[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: Very strange sim results
On Fri, 6 Nov 1998, Biaz Saad wrote:
> Are you using a lossy link ?
That's not apparent from his code (included again as text below for
anyone who can't read attachments, or is browsing the archives.)
I suspect Chris's problem is that he is opening the tracefile, but not
explicitly closing it, so stuff doesn't always get flushed out to it.
L.
-->8--
proc create_net_6_DT { } {
#
# s1 s3
# \ /
# 10Mb, 1ms \ 0.3Mb, 1ms / 0.1Mb, 300ms
# r1 ------------ r2
# 10Mb, 1ms / \ 0.1Mb, 300ms
# / \
# s2 s4 tcl
#
global ns s1 s2 s3 s4 r1 r2
set s1 [$ns node]
set s2 [$ns node]
set r1 [$ns node]
set r2 [$ns node]
set s3 [$ns node]
set s4 [$ns node]
$ns duplex-link $s1 $r1 10Mb 1ms DropTail
$ns duplex-link $s2 $r1 10Mb 1ms DropTail
$ns duplex-link $r1 $r2 0.3Mb 1ms DropTail
$ns duplex-link $r2 $s3 0.1Mb 300ms DropTail
$ns duplex-link $r2 $s4 0.1Mb 300ms DropTail
}
set ns [new Simulator]
create_net_6_DT
set stopTime 30
set tFile [open queue.tr w]
$ns trace-queue $r2 $s3 $tFile
set tcp0 [$ns create-connection TCP/Reno $s1 TCPSink $s3 40]
$tcp0 set window_ 100
set ftp0 [$tcp0 attach-source FTP]
set tcp1 [$ns create-connection TCP/Reno $s2 TCPSink $s4 41]
$tcp1 set window_ 100
set ftp1 [$tcp1 attach-source FTP]
# Enable the next line for different sim results !!!
# $ns at 25.0 "puts \"Very strange...\""
$ns at 0.0 "$ftp0 start"
$ns at 0.0 "$ftp1 start"
$ns at $stopTime "$ftp0 stop"
$ns at $stopTime "$ftp1 stop"
$ns at $stopTime "puts \"Simulation finished.\""
$ns at $stopTime "exit"
$ns run
-->8--
> If so, check that your error model is correctly initialized,
> specially in errmodel.cc if you are using an old version of ns-2.
>
> Saad.
>
>
> On Fri, 6 Nov 1998, Christof Brandauer wrote:
>
> > Dear ns-users!
> >
> > I've come across a (from my point of view) very strange behavior of ns
> > that I can't explain.
> >
> > I run a simple simulation (I attached the script) where I
> > "trace-queue" one link.
> >
> > >From the trace output "queue.tr" I count the # of dropped packets
> > doing a:
> >
> > % grep ^d queue.tr | wc -l
> >
> > which results in 0.
> >
> >
> > Then I run the sim again, but this time schedule one additional event:
> >
> > $ns at 25.0 "puts \"Very strange.\""
> >
> > I would have thought that this should not have any effect on the sim
> > itself. However I get a different "queue.tr", including 43 packet drop
> > events.
> >
> > I'd greatly appreciate it if anyone could clear things up for me.
> >
> > thx in advance,
> > chris
<[email protected]>PGP<http://www.ee.surrey.ac.uk/Personal/L.Wood/>