[ntp:questions] 500ppm - is it too small?

David J Taylor david-taylor at blueyonder.not-this-part.nor-this.co.uk.invalid
Thu May 7 17:28:01 UTC 2009


Richard B. Gilbert wrote:
[]
> I'm not here to make people feel good!

No comment!

> I've checked the hardware available to me and none is worse than 50
> PPM. That's two PC's running Windows XP, three Sun Ultra 10
> Workstations running Solaris 8, 9, and 10, two DEC Alpha workstations
> running VMS.

Thanks for that.  Possibly, on the PCs are representative of the hardware 
class I'm considering.

> The specifications for NTPD say that it will correct errors less than
> or equal to 500 PPM.

Hence the title of this thread.

> I beleive that hardware outside of this limit can properly be
> described as broken!  Would *you* tolerate a clock, computer or wall,
> that gained or lost more then 43 seconds per day?  Our forefathers,
> limited to springs, gears and pendulums could do better than that.

As I hope I mentioned earlier, it may not be hardware, but some BIOS or OS 
power-saving scheme, or something else I've overlooked, which is stopping 
NTP from working.  I am trying to get more details.

I do completely agree that a clock with 43s per day error would be one I 
would reject!  All the non-radio and non-NTP clocks here do better than 
that.

Cheers,
David 




More information about the questions mailing list