[ntp:questions] Re: time

Bjorn Gabrielsson bg at lysator.liu.se
Tue Nov 29 07:55:31 UTC 2005


"Richard B. Gilbert" <rgilbert88 at comcast.net> writes:

> K wrote:

> >I know about NTP v3 and v4 and use it on some systems.  I need a time
> >accuracy of < 1 ms.
> >
> >I know that the linux os is compiled for something like 1 ms precision

?
> 
> So it can be done.   I think that the Motorola GPS has a great deal to
> do with it.  

The motorola should carry you down towards 1us, but K was only asking
for a thousand times less accuracy... :-)

> clock would be much less stable.  10 ms is about the best accuracy I
> would expect from a client synchronized to a server that synchronized
> over the internet.

Even with a DSL-line - if its not overloaded - sub 1ms is easily
attained with good internet servers. Looking at a machine with four
external servers, I get see,

/usr/sbin/ntpq  -c pe 

     remote           refid      st t when poll reach   delay   offset  jitter
==============================================================================
 LOCAL(0)        73.78.73.84      5 l   25   64  377    0.000    0.000   0.001
 xxx             .PPS.            1 u  219  256  377   11.476   -0.235   0.779
 xxx             .PPS.            1 u  102  256  377   17.818   -0.140  75.660
 xxx             .PPS.            1 u  217  256  377   21.575    0.203   0.235
 xxx             .PPS.            1 u  192  256  377   20.889    0.385   0.142

This machine usually have two GPS-refclocks and they show no
systematic offsets wrt the Internet servers.

--

    Björn




More information about the questions mailing list