[ntp:questions] Re: Public servers?

David L. Mills mills at udel.edu
Fri Aug 1 23:12:36 UTC 2003


Huh? There is no intent whatsoever to demonize Harlan in any way, shape 
of form. I am reacting strictly to the content of his message. My 
message said nothing about traceroute or who may have tools based on 
traceroute. How did you get the impression otherwise?

We have had many prior discussions on the newsgroup about the access 
control, service area and automatic configuration in one form or another 
and each and every time the discussion has mired in the access controls 
and service area issues. Discussion would be vastly more productive if 
these issues could be resolve. Upon closure, technical details and 
prototol augmentation are essentialy trivial.

You might start with BGP as a point of departure.


Brad Knowles wrote:
> At 5:51 AM +0000 2003/08/01, David L. Mills wrote:
>>  If the ISP explicitly suggests a victim router, rejoice. If not,
>>  do not rejoice and do not assume in any case whatsoever that a
>>  casual pry of ISP routers for NTP response would be anything but
>>  evil. Please talk to the CSIRO and U Wisconsin and Ultimeth
>>  operator for consensus views.
>     Harlan is not the criminal here.  Harlan doesn't have a tool that 
> does a traceroute from his server to the IP address of the web browser, 
> then does an ntpdate command against each of the IP addresses that show 
> up.  Tim Hogard has that tool, and apparently has had it since 1995.  If 
> you want to yell at someone, yell at Tim and not Harlan.
>     What Harlan is trying to do is open a wider discussion on the 
> possibility of NTP protocol enhancements that might allow us to do local 
> auto-discovery of proper NTP servers, in accordance with the access and 
> usage policies of the individuals or organizations that own the devices 
> in question.  I think this is a good subject to discuss.

More information about the questions mailing list