Here is your base pricing from Truetime:
NTS-150 $2395
NTS-200 $3595
-Mike
Hmm... $2400 is still in the "pricey" range to be throwing out bunches of these across a network in wide distribution. (Pardon me if some of you on the list snicker at my reluctance at the $2400 price - for some of us the "new, new Econcomy" is making things like NTP Stratum 1 clocks a luxury that The Budgeters doesn't see as necessary, since it's an invisible engineering issue.)
One would think that a vendor could come up with a 1u rackmount box with a GPS and single-board computer (BSD or Linux-based) for ~$500 total cost. Add 150% for profit and distribution costs, you're still in the $1300 range, which is more reasonable. I suppose my oversimplification is the reason I'm not in the hardware business. I'd be even happier with a PCI-bus card that I could put into an old (reasonably fast) PC and a CD-ROM with an OpenBSD distribution that automatically did the Right Thing. There is a case to be made about off-the-shelf PC hardware not being accurate enough to handle a true Stratum-1 clock, and that is a valid point. However, if I can get within .5ms, I'm happy since most of my applications don't require anything more accurate than that. (Those of you timing T1's should use the more expensive systems.)
I will go out on a limb and say that a reduction in the cost of stratum-1 servers will increase their use across the Internet. The results of such an increase would be arguably visible, as the current multi-layer timekeeping system seems to be more-or-less keeping clocks correct to the point of usefulness, at least from a layer-4-and-up standpoint. However, accuracy and self-determination for timing are probably things that most organizations would consider "good" by self-evidence, and the lower the price the more possible things become to implement. Perhaps there are reasons that putting stratum-1 clocks in many, many places is sub-optimal; I leave that for others to illuminate.
I know that I would like to not rely on POP-external network connections to keep my clock sources accurate, but these prices (while very inexpensive, compared to other stratum-1 sources I have seen) are still outside the "put-one-in-every-POP" price.
JT
Is it invisible? Proper timing is essential. It's not too
hard to pick a suitable GPS and plug it into a host somewhere if
cost is an issue.
But, more to the point, you don't need a "wide distribution"
of these boxes. 2 or 3 is more than enough. I tend to use
my top level routers, or some distributed hosts (dns, authentication,
logging, you name it) to form a stratum 2 mesh, and then have the rest
of your network talk to them.
A large number of stratum 2 servers talking to each other as
well as a few stratum 1 clocks will result in a very stable distributed
timesource that can support a whole lot of clients.
You've already paid for the network, might as well use it.
--msa
On Tue, Aug 27, 2002 at 11:57:39PM -0400, John Todd mooed:
Hmm... $2400 is still in the "pricey" range to be throwing out
bunches of these across a network in wide distribution. (Pardon me
[...]One would think that a vendor could come up with a 1u rackmount box
with a GPS and single-board computer (BSD or Linux-based) for ~$500
total cost. Add 150% for profit and distribution costs, you're
still in the $1300 range, which is more reasonable. I suppose my
oversimplification is the reason I'm not in the hardware business.
You might be imagining a somewhat larger market for standalone
stratum-1 timeservers than you might imagine. For real accuracy,
you don't want standalone -- you want a locally connected source that
you can use to tightly discipline the local clock. (When I say "real," I
mean sub-milisecond). The difference is .. substantial. Taken
from two of my machines on the same subnet:
remote st poll reach delay offset disp
As I am sure you have noticed from other replies on the list here, the
idea for NTP is not to have a Stratum one device at every single POP. That
would be pricey not only in equipment costs but in roof-rights cost. What
many do for NTP is to have one or two Stratum 1 devices amongst your
network and then distribute it to a box that would then in turn distribute
down to the next layer of equipment and so on. So if you are only spending
$2400 and maybe even $4800 to support NTP across your whole network, I would think
that would be worth it.
-Mike
On Tue, Aug 27, 2002 at 11:07:10PM -0700, Jim Hickstein mooed:
At work, it's all steel studs and foil-backed wallboard, and the windows
(for a patch GPS antenna) are _way over there_. *sigh* I'd love it if
someone would pay for my roof penetration there.
Does your cell phone work in the room? The CDMA time receivers
work in the strangest places. The only places I've had no luck:
- In the bowels of a big building along the mass tech corridor
- when moved to a bad spot inside a fairly steel and concrete-heavy
lab at the university of utah (works in other spots in the lab).
But aside from that, I've got them working in network closets
and labs all over the place. Worth giving a shot if you're really
desperate to play. They're not quite as accurate as GPS (~10 microseconds
vs. ~2-5 microseconds), but what's a few microseconds compared to
sticking an antenna on the roof?
(As a frequency standard, they're quite good. But you can't autocorrect
for the CDMA propagation delay).
-Dave
Hmm... $2400 is still in the "pricey" range to be throwing out
bunches of these across a network in wide distribution.
and why would one want to do so? run one strat 1, two at most (widely far
apart, like on different continents), and chime routers off them, chime
everything else off the routers.
randy
Randy Bush <randy@psg.com> writes:
> Hmm... $2400 is still in the "pricey" range to be throwing out
> bunches of these across a network in wide distribution.and why would one want to do so? run one strat 1, two at most
No.
http://groups.google.com/groups?hl=en&selm=3C32924F.994E1D01%40udel.edu
---Rob
"Every critical organization should run at least four
low-stratum servers configured as above, so dependant servers and
clients can do the same thing. Each critical server should run NTP
symmetric mode (better yet manycast mode) with each of the other
servers at the same stratum, together with at least one peer at the
same stratume in another trusted organization."
By this criteria, a stratum 2 mesh of a bunch of top tier
routers or diverse hosts, with external influence, should be okay.
--msa
BTW: A rather complete list of NTP products:
http://www.eecis.udel.edu/~ntp/hardware.html
Some low price products from random browsing through the list
$ 1,400
http://www.zyfer.com/products/prod_index.html
$ 380
http://www.gpsclock.com/specs.html (looks like serial output only..)