P.R.China. 100084
> > > > >  
> > > > >  
> > > > >  
> > > > >  
> > > >  
> > > >  
> > > > _______________________________________________
> > > > ndnSIM mailing list
> > > > ndnSIM at lists.cs.ucla.edu (mailto:ndnSIM at lists.cs.ucla.edu)
> > > > http://www.lists.cs.ucla.edu/mailman/listinfo/ndnsim
> > >  
> >  
> > _______________________________________________
> > ndnSIM mailing list
> > ndnSIM at lists.cs.ucla.edu (mailto:ndnSIM at lists.cs.ucla.edu)
> > http://www.lists.cs.ucla.edu/mailman/listinfo/ndnsim
>  

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.lists.cs.ucla.edu/pipermail/ndnsim/attachments/20130704/817e1fe7/attachment.html>


More information about the ndnSIM mailing list