The Cloud Has Failed Me

If you’ve been keeping even a moderate eye on the technology industry during the last couple of years I’m sure you have heard of “the cloud.”  While more of a concept than a specific product or service “the cloud”, a term short for “cloud computing” which in turn simply means computing and storage that is handled via network instead of on a user’s desktop computer, has taken off significantly and is no longer just a topic of conversation between geeks.  New service such as Apple’s upcoming “iCloud” are beginning to fuel the popularity of such products and services to the extent that “the cloud” has quite literally become a household term.  And regardless of if you know it or not you’re probably already taking advantage of cloud-based computing and storage as it is, as “the cloud” refers to services like Dropbox and modern fixtures such as web-based email.

Conceptually speaking, cloud products are designed to give users the ability to “outsource” their computing and/or storage needs to another provider with the benefit of being able to access information from a variety of devices from anywhere with an Internet connection, and perhaps more importantly the peace of mind that one’s data is secure and not prone to accidental loss.  But as someone who has used promising cloud-based applications, I must say that what I’ve seen and experienced has been enough to make me think twice about putting my life on the cloud.

To start off with, I think it’s of importance that I start out by discussing the failure with Amazon EC2 – the cost-effective and scalable computing and storage engine utilized by many websites and cloud-used services – earlier this year.  Surely you remember when you weren’t able to access some of your favorite sites or utilize some of your services when a datacenter disruption took countless customers offline.  While these types of outages are supposed to be a rarity in “the cloud” I for one experienced a moment of deja vu when EC2 experienced another outage earlier this week.  And while I don’t personally contract with Amazon Web Services for EC2 utilization, it was somewhat frustrating to go without services like CloudApp that built their offerings atop EC2 backends and thus fell victim to the downtime.

Sure, everything came back online within a half hour, but this event really got me thinking.  Sure, the downtime was controlled by humans this time around, but what would happen to “the cloud” if a single datacenter got hit by a natural disaster?  I mean, even if backup systems were in place how long would users be without their data after a major catastrophe?  Honestly, I don’t want to find out.

And my worry about the cloud stretches beyond just Amazon Web Services as well.  The concept of a single point of failure really does bother me.  Earlier this year, for example, I stopped using previously reviewed IRCCloud because even though the service itself was amazing the fact that it operated off of a single server contributed, in my opinion, to a bumpier experience than I would have liked.  More recently, even, IRCCloud suffered downtime because of a DDoS attack; not because their server was being specifically attacked, but because the datacenter their server was in went attacked.

Need another example?  Look at the melt-down that cloud-based password manager LastPass had earlier this year that left users without access to their passwords while the company scrambled to get things back online.  Worse yet, their mishap turned out to be the result of precautions taken after a data breach; illustrating that the concept of a cloud-based service having a single point of failure extends over to the concept of a cloud-based service having the potential for a single point of compromise.  It’s reasons like this that I stay away from synchronization services such as Opera Link and Google Chrome’s Sync feature, no matter how appealing they may be.

Even looking at web mail, you have to consider the number of times that services like Gmail have gone down, too.

Sure, the cloud probably offers better redundancy for my data than simply storing on my laptop does, but when all is said and done I like controlling my own data.  Of course I subject myself to frustration every now and then, but at least when my local data acts up I know that I can blame myself and not the shortcomings of a product or service that I shouldn’t have put my faith in to begin with.