Database Denormalization

DATABASE DENORMALIZATION (Research Report

TABLE OF CONTENTS

INTRODUCTION

EXAMPLES OF DENORMALIZATION

2 .1 Conditions of Demoralization

3 . DENORMALIZATION APPLICATIONS

4 . CONCLUSION

5 . REFERENCE 1 . INTRODUCTION

Database developers often use normalization process to break raw databases into tables . It is an easy way to manage and organize the data . Normalization also guarantees that the data run through the database is accurate . The process reduces data redundancy . It helps to prevent the data from become functionally dependent so the database uploading is more efficient . However , normalizing database also

has its downside . Normalization could cause a diminishing in database performance . It takes many joins of tables to combine information . When number of joins increases , it takes more time to run the database querying which slows down the completion process

In to have a faster retrieval , sometimes the choice of using denormalization process is considered . The reason of denormalizating a data is to achieve a better and faster performance . Denormalization is the opposite of normalization . The process puts one fact in numerous locations . It speeds up data retrieval with data modification as an expense

Denormalization process can be done by fixing the table structures by creating constraints that synchronized the copies of redundant data in the normalized database . Taking down the normalization level one or two notch . Undo some steps in the normalization process , not completely Another way is by copying data within tables in to reduce the amount of tables that needs to be joined . The process allows controlling redundancy , while the joining increases during the database performance .2 . EXAMPLES OF DENORMALIZATION

A few examples of denormalizations techniques include storing the count of objects in multiple relations as in one relation , adding attributes from different relation to be joined , star schemas , OLAP cubes , and materialized views . Denormalization is not essentially required in performance improvement situation , because it is costly and expensive And it also takes additional efforts to keep track of related data . If the joins exceed than five or six tables , denormalization should do the work

2 .1 . Conditions of Demoralization

There are more conditions that indicate the need of denormalization important queries often rely upon data from more than one table and sometimes need online processing , a group processing is needed for group repetitions , applying calculations is needed to columns before answering the queries , accessing tables is needed in different ways by different users in the same time , primary keys slows in querying , and a large percentage queried by certain columns . Many different types of denormalizations may be used for different needs , such pre-joined tables , report tables , mirror tables , split tables , combined tables redundant data , repeating groups , derivable data , and speed tables

Denormalization , however , leaves a hazardous effect . Data redundancy is increased and application coding gives more complications because the data that have been spread is difficult to locate . And denormalization shows poor performance in inserting , updating , or deleting data , but improved in selecting or reading data

Denormalization should be taken only when normalization failed to give a satisfying...