Cloud computing has been around for several years now, so people should be getting the hang of it and cloud deployments are a relatively straight-forward process, right?
Well, not so much. Cloud deployments have a rather high failure rate, or at best can take a lot longer to deploy than expected. One survey from Gartner found that 95% of private clouds fail in some form or another.
Cloud Storage and Backup Benefits
Protecting your company’s data is critical. Cloud storage with automated backup is scalable, flexible and provides peace of mind. Cobalt Iron’s enterprise-grade backup and recovery solution is known for its hands-free automation and reliability, at a lower cost. Cloud backup that just works.
There was no one overarching or dominant problem, either. It’s a mixture of problems, and you may get hit with more than one problem along the way. Gartner analyst Thomas Bittman listed six reasons in his report, but we’re going one further and listing seven reasons why your cloud deployment may not go as planned.
1) It may not save money
McKinsey & Company first raised this issue back in 2009 with its study Clearing the Air on the Cloud(PDF). Ever since then, it has been clear that public cloud especially is not always a cost saving option.
Another industry expert weighed in. “It does spread the cost of infrastructure, which makes it very attractive to CFOs and CIOs struggling with massive OpEx costs for IT, but similar to a vehicle lease, you end up paying more in the long run because you do not get to write off those capital costs but rather continue to pay the provider long after they are paid for (along with a healthy profit margin). The cost tradeoff is simply not that attractive, especially for large organizations with a sunk cost in their existing infrastructure,” said Andi Mann, vice president of strategic solutions in the Office of the CTO at CA Technologies.
2) Internal reality check
The decision to deploy a cloud service usually requires an internal audit to consider which apps and systems will be affected. From there, the company might start looking at other issues, and pretty soon it’s doing a much larger inventory and reassessment of its needs, said Charles King, principal analyst with Pund-IT.
“A business might be considering cloud to support a specific business unit or work group, like DevOps, only to discover that other business units could benefit from similar kinds of services, resulting in a higher costs than originally considered or approved. That could trigger a delay until a final use case list and tally could be created,” he said.
3) Forgetting some apps don’t work in the cloud
Some cloud deployments fail to appreciate that there are two basic modes of applications and infrastructures in the typical enterprises, those that lend themselves to a cloud deployment very well, and those that have to stay on-premises. People try to straddle the two without recognizing the essential differences, said Gordon Haff, senior cloud strategy marketing and evangelism manager at Red Hat.
“Cloud deployments should focus on new cloud-native workloads while bridging back to existing classic IT services, workflows, and datastores and providing unified management. They should not however try to be all things to all applications,” he said.
4) The loss of control
IT is supposed to have the greater good in mind, but it’s as prone to fiefdoms and power struggles as any other group, and for reasons both altruistic and selfish, it does not like to cede control over content and other valuable company assets.
Whether we are talking about SaaS, PaaS, or IaaS, both IT and business definitely lose control when they move to a public cloud, and even to some degree a private cloud, said Mann. They no longer control procedures to manage an outage (who is told and when, what recourse do customers have, which servers/applications are restarted first), for security and other breaches (e.g. secret NSA letters requiring data without disclosure to the client), firmware or even OS or application patch deployments, or even access to their data after a cloud provider goes out of business.
5) Insufficient in-house skills
This is one of the more well-documented problems facing the cloud in general.
You can hire the help, although consultants are expensive, or you can develop your own in-house talent and hope competitors don’t poach them. “Making use of training programs and/or partners who do possess the skills and real-word experience can significantly accelerate cloud deployments,” said Haff.
6) A shifting market
The fact is that the cloud marketplace is evolving very quickly, to the point that if it takes an organization a few months to determine what kind of cloud solutions it needs, there may be new services and providers available to consider, said King.
The big names in IT like IBM, HP and Oracle are rapidly growing their cloud portfolios, and you might want to ask what they have coming in addition to what is currently available. “Rather than rushing ahead with a cloud service that only partly fit their needs, many would probably delay until more useful, valuable solutions became available,” he said.
7) Perceived insecurity
There is a broad perception that public cloud especially is less secure than on-premises IT. Even a private cloud is more opaque inside the IT department. The numerous public cloud breaches (PlayStation, iCloud, DropBox, and more) don’t help.
“It is clearly not true that cloud is always less secure, but there is enough truth in this to turn off many IT and even business decision makers, especially for mission-critical or commercially-sensitive systems and data,” said Mann. This would make firms more cautious in adopting public or private clouds until they can be assured of security and integrity of sensitive information.
Photo courtesy of Shutterstock.