Intelligent Enterprise

Last week a major outage affected 14% of Google users and caused widespread panic. Okay, it caused frustration, as users could not access their free search engines, free document management systems, and free e-mail systems. Perhaps they should ask for their money back. The comment that I kept hearing was "I had to use Yahoo." Priceless.

Still, the timing could not have been worse, considering that the US Government began discussing how cloud computing fits into their $78 billion IT budget for 2010. Many in the private sector are looking at cloud computing as well. The hype leading them there is the possibility of saving some money.

I blogged a few weeks ago about the cut fiber that disconnected portions of San Jose from anything on the Internet, including SaaS and cloud computing. Now we're seeing instances where the providers themselves fail. While we should consider these risks, the value still seems to be there.

It's an accepted fact that computers and networks go down from time to time, either through hardware or software failures, or in the case of the Google outage, human error. We might as well get over that, and any expectations that cloud computing will make things worse, or better, should be tossed aside with your Thigh Master (which also did not make things worse or better). Moreover, SaaS, a form of cloud computing, has been around for a long time and the number of outages has been miniscule. Moreover, when many enterprises come clean with their own list of outages in a year, it makes their SaaS systems look downright rock solid.

The outage issue will arise when cloud-based systems, new and existing, begin to hit the internet and major companies begin to leverage them as part of their IT infrastructure. This includes storage-as-a-service, platform-as-a-service, and infrastructure-as-a-service, which will prove to be largely reliable but will go 'splat' in scary ways from time to time. This will be largely due to the newness of this technology, and the inexperience of both the providers and the users. Fear not. Each and every one of these failures will appear in the press.

Should this scare you off?

Bob Evans had a great blog that addressed two points that deserve more attention:

• Google reported that its outage was caused by 'human error,' which is often at the heart of corporate data center disruptions as well… As I've stated before, this becomes a quality of support rather than a reliability of service issue.

• I remain convinced that an honest self-assessment by IT and business decision-makers will lead to the realization that their data center reliability, security and performance palls in comparison to today's leading cloud computing vendors. In addition, a thorough evaluation of their time-to-market, flexibility, TCO and ROI would also clearly favor the rapidly evolving SaaS and cloud computing alternatives.

The fact is that SaaS and other cloud computing-based services will hit the dirt from time to time. When all is said-and-done, the value is still there. As time goes on, cloud computing providers will improve, providing better redundancy and management services, and failures will be largely transparent to the users.

In other words, chill out about these outages and pick the best solution for your enterprise.

E-MAIL | FOLLOW US