It's time for the cloud service provider to seriously consider "automation."

Some enterprises or organizations in the adoption of cloud computing architecture, it is found that the response speed of IT systems not only not significantly improved, but slowed down, which in some way affected the enterprise procurement of cloud services enthusiasm. An important reason for this situation is the lack of automation capabilities of cloud solutions or cloud services. "After more than 10 years of development, when cloud computing from the beginning of the local application development to today's scale, has been infiltrated in many areas, the complexity of the cloud with the scale is also gradually emerging, especially for large data centers as the bearer of integrated cloud computing systems, especially in this case, there is an urgent need to enhance the cloud automation capabilities and agility. "This is the right way to deepen cloud applications," said Zheng Jun, chief technology officer of Wuxi Huayun Data Technology Services Limited (hereinafter referred to as Huayun data).

Proving

There are many characteristics of cloud computing products, including high performance, security, stability, usability and so on, but it is very important but few people mention the feature of automation. Why does this happen?

cloud computing from birth to maturity there is a process, simply summed up there are three stages: the first stage, the user first understand the cloud computing infrastructure, the most talked about is the IaaS and each vendor launched the cloud host product, the second phase, the user wants to move more applications and data into the cloud, As a result, manufacturers have introduced more kinds of cloud computing products, in addition to cloud host, cloud storage, data services, software-defined networks and other products are increasingly popular, the third stage, that is, the current stage, cloud computing has been from the initial small-scale attempt to develop to large-scale applications, scale is the inevitable feature of cloud computing. Individual cloud computing products have been difficult to meet the needs of users, users need more features, faster deployment, the use of more convenient cloud solutions and services, which requires a complete cloud computing solution must be in the background of all the individual cloud computing products seamlessly combined, according to the needs of users to achieve automatic deployment, configuration and operations, Reduce capex investment costs and opex operating costs. Current technology trends are also validating this, from the early years of Web service to distributed computing to the advent of today's "microservices" Concepts and technologies.

"Previously, users were concerned about having or not some cloud functionality. Now, users are more concerned about how to use the existing features, and give full play to their value. "This is somewhat similar to the development of smartphones," Zheng Jun said. Previously, in an era when non-smartphones dominated, people were satisfied even with a simple smart phone. Now that smartphones have become mainstream, people are more concerned about which smartphone operating systems are easier to use and more fluid to meet the needs of their users. ”

Gartner conducted a survey, 74% of respondents believe that to achieve business growth, the immediate task is to improve the level of automation of infrastructure.

Enterprises in the building of a unified pool of resources and virtualization has been a qualitative breakthrough, now more need to deepen the private cloud products, can open up the IaaS and PAAs layer, to provide application-driven integrated resource services, rather than just stay in the virtualization phase. From the user's needs, they are increasingly demanding the agility, efficiency, and resiliency of the cloud, requiring a more automated cloud.

"Automation is an advanced feature of the cloud and an important criterion for identifying the maturity of cloud computing. "It's time for users and developers to think carefully about automating the problem," Zheng Jun. ”

As important as the production line

Why is automating this issue on the agenda now?

In the initial phase of cloud computing applications, the primary task is to virtualize. But now, the need for enterprise users for the cloud is far beyond virtualization, and cloud computing needs to provide support for hardware, software, and services that are far more virtualized. With the rise of new applications such as mobile computing, big data, and "Internet +", most businesses are transitioning from a traditional IT manager to an external Internet service provider and internal business service provider.

"Cloud computing goes to a certain stage, from simple to complex applications, and without automation, cloud operations will be very difficult. "In the Zheng Jun," If there is no automation, the enterprise wants to live a complex application of big data, the learning curve is very high, the implementation of the difficulty will increase. Ford Motor Company founder Henry Ford's biggest contribution to the automotive industry is the creation of the world's first automobile production line. This kind of automatic production has played a great role in promoting the large-scale development of automobile industry. Today, the importance of automation to the cloud is also true. ”

Imagine a user entering a simple command line or a "drag-and-drop" for intuitive and simple interactive operations, and thousands of machines in the background can automatically create complex services. This is the charm of automation. For users, the greatest value of automation is that it can mask the technical complexity of the IT backend. Users are most concerned with SaaS applications, app apps, and don't care how the IaaS or PAAs tiers support cloud applications, not to mention how OpenStack and Docker technologies work. "Automation can mask the complexity of technology. As long as the user needs to ask, the background can be automated to provide users with the various cloud resources and functions required. "Zheng Jun told reporters.

Many surveys show that the biggest problem that currently hinders the cloud for Chinese users is the security of the cloud. In fact, the security of the cloud, reliability is not the problem of technology can not be solved, but the user's psychological problems. However, the complexity of the cloud is a problem that can be solved by technical means, that is, to achieve cloud automation. In many enterprises, IT personnel do not understand the business process, which sets up a barrier between business applications and IT solutions. Only by matching applications with it can you streamline business processes and create value. However, it is difficult for every enterprise to equip people with both it and business knowledge, and it is very expensive to cultivate such talents. For example, if an enterprise to buy a large ERP system cost tens of millions of dollars, implementation and system integration and spend two years, then in a need for business rapid innovation in the era, such speed may not be satisfied with the enterprise. The importance of automation is self-evident.

"Take doctrine" and the development of customization

At present, the level of automation of domestic cloud service providers still needs to be improved. Although each cloud service provider can provide dozens of or more cloud computing products, but does not fully achieve the technical background of these products, to achieve automated resource supply and operation. Huayun data will release some new products related to resource management and continuously improve the automation level of its cloud platform.

Previously, applications were relatively small and simple to deploy, and enterprise IT staff could procure software and hardware and then manually deploy and manage them. But now, with the increase in applications, software upgrades are increasingly frequent, manual configuration has not been keeping up with the speed of application development. IT systems have the ability to automatically upgrade and scale out according to the needs of the application. In the case of non-automation, enterprise management of 100 server devices, may require several people or even more dedicated team, but in the implementation of automation, the same number of equipment management may only need one or two people, not only reduce the human intervention can cause errors, and operation and maintenance costs lower. Automation is both the user's demand, but also the manufacturer's efforts to promote the direction of technology development.

"Previously in the enterprise, IT operations to be done by two groups of people, a group responsible for system operations, another group responsible for network operations, two groups of people full-time different. Now, cloud computing to all systems, hardware and software through, to achieve the overall operation and maintenance of automation. "Zheng Jun stressed.

This is an open source era. To create a more optimized cloud platform for automation, open source technology is needed. Using open source technology prevents users from being bound by vendors. However, it is important to note that any open source technology can not be used, but on the basis of open source technology for the enterprise's specific needs of the development, optimization and testing, so as to ensure that open source technology to use the edge. For research and development enterprises should choose suitable for their own open-source technology, Zheng Jun gave advice: first, the Enterprise to master some of the core algorithms, and the open source technology and the company's development strategy tightly integrated; second, enterprises should be based on the open source technology on the basis of customized development, to carry out the necessary system-level optimization and engineering testing, Integrate the open source technology into its own product system; Thirdly, the enterprise should use the agile development method to iterate innovation and enrich the function of the product continuously.

"OpenStack is an excellent platform and a standard in the industry, and development in compliance with this standard will ensure that it is in sync with the mainstream trends in the industry." However, the stability of the OpenStack platform itself has to be improved, and it only provides the basic framework framework platform, the enterprise based on this platform for the development of complex advanced features, must master a certain technical approach. "Zheng Jun said.

Consolidate Cloud data center infrastructure

One trend now is that IaaS and PAAs will be further fused. Zheng Jun that: "From the perspective of technological development, IaaS and PAAs will be integrated, but sometimes in order to meet the needs of different customers, IaaS, PAAs also to separate, because some customers need a separate IaaS or PAAs." But more users are concerned with the application, rather than the specific technology implementation approach. Our cloud computing systems provide unified asset management and services for app applications, and IaaS and PAAs are a resource organization. For an integrated cloud service provider like Huayun data, it is essential to get through the IaaS and PAAs tiers and provide a holistic cloud computing solution. ”

Gartner's survey data shows that since 2013, the adoption rate of hybrid cloud has been increasing globally, the adoption rate of hybrid cloud in Asia-Pacific and Japan has reached 24%, and the importance of infrastructure automation and hybrid cloud-based innovation is becoming increasingly apparent. Many users choose a hybrid cloud because of the agility and flexibility of a hybrid cloud.

Some users who have built a private cloud have come to realize that some noncritical applications and data can be placed in public clouds, not only for convenience but also for lower cost. Zheng Jun believes that hybrid cloud services compete for reliability and availability. Huayun data has 20 data centers in 15 cities across the country, offering a diverse portfolio of cloud-enabled products and a leading cloud-supporting operational capability. "With Internet services, cloud computing does not differentiate between regions. However, the choice of regional cloud data center can optimize the service by considering the specific situation. For example, Huayun data centers in Wuhan can provide local customers with nearby services. And some cloud service providers because there is no data center in the local, can only rely on remote data center to provide services, from the service reliability, connectivity, cost and other aspects of comprehensive consideration, Huayun data occupies an advantage. "Zheng Jun said.

In 2015, one of the business priorities of Huayun data is to continuously optimize and expand existing cloud data center infrastructure while improving its research and development capabilities, consolidating cloud data center infrastructure, and providing customers with better network resources, computing resources and different kinds of cloud services.

This article from "Guo Tao Storage World" blog, declined reprint!

It's time for the cloud service provider to seriously consider "automation."