What is distributed database management system?

A software with a centralized system that handles the distributed database in such a way that all are stored in a single site is known as a database management system (DBMS). The idea of the distributed database system is the centralized database, the data is spread out on a distributed network among the cities each having its computer and data storage facilities. This distributed data management is considered a single logical database.

What is distributed database?

A distributed database is a database of single logic that is widely distributed physically across computers in many localities and these devices are interconnected using the links of data communication. A distributed database is a virtual database. The parts of the distributed database are stored at several discrete localities in several discrete real databases physically. The data can be accessed by the user from any place from anywhere in the network as though the data is saved in the locality of the user. DBMS packages include My Structured Query Language (MySQL), oracle, oracle pl/SQL, SQL Server, Postgres SQL, etc.

Objectives of the distributed database management system

The objective of the distributed DBMS is to control the distributed database so that the centralized database appears to the users. The idea of the distributed database system is the centralized database, the data is spread out on a distributed network among the cities each having its computer and data storage facilities. This distributed data management is considered a single logical database. When a person queries the database on the distributed network from anywhere it is not needed to know where the data on the network that is being sought is located. The user gets the result as the user issues the query.

Example: There are three departments in a college, which are computer science, civil, and electronics departments. If a user needs data from the computer science department, the query "select * from computer science" will list all the details of the computer science department.

Distributed database types

Homogeneous distributed database

• This distributed database has identical software on all sites.
• The database has the awareness of each other and corporate in the process of user request.
• This distributed database system is easy to design and manage.

Heterogeneous distributed database

• This distributed database has various schemes and products on different sites.
• In this database system, the components are not aware of one another.
• This database system provides only limited facilities during the transaction process.

Data storage in case of distributed database

For excessive data inherent distribution and increased availability at the time of failures, distributed database systems are used. The distributed database system designing involves the same process as for centralized database but there are some additional factors to be considered.

Replication of data

If the relation r is repeated multiple times, a copy of r is stored in one or more places. In rare cases, the copy of r is stored fully replicated which means it is stored in every site in the system. There are many pros and cons to replication.

Availability

If anyone's site fails to contain relation r, then it can be found on another site. Therefore, the system continues to process queries related to r despite failing in one site.

High parallelism

When there is high access to the relation r, then the queries involving r can be processed by several sites in parallel. More copies of r improve the chance of providing data for the sites which include r in the transactions. The movement of data between sites is reduced by using data replication.

High overhead on updates

The system should check the copies of r relations for consistency. If not, an error may occur. Whenever r is updated, the updated r must be uploaded to all sites along with its replicas resulting in increased or high-level overhead. Consider the banking system as an example whenever the account details are modified, then it must be replicated and updated on all the sites.

Fragmentation of data

In the distributed database, the logical unit of database distribution and allocations are defined. The database is divided into logical units called fragments and is stored in various sites. The tables are the simplest logic units. Types of fragmentation are,

Horizontal

The table is fragmented horizontally. It selects the subset of rows in the table. In this fragmentation, the table is divided horizontally by selecting suitable rows. Different sites can be assigned by these fragments in a distributed system.

Vertical

The table is fragmented vertically by only keeping certain attributes of the table. In this fragmentation, the table is divided vertically by selecting relevant columns. Every vertical fragment must have the primary key of the original table.

Hybrid

This fragmentation has both the features of horizontal and vertical fragmentation. By combining the select-project operation, each fragment can be specified. We can get the original table by applying operations like Union and natural join in the correct order.

Transparency

The user doesn’t have the requirement to know about the data management i.e. how the data is physically located or how the data can be accessed from various sites. The database system which has these characteristics are called data transparency which has several forms like,

Fragmentation

The fragmentation of relations that occurs is not required to be known to the users.

Replication

Each data is viewed as logically unique by the users. To increase the performance of the system and data availability the distributed system does the replication of the objects. The user should not be concerned with how many objects are replicated and where the replicated objects are being placed.

Location

The physical location of data is not known to the user. Any data should be found by the distributed database system until the user transaction supplies an identifier.

Transaction management fundamentals

Transaction accesses and updates different data items and it is the unit of program execution. Some of their properties:

Atomicity

Transaction operations must be completely reflected or none are reflected.

Consistency

The consistency of the database is preserved by executing transactions in isolation (that means no other transaction occurs concurrently).

Isolation

The execution of transactions occurring at the same time is unaware of others in the system.

Durability

Even if there is a system failure, the changes corresponding to a successfully completed transaction are stored properly in the database.

Features of distributed DBMS

• The distributed database management system is used for creating, retrieving, updating, and also deleting the distributed databases.
• In distributed DBMS, the database is synchronized periodically providing access techniques by virtue this makes a distribution transparent to the users.
• If the data gets modified from any place, it is updated throughout the system and the distributed database system also checks for the updating data.
• It facilitates data-intensive and simultaneous user accessibility.
• The distributed DBMS works for heterogeneous database platforms.
• The distributed DBMS maintains data confidentially in an integrated manner.

Context and Applications

This topic is important for postgraduate and undergraduate courses, particularly for,

• Bachelors in computer science engineering.
• Associate of science in computer science.

Practice Problems

Question 1: In which fragmentation the tables are divided vertically?

A. Horizontal

B. Vertical

C. Hybrid

D. None of these

Answer: Option B is correct.

Explanation: The table is fragmented vertically by only keeping certain attributes of the table. In this fragmentation, the table is divided vertically by selecting columns.

Question 2: ____ distributed database has identical software.

A. Heterogeneous

B. Homogeneous

C. Hybrid

D. None of these

Answer: Option B is correct.

Explanation: This distributed database has identical software on all sites. The database has the awareness of each other and corporate in the process of user request.

Question 3: ____ distributed database has various data schemes.

A. Heterogeneous

B. Homogeneous

C. Hybrid

D. None of these

Answer: Option A is correct.

Explanation: This distributed database has various schemes and products on different sites. This database system does not have the awareness of one another.

Question 4: A ___ is a multiple interconnected databases collection.

A. Distributed database

B. Non-distributed database

C. Heterogeneous database

D. None of these

Answer: Option A is correct.

Explanation: A distributed database is a multiple interconnected databases collection that is widely dispersed physically across different localities that converse through a computer network.

Question 5: The fragmentation of the table horizontally is a _____ in the table.

A. Subset of rows

B. Subset of columns

C. Union of rows

D. None of these

Answer: Option A is correct.

Explanation: The table is fragmented horizontally which is a subset of rows in the table. In this fragmentation, the table is divided horizontally by selecting suitable rows.