Biomedical Engineering Reference
In-Depth Information
In addition to SANs, there is a variety of other network-dependent database architectures. For
example, Network Attached Storage (NAS) is one method of adding storage to a networked system of
workstations. To users on the network, the NAS acts like a second hard drive on their workstations.
However, a NAS device, like a file server, must be managed and archived separately. A similar
approach is to use a Storage Service Provider (SSP), which functions as an Application Service
Provider (ASP) with a database as the application.
With the increased reliance on the Internet, outsourcing storage through Internet-based SANs and
SSPs is often used instead of purchasing huge servers in-house. The advantage of technologies such
as SANs and SSPs is that they can provide virtually unlimited storage as part of huge server farms
that may be located in geographically disparate areas. The downside is loss of control over the data
and archiving process, as well as the risk that company providing the service may fail, resulting in
the loss of valuable research and production data. In addition, like NAS, SANs and SSPs only address
additional storage space, not integration.
Database Management Systems
The database management system (DBMS) is the set of software tools that works with a given
architecture to create a practical database application. The DBMS is the interface between the low-
level hardware commands and the user, allowing the user to think of data management in abstract,
high-level terms using a variety of data models, instead of the bits and bytes on magnetic media. The
DBMS also provides views or high-level abstract models of portions of the conceptual database that
are optimized for particular users. In this way, the DBMS, like the user interface of a typical
application, shields the user from the details of the underlying algorithms and data representation
schemes.
In addition to providing a degree of abstraction, the DBMS facilitates use by maximizing the efficiency
of managing data with techniques such as dynamically configuring operations to make use of a given
hardware platform. For example, a DBMS should recognize a server with large amounts of free RAM
and make use of that RAM to speed serving the data. A DBMS also ensures data integrity by
imposing data consistency constraints, such as requiring numeric data in certain fields, free text in
others, and image data elsewhere. A researcher isn't allowed to insert a numerical sequence in the
space assigned for a nucleotide sequence, for example.
The DBMS also guards against data loss. For example, a DBMS should support quick recovery from
hardware or software failures. A DBMS can guard against data corruption that might result from two
simultaneous operations on a given data item. The most common example is prohibiting two users
from simultaneously manipulating the same data. In addition, a DBMS adds security to a database, in
that a properly constructed DBMS allows only users with permission to have access to specific data,
normally down to the level of individual files. Multi-level user password-protection schemes can be
used to allow only graphic designers to view intermediate graphic data, and those in marketing to
view only final versions. Using intranets that limit data communications within a predefined group of
workstations can add greatly to the security of a database.
Search WWH ::




Custom Search