Storage Virtualization: How Much Is Enough?

Storage virtualization is very different from the virtual world of the Internet, but its goals are similar. Like the Internet, virtualized storage aims for simplicity, transparency, flexibility, scalability, and resilience. But unlike the Internet, which is dispersed, storage virtualization is designed to allow centralized control. And while the Internet is maintained by the collective efforts of millions of individuals, a virtualized storage system is designed to reduce demands on support staff to the barest minimum.

While the Internet has been growing for decades, virtualized storage for open systems was introduced by Xiotech a mere eight years ago. It may seem ironic that, contrary to the normal scheme of things, virtualization was put to work in the vast cyberspace of the Internet so long before it appeared in the smaller arena of corporate systems. Part of the reason may be that the huge warehouses of enterprise data we now take for granted are a relatively recent phenomenon. They may, in fact, owe much of their origin to the Internet itself.

Information that, a decade ago, could not have been gathered at all or that would have taken so long to collect as to be of no use can now be collected almost instantaneously online. Data input that would have required slow, expensive key entry by error-prone clerks can now be pumped directly, accurately, and electronically into databases. And data that might once have languished, disconnected and therefore useless, can now be correlated by smaller, more powerful, faster processors and smarter software into information that is not only useful, but necessary if an organization is to survive in today's competitive markets.

In short, where just a few years ago companies asked themselves whether they could afford to keep massive amounts of data, they now ask whether they can afford not to. Furthermore, companies are now asking themselves how many copies of the data to keep—lest a single copy be lost via human error or natural disaster. Fortunately, while the amount of available data has been burgeoning, storage capacity has been growing (and storage costs falling) at an equally torrid pace. But raw capacity is not enough.