[Click] Click AODV

Daniel Havey dhavey at yahoo.com
Mon Aug 30 18:35:20 EDT 2010


Hi Bart,

Hope you had a good holiday ;^)
Thanxs for the tips.  I got the userlevel AODV working, however, it produces duplicate packets like this:

PING 10.0.100.7 (10.0.100.7) from 10.0.100.1 : 56(84) bytes of data.
64 bytes from 10.0.100.7: icmp_seq=1 ttl=64 time=26.8 ms
64 bytes from 10.0.100.7: icmp_seq=1 ttl=64 time=26.8 ms (DUP!)
64 bytes from 10.0.100.7: icmp_seq=1 ttl=64 time=27.8 ms (DUP!)
64 bytes from 10.0.100.7: icmp_seq=1 ttl=64 time=27.9 ms (DUP!)
64 bytes from 10.0.100.7: icmp_seq=1 ttl=63 time=28.8 ms (DUP!)
64 bytes from 10.0.100.7: icmp_seq=1 ttl=63 time=28.9 ms (DUP!)
64 bytes from 10.0.100.7: icmp_seq=1 ttl=63 time=29.8 ms (DUP!)
64 bytes from 10.0.100.7: icmp_seq=1 ttl=63 time=29.9 ms (DUP!)

The click documentation for ToHost says:
Patchless Installations

On patched installations, FromDevice intercepts packets before they are passed to packet sniffers like tcpdump. However, on patchless installations, packet sniffers receive incoming packets before FromDevice can intercept them. This means that in a configuration such as

  FromDevice(eth0) -> ToHost;

packet sniffers will see eth0's packets twice, once before FromDevice runs and once when the packets are handed back to Linux via ToHost.

My understanding is that this means the duplicate packets will disappear when I run AODV at the kernel level.  However in the aodv.click script there are the following two elements:
[dhavey at node18 aodvscripts]$ click aodv1.click
aodv1.click:22: unknown element class âToSimDeviceâ
aodv1.click:30: unknown element class âFromSimDeviceâ
aodv1.click:41: unknown element class âFromSimDeviceâ
aodv1.click:57: unknown element class âToSimDeviceâ

The click documentation for these elements indicates that they are userlevel elements.  The documentation seems to indicate that these will not work with kernel level click.

How can I get rid of the duplicate packets?

Thanxs ;^)
...Daniel



--- On Mon, 8/16/10, Bart Braem <bart.braem at ua.ac.be> wrote:

> From: Bart Braem <bart.braem at ua.ac.be>
> Subject: Re: [Click] Click AODV
> To: dhavey at yahoo.com
> Cc: click at amsterdam.lcs.mit.edu
> Date: Monday, August 16, 2010, 12:55 AM
> Hi,
> 
> I am just back from a holiday, but some quick pointers to
> get you started:
> Most probably I forgot to add the AODV directory to the
> configure script, have a look at the configure.in or
> configure file to check it.
> ToSimDump was a hand-crafted element I wrote to dump
> information from ns simulations, everything should work when
> you leave out that script.
> 
> best regards,
> Bart
> 
> On 10 Aug 2010, at 00:57, Daniel Havey wrote:
> 
> > Hi,
> > 
> > I am testdriving Bart's AODV implementation ;^) 
> I downloaded 1.7 with AODV from patsgit.  However, I am
> having a little trouble getting it started.  I compiled
> with 
> > 
> > ./configure --enable-aodv --enable-nsclick
> --disable-linuxmodule --enable-userlevel
> > 
> > It seems like all of the ns stuff should be in there,
> but I do not find any of the sim stuff in
> userlevel/elements.conf
> > 
> > What happened?  And what is ToSimDump?  The
> other stuff is in elements/ns but not ToSimDump
> > 
> > Thanxs ;^)
> > 
> > 
> > [dhavey at node0 patsgit-Click-AODV-0a7d06b]$
> ./userlevel/click aodvscripts/aodv.click
> > aodvscripts/aodv.click:10: unknown element class
> âToSimDumpâ
> > aodvscripts/aodv.click:22: unknown element class
> âToSimDeviceâ
> > aodvscripts/aodv.click:30: unknown element class
> âFromSimDeviceâ
> > aodvscripts/aodv.click:41: unknown element class
> âFromSimDeviceâ
> > aodvscripts/aodv.click:44: unknown element class
> âToSimDumpâ
> > aodvscripts/aodv.click:57: unknown element class
> âToSimDeviceâ
> > aodvscripts/aodv.click:139: unknown element class
> âToSimDumpâ
> > aodvscripts/aodv.click:185: unknown element class
> âToSimDumpâ
> > 
> > 
> > 
> > 
> > 
> > _______________________________________________
> > click mailing list
> > click at amsterdam.lcs.mit.edu
> > https://amsterdam.lcs.mit.edu/mailman/listinfo/click
> 
> 


      



More information about the click mailing list