Red Hat CEO of Greater China Pan: cloud Investment and return

The concept of cloud computing is more about turning it into a business-oriented capability than simply reducing it costs. No matter how important "soft cost" is in the enterprise context, it does not mean that cloud computing does not require financial assessment, after all, the types of projects that are specific to each cloud are different.

For example, the use of such public cloud resources provided by Amazon, while reducing capital spending today, is likely to face higher operating costs tomorrow. So we should consider this in any financial analysis. Private or mixed clouds are, in some ways, closer to traditional enterprise internal IT projects. Users also need to buy software (and sometimes even buy hardware), plan and execute operations, train IT staff and users, all of which are combined in an IT environment and architecture to make the business work better. Compared to the virtualization and other projects, from the tactical action, the construction of "cloud" is very different.

Pan, president of Red Hat Greater China

While the cloud uses virtualization in many ways, but it is, after all, a nascent product whose main task is to help IT operations and consumers of its services to do their jobs faster and better, and in the cloud, bloated it or low usage rates are intolerable. So this means that, while users perform more traditional it financial analysis, they also need to implement some complementary measures that can be aligned with the "cloud" provided to the enterprise's main value.

These income-related figures are, of course, only approximate estimates, and vary from company to business, but they do have a financial impact. The metrics that users can use in the analysis include:

The time required to deploy the new service (application). One of the main features of the cloud is the self-service service that users can get computing resources. For example, developers in an enterprise can access the standard templates and tools needed to develop and test applications. This approach can significantly shorten the time required to start a new project or improve the work associated with an existing project. At the same time, self-service services are often implemented in a managed policy-based framework, so the IT department does not lose all visibility when it comes to using a special-purpose public cloud. Ultimately, the productivity of developers is improved, and businesses are able to make revenue-generating services more quickly online. Although this must be a soft cost, but this value can be quantified through a combination of productivity measures (that is, how much work and time can be saved by optimizing the workflow), and it can bring direct business benefits to the enterprise (how much revenue a new customer service can bring to the enterprise one week or one months in advance).

Reduce improper configuration. Research by Gartner, a market research institute, shows that an average of 80% of mission-critical service outages are caused directly by human or process errors. The reason for service outage times usually involves multiple management areas. However, a significant portion of this can be attributed to change management and configuration management, while the concentration of policies and workflow controls in the cloud infrastructure helps to reduce such problems. Gartner also notes that disruption of service time can damage the company's image and reputation. While it's hard to quantify, service outages can also cause companies to lose orders or work overtime to compensate for lost productivity – all of which make it easier to model in financial analysis.

The ratio of administrator to server. There is a big difference between the efficiency of the public cloud provider and the traditional enterprise IT department, and one of the differences is how many servers (or virtual machines) an administrator can manage. In traditional enterprise IT departments, a typical case is an administrator managing dozens of servers. For large cloud service providers, the ratio of administrator to server to thousands of is not a myth. The difference depends to a large extent on the level of standardization that large cloud providers are implementing in their operations. While it is not usually possible for an enterprise to adopt such an efficient approach, a private cloud can also provide a means to develop and deploy a more standardized service category for users, thereby reducing the amount of work that administrators need to perform mirror updates and patches. If the enterprise's hybrid cloud contains both internal and public cloud resources, the focus should be on keeping the environment consistent regardless of the physical location of the workload. Our ultimate goal is not to reduce the number of administrators, but to achieve rapid growth of the server without the linear increase in complexity.

What I'm talking about today is a number of numbers that users can use to provide some quantitative business justification for a live cloud deployment. At the same time, they have other benefits that users can use to set achievable metrics to determine the progress of the project and whether it has been successful. To do this, you need additional metrics to supplement soft and hard cost projects, such as the results of user satisfaction surveys. Now, users not only have the means to prove that you provide value to the business unit, but also have the necessary information to help users in the cloud to launch the next round of outreach work.

(Responsible editor: The good of the Legacy)