Database Software Comparing Microsoft Access, SQL, IBM

¶ … Database Software

Comparing Microsoft Access, SQL, IBM DB2 and Oracle databases is presented in this analysis, taking into account the key features of ACID Compliance, Data partitioning, interface options, referential integrity, operating systems supported, and support for transactions and Unicode. Each of these factors is initially defined followed by a table comparing them across the database types.

Definition of Comparison Factors

At their most fundamental level, all databases have support for relational data models and the ability to index data through the use of a wide variety of taxonomies or organizational structures (Basumallick, Wong, 1996). Relational Database Management Systems (RDBMS) however all have the ability to manage transactions with the greatest efficiency given the design of these systems to support multiple transactions at once, running concurrently from each other. The characteristic of an operating system being able to manage thousands of concurrent transactions at the same time is often referred to as scalability (Politano, 2008).

The concepts of security and scalability are critical to the first comparison element in the matrix, which is the extent to which a given database is ACID-complaint. ACID compliance refers to the extent a given database has the attributes of data atomicity, consistency, isolation, and durability. All of these attributes taken together lead to a more scalable, secure and transaction-capable database over time. From Table 1, Comparing Microsoft Access, SQL, IBM DB2 and oracle databases, it is apparent that the enterprise-level databases have this attribute, while Microsoft Access does not. This can be attributed to the database design attributes of SQL, IBM DB2 and Oracle database technologies as each also supports data models with referential integrity and dedicated memory for each transaction (Basumallick, Wong, 1996).

The next factor in the analysis is partitioning. In database design there are three dominant forms of partitioning, including Range, List and Hash (Politano, 2008). Range partitioning takes a predetermined number of values and defines their values in the data dictionary to ensure that queries can be successfully completed on them once requested (Politano, 2008). List partitioning requires databases to have multiple forms of logic supported within them, including Boolean support for managing nested partitioning requests (Politano, 2008). Hash partitioning is a conditional attribute of a given data element or object which determines its membership in a given group or strata of the data (Basumallick, Wong, 1996). All of these aspects of data partitioning require multiple approaches to managing algorithms and logic workflows, including the ability to create revised data structures based on query results (Basumallick, Wong, 1996). Of the four databases in the comparison, only Microsoft Access cannot support all aspects of data partitioning, as it only supports range-based partitioning across all product versions. Data partitioning is also an area that continually is being augmented and strengthened through the use of new algorithms and developments (Stonebraker, 2010). Database software developers need to have dedicated engineering staffs and teams on these areas to keep pace with the continual improvements being made.