Datasheet

SQL and Relational Database Management Systems 1
Skills are a different story. Database expertise is a costly thing and usually is in short supply.
On average, Oracle expertise is valued a little higher than comparable expertise for Microsoft
SQL Server or DB2. The total cost of ownership (TCO) analysis released by vendors themselves
tends to be biased, so use your best judgment and do your homework before committing your
company to a particular vendor. Make no mistake about it — this is a long-term commitment,
as switching the database vendors halfway into production is an extremely painful and costly
procedure.
Support and persistence
One may ask, why spend thousands of dollars on something that can be substituted with a free
product? The answer is quite simple: For a majority of businesses, the most important thing is
support. They pay money for company safety and shareholders’ peace of mind, in addition to all
the bells and whistles that come with an enterprise-level product with a big name. (As the adage
goes: ‘‘No one was ever fired for buying IBM.’’) First, they can count on relatively prompt sup-
port by qualified specialists in case something goes wrong. Second, the company management
can make a reasonable assumption that vendors such as IBM, Microsoft, and Oracle will still be
around 10 years from now. (Nobody can guarantee that, of course, but their chances definitely
look better against the odds of their smaller competitors.) In addition, ‘‘free’’ products rarely
scale as well as the costlier products, and are rarely as manageable, as robust, or as clever at
optimizing wide varieties of queries over many orders of magnitude of data size. So, the less
expensive (and sometimes free) products by smaller database vendors might be acceptable for
small businesses, nonprofit organizations, or noncritical projects, but very few serious companies
would even consider using them for, say, their payroll or accounting systems.
We should mention that more and more companies are using open-source RDBMS
for certain tasks as the products become more reliable and more functional. Good
examples include Sony Online Entertainment switching to EnterpriseDB (an enterprise-class rela-
tional database management system built on PostgreSQL) or Google, Yahoo, and Ticketmaster
using MySQL for their key projects. Also, RDBMS tools vendors have started to provide support
for some popular open-source database products. For example, TOAD by Quest Software, a data-
base tool popular among Oracle developers, is now available for MySQL. In addition, some
serious companies specializing in software support now offer their services for certain
open-source products.
Major DBMS Implementations
One book cannot possibly cover all existing database implementations, so we’ve decided to con-
centrate on ‘‘the big three’’: Oracle; IBM DB2 for Linux, UNIX, and Windows; and Microsoft SQL
Server. These implementations have many common characteristics. They are all industrial-
strength, enterprise-level relational databases (the relational database model and SQL standards
are covered later in this chapter). They use Structured Query Language (SQL) standardized by
7