Cluster makes use of database timestamps for fail-over and switch-over

Current Situation

• The JOC Cockpit cluster makes use of the database for synchronization purposes.
• However, for fail-over and switch-over operations the OS timestamps of the active JOC Cockpit instance and the standby instances are used. OS timestamps from different servers can run out of sync. If the difference between OS timestamps grows then switch-over can be ignored and an immediate fail-over can happens if a new JOC Cockpit instance is started.

Desired Behavior

• The JOC Cockpit cluster should make use of database timestamps to synchronize the operations to fail-over and to switch-over.
• This excludes any other operations such as logging that stick to use of OS timestamps.