Storage Utility

During the late 1990s, the storage industry caught the same Internet fever that plagued the rest of the computer industry. The concept was deceptively simply: Instead of storing data the old-fashioned way on disks safely housed in the corporate computer room, companies were told that they could store all their data in offsite storage "utilities" and pay only for the actual capacity they used on a monthly basis rather than buying all that storage hardware and software. The data would come to them, from the storage utility, over all of those fat Internet pipes the telcos were investing in as they built-out the Internet. It was just like paying for electricity. The electric utility sold electrons. The storage utility sold data bits. Both were delivered over wires. The model was so compelling that even Enron got into the business.

Then came the Internet meltdown. Most storage utilities died a quick and horrible death, because, in the end, most wound up being housed withinand selling their services tothe very same facilities as other Internet services providers such as Exodus. As compelling as their model seemed during those "frothy" Internet days, the storage utilities failed to address one crucial issue: Storage is as much about data protection as it is about data availability. In addition to looking and acting like electric utilities (plug in, get data), the storage utilities also had to look and act like banks. The storage utilities could not convince enough big customers that their critical corporate data was not only safe from harm, but that they could actually get their data back if the utilities went under. David L. Black, CTO of EMC, puts it this way: "When outsourcing anything, you need to outsource enough of the total process; pulling out too little breaks things and leaves processes uncomfortable."

Bregman resonates with the balance of availability and protection. "Classical utilities, like power or water, are 'fungible'completely interchangeablebut your data and mine are not." Bregman goes on to explain that the ill-fated storage utility model is actually alive and well, but within companies as opposed to an outsourced service. There are significant economies and efficiencies being realized today through the consolidation of storage devices, physical centralization, and the creation of common management structures. This is good, but incremental progress, such as what you read about in Chapter 8. Focusing on achieving 10 percent improvements often misses the 80 percent gain opportunities. Real gain comes from leveraging the entire storage value chain through tighter relationships.

Might this echo what Black is driving at? If enough of the interbusiness processes, along with the data, are carved out and separated from others, perhaps the right economy is realized and without unacceptable risks. If you are a manufacturer tying your fabrication operation closer each year with your plastics and glue providers (and using joint IT tools for messaging, schedule coordination, inventory balancing, and so forth), it might be sensible to have much of those data-driven processes housed elsewhere.

Bregman points out, however, that there are significant challenges, particularly in larger organizations. "For large scale brick-and-mortar companies, outsourcing of computer utilities is somewhat disruptive because all of a sudden they have excess capacity, people and/or real estate and machinery, which creates resistive inertia," explains Bregman. In contrast, smaller companies, particularly start-ups, will more readily embrace an outsourced model. In fact, smaller companies may, as a result, have an improved competitive edge by being able to get to this new efficiency more easily and grow (and shrink) more rapidly. "In an Inescapable Data world, I don't actually care where the data is as long as I can get it back," cites Bregman. Inescapable Data and its networking are once again changing the grounds of competition.