|
TRIMILLENNIUM
MANAGEMENT : INFOTECH
Leading our Digital
future By Bharat Bhaskar
The advances
in managing large volumes of data in the 1980s, the development of
transparent mechanisms for interconnections, and the emergence of global
connectivity based on TCP/IP standards have provided us an opportunity to
manipulate and disseminate information spread across geographic areas and
propelled us into a new economic era. This New Economy-also called the
Digital Economy-which is driven by the Net promises to be the engine of
growth for at least the first quarter of this century. Only corporations
as well as nations that are going to adjust to this reality are poised to
be winners in the future.
The future of cyberspace will be shaped by 3
factors: advances in hardware; the growth of data which will lead to more
complex storage-, retrieval-, and access-techniques; and the Net's reach
and ease of access which will make more and more data available on-line.
The growth in the speed and the architecture
of computer platforms will continue to drive prices downward. Developments
in CPU-technology, memory-technologies, and systems-architecture will
deliver servers with 100 gigabytes of main memory and petabytes of storage
capacity in less than a decade at affordable prices. If present trends
continue, we shall soon be flooded with low-cost, on-chip computers with
several megabytes of main memory. By 2010, technology will enable
enterprises to store and provide access to information at the touch of a
button from any part of the world.
With the availability of these computing
platforms, companies will be able to store data on-line. Database-system
researchers will build prototypes to handle relational tables, structured
data, and unstructured data in a transparent and seamless manner. The Net
will become a medium for creating, publishing, and distributing knowledge
and will make it possible for anyone to publish large amounts of complex
multi-media content inexpensively. While the growth in bandwidth will only
accelerate the growth of content , the management of the growing content
on a Website will pose challenges.
Sophisticated e-Commerce applications will
need efficient ways of tracking customer-profiles, transactions, billing,
and inventory. web publishing, thus, will soon reach a threshold, where
the ease of content-creation and distribution offered by the Net will need
to be backed by the technology capable of managing large pools of content.
Many large content-providers, portals, and retailers have moved the
content management part to XML, a language capable of describing
structured data. XML also supports client-side data caches for updates,
leading to a set of distributed transaction issues: atomicity,
consistency, isolation, and durability (acid).
As we rush into the Cyber Era, organisational
information systems will also become more Net-centric through the use of
intranets and extranets. Net content-creators need tools that can quickly
build data-stores capable of managing structured and un-structured data,
and integrate legacy data as a part of the data-store inexpensively. This
places tremendous demand on database technology, which must be able to
automate the creation, management, search, and security of Net
data-stores. The integration of platform-independent, easy-to-use Net
technology with database technology will need the unification of data and
programme logic within the data store, scalable, plug-and-play database
management systems, a federation of database systems, new, smart, data
architectures, and the integration of structured and semi-structured data.
As hardware changes raise capabilities
through the roof even as they lower costs, there will be a proliferation
of on-chip computers embedded in smart cards, palm-tops, telephones,
set-top boxes, and other information appliances. But since servers with
50,000 concurrent users and petabytes of data-stores will be desirable and
affordable, DBMSs need to be able to scale down to the
information-appliance level, and scale up to the storage-server level.
Database servers running in varied environment must be self-managing and
reliable. As the integration of database and Net technology progresses,
the number of information appliances and database-servers will make human
administration infeasible, leading to plug-and-play database technology.
These self-tuning databases, depending upon the capabilities of the
appliance, will organise themselves for optimal performance.
Over the years,
companies have built a huge inventory of legacy data, residing in multiple
platforms and formats. With the increasing globalisation of enterprises,
the heterogeneity of data will continue to rise. Database-servers will
make it simpler to integrate the information from multiple databases,
rendering the Net a federation of databases. This will need new
query-processing architectures that are federal in nature. Also, the
meaning and interpretation of queries in a federation may be
context-sensitive. It may take a long time to execute under current
database technology which strives to provide accurate answers. In these
situations, it may be appropriate to offer coarse answers. The research
agenda of the database community is now addressing these issues.
To better manage information flows in
companies, many application-level frameworks address workflows. By
compiling workflow rules into a set of triggers in a database system, and
attaching them with the stored data, we can integrate information flows
with data. This will lend workflow processes the high level of security,
reliability, integrity, and concurrency offered by databases. The
development of XML, a hierarchical scheme-oriented language, is likely to
create a plethora of semi-structured documents. Despite efforts to address
the integration, XML is likely to create a new class of documents with
deeply-nested objects. The declarative data-manipulation language for XML
is one such effort.
The grand challenge of information management
is to provide an easy mechanism to store, organise, retrieve, analyse, and
correlate all organisational information on-line. In the era of global
competition, information will be key. To increase corporate memory,
organisational as well as other competitive information from the federated
servers needs to percolate down to all decision-makers. So, the
information management professional faces the challenge of readying his
organisation's systems to fit in these new technologies, and develop a
culture of leveraging knowledge for decision-making in this Knowledge
Economy.
Bharat Bhaskar is a
Professor at the Indian Institute of Management, Lucknow
|