Research Report outlining the general database techniques - Essay Example

16). When a database is designed, one should make sure that all functions performed by the database are correct and quick. There are some performances issues that arise due to the structure and the design of the database. Hence, it needs to be designed and structured carefully. Large tables in the database need to be identified so that special consideration to these tables can be given in terms of performance. Also the database should be tested by increasing a number of users who access the database in order to check performance.

Database performance increases through normalization, database partitioning, using federated database systems and certain hardware considerations such as RAID. Normalization Normalization is a process of removing redundancies of data in a database or in other words it is a technique of producing a set of relations with desirable properties given the data requirements of an enterprise. (Stephens et al., 2008, p. 62). If normalization is not done, then certain anomalies occur which include the insertion anomalies, delete anomalies and update anomalies.

Anomalies occur when at one table of a database there has been an occurrence of change either an insertion of a record or deletion of a record or updating the table; however, unaffecting the duplicate records that exist in other tables in relation to that table. The process of normalization is done basically in three steps. The database is brought to its first normal form (1NF). A relation (table, entity) in a database in which the intersection of each row and column contains one and only one value is said to be in 1NF.

All relations are supposed to be in 1NF. After 1NF, the relations in a database are brought to the 2NF. A relation that is in its 1NF and all non-primary key fields is fully functionally dependent on the primary key is said to be in 2NF. Then a relation that is in 1NF and 2NF and in which no non-primary key field is transitively dependent on the primary key field is said to be in 3NF. These are the three basic steps that normalize the database. Database Partitioning Through database partitioning, database is broken into small physical units that can be managed independently.

In this way, it becomes easier for the designers and operations personnel as the database becomes flexible in data management. Large physical units, on the other hand, cannot be structured, recovered, organized or monitored. Partitioning makes indexing easier and also allows sequential scanning. Through this process, data can be broken up by organization, business unit, region, date and others which help in the desired analysis of data (Connolly and Begg, 2008, p. 119). Through partitioning, the performance of database is enhanced.

This is because when database is partitioned, subsets of data are formed. And these subsets can be managed and accessed with efficiency and are useful for analysis. Partitioning also makes the activities of maintenance operations such as defragmentation and index rebuilds. Partitioning mak