In the configuration of the main engine, the tab ‘Backup Node‘ specifies the failure-proof path. The main engine writes a live status file ‘TAJSALIV.SVC‘ to the directory, and updates its timestamp in the frequency specified in the same tab.

In the configuration of the backup engine, the tab ‘Backup Node‘ specifies the failure-proof path as well. IT SHOULD POINT TO THE SAME PHYSICAL LOCATION AS THE FAILURE-PROOF PATH SPECIFIED BY THE MAIN ENGINE. If the backup engine cannot see the live status file being updated in the past X minutes as specified in the checking frequency, the backup engine will think the main engine is dead and start to take it over.

When the backup engine uses an SQL database as its backend, the backup engine periodically replicates the required SQL data from the primary SQL database to the backup SQL database.

When the backup engine takes over, it first checks to see if the primary SQL database is available. If the primary SQL database is available, the backup engine will use this. If the primary SQL database is not available — i.e. the machine has gone down or there is a network failure — the backup engine will use the specified emergency database.

If the backup engine is using the primary SQL database and the primary engine comes online, control will automatically be switched back to the primary engine. If the backup engine is using the emergency database when the primary engine comes online, control can only be switched back manually.