In this era of the all-pervasive cloud, it’s easy to assume that the data we store will somehow be preserved forever. The only thing to fret about from a posterity perspective, we might think, is the analog information from days gone by—all the stuff on papers, tapes and other pre-digital formats that haven’t been explicitly converted.
Vinton Cerf, often called “the father of the Internet,” has other ideas.
Now chief Internet evangelist at Google, Cerf spoke this week at the annual meeting of the American Association for the Advancement of Science, and he painted a very different picture.
Rather than a world where longevity is a given, Cerf fears a “digital dark age” in which the rapid evolution of technology quickly makes storage formats obsolete thanks to a phenomenon he calls “bit rot.”
In that world, the applications needed to read files we so confidently store today could be lost because they’re incompatible with new hardware technologies that emerge. The result, he contends, could be that many of our those files will be rendered useless, inaccessible to future generations.
Cerf’s proposed solution is something he calls “digital vellum”—essentially, a tool for preserving old technologies so that even obsolete files can be recovered.
“At a high level, the way to solve this would be to maintain at a minimum read compatibility with older data even as new technologies are introduced without worrying about performance, capacity or cost,” said Eric Burgener, a research director with IDC. “The devil, of course, is in the details.”
The Olive project at Carnegie Mellon University could be a prime example. Led by Mahadev Satyanarayanan, a professor of computer science, the project aims to develop the technology needed for long-term preservation of software, games and other executable content.
Until such tools are commonplace, however, companies clearly have at least as much reason to be concerned as individuals do.
“It’s hard to think long-term about this stuff, but I think there are some real risks here that collectively we as an industry should be thinking about,” said Simon Robinson, a research vice president with 451 Research.
The topic is actually not new to the storage industry. Roughly 10 years ago, for example, the Storage Networking Industry Association started talking about the notion of a “100-year archive” to address such issues—though, for the most part, such efforts have “fallen flat,” Robinson said.
The lack of a profit motive is a big part of the problem, he suggested: “This is not seen as a lucrative opportunity, so not really worth the investments required.”
Compounding the problem is its long-term nature.
“I think some businesses definitely ‘get it’ as an issue, but feel slightly helpless in that this is not an issue they can address themselves,” Robinson said. “In this sense, it’s great that Google and others are starting to at least talk about the risks and potential paths forward.”
