The Evolution of Distributed DBMS 4Social and Technical Changes in the 1980’s u Business operations became more decentralized geographically. u Competition increased at the global level. u Customer demands and market needs favored a decentralized management style. u Rapid technological change created low-cost microcomputers. The LANs became the basis for computerized solutions. u The large number of applications based on DBMSs and the need to protect investments in centralized DBMS software made the notion of data sharing attractive.

The Evolution of Distributed DBMS 4Two Database Requirements in a Dynamic Business Environment: u Quick ad hoc data access became crucial in the quick- response decision making environment. u The decentralization of management structure based on the decentralization of business units made decentralized multiple-access and multiple-location databases a necessity. 4Developments in the 1990’s affecting DBMS u The growing acceptance of the Internet and the World Wide Web as the platform for data access and distribution. u The increased focus on data analysis that led to data mining and data warehousing.

The Evolution of Distributed DBMS 4DDBMS Advantages u Data are located near the “greatest demand” site. u Faster data access u Faster data processing u Growth facilitation u Improved communications u Reduced operating costs u User-friendly interface u Less danger of a single- point failure u Processor independence 4DDBMS Disadvantages u Complexity of management and control u Security u Lack of standards u Increased storage requirements

Distributed Processing and Distributed Database 4Distributed processing shares the database’s logical processing among two or more physically independent sites that are connected through a network. (See Figure 10.1) 4Distributed database stores a logically related database over two or more physically independent sites connected via a computer network. (See Figure 10.2)

Distributed Processing Environment Figure 10.1

Distributed Database Environment Figure 10.2

Distributed Processing and Distributed Database 4Distributed processing does not require a distributed database, but a distributed database requires distributed processing. 4Distributed processing may be based on a single database located on a single computer. In order to manage distributed data, copies or parts of the database processing functions must be distributed to all data storage sites. 4Both distributed processing and distributed databases require a network to connect all components.

What Is A Distributed DBMS? 4A distributed database management system (DDBMS) governs the storage and processing of logically related data over interconnected computer systems in which both data and processing functions are distributed among several sites.

What Is A Distributed DBMS? 4Functions of a DDBMS u Application interface u Validation to analyze data requests u Transformation to determine request’s components u Query-optimization to find the best access strategy u Mapping to determine the data location u I/O interface to read or write data u Formatting to prepare the data for presentation u Security to provide data privacy u Backup and recovery u Database administration u Concurrency control u Transaction management

Centralized Database Management System Figure 10.3

Fully Distributed Database Management System Figure 10.4

DDBMS Components 4Computer workstations that form the network system. 4Network hardware and software components that reside in each workstation. 4Communications media that carry the data from one workstation to another. 4Transaction processor (TP) receives and processes the application’s data requests. 4Data processor (DP) stores and retrieves data located at the site. Also known as data manager (DM).

Distributed Database System Components Figure 10.5

DDBMS Components 4DDBMS protocol determines how the DDBMS will: u Interface with the network to transport data and commands between DPs and TPs. u Synchronize all data received from DPs (TP side) and route retrieved data to the appropriate TPs (DP side). u Ensure common database functions in a distributed system -- security, concurrency control, backup, and recovery.

Levels of Data & Process Distribution 4Single-Site Processing, Single-Site Data (SPSD) u All processing is done on a single CPU or host computer. u All data are stored on the host computer’s local disk. u The DBMS is located on the host computer. u The DBMS is accessed by dumb terminals. u Typical of most mainframe and minicomputer DBMSs. u Typical of the 1st generation of single-user microcomputer database. Table 10.1

Nondistributed (Centralized) DBMS Figure 10.6

Levels of Data & Process Distribution 4Multiple-Site Processing, Single-Site Data (MPSD) u Typically, MPSD requires a network file server on which conventional applications are accessed through a LAN. u A variation of the MPSD approach is known as a client/server architecture. Figure 10.7

Levels of Data & Process Distribution 4Multiple-Site Processing, Multiple-Site Data (MPMD) u Fully distributed DBMS with support for multiple DPs and TPs at multiple sites. l Homogeneous DDMS integrate only one type of centralized DBMS over the network. l Heterogeneous DDBMS integrate different types of centralized DBMSs over a network. (See Figure 10.8)

Figure 10.8 Heterogeneous Distributed Database Scenario

Distributed DB Transparency 4DDBMS transparency features have the common property of allowing the end users to think that he is the database’s only user. u Distribution transparency u Transaction transparency u Failure transparency u Performance transparency u Heterogeneity transparency

Distribution Transparency 4Distribution transparency allows us to manage a physically dispersed database as though it were a centralized database. 4Three Levels of Distribution Transparency u Fragmentation transparency u Location transparency u Local mapping transparency Table 10.2

Distribution Transparency 4Example (Figure 10.9): Employee data (EMPLOYEE) are distributed over three locations: New York, Atlanta, and Miami. Depending on the level of distribution transparency support, three different cases of queries are possible: Figure 10.9 Fragment Locations

Distribution Transparency 4Case 1: DB Supports Fragmentation Transparency SELECT * FROM EMPLOYEE WHERE EMP_DOB < ‘01-JAN-1940’;

Distribution Transparency 4Case 2: DB Supports Location Transparency SELECT * FROM E1 WHERE EMP_DOB < ‘01-JAN-1940’; UNION SELECT * FROM E2 WHERE EMP_DOC < ‘01-JAN-1940’; UNION SELECT * FROM E3 WHERE EMP_DOC < ‘01-JAN-1940’;

Distribution Transparency 4Case 3: DB Supports Local Mapping Transparency SELECT * FROM E1 NODE NY WHERE EMP_DOB < ‘01-JAN-1940’; UNION SELECT * FROM E2 NODE ATL WHERE EMP_DOC < ‘01-JAN-1940’; UNION SELECT * FROM E3 NODE MIA WHERE EMP_DOC < ‘01-JAN-1940’;

Distribution Transparency 4Distribution transparency is supported by distributed data dictionary (DDD) or a distributed data catalog (DDC). 4The DDC contains the description of the entire database as seen by the database administrator. 4The database description, known as the distributed global schema, is the common database schema used by local TPs to translate user requests into subqueries.

Performance Transparency and Query Optimization 4The objective of a query optimization routine is to minimize the total cost associated with the execution of a request. The costs associated with a request are a function of the: u Access time (I/O) cost involved in accessing the physical data stored on disk. u Communication cost associated with the transmission of data among nodes in distributed database systems. u CPU time cost associated with the processing overhead of managing distributed transactions.

4Query optimization must provide distribution transparency as well as replica transparency. 4Replica transparency refers to the DDBMSs ability to hide the existence of multiple copies of data from the user. 4Most of the query optimization algorithms are based on two principles: u Selection of the optimum execution order u Selection of sites to be accessed to minimize communication costs Performance Transparency and Query Optimization

4Operation Modes of Query Optimization u Automatic query optimization means that the DDBMS finds the most cost-effective access path without user intervention. u Manual query optimization requires that the optimization be selected and scheduled by the end user or programmer. 4Timing of Query Optimization u Static query optimization takes place at compilation time. u Dynamic query optimization takes place at execution time. Performance Transparency and Query Optimization

4Optimization Techniques by Information Used u Statistically based query optimization uses statistical information about the database. l In the dynamic statistical generation mode, the DDBMS automatically evaluates and updates the statistics after each access. l In the manual statistical generation mode, the statistics must be updated periodically through a user-selected utility. u Rule-based query optimization algorithm is based on a set of user-defined rules to determine the best query access strategy. Performance Transparency and Query Optimization

Distributed Database Design 4The design of a distributed database introduces three new issues: u How to partition the database into fragments. u Which fragments to replicate. u Where to locate those fragments and replicas.

Data Fragmentation 4Data fragmentation allows us to break a single object into two or more segments or fragments. 4Each fragment can be stored at any site over a computer network. 4Data fragmentation information is stored in the distributed data catalog (DDC), from which it is accessed by the transaction processor (TP) to process user requests. 4Three Types of Fragmentation Strategies: u Horizontal fragmentation u Vertical fragmentation u Mixed fragmentation

A Sample CUSTOMER Table Figure 10.16

Data Fragmentation  Horizontal Fragmentation Division of a relation into subsets (fragments) of tuples (rows). Each fragment is stored at a different node, and each fragment has unique rows. Each fragment represents the equivalent of a SELECT statement, with the WHERE clause on a single attribute. Table 10.3 Horizontal Fragmentation of the CUSTOMER Table By State

Table Fragments In Three Locations Figure 10.17

Data Fragmentation  Vertical Fragmentation Division of a relation into attribute (column) subsets. Each subset (fragment) is stored at a different node, and each fragment has unique columns -- with the exception of the key column. This is the equivalent of the PROJECT statement. Table 10.4 Vertical Fragmentation of the CUSTOMER Table

Vertically Fragmented Table Contents Figure 10.18

Data Fragmentation 4Mixed Fragmentation Combination of horizontal and vertical strategies. A table may be divided into several horizontal subsets (rows), each one having a subset of the attributes (columns).

Table 10.5 Mixed Fragmentation of the CUSTOMER Table

Figure 10.19

Data Replication 4Data replication refers to the storage of data copies at multiple sites served by a computer network. u Fragment copies can be stored at several sites to serve specific information requirements. u The existence of fragment copies can enhance data availability and response time, reducing communication and total query costs. Figure 10.20

Data Replication 4Mutual Consistency Rule u Replicated data are subject to the mutual consistency rule, which requires that all copies of data fragments be identical. u DDBMS must ensure that a database update is performed at all sites where replicas exist. u Data replication imposes additional DDBMS processing overhead.

Data Replication 4Replication Conditions u A fully replicated database stores multiple copies of all database fragments at multiple sites. u A partially replicated database stores multiple copies of some database fragments at multiple sites. 4Factors for Data Replication Decision u Database Size u Usage Frequency

Data Allocation 4Data allocation describes the processing of deciding where to locate data. 4Data Allocation Strategies u Centralized The entire database is stored at one site. u Partitioned The database is divided into several disjoint parts (fragments) and stored at several sites. u Replicated Copies of one or more database fragments are stored at several sites.

Data Allocation 4Data allocation algorithms take into consideration a variety of factors: u Performance and data availability goals u Size, number of rows, the number of relations that an entity maintains with other entities. u Types of transactions to be applied to the database, the attributes accessed by each of those transactions.

Client/Server vs. DDBMS 4Client/server architecture refers to the way in which computers interact to form a system. 4It features a user of resources or a client and a provider of resources or a server. 4The architecture can be used to implement a DBMS in which the client is the transaction processor (TP) and the server is the data processor (DP).

Client/Server Architecture 4Client/Server Advantages u Client/server solutions tend to be less expensive. u Client/server solutions allow the end user to use the microcomputer’s graphical user interface (GUI), thereby improving functionality and simplicity. u There are more people with PC skills than with mainframe skills. u The PC is well established in the workplace. u Numerous data analysis and query tools exist to facilitate interaction with many of the DBMSs. u There are considerable cost advantages to off-loading application development from the mainframe to PCs.

Client/Server Architecture 4Client/Server Disadvantages u The client/server architecture creates a more complex environment with different platforms. u An increase in the number of users and processing sites often paves the way for security problems. u The burden of training a wider circle of users and computer personnel increases the cost of maintaining the environment.

C. J. Date’s 12 Commandments for Distributed Database 1.Local Site Independence 2.Central Site Independence 3.Failure Independence 4.Location Transparency 5.Fragmentation Transparency 6.Replication Transparency 7.Distributed Query Processing 8.Distributed Transaction Processing 9.Hardware Independence 10.Operating System Independence 11.Network Independence 12.Database Independence