Relational Database Implementation

Moving from a conceptual data design to a physical database implementation is a critical process, benefiting from both conceptual relational principles and extensive knowledge of database vendor-specific features.

Assistance can be provided with execution or verification of the following tasks.

Normalization: domain/key normal form (a refinement of 3NF) is the goal for most situations.

Primary key selection: relies on an understanding of the relative benefits of surrogate vs. candidate keys, the strengths & weaknessses of multipart keys.

Indexing: much basic indexing is rote, but the creation of multipart key indexes and the optimal order of keys (for example) must be based on an appreciate of application requirements.

Denormalization: an unpleasant reality for many system, performance and programming considerations sometimes make this a requirement. Such data should always be maintained automatically via triggers if possible.

Referential Integrity declaration: sometimes regarded as an inconvenient luxury, declarative RI is the most important tool for guarding against application errors.

Security and authorization strategies: modern database offer an array of approaches to database object access control in addition to the traditional

Hardware selection & utilization: performance and reliability can both be optimized by appropriate allocation of database files across available devices, and appropriate use of technologies such as RAID.

SQL Development >>>