Data Management Systems Overview | MGMT 329, Study notes of Data Communication Systems and Computer Networks

Download Data Management Systems Overview | MGMT 329 and more Study notes Data Communication Systems and Computer Networks in PDF only on Docsity! THE ROBERT O. ANDERSON SCHOOL AND GRADUATE SCHOOL OF MANAGEMENT The University of New Mexico MGT 329/637 Database Management Systems Overview Last Revised: 1/22/2008 6:59 PM Formal Definitions Database (DB) – Integrated collection of data and metadata Metadata – Data that describes other data Database Management System (DBMS) – Application or system software that stores and provides access to data stored in one or more databases Database Definition Revisited Database (DB) – Integrated collection of data and metadata. Integration implies that relationships among data items such as names, addresses, course titles, and grades are recognized and stored within the system For example: A particular paycheck was deposited to Burd’s bank account Burd enrolls in a particular course during a particular semester and receives a grade for that course The course Burd completed satisfies one requirement of a particular degree program Burd teaches a particular course and assigns grades to all students enrolled in that course Metadata Definition Revisited Metadata – Data that describes other data Some types of metadata: Single and multiple data item value constraints – for example: Salary must be a positive number Allowable grades are A, B, C, … If grade is A then points earned is 4.0 If student status is “suspended” then currently enrolled credits must be zero Data naming and organization – for example, the data item Name is part of the Person table and is indexed alphabetically to speed alphabetic searches. Computation of derived values – e.g., how is grade point average computed from grades earned Security – What users can view, change, or delete grades and salaries? Location – Where is a particular group of data items stored (on what disk and machine)? If they’re stored in multiple places, which holds the “original” or “master” and which are the copies? Review Exercise Pick a type of business or organization and consider the data that it stores and uses in ordinary operations List 3 physical things about which it stores data Describe 3 relationships among those things that an “integrated collection of data” should store or represent Describe 3 instances of metadata that should be represented within the database Review Exercise What’s the catch? (What do all of the benefits on the previous page “cost”?) Trains of thought Friedman’s free lunch rule – there is no free lunch Fisher on genetics and adaptation – highly specialized organisms are efficient in their environment but can’t easily adapt to other environments The golden rule – s/he who has the gold makes the rules Database/DBMS - Prehistory Prehistory - before 1965: Data are stored in computerized files. Files contain records and records contain fields (numbers, characters, and strings) Computerized files are created for specific application programs and groups of application programs (for example, student grades, payroll, accounts payable) Connections among files across application areas are weak or nonexistent (for example, Burd has separate student, employee, and vendor ID numbers with different values and formats) Redundancy among files is high (for example, Burd’s address is stored in three different files which must all be updated if he moves). Database/DBMS History – First Era First era – 1965-1980 CODASYL Committee defines a standard commonly called “network database” Common data type definitions Pointers among records represent relationships Common access methods Several mainframe products are matched to the standard – IMS and IDMS become the dominant DBMSs Rapid computerization of business processes fuels demand for DBMSs – Most business applications use them by the early 1980s Standing at the Precipice Organizations are overwhelmed by the volume of data Data has value independent of its support for “ordinary” business processing (e.g., marketing analysis for trends and new products) Modern software views data and software as an integrated whole (the object-oriented view) but relational DBMSs treat them separately Relational DBMSs are a poor fit to modern software and applications – limiting the kinds of data that can be stored and the types of applications that can be feasibly built and maintained (e.g., genome analysis and genetically-customized drug design and production using a relational database?) The sheer volume of data stored in relational databases and the dependence of “everything” on those databases and their host DBMSs makes fundamnetal change almost unthinkable. What is a DBMS? DBMSs can be small-scale (e.g., Microsoft Access) or large-scale (e.g., Oracle). For medium- and large-scale applications, a DBMS is a large, complex, and expensive program Characteristics of that program include: Tightly bound to the operating system since both share responsibility for accessing disk storage, network I/O, and control of related hardware resources. Resource-hungry – industrial strength DBMSs require industrial-strength hardware costing thousands to tens of millions of dollars. Difficult to manage – requires an army of highly-trained specialists. Some DBMS Products Desktop Access (Jet) FoxPro Paradox Enterprise Oracle SqlServer Open Source: mySQL, PostgreSQL, Firebird Embedded SQL Server Mobile Open Source