Schneider Shevigang: Cloud computing brings four problems to data centers

The 2nd Session of the "2012 China Data Center Industry Development Conference" was held in Beijing on the 20,112-month 17th, and Shevigang, director of strategic marketing at Schneider Electric IT division Greater China, released the keynote speech "data center, at fingertips". Shevigang said that in 2050, global energy demand would double and share some of the energy problems of data center users in current data centers or cloud computing situations through effective solutions. And said cloud computing will bring the entire data center industry should be four aspects of the problem: power, possibility, manageability, and energy-related problems.

Shevigang: Distinguished leaders and some of our data center's senior experts are good. It is a great pleasure to have this opportunity to share with you our solutions to the whole of this part of the data center. I think we have mentioned earlier that just now a colleague and I changed my business card and said, "Busy today." However, in the current situation, I want to do the data center is not busy, and in this context, now we have to face not only in the data center of the new technology brings some challenges. At the same time, we need to consider the impact of the data center on our entire environment under the task of the whole society.

Schneider Electric IT division Greater China Strategic Marketing Director Shevigang

The first thing I want to mention is that we have mentioned before that in the coming 2050, our entire global energy demand will double, and if we do not halve our current carbon dioxide emissions, our entire climate will change seriously. Therefore, as a responsible enterprise, we must consider how to effectively avoid frequent power outages, curb the rise in energy prices and supply and demand contradictions at the same time, how we through our effective solutions, the current data center or in the cloud of data center users of some of the energy problems.

To introduce Schneider, Schneider's 2010 turnover reached 20 billion, the total turnover in the new country is now 37%. Schneider Electric currently has 23,000 employees in China, has become the world's largest number of employees, and in China already has 77 offices, 26 production base. We have equipment for buildings, which means that in this area of the current data center, Schneider Electric is available in addition to IT equipment.

The next step is to say that today's data centers have an impact on cloud computing or IT integration or virtualization for the users of our current data centers. If we look at the entire data center, it should be said that the entire data center is divided into three categories, one is the enterprise's own data center, this kind of data center is a typical feature of our data center as we optimize our business processes to support the implementation of our business, such an important tool, So we built it as a data center as a tool.

Another category, that is, early in the financial, insurance, securities industry, the data center development is more advanced. In addition, in the last year, we also heard that the two years ago, the Tianhe in Tianjin, was the world's fastest super center. Data centers like this one are more likely to use data centers as an important means of improving the competitiveness of their businesses, attracting their customers, and then upgrading their business's competitiveness. So, from this point of view, the data centers in this category should be their data centers as an investment. But there is also a growing number of government-led or High-tech parks or our carriers involved, and then the data centers in this category are dedicated to the data centers of cloud computing, which they regard as their main business, So we call it the data center is business.

For different businesses, because the data center is not the same as the enterprise's business relationship. So we can say that the need for data centers is different. However, an important trend we have seen is that, in the entire enterprise-class data center market, with the maturity of cloud computing, security continues to improve, we see that hosting has gone into more and more small and medium-sized enterprises. If we look at what Schneider is playing in the data center, the interface is an operating platform at this level, and the next round is it hardware, which involves networking and hardware. Another aspect of the overall information management aspect of cloud computing is that we refer to the physical infrastructure of the data center.

For now, we can provide complete data center services for both enterprise-level data centers, cloud computing, or data center hosting. When it comes to cloud computing, people will mention, now I think in this National Convention Center, there is another meeting today, also talk about cloud computing. And we see that in 4 May, cloud computing will be like the data center of cloud computing, really is surging. When it comes to cloud computing, it's not really a new technology, it's just a business model. Virtualization is one of the core technologies of cloud computing, and virtualization can have three effects on the data center, high density. In addition, cloud computing resources to the entire pool of resources to allocate, because it runs a different time, he will be the development of W time, the actual it complex and dynamic migration. In another part, cloud computing requires virtualization to treat all computing resources as a resource pool, that is, any one it server, or a network of devices, will assume more computing tasks or information exchange tasks. This time requires a significant increase in the actual server's criticality.

At the same time, cloud computing can effectively reduce the number of IT equipment, from this point of view, is very responsive to the current overall trend of energy saving and emission reduction. But, in short, cloud computing can bring four problems to the entire data center industry: power, probability, manageability, and energy-saving problems.

If you look specifically at the data Center for all of us here, what if you build a datacenter? If I need to operate, how should I operate, if need to manage, how should manage? From this point of view, Schneider Electric has a very sophisticated tool. If you need planning, you can use our traditional tools online, which is completely free and can be used directly online on our traditional website. If you want to operate, you have to manage, or even maintain a data center, we will provide you with such a tool today. On the one hand, we can based on the actual your IT site in the corresponding operation of data, including the corresponding power supply data, temperature and humidity data, as well as energy consumption, IT equipment numbers, and finally concentrated in the data center management platform. Then we analyze the energy consumption of each subsystem in your data center and the waste of resources through effective analysis. If you think about two days to do the corresponding repair and maintenance of the room, you can through our platform to understand some of the problems and further optimize your infrastructure.

The actual operational level is divided into several parts, on the one hand, at the infrastructure level of the data center, which is what we normally talk about, as well as the power distribution facilities. Another part involves the actual operation and management of the data center. There is also the section on data center upgrades and further equipment management. In the end, perhaps for different people, for example, what management may need to know is an intuitive, understanding through an interface, which we have a corresponding energy-efficient dashboard can help you just show the value of the direct pue of your data center, and the energy consumption of each subsystem. At the same time, remote monitoring is also possible.

Further, we are actually in the entire data center, we can on the one hand for different decision-making level of people, or in the data center involved in the different members of the corresponding management. If you are just a data center management, is the CIO level, is the actual data Center manager, this time we can through the effective dashboard, help you to render the data center Pue value, also can help you to present the data center in each subsystem of energy consumption, It can also help you understand your peak situation. If you are a facility for this level of management, then it is possible for each device to actually load the situation, and then the actual topology diagram and pipeline links may be very important to you. This intuitive graphic will help you understand the operation of data center failures. If you go to the IT device in the data center, it's another level, on the one hand, if you're going to deploy a new server, you have to be fully aware of the deployment of the various parts of the data center, the actual deployment diagram, and the general direction of his stay, as well as your power supply, refrigeration, and overhead racks, The actual use of cyberspace.

This part in fact we not only exist through the actual use of the computer interface, but also in the iphone, or now behind the smart terminal to do the corresponding use. In addition, if you have a corresponding data center at every point in the country, further through this section, you can make a corresponding presentation of this situation in the National Data Center.

The problem is, in the actual operation process, I now have a few data centers, all the data center in the country, if found the problem, how do I understand the actual problem, what data center? Where is the actual operation? We can further understand the operation of each cabinet through a further layout diagram. If we have just seen an alarm, this alarm when we can find that, in the C air conditioning this level, specifically, you can be specific to the situation of the failure of a detailed description, so as to help you eliminate the actual operation of the fault. The other part, you can also understand your current opportunity, what the above line is, the corresponding distribution line exists in which cabinet, this can have a very intuitive understanding.

If you're going to overhaul a space right now, you can do a corresponding simulation between repairs. If I have to do a maintenance on this device, that is, the UPS, you can get a closer look at the device list for all the speakers by simulating. For the equipment list, if your air-conditioning to carry out the corresponding maintenance, you can also do the corresponding simulation, to simulate the corresponding lower limit of the air-conditioning, the surrounding equipment temperature caused by the impact. You can see in this, if this C air conditioner, we turn it off, you can know, in the entire data center, which cabinet of the air inlet temperature will increase, you can determine what the next step to take action.

Further, if you are the actual operator of a data center, you can use the entire software to understand the entire power consumption of the data center in real time. On the one hand, you can use the corresponding layout diagram, this layout of the toilet, and then understand his cabinet inside the whole of this configuration is what kind of, what kind of power lines, probably provide how much capacity. On the other hand, you can also do the actual physical operation of the data center, like I'm in the data center, I want to know what the actual operating temperature of the data center is? What is the corresponding humidity? We are here to provide real-time wiring, to help you understand, and to help you determine the appropriate security domain governance, once more than this domain governance, you can get the first time the corresponding notice.

Now we see in "Twelve-Five", in the whole data center related to this part, on the one hand is cloud computing, on the other hand is the whole energy-saving emission reduction, or energy saving this part. What do I think is the actual energy efficiency metric for any data center? What is the energy efficiency report of our actual phased operation? Directly affect the efficient management of the data center, we can help you determine the energy consumption of each subsystem in the datacenter, and we can also help you understand the carbon dioxide emissions corresponding to your energy consumption throughout the datacenter. To learn more about your space for ascension.

On the other hand the purpose of the whole management is to further optimize the operation of your data center through simulation. Now, for example, you now know that with the Pue value and the actual situation of each subsystem, you know that the consumption of air conditioners now accounts for nearly 50% of your data center. What kind of things can you do to lower your data center's energy consumption? In this case, which air-conditioning is can be turned off when unnecessary, in this area, we can through this simulation, and then further understanding. If we turn off the number 11th air conditioner, you can see that the actual operating conditions throughout the data center have not changed significantly in temperature. If you turn off the other two air conditioners, the whole actual operation of the data center is still in a lot of domain governance, you can know that in the data center, the previous closed to the several air-conditioning, now is not necessary when it can be turned off, which can greatly reduce the data center air-conditioning consumption. Also, you need to understand the entire, actual load, peak value of a cabinet, and if so, how to adjust the hotspot. So you can go through a simulation like this, and learn more about what the IT load is causing the hotspot? After this load is caused, how do we get the actual operations that exist on this it load and then move on to other IT devices, we can see that in the green area of the data center, there are actual resources, we can put it equipment, the corresponding commitment to the budget, Migrate further to this area. This time, you can further promote the analysis of airflow organization to further understand that this hotspot has disappeared, there is no hot spot in the data center, so as to further optimize the operation of your data center.

If you know that cloud computing brings two challenges, the first scale will be infinitely magnified, and this solution is through modularity and standardization. The other is that cloud computing requires a significant increase in your deployment speed. It's possible that your sales went outside today and signed a list of 500 servers, or 200 servers, or 10000 servers. This time, if the small temperature of the server online, how do you in the data center where the most reasonable placement? We can add this device first, and we'll be smart to help you find out where the device should be in the data center. Then we will further generate this dispatch Tampai sent to your corresponding IT equipment supplier, let it directly according to this to you. Also, you can define periodic reports based on your needs.

I just mentioned the security implications of virtualization, how do I put security through this effective management? We have a simple example here, if you are in the data center, you now have UPSA and B to power, if there is a switch to battery power, you can through our entire central manager to further inform your actual this virtualization manager, which devices are currently in danger, You can find this risk in the case of a high-density chassis of both servers in yellow. We can further through the virtualization of the management of a corresponding migration, and finally return to our entire central manager. In this case, if you have a UPS power supply some fault, so you can intelligently get the corresponding notice, from the preventive, leading to a larger failure occurred.

For the entire data Center manager, at a higher level, you can use a simple dashboard to understand the actual pue value of the data center, and then you'll learn about your power and refrigeration, and then the corresponding capacity changes applied, the actual peak of the operation, At the same time you can also generate some customized reports, including your failure report, and your new device report. In the entire data center virtualization brings a rapid concept, basically a device failure, we can through the central manager to do the failure of equipment migration, and finally achieve a balance. So in our entire data center management software, can from the fault simulation to the actual troubleshooting, to the actual if the fault occurs to achieve a fault balance, can actually do.

What's more, what we're offering now is the level of all intelligent terminals, including your black mold phone, and your iphone, which can provide you with real-time viewing. As long as your data center can be directly connected to the corresponding Internet, and your corresponding intelligent terminals can be connected to the network can be achieved. You can download directly from the Apple App Store, and you can download it directly to get an idea of the actual operation of our data centers in various parts of the world.

Specifically, now since last September, we have launched the data center as a whole management platform such a software system, there are already a lot of applications around the world, in China now has the corresponding 2 operators in the actual data center, has used our energy efficiency management platform. This is part of a global financial institution that, through our software system, finds that the extra traffic in the data center can be deployed to more IT equipment, extending its data center life by 3 years. On the other hand, a software platform like this can help you reduce the number of people in your datacenter. Currently in China, on the one hand, in Guangdong, another in Henan, two different operators have been mature use of our Schneider Electric energy efficiency management platform.

Today's sharing hopes to tell you that, on the one hand, we use the latest technology in the entire data center, virtualization, can also use our high-density, this is not enough. On the other hand, we need to optimize the operation of this infrastructure within the data center as a whole. In other words, we are in the data center, we can through the mouse, even at the intelligent terminal, through the finger operation, the entire data center management to do real-time, further understanding of the data center inside the details, or each of the actual operation of IT equipment. My share will be here, thank you.

(Responsible editor: The good of the Legacy)