[LEAPSECS] Leap Sec vs Y2K

[Ian Batten]

On 11 Dec 2010, at 16:23, Paul Sheer wrote:


> 

>> I'm surprised by your claim that Telcos don't do NTP, [...]

> 

> I'm sure many do.

> 

> My point is that, if one's starting premise is that most systems in the

> world *require* second-accurate syncronization, this is simply untrue.

> 

> In fact most work perfectly well even when they drift by minutes,


Well, except for Active Directory, which sets an upper bound of five minutes on the maximum error.  In practice, an AD deployment in which clocks were allowed to drift apart by minutes would behave very badly, so the typical target is less than that.  Yes, in principle, an AD deployment could synch to an artificial time, or indeed to whatever time the AD masters drifted to (although, as they'd be drifting independently, there would be all sorts of problems with doing that, such as what happens to the clocks of clients when they switch from one slave server to another).  In reality, it's much easier to synch to UTC or its local realisation than to some private timescale, especially in multi-site deployments.  So I doubt there are many Windows networks of any scale that don't synch their clocks.

Still, it's not as though either Windows or AD are widely used.

ian