[dns-operations] ID of IPv4 fragments and DNS and the future RFC

[Stephane Bortzmeyer]

On Sun, Jan 13, 2013 at 08:59:39PM +0100,
 Florian Weimer <fw at deneb.enyo.de> wrote 
 a message of 30 lines which said:

> A typical initial TTL is 64, so the packet lives for at most 64
> seconds.  (Originally, the TTL was measured in seconds,

It was a very long time ago. RFC 1122, in 1989, already said that the
TTL field is more a hop count than a real TTL.

Future RFC 6864 speaks about MDL (Maximum Datagram Lifetime) and,
relying on things like the reassembly timeout (RFC 1122, section
3.3.2), estimates it to two minutes.

> 1000 responses per second doesn't seem that much, though.

To the *same* destination? (The ID only has to be unique per each tuple
{src, dst, protocol}.) It looks a lot.
 
> (Fortunately, IPv6 comes with a 32 bit fragment ID...)

That's an answer to my question: move all the traffic to IPv6 and the
problem will disappear.