 Okay, thanks everyone. We're just going to go through quick introduction because you have three presenters today and And then I'll go through the presentation in three parts. So that's what though if you can introduce yourself I'm Kentaro Matsumoto from KDDI in KDDI. I'm doing cloud architect Hi, this is the Wesaka I'm coming from the NEC I'm a general manager of the Platon service division. Thank you Thank you. Hi everyone. My name is Pierre Mathis director of telco solutions globally at Red Hat and Basically my responsibility to charter Red Hat is to work with all the tier one operators globally and partners in the telco ecosystem The way I got into Red Hat was before the Red Hat I was it through a company called Inavance and then we got acquired by Red Hat and Inavance was at the time the 7th largest open-stack contributor And so we're doing a lot of really interesting advanced thing And we were doing this with about 50 engineers So it's pretty remarkable with the amount of people how we were able to transform In the way we were contributing code. We're really happy that we made the transition to Red Hat And I'll show you a few things of how these emanated into the platform But the first thing I'd like to announce is I mean, this is fresh off the press As you can see dated May 8 in Tokyo So we announced today that we successfully completed Our trials with NEC and KDDI on their next-gen platform for IOT and 5G So it's great to see how far open-stack has come along And now we're really looking to do bleeding-edge things on the open-stack platform So it was a very very successful launch Namely, you know if I go back to my days of open-stack version 6 I remember you know at Verizon and AT&T really struggling with the various components of open-stack I remember vividly Keystone kept crashing because it was flooded by the network I mean, and if you look at where we're at today with open-stack, it's been pretty remarkable And one of the things that I don't know if people understand is this was a frustration I had a while back was how do we have people Participate in the upstream and downstream in contribution to open-stack and you know a lot of times You know companies and telcos view us as kind of the brokers To help them to get the features and functions and open-stack But what we really want to do is to go ahead of this and have a collaboration model and a platform to do so And so you know a lot of times people don't understand the mechanism Of how the secret sauce gets made an open-stack right how do we take it from trunk package in an RDO And then package it into our red-out open-stack platform that takes a lot of time and you know One of the things that we work really really hard is to make sure that we take the bugs and the fixes and make sure We have stable code and that takes You know a good six six months to do so you know one of the ways to speed up that process is Through the next platform that we're going to be launching soon is around DCI so it's taking the concept of CI CD and Basically baking it in into a platform into open-stack And this is what we've really worked hard with the NEC team And I think it helped in the evolution and the deployment of open-stack at in the KTDI implementation and so this enables us this platform that we use with partners enable us to Essentially participate so when we see when we have bug fix we're able to put it into the platform Another thing that we're able to do is if you see through the repository We have this continuous integration repository and this is where we put all our tests, right? We put a rally tempus browbeat and we even put our certification from red at inside a test And so the benefit of this is that we use the notion of triple o under cloud over cloud And this is where we work with our OSP director to basically run those tests The beauty of this is that you used to take about a month or so to Certify open-stack we can now do it over a weekend And so this has been very very effective in deploying and managing and now what we can start doing is test Continuously right we don't have to wait for the next version if we do something from a VNF perspective We need to add we can then test how is this going to break where the things are going to break what version and so we have a Very efficient way of collaborating with our partners in making sure that we have a way to basically do the patches and fix upstream and then downstream as part of our open first stream first strategy So this has been a very very successful platform and we're going to see a lot of this So the plan for red hat is to bring this into the mainstream In version 12 and so you'll see a lot of that coming in the near future But this is one of the areas. I think was very beneficial for us to working and collaborating So this is what I had for the moment So I'll hand it over and then maybe we can go dig Through the NEC piece and then we're going to go through the KDDI Talk a bit about the architecture and then we'll open up for Q&A sessions. Okay Next turn so I will explain the NEC cloud system OSS building model. First of all reason why NEC to collaborate Adapter Red Hat OpenStack platform because we have a Over the 10 years relationship to support the Open Mission political system from the line-ups operating system support right now We are supporting a open-source open-source platform so This is a OSS based cloud infrastructure building solution we call the NEC cloud system and as you know the OSS is a OSS that means you can download and you can integrate by yourself But it's very difficult to optimize for users demand NEC have a lot of experiences to integrate and to manage by ourselves NEC launch the public cloud solution since 2014 and NEC announced to promote and accelerate OP NFV solution with a Red Hat since 2015 and then We are Announced a NEC cloud system 2016 such kind of way we are We have some knowledge how to Integrate and manage and operate to use open-source software That is not only open-source but also the XAVIX type of monitoring system and the bacula type of backup system so NEC integrates such kind of a proprietary Open-source software to realize cloud environment so NEC provide three type of services winds up building integration service and Second one is a support service NEC is a one of hardware vendor Definitely we can support our NEC server storage network, but Our hardware not so popular out of Japan Such case maybe user want to use as a third-party hardware Server storage network, but NEC cloud system or support such kind of much vendor platform to integrate open-source and then Do you know that if user want to launch the cloud environment by themself They don't have enough management knowledge NEC have such knowledge and NEC provider such managed service by NEC itself or NEC are offered to adopt our knowledge by full customer environment such kind of way Open source is an open source everyone can use that but if user want to use or this Open-source software environment by themselves NEC can help this Environment Next slide is a NEC cloud system and the NFV stack This is a very busy, but the base is Etsy NFV architecture Our right side is a mono area NEC network provider this mono capability and then our beam Red Hat OpenStack platform present this beam capability and NFV area Red Hat virtualization that is a KVM or Red Hat support it and SDS software defined storage area we adapt Red Hat to safe storage and Network area NEC is a one of Pratchina member of OpenDirect we adapt the OpenDirect SCN controller to that and Teleco area are Their key factor is a high performance or we provide a DPDK OBS DPDK and the SRO IOV type of pass-through network Capability such kind of way NFVI and the beam NEC cloud system cover this area and Top of Box which is OSS, VSS area NEC network provider this capability and NFV VNF NEC itself has a several VNF, but You don't want to use another other third-party VNF Definitely, we can support such or third-party VNF on board Next slide is a overview of NEC cloud or solution Left side is a NEC cloud years which is a Public cloud solution We have a two type of environment one is a HA which is based on the VMware and Second one is a standard which is based on the open stack and the right side NEC cloud system is a on-promise cloud solution We have a three type of On-promise cloud solution one is a build OSS building model Second one is a commercial products building model. This is a VMware base and the cloud platform suite is a Vertical solution to realize a cloud solution such kind of way NEC support Public cloud and on-promise Cloud and we can definitely support the hybrid cloud environment and then NEC support the SOR type of applications Environment and the SOE type of cloud native application environment by this NEC cloud solution and The OSS or part we collaborate with the red hat and we adapt Open stack and the safe and the open shift and the answerable open shift. So as you know the User want to realize that is not only ears, but also the parts and the sauce such kind of way We need to adapt the more Upper side solution To for customer demand such kind of way we can support Such kind of requirement Next slide. As you know the open stack Will release two times per year Version up is a very serious issue And the user don't want to stop the current system NEC or try to achieve rolling upgrade without Zero downtime So Several steps are there But we prepared some automation tool which is presented by Ansible Prevok This operation will be done by within the one hour Such kind of way We are prepared some Best practice and we contribute to this Process and the tool to the upstream So next slide is an ongoing project with KDDI After my presentation Matsumoto-san will explain in detail Right now we're collaborating to project With KDDI. Why is a distributed monitoring and analysis project? And the second one is a zone migration project Especially zone migration project. We are contributing open stack watcher project to realize a How can I say If you want to Maintain answer hardware some VM need to move to the another hardware that is not only VM image but also the data image Watcher have some capability to realize such kind of Requirement we are but before watcher Okata cycle watcher version not enough to realize such kind of capability NEC contribute to new Enhancement to realize such zone migrations capability and we hope to To adapt and Apply this our new technology Pike cycle I hope Yeah Excuse me. I'm Kentaro Matsumoto from KDDI So we had a big POC open stack POC with Red Heart and NEC and Today I will introduce about our achievement This is my agenda of my 10 minutes. So about KDDI and Overview of this POC and what we achieved in POC So at first about KDDI So, do you know KDDI? So could you raise your hand if you know KDDI? Okay, thank you very much. I'm happy so many people know my company So I'm in home. I feel home though. It is United States anyway we are second biggest telecom company in Japan and Our main business is Japanese domestic and mobile, but we are doing global business so we have over a hundred offices all over the world and We have many data centers We call it telehouse. So we have 45 data centers all over the world Okay, next topic Overview of this POC So at first as background we have hundreds of internal IT systems Separately on different hardware. So it is a typical silo type structure so we spend We spend a lot of cost for operation and maintenance So we need to reduce these costs so this is our concept of new infrastructure We will have huge hardware pool maybe it will be thousands of servers and We developed infrastructure process communicator and using this We can easily assign hardware to OpenStack environment or VMware ESXi environment or bare metal environment So many systems on huge one hardware pool We will reduce the cost of management and using this We developed OpenStack POC environment and we did various kinds of tests to check if our workloads can learning can run on OpenStack environment for example, we did IPv6 test and Maruchi site authentication test and so on But today I want to introduce these three things Our special achievement first is flexible hardware assignment and second is harder maintenance without service outage we call it zone migration and Distributed monitoring and analysis So first thing we achieved is flexible hardware assignment So Let me explain about use case So in this use case, please see the picture There is OpenStack environment and we are running application number one and number two and We want to run new application number five but We don't have enough Resources enough servers on OpenStack environment So what can we do at first? We add hardware from hardware pool to OpenStack IS environment by IPC for extension and Next Configure these hardware as OpenStack compute node or OpenStack controller node or SDS like safe in OpenStack environment, so we can Do by a setting or install operating system install operating operating systems we can install Middleware and configure various kind of things and ready to use And at the last we can run new application number five on new compute nodes This is a concept or Flexible hardware assignment and if we want to quit application number five We can easily remove hardware from OpenStack environment and return it to huge hardware pool Okay Second thing we achieved is hardware maintenance without service outage. We call it zone migration So as background We already have IS environment and we have many IS tenants owner of applications if we want to Do maintenance of hardware for example bias problem or KBM security problems We have to negotiate with our tenants application owners about Maintenance because we need sometimes we need Reboot the hardware and if These tenants are not cloud native applications It there is service outage So for us it is very hard to negotiate a lot of tenants about maintenance So we are thinking of this structure. So let me explain this use case so in this use case There is OpenStack environment and We are running application number one and application number two and we want to do hardware maintenance of OpenStack environment, so what we can do So at first add hardware or OpenStack IS environment for maintenance and Configure these hardware as compute nodes and next move Applications application means virtual machines or shinder volumes So these move these applications to new compute nodes or strategies Like this So after that the left side compute nodes original compute nodes There is no application no virtual machines or no Volumes so we can do maintenance very easily so So there is original zone and we add maintenance zone and we move Resources workloads from original zone to maintenance zone. So we call it zone migration So to realize this structure we use various kinds of OpenStack technology technologies like Nova live migration, Nova cold migration, shinder volume migration and to schedule these migration we are using OpenStack Watcher project as Wesaka-san explained and So we have a plan to contribute this structure to OpenStack upstream And last thing we achieved is distributed monitoring and analysis So usual monitoring system like Xavix there is monitoring servers and there are monitoring agents and Monitoring process is running mainly at monitoring servers but in this structure monitoring process is Running at monitoring agents each compute node and only result is sent to central DB So monitoring process is distributed to each computer node So we can realize Real-time monitoring and it is very scalable solution So we have dedicated a session about it on Wednesday At noon distributed monitoring and analysis for telecom requirements So if you have an interest, please come to this session So I explained about POC result So now Using this result and experience we got We continue to design and Develop our new IT infrastructure Thank you Any questions? Yes, I would just ask you to go to the microphone That would be great because we're recording. So want to record the questions as well So at this right, there's a diagram called DCI. Correct. What DCI stands for? Oh Distributed continuous integration Same for CI right continuous integration continuous development. So that's that's how we call Any other questions? That was an easy one Okay, great Fred Herman from red app I'm curious about the pool of hardware you have and the kind of connectivity you have set up in order to be Able automatically to allocate resources like that to open stack clusters and so on I mean, is it a special all these hardware already interconnected together? So basically, you know, so friend so Essentially if I understood correctly, so you want to know the interconnectivity Between the various nodes and are they are pre-configured or do you add the hardware and the network? Or you just add pools with an existing network infrastructure So how how is KDDI? Managing this Let me answer very roughly Of course The physically all servers are connected to physical network So we can dynamically set up a virtual network Maybe be run of excellent these kinds of things. So and we can Assign hardware and separate each tenant Answer your question. Okay. Thank you very much. Thank you You know one of the things we we didn't talk about you know, I think a lot of times and I've seen this Not only KDI, but a lot of other Partners and clients is you know the notion that we think that NFV and going down the NFV path is just a technology play and What we see what we tend to see is There's a lot of Transformation occurs from a process standpoint. How do we work? How do we work in agile teams, right? So that there's a big change of how we collaborate together So the process is one of them, but also You know the mindset and what we can do with this, right? And so one of the key things with this NFV platform is really how can we go faster to market with new services and new? platforms and it really changes the way how we drive business and You know for a lot of clients NFV is a very transformational project Both how we work inside the company and how we develop and deploy new services And that's something I think we kind of glance over a lot We think that it's just a technology play but actually it really changes the culture and the way we do things inside the company and also from a business standpoint of how we deliver services faster and And you know at first a lot of clients think that you know open stack is really a cosplay And actually it turns out that it's a revenue play Because it enables you to develop and deploy these services at a much faster pace and cadence and you know If you ask the telcos today, you know Who are they competing with and a lot of them will tell you the folks that they're competing with are really the fangs or the Bat depending on which side of the ocean you are the fangs or the Facebook Amazon Netflix and Google and then you know You got the the the Baidu 10 cents and Ali Baba over on the other side And so you have this pressure and you know and they use NFV functions to deploy these new services in a very very effective way And this is why I think these projects are very exciting to see them emerge is that we're gonna start seeing a lot of new Services coming from these platforms from KDDI the question for the KDDI Person, so this is gonna be a POC right then when this model or this technology gonna be deployed into the production system Okay, I'm not sure so fan it is but maybe soon Very lovely answer sorry He must be a KDDI customer Yeah, right, so I think That's it no more questions going once going twice. All right soul. All right. Thank you very much