What is a Database Management System?

What is a Database Management System?


Database management is one of the essential skills computer science experts need to understand. A database, of course, is a structured data set organized with future retrieval or calculations in mind. Something as simple as a telephone directory can be considered a database. Some of the most basic electronic databases are organized using Microsoft Excel or similar programs.

In effect, databases may be thought of as the fundamental unit of information storage in digital information systems. As databases become more complex, however, it is necessary to introduce a more sophisticated method for managing, maintaining, and disseminating their contents. Without an effective database management system, it becomes impossible to keep data structured and accessible to those who rely on it.

An effective database management system (DBMS) provides best practices for all the major elements of the system:

  • Data
  • Hardware
  • Software
  • People
  • Processes

What Features are Needed in a Scalable, Secure Database Management System?

In an enterprise, a single DBMS solution needs to be able to account for all the organization, storage, retrieval, data security, and data integrity related to a function or division. Ideally, either all functional areas will use the same solution or all deployed solutions will interface seamlessly thanks to a set of interoperable, vendor-neutral standards.

There are several key features a DBMS must have to be effective:

Data Security

Data security functions must be integral to all operations in the database management system. On the most fundamental level, that means ensuring only authorized users can access, change, or reproduce the information represented in the system. This requires implementing the “Principle of Least Privilege” into the database by developing task-oriented subschema. A database can be split into an arbitrary number of subschemas, each one representing a subset of the total data based on its functional purpose. For example, in an employee database, some users may be able to access salary and benefits information while others may only access performance data.

Data Integrity

Without data integrity, personnel affected by the database cannot operate with full assurance that the information it provides is accurate. Data integrity measures must protect against both human error and technical problems. For example, user input verification should ensure that users can’t input the wrong type of data for a given entry. Likewise, change control should be in place to lock entries while in use and prevent multiple users from editing them at the same time. Some aspects of database integrity, such as regular backups, are outside the scope of the DBMS itself but are still crucial to efficient operation and continuity in case of an emergency.

Interactive Query Language

Each DBMS has its own interactive report writer, typically based on a common query language. In recent years, open source solutions for interactive query management have become quite common, reducing the licensing fees associated with database software and accelerating the pace of development. Unfortunately, open source solutions can be vulnerable to security issues, as any design flaws will be quickly publicized and exploited. Large enterprises managing sensitive data should consider encryption and other proactive security measures – in some compliance-focused industries, such as healthcare, such security is required by regulation.

Data Entry and Updating

Interactive data entry is the essential function that provides databases with their usefulness. However, not all database management systems offer the change control and process management that large organizations require to comply with auditing standards. Off-the-shelf solutions should be modified to provide the level of visibility an organization requires. This also allows for clearer understanding of how file systems and individual transactions interrelate, empowering a database administrator to optimize the system for greater efficiency based on actual usage patterns.

Data Independence

A database system might be accessed by dozens, hundreds, or even thousands of stakeholders. Each staff member accesses the information for their own purposes and may even use their own application software. As in object-oriented programming, a good DBMS essentially encapsulates the data and keeps its own inner workings opaque while providing each end user application with the expected functionality. A DMBS without data independence would be fundamentally useless, as each end user application would have to store large quantities of the data. Various versions of the database’s entries would quickly become mismatched, undercutting the system’s purpose as a centralized depository of records.

DBMS and the Importance of Efficient Database Design

The functional requirements of data independence lead to an important technological quandary that database administrators need to be aware of. End user applications need to be able to access data efficiently and effectively across many different platforms and configurations. As a result, the actual structure of the data itself will go a long way toward determining the level of performance that users can expect from the DBMS. A database with an inefficient design is responsible for introducing unnecessary complexity and, therefore, technical overhead.

It’s crucial to ensure that the design of a database is aligned with its intended use. Some options include:

  • Hierarchical: Hierarchical databases are the simplest and most fundamental, but aren’t suited to all applications. Essentially, these databases provide a single link between a record type and an individual owner. While these can be simple and efficient, they fall short when it comes to mapping complex, real-world structures.
  • Relational: Relational databases have rapidly become a popular solution for modeling of sophisticated systems—however, they require much more knowledge on the part of the database administrator. Relational records are not “physically” linked; instead, a common field must be represented in the record design to permit matching.
  • Network: In their simplest form, network databases are similar to hierarchical ones. They differ from the hierarchical model in that each data record can have multiple owners. This makes it more important that database administrators ensure no “cross-talk” when several users try to access a record. Likewise, owner and record names must be distinct.
  • Object Databases: Object databases are ideal for situations where the data system must model multiple many-to-many relationships. Although these can be complex from the technical and operational standpoints, they most closely match the internal workings of standalone programs and scripts in a variety of modern OOP languages.

As systems have been developed to correlate and manage larger, more complex data-sets, computer science experts with database management knowledge have become essential to many large enterprises. A fluent command of mathematical operations and formal logic can help to design and manage databases that are apt for their current use and adaptable to future needs. As enterprises grapple with the challenges of big data, it will become increasingly essential that database architecture incorporates best practices at every level.

New Jersey Institute of Technology offers an online master in computer science program so students can learn the latest concepts and techniques in Computer Science, and develop the skills to take the world in imaginative new directions.