Performance Anxiety: What the Intuit Problem Tells Us About the Future of Cloud Computing

As with many things in life, like heart attacks or giant oil spills, certain problems aren't necessarily problems until they are. That seems to be the case with the issue of performance in cloud computing, to judge from the negative reactions to the Intuit outage last week. So what are the lessons to be learned from this snafu?

Last week the web sites for Intuit's small-business online applications services were down for approximately 36 hours. Intuit has more than 300,000 customers using those sites, according to published reports.

Intuit president and CEO Brad Smith apologized in a letter on the company's blog site: "We hold ourselves to the highest standards in dependability and customer service, and over the past two days, we have failed to live up to those expectations." Smith revealed that the outage was caused by a power failure related to a "routine maintenance procedure" which took out both primary and backup systems.

Several commentators questioned Intuit's IQ level, and appropriately so, for housing primary and backup systems in the same location. One commentator, however, made the most salient point about it:

"Everyone complaining about the primary and backup systems in the same place, did you ask this company, who is holding your companies life blood, what the disaster plan was and is and how and where the primary and backups are and how the power is[?] ... Did you do your due diligence?"

Amen, brother.

In a recent blog about the weaknesses of cloud computing, I cited performance as the number one issue. Several commentators took me to task for that, dismissing performance issues with cloud service providers as overblown or less worrisome than those related to the operation of internal data centers.

There are two important points (at least) to be made about the Intuit problem:

First, performance is an issue with cloud computing … a big issue. Major corporations and institutions will remain hesitant to embrace cloud computing in any significant way as long as there are news stories every other week about service outages.

Second, customers have to take responsibility for their half of the partnership represented by cloud computing. That's why IT must be involved. The CIO, or members of his or her staff, must vet each and every SaaS, IaaS, and PaaS engagement to ensure that they meet the standards of a well-run IT organization in terms of performance, security, and compliance.As with many things in life, like heart attacks or giant oil spills, certain problems aren't necessarily problems until they are. That seems to be the case with the issue of performance in cloud computing, to judge from the negative reactions to the Intuit outage last week. So what are the lessons to be learned from this snafu?