Re: Randomness requirements for message ID generation

[Edward Lewis <Ed.Lewis@neustar.biz>]

At 10:56 -0400 7/12/05, Ólafur Guðmundsson /DNSEXT  co-chair wrote:

> Good randomness should be used by all DNS resolvers on query ID.

During some recent testing, I looked at the 
message ID's used by a popular and recent release 
of DNS software.  Over the time period I looked, 
some id numbers were used 6 times, a lot of the 
numbers were not.

Not all that random - without loss of 
functionality.  Although not random, the id 
numbers were somewhat unpredictable.  Well, come 
to think of it, no single id number ever had two 
simultaneously outstanding requests.  It's 
certainly not random, and you can predict that a 
number won't appear if it's outstanding.

I know that this is a dead issue in the WG 
(fortunately), but I had some real data to throw 
in.  And I wanted to kill once and for all the 
notion that the message id had to be "random."

--
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
Edward Lewis                                                +1-571-434-5468
NeuStar

If you knew what I was thinking, you'd understand what I was saying.

--
to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.
archive: <http://ops.ietf.org/lists/namedroppers/>