Unit 9 Database Design using ORACLE and SQL.PPT
- 1. Database Systems Slide 1 Database Design Asif Sohail University of the Punjab Punjab University College of Information Technology (PUCIT)
- 2. Database Systems Slide 2 Database Design • It is a process of creating a design that will support the organization’s objectives for the required database system. • There are two main approaches to database design: a) Bottom-up Approach b) Top-down Approach
- 3. Database Systems Slide 3 Bottom-up Approach (design by synthesis) • It begins with the fundamental level of attributes, which through analysis of the associations between attributes, are grouped into relations. • The process of Normalization is used to find the normalized relations based on functional dependencies between the attributes. • The bottom-up approach is appropriate for the design of simple databases with relatively small number of attributes. • This approach becomes difficult to use with a larger number of attributes, where it is difficult to establish all the functional dependencies between the attributes. Database Design
- 4. Database Systems Slide 4 Top-down Approach (design by analysis) • It starts with a number of groupings of attributes into relations that exist together naturally, for example, an invoice. • The relations are then analyzed through normalization. • One may start with the development of ER model, beginning with the identification of entities and relationships between the entities, which are of interest to the organization. • The ER model is then transformed into a set of relations using the appropriate rules for conversion. Database Design
- 5. Database Systems Slide 5 Phases of Database Design • Database design is an iterative process. • Database design is made up of three main phases: 1.Conceptual database design 2.Logical database design 3.Physical database design
- 6. Database Systems Slide 6 Conceptual database design • The process of constructing a model of the data used in an enterprise, independent of all physical considerations, such as the target DBMS software, application programs, programming language, hardware platform etc. • The data model is built using the user’s requirements specification document. Phases of Database Design
- 7. Database Systems Slide 7 Steps of Conceptual database design 1. Identify entity types 2. Identify relationship types 3. Identify and associate attributes with entity or relationship types 4. Determine attribute domains 5. Determine candidate, primary, and alternate key attributes 6. Consider use of enhanced modeling concepts (optional step) 7. Check model for redundancy 8. Validate conceptual data model against user transactions 9. Review conceptual data model with user Phases of Database Design
- 8. Database Systems Slide 8 Logical database design • The process of constructing a model of the data used in an enterprise based on a specific data model. • The logical data model is based on the target model for the database, such as the relational data model. • The technique of normalization is used to test the correctness of the logical data model. Phases of Database Design
- 9. Database Systems Slide 9 Steps of Logical database design 1. Derive relations for logical data model 2. Validate relations using normalization 3. Validate relations against user transactions 4. Check integrity constraints 5. Review logical data model with user 6. Merge logical data models into global model (optional step) 7. Check for future growth Phases of Database Design
- 10. Database Systems Slide 10 Physical database design • The process of producing the description of the database on the secondary storage; it describes the base relations, file organizations, and indexes used to achieve efficient access to the data. • Physical database design is tailored to a specific DBMS. • Logical database design is concerned with what, physical database design is concerned with how. Phases of Database Design
- 11. Database Systems Slide 11 Steps of Physical database design 1. Design base relations 2. Design representation of derived data 3. Design general constraints 4. Design file organization and indexes 5. Coding and Compression Techniques 6. Analyze transactions 7. De-normalization 8. Partitioning 9. Estimate disk space requirements 10. Design user views 11. Design security mechanism Phases of Database Design
- 12. Database Systems Slide 12 Design base relations • For each relation defined in the logical data model, we have a definition consisting of: – The name of the relation and its attributes – Key attributes – Integrity constraints • For each attribute, we have: – Its domain, consisting of its data type, length, and any constraint on the domain – An optional default value for the attribute – Whether the attribute can hold nulls – Whether the attribute is derived and, if so, how it should be computed? Physical Database Design
- 13. Database Systems Slide 13 Derived data • Whether a derived attribute is stored in the database or calculated every time it is needed is a tradeoff. The designer should calculate: – The additional cost to store the derived data and keep it consistent with operational data. – The cost to calculate each time it is required. • The less expensive option is chosen subject to performance constraints. • Derived attribute should be stored when: – A frequent query is made against it. – The DBMS’s query language cannot easily calculate the derived attribute. Physical Database Design
- 14. Database Systems Slide 14 Design general constraints • Constraints are designed to enforce business rules and the correctness of data. • The constraints defined at physical database design can greatly reduce the programming effort at the application level. • We can introduce constraints such as DEFAULT value, CHECK constraint, Referential Integrity, database triggers etc. Physical Database Design
- 15. Database Systems Slide 15 File organization and indexes • To determine the optimal file organizations to store the base relations and the indexes that are required to achieve acceptable performance; that is the way in which relations and tuples will be held on the secondary storage. • In many cases, a RDBMS may give little or no choice for choosing file organization, although some may be established as indexes are specified. • An index should be used against an attribute: – That is used most often for join operations. – That is used most often to access the tuples in a relation. – The columns frequently involved in ORDER BY and GROUP BY clause. – The column contains a wide range of values. • An index on primary key is called primary index and an index on non key is secondary index. Physical Database Design
- 16. Database Systems Slide 16 Coding and Compression Techniques • Coding techniques are useful for compression of data values, by replacing those data values with the smaller sized codes. • Consider the following relation: • EMP (EmpNo, Ename, Job, City,…….) • If we have large number of records, then the wastage of small amount of space in each record can lead to loss of huge amount of data storage space. • To avoid the above problem, we can use the following relations: – EMP (EmpNo, Ename, JobCode, CityCode,…….) – Job (JobCode, JobTitle) – City (CityCode, CityName) Physical Database Design
- 17. Database Systems Slide 17 Analyze transactions • To understand the functionality of the transactions that will run on the database and to analyze the important transactions. • For efficient database design, it is necessary to have knowledge of the transactions or queries that will run on the database. It includes the identification of frequent transactions, the transactions that are critical, peak load time etc. • We use the transaction information to select appropriate file organizations and indexes. • For transaction analysis, we can use a transaction / relation cross- reference matrix, which shows the operation a transaction performs on a certain relation. • The above information will be used to determine the indexes that are required. Physical Database Design
- 18. Database Systems Slide 18 Transaction / relation cross-reference matrix Transaction / Relation T1 T2 . . . . . . Tm I R U D I R U D . . . . . . I R U D R1 X X X X X R2 X X X X X X . . . . . . . . . . . . . . . Rn X X X X I = Insert; R = Read; U = Update; D = Delete
- 19. Database Systems Slide 19 Denormalization • Denormalization is a technique to move from higher to lower normal forms of database modeling in order to speed up database access. • A fully normalized database schema can fail to provide adequate system response time due to excessive table join operations. • Denormalization results in combining two relations into one relation. More specifically, in this step, we consider duplicating certain attributes or combining relations together to reduce the number of joins required to perform a query. • As a general rule of thumb, if performance is unsatisfactory and a relation has a low update rate and very high query rate, denormalization may be a viable option. Physical Database Design
- 20. Database Systems Slide 20 Denormalization • Denormalization Situation 1: • Many to many binary relationships mapped to three relations. – EMP (empID, eName, job, Sal) – PROJ (pjId, pjName) – WORK (empID, pjId, dtHired) • By de-normalizing these relations, we merge the WORK relation with PROJ relation. In this case it is violating 2NF and anomalies of 2NF would be there, but there would be only one join operation involved by joining two tables, which increases the efficiency. – EMP (empID, eName, job, Sal) – PROJ (pjId, pjName, empId, dtHired) Physical Database Design
- 21. Database Systems Slide 21 Denormalization • Denormalization Situation 2: • We have attributes that are almost always used together in application and they end up in different relations, then we will always have to perform a join operation when we retrieve them. • Example: Consider the following relation: – EMP (empID, eName, street, postcode, city) • The relation isn’t in 3NF, as it has a transitive dependency postcode -> city. • After converting the above relation into 3NF, we have – EMP1 (empID, eName, street, postcode) – EMP2 (postcode, city) • However, this would mean that we would have to do a join whenever we want a complete address for a person. In this case, we would settle for 2NF and implement the original EMP relation. Physical Database Design
- 22. Database Systems Slide 22 Partitioning • De-normalization leads to merging different relations, whereas partitioning splits same relation into two. • The general aims of data partitioning and placement in database are to – Reduce workload (e.g. data access, communication costs, search space) – Balance workload – Efficiency • There are two types of partitioning:- – Horizontal Partitioning – Vertical Partitioning Physical Database Design
- 23. Database Systems Slide 23 Partitioning Horizontal Partitioning • Table is split on the basis of rows, which means a larger table is split into smaller tables. • Now the advantage of this is that time in accessing the records of a larger table is much more than a smaller table. • It also helps in the maintenance of tables, security, authorization and backup. • These smaller partitions can also be placed on different disks to reduce disk contention. Physical Database Design
- 24. Database Systems Slide 24 Partitioning Vertical Partitioning • It is done on the basis of attributes. Same table is split into different physical records depending on the nature of accesses. • Primary key is repeated in all vertical partitions of a table to get the original table. • You can replicate a limited subset of a table's columns to other machines. Physical Database Design
- 25. Database Systems Slide 25 Estimate disk space requirements • This step is required so that the appropriate disk size can be selected. • In general, the estimate is based on the size of each tuple and the number of tuples in a relation. • The growth of the relation should also be taken into consideration. Physical Database Design
- 26. Database Systems Slide 26 Design user views • A View is a subset of the database that is presented to one or more users. • A view is often referred to as a virtual table. • Views are created to satisfy the requirements of multiple users in an efficient way. Physical Database Design
- 27. Database Systems Slide 27 Design security mechanism • To design the security mechanism for the database as specified by the user during the requirements specification. • Authorization Table is used to restrict access to data and to restrict the actions that the database users can perform. • Fernandez, Summers and Wood developed this conceptual model of database security. • It expresses the Authorization Rules in the form of a table or matrix that include Subject, Object & Privileges Physical Database Design
- 28. Database Systems Slide 28 Design security mechanism Physical Database Design OBJECT SUBJECT Student Course Result Faculty Entry Read Read Read Null Middle Read/ Write Read/ Write Read Read Admin All Update/ Delete All Read/ Write/ Update
- 29. Database Systems Slide 29 Design security mechanism • The Column Headings represent database objects, which may be tables, Views, Sequence, Indexes etc. • The Subjects are written on the left side of the table. It may be an individual or a group of users. • The Cell entries of the table specify the privileges. These include INSERT, UPDATE, DELETE, ALTER, SELECT • Once the Authorization table or Access Control Matrix is complete, the DBA grants the privileges accordingly. Physical Database Design
- 30. Database Systems Slide 30 Thank you for your attention. Asif Sohail Assistant Professor University of the Punjab Punjab University College of Information Technology (PUCIT) Allama Iqbal (Old) Campus, Anarkali Lahore, Pakistan Tel: +92-(0)42-111-923-923 Ext. 154 E-mail: [email protected]
Editor's Notes
- #5: Chapter 16 – 18 (Connolly)
- #7: Check model for redundancy (1) re-examine one-to-one (1:1) relationships; (for entity merging) (2) remove redundant relationships; (3) consider time dimension.
- #17: When Not to Create an Index The table is small. The columns are not often used as a condition in the query. Most queries are expected to retrieve more than 2–4% of the rows. The table is updated frequently. If you have one or more indexes on a table, the DML statements that access the table take more time. Class Management Note Null values are not included in the index. To optimize joins, you can create an index on the FOREIGN KEY column, which will speed up the search to match rows to the PRIMARY KEY column. The optimizer does not use an index if the WHERE clause contains the IS NULL expression.
- #19: When Not to Create an Index The table is small. The columns are not often used as a condition in the query. Most queries are expected to retrieve more than 2–4% of the rows. The table is updated frequently. If you have one or more indexes on a table, the DML statements that access the table take more time. Class Management Note Null values are not included in the index. To optimize joins, you can create an index on the FOREIGN KEY column, which will speed up the search to match rows to the PRIMARY KEY column. The optimizer does not use an index if the WHERE clause contains the IS NULL expression.
- #20: When Not to Create an Index The table is small. The columns are not often used as a condition in the query. Most queries are expected to retrieve more than 2–4% of the rows. The table is updated frequently. If you have one or more indexes on a table, the DML statements that access the table take more time. Class Management Note Null values are not included in the index. To optimize joins, you can create an index on the FOREIGN KEY column, which will speed up the search to match rows to the PRIMARY KEY column. The optimizer does not use an index if the WHERE clause contains the IS NULL expression.
- #21: When Not to Create an Index The table is small. The columns are not often used as a condition in the query. Most queries are expected to retrieve more than 2–4% of the rows. The table is updated frequently. If you have one or more indexes on a table, the DML statements that access the table take more time. Class Management Note Null values are not included in the index. To optimize joins, you can create an index on the FOREIGN KEY column, which will speed up the search to match rows to the PRIMARY KEY column. The optimizer does not use an index if the WHERE clause contains the IS NULL expression.
- #22: When Not to Create an Index The table is small. The columns are not often used as a condition in the query. Most queries are expected to retrieve more than 2–4% of the rows. The table is updated frequently. If you have one or more indexes on a table, the DML statements that access the table take more time. Class Management Note Null values are not included in the index. To optimize joins, you can create an index on the FOREIGN KEY column, which will speed up the search to match rows to the PRIMARY KEY column. The optimizer does not use an index if the WHERE clause contains the IS NULL expression.
- #23: When Not to Create an Index The table is small. The columns are not often used as a condition in the query. Most queries are expected to retrieve more than 2–4% of the rows. The table is updated frequently. If you have one or more indexes on a table, the DML statements that access the table take more time. Class Management Note Null values are not included in the index. To optimize joins, you can create an index on the FOREIGN KEY column, which will speed up the search to match rows to the PRIMARY KEY column. The optimizer does not use an index if the WHERE clause contains the IS NULL expression.
- #24: When Not to Create an Index The table is small. The columns are not often used as a condition in the query. Most queries are expected to retrieve more than 2–4% of the rows. The table is updated frequently. If you have one or more indexes on a table, the DML statements that access the table take more time. Class Management Note Null values are not included in the index. To optimize joins, you can create an index on the FOREIGN KEY column, which will speed up the search to match rows to the PRIMARY KEY column. The optimizer does not use an index if the WHERE clause contains the IS NULL expression.
- #25: When Not to Create an Index The table is small. The columns are not often used as a condition in the query. Most queries are expected to retrieve more than 2–4% of the rows. The table is updated frequently. If you have one or more indexes on a table, the DML statements that access the table take more time. Class Management Note Null values are not included in the index. To optimize joins, you can create an index on the FOREIGN KEY column, which will speed up the search to match rows to the PRIMARY KEY column. The optimizer does not use an index if the WHERE clause contains the IS NULL expression.
- #26: When Not to Create an Index The table is small. The columns are not often used as a condition in the query. Most queries are expected to retrieve more than 2–4% of the rows. The table is updated frequently. If you have one or more indexes on a table, the DML statements that access the table take more time. Class Management Note Null values are not included in the index. To optimize joins, you can create an index on the FOREIGN KEY column, which will speed up the search to match rows to the PRIMARY KEY column. The optimizer does not use an index if the WHERE clause contains the IS NULL expression.
- #27: When Not to Create an Index The table is small. The columns are not often used as a condition in the query. Most queries are expected to retrieve more than 2–4% of the rows. The table is updated frequently. If you have one or more indexes on a table, the DML statements that access the table take more time. Class Management Note Null values are not included in the index. To optimize joins, you can create an index on the FOREIGN KEY column, which will speed up the search to match rows to the PRIMARY KEY column. The optimizer does not use an index if the WHERE clause contains the IS NULL expression.
- #28: When Not to Create an Index The table is small. The columns are not often used as a condition in the query. Most queries are expected to retrieve more than 2–4% of the rows. The table is updated frequently. If you have one or more indexes on a table, the DML statements that access the table take more time. Class Management Note Null values are not included in the index. To optimize joins, you can create an index on the FOREIGN KEY column, which will speed up the search to match rows to the PRIMARY KEY column. The optimizer does not use an index if the WHERE clause contains the IS NULL expression.
- #29: When Not to Create an Index The table is small. The columns are not often used as a condition in the query. Most queries are expected to retrieve more than 2–4% of the rows. The table is updated frequently. If you have one or more indexes on a table, the DML statements that access the table take more time. Class Management Note Null values are not included in the index. To optimize joins, you can create an index on the FOREIGN KEY column, which will speed up the search to match rows to the PRIMARY KEY column. The optimizer does not use an index if the WHERE clause contains the IS NULL expression.