About four years ago certain industries such as BFSI and Petroleum saw large-scale IT deployments or massive transformations in IT infrastructure. This were driven mainly by a spurt in business and the need to increase the customer base through new services such as Internet banking.

This transformation started happening again last year, in the worst moments of the recession, when IT budgets dived to all-time lows. Suddenly everyone realized they had to get more out of the existing infrastructure—instead of doing the standard incremental upgrades. That called for a radically different approach to IT (explained at the end of this story) and the place to start is the data center.
There are a couple of other trends driving this transformation and eventually, it will lead organizations to the cloud, as we explain in the next story.
To understand this, you’d have to review some trends.

1. First, a problem with Infrastructure itself:

The old way of setting up infrastructure was like this: the organization first decided to implement an enterprise application such as ERP, CRM or messaging. Once brands were shortlisted, internal and external experts (consultants and ISVs) would compare these across various parameters.

The brand and version were selected based on specific business requirements and the application’s system requirements were carefully considered. The organization would buy best-of-breed products for compute, storage and networking, specific to that application. The application was then deployed and configured, to everyone’s satisfaction, within that department. If another department raised a request for a different application, the infrastructure was procured and deployed in a similar manner. Over time, the organization deployed many applications—each with specific infrastructure built around it. This is what is regarded as Infrastructure 1.0 (see illustration on previous page).
This application-centric approach created silos of incompatible infra-structures. It was not only expensive in the long run, but proved to be highly wasteful, and created many problems that had a negative impact on the top line. Since these infrastructures were usually built on different platforms it posed serious problems in the areas of capacity utilization and provisioning.

For instance, if just 30 percent capacity was used by a particular server hosting a certain application in a certain department, and another application was begging for compute power because its server was overloaded, the extra capacity could not be provisioned from an underutilized server, because of the heterogeneity of the environment.

2. Infrastructure 1.0 cannot address the exponential growth of unstructured data.

Creaky IT infrastructures (with older technologies) built 10 - 15 years ago cannot support today’s network traffic rates. Older network plumbing was built to support mainly textual data. Today, data exists in other forms such as video, audio and graphics. As this data seeps into an organization’s network, mainly through user-to-user exchanges, it slows performance down to a frustrating crawl. It also makes huge demands on storage resources.

3. The third trend is the need to cut costs and have tighter control of operational expenses.

Electricity and real estate costs are high. Low footprint, power efficient systems that require lesser cooling can reduce these overheads. Consolidation of infrastructure can also reduce costs significantly.

4. The fourth trend is the shortage of skilled IT manpower to manage IT resources.

This has resulted in demand for managed services and simplified management through unified consoles.

"By consolidating you can take an infrastructure approach to information rather than a tactical approach” - Manoj Chugh, President – EMC India & SAARC

5. The fifth trend driving the need for new IT infrastructure is an increasingly mobile and dispersed workforce.

Global teams collaborate on projects using Unified Communications (UC) tools over the Web. This is taking an organization’s IT infrastructure beyond its firewall and into the cloud, as we illustrate in the third cover story on UC. That also means that resources within the data center must be available 24/7. High availability necessitates certain levels of redundancy.

These five trends (among others) are driving major changes in enterprise infrastructure. In fact organizations are moving from the erstwhile application-centric approach to a service-centric approach. And all the transformations begin in the data center.

No wonder IT vendors such as HP, Microsoft, Cisco, EMC, Fujitsu and NetApp are spending millions of dollars to acquire data center expertise, and on setting up data centers worldwide. Telcos and ITES companies are also gearing up to offer infrastructure services.