Get Off Of My Cloud
If you use the Internet, a smart phone, a tablet, or any electronic device (aside from a dumb phone, perhaps) you have by now probably heard of this thing called “cloud computing”. Now if you’re as technologically illiterate as I am, you probably also don’t know what this expression “cloud computing” means. Perhaps you imagine some grid of mega-computers (all of which resemble the classic IBM mainframes from the 1950s and 1960s) hidden deep in a subterranean vault. Or maybe you imagine something more modern – sleek servers in a sterile white environment. Perhaps you “see” the cloud not as a thing at all, but as an abstraction for something that cannot be seen.
Cloud computing provides users access to things like networks, software, and applications from remote servers, without the user having to know where those servers are located. One common way to think of it is like a power grid. There, you are using energy “on demand” (turning lights on and off, for example), while that energy is coming from a remote location or locations. Google Apps is an example of this. When you use a Google app like Gmail, you are using something located on a remote server as if it were a program on your computer. Your e-mails are saved on a remote server and not on your hard drive. You access cloud-based applications through your mobile devices or personal computer while all that data is being stored somewhere else.
Where does cloud computing come from? With the mass proliferation of data, the idea of utilizing unused server space takes on an increasing importance.1 How massive is world computer data? “According to one estimate, mankind created 150 exabytes (billion gigabytes) of data in 2005. This year , it will create 1,200 exabytes.” (The Economist) That is exponential growth (if it were graphed, it would look like this). Doesn’t it make sense that government, companies, and private individuals would try to find new ways to store their data without sinking excess money into hardware costs?
Now that you know what cloud computing is and why it exists, the next question you may want to ask is: “Who cares?” or perhaps: “Why do I keep hearing about it?” Those are questions that get at the heart of modern computing and data storage. If things are being stored on shared servers, you have to imagine that some of those things are actually important. Only last year, Sony admitted that hackers had stolen information from over 77m users of its online gaming networks. It was also last year that Amazon suffered a breakdown at one of its server centers, jeopardizing the data of the companies using those servers on the cloud.
More seriously, cloud computing is at serious risk of hacking. The U.S. Government is so concerned about computer security that it has opted “to host ‘mission critical’ information on its own servers.” Former head of the National Security Administration, Mike McConnell, has repeatedly said that cloud computing services ought to take a political stance and begin protecting American companies from cyber espionage. He’s gone on record as saying: “We are the most digitally dependent nation on earth, we have more to lose. There isn’t an entity on the globe that’s safe from penetration, not one.”
These and other security concerns have spawned the growth of a cloud security industry. Companies like Vaultive encrypt data before it goes to the cloud server. The Economist notes that “the lesson for companies let down by Amazon’s outage is that they need to be aware of the risks of being too reliant on a single supplier, with cloud computing as with anything else. Firms that use cloud-based systems should be looking at ways to distribute work across multiple providers.” Cloud computing is far from safe, reliable, and comforting.
Amazon’s malfunctioning server center demonstrates another potential danger of cloud computing. Martin Casado, who wrote his PhD at Stanford on cloud computing, notes that ““if you actually do a power outage [at the server center], you’ll lose all of them. If the cloud was really as elastic as we say it is, we should be able to take all those workloads and move them to another data center.” This is because the workloads that Casado mentions are still processed at the network data centers. That is, they aren’t divorced from the hardware. Nicira, Casado’s company, uses software to divorce this data from the hardware (watch the video on their website, linked above, for an informational overview of how this works). They create a digital world that houses the same data stored on the physical hardware (Casado compares it to the virtual world of The Matrix).
The take-away lesson from all this is that cloud computing is a necessary data storage solution, but one fraught with difficulties and dangers. It is the immediate future of computing.
1 According to one commentator at The Economist, the technology behind cloud computing existed as far back as 1994! This suggests that no one had thought of utilizing the technology because there wasn’t then a need for that kind of large-scale data outsourcing.