[Ndn-interest] Fib-entry modified

[Klaus Schneider]

Hi Ahmed,

This function sorts the fib entries by cost from lower to higher. This 
will then determine which default nexthop your forwarding strategy will 
use, which can explain your lower latency results. (for details: please 
attach your simulation and topology file)

It doesn't make much sense to change the sorting function, since it's 
cleaner and more intuitive to just change the link metrics.

Best regards,
Klaus


On 6/5/19 3:23 AM, Ahmed Malik via Ndn-interest wrote:
> Hello everyone,
> 
> I was trying to modify fib-entry.cpp and i ended up only modifying this
> void
>     77 Entry::sortNextHops()
>     78 {
>     79  std::sort(m_nextHops.begin(), m_nextHops.end(),
>     80  [] (const NextHop 
> <https://ndnsim.net/2.0/doxygen/classnfd_1_1fib_1_1NextHop.html>& a, 
> const NextHop 
> <https://ndnsim.net/2.0/doxygen/classnfd_1_1fib_1_1NextHop.html>& b) { 
> return a.getCost 
> <https://ndnsim.net/2.0/doxygen/classnfd_1_1fib_1_1NextHop.html#adc2961b0ddd341510bf4e49f93b4cf5b>() 
> < b.getCost(); });
>     81 }
> 
> by replacing the "<" by ">",  i used best route as a strategy and 
> changed the metric of the manual fib entry to a higher value (100), when 
> i ran this scenario and analyzed the trace file i found better RTT 
> results than with same fib entries with a metric equal to 1, the latency 
> is decreased by only changing this. Can anyone, please explain to me 
> what does this line do exactly ? and why is it producing better resulats 
> ? thank you in advance.
> 
> Kind regards.
> 
> _______________________________________________
> Ndn-interest mailing list
> Ndn-interest at lists.cs.ucla.edu
> http://www.lists.cs.ucla.edu/mailman/listinfo/ndn-interest
>