In this section we will examine Oracle database, how this product can be suitable for our GIS projects.

Oracle products are extraordinarily careful with data: For safety, they write all changes to data in two separate places—first to a transaction log and then to the database data file. Support transactions, which guarantees that if a mistake or an error happens as part of a series of related database commands, every related change will be undone in reverse order (an action called a rollback) to preserve database integrity.

Extensive SQL support, transaction support, and scalability are three key features that distinguish Oracle from other products which are adequate for desktop work, such as Microsoft Access.

Elements that are important when you’re choosing database server software for mission-critical projects are…..

Support for security, backup management, replication, programmability, extensibility, fault tolerance, and manageability.

A commitment to open-source standards like Java and XML is another trend. Oracle has all added extensive Java features to its databases…

This move was realized by Microsoft, which was the first to include features for data warehousing and data analysis (also called online analytical processing, or OLAP) into its database, starting with SQL Server 7.0. Oracle and IBM later did the same.The latest innovation—offered in Oracle9i—is integrated data mining, which helps business people do advanced data analysis.

Most database servers are on a collision course with XML. The move to XML is a natural tie-in with a new approach to computing—deploying database-enabled Web applications running on application servers in place of standalone Windows programs. XML is causing database vendors to rethink their direction from the ground up, and the SQL language itself could well be on the way out in a few years, potentially to be replaced by an XML-based language called XML Query, now in development.

Clustering several database servers for increased speed and reliability is another significant development and was a major focus of the Oracle9i launch. Organizations that can tolerate database downtimes of only seconds per year can turn to clustering.

As well as considering cost, capabilities, and speed, keep in mind the need for specially trained staff to manage your database system. A pro at operating systems and network hardware who wings it on database design and index layout will probably create something that works fine for 1 user but totally melts with 100 users.