DATABASE MANAGEMENT SYSTEMS

Comparisons of Relational Database Management Systems



Data Base Management Systems are the backbone for running a computer operating machine. It is the whole system that a computer runs on. It is a very effective means of data management for the business organizations, in order to store their data and establish their entire working on those systems. There are a lot of data base management systems in the world today, but, a contrast and comparison has been done on the 3 most common data base systems such as Oracle Database 11g, Microsoft SQL server compact and MYSQL.

Oracle Database 11g

Oracle 11g fully automates the process in a daily maintenance task to provide an end-to-end automatic tuning solution. Sophisticated safeguards are built-in to prevent the tuning session from having a negative impact on the application workload. These capabilities complement each other to provide a unique solution that allows the system to automatically fix performance issues while protecting itself from harm (Schek, 2002). Oracle 10g introduced the SQL Tuning Advisor, a feature that performs a comprehensive root-cause analysis of one or more SQL statements and recommends a fix to the end user. When the performance issue is related to the execution plan, it can provide an application-transparent fix in the form of a SQL profile. This was a huge improvement for our customers, but it still involved several manual steps. DBAs had to find problem periods on their system, invoke the advisor after the problem occurs, and review and test the recommendations themselves. SQL Tuning is one of the most challenging and time consuming database management tasks for DBAs.

It requires an extensive breadth of knowledge, involves considering the merits of many possible solutions, and is an ongoing task that is never complete once and for all. Often none of the typical solutions are attractive because they either require rewriting a SQL statement, which necessitates an application patch, or changing the database configuration, which could have a negative impact on other SQL statements (Ramacher, 2005). The Oracle query optimizer uses object and system statistics to choose the best execution plan within the search space it considers. Best is relative to a cost model that provides a cost for various database operators such as hash-join. Object statistics include information such the number of rows and column distribution in a table. The accuracy of object statistics affects the quality of the execution plan selected by the query optimizer. In very large databases, tables are often decomposed into smaller pieces called partitions. For example, suppose a table line item contains line item records.

A DBA can partition the line item table on the ship date column so that line items that are shipped in the same month fall into the same partition. Table partitioning was first introduced in Oracle 8.0. Partitioning is used to improve the query performance, as well as the availability and manageability of the database. Reorganization is performed on tables as well as on ...