[Olsr-users] Question about hidden node behavior

[Markus Kittenberger]

On Thu, Jan 6, 2011 at 10:04 PM, Mitar <(spam-protected)> wrote:

>
> Yes, of course they are not 100 % hidden, as they are still close. But
> OLSR has never created a link between them yet. So it seems that there
> is not much traffic going on. But I will tcpdump a little and see for
> myself.
>
even if you can not tcpdump anything, they might still affect/disturb each
other easily,..

imho the easiest method to find out which link disturbs what, is to create
(unidirectional) test-traffic from one node to a neighbour, and monitor the
changes in other links,.. (and so on,..)

furthermore i don`t think that fri cannot receive an olsr packet from
glinska, because of trzaska is sending at this moment aswell,..
because if so, than trzaska would not be able to receive glinksas message
aswell (as it was sending instead of listening/receiving) (but as afaik
trzaska-glinksa is etx 1.0, it`s link sesning seems to work, so it does not
talk while glinksa is talking)

i guess fri cannot receive from glinska because it e.g. receives data from
the uplink (probably targeted to trzaska)
(if this theory is correct, rts/cts could improve it a bit)


> > furthermore what is the "uplink" of glinksa (another wireless link?)
>
> Only fri has uplink from those nodes.

ok, what kind of uplink:
also a wireless link?
  if so: same wireless interface?
  if not: same band?, or maybe even a neighbouring channel?

e.g. i just "found" 2 old wrt54gl (that share the same mast/box) configured
at channel 1 and 13 this week
one of them can not receive a single packet from its neighbours (normally
etx around 2), if the other one is sending continuosly,.. *g
(in this case i will have to exchange at least one of this routers,..)

>
> > furthermore verify if u have QOS priorizing olsrd traffic,..
>
> We do not do that. Is this a good idea?

priorize olsrd traffic is a very good idea!

even when priorizing olsrd packets the effects of traffic to link-quality
can be far bigger as u might want *G

> I thought it is good that when
> link is consumed a lot that LQ falls a bit, so that it is visible to
> OLSR that link is not as good as it was when it was free.
>

for changing linkcost/routing based on actual traffic, it needs other
measurement and statistcal methods as olsrd is doing for lq based on
packetloss,..

Markus
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.olsr.org/pipermail/olsr-users/attachments/20110107/6d366a85/attachment.html>