Exploding amounts of digital information and an ongoing transition to a so-called "information economy" will necessitate a new, more holistic approach to storage, said speakers at the Storage Networking World conference in Dallas last week.
Just as businesses employ a chief financial officer to make money, save money, and stay out of legal trouble, organisations will need an executive focused on the risks and rewards of handling digital information, said Chuck Hollis, global marketing chief technology officer at EMC.
A good CFO leverages a company's economic portfolio as an asset, and "I'm starting to see more and more CIOs that see their job the same way, but around information", Hollis said.
Hollis was one of numerous speakers at the conference, hosted by Computerworld US and the Storage Networking Industry Association.
Information is fast becoming the world's "single most valuable asset", Hollis said. As such, it's important to oversee the entire information portfolio, understand where it's being stored, how it's being used, and to stay out of trouble by complying with all security and data retention regulations. Hollis also predicted that consumers are going to demand ever more detail about their personal information and where it's being stored, and will want as much control over that information as they have over their personal finances.
"I think the next five years are going to be more interesting than the last 20 years put together," Hollis said. "We're becoming an information economy."
More than 369 exabytes of information have been created since the beginning of 2008, more than double the amount in all of 2006, according to EMC's Digital Footprint Calculator.
Data creation is growing by 60% each year, in good economic times and bad, Hollis noted. Another speaker, executive vice president Madge Meyer of State Street Corporation in Boston, said her organisation acquires 40TB of new information each month.
The burden is falling heavily on storage administrators, particularly when it comes to securing all this information, said IT industry consultant Richard Austin.
"The information that has become the crown jewels of the modern enterprise is completely coming under the control and responsibility [of storage professionals]," he said.
Even as storage costs go down on a per-byte basis, businesses are spending more on storage each year because of data growth and power consumption. Diane Bryant, vice president and CIO of Intel, said the chipmaker predicts it will double its storage spending by 2012.
"It's daunting," Bryant said. That type of growth is "not something we have the ability to maintain or support. This is unsustainable".
Increasing complexity of devices, electronic forms of collaboration, the proliferation of the web as a business tool, and all types of regulations are contributing to the amount of information that enterprises must store, Bryant said.
Intel is creating information lifecycle management policies for all of its data, describing how quickly it must be accessed, how long it must be retained and when it should be deleted. Virtualisation and thin provisioning, the latter of which prevents over-allocation of storage, is also in the works, Bryant said.
Intel is further trying to cut costs by using tiers of storage, because not all applications need ultra-fast access to information. Solid-state disk, for example, is devoted to the applications requiring the highest performance and reliability, while optical tape is set aside for bulk storage, she said.
There are numerous roadblocks preventing optimal handling of data, from aging equipment to lack of mobility and low storage utilisation rates, noted speaker Jack Domme, chief operating officer of Hitachi Data Systems. Complicating matters is the poor economy, which has business executives demanding spending reductions.
Costs can be reduced significantly by improving data mobility with storage virtualisation, use of thin provisioning and lower-cost storage tiers, as well as de-duplication and archival of "stale" data that's rarely accessed, Domme said.
"We all know our utilisation rates of our assets, our data and our storage are fairly low," he said. "Everywhere I go [people say] the key is utilisation increase."
It's also crucial to avoid storing redundant copies of data, Wayne Adams, an EMC official and chairman of the Storage Networking Industry Association, said. This requires a close accounting of data and metadata with tagging and search capabilities, allowing all copies of a given piece of data to be deleted at the end of its lifecycle, he said.
While many of these measures can be deployed immediately, Xiotech CEO and president Casey Powell predicted a complete revamping of how storage is organised and allocated in the coming years.
The storage-area network won't provide the flexibility needed in the future, Powell said. Applications themselves must be able to control their own storage resources, automatically provisioning and de-provisioning storage as needed, he said. "We need a system that is integrated to the point that the application, in conjunction with the operating system, controls its own destiny," he said. This will reduce the potential of human error and let storage pros focus on higher-level tasks, he said.
Powell acknowledged this will be difficult, particularly when it comes to creating applications that can control unstructured data.
"If you allow humans to control the system, it becomes cumbersome to the point where it probably won't work," he said. Some applications are already self-provisioning regarding storage, but wholesale changes will take time.
"In 2020 we can come back here and see what's actually happened," Powell said in response to an audience member who asked for more details on how application-controlled storage will work. "The reality is we start small. ... We're not going to do a massive, across-the-board change in a short period of time."