Establishing a disaster tolerant infrastructure can be a very complex and costly undertaking. Having the ability to recover all mission critical applications, retain all data without any loss and ensure that end users have access to these applications is extremely difficult. Fortunately, technology has advanced and costs have come down.

Since 9/11 and the 2005 hurricanes, business continuity and disaster recovery have come off the back burner. The events of September 11, 2001 didn't change everything - but they did bring into stark relief a collection of trends in business continuity/disaster recovery (BC/DR) that had been building since the 1990s: Businesses had become increasingly reliant on I.T. and communication links, while the Internet and, more recently, Web 2.0 were changing the way they interact, collaborate and conduct business.

At the same time, datacenter and communications technologies that enable continuous operations had improved, and they had become more affordable.

Among the painful lessons of 9/11 was the realization in many I.T. shops that they needed a more robust BC/DR plan. Boardroom-level directors and officers also began to pay increased attention to BC/DR, and many made it a CEO-level mandate to improve the survivability of the I.T. and communications systems.

This article explores these trends and their impact on today's approaches to BC/DR. The current focus, as we shall see, is on keeping continuity of information access and communications flows. This is why the most prominent new term in the field is "disaster tolerant" environments. We no longer want to "recover" and "continue" - we want to avoid outages and ensure that our systems and services continue running no matter what.

What It Used To Be Like

The bottom line in today's business environment is simple: Companies cannot afford to have downtime. But it wasn't always like this.

Fifteen years ago, most companies had only one datacenter. Typically, this was a hardened facility with no single points of failure, including redundant power and communications connections, backup power, etc. Being disaster tolerant in the early 1990s meant making arrangements for disaster recovery, by performing a nightly tape back-up and then shipping that off to a secure vault somewhere. True business continuity was rare, even for organizations that could fail over to redundant components within the datacenter.

If and when a disaster struck, tapes and personnel were shipped to a designated recovery center, usually a shared facility operated by IBM, Sungard or another regional provider. The tapes were loaded and the people went to work to resume operations as quickly as possible.

Source