Genetic Diversity… Not Just for Peas and Fruit Flies Any More

Reading time for this article .

By: Bob Cone 

Yes… I’m referring to Gregor Mendel’s experiments with peas in the 1800’s, which led to a better understanding of genetics, and possibly our own study of fruit flies in middle school biology which drove the point home. So how have we completely forgotten about the importance of genetic diversity when it comes to one of the most important aspects of modern day life – protecting our data?

Gregor Mendel

Data storage is one of those things we’ve taken for granted for a long time. We back it up. We replicate it. We archive it. It’s all good. But what do we backup, replicate and archive to? Does it matter? It used to be tape which is quite “genetically” diverse from disk where the primary data lived.  There were reasons we might have favored keeping it all on disk. It’s quicker for a short restore. We could put the data there without “middleware,” software that understands both disk and tape. And even though disk creates a lot more heat than tape, it just somehow seemed cooler.

So why did we use tape? It was the cost leader. Some studies get a bit crazy on just how much less expensive tape is, but even today comparing it to disk with deduplication, the more conservative studies even put it at about two times less expensive to buy than disk and easily 6 to 10 times less expensive to own, power, and operate. But the cost is way closer than it used to be, so that was enough impetus for many to start moving those long term stores over to disk. And for offsite? Just replicate from disk A to disk B offsite. What’s not to love?

Disaster. That’s what’s not to love. Kind of like when the potato blight killed millions because there was virtually one food source which was wiped out by a blight or rampant hemophilia in royal families all over the world because they were all intermarried. History forgotten is doomed to be repeated? Maybe. When I say, “disaster,” most people in data protection think of natural disaster like flood, fire, or earthquake.  I think any one of the numerous cyber-attacks we read about on a weekly basis would be much more of a disaster. It could be ransomware, hack and destroy (ask Sony Corporation about that one), malware, you name it.

Natural disasters don’t “think.” It’s not a fire or flood’s intention to destroy data when it strikes. Cyber attackers who are after your data can figure out where and how it’s stored and take many paths to get there. Attacks on disk are rarely able to destroy data on tape and vice versa.

As we’re moving more and more data to disk, I might add that even the disk we’re putting it on has become less genetically diverse.  Whereas there were over a dozen disk manufacturers in the late 80’s, today there are only three, and only two of those really play in the enterprise area. If we’re talking about malware that gets embedded in disk firmware (ask the NSA about that one), you only have to hack into two sources of firmware to take out roughly 90% of the data that’s stored on disk world-wide.

Although tape still has several advantages over disk in some areas and applications, it’s the fact that tape is not disk that really gives it an advantage. It’s built in genetic diversity. Tape and disk together give a more robust and longer lived digital preservation plan than any single media on its own ever could. From “cost leader” to “security leader.” Long live the ever long lived tape.

For more information about the term “genetic diversity” and what it means for your data center, check out our newest whitepaper: Genetic Diversity in Deep Storage.