Using a Time Server and my rp7440 drifts

[harmt01]

Here is my ntp.conf:

driftfile /etc/ntp.drift
server timeserver
broadcastclient yes
restrict default ignore

Drift file:
0.000 0

This server lags behind all the others about 5 minutes over the weekend.

I have all the lastest and greatest patches and I have 9 HP-UX 11.11 servers that work just fine. HELP!!!

[A. Clay Stephenson · Accepted solution]

I would comment out you "restrict default ignore" entry.

You have also violated one of the fundamental rules of good NTP design in that you have a single timeserver. The NTP client has nothing to comapre as to stability of the data. Normally, you want at least 3 timeservers.

I would also add a timeserver entry to the ntp.conf file that instructs ntp to use the built-in clock as a last-resort.

server 127.127.1.1 stratum 10 # localhost

Because you have a driftfile configured and the drift is 0. I rather suspect that the local clock is the only thing that is actually begin used.

We can learn a lot by your posting of the output of ntpq -p. That will tell us what is going on.

[James R. Ferguson]

Hi:

# ntpq -pn

...what does that show?

If your server's time is more thatn ~ 1000 seconds from correct time, your 'xntpd' daamon will die shortly after startup.

Regards!

...JRF...

[Steven E. Protter]

Shalom,

Run a series of nqpq commands to track the drift.

I'd think about getting the system hardware checked if its running that badly off in such a short time.

ntpdate is a command you may be able to put in cron (Have it in Linux no idea where it came from) that could keep the system more in line.

ntpdate timeserver
Will force an update.

Also see if ntpd is even running right.

SEP

[harmt01]

Thanks, added more servers and restarted XNTPD. Before doing this I got "Server Timed Out" Now I get a reading off all the time servers.

Much Thanks,

[harmt01]

Thanks