Background concepts

A relational database management system (RDBMS) must be able to respond to the requests of many concurrent users. An RDBMS must also maintain its transaction state while ensuring all transactional properties. Adaptive Server is based on a multithreaded, single-process architecture that manages thousands of client connections and multiple concurrent client requests without overburdening the operating system.

In a system with multiple CPUs, enhance performance by configuring Adaptive Server to use multiple Adaptive Server engines. In threaded kernel mode (the default), each engine is an operating system thread. In process mode, each engine is a separate operating system process.

All engines are peers that communicate through shared memory as they act upon common user databases and internal structures such as data caches and lock chains. Adaptive Server engines service client requests. They perform all database functions, including searching data caches, issuing disk I/O read and write requests, requesting and releasing locks, updating, and logging.

Adaptive Server manages the way in which CPU resources are shared between the engines that process client requests. It also manages system services (such as database locking, disk I/O, and network I/O) that impact processing resources.