©Silberschatz, Korth and Sudarshan1.1 Chapter 1: Introduction Purpose of Database Systems View of Data Data Models Data Definition Language Data Manipulation Language Transaction Management Storage Management Database Administrator Database Users Database Management System Structure

©Silberschatz, Korth and Sudarshan1.2 What is a Database? According to the book:  Collection of interrelated data  Set of programs to access the data  A DBMS contains information about a particular enterprise  DBMS provides an environment that is both convenient and efficient to use. Another definition:  A database is a collection of organized, interrelated data, typically relating to a particular enterprise  A Database Management System (DBMS) is a set of programs for managing and accessing databases

©Silberschatz, Korth and Sudarshan1.3 Some Popular Database Management Systems Commercial “off-the-shelf” (COTS):  Oracle  IBM DB2 (IBM)  SQL Server (Microsoft)  Sybase  Informix (IBM)  Access (Microsoft) Open Source:  MySQL  PostgreSQL Note: This is not a course on any particular DBMS!

©Silberschatz, Korth and Sudarshan1.4 Some Database Applications Databases touch all aspects of our lives:  Banking – accounts, loans, customers  Airlines - reservations, schedules  Universities - registration, grades  Sales - customers, products, purchases  Manufacturing - production, inventory, orders, supply chain  Human resources - employee records, salaries, tax deductions Anywhere there is data, there could be a database. Course context is an “enterprise” that has requirements for:  Storage and management of 100’s of gigabytes or terabytes of data  Support for 100’s or more of concurrent users and transactions  Traditional supporting platform, e.g, Sun Enterprise server, 2GB RAM, 10TB of disk space

©Silberschatz, Korth and Sudarshan1.5 Purpose of Database System In the early days, database applications were built on top of file systems – coded from the ground up. Sometimes this approach is still advocated. Drawbacks of this approach:  Data redundancy and inconsistency  Multiple files and formats  Difficulty accessing data A new program is required to carry out each new task  Data integrity problems Integrity constraints (e.g. account balance > 0) become embedded throughout program code Hard to add new constraints or change existing ones  Plus others…

©Silberschatz, Korth and Sudarshan1.6 Purpose of Database Systems (Cont.) Database systems offer solutions for the above problems. Database systems also support:  Atomicity of updates  Failures may leave database in an inconsistent state with partial updates E.g., transfer of funds from one account to another should either complete or not happen at all  Concurrent access by multiple users  Concurrent accessed is needed for performance  Uncontrolled concurrent accesses can lead to inconsistencies E.g. two people reading a balance and updating it at the same time  Data security Recoding this functionality from scratch is not easy!

©Silberschatz, Korth and Sudarshan1.7 Levels of Abstraction Physical level: defines low-level details about how a data item (e.g., customer) is stored on disk. Logical level: describes data stored in a database, and the relationships among the data. type customer = record name : string; street : string; city : integer; end; View level: defines how information is presented to users. Views can also hide details of data types, and information (e.g., salary) for security purposes.

©Silberschatz, Korth and Sudarshan1.8 View of Data An “architecture” for a database system:

©Silberschatz, Korth and Sudarshan1.9 View of Data, Cont. In general, the interfaces between the three levels should be defined so that changes in some parts do not seriously influence others. Carefully defining the interfaces helps enhance Physical Data Independence, which is the ability to modify the physical schema without changing the logical schema.

©Silberschatz, Korth and Sudarshan1.10 Instances vs. Schemas Similar to types and variables in programming languages Schema – defines the structure or design of a database  Analogous to type information of a variable in a program  E.g., a database might consists of information about a set of customers and accounts and the relationship between them  More precisely:  Logical schema: database design at the logical level  Physical schema: database design at the physical level Instance – a database and its’ contents at one point in time  Analogous to a variable and its’ value

©Silberschatz, Korth and Sudarshan1.11 What is a Data Model? The phrase “data model” is used in a couple of different ways. The phrase (use #1) is frequently used to refer to an overall approach or philosophy for database design and development. For those individuals, groups and corporations that subscribe to a specific data model, that model permeates all aspects of database design, development, implementation, etc. Current data models:  Entity-Relationship model  Relational model  Object-oriented model  Object-relational model  Semi, and non-structured data models Legacy models:  Network  Hierarchical

©Silberschatz, Korth and Sudarshan1.12 What is a Data Model, Cont? During the early phases of database design and development, a “data model” is frequently developed (use #2). The purpose of developing the data model is to define:  Data  Relationships between data items  Semantics of data items  Constraints on data items In other words, a data model defines the logical schema, i.e., the logical level of design of a database. A data model is typically conveyed as one or more diagrams. The type of diagrams used depends on the overall approach or philosophy (i.e., the data model, as defined in the first sense). This early phase is referred to as data modeling.

©Silberschatz, Korth and Sudarshan1.13 Entity-Relationship Model Example of an entity-relationship diagram: Widely used for database modelling. An ER model is converted to tables in a relational database.

©Silberschatz, Korth and Sudarshan1.14 Relational Model Example of tabular data in the relational model From a data modeling perspective, which approach is preferable? The ER model, or the relational model? Attributes customer-name customer-id customer-street customer-city account-number Johnson Smith Johnson Jones Smith Alma North Alma Main North Palo Alto Rye Palo Alto Harrison Rye A-101 A-215 A-201 A-217 A-201

©Silberschatz, Korth and Sudarshan1.15 A Sample Relational Database

©Silberschatz, Korth and Sudarshan1.16 Data Definition Language (DDL) Notation and program for defining a (physical) database schema (see page 129 for a more complete example): create table account ( account-numberchar(10), balanceinteger) Given a DDL file, the DDL compiler generates a set of tables A description of those tables is stored in a data dictionary:  Contains information from the database schema  Frequently referred to as metadata (i.e., data about data) Data storage and definition language:  Language in which the storage structure and access methods used by the database system are specified  Usually an extension of the data definition language

©Silberschatz, Korth and Sudarshan1.17 Data Manipulation Language (DML) Language for accessing and manipulating the data  DML is also known as query language Two classes of DML languages:  Procedural – user specifies what data is required and how to get that data  Non-procedural – user specifies what data is required without specifying how to get that data SQL is the most widely used query language  Usually referred to as a non-procedural query language

©Silberschatz, Korth and Sudarshan1.18 SQL Examples Find the name of the customer with customer-id : selectcustomer.customer-name fromcustomer wherecustomer.customer-id = ‘ ’ Find the balances of all accounts held by the customer with customer-id : selectaccount.balance fromdepositor, account wheredepositor.customer-id = ‘ ’ and depositor.account-number = account.account-number Databases are typically accessed by:  Users through a command line interface  Application programs that (generally) access them through: Language extensions to allow embedded SQL An application program interface (e.g. ODBC/JDBC) which allow SQL queries to be sent to a database

©Silberschatz, Korth and Sudarshan1.19 Database Users Users are differentiated by the way they interact with the system Naïve users – invoke application programs that have been written previously  E.g. people accessing a database over the web, bank tellers, clerical staff, ATM users Application programmers – interact with the system by making DML calls through an API, e.g., ODBC or JDBC from within a computer program Sophisticated users – form requests in a database query language, typically submitted at the command-line Specialized users – write specialized database applications that do not fit into the traditional data processing framework

©Silberschatz, Korth and Sudarshan1.20 Database Administrator (DBA) Coordinates all the activities of the database system; the DBA has a good understanding of the enterprise’s information resources and needs. DBA duties include:  Granting user authority to access the database  Acting as liaison with users  Installing and maintaining DBMS software  Monitoring performance and performance tuning  Backup and recovery According to the book, the DBA is also responsible for:  Schema definition  Storage structure and access method definition  Schema and physical organization modification  Specifying integrity constraints  Responding to changes in requirements These latter tasks are typically performed by a DB designer.

©Silberschatz, Korth and Sudarshan1.21 Transaction Management A transaction is a collection of operations that performs a single logical function in a database application The backup and recovery components of a DBMS ensure that the database remains in a consistent (correct) state despite failures:  system, power, network failures  operating system crashes  transaction failures. The concurrency-control manager in a DBMS controls the interaction among the concurrent transactions, to ensure the consistency of the database.

©Silberschatz, Korth and Sudarshan1.22 Storage Management The storage manager in a DBMS provides the interface between the low-level data stored in the database and the application programs and queries submitted to the system. The storage manager is responsible to the following tasks:  interaction with the file manager  efficient storing, retrieving and updating of data Note that the DBMS may or may not make use of the facilities of the operating systems’ file management facilities.

©Silberschatz, Korth and Sudarshan1.23 Overall System Structure Query Optimizer

©Silberschatz, Korth and Sudarshan1.24 Application Architectures Architectures:  Mainframe – client programs and DBMS reside on one platform.  Two Tier – client programs and DBMS reside on different platforms; clients connect to DBMS via an API such as ODBC/JDBC.  Three Tier – client programs, application server (or other “middleware”), and DBMS; clients connect to DBMS indirectly through the application server (also via an API). Typically used in web-based applications.  N Tier – recent generalization of 2 and 3 tier architectures.