See the other replies to your comment that have been sent to the list, and let me emphasize, DO NOT DO THIS. 

Historically, it has been shown to create problems, as noted in the reply from Björn, referencing http://en.wikipedia.org/wiki/NTP_vandalism

This is entirely a "me-too" posting, I realize, but at least it's a bit stronger than your average AOL kind of thing. Your application should absolutely get capability to handle DNS before shipping. Even shipping with a root hints file and self-updating it would be acceptable, even encouraged, compared to hardcoding ANY ip address.

The servers listed in that link are used by many of us in the pool to run our servers, which redistribute time to other hosts. The pool exists to help lower the load on those same servers you have sent us the link to, and they are absolutely not to be used by machines that don't exist to redistribute time.

-Jeff
SIG: HUP

On Jul 19, 2006, at 2:19 PM, Jon Tandy wrote:

Thank you to those who responded to my question.  I've found what I think is a solution to my issue.  The NIST publicly lists IP addresses for several time servers, which leads me to believe they should remain relatively stable.  I know it's not guaranteed, but it should hopefully be relatively consistent.  I'm passing it along in case you receive any future requests for such thing.
 
 
 
Jon Tandy
 
-----Original Message-----
From: Jon Tandy [mailto:[EMAIL PROTECTED]]
Sent: Friday, July 14, 2006 5:38 PM
To: '[email protected]'
Subject: Fixed IP addresses for SNTP

Hello,
 
We have some embedded devices which require time synchronization using SNTP, but which presently do not have the ability to use a named DNS server address.  Are there any known public, fixed IP sources which would be available for end users, which are not likely to change IP addresses?
_______________________________________________
timekeepers mailing list
[email protected]
https://fortytwo.ch/mailman/cgi-bin/listinfo/timekeepers

Reply via email to