The Problem of Self-Tuningdatabases Has Been Very Important For the Last 10 Years. Commercial Databasesystems Have Been Adding Features For a Long Time and Have Now Reached a Levelof Complexity That Makes Them a Difficult Choice As a Central Platform Forinformation Services or E-Commerce. It Is Critical That Database Systems Beeasy to Manage, Predictable In Their Performance Characteristics And,Ultimately, Self-Tuning. Self-Tuning Database Technology Is Bound to Build Onmathematical Models and to Provide Proper Engineering into Systemcomponents.Today’S Dbmss Should Be Knob-Free and Must Be Able to Adapt Asconditions Change, Without Human Intervention. an Autonomic Dbms Should Becapable of Managing and Maintaining Itself As Well. This Paper Examines Thecharacteristics That a Dbms Should Have In Order to Be Consideredautonomic.Optimization of Self-Tuning Models Strives to Minimize a Programmablegoal Function That Reflects the System Designer’S Preferences and the Systembehavior.