Thursday, October 28, 2010

What's in a Cloud?

If you're paying attention in the technology workplace these days, you cannot help but encounter a lot of talk about Cloud Computing.  The term is not new, but its increased usage is a characteristic of the modern IT discussion.

So what do we mean by Cloud Computing? Let's take a look

Larry Ellison said it once in a very creative way at an Internet World conference where I was also speaking (circa 2001). He said:

  • How we relate to software today is sort of insane. We get in the car and go down to the Fry's Electronics, then we hand over $200 for the newest version of Microsoft Office. They hand us a box filled with bits and bytes, and we take it home.  Then in another 18 months, we get in the car again to go get the newest bits and bytes.  A car is not a very efficient means to transmit information.

At the time, he was talking about the use of the Internet to deliver software to the end user.  But just beyond the horizon at the time was the idea that we wouldn't even move the software to our own computers. Instead, it could rest on a host machine that would be maintained professionally, and would have the latest version of the software always available -- on demand.

This is the principle of Cloud Computing.  We move the administrative overhead of installing and maintaining software to a centralized location, and we spend our time working, instead of maintaining one of our tools for doing the work (ie. the computer/software combination).

What's ironic, is that in spite of what Ellison said in his address, he has shown some reluctance to embrace the idea of Cloud Computing. (At least openly. Some of the recent acquisitions by his company look like they're hedging their bets.)

The primary reason people sometimes resist this revolution in computer/software use is that we like to feel that we have control over our own data and software installation.

But here's the truth -- you don't have that much control over your data and software under the old system either.

Some days, your system crashes and you have to spend a lot of time re-installing your software and your data from backup. (You DO have a backup of your data, right?)  Some days you spend a lot of time installing an upgrade to the software, occasionally with delays while you research online to find facts omitted from the installation guide, or to convert your data into a format compatible with the new version of the software. And occasionally it's necessary for you to take a half-day to rid your system of viruses before your existing software will work properly again.

So if you are resisting using Google Docs, or Salesforce.com, or some other element of the move to Cloud Computing, just remember that you really didn't have control of your own data before (very much) and the truth is that you probably (in business, at least) have a lot of your necessary data hosted on vendor's host computers right now anyway.  The real trick is simply to find a vendor you are willing to trust.

So what becomes possible when you embrace the promise of Cloud Computing?  Well the biggest win for you is that you can focus on doing your real business and leave the IT infrastructure to people whose business THAT is.  Your data can be hosted in a secure, regularly backed-up, power and connection redundant data center, probably with a reliability index that you would pay a pretty penny to achieve in your own computing center.

I know, this only touches the very high-level implications of the Cloud, but starting from there, it becomes possible to consider the nuances and challenges that the approach introduces.

For me, there's nothing better.  I never worry any more about whether my hard drive will crash (mostly what it contains now is a giant browser cache) or whether I have the most up-to-date software. (As long as Chrome is up to date, I have everything else I need.)

Perhaps most valuable to me is that where ever I go, as long as I can get online, my stuff is available.  That's the sort of world I'm hoping for.

No comments:

Post a Comment

Reduce to Dashboard

When developers use DataWeave, they often come to rely on the reduce() function to fill in any gaps left by the standard Core library. Altho...