Solutions Has Your Back(Up)

Nov. 1, 2003
Have you ever lost all or part of your database to an electrical blackout, fire, a lightning strike or a severe computer system failure? It happens more

Have you ever lost all or part of your database to an electrical blackout, fire, a lightning strike or a severe computer system failure? It happens more often than you think and it can't be predicted. In answer to the problem, Springfield, Mass.-based Solutions by Computer recently introduced Fail-Safe, a data-recovery program designed for rental companies to archive system data in real or near-instantaneous time, depending on the option they choose.

Since most systems do data backups every night, a failure during the day results in loss of all data since the last night's backup. Fail-Safe creates a separate log of transactions on a set time interval throughout the day. So if a system failure occurs at 3 p.m., for example, combining the backup from the night before with the latest log file created that afternoon facilitates recovery. As a result, very little, if any, data is lost because of the failure.

Several different levels of coverage are available for customers to choose from with the Fail-Safe system.

The first option provides a stand-by system within the rental company that is networked to the main server so that all transactions and database changes are copied and sent to the standby database in real time. If a server fails, the standby is current and could become the main server in less than one hour.

This option is recommended for customers with a backup system available or who may be upgrading their old systems. In this case, the old system is well suited to serve as the standby.

The second option provides software to log all transactions and database changes automatically. This log file is then copied on a set time interval to a floppy disk in the server — if the system is no larger than four terminals. All systems larger than four terminals will copy the log file to a tape drive instead. Upon server failure, recovery consists of mounting last night's backup on a Swapback system and merging the log files stored on the backup media.

“We're always looking for ways to minimize downtime,” says Jack Shea, president of Solutions. “That's why we implemented this Swapback program. If we can't resolve the problem that day, we'll ship out a new server to be delivered the next morning via FedEx.”

The Swapback server is preloaded with the company's system configurations so when it arrives all they need to do is plug in the cables, restore the backup and return to business as usual.

The third option provides a standby system located at Solutions' headquarters. The data log file is sent there via Internet FTP (file transfer protocol) and merged with your company's database on a set time interval throughout each day. In the event of server failure at a business location, a Swapback system would be shipped overnight to the business loaded with a near-real time database. For multi-location businesses, the standby system could be located at a branch location instead of at Solutions, and could be moved to the headquarters store to replace the main server.

Shea recommends that really large operations should opt to have standby servers located offsite or hosted at Solutions offices.

“Every year we have a customer or two who has a fire,” Shea says. “Backup media are often left in the system overnight where they are vulnerable to fire and other disasters. When this happens, data is restored from a two-day-old backup at best, resulting in the loss of another whole day's transactions.”

Solutions offers another alternative with option three, providing CProRent operation for the affected business until the server is repaired. Recovery from a system failure takes about one hour.