NIST announces new time servers

[Judah Levine]

In collaboration with TrueTime, Inc. and AOL Time Warner, Inc. the Time and
Frequency Division of NIST announces two new stratum-1 time servers. These
servers will support open access from any address that queries the server no
more frequently than 20 times per hour (using a one-day average). Users who
need more frequent access should contact NIST at the e-mail address
ti...@time.nist.gov.

The names and addresses are:

nist1-aol-va.truetime.com, ip address: 205.188.185.33, located in Northern
Virginia

and

nist1.aol-ca.truetime.com, ip address: 207.200.81.113, located in the San
Francisco Bay area

These servers operate at stratum 1 and provide time stamps that are directly
traceable to UTC(NIST). In addition to the Network Time Protocol, these
systems will respond to requests in other formats, including daytime (tcp/ip
port 13) and time (udp/ip) port 37. You can download documentation
and example source code from either system using anonymous ftp. Look
at URL: ftp://<server name>/pub for more information. (When using this
URL, replace <server name> with the name of one of the servers listed
above.)

For additional information about other NIST services, look at the Time
and Frequency web page at: www.boulder.nist.gov/timefreq. Information about
TrueTime is at www.truetime.com. The web page for AOL Time Warner is
www.aol.com.

Questions or comments about these servers may be directed to:

ti...@time.nist.gov
28 February 2001

[James Kirkpatrick]

Judah Levine wrote:
> In collaboration with TrueTime, Inc. and AOL Time Warner, Inc. the Time and
> Frequency Division of NIST announces two new stratum-1 time servers.

...
> nist1-aol-va.truetime.com, ip address: 205.188.185.33
>
> and
>
> nist1.aol-ca.truetime.com, ip address: 207.200.81.113

>
> Questions or comments about these servers may be directed to:
>
> ti...@time.nist.gov

FWIW their DNS is messed up. Reverse lookup of those addresses works but
name-to-address is bungled. I sent them email.

Jim

[James Kirkpatrick]

Judah Levine wrote:

> In collaboration with TrueTime, Inc. and AOL Time Warner, Inc. the Time and
> Frequency Division of NIST announces two new stratum-1 time servers

First, this sounds like "We're the US government, brought to you by AOL
Time Warner TrueTime." Oh Well. I suppose they deserve credit for their
contribution.

Anyway, the Virginia server is unimpressive from here. Over the last 18
hours or so it was a "false ticker" ("x" in ntpq -p) well over half the
time, with reachability problems (dropped packets) at least 30% of the time
(eyeball scan of output from ntpq -p every minute). Dispersion wasn't too
bad, well under 1.0. Delay about 88 ms, pretty consistent, but offset was
between 15 and 17 (ms I assume). Of course this could be due to network
issues which is why I said "from here."

The California server isn't much better. ntpq always reported "discarded
by clstering algorithm" ("-"). Reachability about the same, delay around
55 ms. Dispersion seemed worse and offset about -6 to -7.

We have an onsite Datum/Bancomm TymServe 2100 (GPS) and use a few others
for failover. We'll probably stick with what we have and remove the
new servers from this trial.

I do appreciate the effort, and undoubtedly these servers will be of
great use to sites "closer" to them. I'm not knocking them, just reporting
some results from the middle of nowhere (Wyoming -- that big square state
just north of Denver :-).

Jim