[Olsr-users] Too big packets | > 1420 Bytes!

[Teco Boot]

Op 12 aug. 2012, om 22:26 heeft Markus Kittenberger het volgende geschreven:

> 
> 
> On Sun, Aug 12, 2012 at 2:06 PM, Michael Rack <(spam-protected)> wrote:
> All right! That figures :)
> 
> So i try to adjust the vpn-interface mtu settings. I hope OLSR splits messages, that gets bigger then the maximum mtu / mss.
> afair olsr does not split existing messages!
> 
> it will just not combine olsr-messages to packets that are too big for the local interface mtu
> (usually/often olsr-packets consist of multiple olsr-messages)
> 
> but if you have nodes that have enough neighbours to create > 1420 byte long messages, you have a problem.
Why can't IP fragmentation handle >MTU packets?
If IP stack is unaware of MTU, there is a problem for all IP traffic.

Teco

> 
> Markus
> 
> 
> Currently i use OLSR v0.6.2.
> 
>> DebugLevel      0
>> IpVersion       4
>> FIBMetric       "flat"
>> ClearScreen     yes
>> 
>> RtTable 100
>> RtTableDefault 253
>> 
>> AllowNoInt      yes
>> TosValue        16
>> 
>> Willingness             3
>> LinkQualityLevel        2
>> LinkQualityAlgorithm    "etx_ff"
>> 
>> UseHysteresis   no
>> Pollrate        0.05
>> NicChgsPollInt  3.0
>> 
>> TcRedundancy    2
>> MprCoverage     1
>> 
>> InterfaceDefaults
>> {
>>   ...
>> }
> 
> Liebe Grüße aus Freilassing,
> 
> Michael Rack
> RSM Freilassing
> -- 
> RSM Freilassing                 Tel.: +49 8654 607110
> Nocksteinstr. 13                Fax.: +49 8654 670438
> D-83395 Freilassing            www.rsm-freilassing.de 
> Am 12.08.2012 01:09, schrieb Song Gao:
>> Hey Michael,
>> 
>> I am probably not the best person to answer this but from my understanding, OLSRd uses MTU of the interface to decide packet size. The interface created by VPN tunnel should be setting proper MTU.
>> 
>> Could you post your OLSRd config file as well as the version of OLSRd you are using?
>> 
>> 
>> Kind regards,
>> -- 
>> Song Gao
>> https://profiles.google.com/song.gao.beta/about
>> 
>> 
>> 
>> On Sat, Aug 11, 2012 at 3:49 PM, Michael Rack <(spam-protected)> wrote:
>> Hi OLSR-Users!
>> 
>> I have a VPN-Tunnel that runs over a PPPoE-Link, so i have only 1492 Bytes available. The VPN-Tunnel is established via             UDP, so i have 8 Bytes (MAC), and 12 Bytes (UDP) overhead. So the remaining MTU is 1472 Bytes for the inner Layer-2.
>> 
>> So there are 1452 Bytes left for OLSR.
>> 
>> But OLSR currently send out packets with more then 1452 Bytes, so the VPN-Tunnel goes down and automatically get reestablished.
>> 
>> Aug 11 21:18:50 vtund[6598]: Session neubichl-dsl-backup[79.192.xxx.xxx:46105] opened
>> Aug 11 21:18:50 vtund[6598]: UDP connection initialized
>> Aug 11 21:19:29 vtund[6598]: Message too long (90)
>> Aug 11 21:19:29 vtund[6598]: Session neubichl-dsl-backup closed
>> 
>> currently i drop packages over 1420 bytes that are passing the DSL-Interface:
>> iptables -A OUTPUT -o tap107 -m length --length 1420: -j DROP
>> 
>> since that, the VPN-Tunnel is stable.
>> 
>> ~ # ip route show table olsr-network | wc -l
>> 84
>> 
>> currently there are 84 routes spreaded over the whole network via OLSR. That can not result in a so high package size? How can i reduce the packet-size, or what have to be done, so olsr does not eat up more then 1400 Bytes?
>> 
>> 
>> Liebe Grüße aus Freilassing,
>> 
>> Michael Rack
>> RSM Freilassing
>> -- 
>> RSM Freilassing                 Tel.: +49 8654 607110
>> Nocksteinstr. 13                Fax.: +49 8654 670438
>> D-83395 Freilassing            www.rsm-freilassing.de
>> 
>> 
>> -- 
>> Olsr-users mailing list
>> (spam-protected)
>> https://lists.olsr.org/mailman/listinfo/olsr-users
>> 
> 
> 
> --
> Olsr-users mailing list
> (spam-protected)
> https://lists.olsr.org/mailman/listinfo/olsr-users
> 
> -- 
> Olsr-users mailing list
> (spam-protected)
> https://lists.olsr.org/mailman/listinfo/olsr-users

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.olsr.org/pipermail/olsr-users/attachments/20120813/c5e84acd/attachment.html>