 Now the next person that's going to come up is going to be Yufeng Zhang. He's a system engineer with Kihu 360. Another giant in China, and I'd love to hear what he's doing with OpenStack. So let's welcome him up. Thank you. Hi everyone, I'm Yufeng and this is my engineer from Qihu 360 and it's pleasure to talk about OpenStack at Qihu. Qihu is a top three internet company as made by Yufeng based in China and we are the number one provider of internet and mobile security products and services in China. And we provide the number two browsers and the number two search engine in China. By now we are searching 450 million users. So here are some of our products and some of the products are our core security product that protect our users, computers and mobile devices against any kind of malware and do system optimization we need it. And we have also browsers and it's the number two browsers in China and they provide user security and optimized access internet at point and we have also mobile assistant. It's a comprehensive and Android based internet that enable users to install, manage their applications on their smartphone and all the other services are web services that fully satisfy users access to their internet. So all these products and services are supported by our cloud based technology. So cool, but what does it mean to us especially to our engineers and here are some interesting numbers. For each day there are 590 million clicks on the start-up page each day and there are 13 to 15 million files being identified on the cloud antivirus engine and 2 million pve on mobile assistant each day. So there are many numbers to be listed here and I think it is a great challenge for our engineers to build great infrastructure to fully satisfy all this need, all this heavy load and all these products to deliver great user experience. So what we need from our infrastructure? I think we need agility. We need agility to give quick response to our customers requirement and we need the stability to make sure our customer to get high quality services and we also need elasticity to make sure we have high resource management in the infrastructure. So over stack makes our life easier and it gives us a chance to build infrastructure with agility, scalability and elasticity and that's where we start our open stack journey. So here is where we are. We start our open stack journey since Q3 last year and all of our open stack fosoms branch on the production system in our company and we would back part patches from upstream when needed which would host a CentOS 5.4 and a Vistin as hypervisor and by now there are four sorting instances around 20 data centers and the instance serves a chance of product each of which have numerous users. By now there are over 40% of all instances managed by open stack in Chihuahua and I think the number would get to 60% in the end of the year I think. And as shown in this picture, Chihuahua's infrastructure have heavily relays on open stack and many important services and products are supported by open stack. For example, all replications are hosting within instance managed by open stack and we leverage Glance replica to replicate images across all data centers so that we can make sure a new web services could be online within hours across all channels and online gaming is another story. Open stack gave us a chance to manage resources and instance in a fine-grained manner and as a security company we have many instance windows instance to be managed and with OpenStack API we could manage instance with automation and we can do scale out instance or scale in on demand and there are many examples, use cases such as infrastructure tools Hadoop Glance, automation building and development testing and we are looking for the next user. And finally, Sanctuary's great documentation and great support from the community a really small team could handle the daily job for OpenStack data operation and troubleshooting for now we have only two full-time engineers and we could handle the work to build and manage OpenStack based private cloud and we get good and positive feedback from our users. So here's our cloud antivirus engine all the files are extracted from the support database and pushed into the task pool and the instance are separated into several groups to execute some type of dedicated jobs and based on the information gathered from the monitor we could, the scheduler could scale out instance on demand for each day there are 13 to 15 million files being identified in this OpenStack based cloud antivirus engine. So many of Chihu's software and are based built on open source software so that we can have control of our software and we can do customization when needed and with OpenStack we have control of our infrastructure and we don't have to care about a weather or host login and we can do customization when needed to satisfy our special requirements. So here are some customizations and we reflect in the network model from layer 3 to layer 2 so that we could get high performance and get cost reduction and we introduce a neural table based image that we could build instance or instance image more quickly and we also add some QS parameter into instance so that some important instance could get a bad base assurance and we also implement instances such as being on our customized dashboard so that some instance could be resolved on circumstance that are deleted by mistake and we also integrate authentication with Kistume so that everyone has a Chihu account that they can get into OpenStack and we also do a lot of improvements in the dashboard also we would encounter bugs or facial requests in our daily job and we are happy to work on it and with the help of community we have tons of patches being merged into upstream and all of our team members are contributors of OpenStack So where are we going? I think we are still on the early stage of an OpenStack deployment and a lot of work to be done to make the OpenStack-based private cloud more effective and more stable and we are looking forward to move Havana from some regions and we need to add storage support in our OpenStack cluster and finally, thank you OY in community your great work makes our life easier and make all these great things happen Thank you Thank you so much Alright, wow So one of the largest internet and security companies in China running thousands of instances and 20 data centers of OpenStack that is incredible and I think that when we hear from users who love the flexibility of OpenStack but they are also contributing back that is just amazing to me and we think about some of that data I shared earlier where we see contributors and users foundation members those are not silos these are active members of the community who are all three really and that's the best kind of users the ones that are ready to get in there and help us make OpenStack better for everybody