Yogi Berra- The Future Ain't What It Used to Be

[This is my first excerpt from my The Future of Backup whitepaper. Download the full whitepaper for instant gratification.]

When Yogi Berra said, “The future ain’t what it used to be,” he could have been talking about backup rather than baseball.  Despite enormous progress in backup technologies over the last decade, the “secondary” technology of backup has barely kept up when compared to the incredible gains in the “primary” technologies associated with computer and storage platforms. Most backup vendors as a whole have failed to make enough progress in delivering increasing fundamental value to customers.

In today’s post, I am going to discuss the fundamental challenges associated with backup.


Disk Capacity and Bandwidth Challenge

The most critical challenge of backup has now existed for decades – primary data is growing at a much faster rate than the bandwidth of the interconnects available for copying that data to a secondary storage location. Disk drives have increased in capacity from approximately 100MB in 1990 to 2TB in 2010 – a factor of 20,000X!  At the same time, interconnect technologies have evolved at a much slower rate.

The chart below illustrates the gap that has developed in the last twenty years between primary storage capacity on a per-disk basis and common interconnect technologies.

Gap between primary storage capacity on a per-disk basis and common interconnect technologies.

Primary Versus Secondary Storage Capacity Challenge

In 1990, disk drives were approximately 100MB in capacity and tape drives were approximately 5GB.  This represented a 1:50 ratio.  In 2010, disk drives were roughly 2TB in capacity while tape drives (LTO-5) were 1.5TB, representing a 1:0.75 ratio.  This diminishing capacity ratio is one of the fundamental reasons that tape-based backup is much less viable than it was 20 years ago.

However, disk-based backup presents its own challenges.  Disk-based backup has at worst a 1:1 ratio (when the same disk types are used in both primary and secondary storage) and at best an approximate 1:3 ratio (when higher-speed 15K RPM drives are used).  When solid state disk drives are considered, the ratio doesn’t get better. While expensive, there are 2TB SSDs available today while the largest commodity rotational drives available are 2TB.

Data Reduction Techniques and Reliability Challenge

Backup works by creating copies of data, which increases the overall reliability of the protected system.  It’s that simple.  Data reduction, on the other hand, works by eliminating redundancy.  The two work against each other.

There is a continuing drive by vendors to incorporate data reduction into primary storage.  In the past, this hasn’t gained much traction because of the performance penalties associated with data reduction techniques.  However, the strong venture capital, mergers and acquisition activity in this area (as well as global 1000 vendors’ statements) indicate a renewed interest and focus in primary storage deduplication.

What does primary storage data reduction have to do with the challenges of backup?  The answer concerns primary versus secondary storage capacity ratios.  When data reduction is used for secondary storage, it can lower the price per effective terabyte of backup capacity when compared to the primary storage being protected.  However, when primary storage data reduction is employed, it reduces the aforementioned primary versus storage capacity ratio.

Complexity and Operational Costs Challenge

The challenges previously discussed are based upon the technology trends associated with storage and with backup.  However, there is another challenge that users have been increasingly facing – the sheer complexity of data protection and its resulting rising operational.

The complexity of data protection is driven by several factors:

  • Few data protection products handle the major functions associated with backup, which include not only backup but archiving and disaster recovery in an integrated solution.  Licensing on a piecemeal basis greatly exacerbates this complexity.
  • Many data protection products push the task of purchasing, assembly, integration, management and monitoring of backup servers, networking, storage, operating system, anti-virus software, backup software and other components back to their customers.  This increases operational costs dramatically.

In my next two upcoming postings, I will tackle the past and present state of the disaster recovery industry. I will also explore several data protection future scenarios. Remember, if you do not feel like waiting for future postings, you can download my The Future of Backup whitepaper to read this series in its entirety.