[ntp:questions] testing slew only mode (-x), not slewing correctly (linux sles10, ntpd v 4.1.1)

Richard B. Gilbert rgilbert88 at comcast.net
Fri Oct 23 17:18:59 UTC 2009


Unruh wrote:
> E-Mail Sent to this address will be added to the BlackLists <Null at BlackList.Griffin-Technologies.invalid> writes:
> 
>> Richard B. Gilbert wrote:
>>> BlackList wrote:
>>>> I guess I could see 10 minutes (or more) if the time had
>>>>  never been set to anything reasonable;  However 10 minutes
>>>>  seems really far out for something that was previously
>>>>  syncing with NTP servers (or any other source of time).
>>>>
>>> I don't see what "previously synching" has to do with anything.
>>> If previously synching means three weeks ago, it's meaningless!
>>> If it means thirty minutes ago, I would expect the clock to be
>>> within a few milliseconds of the correct time.  If not I would
>>> regard the clock as broken!  I'm assuming reasonably constant
>>> temperature!
> 
>> If over three weeks, a cheap generic RTC gets out by 10 minutes,
>>  it is broken.  Thats an average of 1.6 seconds drift, in the
>>  same direction, every hour?
> 
> 10 min in three weeks is about 300PPM, which is not that ridiculous. You might call it
> broken but not seriously.
> 

I'd call it "lunatic fringe"!  Most clocks should be within 100 PPM or 
less.  My Sun hardware reports values in the +/~ 50 range.




More information about the questions mailing list