clock problem

[Jeffrey Goldberg]

On May 8, 2007, at 8:33 AM, Martin Dieringer wrote:

> On Tue, 8 May 2007, Oliver Fromme wrote:
>
>> Martin Dieringer wrote:
>> > My clock is now _15 minutes_ late, after about 1 day with powerd
>> > running.
>> > ntpd is running also.
>> > Can nobody tell where the problem is here?
>>
>> Are you sure that your /etc/ntp.conf ist correct?
>
>
> # cat /etc/ntp.conf
>
> server time.fu-berlin.de iburst maxpoll 9

Add other servers but don't use iburst, nor reduce the default  
maxpoll.  In your case try something like

    server 0.de.pool.ntp.org
    server 1.de.pool.ntp.org
    server 2.de.pool.ntp.org
    server 3.de.pool.ntp.org

Keep in mind that you will get better quality time by using multiple  
servers at higher strata than you will by using just one stratum 1  
server.  Well managed stratum 4 servers really are good enough (and  
often a better choice given availability).

>> Are there any messages from ntpd in /var/log/messages?
>> What's the output from "ntpq -p"?
>
> # ntpq -p
> No association ID's returned
> #

That's not good.  I think that adding more servers (as suggested  
above) will help.

Also if you have other machines on your local network, you can use  
them to peer with the one with the bad clock.

So add

   peer some-local-net-machine
   peer some-other-local-net-machine

to ntp.conf

I recall some discussion earlier on this list (I think) about these  
problems with powerd and clocks.  The solution involved some tuning  
with sysctl or something similar.  I'm sorry I can't be more specific.

You may also try to set things to restart ntpd on dial-up.

-j