[58450] in North American Network Operators' Group
Re: NTp sources that work in a datacenter (was Re: Is latency equivalent
daemon@ATHENA.MIT.EDU (bmanning@karoshi.com)
Wed May 14 11:29:12 2003
From: bmanning@karoshi.com
To: steve@expertcity.com (Steve Francis)
Date: Wed, 14 May 2003 08:28:37 -0700 (PDT)
Cc: Michael.Dillon@radianz.com, nanog@merit.edu
In-Reply-To: <3EC25A84.1040700@expertcity.com> from "Steve Francis" at May 14, 2003 08:02:28 AM
Errors-To: owner-nanog-outgoing@merit.edu
> >in every PoP to do measurements. In that case, the difficulty isn't in
> >measuring one-way latency, it's in synchronizing the time on all the
> >servers. And with fairly cheap GPS and CDMA clocks that is a lot
> >easier/cheaper than it once was.
a robust mesh of strat-2 chimers gives one more resilence
and more accuracy than syncing off a single source.
> But what GPS clock can you install in a datacenter? AFAIK, they all
> require roof (or at least window) access in order to install the
> antenna. (At least, all the GPS based ntp servers I've looked at do).
> Is that not true of CDMA servers?
some GPS, some PPS, and an atomic source here and there
give great diversity and only a few need roof access.
> How have others solved this issue? (Short of owning their datacenters.)
Use NTP, run most systems as strat-2
--bill
