Databases

The assigned article from Access Science about Database Management Systems (DBMS) gives information about characteristic features and use of database systems. A database consists of hardware, the computer system where the data is stored, accessed, and processed; software, the DBMS that allows users to perform operations on the data; data, the set of information; users, the people for whom the database was specifically created; and procedural components, the procedures users follow to perform certain application on data. A data model is necessary and important in order to allow users to visualize the data in certain ways without showing all the details about the data. The relational model is the most popular of all DBMSs where relations or two-dimensional tables form the logical structure of all data. Each relation in a row of a table must have a set of attributes that is called primary key and each key value must be unique and cannot be null. On the other hand, a foreign key is a column in a table with data coming from a column in another table that is a primary key. The article states that relational DBMSs are the most widely used among other DBMSs and it seems that will continue for a time.

The first article I read is "Emergence of the RDBMS", part three of a series of articles called A Brief History of Modern RDBMS IT Management. In this part of the series the author gives Codd's Twelve Rules, the thirteen rules proposed by the pioneer of the relational model Edgar F. Codd  explaining what is necessary for a DBMS to be RDBMS. He also gives the first companies claiming to make RDBMS such as Oracle and IBM with SQL, and later on Microsoft with SQLServer. To define a fully relational database all of those rules must be satisfied. However, rules 6,9,10,11, and 12 are difficult to satisfy and the author focuses on Rule 10: Integrity Independence. Data integrity is a big issue in practice and that rule is generally cannot be established in a DBMS to make it fully RDBMS. The article helped me understand the rules and especially the challenges with some of the rules in practice and according to those there are no fully relational database management systems available today.

The second article I read "Rethinking The Relational Database" is about the capabilities of relational databases and why they don't match the data generated by businesses today. The author talks about the two big IT issues, the requirements of RFID and Sarbanes-Oxley Act. Those requirements make companies store huge volumes of corporate data, in some practices from tens of terabytes to millions of terabytes daily. Relational databases with their transactional, dynamic and multi-user capabilities have excess functionality when what's needed is to only write-once and read-maybe data; and the cost increases in this way. The author gives the solution as using a flat file with a RDBMS, a file with records that have no structured interrelationship, such as a text file (.txt). She states that by applying an index to a flat file results in a very quick and accessible repository which will reduce the need for a high-end server hardware. Since the amount of data generated is huge and according to the author only 80 percent of the data stored in relational databases is never accessed after they are written, I assume the author's solution would be satisfactory.