[Olsr-dev] mDns plugin improvement

[Teco Boot]

Let's try to comply to standards. We should not break nodes that 
discards mDNS packets not having TTL=225. And support protocols 
that have TTL=1. 

Packet rate of mDNS would be low, so hash calculation looks
acceptable to me. This is the way SMF works.
http://tools.ietf.org/html/rfc6621#section-6.2.2
I prefer having the standards based method implemented. In
IETF, we had a long discussion. With this outcome.

It would be nice having only one copy of the mDNS (or whatever 
multicast) packet flooded. This needs some coordination between
the leaf routers. Could be active / backup election. Could be an 
extended dup detection in the flooding mechanism, with a plugin.

Teco

Op 30 mei 2012, om 20:00 heeft ZioPRoTo (Saverio Proto) het volgende geschreven:

> Hello Teco,
> 
>> mDNS packets SHOULD have TTL 255
> 
> This is true for the application that generates the packet. But once
> the packet is out in the network routers can alter the TTL value.
> 
>> Why not the more standard method using a hash table, like bmf? Costs more
>> CPU / memory, but it is much safer. p2pd has / had this. We didn't push
>> this version I think, because there was an issue with latest modifications.
>> nyway, we should get rid of 2 out of 3 plugins.
> 
> please note that this patch is NOT for checking if a "OLSR message" is
> for forwarding. This patch is to check if the IP packet should be
> captured from the users' LAN and passed for encapsulation to mDNS
> plugin.
> 
> this way we do not capture a packet that another OLSR router pushed to
> a ethernet LAN, we only capture packets generated from user
> applications that have TTL 255
> 
> we believe that using hash table is just waste of CPU when you can
> check if the packet is good for forwarding just checking the TTL
> field.
> 
> Saverio