From: Gary Lopez (gary@catapult.com)
Date: Wed Oct 01 2003 - 14:55:00 EDT
Thanks to Darren Dunham, Paul LaMadeleine, Frank Smith, Stephen Kives,
Dan Astoorian, and Jay lesset.
>On Fri, Sep 19, 2003 at 01:46:26PM -0700, Gary Lopez wrote:
>
>
>>one sever that is NOT nis'd but running solaris 8 as well. All the other
>>machines keep time fine with the NTPserver. Very little discrepancy. As
>>much as I do an rdate and sync the non-nis'd machine to the NTPserver,
>>it always runs out about 2-3seconds. My option is to run a cronjob that
>>does an rdate every 15min. Is there a reason why this machine continues
>>to run ahead so consistently ? If anyone would like logs or any other
>>info to help, I would gladly provide.
>>
>>
>First, I learned that rdate should NEVER be done when xntpd. Thanks to everyone who so passionately pointed that out. The following is what jay sent me and I tried it and it has worked perfectly since. Thanks again all. I really appreciate all your knowledge ..
>Gary,
>
>Running rdate on a machine that is already running xntpd will booger
>things up. So stop running rdate and let's just debug ntp.
>
>First let's:
>
>ntpq -p 192.168.0.10
>/etc/init.d/xntpd stop
>/etc/init.d/xntpd start
>tail /var/adm/messages
>sleep 60
>ntpq -p
>sleep 300
>ntpq -p
>sleep 300
>ntpq -p
>
>This will confirm that you're actually running xntpd and pointing
>at an NTP server that is functional.
>
>Now wait until it is "2-3 seconds out" and:
>
>date
>grep ntp /var/adm/messages | tail
>ntpq -p
>
>There is some Sun hardware (like Sunblade 100, and whatever the
>equiv Netra is) that has very bad RTC which needs to be tweaked
>in /etc/system. If this is going on, you'll see it in
>/var/adm/messages xntpd log entries.
_______________________________________________
sunmanagers mailing list
sunmanagers@sunmanagers.org
http://www.sunmanagers.org/mailman/listinfo/sunmanagers
This archive was generated by hypermail 2.1.7 : Wed Apr 09 2008 - 23:27:13 EDT