company logo

Running a replication server

When running a replication server, clients are downloading the replication database when starting up the first time. While running, update transactions are sent to the server and updates from other clients are received from server.

When restarting a client, all update transaction made meanwhile will be loaded to the client database. The number of maximum transactions is defined for the replication database when enabling it. When reaching the transaction limit, the database must be re-enabled explicitly or implicitly.

Transaction limit

When the number of transactions becomes very large, starting up an application may take some time. Hence, the number of transaction stored for synchronization can be limited (default: 5000 ). When a replication client starts that cannot be updated by the transactions registered, a backup of the database is sent to the clients.

The backup is an ordinary database backup file and must be located on the server in the same directory as the database or in a separate

either a WEB folder which is defined as SERVER_URL in the server.ini file. The backup file name must correspond to the database file name with the extension replaced by .ozi (ODABA back-up format). It is suggested to run an automatic service, which creates a database backup each day (or night) to reduce the server burden at run-time.