[Olsr-dev] Freifunk Testing

[Henning Rogge]

Am Montag 16 Juni 2008 23:29:46 schrieb Daniel Nitzpon:
> hi!
>
> i don't grasp the scenarios for fisheye routing loop production:
>
> Henning Rogge schrieb:
>  > The problem can happen at the edge of a fisheye zone. You get an
>  > "update border" at this point because some packages are not flooded to
>  > the edge of the network.
>
> why is that? can you give an example topology?

let's just look at a very extreme example.

we have a linear string of nodes (a total of 6 nodes) and a long connection 
from node 1 to 6 through other parts of the network.

1 -- 2 -- 3 -- 4 -- 5 -- 6
|                        |
+.....(many nodes).......+

Node 1 and 2 loose connection, so node 2 sends a new TC with the changed 
topology information. Unfortunately fisheye sends this TC with a TTL of 3, so 
only nodes 2, 3 and 4 get it.

Node 1 sends a package to 6, because it does not know about the lost 
connection between 5 and 6, it sends the package to node 2. Node two knows 
about the lost connection and about the other way between 1 and 6, so it 
sends the package back to 1.

=> Routing loop

Henning


*************************************************
Diplom Informatiker Henning Rogge
Forschungsgesellschaft für
Angewandte Naturwissenschaften e. V. (FGAN) 
Neuenahrer Str. 20, 53343 Wachtberg, Germany
Tel.: 0049 (0)228 9435-961
Fax: 0049 (0)228 9435-685
E-Mail: (spam-protected)
Web: www.fgan.de
************************************************
Sitz der Gesellschaft: Bonn
Registergericht: Amtsgericht Bonn VR 2530
Vorstand: Dr. rer. nat. Ralf Dornhaus (Vors.), Prof. Dr. Joachim Ender 
(Stellv.)