Fail-over in a backup cluster should work even when order configuration files do not have the same modification date

Current Situation

• In the fail-over situation for a passive JobScheduler Master cluster orders will not be continued when the order configuration files in both instances have different modification dates.

Reason for this behaviour:

• If a JobScheduler Master starts then it loads the configuration files and some information about the states of the objects from the database such as the job chain node of an order, if it is suspended, etc.
• If an order configuration file is changed in between two starts or the Master then the file beats the information from the database so that the database information gets lost and the order will be reset during a fail-over or when reactivating the primary JobScheduler Master.

Desired Behavior

• In a fail-over situation for a passive JobScheduler Master cluster it should not be required that configuration files have the same modification date.