MMIS 630 Database Systems – Final Exam
1. What are the issues in the normalization vs. performance trade-off? Discuss. Your essay should (1) define and describe the characteristics of normalization, (2) define and describe the characteristics of performance, and (3) define and describe what is meant by “performance trade-off”, (4) summarize your essay with recommendations and/or guidelines regarding use of normalization in database design.
When we take a close look in designing a database one may not be thinking of normalization and performance, we may just be thinking, “I have this major project due in several months, I do not have time to be considering about formal techniques and performance, I will worry about this later.” Though this may occur in industry, database designers need to keep in mind that utilization of scarce resources such as processor time, network utilization, memory, etc., are very critical and can have a significant impact on database access and retrieval. As I had previously mentioned in our forum, one of the main concerns to keep in mind is that in other fields being 2 or 3 times slower may not be noticeable and certainly may not have a long term affect but in the realm of computers, having a multiplier can be detrimental in our resultant. What does this mean? It would simply mean that 1) Your database is not consistent OR 2) Your database is lacking the performance. Finding the equilibrium in this case can be a task like working with two equations setting them equal to each other for all else to be equal.
What is Normalization? Normalization as stated in our text is, “A technique for producing a set of relations with desirable properties, given the data requirements of an enterprise.” What does this mean? It means that normalization is a process and it is often to perform a sequence of tests to see whether it satisfies or violates the requirements of a given normal form. The five normal forms that we are mainly concerned about would be 1NF – 5NF. These normal forms in database design are based on practical/functional dependencies among the attributes of a relation. These are the stronger forms and are know as the Boyce-Codd Normal form. As you see that we have 1NF – 5NF, note that the 4NF and 5NF are often used in very rare conditions. With all of this said, we can conclude that normalizations’ main objective is to identify relations based on primary and candidate keys and their functional dependency. It also aids database designers by presenting a sequence of steps to help reduce data redundancy and to avoid update anomalies. The anomalies that we consider are insertion, deletion and modification. In these categories a unique situation can arise where data was no inserted, deleted and/or removed or a not modified. In any case, these situations can have an adverse affect on our database performance, it can severely impact our bottom line. Using normalization eliminates and/or removes any of these consistencies that arise when database designers do not take the time to properly design their database due to lack of time.
Performance in the database perspective can be defined as the effectiveness and speed in a database performs a transaction. Of course this is going to be dependent as to how this database system will make use of its scarce resources. As mentioned in our text some of the variables that we should be concerned with is 1) the access time (I/O) cost involved in accessing the physical data on the disk, 2) the CPU time cost incurred when performing operation on data in main memory and 3) the communication cost associated with the transmission of data across the network. If you do not have your hardware setup properly either it be in a centralized system or distributed environment, it may not perform to your necessary standards in spite of your design. You need to have both equations of normalization and performance equal to each other in...
Please join StudyMode to read the full document