[ntp:questions] Re: Philosophical question about strata

Richard B. Gilbert rgilbert88 at comcast.net
Fri Oct 14 14:13:09 UTC 2005

Wolfgang S. Rupprecht wrote:

>"Richard B. Gilbert" <rgilbert88 at comcast.net> writes:
>>I think it makes a lot of sense for stratum one servers to peer with
>>each other!
>Will that have any effect?  Won't each stratum-1 only believe it's own
>directly attached stratum-0 reference?  Or will a large enough swarm
>of stratum-1's all with similar times be able to win the selection
>contest against a failed stratum-0?
If I correctly understand what I have read, a stratum one server with 
three or more peers could be convinced that its reference clock was less 
than perfect.  If the reference clock says it's 0301:03.01 and the three 
peers all agree that it's 0301:04.00 then server should tell its clients 
that the time is 0301:04.00.  If the peers say it's 0301 and the 
reference clock says it's 0401 the server should declare its reference 
clock insane!

More information about the questions mailing list