Re: Multicast Packet Counter

[Lori Flynn <lori@cse.ucsc.edu>]

Abdelhamid,

Thanks very much for the help, it's helpful to know there's a way
(trace-topo) to have packet counts greater than 0 come out of
McastMonitor.  

-Lori

--------------------------------------------------

On Mon, 24 May 1999, Joumdane Abdelhamid wrote:

> Hi Lori,
> 
> I tested the McastMonitor in ns2.1b4 with the example file ~ns/tcl/ex/newmacast/pktintran.tcl. It works fine. 
> 
> When I tried with your script. Yes, it gives 0 packet all the time. So I tried trace-topo, and I use the node Id of $n0 :
> set mcastmonitor [$ns McastMonitor]
> $mcastmonitor set period_ 0.01
> $mcastmonitor trace-topo [$n0 id] [expr $group]
> 
> And it's working fine.
> 
> I tried trace-tree on the example file pktintran.tcl. It gave me the same result : Packet count = 0.
> 
> I also tried to see if trace-tree in ns2.1b5. This function has desappered. The class McastMonitor has chaged a bit.
> 
> Hope it helps,
> 
> Abdelhamid JOUMDANE
> 
> 
> --
> 
> On Thu, 20 May 1999 16:25:26   Lori Flynn wrote:
> >Hi,
> >
> >I'm trying to use McastMonitor and having trouble. I set it up like this:
> >set mcastmonitor [$ns McastMonitor]
> >$mcastmonitor set period_ 0.01
> >$mcastmonitor trace-tree $n0 $group
> >
> >but it doesn't print out counts of packets for the fields (data, prune,
> >join, etc.).  Instead, everything is 0, incorrectly.  (even the nam
> >simulator shows packets flying, plus the trace-all shows packets as well.)
> >
> >You wrote to the ns2 newsgroup about the McastMonitor.  I have tried to
> >use it, using 'trace-tree' mode, and it does indeed print out a trace for
> >every time period as the simulation runs... however, the counts of packets
> >are 0 for each and every field (data, prune, etc.).  Even though true
> >counts are above 0.
> >
> >Here's part of what I get:
> >1.2200000000000009 0 0 0 0 _o10 0x8000
> >1.2300000000000009 0 0 0 0 _o10 0x8000
> >1.2400000000000009 0 0 0 0 _o10 0x8000
> >
> >The _o10 is not what I would expect the source to be (I have defined the
> >source as $n0, which is a UDP CBR sender, if that makes a difference).
> >I have attached my code (which is *very* short) in case you have any
> >questions about it.
> >
> >
> >Thanks for any advice!
> >Lori
> >