Problem solve Get help with specific problems with your technologies, process and projects.

Bracing for the failure of your private cloud architecture

The many misconceptions regarding private cloud can be dangerous. Why do private clouds fail and will yours be one of them?

A common misconception among cloud consumers is that private cloud architecture doesn't have the same issues that arise with public cloud services. Due to its private nature, some believe that security is a given and failure is not an option. Those users learn the hard way that is not the case. But why do private clouds fail?

The core reasons that private clouds fail include a lack of planning on the part of IT and a failure to understand that cloud security requires attention and maintenance. Private cloud architecture is typically much more dense and intense than traditional systems, with processing and storage concentrated within a small cluster of servers.

Most patterns of failure link back to three problems during private cloud builds.

1. Re-using existing hardware

There are two types of clouds: clouds that are managed properly, and clouds that don't work.

Many private clouds are built using repurposed hardware and some existing software licenses. While this seems like a way to maximize the value of your private cloud, you'll find that the hardware requirements for most private cloud OSes are demanding.

If hardware doesn't have enough horsepower, the system will begin thrashing, which causes poor performance and likely a system crash. These service interruptions may cause users to lose confidence, and the private cloud fails.

2. Security oversights

Many IT pros who implement private clouds fail to consider security issues. A common misunderstanding is that private clouds, such as an OpenStack distribution or Eucalyptus, come with enough pre-built security to meet requirements. Another misunderstanding is that security is a non-issue due to private clouds being innately physically secure.

Neither could be further from the truth. Just ask Target and Sony about breaches within internal systems. Your worst enemy is your user base because they are the most likely to walk out with data they shouldn't.

3. Management issues

The same rules as security apply to management of the services -- such as a storage APIs -- or the resources. Private clouds are still clouds, and thus need to be managed effectively, which includes controlling access to core private cloud services, or provisioning and deprovisioning resources. There are two types of clouds: clouds that are managed properly and clouds that don’t work.

While these are a few basic mistakes that IT makes when operating private clouds, the reality is that private clouds are like any other system. Their success is dependent upon the planning, architecture and testing that goes into the deployment of the private cloud.

About the author:
David "Dave" S. Linthicum is senior vice president of Cloud Technology Partners and an internationally recognized cloud industry expert and thought leader. He is the author or co-author of 13 books on computing, including the best-selling Enterprise Application Integration. Linthicum keynotes at many leading technology conferences on cloud computing, SOA, enterprise application integration and enterprise architecture.

His latest book is Cloud Computing and SOA Convergence in Your Enterprise: A Step-by-Step Guide. His industry experience includes tenures as chief technology officer and CEO of several successful software companies and upper-level management positions in Fortune 100 companies. In addition, he was an associate professor of computer science for eight years and continues to lecture at major technical colleges and universities, including the University of Virginia, Arizona State University and the University of Wisconsin.

Next Steps

Should you invest in private cloud?

Can your private cloud beat AWS?

Dig Deeper on Cloud architecture design and planning

Join the conversation


Send me notifications when other members comment.

Please create a username to comment.

I would add to this; the lack of experience the IT staff has in the design, configuration and operation of the Cloud Stack itself. Highly virtualized environments are not Converged Infrastructures. To achieve levels of Cloud Maturity that provide for true "Self-Service" private cloud environments you need to build a fully integrated, automated and orchestrated Cloud Stack. Too many IT shops have considerable "Shelf-ware" that they have invested in but have not followed through with the development of the appropriate Use Cases for a Private Cloud that meets Business requirements. These Use Cases must meet those requirements for the full integration and automation of workflows. Most IT departments do not have the time and resources to invest the expertise needed for this effort. Consequently, only pieces of automation exist that do not fully realize the benefits of a Private Cloud environment.
I'd say that reason #1 is that the learning curve of private IT operations can't keep up with the learning curve of public cloud operations.