One of the main objectives when using UDS/SQL is to ensure that existing processors are utilized efficiently. This is advantageous not only to the UDS/SQL user, who directly benefits from the quicker response times and increased throughput, but also to all other applications and programs on the system, since the available CPU resources are used economically.
The UDS/SQL server tasks of the independent DBH accept requests from application tasks without any fixed association. Requests are maintained in a common request pool, from which they are retrieved equally by all servers and processed. Requests that have not been fully processed because they are waiting for the completion of an I/O, for example, are returned to the same pool and can then be completed by any server. No load control and distribution is required by the database administrator.
Every server task uses “multithreading”, i.e. interrupts any request that cannot be processed further, returns the request, and begins a new one. This method ensures high processor availability for the server tasks and thus guarantees excellent performance.