[p2p-hackers] UDP Keep-alive

[Travis Kalanick]

Thanks for the input!  

Here's something I found:

http://www.tomax7.com/mcse/cisco_ipcommands.htm

Cisco gear does a 300 sec default timeout on NAT translation tables for UDP
connections.


-----Original Message-----
From: p2p-hackers-bounces at zgp.org [mailto:p2p-hackers-bounces at zgp.org] On
Behalf Of Eugen Leitl
Sent: Tuesday, January 11, 2005 1:14 PM
To: Peer-to-peer development.
Subject: Re: [p2p-hackers] UDP Keep-alive

On Tue, Jan 11, 2005 at 01:01:33PM -0800, coderman wrote:

> Most NAT's appear to give a 1-5 minute timeout since last
> packet seen.  It would be nice to find a market summary of
> various NAT behaviors.  Anyone know of such a thing?

Here's a data point: I've spent the better part of the day trying to remove
default NAT (60 sec TCP, 180 sec UDP) idle connection decay from a Draytek
Vigor 2900G with latest firmware.

To no avail, had to send email to support (probably, /dev/null).
Proprietary NAT boxes are evil, period.

Which reminds me (since I've got my IPv6 subnet approved a few days ago):
what's p2p application situation for IPv6? Can anyone give a brief summary?
 
> The timeout is trivial to avoid with loose UDP NAT (simply
> send / recv a packet from any peer within the few minute
> window).  For symmetric NAT it is more of a pain. Each
> session must send traffic within the timeout window which
> raises overall communication significantly for large numbers
> of logical NAT-UDP connections.

-- 
Eugen* Leitl <a href="http://leitl.org">leitl</a>
______________________________________________________________
ICBM: 48.07078, 11.61144            http://www.leitl.org
8B29F6BE: 099D 78BA 2FD3 B014 B08A  7779 75B0 2443 8B29 F6BE
http://moleculardevices.org         http://nanomachines.net