[time-nuts] NTP stratum 1 appliances with different (GPS, etc) cores...

[Robert Seastrom]

On Mar 9, 2015, at 4:31 PM, Robert Seastrom <rs-lists at seastrom.com> wrote:

> Anyway, I couldn't keep my mouth shut and suggested that at the very least, trying for multiple vendors (we have a strong preference for appliancey stuff here due to internal fiefdoms, ownership issues, and balkanization typical of a Fortune 200 company) with different GPS cores would be a really fine plan.
> 
> So of course, now it's my job to come up with proper nominees for this list.  Any suggestions would be greatly appreciated.  Hopefully someone's already done the legwork for this.  I will likely send this same request to the ntp dev list...

Appreciate the feedback I've gotten on and off list so far.  It occurs to me that I may not have made it completely clear what I'm looking for.

Specifically, I'm looking for information of the form "We are using X radio clock appliance which is still available/not EOL and it uses Y core and has a source of Z (likely limited to GPS and WWVB)".

Or better yet, "Last year we deployed one each of A with A' core, B with B' core, C with C' core, D with D' core, and E which unfortunately has the popular A' core but those are the breaks, and unfortunately ".

Roof access is not a problem.  We own the datacenters in question and the land underneath them; I just write up engineering docs and a MOP and open a ticket with Critical Infrastructure and voila!  antenna on roof, properly mounted and grounded, with cabling that comes to the rack where the radio clocks live.  I don't even have to climb ladders with a tool satchel.  Whenever I am annoyed by certain aspects of working for a big company, I remember things like this. 

Besides GPS, we have a datacenter around 115 km from WWV and WWVB, so those are viable choices too.  Alas, no more GOES, LORAN, etc.  If anyone knows of over-the-air solutions or a pure GLONASS solution that unwinds cleanly to UTC within the tolearances I outlined in the original post so much the better.

Money is not a problem, though it would be a mistake to regard us as an organization with more money than clue.

Unfortunately, lovingly hand-built solutions, even OS-on-commodity-server-with-board, aren't going to fly due to the aforementioned ownership issues.  I'd love to be assembling stuff myself, but then I'd be stuck with supporting it.  I suffer from both the "hit by a bus" problem and the fact that getting to Operations from where I sit in the org chart involves traversing a pair of senior vice presidents.  When one is purely on the R&D/Architecture side, it won't do to deliver a solution to Operations that they can't support...  so commercial product is 100% a must.

Ditto chiming to anyone as a source who we don't have a contractual relationship with (and probably a direct connection to).  Our change control process makes this unpalatable.

I also get it as far as network time infrastructure goes and how to put it together and monitor it.  Still grateful to Dr. Mills for failing to chase me away from his lab when I was an ankle-biting high school student.

If anyone's done FMECA for time infrastructure and can share the document, it would save me a fair amount of work and get me additional traction internally.

Thanks,

-r