[ndnSIM] segmentation

[Alex Afanasyev]

Hi Masri,

If I understand correctly, you want to do some kind of "prefetching"/Data "pre-publishing".  What you can do is to enable unrestrictive caching from local applications using CacheUnsolicitedDataFromApps attribute.

Config::SetDefault("ns3::ndn::ForwardingStrategy::CacheUnsolicitedDataFromApps", BooleanValue(true));

After that, you can simply "publish" any number of data packets into local node's cache as a response to the incoming Interests; and any further Interests for the published Data will be answered directly from the cache.

Would that work for you?

---
Alex

On Nov 17, 2013, at 4:45 PM, nor masri sahri <masri.sahri at gmail.com> wrote:

> hi alex, thank you for the clarification. 
> need idea from sifu Alex ;)
> 1. i plan to break packets into segment. 
> default size of payload is 1024 and i plan to break into 4 packets BUT named in one segment. For example, /prefix 0 to /prefix 4 is named under Segment 1. When Consumer send FIRST Interest , it will fetch the segment name (in this case Segment 1, and Producer will reply packet /prefix 0 to /prefix 4, 256 bytes each). And the following Interest will fetch Segment 2 and so on.
> Can you give some advice on how to do this, your help is really appreciated.
> 
> 
> 
> On Fri, Nov 15, 2013 at 1:41 AM, Alex Afanasyev <alexander.afanasyev at ucla.edu> wrote:
> Hi Masri,
> 
> You're seeing the correct behavior.  The consumer application explicitly requests (as it is done in majority cases in NDN) specific Data packets, which can be interpreted as a part of a large content.
> 
> Another small comment.  None of the implemented consumer applications in ndnSIM wait for the arrival of Data before sending another Interest.  Instead, the number of sent interest is controlled either by rate (ConsumerCbr) or "outstanding Interest window" (ConsumerWindow).
> 
> ---
> Alex
> 
> On Nov 12, 2013, at 9:54 PM, nor masri sahri <masri.sahri at gmail.com> wrote:
> 
>> hi alex and all,
>> correct me if i am wrong about my interpretation.
>> from the example file (ndn-simple.cc), i just did simple rule to set maximum number of prefix to be requested (i set to max = 200) to see how it works.(assuming the size of /prefix name file is consist of 200 segment)
>> 
>> 1. consumer send interest for data name /prefix
>> 2. Producer will reply /prefix 0 (part or chunk or segment of /prefix named file)
>> 3. when consumer receive /prefix 0, consumer will send another interest for the same file name (/prefix) 
>> 4. producer will reply /prefix 1 since that is the sequence number for the complete named data of /prefix
>> 
>> im quite confused when i see the ns log file
>> 
>> 0s 0 ndn.Consumer:SendPacket()
>> 0s 0 ndn.Consumer:SendPacket(): [INFO ] > Interest for 0
>> 0s 0 ndn.Consumer:WillSendOutInterest(): [DEBUG] Trying to add 0 with +0.0ns. already 0 items
>> 0.020496s 2 ndn.Producer:OnInterest(0x8c01308, 0x8c0db70)
>> 0.020496s 2 ndn.Producer:OnInterest(): [INFO ] node(2) respodning with Data: /prefix/%00
>> 0.057472s 0 ndn.Consumer:OnData(0x8c00d38, 0x8c0d260)
>> 0.057472s 0 ndn.Consumer:OnData(): [INFO ] < DATA for 0
>> 0.1s 0 ndn.Consumer:SendPacket()
>> 0.1s 0 ndn.Consumer:SendPacket(): [INFO ] > Interest for 1
>> 0.1s 0 ndn.Consumer:WillSendOutInterest(): [DEBUG] Trying to add 1 with +100000000.0ns. already 0 items
>> 0.120512s 2 ndn.Producer:OnInterest(0x8c01308, 0x8c0f7f0)
>> 0.120512s 2 ndn.Producer:OnInterest(): [INFO ] node(2) respodning with Data: /prefix/%00%01
>> 0.157504s 0 ndn.Consumer:OnData(0x8c00d38, 0x8c0f5b0)
>> 0.157504s 0 ndn.Consumer:OnData(): [INFO ] < DATA for 1
>> 
>> Is Consumer request specifically for /prefix 0 at the first time and when he get the data, the consumer send another interest for next data which is /prefix 1?
>> 
>> In this scenario, it is presumed that the size of /prefix is max 200 segment.
>> 
>> 
>> Sorry about my question, little confused. 
>> 
>> I am planning to do research on window on NDNsim. Need advice. Many thanks
>> 
>> 
>> 
>> -- 
>> Regards,
>> 
>> masri.sahri
>> 
>> _______________________________________________
>> ndnSIM mailing list
>> ndnSIM at lists.cs.ucla.edu
>> http://www.lists.cs.ucla.edu/mailman/listinfo/ndnsim
> 
> 
> 
> 
> -- 
> Regards,
> 
> masri.sahri
> 
> _______________________________________________
> ndnSIM mailing list
> 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/20131118/bda8e404/attachment.html>