Top 10 Reasons Cloud Computing Deployments Fail: Page 2

4. Failing to hold yourself accountable.

Even if you have a solid SLA that has provisions for remediation, that doesn’t mean you are off the hook if something goes wrong.

For instance, what happens if you store sensitive customer data in the cloud and someone breaches it? Do you really think it matters what your SLA says? Who will your customers hold responsible?

You, that’s who.

Earlier this month a security breach at AT&T exposed the email addresses of more than 100,000 iPad users. Most customers blamed Apple, but the problem was with AT&T’s cloud service.

The breach was a minor one. After all, most people’s email addresses have been farmed by spammers many times over. However, had the leaked information been credit card or personal information, Apple would have had a problem that made the iPhone 4 antenna problems seem trivial.

“You can never abdicate responsibility to a service provider,” Stroud said. “The cloud provider may be a custodian of your information, but the reality is that it is your reputation that will suffer if something goes wrong.”

5. Failing to scrutinize vendors.

Pretty much every service provider, hosting company and ISP is rebranding itself as a “cloud provider.” However, not all cloud providers are created equal. While it’s a pretty safe bet that Google, Amazon and IBM will be around in the years to come, you can’t say the same about numerous cloud computing startups.

What happens if your cloud provider fails? The collapse of cloud startup Coghead last year shows just how dangerous skimping on due diligence can be. Coghead wooed customers with low prices. Then, when it ran out of money and failed to raise additional VC capital, it gave its customers a few short weeks to get their data off of its systems.

It could have been worse. What happens if your cloud provider shuts down with no notice? What happens if disgruntled employees smuggle servers out the back door after they get their pink slips? What happens if the local sheriff chains up the building under order from a bankruptcy judge?

If you don’t do your due diligence, you might find out.

6. Failing to understand the service supply chain.

Even if your cloud provider is stable, do you know how stable their service providers are? Cloud providers are increasingly outsourcing components of their services to third-parties. It’s important to understand the entire service supply chain in order to accurately judge the viability of the service you are signing up for.

If you’re dealing with a large, established cloud provider, at least you have a single neck to choke if something goes wrong, and you can bet that bad press will motivate them to solve the problem. With smaller vendors, you might be on your own.

7. Failure to manage and monitor applications.

Many organizations have made the mistake of believing that management and performance problems disappear when they move to the cloud. “With traditional applications, eighty percent of your time and resources are spent on management and monitoring,” Sebastian of Heroku said. “The cloud puts a big dent in that, but it doesn’t go away entirely.”

If your application performs poorly, your customers won’t blame the cloud provider, they will blame you. “There will be mistakes in your application. There always are,” Sebastian said. “With the proper performance management and monitoring tools in place, you’ll have a better chance of catching those mistakes before they become a disaster.”