NTP problem - Clock too fast for NTP to keep up?

[jdow]

From: "Gene Heskett" <gene.heskett at verizon.net>

> On Wednesday 09 February 2005 23:25, Peter Kiem wrote:
> >> I too had the fast clock problem, and by tickling the tick counter
> >> with the 'tickadj' command, was able to stay well within 1 second
> >> per hour.  The bare command will return the current tick value,
> >> default is 10000, but I'm curently setting it at 9926 for accurate
> >> time keeping.  This seems to be 2.6 problem as the default is
> >> pretty close on a 2.4.29 system running on an elderly TYAN mobo at
> >> 500mhz.
> >
> >Where and how would you set the tickadj parameter?
> >Would you set it higher or lower for a system running too fast?
> 
> Lower.  Default is 10000.
> 
> Setup an ntpdate to hit the net for the time, say hourly.  Put
> a tail on the log so you can see how far off it is.

This one is better. I found that ntpdate, even when told not to mess with
the system's timing, still causes problems with the smooth operation of
ntpd from time to time. Using ntpq removes any interference potential.

===8<---
#!/bin/bash
echo >>ntp_log "NTP Log for about 16 hours every 64 seconds"
for (( i = 0 ; i < 900 ; i++ )) ; do
    date >>ntp_log
    ntpq -p >>ntp_log
    sleep 64
done
===8<---

{^_^}