[ntp:hackers] 4.2.5p127 odd behaviour.

David L. Mills mills at udel.edu
Thu Sep 4 01:53:36 UTC 2008


I am running p127 here on three machines at present. All three machine 
have consistent and stable offsets well below the millisecond; however, 
they wobble a few tens of microseconds following the air conditioner 
cycles. Just to be ornery, I torqued one of them a few tens of 
milleseconds off-track and restarted it. It behaved as expected - zero 
crossing in about 3000 s with a few percent overshoot and slow return to 
zero with no oscillatory behavior.


I regularly see mail from the hackers list. It's the bugs and questions 
lists I prefer to opt out of.


Harlan Stenn wrote:

> Reg,
> Please add Dave as a Cc: on things like this. He is not on (does not
> choose to be a subscribed member of) hackers.
>> This is 4.2.5p127, running since early yesterday.
>> There was a -3ms offset in the loopstats when it
>> started up, and it has been oscillating ever since,
>> starting with an amplitude of 35ms and SLOWLY
>> decreasing.
>> This is not good.
>> A plot is attached of the current 24hr period.
> I bet the email system scrubbed that attachment. Could you please post
> a URL to it, and if people think picture(?) attachments are generally OK
> for the list we can ask the listmaster to allow them.
> Personally, I'd rather have a URL to the plot than have it be in the
> email, but that's really just my personal opinion/preference.
> H

More information about the hackers mailing list