Recently, Jamais Cascio wrote an article on Fast Company pointing out to the recent Orwellian Kindle fiasco by Amazon and, then, arguing that one should not trust the Cloud Computing. The article can be categorized as borderline fear mongering with somewhat incorrect representation about the actual nature of the Cloud Computing. However, there are some important lessons that we should learn from the points he has raised in that article. Let me use this post to debunk the myths and highlight what is important for the Cloud users.
The biggest problem I have with the Fast Company article is the way in which he has generalized the Amazon incident giving an impression that every single Cloud vendor could possibly do something like this. This is plain absurd. First, this is not unique to Cloud Computing alone. Even in the traditional computing model, an OS vendor could push some code that could delete any file on the local desktops. In fact, we have seen some OS vendors restricting the very use of desktops by using the proprietary nature of their software. Remember the kill switch on iPhones? Apple installed a kill switch on the iPhone OS so that they can remotely kill any application installed on the device that they consider to be a security threat. How is this different from the kind of control Amazon exerts on Kindle devices? In its hay days, we have seen Microsoft pushing stuff up the throats of their customers remotely. How is that different from the kind of control Amazon has on Kindle. It is absurd to tie this type of control, exerted by certain vendors, to just the Cloud Computing alone. Such actions, by certain overzealous vendors, can happen in the traditional distributed localized computing systems as well as in some non computing systems. You don’t need any centralization to execute such ridiculous business practices. Let us debunk this first myth which says that such incidents can happen only on Cloud Computing services.
Now it is time to debunk the second myth in that article. It is not actually a myth. Rather, it is an incorrect factual statement on the nature of Cloud Computing. The author portrays the “centralization” in the Cloud Computing in the same way as it is in the mainframe model. It is time for pundits to understand that the centralization in Cloud Computing is totally different from the one we had in the mainframe era. In the mainframe era, the centralized nerve center was a single powerful computing device to which the clients connected. The Cloud Computing world might appear to have a centralized system but the nature of this centralized system is completely different from the one we had earlier. It has a distributed component to it. Actually, it is a collection of several thousand x86 machines that shows a centralized character due to the advances in the virtualization and the other “fabric” technologies. In a well architected Cloud ecosystem, there will not be a single point of failure like how it was in the mainframe era. Jamais Cascio’s characterization of Cloud Computing is factually incorrect.
However, he also makes two important points which the Cloud users can only ignore to their peril.
- There’s plenty of precedence for courts to order the removal (or evenbricking) of devices attached to centralized services when there seemsto be an infringement of some kind.
- If a cloud system locks up–or if a legal decision, change inownership, or service provider whim alters the rulesunilaterally–potentially millions of people will lose access.
These are valid arguments. In fact, in my early days at Cloud Ave, I wrote a post highlighting some of the questions we need to ask the Cloud vendor before we trust our data to them. I pointed out to 13 questions that are crucial to protecting our data in the Clouds. It ranged from data ownership to data portability to protection against subpoena to SLAs. I have covered the above two points raised in the Fast Company article in my list. When we move from a desktop world to a Cloud based world, we give up some control of our data in order to take advantage of the inherent benefits of Cloud Computing. It is important that we do the necessary groundwork before we move our data to the Clouds. Part of it includes finding answers to the questions I have highlighted in that post. Unless the vendor offers a satisfying response, we should not trust them (please note that I am only talking about trusting the vendors not the cloud technology per se) with our data.
Let us not get carried away by any fear mongering about a new technology. At the same time, let us understand what we are doing in the Clouds and protect ourselves from any miscues on the part of the Cloud vendors. A smart user is one who takes advantage of the benefits of a new technology while being careful about the potential pitfalls that might come on the way.