The service fail time detected by clustat is different from logged failure time in /var/log/messages on Red Hat Enterprise Linux

Issue

With command Clustat -i 1 :

Wed Jun 13 03:45:17 KST 2012
Service Name                   Owner (Last)                   State         
 ------- ----                   ----- ------                   -----         
service:was-tomcat5            (node1.rhcs.com)               recoverable   

Wed Jun 13 03:45:28 2012
service:was-tomcat5            node1.rhcs.com                 starting    

Wed Jun 13 03:45:31 KST 2012
service:was-tomcat5            node1.rhcs.com                 started   

In the file /var/log/message:

Service fail time:

Jun 13 03:45:16 node1 clurgmgrd: [20561]: <err> script:tomcat5: status of /data/nimadmin/tomcat/bin/tomcat5-rhcs.sh failed (returned 1) 
Jun 13 03:45:16 node1 clurgmgrd: [20561]: <err> script:tomcat5: status of /data/nimadmin/tomcat/bin/tomcat5-rhcs.sh failed (returned 1) 
Jun 13 03:45:16 node1 clurgmgrd[20561]: <notice> status on script "tomcat5" returned 1 (generic error) 
Jun 13 03:45:16 node1 clurgmgrd[20561]: <notice> status on script "tomcat5" returned 1 (generic error) 
Jun 13 03:45:16 node1 clurgmgrd[20561]: <notice> Stopping service service:was-tomcat5
Jun 13 03:45:16 node1 clurgmgrd[20561]: <notice> Stopping service service:was-tomcat5
Jun 13 03:45:16 node1 su: PAM unable to dlopen(/lib/security/pam_wheel.so)
Jun 13 03:45:16 node1 su: PAM [error: /lib/security/pam_wheel.so: wrong ELF class: ELFCLASS32]
Jun 13 03:45:16 node1 su: PAM adding faulty module: /lib/security/pam_wheel.so
Jun 13 03:45:17 node1 logger: Still WAS Process is running.
Jun 13 03:45:17 node1 logger: Still WAS Process is running.

service recovering time:

Jun 13 03:45:27 node1 clurgmgrd[20561]: <notice> Service service:was-tomcat5 is recovering
Jun 13 03:45:27 node1 clurgmgrd[20561]: <notice> Service service:was-tomcat5 is recovering
Jun 13 03:45:27 node1 clurgmgrd[20561]: <notice> Recovering failed service service:was-tomcat5 
Jun 13 03:45:27 node1 clurgmgrd[20561]: <notice> Recovering failed service service:was-tomcat5 

service starting time:

Jun 13 03:45:30 node1 logger: WAS Process is started.
Jun 13 03:45:30 node1 logger: WAS Process is started.
Jun 13 03:45:30 node1 clurgmgrd[20561]: <notice> Service service:was-tomcat5 started
Jun 13 03:45:30 node1 clurgmgrd[20561]: <notice> Service service:was-tomcat5 started

We can found the time in log is 1 second faster than the time we found with command cluster -i 1