Databases Reference
In-Depth Information
These can be different instances of the same DBMS, or could even be different
DBMSs, such as having SQL Server and My SQL running on the same computer.
Each instance, though running on the same hardware, is treated as a completely
separate database server. Each is managed separately, with separate configura-
tions, and using computer resources separately.
Using separate server instances should almost always be limited to a devel-
opment environment as a way of minimizing hardware when testing different
database servers or different server configuration parameters. However, there are
rare instances when you might use this configuration in a production environ-
ment. For example, you might need multiple database servers that need different
security configurations, but the resources requirements are such that they can
share a hardware platform without interfering with each other. As with running
different server applications, this configuration needs to be carefully monitored.
12.1.4 Consolidating Data Sources
Most database management systems are designed to support multiple user data-
bases. There are usually a specified maximum number of supported databases,
but this is usually large enough not to be a concern, allowing for hundreds,
thousands, or even more databases. The realistic limiting factors are available
disk space and memory. Each database will require a minimum amount of space,
and, even if the database is not currently being used, the DBMS will typically
set aside a portion of its dedicated memory for database support.
It's common to see a database server hosting multiple databases. It might be
that different databases are required for different divisions or departments within
a company. Your company's security guidelines may require that sensitive or con-
fidential data be stored on a database separate from other data. When supporting
a decision support system (DSS) based on a data warehouse, you will often keep
the DSS data and the production data on which it is based on the same database
server. As you've already learned, the term data warehouse refers to a shared data-
base containing accumulated data that is used to support management decisions.
In a strictly centralized data environment, one where all of an application's
data resides on a single database, the assumption is that each database is essen-
tially a stand-alone data store supporting its own unique application or set of
applications. At most, you need to support periodic data transfers, such as load-
ing a decision support data warehouse with data from a production database.
However, there is little or no direct interaction between the databases, at least
between the data in the databases, during normal operations.
When you do have interaction between the databases, you move into a dis-
tributed data environment. Because the data is spread across multiple databases,
even though they are located on the same physical server, you have many of the
same data support issues you would have if using
distributed database servers
(a multiple server configuration supporting related data).
