rh-l] NTP Problems Redux

James Cooley jcooley at fit.edu
Mon Apr 11 20:43:36 UTC 2005


Also, try to keep in mind that if you are running ntpd, you do not need
to manually run ntpdate.  ntpd will slowly sync your clock to the remote
time server clock so as to avoid a drastic time change.  Also, ntpd
needs to be running for about 15 minutes before your computer will trust
its time sources for synchronization.

To see the status of the ntp synching, use the command ntptrace.  You
shouldn't really ever have to run ntpdate manually if ntpd is running.

--James Cooley

P.S. I like top-posting :)

R P Herrold wrote:

> On Mon, 11 Apr 2005, Gavin McDonald wrote:
>
>> I stop the ntpd service, and run ntpdate to check the configs:
>> # ntpdate
>> 11 Apr 09:52:36 ntpdate[31314]: no servers can be used, exiting
>>
>> But If I manually specify the servers, it works fine.
>> # ntpdate 0.ca.pool.ntp.org 1.ca.pool.ntp.org 2.ca.pool.ntp.org
>> 11 Apr 09:52:49 ntpdate[31316]: adjust time server 142.179.100.217
>> offset
>> 0.0177 63 sec
>
>
> a manual vs an initscript variance sounds like something in your path
> mungeing may be in play, although it may be a drop userid permissions
> issue, or that the resolver setup is somehow having issues.  I think I
> would add the strace package to your server, and wrap the initscript,
> capturing the strace and seeing where it falls apart.
>
> -- Russ Herrold
>





More information about the redhat-list mailing list