[ndnSIM] PIT size
alexander.afanasyev at ucla.edu
Thu Mar 13 16:38:11 PDT 2014
An excessive DropData could be a result of the way you set up the limit. You didn't show the detail, but if you have used "Lru" policy for the limit, then this would be the core problem---whenever limit is reached, you would evict an older entry, instead of keeping it and rejecting new interest. I suspect that you will get more reasonable results if you change to "Persistent" policy (which is actually a default).
On Feb 26, 2014, at 5:43 AM, Hovaidi Ardestani Mohammad <mohammad.hovaidi.ardestani at aalto.fi> wrote:
> Hi all,
> I have a basic question about PIT size and its role to control the traffic flow.
> In my scenario when I change the PIT size into the lower value, the result shows more DropData.
> I reckon the small size of PIT means the small number of data requests and consequently the small number of DropData but large number of InterestData. Is It so?
> On the other hand, when PIT overflows, we would expect an increase rate of retransmissons from the downstream nodes. What would be the role of increasing retransmission rate on the packet drops both on layer 2 and 3.
> I really appreciate your response in advance!
> ndnSIM mailing list
> ndnSIM at lists.cs.ucla.edu
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the ndnSIM