My friends at Stratus brought my attention to a recent survey done by Laura Didido of Information Technology Intelligence Corporation (ITIC) that reminds me a great deal of surveys executed by my former employer, International Data Corporation (IDC).
Early in the adoption cycle of client/server computing, IDC fielded a number of comprehensive surveys that indicated that IT decision makers had no interest in deploying this approach to distributed computing. IDC then fielded surveys every year or two that tracked the change in thinking and the overwhelming adoption of client/server computing.
ITIC's survey finds similar sentiments today for the adoption of cloud computing. I'm not at all surprised about the findings considering how new this approach is for mainstream applications and the trajectory of cloud computing adoption.
The statements from Stratus were measured and reasonable. ITIC's statements, on the other hand, seemed a tiny bit more extreme to me. I wasn't told where to find the report containing all of the survey findings so, it is impossible to determine if the survey instrument was clear and unbiased. I would bet that ITIC hopes that we'll all rush out and purchase a copy of the survey findings and analysis. Let's look at that Stratus and ITIC have to say about the study.
Here's what Stratus/ITIC had to say to introduce the survey findings
A recent rash of downtime in major data centers highlights one of the main reasons only 15 percent of IT managers in a new Stratus Technologies-ITIC survey are implementing cloud computing projects: doubts about performance, redundancy and fault tolerance in cloud infrastructures.
The “IT Performance Trends Survey” of nearly 300 IT professionals in industries ranging from construction to financial services to health care found that an overwhelming majority of the corporate respondents are concerned about data integrity and data access in cloud computing environments – if not outright skeptical.
“Putting all our eggs in one basket – a basket that we have limited control over and limited access to – is not high on our list of priorities,” replied an IT manager at a financial firm with 5,000 desktops and 500 servers.
His was one of the blunter assessments of cloud computing, but it wasn’t out of the survey’s mainstream. Questions about basic operational issues dominated responses. Data integrity was a major concern for 83 percent of the respondents, as security remains a significant obstacle for cloud computing environments. Sixty-three percent are concerned about latency/response times in cloud environments. Outages occurring within days of each other at Google, Rackspace Hosting, and Equinext Inc. in late June and early July validated that concern and others about data and application availability in the cloud.
“As attractive a concept as cloud computing is, the survey results show that IT groups are being very hard-eyed about practical issues before they put serious resources into it,” said ITIC Principal Laura DiDio. “They need to know that cloud environments are reliable and fault tolerant before entrusting them with strategic data and applications.”
Closely related to data integrity were concerns about unscheduled downtime. More than half (52 percent) of respondents can’t accept downtime for contracted services, and 43 percent want outclauses and penalties to discourage downtime.
The reasons given for avoiding the new approach in both surveys centered mainly on reliability, availability, security, the need for demonstrations of real cost savings or cost avoidance, problems administrating distributed solutions and the like.
In my analysis, however, the real reason in both those old surveys on client/server computing and this survey on cloud computing adoption, however, is more likely to be that IT decision makers don't like change. Pure and simple.
That's really reasonsible since they're the keepers of the status quo and only embrace change when one of two conditions is in force - the new technology drops in to an established environment and requires no change or outside conditions are forcing the adoption of the technology. Change translates in their minds to risk, increased costs and they're simply too busy keeping things running to change things that are working just fine now. Why change, they would say, change breaks things.
Here we are in the early phases of the adoption cycle of cloud computing and the the suppliers are presenting glowing statements that add up to "cloud computing is inevitable" and some mainstream uses of IT systems are saying "No way, Jose." (As an aside, I have tried to meet Jose and he's always overbooked.)
The early adopters in both client/server computing and now cloud computing state that they've obtained increases in flexibility, cost avoidance and the like. They often comment that they have plans to expand upon the early use of the technology because of the benefits they have obtained.
This is why I think that start ups and small to medium sized organizations are likely to adopt cloud computing early on. Large and very large organizations are likely to conduct a pilot project or two using workloads that are not business or mission critical. Organizations that made heavy use of grid computing/high performance computing are likely to move a bit more rapidly than those who have not because their current environments are likely to be more similar to that offered by the suppliers of cloud computing products and services.
I enjoyed reading about the survey, but understand that these are the early days and so, we're going to hear about surveys from many research firms that merely echo the thoughts of those who live by the golden rules of IT (see Reprise of the Golden Rules of IT) not the people who are likely to move forward because they have a need and cloud computing appears to be a way to address that need.