Thursday, January 7, 2010

Cloud Computing in Simple Terms

Have you heard this term being bandied around in your office or in the media, but don't have a good grasp of what it really means? This article might make the truly hardcore ├╝bergeeks wince, but then it's not being written for them.  I'd like to help you get a grasp of the concept: a foundation to build a better understanding on.

The term "cloud computing" is an abstraction — a term that describes a complex combination of technologies using a readily understandable mental image — but as comparisons go, I think it's a good one.

When I think of the nature of clouds, a number of defining characteristics come to mind — their mobility, variability in size and variety of shape.  These can be also easily applied to computing.



Cloud System
Cloud Computing
Mobility
Global, like weather.
Global, like other forms of electronic communications.
Size
Expandable or contactable. Size is variable and changes to it do not affect its fundamental nature.
Dynamic and scalable; capable of growing and changing to accommodate needs and demand.
Shape
A cloud is a cloud.  It can be regularly or irregularly shaped and change at any time without altering its identity as a cloud.
Variable.  Comprised of clusters of computer servers in various remote geographical locations, but the locations themselves are unimportant.

So, when some tech geek tells you that data or a software application is "up in the cloud", what she means is that it's hosted on a highly mutable global computer system.  It can be made up of computer servers which are physically far apart — like Canada, Singapore, and Australia — but which are working together with their combined resources.

Broadband Internet speed and increasingly inexpensive computing power and storage space are what makes this worldwide information processing and storage "cloud" possible.  Another critical contributing technology that allows for this cloud is something called virtualization, but a discussion of that would be another article in itself.

When hardware was expensive and transmission speeds were slow, computing was an activity that occurred primarily individually and locally.  That is, applications and data resided on a computer you directly used, or on a server in a local area network (LAN).  If the local network couldn't handle the demand from the number of users, additional hardware would have to be purchased and installed locally.

Now, it's no longer necessary for your applications and data to reside on a local computer that you have in-person access to, nor is it necessary for you to purchase additional servers or storage if you run low. Obtaining additional "infrastructure" (with or without software installed on it) is about as simple as picking up the phone and ordering a pizza.

It might sound revolutionary, but it's not.  Purchasing services that you can't easily produce yourself is nothing new.  The electric power grids that service cities nationwide operate on the same principle: those that have a surplus can sell to those that produce too little (or none at all).  Computing today has evolved to become just another form of utility, so feel free to stick that invoice on the bulletin board, next to the heat, gas, and water bill.

For this added convenience, there are risks.  While cloud computing makes for a beautiful mental visualization, in person the hardware installations amount to little more than ugly, power hungry server farms.  Not great for the environment.  Now, if only Google could harness solar power to help fuel their massive computing warehouses, the way that they offset their electricity consumption at their Mountain View, CA headquarters.

On the subject of privacy and security: while there is no guarantee that information that you store locally is safer from intrusion by hackers than information stored remotely in the cloud, there are complicating legal issues.  Data that you store "in the cloud" resides in the legal jurisdiction of the computers that are providing the processing and storage, or the network routers that handle the data transfer.  Should some third party request access to that data under the powers of  a law (e.g. The Patriot Act), the location of your information and its pathway back to you is critical; your residential location is irrelevant.

"Do you know where your personal information is, or how it is being used in the Clouds?" If you'd like to learn more, Ontario Information and Privacy Commissioner Ann Cavoukian published a white paper last year examining the rise of cloud computing.  A presentation by Assistant Commissioner Ken Anderson (delivered at the University of Alberta) is also available for download (in PDF format).

Meanwhile, the US Federal Trade Commission has just announced that it will be examining the risks posed to privacy and security by this emerging technology.  A round-table discussion is scheduled for January 2, 2010.