[time-nuts] Legal Time dissemination

Miguel Barbosa Gonçalves m at mbg.pt
Tue Jun 4 22:37:59 EDT 2013


Hi Chris!

I have a week's worth of data now... Take a look at these graphics

http://www.flickr.com/photos/miguelbarbosagoncalves/8955346508/ and

http://www.flickr.com/photos/miguelbarbosagoncalves/8955346522/

One of them shows is the plot of the loopstats of my stratum 1 server. The
other one shows the offset determined every minute of their servers
compared to mine.

On 28 May 2013 16:11, Chris Albertson <albertson.chris at gmail.com> wrote:

> The "stratum 2" servers are by definition not connected to GPS.  They get
> their time from some other NTP server that is connected to an authoritative
> clock which may or may not be GPS.    It looks like the "red server" has
>

I know that but I would expect good stratum 2 servers if they are keeping
their stratum 1 servers private and not available to the public.


> rather smooth swings over around a "handful" of milliseconds.   This is to
> be excepted.  It is within the normal range of what NTP does.  Perhaps it
> is a Windows PC running in some room where the temperature changes and the
> network that connects it to the strum 1 server is loaded.  I don't know but
> a handful of milliseconds is in the normal range.  Yes it could be better.
>

I don't believe it is a Windows server but I could be wrong...


> Also the curve is somewhat smooth.  It does not look like noise from a busy
> network.  It looks like their server really is moving around.
>

I believe they are on the same network. It looks like a temperature problem
really.

Thanks for all the input!

Kind regards,
Miguel


More information about the time-nuts mailing list