Mistakes done while programming the data base application, storing, modifying, updating, deleting or retrieving are sure to reflect on the performance and the quality of the data base and data base applications.
In this article we shall state about what kind of bugs can be introduced if the normalization of the data base has not been done properly. Below is a brief description of what all problems can be caused for a certain reason:
1. Poor Design and Planning
– Sufficient planning, solid foundation, blue prints etc are must needed while building a house or a database.
– Design gives an assurance that the desired application or software system is built and the foundation on which the application is being built will not falter in the future days.
– The data base is developed with so much thoughts, care and attention as per the needs of the data that it is going to handle.
– Development of a data base cannot be taken to be easy.
– If time is not given for planning how the data base is to be built, then it is likely to happen for the project to lose direction and wear off.
– Moreover, planning is ignored for just the sake of building the data base.
2. Ignoring the Normalization
– Normalization defines a set of methods for breaking down the tables in to their smallest constituent parts until and unless one and only one value is represented by each of them and the description of the represented by the table are given by all of its columns.
– It has been 30 yrs since we adopted process of normalization and it has formed the basis for the implementation of the relational as well as SQL data bases.
– It is a fact not known to many of us that structured query language was actually invented to be used on the data structures that have been thoroughly normalized.
– Normalization is not to be considered as a plot for annoying the application developers by the data base programmers rather it is a side effect that satisfies proper working of a data base.
– The nature of the SQL is quite additive.
– We mean to say that if the pieces and bits of data are available then it becomes quite easy to create a set of results or values.
– For example, using the FROM clause you can join one table to another table.
– As many sets of the data can be added so as to the produce the desired final set.
3. Poor Naming Standards
– Additive nature is extremely important for the performance as well as for easing the development process.
– In some cases, you have to use CHARINDEX, LIKE, SUBSTRING and so on for parsing out a particular value in a single column, here a break down occurs in the SQL paradigms making the data lose its search-ability by more and more margins.
– This indicates towards the importance of normalization for the better performance of the data base and its development.
But one question is always raised ‘how much normalization is enough?”
– 3rd normal form is considered to be the most essential however the higher normalization forms i.e., 4th form and 5th form really help, if you can learn to handle them.
– The time taken by the normalization is worth the results.
– It is quite horrifying to see the tables having repeating columns appended with numbers.
– However, it is usual to find that even the first normal form in not implemented properly.
4. Lack of proper documentation
5. Putting all domain values in one table
6. Using guide/ identity columns as the key
7. Lack of use of SQL facilities for protecting data integrity
8. Lack of use of stored procedures for accessing data
9. Building generic objects
10. Lack of efficient testing