This content has been marked as final. Show 2 replies
You can check and modify /etc/ntp.conf to use the NTP server of your choice. By default it is using public servers from the pool.ntp.org project. For more details http://www.pool.ntp.org/join.html.
Then type "ntpq" and enter "peers" to see if you can connect.
The ntpd behavior at startup depends on whether the frequency file, usually /etc/ntp.drift, exists. It should be created automatically after 15 minutes. It may take some hours for the frequency and offset to settle down. See "Frequency Discipline" at http://doc.ntp.org/4.1.0/ntpd.htm
If there is too much offset between your CMOS clock and the NTP time, you can use the following command to update the CMOS clock.
You can also schedule this in cron if your hardware clock is bad. Maybe you need to replace the CMOS battery.
The hwclock --systohc command worked. I'm using an ODA where the secondary node was about 4 minutes off.
This was generating errors of "CRS-10051:CVU found following errors with Clusterware setup : PRVF-5408 : NTP Time Server".