Prime Optical Server not keeping correct time.

Question

OK this is an issue that I have been having since we loaded Optical Prime 9.3.1. I thought it might have been solved in OP 9.6, it hasn't.  We are running both on two separate Red Hat 5.8 servers that are spot on to the NTP server we have.  The NTP is tied to a GPS Strat 1 source. We have tried every setting we can think of.  It's not a small drift either.  Current time on both RH servers is 1956 UTC. the 9.3.1 OP says 0207 UTC tomorrow, and the 9.6 OP server shows 0443 UTC tomorrow.   It is frustrating to my Techs and me.  I have told them to use the NE alarm time on the screen for tracking tickets. 
I need help. it is causing problems with data collection and analysis. things I can do in CPO I have to collect from 117 different NE's.
Thanks for any suggestions,
Steve

Answer 1

try tp cheack nymber of ntp procceses and leave only one like in example:
[root@CPO ~]# ps -ef | grep -v grep | grep "ntp" | wc -l
2
[root@CPOn ~]# ps -ef | grep -v grep | grep "ntp"
ntp       2040     1  0 Jan23 ?        00:00:01 ntpd -u ntp:ntp -p /var/run/ntpd.pid -g
root      2051  2040  0 Jan23 ?        00:00:01 ntpd -u ntp:ntp -p /var/run/ntpd.pid -g
Stop CPO - opticalctl stop
[root@CPO ~]# kill -9 2051
[root@CPO ~]# ps -ef | grep -v grep | grep "ntp"
ntp       2040     1  0 Jan23 ?        00:00:01 ntpd -u ntp:ntp -p /var/run/ntpd.pid -g
[root@CPO ~]# ps -ef | grep -v grep | grep "ntp" | wc -l
1
Cheack command "prefer" after IP of NTP server in /etc/ntp.config file.
Start CPO - opticalctl start