[OLSR-users] olsrd-0.5.0: LQ larger than 1.0

[Victor]

Hi,
I didn't uncomment LinkQualityWinSize, so it must be default value 10 for
all nodes in domain.

I do not see correlation between packet loss reported by application, and
packet loss reported by olsrd. I created congestion between 2 olsr nodes
causing 80% UDP packet loss ( reported by application). That doesn't seem to
change "Hello" packet loss reported by olsrd for that link. From another
side, I see "Hello" packet loss on the link where application ( and iPerf )
don't report any packet loss. Is there any way I can make olsrd packet loss
reporting more accurate?
  
Thanks,
-V

-----Original Message-----
From: (spam-protected) [mailto:(spam-protected)] On
Behalf Of Marco Tidow
Sent: Friday, May 11, 2007 11:52 AM
To: OLSR discussion and development
Subject: Re: [OLSR-users] olsrd-0.5.0: LQ larger than 1.0

On Fri, May.11. 07:59 -0700, Joerg Pommnitz wrote:
> Hello all,
> as already reported in
http://www.olsr.org/pipermail/olsr-users/2007-April/887695.html 
> the latest olsrd release reports LQ values larger than 1.0. 
> 
> Is this intentional or a bug? 
> If it is intentional: what does a LQ > 1.0 mean? 
> If this is a bug: any idea how to fix this?

Hi,

Did you check the setup for not using different HelloInterval´s a/o
LinkQualityWinSize in the same network segment / broadcast-domain?

Did see myself ETX < 1  due to LinkQualityMult >1, even w/ older olsr
versions,
that´s quiet "logical", useful in the sense of "intentionally prefer link as
long it
exists".

cu marco


_______________________________________________
olsr-users mailing list
(spam-protected)
https://www.olsr.org/mailman/listinfo/olsr-users