How Forrester:it enterprises to control cloud computing

Developers love cloud computing, because cloud computing makes their work easier. Infrastructure and operations should not resist this trend, but should try to accept it. The recent Forrester report provides some advice on controlling cloud computing.

The report from Forrester, "a new era of it responsiveness and efficiency in cloud computing", describes how it organizations should deploy and control cloud computing and provide some advice to it organizations.

The report is divided into two parts:

1. How should application engineers and software engineers respond to

2.IT infrastructure and operations should respond to enable the entire enterprise to meet its needs through cloud computing

Abandoning concerns and accepting cloud computing

The function of cloud computing is what software engineers expect. Public cloud computing (essentially Amazon Cloud computing services) enables rapid resource access, high productivity, and cheap development. With these features, the developer's work becomes easier. If the private cloud established by the enterprise does not provide these capabilities, the developer will refuse to use the enterprise private cloud. This means that it organizations need to evaluate their private cloud deployment plans to assess whether they meet the requirements of developers.

Forrester analyst James Staten "scoffed" at some IT companies deploying virtualized environments as private clouds. Crucially, developers now know what they want and cannot be satisfied if they do not meet their requirements. According to a Forrester survey, most developers are dissatisfied with private cloud deployments. Their attitude can be summed up as: "If this can help my work, I will use it, but if not, I will continue to use the cloud environment I have been using." ”

It is impossible to increase efficiency if the enterprise forces developers to use "official" cloud computing.

Solution Choice IaaS, PAAs or others?

One interesting finding in the report is what software engineers are looking for in the development framework. The infrastructure for virtual machines, storage, and network connectivity-services not only hinders productivity, but also forces developers to manage pipelines rather than focus more on application functionality.

The solution to this IaaS environmental problem is the platform, the service. The logic is that the programming framework provides a common service for developers, freeing developers from handling details, and relying on frameworks to perform the necessary tasks, including persistent data storage, identity management, and more.

If all you want is a framework that provides these features, the framework works well. However, applications often require features that the framework cannot provide, and sometimes require direct access to lower-level functionality, and a high productivity framework becomes a shackle.

Forrester believes that developers need to mix capabilities, that is, they need to help them solve the tedious work of the service, but also need to be able to access the lower level of functionality. Amazon's cloud computing services are in line with their requirements, with AWS's rich development services, as well as direct access to lower basic functionality for developers to deploy or install features other than AWS.

The report notes that independent software vendors (ISVs) have also entered the market to build SaaS versions of their infrastructure products. For example, the Informatica company now offers cloud-based consolidation services that allow developers to consolidate different applications. As a result, enterprise applications are becoming more and more rich, including Home-grown code, cloud services, and the functionality provided by independent software vendors. Developers are more about assembling than writing applications.

Five things the infrastructure and Operations department has to do to control cloud computing

The report conveys an unmistakable message that infrastructure and operations must respond to the expectations of these developers, or they will be eliminated by the times. What we need to do now is to achieve flexibility, not just managing assets.

To this end, Forrester offers five recommendations:

Become a service provider. Infrastructure and Operations (I&O) are now under pressure from external service providers, who are fully committed to trying to help developers. I&o need to respond positively. My point is that I&o needs to simplify its approach and avoid traditional custom service methods. While providing a custom configuration appears to be customer centric, this can increase overhead and delay time. To become a service organization, I&o needs to provide a rich development environment and a variety of important application services. I&o should also create a list of pre-approved SaaS vendors ' products, allowing developers to quickly view their products.

Create a service directory. It's very powerful. Creating a service directory can provide developers with pre-configured and repaired mirrors for quick access, rather than forcing developers to deploy their application configurations from scratch, and may also be misconfigured or ignore critical patches. Pre-configured mirroring accelerates development and is extremely attractive to developers. This allows I&o to ensure that the application runs according to pre-set configurations and the latest version, which is more effective than post development audits.

Let enterprise architecture design patterns for developers to use. It is recommended that you use an enterprise architecture to create application designs and patterns and provide them to the development team. This can bring a lot of benefits, reduce operating costs, better skills sharing.

Cost transparency. According to the report, a pay-for-use model is attractive and allows you to make sure that resources are used efficiently.

Unlike enterprise internal equipment (involving initial investment), paying by use involves the entire lifecycle of the application. Forrester recommends transparency of all the costs of the application, which allows developers to scale and shrink applications as appropriate, while operations can run applications more efficiently.

Enables application lifecycle flexibility through DevOps. The developer's self-service is very attractive and certainly offers many advantages, including improving engineering efficiency and making engineers happier. However, failure to achieve flexibility in the application lifecycle means that the enterprise cannot gain the benefits of all the financial and market responses provided by cloud computing.

Achieving lifecycle flexibility requires optimization and coordination of all parties. Even if you hate devops, you still need it. Now that cloud computing can provide resources without delay, it processes also need to eliminate latency, which is also the role of DevOps only if development and operational collaboration can be implemented.

The message from the report is that the rapid development of cloud computing makes it imperative to reform existing processes, and the development team that expects cloud computing will not wait long.

(Responsible editor: The good of the Legacy)