I had the same problem as Partrick, building a new kernel as of today 
also solved the problem.

Marcus

Frederick Bruckman wrote:

>On Mon, 5 Jul 2004, Patrick Welche wrote:
>
>  
>
>>On Sun, Jul 04, 2004 at 04:46:16PM -0500, Frederick Bruckman wrote:
>>    
>>
>>>Simply put, if the value in
>>>"/var/db/ntp.drift" is way off base, it could take days to straighten
>>>out, even if nothing else is wrong. If that's the only problem, you
>>>just need to manually change it to something close to where it's
>>>supposed to be before starting "ntpd".
>>>      
>>>
>>That was it - in fact without touching anything, things have settled. I
>>thought that after an ntpdate all would be well, but forgot about
>>ntp.drift..
>>
>>Thank you,
>>    
>>
>
>You're welcome, and thanks for the feedback.
>
>Hey, is your frequency changed after clock.c 1.83-1.84? My best
>timekeeper fell from ~157 to ~143 in 8 hours, and still falling.
>That's not bad, by itself, but together with the "too stiff" issue,
>it could lead to more problem reports.
>
>Frederick
>  
>