Shevigang: Dispel, the path of data center construction under cloud strategy

At the "2011 China Cloud Computing and Cloud Services Summit", Mr. Shevigang, director of strategic marketing at Schneider Electric IT division Greater China, made a speech on the theme of "dispel, the building of data center under cloud strategy", and the following is a transcript.

Hello everyone! The topic is simple, the impact of cloud computing on our future. A brief introduction to Schneider, like cloud computing, Schneider is also around us, the switch at home can see our brand, or we can use the wiring board. We are not the company that produces electricity, we mainly produce the control equipment of power transmission. At present, our solutions account for 72% of the world's electricity consumption, so that we become the global energy efficiency management experts, we have received the attention of the Chinese government, last year in November, when our Premier Wen visited our factory. And at the earliest time, Deng Xiaoping worked in France for 6 months.

The topic I'm sharing today is still related to cloud computing, and a lot of people are talking about what cloud computing has to do with me, and I have a little story on my side. The story happened in the United States, accounting firms often work overtime, so ordered a lot of fast food, and then sent downstairs to the people ordered to come down, but forgot to bring the best general ledger bills, pizza how much money, cola how much money, but no sum. How much is the last question? Courier saw the day said 850 dollars, people ask how can it be so fast? Because of cloud computing.

When it comes to cloud computing and the relationship between us, don't think that you are a small business will not use cloud computing, cloud computing into several stages, for ordinary enterprises, the data center below cloud computing is just a tool, will support your business process. For banks, electrical, or high-performance enterprises, cloud computing is an input to them, will allow a significant increase in competitiveness, can provide users with better services. We see a lot of cloud base today, Cloud Valley This kind of, regard it as the business, including electric company, still have a lot of trusteeship.

In the future, more and more small and medium-sized enterprises will accept from the cloud computing, or data center of the IDC hosting services, this is the future direction of development. The construction of any data center, if based on the original idea, will take at least 2-3 years. Looking back at what has changed over the past 2-3 years, we believe that with the green IT application, the 2009 virtualization has come into our every vision, and we say that we need to integrate resources with virtual machines to manage our IT applications. From 2010 onwards until 2011, that is, this year, including today everyone together, look up the sky, cloud computing is around us.

What is cloud computing? It's really simple, it's a whole new business model, all things, bandwidth, wireless, storage, networking, equipment, everything is already there. What changes have been brought to us by this new model to all of you here, and to the core of what the data centre is going to talk about today? 40 years after the cloud came out, the data centers, like mainframes, PCs and IBM-centric technologies, have emerged, including existing data centers based on the basic architectural approach of the data center 40 years ago. At that time a very typical feature, I want what kind of it application, I these more than 10 servers will provide applications. Cloud computing Today, this business model has broken the boundaries of the original physical equipment, unified IT resources, whether storage or operation, or the entire data transfer, it is integrated into a pool of resources, the unified call by the cloud. Regardless of how the cloud is presented, it ultimately needs to be delivered through the data center.

For the enterprise application is not the same, the private cloud, mixed cloud or public cloud, the service model is not the same, summed up, virtualization brings us the challenge is manageability, deployment of a significant increase in speed, once deployed, as the user more and more, need to quickly deliver it. At the same time, industrial density needs to be greatly improved.

All of this poses a problem for you, if you're going to have to avoid any bugs if you deploy quickly, how to determine where to add the equipment, anticipate future needs, and how to ensure that the data center supports business applications and extensions. Perhaps this problem can not be solved to seize the opportunity, a very important premise, how effectively we manage the data center facilities and data center based applications. How do we apply a framework that can be standardized and easily expanded to provide a more solid foundation for cloud computing? Based on this premise, how to achieve the company's overall value maximization, as a tool, quickly allow your business processes to be better supported by tools, or to make it more productive, or as a business, to quickly deliver the business to users, to enjoy the value of the business, all of which have a very big business with the company's business. Summed up, is how to improve the availability of data center, improve the flexibility of data center, reduce upfront input, this is the core of each enterprise in the actual data center operation process.

Just now the guest has talked about, virtualization is the core technology of cloud computing, the first problem that virtualization brings, let your server, or IT equipment to consolidate, density increases, load will increase, so need more power and refrigeration. In addition, because virtualization is a unified call to the IT hardware of the entire datacenter, it is possible to appear on service devices at different points in time, and it is also a question of how to improve the possibilities effectively. In addition, virtualization originally hosted a single it application, but after virtualization it could be more than 10 servers or dozens of servers moving to a server.

How to start with the power supply, give your data center infrastructure, and build a standardized power supply architecture. On the other hand, based on the "green" trend, how to effectively enhance the flexibility of the architecture, if you want to save energy, where energy saving? Where to cut emissions on the other hand, once the data center is built, it is also important to discover how the customer's volume has increased, how quickly it expands, and provides new cloud services to the user. On the other hand, based on tradition, I just mentioned the architecture of the data center 40 years ago, we are surrounded by refrigeration methods, but it is difficult to support high-density applications and load, which inevitably leads to the inability to abandon the original data center, from the traditional data Center architecture to the new cloud below the data center architecture to overdo. If you just plug in some new refrigeration equipment, you might also have a red section in the datacenter, which is hot. The only way to put your thermal load with your refrigeration facilities to do an effective connection, shorten the hot and cold air transmission distance, dynamic increase in refrigeration capacity, effectively improve refrigeration efficiency, can also achieve energy-saving emission reduction.

Virtualization can dramatically reduce it's physical equipment, which is a good thing to do in terms of energy saving and emission reduction, and the efficiency of many data centers can be lost if the mere virtualization of it devices does not make the architectural alterations to the physical facilities of the data center. On the diagram, the efficiency of the data center is measured by the PUE value, the lower the Pue value, the higher the efficiency of the data center. As an example of this datacenter scale, if you don't change the infrastructure, your Pue value increases, the power consumed by the unit KW load, you add 0.56 kilowatts to the infrastructure, which is a waste. If you implement the infrastructure optimization, it will lead to greater improvement.

There are two small tools, one to help you evaluate the efficiency of the datacenter, how do you know how much electricity is used in your data center, how much is the Pue value, and this is dynamic data. We can login on our website, if interested can email me, I will send the link to everyone.

After virtualization how to foresee the savings that virtualization brings to you, when you think about implementing the cloud strategy, the data center can reduce the amount of electricity, this is the dynamic tool we can provide, can help you anticipate it ahead of time. Once you take a cloud strategy, your energy efficiency will increase.

Another topic, after the cloud, you may face another situation, I want to add equipment, I do not know where to put, I want to understand the data center of power consumption, I want to know which part of the bill is consumed by which system, this series of problems need to consider the data center management of the upgrade. At present, we consider the management of three parts, the first is to plan, the construction of a data center, to understand the requirements, that is, the future of how to avoid potential new technology problems. The second part is the operation, the operation of how to consider the dynamic addition of new equipment, taking into account the needs of equipment, if the reduction of equipment, may affect the relevant load and the relevant structure is what. and later actual capacity management is also an important part.

The core part of the entire cloud that we have to consider is to accelerate deployment. A core direction, what we have to consider is modularity, building blocks, such as container-type data centers, a picture of our Schneider data center, with a power of 20 MW, each small data center, which is also made up of smaller modules that can be substantially increased.

To summarize, under the cloud strategy, what is the impact on the datacenter? First, data center standardization is required. In addition, the need for data centers to be flexible, to increase today to increase, to reduce the reduction. How to meet rapid deployment requires data center automation, tailored to your needs. For the entire cloud, the core requirements of our datacenter are that data centers are not only based on cloud computing, but also on the whole of the human race, and are influenced by government systems and the social environment. Now everyone has smartphones, and many applications are based on data centers. Now the data center is based on the original business-oriented, support the company's business, now affect our quality of life, if one day your mobile phone Internet function can not be realized, can not micro-bo, can not QQ, this may be a big problem. So many things will drive the data center more and more, for different data centers, the focus is not the same, I do not give you detailed introduction.

Finally share our Schneider solutions. In this part of the data center, the overall goal is to reduce unnecessary rework and ultimately provide an integrated solution. Examples of this solution include power distribution from prophase to monitoring, to building automation control, and to final distribution and refrigeration parts. Based on just a few big challenges, we offer the products and solutions.

Finally, we are doing some cloud computing projects. Also our Chiyu group project in Harbin, we have a project at the invitation of Harbin Economic and technological Development zone on June 18. North of the Disaster preparedness Center is also our unified construction of Schneider, from the software management platform to the actual application platform.

Time relationship, I will give you share so much, if you are interested in our products can see my contact, thank you!

(Responsible editor: Duqing first)