Data Protection and System Recovery

By Tom Schmidt

Nothing tests an enterprise quite as thoroughly as a full-blown IT disaster. Cut off from its vital information, a company faces a dire situation almost immediately. After all, everything about a company -- its product development, sales, relationship management, marketing, competitive analysis, investor relations, finance, and human resources -- is managed through that information.

It's no surprise, then, that disaster recovery planning plays an increasingly vital role in keeping corporate information both secure and available. However, given the exponential growth rate of data volumes, shrinking backup windows, the demand for more effective change management, and the need for fast, reliable recovery, IT departments are finding themselves seriously tested in their disaster recovery efforts. Increasingly, today's enterprises are coming to the realization that they need to embrace solutions offering both best-of-breed data protection and best-of-breed system recovery.

This article examines how enterprises that do so are in a better position to dramatically minimize downtime.

The prohibitive cost of downtime
Just how much is at stake when critical business information is unavailable? Precise estimates vary. A 2005 study of 80 large organizations by Infonetics Research found that overall downtime costs averaged an astounding 3.6% of annual revenue. In another study, Forrester Research estimated the average cost of downtime for e-commerce sites at $8,000 per hour; at larger sites, such as eBay and Amazon, the costs soar to hundreds of thousands of dollars per hour. According to Contingency Planning Research, system downtime costs anywhere between $14,000 and $6,450,000 per hour.

And recent events leave little doubt about the consequences of such outages. The cancellation of 1,100 Christmas Day flights in 2004 by Comair because of a computer meltdown prompted its president to resign shortly thereafter. An eight-day system outage at Sunoco's Haverhill, Ohio, plant reduced its 2005 fourth-quarter results by approximately $2 million.

According to IDC, hardware failure, application-related failure, and human error are the leading causes of IT downtime, not natural disasters such as hurricanes or earthquakes. While the latter grab the headlines, everyday occurrences put business at risk.
In any event, the conclusion is unavoidable: extended downtime has potentially disastrous consequences for a business.

Continuous data protection
Traditionally, tape has dominated as the backup and recovery medium of choice. But as the cost of disk-to-disk backup decreases and the need to restore rapidly becomes ever more important, disk is increasingly finding its place in what IDC calls the "backup hierarchy."

That's due in part to the impact of today's regulatory climate. To give just one example, Section 409 of the Sarbanes-Oxley Act explicitly requires companies to make "rapid and current" disclosures concerning "material changes" to their financial conditions. A number of industry observers have interpreted "rapid and current" to mean 48 hours, noting that rapid recovery will be used increasingly as a measurement of the soundness of a company's records management.

Thanks to advances in disk-based backup technology, enterprises can transcend the limitations of traditional backup and recovery practices to take advantage of the following benefits:

• Speed In most cases, backing up to and restoring from disk is faster than using tape. Also, because disks are random-access devices, the drives can instantly start to transfer files, whereas tape drives require that the tape be loaded, accessed, and sequentially written.
• Flexibility The faster, more efficient storage capabilities of disk drives let administrators schedule more frequent backups, lowering the exposure from any data loss. Disks used as backup devices can also support simultaneous backups, restores, and duplication operations. This is impossible with sequential tape drives.
• Efficiency Disk-based backup's ability to write multiple backup jobs simultaneously to individual backup files (one file per job) provides both exceptional performance and storage granularity that tape solutions cannot approach even with the most sophisticated multiplexing, multi-threaded solutions.
• Expense Because disk-based backup increases the speed, flexibility, and efficiency of a data protection strategy, it also lowers the total cost of ownership of storage management.

Rapid system recovery
As important as data recovery is, it isn't sufficient to help a business resume operations if there has been physical corruption of a system or application, or loss of a server. Rebuilding systems from bare metal can take hours or even days.

That's because system recovery traditionally has been a very manual, intensive process. Steps include repairing the hardware and reinstalling the OS, applications, patches, system updates, and other system requirements. After these were completed, administrators would then need to try and reconfigure the system back to the exact state before the disaster. Several tests would be needed to ensure the system was ready for production. Then administrators would need to load the data onto the system from the last backup. Finally, the system would be introduced back into the live environment.

In contrast, today's recovery solutions combine the speed and reliability of disk-based, bare-metal Windows system recovery with new technologies for hardware-independent restoration. This allows organizations to perform a full system recovery in just minutes, and to create real-time system recovery points without disrupting accessibility. In addition, this delivers the flexibility to recover to dissimilar hardware. By removing the need to recover systems to the identical hardware where recovery points were created, this eliminates the cost of having to maintain duplicate hardware solely for the purposes of system recovery in the event of an emergency.

Additionally, administrators can perform system restorations even if there's no hardware available by restoring recovery points to virtual environments. When new hardware is available, or existing hardware is repaired, the servers can then be restored from a virtual machine back to a physical machine without impacting business continuity.

Conclusion
Traditionally, enterprises have struggled to recover whenever their IT operations were disrupted. All too often critical data and systems tools weren't interoperable. IT operations and security functions overlapped or had conflicting priorities. And problems could require dozens of vendors to fix. The result has been higher costs, slower response times, and an inability to achieve business objectives.

Better and faster backup and recovery are required because IT downtime has become less predictable given all the dynamics that can affect IT today. As they increasingly expose their networks to customers, partners, and suppliers, enterprises must be able to isolate a threat or outage and know precisely what steps to take to recover.

By deploying both best-of-breed data protection and best-of-breed system recovery solutions, enterprises are in a better position to keep their business up, running, and growing, no matter what happens.

Tom Schmidt writes frequently about information security topics. He has more than 15 years' experience as a writer and editor in high-tech publishing.