Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!rice!sun-spots-request
From: libes@cme.nist.gov (Don Libes)
Newsgroups: comp.sys.sun
Subject: Re: Looking for a good source of time on the net.
Keywords: Networks
Message-ID: <5957@brazos.Rice.edu>
Date: 21 Mar 90 06:24:52 GMT
Sender: root@rice.edu
Organization: Sun-Spots
Lines: 24
Approved: Sun-Spots@rice.edu
X-Refs:  Original: v9n87
X-Sun-Spots-Digest: Volume 9, Issue 87, message 10

In article <5933@brazos.Rice.edu> vaillan@ireq.hydro.qc.ca (C.Vaillancourt Hydro-Quebec QC 514-652-8238) writes:
>My question is: Is there any installation of Sun on the net synchronized
>with a receiver to the Universal Coordinated Time?

Our lab in Boulder maintains the civilian Time-Of-Day.  They have a
computer which they guarantee is within 30 milliseconds of the correct
time.  (At the moment it is 17 msecs off.)  It happens to be a Sun running
UNIX, so naturally it runs 'rdate'.  The hostname is india.colorado.edu.
(Don't ask what it's doing in .edu)

This means you can put 'rdate' into your server's rc.local.  To reduce
network traffic, the 'rdate' in your client's rc.local should point to
your server.  (See rdate(8) for more info.)

Of course, 'rdate' makes no guarantees of accuracy, but sample tests we
ran indicated the results were less than a second off.  (We're in
Gaithersburg, Maryland!)

Boulder is working on a server to provide time service to better than
millisecond accuracy taking into account network and OS latency, but it is
not yet available.  They seemed disappointed when I said "within a couple
of seconds would be just fine."

Don Libes          libes@cme.nist.gov      ...!uunet!cme-durer!libes