Database management systems (DBMSs) are sophisticated software applications that enable businesses to centrally manage their data using a single computer. A centralized database management system (DBMS) refers to a database that is housed, accessed, and updated from a single location (Geeks for Geeks, 2022). This system is typically implemented in a business, institution, or organization with the intention of centralizing business operations. A centralized database system facilitates convenient data sharing and distribution amongst users over a local area network (LAN) or wide area network (WAN). An example of a centralized database management system is a mainframe computer (Database Town, 2022).

The distributed nature of data in various organizations makes a centralized database management system (DBMS) inadequate. In response, a novel class of database management software called distributed database management systems (DDBMS) was created to handle data from dispersed locations. A distributed database is a system that stores data in various databases spread over multiple locations connected over a network. (Rahimi & Haug, 2010). DDBMS are used to maintain and manage data from dispersed locations over a network of computers. A distributed database management system exposes a collection of diverse data stores to application users or developers as a single logical entity. The most fundamental idea of a DDBMS is location transparency, which means that the user is oblivious to the physical location of the data because the system functions logically as a single unit (Moore, 2018). Therefore, computer networks facilitate communication across multiple databases located at different physical locations (Geeks for Geeks, 2022). Data in a DDBMS is logically integrated and managed as if it were stored in the same place. To ensure data integrity, the DDBMS synchronizes data regularly, ensuring that modifications and deletes performed at one site are automatically reflected in data stored in another location. Some examples of distributed databases include Apache Ignite, Apache Cassandra, Apache HBase, Couchbase Server, Amazon SimpleDB, Clusterpoint, and FoundationDB.

What are the Suitable Applications used with Distributed and Centralized Database Systems?

Distributed database management systems (DDBMS) are frequently used in data warehouses because they enable the simultaneous processing and access of vast volumes of data by many users or database clients (Abid Ahmad, 2010). Distributed database management systems are the best choice for safe network data management, data confidentiality, integrity, and dependable information retrieval.

On the other hand, several websites continue to employ mainframe computers which are centralized database management systems to house production databases (IBM, 2010). New mainframe hardware and software technologies are well suited for online transactions due to their capacity to support many users and applications accessing the same data fast and without interference. Also, due to their security, scalability, and dependability, mainframe computers are critical for the safe and effective operation of modern information processing (IBM, 2010). For example, a bank’s client account information, for which transactions can be submitted from thousands of ATMs worldwide, could be housed on a mainframe. Enterprises use the mainframe because it can execute transactions on a vast scale (thousands of transactions per second) while allowing thousands of users and application programs to access various resources concurrently. Mainframe computers are also capable of managing terabytes of data while handling high-bandwidth data communications.

System Architecture for Distributed Database Management Systems (DDBMS)

Client-server, collaborating server, and middleware are the three architectural designs for a distributed DBMS that isolates functionality across distinct DBMS-related operations (Subramanian, 2021). A general outlook of a distributed database management system design is explained by these three architectural designs.

In a Client-Server system, there are one or more client processes and one or more server processes, and a client process can send a query to any of the server processes. This is a two-tiered architecture with functionality segregated into servers and clients (Shah, 2021). The server functions include data management, query processing, optimization, and transaction management. Client functions primarily concern user interfaces, while servers maintain data and process transactions.

Using a collaborating server system, we can conduct transactions spanning numerous servers with the help of a group of database servers, each of which can run transactions against data stored locally (Subramanian, 2021). When a server receives a query that references data located on other servers, it breaks the query down into smaller, more manageable subqueries, send them out to be processed, and then compiles the results to form an answer. The query should be decomposed using cost-based optimization, which considers both network communication and local processing costs.

The Middleware architecture is made so that a single query can span several servers without every database server having the capability of managing multisite execution strategies (Subramanian, 2021). This architecture is especially appealing when connecting many outdated systems whose core functionality cannot be enhanced. The goal is to have a single database server that can coordinate requests across numerous machines.

Centralized Database Management System Architecture

A centralized database resides on a single system, for instance, a mainframe computer. All updates and changes are made directly to this single system via a local area network (LAN) or wide area network (WAN)(Database Town, 2022). Organizations like colleges and banks mainly use the centralized Database.

What are the System Functions of Centralized DBMS and Distributed DBMS

One of the key functions of a centralized database management system (DBMS) is to provide distributed query processing by making available all of the networked computers that satisfy the needs of a single node (Database Town, 2022). There is just one centralized database management system where all the data is managed; thus, data consistency is obtained. The server is the computer that services the needs of all the other connected computers, while the clients are the computers that access the server. The server is a central computer that handles all the requests. Data in this management system is not duplicated, and redundancy is minimized. Given that with a centralized DBMS, data is stored in one location, access and processing speeds are slow (Geeks for Geeks, 2022). As several users attempt to obtain and modify data queries on the same server simultaneously, the central computer’s processing performance degrades. If a centralized database management system fails or data is lost, it cannot be restored without backups.

On the other hand, a distributed database management system comprises various databases linked together over a network and stored in different locations (Geeks for Geeks, 2022). A distributed database allows multiple users to manage data at the same time. As such, a single query can be executed against numerous local databases, and the combined results provide the impression of a single database. With DDBMS, maintenance is challenging as data is scattered across multiple sites. As a result, identifying data duplication concerns and determining how to keep the multiple nodes in sync is crucial. In the event of a disaster, distributed databases are excellent at preventing data loss. This is because users will still have access to other databases if one or more of the databases fails.

Example

Data warehouses often use distributed database management systems (DDBMS) because they allow for the simultaneous processing and access of enormous volumes of data from diverse sources  (Rahimi & Haug, 2010). As the quantity of data sources within large enterprises expands, so does the complexity of data warehouses. The usage of distributed database technology during data warehousing development aids in the simplification of the ETL component, which is a critical component of data warehousing (Abid Ahmad, 2010). A data warehouse gathers data from multiple sources at the same time. This information is extracted, transformed, and loaded (ETL) into the data warehouse. Numerous servers supply the data warehouse with structured and unstructured data from various sources linked by a network. As a result, distributed database management systems are widely used in data warehouse infrastructures.

References

Abid Ahmad, M. Z. (2010). Using distributed database technology to simplify the ETL component of data warehouse. Proceedings of WSEAS International Conference on Applied Computer Science (ACS’10). Iwate,           

Database Town. (2022). Centralized Database. Database Town. https://databasetown.com/centralized-database-functions-advantages/#:~:text=Centralized%20database%20management%20system%20is,institution%20to%20centralize%20the%20tasks.         

Geeks for Geeks. (2022). Difference between centralized Database and distributed Database. Geeks for Geeks. https://www.geeksforgeeks.org/difference-between-centralized-database-and-distributed-database/      

IBM. (2010). Mainframe concepts. IBM. https://www.ibm.com/docs/en/zos-basic-skills?topic=vmt-who-uses-mainframes-why-do-they-do-it

Moore, L. (2018). distributed Database. TechTarget. https://www.techtarget.com/searchoracle/definition/distributed-database   

Rahimi, S. K., & Haug, F. S. (2010). Distributed database management systems: A Practical Approach. John Wiley & Sons. https://books.google.com/books?hl=en&lr=&id=0TqLzJBGImcC&oi=fnd&pg=PR7&dq=Distributed+database+management+systems:+A+Practical+Approach&ots=c4nh8zYBGY&sig=7vgpyXxluJEssqZZmJoGAdJk1RA#v=onepage&q=Distributed%20database%20management%20systems%3A%20A%20Practical%20Approach&f=false

Shah, R. (2021). Architecture model of distributed database management system (DDBMS). Bench Partner. https://benchpartner.com/architecture-model-of-distributed-database-management-system-ddbms    

Subramanian, V. K. (2021). Distributed DBMS architectures. Medium. https://medium.com/@VijayKumarSubramanian/distributed-dbms-architectures-a79c1c1e449d