[ntp:questions] Re: server's address in ntp payload?

David L. Mills mills at udel.edu
Tue Nov 22 22:41:42 UTC 2005


Not at all to prolong this discussion, but there is a simple test to 
determine if multiple addresses belong to the same machine. If a 
certificate with public key works and is verified for a number of 
addresses, they are all on the same machine and that machine has the 
private key. I do the same thing here as you do, trusted with no 
authenticate and untrusted with authentication. There is a discussion on 
secure hierarchical groups on the autokey protocol page and briefing at 
the NTP project page.


David Schwartz wrote:
> "Danny Mayer" <mayer at gis.net> wrote in message 
> news:4380B20A.9010804 at gis.net...
>>David Schwartz wrote:
>>>    For example, suppose I have two servers, A and B. A is talking to a
>>>third server C. I wish to determine if a server B is talking to is that 
>>>server C, or a distinct server, D. How will a session hash help me do 
>>>On the other hand, a public server ID that is globally unique *will* help 
>>>do that. The source IP won't, because a single machine can have multiple
>>That's what the refid is for.
>     I've seen many servers with the same refid. It does not appear to be a 
> globally unique server identifier.
>     DS

More information about the questions mailing list