Database Management Systems CSE530a

1
Database Management Systems CSE530a

• C. David Butler, Pharm.D., M.B.A.
• cdbutler_at_wustl.edu
• Office Number 314-935-9824
• Office 221 Sever
• Office Hours 3-4 TuTh or call to arrange a time.

2
Introduction

• Course Overview
• Course Expectations
• Database Management Systems
• Hierarchical
• Object
• XML
• Relational

3
Course expectations

• By the end of this course, you should be able to
• Grasp and apply mathematical concepts supporting
  the relational database structure
• Design and implement a relational database
• Differentiate relational from object,
  hierarchical and xml data storage
• Describe the physical storage operations of a
  dbms
• Write optimal queries using structured query
  language, relational algebra and relational
  calculus
• Describe the transaction management process
• Methods used in this class will focus on
• Concepts vs cookbooks
• Hands-on project vs rote memorization
• Local Resources
• http//cec.wustl.edu
• http//webdev.cec.wustl.edu
• http//classes.cec.wustl.edu/cse530/2006/

4
Overview

• A broad field a microcosm of computer science,
  including
• Languages
• Data structures
• File storage
• Security
• Operating systems
• Load balancing
• User interfaces
• Areas we could cover include
• Data mining and data warehousing
• Expert systems
• Object-oriented databases
• Deductive reasoning
• Client/server applications
• Transaction processing
• Financial analysis
• Statistical analysis
• Enterprise resource planning
• Decision analysis

5
Areas we will cover

• Database Design
• How to describe a real-world situation in terms
  of a DBMS
• Data Analysis
• How to answer questions
• Concurrency, Robustness, Security
• How to share access fairly
• Efficiency and Scalability
• How to optimize
• Application
• Making a real-world application

6
Hierarchical

• A hierarchical data model, such as IBMs
  Information Management System (IMS) consists of
• Records
• A collection of field values providing
  information on an entity or relationship instance
• Records of the same type are grouped into record
  types
• Parent-child relationships
• A 1N relationship between two record types
• Fast for searching stored records within the
  forest D1P1E11,E12,E13P2E21,E22 D2BI1,BI2

Example
Department
Project
Budget Item
Employee
7
Object

• An object data model (i,c,v), such as in Versant,
  Objectivity or db4Objects, consists of
• Objects a concept or abstraction of something
  with meaning for an application
• Class a description of a group of objects with
  similar properties and behavior
• Values a piece of data
• Methods an implementation of an operation (a
  function or procedure) for a class
• Links and Associations physical or conceptual
  connections between objects

o1(i1,atom,Clayton) o2(i2,atom,St.
Louis) o3(i3,atom,Ladue) O4(i4,atom,5) O5(i5
,atom,Research) 06(i6,atom,08/01/2001) O7(i7
,seti1,i2,i3) O8(i8,tuple,ltDNi5,Dnumbi4,Locat
ionsi7gt)
8
Relational

• A relational data model, found in Oracle, DB2 or
  SQL Server consists of
• Relations a set of unique records
• Schema
• A description of data in terms of a data model
• Specifies the name of the relational model, of
  each field and of each field type
• Integrity constraints may or may not be present

9
Database Management Systems

• Database
• A self-describing collection of integrated
  records
• For relational dbms a collection of logically
  related data that contains
• Entities a distinct object
• Attributes a property of an object or
  relationship
• Relationships an association between objects
• Allows data abstraction
• Database management system (DBMS)
• An application providing program-data
  independence
• Software designed to assist in the capture,
  storage, manipulation and retrieval of data from
  a database

10
Describing and Storing Data a comparison of
terminology
11
When is a db relational?

• According to Codd, when it complies with rules
  affecting the following areas
• Foundational
• Structural
• Integrity
• Data manipulation
• Data independence

12
Foundational rules

• It must use relational operations to manage
  databases
• Low-level (single-record) languages must not
  circumvent integrity rules and constraints for
  queries

13
Structural rules

• All data are represented
• at the logical level
• by tables and values
• Theoretically updatable views must be supported

14
Integrity rules

• Null values representing missing values must be
  supported consistently across all data types
• Integrity constraints must be definable by the
  query language and stored in relations

15
Data manipulation rules

• All data are guaranteed accessible logically
  using the relational model
• The system catalog is represented in the
  relational model and accessible
• A sublanguage will allow data and view
  definition, data manipulation, integrity
  constraints, authorization and transaction
  boundaries
• Insert, update and delete of data required

16
Data independence rules

• Applications must be logically unimpaired by
  changes in storage or access methods
• Applications must be logically unimpaired by
  information-preserving changes to the base tables
• Applications must be allowed to be logically
  unchanged for physically centralized vs
  distributed storage changes