Theory and Practice in Backup and Archiving

For years, backup and archiving have been relegated to the back stage of the storage infrastructure while increased storage capacity, access speed, and efficient primary storage management have commanded users' and senior technical staffs' attention. Backup and archiving just didn't generate the same kind enthusiasm. Some industry experts argue that although most storage administrators talk a good game, backup processes have often consisted of a hope and prayer as much as a well-tested, robust infrastructure. As for archiving, it was thought that archived data was put away for a long period of time, if not forever. Not many people anticipated rummaging through their metaphorically dusty data archives for the last 5 years' worth of email messages.

Over the past couple of years, backup and archiving have begun to move closer to center stage. Several factors have fueled the new interest in backup and storage. First, new technologies such as low-cost hard disks are now considered efficient backup technologies, and companies are promoting new approaches to backup processes. Second, the implementation of large-scale enterprise applications has generated the need to more effectively manage the data produced. Many companies have found upgrading to the latest enterprise application versions requires archiving large amounts of data.

As various recent high-profile court cases have demonstrated, archived data today is accessed more frequently than in the past and needs to become easier to access. And the introduction of Storage Resource Management (SRM) tools has led to the emergence of an interest in handling data life-cycle concerns more cost-effectively. Storage administrators want to be able to put policies in place that automatically move less frequently used data to less expensive storage repositories.

Vendors promoting products to address users' needs have ballyhooed each technology development. But how far have the changes actually penetrated the corporate world? Peripheral Concepts, a market research company, recently surveyed more than 1000 organizations that had a combined total of 320 petabytes (PBs) of disk storage and 1050PBs of near online storage. The results were intriguing. Perhaps the most interesting finding is that only 40 percent of the companies surveyed back up all of their file and application servers. Although many companies don't back up their application servers (rather, they reinstall the application software in the wake of a major malfunction), Peripheral Concepts' findings indicate that organizations still have too much data that can be lost and never retrieved again. Using backup technology is similar to carrying insurance: You hope you never have to use it, but prudent administrators know that major crashes are virtually inevitable.

Peripheral Concepts also found that although organizations show a growing interest in using low-cost hard disk technology for nearline and offline storage, most organizations still rely primarily on tape as the backup medium of choice. About 60 percent of the respondents indicated they're using hard disks for some backup applications, but those applications accounted for only about 20 percent of their data. The findings also show that respondents often use hard disk technology primarily as a staging area as data moves from online storage technology to tape. The most significant barriers to the increase in the use of hard disk technology for backup are users' reliability and performance concerns.

The findings also indicate that another significant trend is the conceptual separation of backup and archiving operations. Historically, organizations have managed such applications together, but that's no longer the case. Approximately 50 percent of the respondents indicated that they treat archiving as a separate application from backup, and another 21 percent stated that they plan to set up archiving as a separate application within the next 2 years.

The Peripheral Concepts survey shows that the development of best practices in storage management is a top-down process. The largest enterprises are often the first to adopt new technologies and techniques, which then migrate to smaller organizations. The final analysis is that new technologies and techniques have generated renewed attention to backup and archiving over the past couple of years. But many organizations have still yet to take advantage of these new opportunities.