Cloud Base Reitao: The great wisdom of the big data age

Good afternoon, everybody! Thank you so many colleagues can return to this meeting this afternoon. Let's introduce our cloud base, Cloud Base is an organization, the field is always the leader of our base, with capital and base of this new industrial model invested 13 companies, from infrastructure, such as container-type data center to SaaS software, our topic today is more specialized in a certain field, Because we have developed rapidly in the past two years, since last October, has produced a lot of economic benefits, the production of these servers have a certain scale. But we have also seen in the large-scale IDC data center at the same time, what is the real drive we can let cloud computing truly landed, can give the country to make some contribution, below I share our cloud base in cloud computing this area we will focus on in which areas to contribute.

In the context of the former globalization, Tian is now in Silicon Valley with the more advanced North American plants to discuss the acquisition of cooperation. The whole industry chain has been full, but we also see cloud computing at present from our large-scale construction there are some obstacles and cycles, I share a report, what factors affect us? such as security and other issues. IaaS we've done a couple of big successes since last October and we've gained some experience, but we've also had some problems, and some of the recent earnings reports have revealed a problem that, before the scale effect, its profitability model is questionable. Now that profits have fallen, cloud computing is the biggest problem.

What does cloud computing offer after walking through the IaaS level? This cycle may develop very quickly, some of the domestic advanced manufacturers, such as the morning of the Cloud Express line, has provided the IaaS service scale, then the next wave of cloud computing tide where? Today, director Xu also said big data economy, in fact, I would like to talk about the big data today. Big Data is one of the hottest topics in North America, where there are a lot of seminars, and there are different levels of applications that are related to big data, so let's take a look at this big data.

What is big data? IDC gives this forecast, what is EB? Now the data nine years later only 2%, I also saw another set of data, is seen on the micro-blog, there is no textual research, writing is the next 200,000 households online, you can see the data is very rapid growth. If this data is rubbish, if it is in the data center, how can we turn it from data to information to produce value? That's what we're going to talk about in a new core productivity of cloud computing, a use of big data.

McKinsey gave a report this April, not an IT report, an economic model analysis that details the impact of the data itself on the economy in several industries. To share, for example, in the healthcare industry, the big data itself is based on McKinsey's plan to generate 300 billion of millions of dollars in productivity, equivalent to twice times the size of Spain's entire healthcare industry. Public utilities, equivalent to 250 billion euros, basically equals Greece's GDP. From a global perspective, the economic value of this data is estimated to be 600 billion dollars, resulting in 140,000 of jobs in North America, and data assessment-related jobs, the data itself is creating economic benefits.

What are the characteristics of the data? Let's look at the data itself in the cloud era. Here's an example of a hospital in North America that has generally started using a community network, where a doctor's advice is not written on paper, but on Twitter and on his personal web page, blog. The corresponding information is used as a medical diagnosis, which is not accessible by traditional IT systems. Some professional database-level systems are certainly unlikely to capture information on these other community networks, so large data itself has some challenges.

The second challenge is the massive nature of the data. We also see that a person's life may produce a PE data, including the first book that you like, the first music, or a piece, or a photo for your children, or a commemorative note that's stored, everyone can reach a PE, and in this environment it's not about the capacity. But how can I put it in, how can I get it back.

I read a message yesterday that Amazon evaluates storage energy, not the PE level concept, they have 450 billion objects, described in this way. Why is this description? This is a new type of mechanism, how it can be put in and how it can be taken out, millions of things can be singled out this is the next generation of storage challenges.

At the same time, also appeared the economy itself innovation, appeared the new economic form. For example, like Twitter, the last Beijing Economic and Trade Commission also talked about the first day is Li Na won the championship, she won the title within five minutes Sina Weibo has 300,000 of the forwarding. So much information itself, from the perspective of government functions, how to control this information? How do you use this information? For example, to do the public opinion analysis system, also talked with the Beijing Economic and Trade Commission. These systems can help you in the event of public events, we can not always do afterwards, through the information itself tracking and digging, we could do the preliminary work.

At the same time, we have discussed with the operators, but also with mobile communication, there are many based on the data itself, we are doing the project is to provide intelligent analysis, behavioral analysis. For governments, such as real estate transactions, there are also public utilities we have done some success stories, such as the smart grid field, each information is a local calculation after the implementation of the overall submission, previously through the traditional architecture to operate such large-scale data is not possible, these must use the cloud technology of large data technology.

We look at the big data technology is not spring? Everyone reads the financial news every day, each rolling data, such as why oil is 117 dollars instead of 116 dollars? To further say Greece, the sovereign funds have lowered two levels, they are more energy than the U.S. military, how to achieve? What size do they look like? For example, 17,000, can be implemented on 100,000 processors, while in less than a millisecond to achieve, break the traditional data center structure framework. Why the age of SNP? Our generation grew up in the age of SNP because it was fast enough and strong enough. If cloud technology organizes thousands of servers, the computing power gained is greater.

How to make big data itself produce great wisdom? Let's explore the areas in which cloud bases have been involved at this level. We've defined a number of areas that we can relate to, such as the top right-hand side of the product, the personal cloud, some of the small companies we've acquired and some of the Finnish manufacturers. The infrastructure itself, these four blocks, including infrastructure, our goal is to do Asia's largest, the Beijing municipal government has also invested, is currently more than 60,000 capacity, planning capacity is 500,000 servers, as well as infrastructure, containers, as well as virtualization products, virtualization platform, our leadership is also relatively strong. We also implement automated deployment tools internally, at the infrastructure level.

The other two are our focus on cloud computing and data storage, which now has two companies working on the product.

We see this area is very hot, divided into several major branches, (algorithmic trading) We are involved in more SOA, this area to do more. So many applications are originally run on the server, how do we let it to optimize? This is the key to consider, is how to make it more processing power. When we talk to the manufacturers, they say that our current stock trading, these transactions can only do little, if we can break through this way, is the operation can be fragmented, with this to help him achieve distributed computing. The traditional business, the right red flag is can be cloud business, transaction types such as complex event processing applications, as well as the transaction itself analysis and the behavior of the data, these are the big data involved in the application area. At present, large data direction, a few big companies put more in the above, also are not some troupes, field total investment are some teams in North America.

We can provide two optimization systems, for example, many enterprises are in the use of distributed computing, open source is to help everyone to try the first way, but open source itself does not have the ability to expand, there is no enterprise-class features, we have a lot of samples of the base, made a platform, is that customers do not need to write their own We help you with your career practice, which is the current business practice mechanism that can be implemented.

The other is more subversive, that is, we use the new peak calculation plus storage mechanism to achieve. We can see this graph, the left and right contrast is the traditional storage and the new generation of storage characteristics, traditional storage can not do too much, if the extension how to achieve? Because there are some data inside, not information. What is information? When I add this information to the descriptive label, I know what this is and then the information, so we use the information store to distinguish it from the traditional pattern. With this container, it is completely a flat structure, through the provinces, across the region, this can achieve the bottom call, break the capacity limit.

The last time you can implement a distributed computing platform to trade, for example, the right side of the blueprint, I used a lot of terminology, maybe we listen to more headaches, if you can see clearly this, a simple example, how to count how many triangles in this box, if the number on the left, we use a distributed framework to open a lot of Count yourself first, and finally merge, this calculation is a distributed computing. For example, real estate transactions, I do not need to aggregate all the data into the data center to calculate, the large database to calculate, this is unrealistic, and now many operators also see that there is no such ability to put so much data into all the storage. How do you do that? This is done locally, in the housing exchange, by making full use of local resources and using cheap platform resources. The difference from the traditional structure is the picture on the left, which is not the structure of the past, but the two directions of use. The more on the left is to see the application server, what is the right face? is business logic. These two are not the same, the calculation is very large, this is the characteristics of cloud computing.

What is the benefit of this framework? I take an example of the medical industry, such as regional health care, how do you want to do disease protection, from a hospital to another hospital, how can the original X-ray film continue to use? is the existence of a system that is managed through a database. Now regional medicine is difficult to integrate, if the use of this new structure, each X-ray does not need a database, like Google to see the page, how to find the page? This flat structure is what we use. Direct-Body web interface directly after the Internet to obtain the data you need, the data center does not need to set up a system, you can obtain different customized in the data center of these videos.

This is where I share some of our current technology implementations and successful cases in the Big data field.

Cloud base This idea according to our leader field total compares, is to use the price of the book to obtain data center computing ability, this is it his mission. His last mission was to get everyone on the internet, and he helped to make the IPO a success, and now the third industry is hoping to bring the dozens of companies with the cloud base to help them acquire the capabilities of these mega-data centers in North America at a cheaper cost.

This is our base situation, thank you!

(Responsible editor: Duqing first)