By any calculation, the demand curve for data storage is rising off the chart. In each of the three years to 2008, the average user will have increased their requirements for holding data by 35% to 40%, according to storage industry analyst group Horison Information Strategies.

But IT managers widely acknowledge that the doubling of storage volumes every three years cannot continue, especially when much of the data stored is of little or no business value. Often the sheer volume of data available clogs business processes, hindering regulatory compliance efforts and dragging down corporate efficiency.

As Jon Toigo, CEO of storage consultancy Toigo Partners International (TPI), puts it: “Undifferentiated, unmanaged data is the bane of contemporary IT.”

Stemming the flow

Over the past three years information lifecycle management (ILM) has been widely touted as a strategy for taming that untrammelled growth of data. And yet more recently, storage vendors have chosen to focus on more exotic propositions, talking up such concepts as thin provisioning, content-addressable storage or de-duplication technologies.

Given ILM’s positioning as a corporate-wide strategy rather than a set of point solutions, the message has not always been well received. But that is short-sighted, says Toigo. “Current panaceas such as storage virtualisation, thin provisioning and de-duplication are just rearranging deckchairs on the Titanic.” To meet today’s business demands, strategic data management based on classifying and managing data via policy is essential, he adds. And that is something demonstrated by the 2008 Effective IT Survey.

Though ILM may still not be widely adopted (only 22% of the respondents have gone down that route), where companies have implemented it the impact has been overwhelmingly positive. Around 56% rated ILM as either effective or very effective (see graph).

What they are buying into is far reaching. Essentially, an ILM strategy recognises that storage spending has to be optimised. To achieve this, ILM technologies and processes move different data between storage devices depending on its changing value. So high-value, frequently accessed data is assigned to high-end arrays that are constantly available to servers, while lower-value data cascades down, perhaps ending up in a tape archive or being deleted altogether.

At US healthcare provider Wellmont Health System the results have been dramatic. ILM is helping drive improvements in patient care. Wellmont has, perhaps, a greater appetite for storage than most: its data archive has grown 1,500% since 2003, driven by its need to store high-resolution medical images such as X-rays, ultrasounds and MRI scans. Today, its storage system – spread across hospitals in North East Tennessee and South West Virginia – can hold a mind-boggling 542 terabytes of data.

That system also has to meet the needs of some demanding users. Doctors want to be able to compare and contrast new patient images with ones previously stored. The only way to build an affordable, high-performance storage network was to introduce the principles of ILM, to ensure the system was “scalable, affordable and tiered to protect all our critical information at the right service levels”, says Wellmont storage engineer Ray Lewis.

Classified information

ILM can potentially deliver further benefits. With environmental concerns inching up the business agenda, many IT managers are already under pressure to find ways of making storage more energy efficient. The most effective way to do this is through the implementation of ILM and tiered storage, says Mark Lewis, president of storage company EMC. Only data that is business critical should be stored on high-end arrays, he explains. Other data can be transferred to tiers that use slower-spinning disks, thus using less energy.

Nevertheless, it would be wrong to assume that implementing an ILM strategy will instantly cure all of today’s storage ills, says Barry Runyon, an analyst with IT advisory group Gartner. Today, many organisations have established storage tiers and invested in storage management tools. This has allowed them to implement some basic ILM practices, but many “find it easier and cheaper” simply to buy more storage arrays than to go through the painful task of data classification, he says.

Getting to grips with data

The hard part of ILM is what TPI’s Toigo describes as getting “up close and personal with our data”. Organisations’ vast pools of structured and unstructured data make it exceedingly difficult for them to really know what data they have, never mind determine the value of that data to the business. The extent of this challenge is underlined by the number of respondents to the Effective IT Survey that said they had no timeline for implementing ILM – around 59% said they had no plans to do so.

Help is on the way, though. A new breed of data classification tools from start-ups such as Abrevity, Arkivio, Kazeon, StoredIQ and Scentric are making the process of understanding corporate data more manageable, says Gartner’s Runyon. The approach varies; some focus on e-discovery or records management, while others take a data management or information governance approach. All of these tools provide a method for building metadata repositories, which can form the bedrock of an ILM strategy, he adds.

“However, determining the value and importance of data to the business cannot be done entirely by a software tool – it also requires the peculiar judgement of departmental stakeholders intimately familiar with the information that sustains and drives the enterprise,” Runyon says.

To date, some of the storage heavyweights, such as EMC and Hitachi Data Systems, have chosen to partner with data classification start-ups. Others, such as IBM and Symantec, are building their capabilities in-house. Much of this work sits apart from pan-enterprise ILM initiatives, adding to the sense that ILM is not for everyone.

Dangerous data

While the term ILM may be losing some of its cachet for IT industry marketing executives, some of the drivers behind its conception have only grown in importance – not least regulatory forces.

Sarbanes-Oxley and Basel II may have kick-started the first wave of interest in data governance but new legislation, such as MiFID and the changes in the US Federal Rules of Civil Procedures (FRCP) or the UK’s Civil Procedure Rules are reinvigorating corporate efforts to get to grip with the unwieldy volumes of data.

Landmark legal cases have already alerted business leaders to the risk of not knowing what data resides within their enterprise stores. But again, tactical implementations, in this case e-discovery tools, will help reduce corporate exposure. Indeed, IT industry watcher Forrester Research predicts that spending on e-discovery tools will rise from $1.4 billion in 2006 to $4.8 billion in 2011.

Such tactical approaches to information management may not initially look much like ILM, but in fact they follow the same principles of implementing processes and software to effectively manage the retention, migration and, ultimately, disposal of data. So while a significant proportion of organisations may not formally adopt ILM, they are working towards the same principles – and the real business benefits.

Further reading

Back to the Effective IT 2008 Report contents page

The ILM tug-of-war

From cradle to grave Find more stories in the Information Management Briefing Room