Database Design – Lecture 16 Distributed Databases

Lecture Objectives Distributed Processing and Distributed Databases Distributed Database Management System (DDBMS) Distributed Database Design

Distributed Processing Shares the database’s logical processing among two or more physically independent sites that are connected through a network. Note: data resides at only one site and is shared by other sites (“centralized”)

Distributed Databases Stores a logically related database over two or more physically independent sites. The sites are connected by a computer network. Note: database is composed of several parts know as database fragments. These fragments are located at several different sites.

Distributed Processing and Distributed Databases In a distributed database environment, the users do not need to know the name or location of each database fragment in order to access the database – transparent to the user Distributed processing does not require a distributed database but a distributed database requires distributed processing Both distributed processing and distributed databases require a network to connect all components

Lecture Objectives Distributed Processing and Distributed Databases Distributed Database Management System (DDBMS) Distributed Database Design

DDBMS Advantages Data are located near/at “greatest demand” site – improved performance Improved reliability – data replication Growth facilitation Reduced operating costs

DDBMS Disadvantages Complexity Cost Database design more complex

Distributed Database Management System(DDBMS) Governs the storage and processing of a single logically related database over interconnected computer systems in which both data and processing functions are distributed among several sites.

Distributed Database Management System(DDBMS) A DDBMS must have at least the following functions to be classified as distributed: - Application Interface - Validation - Transformation - Query Optimization - Mapping - I/O Interface - Formatting - Security - Backup & Recovery - DB Administration - Concurrency Control - Transaction Management - Computer Workstations (sites or nodes) - Network Hardware & Software - Communications Media

Distributed Database Management System(DDBMS) A DDBMS must have at least the following functions to be classified as distributed: Application Interface Allows the interaction with the end user or application programs and with other DBMSs within the distributed database Validation Able to analyze data requests Transformation To determine which data request components are distributed and which ones are local

Distributed Database Management System(DDBMS) A DDBMS must have at least the following functions to be classified as distributed: Query Optimization To find the best access strategy Mapping To determine the data location of local and remote fragments I/O Interface To read or write data from or to permanent local storage

Distributed Database Management System(DDBMS) A DDBMS must have at least the following functions to be classified as distributed: Formatting To prepare the data for presentation to the end user or an application program Security To provide data privacy at both local and remote databases Backup and Recovery To ensure the availability and recoverability of the database in case of a failure

Distributed Database Management System(DDBMS) A DDBMS must have at least the following functions to be classified as distributed: DB Administration To allow the Database Administrator to maintain the databases Concurrency Control To manage simultaneous data access and ensure data consistency across database fragments in the DDBMS Transaction Management To ensure that the data move from on consistent state to another – synchronizing transactions

Distributed Database Management System(DDBMS) A DDBMS must have at least the following components: Computer Workstations (sites or nodes) Form the network system Network Hardware and Software Components that reside in each workstation Allows all sites to interact and exchange data Communications media Carries data from one workstation to another

Distributed Database Management System(DDBMS) A DDBMS must have at least the following components: Transaction Processor (TP) Software component found in each computer that requests data Receives and processes the application’s data requests (remote and local) Data Processor (DP) Software component residing on each computer that stores and retrieves data located at the site

Distributed Database Environment

Lecture Objectives Distributed Processing and Distributed Databases Distributed Database Management System (DDBMS) Distributed Database Design

Distributed Database Design Designing for a relational data base structure does not change – start with a top down approach HOWEVER, need to consider the following as well: How to partition the database into fragments Which fragments to replicate Where to locate those fragments and replicas More frequently used fragments should be stored locally Fragments used by all users should be stored centrally

Distributed Database Design Data Fragmentation: Allows a single object to be broken into two or more segments or fragments Each fragment can be stored at any site on the network Data fragmentation information is stored in the distributed data catalog (DDC), from which it is accessed by the TP to process user requests

Distributed Database Design Types of Data Fragmentation: Horizontal Vertical Mixed

Distributed Database Design Types of Data Fragmentation: Horizontal The division of a relation into tuples (rows) Each fragment is stored at a different node and each fragment has unique rows Each tuple has the same attributes (columns) but the rows are fragmented

Distributed Database Design Example of horizontal fragmentation Original structure: 5th Edition Fragmented structure: Split by state 6th Edition

Distributed Database Design Example of horizontal fragmentation Resulting structure: Fragmented structure: Split by state 5th Edition

Distributed Database Design Types of Data Fragmentation: Vertical The division of a relation into subsets by attributes (column) Each subset is stored at a different node, and each fragment has unique columns – with the exception of the key column, which is common to all fragments Transaction issues here because same record may need to be inserted into two tables (part of record into 1 table and other part into another table). If only 1 insert is successful; end up with inconsistent data.

Distributed Database Design Original structure: 5th Edition Fragmented structure: Split by location 6th Edition

Distributed Database Design Original structure: 5th Edition Example of Vertical Fragmentation Fragmented structure: Split by location 5th Edition

Distributed Database Design Types of Data Fragmentation: Mixed A combination of horizontal and vertical strategies

Distributed Database Design Example of Mixed Fragmentation:

Distributed Database Design Example of Mixed Fragmentation:

Data Replication Storage of data copies at multiple sites served by a computer network Fragment copies can be stored at several sites to serve specific information requirements Can enhance data availability and response time Can help to reduce communication and total query costs

Replication Scenarios Fully replicated database: Stores multiple copies of each database fragment at multiple sites Can be impractical due to amount of overhead Partially replicated database: Stores multiple copies of some database fragments at multiple sites Most DDBMSs are able to handle the partially replicated database well Unreplicated database: Stores each database fragment at a single site No duplicate database fragments

Data Allocation Deciding where to locate data Allocation strategies: Centralized data allocation Entire database is stored at one site Partitioned data allocation Database is divided into several disjointed parts (fragments) and stored at several sites Replicated data allocation Copies of one or more database fragments are stored at several sites Data distribution over a computer network is achieved through data partition, data replication, or a combination of both

Distributed Database Design How is a distributed database managed? Distributed Data Catalog (DDC) Contains the description of the entire database as seen by the DBA Translates user requests into sub-queries (remote requests) that will be processed by different DPs DDC is distributed and replicated at network nodes (the location of a database fragment)

Examples of Distributed Databases Banking Account data distributed at each local branch Loan data distributed at each local branch Corporate data at head office (summarized branch information) Insurance Policy data with each branch Corporate data at head office

Examples of Distributed Databases Retail Inventory data distributed at each local store Employee Scheduling data at each store Corporate data at head office (summarized store information) Payroll data at head office Utilities Utility monitoring data at each location (I.e. nuclear station monitoring – air, water etc at each location) Corporate data at head office

Distributed Database vs Client Server Client/Server is really an architecture which models a computerized solution based on the distribution of functions between servers and clients. A client requests specific services from a server and a server provides requested services to clients Distributed processing could be one aspect of client/server architecture – data ‘centralized’ The DDBMS distributes data to different locations – could be used in a Client/Server architecture

Distributed Database Design Steps: Always start with a centralized view design Consider horizontal fragmentation of a centralized database Consider vertical fragmentation of a horizontally fragmented database Re-consider PK for all fragments of the database Define data replication rules (scenarios) Complete Design