Web-Based Database Technology

PDF version of this report
You must have Adobe Acrobat reader to view, save, or print PDF files. The
reader is available for free
download.

Web-Based Database Technology

by Kirk Woodward

Docid: 00011541

Publication Date: 1805

Report Type: TUTORIAL

Preview

Web-based database technology offers database services independent of
location and platform. Database-as-a-Service, or, in other words, database
services outsourced to the Internet, takes advantage of Web-based database
technology in order to free staff to concentrate on core functions.
Nevertheless, issues of control, interfaces, compatibility, and,
increasingly, security present challenges for an organization.

Executive Summary

[return to top of this
report]

The Web-based database makes Database-as-a-Service (DaaS), or database
outsourcing, possible. It offers potential advantages in platform and
location independence. Since database administration is a complex skill,
outsourcing the function onto the Web may be advantageous for an
organization that lacks depth of experience in database administration.

Nevertheless, along with the choice to use Web-based database technology,
the enterprise must make its own decisions about many related issues.
Among these are design, both of the database itself and of the overall
architecture of the enterprise including possible expansion; security,
both on the home and the remote site; and administration, since the
relationship between the remote database and the enterprise user is of
primary importance.

A split in web-based database types is found in the use of SQL
(Structured Query Language) and NoSQL, a catchall label for databases that
are otherwise structured, and are now widely available through major
database vendors.

The Web-based database market is active and growing with participation by
major as well as numerous smaller software vendors and with continuing
advances in technology. Ultimately, though, the organization is
responsible for its own data and must approach DaaS with care.

Description

[return to top of this
report]

A database is a collection of records or pieces of data organized by
software (sometimes known as a database engine) for the purpose of
allowing that information to be retrieved in different combinations
depending on the search model. These records can be stored in a database
on the Web and managed remotely by the enterprise.

This tutorial links two technologies, the Web-based database and
Database-as-a-Service, or database outsourcing. The linkage is logical
since DaaS exists only because of the World Wide Web (also often referred
to as the “cloud”) and because ordinarily if an organization does not want
to maintain its database operation in-house, it will turn to outsourcing
services.

A Web-based database (sometimes referred to as Database 2.0) is a
database that is located on a Web site and can (1) receive queries
addressed to the database across the World Wide Web, (2) process the data
to obtain results from those queries, and (3) make the results available
using a Web browser. Database-as-a-Service is a database that is
outsourced to a Web-based database that can be located onshore or offshore
but, in any case, is hosted on the Web site and not locally in the
organization’s own facilities.

However, it should be noted that an enterprise could locate a Web
database product on its own intranet rather than outsourcing it.
Also, although Microsoft’s Access database product is frequently
considered a consumer product for a local user, Microsoft encourages Web
use of its popular database as well.

An advantage of DaaS is that it reduces administrative burdens. The major
issue with DaaS is one of control, because the enterprise’s data –
critically important to its operation – is remotely located and subject to
the security concerns that are increasingly the subject of news reports,
as large data breaches occur.

The Web-based database offers a number of advantages to the enterprise.
It is platform-independent; any computer that can access the Web can
access the database. It is also location-independent, a particular
advantage for an organization spread over a wide geographical area.
Typically the Web-based database may use Web services that employ standard
Internet protocols like SOAP, WSDL, and UDDI.

There are a number of database models or ways of
configuring data in a database. Some of the best known of these are
the relational database, or a database built on tables of data;
the hierarchical database, in which data is arranged in a virtual
tree structure; and the network database, with an interlocking series of
relationships among data. Other models also exist. A basic understanding
of database models is important in order to know how to choose an
appropriate Web-based database and/or Database-as-a-Service.

Among the major database administration software programs and
manufacturers are Oracle Database, Microsoft SQL, IBM’s DB2, and Amazon
Relational Database Server (RDS). However, NoSQL, a term standing for
alternatives to traditional Structured Query Language (SQL) database
organization, is making some headway in the market. It should be noted
that “NoSQL” can stand for not using SQL as well, or for “Not only
SQL,” meaning a hybrid approach.

Current View

[return to top of this
report]

The Web-based database, along with DaaS, has expanded its market share as
the Web itself has grown and is an established business today. Two
principal variations are a locally held database that can access Web-based
database services from within itself, and an enterprise that can access a
remotely located database and its functions. In either case, the
technology has evolved to the point where it can be considered mature.
However, change is the nature of web based services, and new applications
may demand new database forms or functions.

The large software vendors all offer DaaS. Among them,
Microsoft’s SQL uses its SQL Server 2017 and IBM offers its
Cloudant NoSQL DB – IBM Cloudthat utilizes a file-based approach.

Oracle, with nearly half the database market (followed by IBM and
Microsoft), positions itself around its database offerings including MySQL
Cloud Service, a widely popular an open-source database that is
basically free. To further its advantage, Oracle in 2011 purchased Endeca,
a manager of unstructured data. Oracle was initially slow to develop its
own cloud-based database, and the DaaS phenomenon has presented Oracle
with increased competition, but with Oracle Database 12c and its open
source products, it appears to be maintaining its ground.

Other large vendors such as Amazon with its RDS (Relational Database
Service), DynamoDB, ElastiCache, Neptune, and RedShift products;
Salesforce.com; Enterprise DB; and Caspio all offer database products, and
numerous third party companies also offer enhancements for database
products from the leading vendors. In addition, there are numerous
other large and small vendors that offer DaaS. Some of these vendors are
more suited for the enterprise and some for personal use.

NoSQL products (using other than the tabular relations of Structured
Query Language, or SQL) include Oracle NoSQL, DynamoDB from Amazon, and
MongoDB’s widely used product. Although NoSQL is overshadowed by SQL in
market share, many analysts suggest that its importance is growing along
with the development of applications more suited to its use.

Security concerns are paramount in any computer network system and there
are a number of reports of attempted and successful intrusions against
remotely hosted databases, as well as security weaknesses in some remotely
hosted code. Methods of security breaches include poor security, lost or
stolen media, the accidental publishing of data, and hacking from both
outside and inside the vendor. Millions of database records have been
compromised. Based on current trends, it seems certain that this
situation will not only continue but will intensify.

Outlook

[return to top of this
report]

Much of the market success of DaaS will be linked to the rate of
outsourcing of IT in general. The increasing popularity of Service
Oriented Architecture (SOA) as a model for the enterprise also will
continue to encourage the use of the Web-based database because of the
platform-independent nature of SOA. On the other hand, since economic
difficulties continue to have an impact on IT spending, this factor
can be expected to have the same kind of impact on database spending as on
other areas.

On the technological front, the Web-based database will continue to
evolve. In particular, the technology will continue to add to the complex
data types it is able to support, as will administration tools. The widely
used SOAP bindings will be adapted to protocols besides HTTP, increasing
the flexibility available to the technology. In addition, it is likely
that vendor specific database products will evolve to accommodate
interoperability, reducing or ultimately eliminating the need for XML
schema definition information, utilizing dynamic Web services that will be
used only when called for.

Many vendors will increasingly stress the “friendly” nature of their
database interfaces, although such claims require investigation to see
whether the database management service is sufficiently robust for the
enterprise or is a more consumer-level product. It is also likely that
progress will be made in designing DaaS as a data warehousing service with
defined and reusable database types.

A number of vendors are beginning to offer specifically tailored database
sites, such as software lifecycle databases, expense tracking databases,
and client relationship manager (CRM) databases. The appropriateness of
these tailored databases to the enterprise will depend on the nature of
the business. Expansion and compatibility issues should always be kept in
mind.

Competition in the DaaS market will continue to be robust. Companies such
as Amazon with DynamoDB are attempting to gain share in the market,
although often with a slant toward the consumer rather than the
enterprise. The long-established companies are competing aggressively,
with Microsoft’s Master Data Management (formerly SQL Server Data
Services) using SQL Server 2017 as its database engine, and Oracle
maintaining its market lead through significant acquisitions and
aggressive marketing of its cloud product. Meanwhile, NoSQL products add
further complexity to the market, with an increasing market share. The
larger vendors are also promoting their own entries in the NoSQL market
niche.

Numerous combinations of databases and tools among Web database
components are likely to be developed, many of them ad hoc. The strength
of this trend is yet to be determined but appears likely to be
significant.

Recommendations

[return to top of this
report]

Because control issues are preeminent in the linked technologies of
Web-based Databases and database-as-a-Service, recommendations concerning
those technologies must specifically address control issues. An
experienced database administrator (DBA) is mandatory for this process.

The security and privacy of data is of paramount importance, since
storing data online and remotely means that the data is only as safe as
the hosting site makes it. Furthermore, external attacks are not the only
security danger faced by Web-based databases; the enterprise needs to know
that its data is also safe from improper actions by the service providers
who are hosting the service.

Even the size of the vendor is not a guarantee of security; the hacking
of databases of large companies and security flaws in online databases are
increasingly in the news. Vendor assurances of data security can no longer
be taken at face value.

One logical answer to this situation is encryption of the data. However,
encryption has its own costs, both in the price of the encryption process
and in processing time, which may degrade significantly depending on the
degree of encryption. Managing encryption keys is a specialty and must be
carefully administered.

In any case, the enterprise must carefully investigate the protections
offered by the Database-as-a-Service provider against both external and
internal intrusions on the data, for example the degree to which queries
themselves are encrypted and the degree to which the provider is equipped
to defend against malicious intrusions. Also of critical importance are
backup systems and the ease of retrieving backed up data.

The enterprise must not assume that its data is safe because its own site
is adequately secured. The site hosting the database must be adequately
secured as well.

Another important issue is the scalability of the database, since both
business needs and the nature of the data may change. A database provider
must offer alternatives both in the growth of size of the database and in
its structure. Case studies will be particularly useful in this regard,
demonstrating how these issues have been handled with other clients.

The organization will very likely want to keep its own in-house database
administrator, who will perform essentially the same functions as at
present, only on a remote database.

Database-as-a-Service may present design issues such as the separation of
database logic and business logic that must be handled by technical
experts.

Regardless of who hosts the database engine, the enterprise itself will
need an interface to the database – a way of entering and viewing
information – appropriate to its requirements. This interface should be
configurable and flexible enough to be redesigned for future needs when
they arise.

The database design should not be done ad hoc but should be based on a
strategic analysis of the needs of the enterprise, in particular because
it is possible that over the long run one database service may not be able
to accommodate all an organization’s needs. If this is the case, and if
additional external databases are needed, issues of compatibility will
arise, and designing interfaces between proprietary databases may be an
arduous task. Again, a skilled database administrator is a prerequisite
for this development work.

Service and support are of particular importance. If the organization’s
database is in constant operational use, support should be available on an
all-day-every-day (24/7/365) basis. Reliable performance measurements are
also essential, including metrics of the time the network is connected to
the database server per call, and the response time of the database
server. Management tools for web based database services are improving in
ease of use.

In the much longer range, the enterprise must consider the dependability
of the Web as a medium. Although the Web is so ubiquitous that many take
it for granted, according to current estimates it is growing in size by up
to 20 percent per year, which in the long run may pose risks. While
improvements in the Web’s infrastructure have to some extent kept pace
with this expansion, the highly anticipated boom in video conferencing and
informal video communications will further use up broadband available for
the Internet and could conceivably lead to slowdowns or performance
degradation, which could have serious implications for an operational
database.

Web Links

[return to top of this
report]

IBM: http://www.ibm.com/
Microsoft: http://www.microsoft.com
MySQL: http://mysql.com
Oracle: http://www.oracle.com/database/index.html

About the Author

[return to top of this
report]

Kirk Woodward is a technical writer. In addition to
project management, Mr. Woodward’s areas of expertise include enterprise
software, hardware systems, and the use of Internet resources.

[return to top of this
report]