The basic method of building private cloud

When you use enterprise software, many experts will recommend you try the private cloud first if you are not comfortable with running enterprise applications on the public cloud. But the company's cloud computing director Joe Tobolski says it's not just an infrastructure solution that wants to build and manage the cloud inside the datacenter.

"There are a lot of technology companies that mistakenly view private cloud as a business to buy-and then insulate some parts of the data center from the cloud to form an internal cloud," explains Tobolski. "This would violate the true meaning of the private cloud. Those IT companies have been in the direction of women is the internal default private cloud. If you want to build a private cloud, you should consolidate from the data center, the operating system rationalization, should also be from the hardware and software platform and virtualization software on the server, storage and network applications start.

Flexibility and Pay-as-you-pay pricing patterns are the main advantages of cloud computing, including standardization of it, automation and commercialization, as well as infrastructure and configuration resources, Tobolski added, "also involving application development and user it application experience".

Although cloud computing is a hot spot in the marketplace, the current mention of the internal cloud is at a very early stage. According to James Stanley, chief analyst at Forrester, a company, perhaps more than half of the companies do have internal clouds, but only 5% of the world's large companies have the ability to run internal clouds.

But if you're interested in developing private cloud computing, here's what you need to know.

First step: standardization, automation and resource sharing

The three principles that Forrester builds inside the cloud are similar to the application of Accenture's new generation of it.

To build a preset cloud, you have to develop standardized, documented processes and procedures for running, configuring, and maintaining the cloud environment, Stanley explains.

Most enterprise users are not standardized, although businesses that move it repositories down in order to implement IT service management are closer to the goal of standardizing the cloud process than other users.

Making the process more efficient and continuous standardizing the process is the next fundamental step-the key to automation. "You have to trust top users with automation, which is a big hurdle for most businesses," Stanley said.

Automated configuration may be the best starting point for this phase, because users can implement self-service services. For a private cloud, this is not the Amazon model where any developer can configure a virtual machine at will. "This is an internal confusion and is completely unrealistic," Stanley said.

For a private cloud, self-service services mean that enterprise users can build automated workflows through which the resource requirements can pass the approval process.

Once approved, the cloud platform automates the configuration of the specified environment. More often, private cloud self-service services involve developers ' demand for "three virtual machines of this size, the size of storage capacity and these bandwidths," Stanley said. For end users who want to get resources on the internal enterprise cloud, the self-service service will be "I need a SharePoint or file share."

See "Hands-on teaching you how to build a private cloud"