[Olsr-dev] reduce OLSR bandwidth by tuning emission intervals

Teco Boot (spam-protected)
Tue Feb 26 17:35:40 CET 2013


We had this discussions on large scale VPNs before. I suggested to not using OLSR on VPN tunnels. I use the tunnel keep-alive for state, and let all routers set up their own tunnel. I have any to any connectivity with only few keep-alives on each VPN tunnel. This works well with many borders. Scales to thousands of nodes, not to mention millions. Depends on backbone routing and number of VPN servers. We know how to make that work nicely.

Today, I use route injection in backbone with OpenVPN scripts (and quagga to inject in OSPF backbone). I use certificate common_name and transform to route prefix. Generic implementation may need some adjustments here, either better suited attribute in certificate, lookup in table or signaling the prefix with some kind of TC message.

Tuning OLSR doesn't help that much. As Hanning says: because sender of TC (etc.) messages mandates the interval for the whole OLSR network.

Teco



Op 26 feb. 2013, om 14:39 heeft Henning Rogge <(spam-protected)> het volgende geschreven:

> Hi,
> 
> incoming OLSR messages like TCs/MIDs/HNAs are delayed by a few hundred milliseconds (to allow message aggregation), but besides this they are (and have to be) forwarded instantly.
> 
> The emission intervals are only responsible for the generation of messages at the originator, not for forwarding messages.
> 
> Henning Rogge
> 
> On 02/26/2013 02:35 PM, Saverio Proto wrote:
>>> [*] RFC 3626 paragraph 18.1:
>>> """
>>>      -    the emission intervals (section 18.2), along with the
>>>           advertised holding times (subject to the above constraints)
>>>           MAY be selected on a per node basis.
>>> """
>>> 
>>> [^] http://zumbi.netgroup.uniroma2.it/~clauz/tmp/bigemission3.pcap
>> 
>> is it correct that the values in the configuration file are for the
>> OLSR messages generated by the router ?
>> 
>> the OLSR messages that are forwarded (i.e. originator id is another
>> router) are forwarded asap without respect of EmissionInterval.
>> correct ?
> 
> 
> -- 
> Diplom-Informatiker Henning Rogge , Fraunhofer-Institut für
> Kommunikation, Informationsverarbeitung und Ergonomie FKIE
> Kommunikationssysteme (KOM)
> Fraunhofer Straße 20, 53343 Wachtberg, Germany
> Telefon +49 228 9435-961,   Fax +49 228 9435 685
> mailto:(spam-protected) http://www.fkie.fraunhofer.de
> 
> -- 
> Olsr-dev mailing list
> (spam-protected)
> https://lists.olsr.org/mailman/listinfo/olsr-dev





More information about the Olsr-dev mailing list