 I'm going to do a rather challenging thing, I'm trying to figure out the truth at the state of cloud computing in China. The cloud itself is often confused by many people, but the China cloud is usually more confusing to people because China has not all four unwritten rules, so it's very hard to figure out some things through just reading papers and reports. Cloud computing either is business development and technology is still in a very early stage in China, so as far as I know, there are not a real good cloud product in China now, so as a new cloud entrepreneur, I must be very aware of the state of cloud computing in China, so here today I will share what I have learned from my ventures in last year. This talk will cover three sections, the first is I will talk about some significant cloud, China cloud players, and the second, I will talk about overstake development in China. Now, because of overstake's dominant capabilities, in China we can regard that overstake equals the cloud. In order to understand overstake in China, you must know one company, that is United Stack, that's because you will, in my last pages, I will have some insights about overstake and my company's development. Before I talk about China cloud players, I will share an idea which I raised in my topic abstract, that is the comparison of enterprise IT versus consumer internet. In consumer internet market, almost all the domestic companies like Tencent, Baidu has dominated the whole internet industry, and the whole netizens in China, were leaving very little space to the foreign counterparts, and we all know that this attributes most to China's great firewall, but here I don't want to talk about political topics. At the same time, we can see in Chinese enterprise IT market, the foreign enterprise giants have very stable and strong positions in Chinese market, and few local companies have the ability to compete directly with this traditional IT giant. It will be very interesting to dive into this phenomenon, so I think for a long time about the background reason of this, here I picked one reason. We can see, here I have a diagram to show the investment overview in 2012. We can see only 2 to 5% of the investment goes to enterprise IT startup in 2012. As you know, the investment is very important, the incentive to the innovation of the IT industry, because enterprise IT or user needs a very long-term development and needs much patience, so here the venture capitalist investment plays a very critical role to development of enterprise software. So 2% investment of enterprise IT, and we can see 60 to 70% investment to consumer internet. So this is why so many new internet companies in China rise almost every day. In cloud area, we can see very similar status for the public cloud service hosting business. Almost these 4 biggest players, their first companies are almost the internet company. So that's because public cloud and service hosting business are more like consumer internet, while the public cloud business are more like traditional IT business. So the traditional IT companies almost own the whole private cloud business in China. I will talk about this public cloud hosting companies. Baidu Cloud, Baidu is a web giant in China and it provides a large range of services including Baidu Cloud. Baidu Cloud is not a very typical cloud. It's under the Baidu open platform, which is beautiful, which is beautiful core business like a search engine. Its major product is Baidu App Engine, is a platform as a service, while Baidu also provides a very simple infrastructure service like virtual machine management, but it's only open to its partner, not public available. AniCloud looks like a relatively successful cloud player in China. AniCloud once claimed that this platform has 1,000 developers for about three years. And Baidu Cloud, AniCloud is an independent business unit for Anibaba. We now own the big name and huge investment from the Anibaba Group. AniCloud gets a very huge customer base from Taobao e-commerce platform and also gets enough progress in the last two years. Now AniCloud should be the number one public cloud provider in China. As for the Tencent Cloud, Tencent now sees the number one seat for Chinese Internet. It touches almost every domain of the web services. From desktop to mobile, Tencent has hundreds of products online, one of them is the public cloud service. But there is a pity that Tencent Cloud is not publicly available. You can not experience the WebGiants Cloud unless you are the partner who runs your own business on Tencent social platform. As for Sina, Sina is my former employer and I just work there for Tencent Cloud. Sina has the very most popular cloud product that is Sina Avengy. It's the first most popular platform service in China. It supports multiple web language and framework like Python, PHP. And now this platform has begun about 2009. After several years development, it has begun the most important and biggest platform service in China. So I do a quick summary for these companies. Here I have some observations and personal comments. The first cloud hosting business is not a very serious business for Internet giants. That's because the revenue generated from the cloud business could be ignored by the company. They are major business. So they do not play very seriously on cloud. So the second, because this company has a very long history in Internet, they will face the cloud problem many years ago. But then there are no open stacks. There are no very good open source solutions. So most of these companies have developed their own cloud platform from scratch. The third observation is that many of my friends often ask me which public cloud should we choose. But I often say that I have no answer because there is no open cloud neither in China. Even any cloud is the number one cloud provider that intends for their size, but not intends for its quality. And I didn't mention before is that there are dozens of other small public hosting providers now most of them are struggling to survive. And some of them already died. This is a reality for public cloud in China. Even though we have such a situation, but I am also very optimistic about public cloud in China. That's because OpenStack, I think, give us a very huge opportunity in this area because it enables us to build a public cloud in a short time using the material technologies. For the public cloud, I just need these companies because most of you have already know their product portfolios or services. And for some sources, as I know, for some sources, IBM and HP, I think, is a big winner of Chinese private cloud market. The reason why IBM and HP are the most successful in the private cloud business, maybe because they have very strong, there were many years of relationship with the government. So in the last several years, from 2009 to now, Chinese government has invested a very huge money on the cloud development. So you can see many cities have built their own so-called cloud cities or cloud science parks. So almost all of these government cloud projects are built by the IBM and HP technologies and sometimes we can see where we are. So about OpenStack in China, I'm more willing to share about OpenStack in China because the market seems very clear for me. How many people have heard my topics yesterday? How many people have heard my topics? Oh, that's good because the following content is much similar to my yesterday's topic. So first, I want to talk about the OpenStack community in China. I spend some time to think about what is community. Because of OpenStack, we usually get together to share ideas, to communicate with each other, even to meet some business opportunities. I think this is the community. And it was mentioned that OpenStack community have brought me a lot of changes in my last two years in the beginning. Two years ago, when I first joined the OpenStack community, it was just changed my way to need an again, which we came from our in-house development to join the global community. And work with the global engineers and developers to work with them together to develop one platform. So this is a very great experience. And the native OpenStack have changed my career and even my life because I started a new company on OpenStack. So community is very great. I was often asked how big is China OpenStack community? Here, I have some numbers. According to the foundations, that is about the division number we have for 2,200 foundation division members from China, accounting for 20% of the global division members registered. And we have 1,400 members in China. OpenStack Google mining is not an easy number because, as you know, Google group service is totally blocked in China. So many stackers using VPN to access Google group service. And this is, 2,200 group is very active, about 2,500 people. There are about 100 messages per day. And OpenStack Vivo is the most important channel for us to deliver the foundation's message and the global OpenStack activities to many China's users. As for the, if I compare it to the world, China is number two hot area in these countries. Before, it's very, it's worth mentioning that before we can see few Chinese companies are willing to join the global collaboration because many companies, many Chinese companies don't have much sense on open source and the collaboration. But now the things have changed. In the Google, in the foundation's many list, we can see more and more messages from Chinese user group and Chinese companies. That's a very good sign. This is the most exciting page in my horror presentations. We can see almost all the Chinese top internet companies have already put an OpenStack in production or testing use. About two months ago, I began to visit these companies to pitch our OpenStack product and services. And to my surprise, I found that almost all these companies are running OpenStack well and they, that because internet companies in China have the best infrastructure engineers to do OpenStack development, to do troubleshooting, to do OpenStack operation, and to solve problems by themselves. And fortunately, after the third round of communication, some of them have become our early customers. Before I started a new business of OpenStack, I can only work for one company. But now I have a chance to work almost all these internet companies. So my team have given understanding of how this company is using OpenStack and how OpenStack influences their infrastructure. Here I have several observations. The first, OpenStack is replacing their legacy infrastructure software stack. Because these companies have very strong R&D ability to be developed by themselves. So I just mentioned that before OpenStack, they don't have very good infrastructure software to use. Neither commercial software nor the open source solutions. That's because before these internet companies, few companies have faced such massive workload and users. So this company has no choice to develop by themselves. So after many years of development and accumulation, they have communicated tons of infrastructure software, such as the machine management platform, cloud storage platform, security networking, and tools. But when they began to use OpenStack, they found that OpenStack can replace almost all their legacy infrastructure software. For example, NOVA can replace their machine management and SWEAV center can replace their server development storage system. And as for networking security, NUTRA can handle this problem. CHO to solve the DBA issues, Kistone solve, to replace the single sign on system. And Savannah can, even to management of the Hadoop system, and Aronic can management of the physical resource provision system. So OpenStack from only three projects to about 10 corporate projects and 100 community projects surrounding it. So OpenStack now has become very mature infrastructure software stack. And the internet companies has become the early adopters to this. So I think this is the future. And the second observation is that OpenStack restructures are infrastructure king. Before OpenStack, these companies have separated different teams to do different things. This is a security networking operation given OpenStack. They have different tool chance, different code base, different coding style, and different standard. But now they're working on one team based on OpenStack, same code base, same workflow. And they can also work together very well. The third observation is OpenStack actually has a fabricated that connects almost all the vertical component, or vertical infrastructure software that have visual database networking storage. So in my experience, in the last 30 years, IT development, this is the first that one single platform that can connect almost all the other open source and commercial platform. So this is why in order for some storage company, networking company, they embrace OpenStack. Next, I want to talk about OpenStack users in China. I gave an OpenStack user survey two weeks ago. And to my surprise, in the first week, I got more than 400 responses from 50 companies. These companies are from ID vendors and internet companies as we serve as provider education. We can see here internet companies are very active and the most active user group in OpenStack. And regarding the problem, what's your current stage of OpenStack company? This result made me a little surprised. We can find that 30% of people have tried to give an OpenStack but failed. And 30% of people have tried OpenStack, but not yet started to use OpenStack. And only 3% of users have successfully deployed for production use. And most of them are internet companies. This is a very interesting result. So as an OpenStack company, if I can make my customer to easily deploy OpenStack and easily operate OpenStack, we can be very successful in OpenStack market. So in order to understand the background of why they are failed and why they do not start, I designed the next question. Was hearing the development of OpenStack in your company? You can see the top three reasons. The first one is snacking OpenStack engineers accounting for 50% and very hard to deploy OpenStack very difficult to maintain and operate OpenStack. So this is very much our expectations. So before this user survey, my company, United State, have development of OpenStack distribution that exactly solved the top three reasons. And for rich OpenStack engineers, most people are using Grizzly. And Havana comes second place. The results shows that most people are very active to try new versions. As for which OpenStack project are you using, the traditional core, so even the neutral and standard most of this core project are accounting for more than half of our users. And as for which workload and application are running OpenStack, this is a very important result for the OpenStack company who are building applications of OpenStack. So you can see what the people need. So the first one, the first three applications are web service and storage and the virtual desktop. So next, I want to talk about the OpenStack vendors in China. In order to have better understanding of the status of Chinese players, I need to use the US OpenStack players. This size comes from many boughs from skating. They have a very good category that separates the OpenStack companies to this category, hardware vendors, component vendors, and the system. So this is the US players. So how about the Chinese status? We can see that the two kinds are the same, but examples are purely domestic companies. We can see the most crowded area is the hardware vendor. This company is most of them a security company or networking company that want to integrate their hardware to OpenStack. We have no component vendor because very component often requires very strong technical abilities. And we have no code distributions. We have no service companies. We have no significant OpenStack public cloud based on OpenStack. We have no ISV. Not only we have a nod off for private cloud environment. So this is why I say in order to understand China, OpenStack in China, you must understand the United States because we are the only real OpenStack players in China. So we often can see that we are an open source company that because we have a very strong commitment on the upstream. And according to the committees, we are the number two committed to the OpenStack core, and including the community projects. And we have core developers in OpenStack. And as far as we know, we are the biggest contributor from APAC. And besides OpenStack, we have a very diverse footprint on all these infrastructure sort of where. Because if you want to run the infrastructure well, you must have a better understanding of this vertical component. OpenStack is just a framework. So this component plays a very critical role in the infrastructure. For example, KVM hosting your virtual machine serve in our distribution hosting your data. And in us, management many lower function components. So if we are a purely open source company, my virtual capital investors, that because in China, open source cannot make money. So how do we make a business on OpenStack is a really, really hard question, especially in Chinese market. In order to give some quantitative image of Chinese business to OpenStack business, I insert this question to my OpenStack users away, which this model do you prefer for United State charges? We can see here 60% people want us to provide a free product with paid service like a red hat. And 30% want us to provide free product and business paid product like a cloud error. So about 90% want us to provide a free product. And as for the paid service, for foreigners here, maybe fear very optimistic that so many Chinese companies are willing to pay the service fee. But the reality is that for you companies are willing to pay service fees to open source companies. So this is why in last 10 years of open source history, we can see for you, we can see no one Chinese open source company survived on the Chinese market. So that's because even though he responds that he wants to pay the service fee, but the reality, they don't want to. They just want to free product. So the reality, so the result is that we can, neither we can make money from the paid service with free product, nor we can make money from paid product like a VMware, because people don't like it. Only 2% are willing us to be another VMware. Even though I said I can provide a very mature, very stable product, they do not buy it. So we cannot make money from service. We cannot make money from product. But how do we do that? So about several months ago, I raised this solution. I want to provide free software. And I want to make our service free. Oh, no. As for the service free, my explanation is that it's not that our service is free of charge. That is I want to make our free product as easy to use and easy to set up, easy to operate without buying any services from us. And we also provide a very human UI and easy management portal. So as a free software, it doesn't mean the software is free. It means the cost is really low, and there are no barriers to using this product. So another important side is that we want to deliver our product directly to our customers without broker or without SI. Because if our product is delivered by this company, the fees will be very high. And the user experience and the user were not very happy with our product. So we don't want to our product deliver to our broker. So we were provided our free oversight distribution available in our website. Now we have already released our first version, but now it's in internal testing. And now we will release our product in two months later, and then it will be freely downloaded. So in order to implement this strategy, we designed United States. This is a revolutionary overstag distribution for private cloud. We just did a demo in the demo theater, and I think some people have seen the whole information process. Before us, there are many other overstag distributions, like piston, merengue, and red hat. And the big difference is that it's a difference of business model. These companies, red hat, red space, they have opposite distribution, but their business model is essentially the service. So they can't make it too much too easy, because if you can do everything by yourself, how do we charge fees from our customers? But we do not want my users to buy our service soon. So I think this is a big difference. This is all you need to build such a private cloud when you disk. This may remind you that piston has also provided a UDISC installation. But as far as my whole team do not manage to set up for piston in our environment, because here it's a very strong requirement on the hardware, and it even requires you have very specific switches. So we only want disk. We do not have any requirement on your hardware. I'm not going to give you hundreds of pages of documentations to guide you how to install it. One is nice. And there are three steps, 15 minutes. The first is just to design VNN with CIDR network address. And then insert your USB disk to the first node. And then after 15 minutes, you will have the first node ready. And then you will have a human web portal. And then in the web portal, you can add the other physical server. Just the power on the other servers, the other servers will automatically integrate it to the previous server. And all the configuration has been done in the background without any in the fields. And we can add in more servers. So this is our deployment permutation. It's really easy. OK, so much about the families. Do you guys have any questions? You can ask in Chinese, so I can answer in Chinese. This is very interesting. I know the question. We ask whether I have already made money from Olmstack. Actually, the only money we made is in October 20, we hold a press release. And we're sending the ticket to the users. And as far as I know, this is my first revenue into a certain press release ticket. You know, in Chinese enterprise history, we are the only one enterprise company who are selling ticket for their product release press. So this is my first revenue. So this is a joke. So, frankly, we do not actually make money from Olmstack. But we have some internet companies. Now, we, part of them, even though this company is an instant, an instant internet company, most of them has never, never bought open source services before. But not many, some of them have bought our Olmstack service. And we made a little money from them. Yeah, service free and free software. Yeah. OK, I know you will ask this question. So I will explain the background of our strategies. Many friends may know that my founder team, most of our members, come from the internet companies. For internet companies, we want to make our product as scalable as possible. So we do not want to make money from case by case. We want our product to be used by as many as users. So my assumption is that if 1,000 users have successfully deployed our product, I believe that at least 1% will pay for the premium services. And this company will be very big companies. And this is one assumption. The second is we want to keep our customers happy and give the best reputations, and give as much as the installation base, and make all of them happy. So we have very good reputation. So this is the second. So the third, why we do not want to make money from almost every in this early stage? That's because in our opinion, infrastructure software usually needs a very long time to implement and very huge customer bases and very sony testing processes. So we made this product free in about one or two years. And after my product has been approved by many users, we can release our rear enterprise version. So this is, in the end, we will change our pure free model to a freemium model, free versions and the enterprise version. So this is a very typical model for over source business. So OK, this is our explanation of our strategy. We have enough time, OK? Yeah, OK. From my, I just, I didn't explain why. I just mentioned that I'm still optimistic about the future of public cloud in China. But I don't explain why, but I answer your question. So the reason why the previous development of public cloud development is not so good in my opinion is that the product and the product is not mature. Public cloud product is not mature. And the technology is not ready before. But in this stage, especially using overstack, you can have a very mature product in a very few years, maybe one or two years. I guess in one or two years, the overstack product will be very mature. In the second, the overstack will be very stable. So overstack can solve the two main reasons why public cloud didn't run well before. So this is why I'm still very optimistic about overstack public cloud in the future. Yeah. Yeah. Yes. I have seen a lot of universities and academies have brought overstack to their internal IT or to make overstack as their research topic. And as for students, many students have learned cloud computing technology through overstack. So I think this education academy is very active in China. I don't know anything here. But the reality here is. Any questions? OK. Thank you all.