What to consider when protecting data

Losing access to key applications and critical business data can be devastating. With downtime costing companies from thousands to millions of dollars per hour in lost productivity and opportunity costs, it is imperative to keep the IT business infrastructure available at all times.

IT departments must meet the challenges of real-time business continuity, disaster recovery, compliance and governance requirements with higher service levels, fewer resources, and ever tightening budgets.

The need for a complete, high-performance, easy-to-use solution for protecting business-critical data is clear. Here are some of the key concepts to consider when you're evaluating data-protection solutions:

Recovery Time Granularity (RTG) dictates the data-protection tool's ability to recover data down to the day, hour, minute or second. The finer the granularity, the more control you have over the ability to recover usable data, so RTG is an important parameter for recovering from a logical failure. Should data corruption happen at 10 a.m. while data protection is occurring, for example, the data is also corrupted. A data-protection tool that provides RTG in seconds could recover the data to 9:59 a.m. and 30 seconds, providing a recovery point as close as possible to the logical failure.

Recovery Object Granularity (ROG) measures the level of object granularity a data-protection solution is capable of recovering. Object granularity may be a storage volume, a file system, a database table, a transaction, a mailbox or even a single message.

Recovery Event Granularity (REG) measures the capability of a data-protection tool to track specific events, such as the opening, closing or saving of a file. Additionally, REG measures the ability of a data-protection solution to recover a failed application or data set for such a specific event. For instance, REG measures the capability of a protection tool to recover to the saving of a specific file. The better the ability to recover to a specific event the better the data-protection tools' REG.

Recovery Consistency Characteristics (RCC) defines the usability of recovered data by the associated application. The ability of a data-protection solution to return consistent data depends not only on how data is captured and stored but also on the data type being protected.

In a true real-time data-protection solution -- one that supports robust granularity requirements -- a data object can be of any granularity, and may have hierarchy. Metadata-, time-, and event-indexing capabilities enable tracking of real-time continuous object history, locating missing information and delivering object recovery of different granularities.

Data-protection services must be more reliable than the applications they protect. When a data-protection service fails, the data service must fail over to another data-protection instance, such that an application would be continuously protected.

A protection tool also must be secured, such that individuals without the proper authorization cannot freely configure its policy. Unlike an application, a data-protection solution should not allow any individual to alter its protected data. Data history can only be purged by policies.

Most data-protection solutions require that the protected data be presented locally before it can be recovered to the primary storage. As businesses become more global, and government regulatory requirements more stringent, data-protection solutions must be able to support both LAN and WAN recovery location scope.

A true real-time data-protection solution avoids recovering inconsistent or corrupted data through the use of comprehensive journaling (real-time data, metadata and event) and continuous object-store indexing techniques. These elements are combined to preserve a true data history of the application with consistency marking to ensure application recoverability across multiple dimensions of time (second, minutes, hours and days).

In most industries, service-level agreements for data protection and recovery include no time for backup windows, no tolerance for data loss, and very little margin for recovery downtime. Careful selection of the right solution can save time, money and the business.

Joyce is Senior Director of Marketing for Asempra Technologies. He can be reached at mjoyce@asempra.com.

• Bookmark this page
• Share this article
  •  
  • facebook
  • slashdot
  • digg
  • Reddit
  • stumbleupon
  • linkedin
  • twitter
• Got more on this story? Email Computerworld
• Follow Computerworld on twitter

• Share
  Share this article

  •  google+
  • facebookfacebook
  • slashdotslashdot
  • diggdigg
  • RedditReddit
  • stumbleuponstumbleupon
  • linkedinlinkedin
  • twittertwitter
• print
• email
• Bookmark