Murphy's Law and Continued Innovation Ensure Enduring User Interest in Tape

Recently a blog entry appeared on the Byte & Switch website that asks the question if tape will be cancelled due to a lack of customer interest. In short, the author of the article, George Crump, postulates that customers are losing interest in tape partly because tape manufacturers are taking more interest in selling disk than tape. As a result, more innovations are occurring in disk libraries while tape libraries languish. But has tape in fact outlived its usefulness?

While I can see Crump's point and agree that a great deal of innovation is occurring in disk libraries, I have heard enough customer stories and seen enough new technologies emerge to convince me that tape will outlive forecasts for its ultimate demise though its role is certainly going to change.

As anyone who follows the data protection space at any level probably knows, disk and deduplicating disk systems are on their way to replacing tape as the primary backup target in many shops. The most often cited reasons for this change include:

• Larger capacity disks and deduplication put the cost of disk on par with tape
• Faster backup and recovery times
• Reduced backup windows
• More efficient use of disk (deduplicating disk systems)
• Facilitates offsite data replication

But why will tape survive, and possibly even thrive, in spite of these obvious benefits of using disk? Here are a few stories and tidbits of information that I have picked up over the last year as to why tape will continue on.

• Murphy's Law. Most of us are familiar with the short version of Murphy's Law that states, "If anything that can go wrong, it will." However, there is another version of it that goes along the lines of "Whatever can go wrong will go wrong at the worst possible time and in the worst possible way." I only bring this up because last spring that I was attending a conference (sponsored by a disk vendor no less) where a user was extremely relieved that he had copied his production data from disk to tape.

It turns out this individual had two data centers - a production data center in the New Orleans area and a DR site in the Houston area - and they were both hit by hurricanes within the space of a month and he had to shut down both of them at the same time. Thankfully he had copied all of his production data to tape so he could move it to yet a third data center and recover his production applications.

It is because of circumstances like this that companies like Quantum are hearing from conversations with its customers that tape continues to hold a strong role in disaster recovery and long term data retention in its customer accounts. Its customers are also seeing a growing role for managed, near-line archives of indexed, tiered file data.

• Innovation in tape continues. Granted, tape library vendors are giving more attention to disk-based backup solutions but that does not mean innovation in their tape libraries has stopped either. Instead, they are currently putting a greater focus on the data management software found on tape libraries, but that innovation does not occur overnight.

The customer requirements associated with the emerging use models for tape are primarily about software. There are some needs around traditional hardware, but the most important needs are about data movement and data and media management. The ability to tune these elements together is key to creating a complete answer for the customer. You'll continue to see enhancements in drive technology and automation, but the big news going forward will be how tape library vendors use the drives and the automation - in conjunction with disk - to solve real customer needs and integrate them.

We now see similar types of innovation occurring in products like the Quantum DXi7500. It can already manage the movement of data from disk to tape and back again since not every company has a secondary location or second disk system to which it can replicate data. Already the DXi7500 works with enterprise data protection software such as EMC NetWorker and Symantec NetBackup which can track the movement of backed up data.

Even assuming all innovation in tape technology stopped today, I sincerely doubt companies would stop using tape. Evolutions in storage take years if not decades to occur (I am still aware of banks that use reel-to-reel tape systems and innovation in that technology ended eons ago). In addition, with the increasing focus on energy efficiency in data centers, tape is arguably the most efficient means for storing archive data that must be saved for compliance reasons, but may never see the light of day again. So to declare that tape is dead when innovation is occurring is just downright irresponsible.

Neither disk nor tape technologies, when viewed outside of the context of how companies are using them to solve problems, are very interesting of and by themselves. It is only when they solve real world problems and how companies utilize them in those capacities that companies perk up and take notice. So just because tape is no longer used as the primary target for backup doesn't mean that tape is dead, it just means that it is evolving to solve new sets of corporate problems. As those specific problems become better understood and documented, expect the products that support tape to experience similar forward movements in innovation.