The Backup and Recovery Rut

A recent survey of 366 IT managers conducted by Double-Take Software, a provider of data protection software that is in the process of being acquired by Vision Solutions, a provider of high-availability and disaster recovery software, found that the vast majority of them were experiencing significant problems when it comes to backup and recovery.

About 25 percent of them were still relying on tape as their primary backup and recovery medium. Only about 23 percent said they were using disk backup, while 18 percent were relying on high-availability solutions, 17 percent had real-time data replication in place, and 12 percent had nothing at all.

Despite this impressive array of backup and recovery technologies, 79 percent reported they had trouble completing  backups within their allotted period of time. Chances are good that most of the people having this problem are still relying too much on tape backup even as they move to adopt other data protection technologies. In fact, when it came to identifying challenges, 87 percent said improving the reliability of a backup was still an issue; 92 percent cited the time it takes to recover; 92 percent cited cost; and 89 percent cited the potential for data loss.

Even though many IT organizations are using disk-based backup, tape-based backup appears to be still carrying most of the backup load. And just to make things even more interesting, the advent of virtual servers is further complicating the backup and recovery process as more virtual servers start to share the same backup storage process. In many cases, it’s not unusual to find an IT organization moving aggressively on virtualization, but still relying in many cases on antiquated tape backup for their servers.

The end result, said Christian Tate, vice president of worldwide marketing for Double-Take Software, is instead of being able to recover a virtual server in under 20 minutes, the IT organization winds up with more servers than ever to recover using outmoded recovery processes that can take as long as half a day per server.

In addition, Tate notes that most of the IT organizations relying on tape don’t know for sure that the recovery process is going to work because there really isn’t any effective mechanism for testing whether the backup process really works.

It’s not like tape backup is ever going to completely disappear. But as the amount of data that needs to be managed grows and has to be recovered faster, the role of tape is changing in terms of becoming more of an archive medium. In the meantime, IT organizations would do well to remember that as one element of the IT infrastructure advances, chances are there is a cause and effect on some other part of the Infrastructure that will inevitably come back to haunt them when they least appreciate it. In the case of backup and recovery, that means relying too much on tape at the wrong time for all the wrong reasons.