[Olsr-users] OLSR routing issue with highly connected node

[Henning Rogge]

Hi,

after looking a lot through the Hello code I have to admit I don't see
anything unusual.. which doesn't mean much, the code is horrible to
read.

One last thing which might help... could you try to get a binary
tcpdump with the fragmented Hello? the 160+ node should send out two
Hellos directly after each other because it has to fragment it. The
fragmentation is done on OLSR level, not on UDP level.

This might help to see if the problem is in the generation or parsing code.

Henning

On Tue, Dec 23, 2014 at 8:24 PM, Jernej Kos <(spam-protected)> wrote:
> Hello!
>
> On 23. 12. 2014 20:20, Henning Rogge wrote:
>> and the txtinfo output only mentions a single link... but it should
>> have more, right?
>
> Yes, it should have one additional link to the VPN server that has
> around 160 neighbours. Also note that the routing table is flapping --
> sometimes the routes received via the VPN server with less neighbours
> are there and sometimes they all disappear. But routes via the other VPN
> server are never established when the MTU is low. As soon as the MTU is
> increased, everything is back to normal.
>
>
> Jernej
>