Because the VM is not the real host hence does not have the real clock to sync. The ntp daemon always has the large offset. Some application, like Oracle RAC requires the NTP daemon running and if the offset value is too large, we may get the warming or system failure. Here is what I did to remedy the issue.
1. Update the /etc/sysconfig/ntpd
OPTIONS="-x -u ntp:ntp -p /var/run/ntpd.pid -g"
NTPDATE_OPTIONS="-x -u ntp:ntp -p /var/run/ntpd.pid"
After change the setting, you may need to reboot the host to make it effective.
2. Disable the USB/Serial Port on the VM setting.
If above setting does not help, try to disable the USB and serial Port in the VM hardware
3. use ntpdate to sync manually.
If all the above fail, use ntpdate to sync manually and resart the ntpd .