Data explosion is one of the biggest issues facing IT today. The amount of data that organizations store has grown exponentially in the last 10 years. According to Gartner research director April Adams, data capacity on average in enterprises grows at 40 percent to 60 percent year over year.

Data is the lifeblood of any business, and companies of all sizes are struggling with the increasing amount of data stored on their networks. Because storage capacity has increased and costs have declined, many IT administrators have become more lax about what they allow their users to store on the corporate network and for how long. While the ability to store increasing amounts of data empowers organizations, it also presents them with the challenge of managing all of that information. As network storage grows, users are also adding an additional layer of complexity as they become increasingly dependent on ubiquitous access: they want to be able to access their data from wherever they are and from a variety of devices, including smartphones, tablets and laptops.

One approach is to just back everything up, but this tactic actually impedes your ability to get operations back up and running when a failure takes place. Going through mounds of unorganized data just isn't feasible and can cause companies to waste valuable time during a disaster. Businesses simply can't afford to treat all data equally, and prioritization is key. Companies may encounter serious issues if they store huge amounts of data onto tapes or into the cloud indiscriminately.

In sum, tougher recovery demands compound the problem of growing data. Organizations are intolerant of any data loss or downtime, putting a lot of pressure on IT managers, who are working in environments in flux thanks to evolving technologies and a growing variety of endpoints that need to be protected.

The 10 Percent Rule
Not all data is created equal. There is some critical data that, when lost, will bring a business to a halt. On average, only 10 percent of an organization's data is critical. "Critical" means that a file is in active use or changes frequently. That's typically about 10% of a company's information and represents the items they access daily and need immediately when a disaster strikes. Critical varies from organization to organization, but every minute spent recovering this data means lost productivity and lost revenue.

Of course, this doesn't mean that you don't need to protect the other 90 percent. It just means that you should prioritize. Arguably, all data is important, but organizations need a structured or tiered approach to ensure critical applications and systems are operational first in the event of data failure. They should plan and prioritize their information in advance, ideally with the help of professional data support personnel, so that they can recover information efficiently in the event of a disaster.

This approach will reduce downtime in the event of a widespread failure. If data is not prioritized, much time will be squandered recovering non-critical data, extending the length of a down period.

A Real Life Example
The benefits of a well-planned recovery strategy are best illustrated using a real world scenario. Let's consider a management consulting firm that has over one terabyte of data. Some of that data is Microsoft Exchange email, some resides on a file server and some of it is from a proprietary application for their business, which runs on a SCO UNIX server.

Using the 10 percent rule as a guide, the firm determines that if it were to experience data loss as the result of a server crash or other disaster, they would need to recover the last three months of their email, the last year of their file server data and the last three months of their UNIX data in order to get their business back up and running immediately. The rest of their data could be restored a day or two later without interruption to their productivity.

Armed with this information in advance, the organization uses a cloud-based backup vendor to design the backup and construct archiving rules to reflect their recovery time objective (RTO):

1. Local Storage for Instant Recovery

This firm has a dedicated network storage location, so their cloud vendor pushes a copy of the backups to this location while simultaneously sending encrypted data to its data center facility. Using local storage, the organization can restore files from the local copy over its local area network, making recovery as fast as a file transfer.

2. Time-Based Archiving Rules

In order to control the amount of critical data that remains in the cloud vendor's online vault and manage costs, they create rules that automatically push older data to archive after a specified period of time.

3. Delta Blocking for Short Backup Windows

Although the cloud vendor is protecting over 1TB of data for them, nightly backups usually run in under one hour, sometimes as fast as 20 minutes. This is due to delta-blocking technology, which identifies changes made to a file and backs up only those changes, rather than the entire file.

By designating which data needs to be restored immediately and which does not, the organization receives a customized backup and recovery strategy that fits their recovery objectives and cost requirements.

Conclusion
Putting together a comprehensive recovery strategy like the one outlined above requires a certain amount of expertise and lots of upfront planning. While the "set it, and forget it" mentality is very attractive, data is growing too quickly and technology is changing too rapidly for companies to simply entrust their backups to just any cloud provider. You may have access only to a written Q&A or a junior technology staff member reading from a script when you need help restoring your critical data. Recovery could take a long time if you try to bring back all of your data at the same time. That's why advance prioritization of data is so essential.

When disaster strikes, the last thing an IT administrator wants is to fill out online forms or talk to someone who's reading from a script. Companies need competent providers who know their data environment, understand their business needs and can help walk them through the process.