 Good afternoon and good evening. I know it's a prime time in EU and You know, I hope you have gotten some lunch on the east coast and for those of you who are waking up and joining on the west coast welcome To the to the second session on day three It on if has been a such a great success and thank you all for the feedback You have been giving us on the quality of speakers and the content. So we're really excited to kick this session off I would say we're saving absolutely the best for this session some of the top-notch leaders in the industry are going to give us a view of What we should all be thinking about in 2021 and beyond and so let me kick it off by Introducing our first speaker today You know, obviously he works at Ericsson, which needs no introduction a global leader there Anders Rosengren is the head of architecture and technology at Ericsson and Let's see what his views are on what we should all be looking at. So please welcome Anders Thank you very much Arpit, I think always great to be here. I love these conferences I and the combination of technology and open source and the industry Ecosystem, it's just such a great place to be so I'm here today to give some view on on how Ericsson looks at the future of 2021 as you say and going forward and we see the Year of 2021 as the year of 5g but also how we see edge orchestration and exposure that is key to develop and deliver on key Enterprise use cases beyond mobile broadband So let me start by talking about some of the use cases and From an opportunities as an telco operator that exists beyond mobile broadband So going towards B2B and B2B to X and of course private networks comes quite naturally as a As an opportunity for operators and then of course you also look at Combining that with edge compute to bring on not only the private network connectivity, but also the capability of hosting applications and combining that with connectivity and then of course Industrial IoT as you see in the middle here is a great opportunity to combine those capabilities of edge and private 5g networks into Addressing low latency and ultra lie ultra high reliability Then of course big data Another of these capabilities that we will see much more of building on top of connectivity Bringing the data back from the sensors from the IOT systems Utilizing that to optimize your processes and then of course security. I think we all know about increased Threat vectors, but also the increased requirements on distributed working and The necessary functionality to be built on top of that. So we see a lot of Opportunities for operators in this space building on combining connectivity with additional applications with additional features To deliver really great service to the enterprise and 2021 is the year to start the journey on this From Ericsson, we definitely see 5g as an transformative event As we move forward here, not just for the speeds and feeds of course, but really as a connected innovation platform for Applications and users and of course it starts with key technologies that we built into the system to really live up to the promises of 5g Around massive my MO around AI around cloud native Around intent and model driven Around energy efficiency all of the capabilities and the technologies that we use to build the function and characteristics that make up that platform And of course then you come to it should be highly efficient Of course multi-purpose to run different type of connectivity and applications Mobile radio access to cut the cord then of course zero touch management and orchestration that is close at at the heart of the LF networking foundation and own up, of course, so that's that's something that we really are building on Cloud native we have CNC FS the sponsor here I think that's fantastic and we see how CNC F is really creating a fantastic ecosystem around the cloud native Which we are using to build a cloud native 5g core and delivering that and with all the promises and capabilities that cloud native bring On top of that, of course, we see then that this is creating a set of values for the users So predictable connectivity Distributed processing we see Embedded intelligence and well as well as continue to build on that security as a high on the on the agenda and privacy and of course cost efficient and With this we enable diff the operators to address different type of users not only the consumer But definitely the industries larger group of people's and and also Government and the first responder networks, for example as different users and combining enabling that on top of This mobile network on top of this network platform So what kind of use cases do we see there and what does this I mean if we dive a little bit deeper into these use cases and and how the Network platform is then supporting that so if we start with warehouses and factories I think Manufacturing and digitalization of the manufacturing space has been very high on the agenda down to the right here You see smart manufacturing Eric's and as the first automated factory in Texas coming came a line end of last year and We see there the requirements are quite stringent of course. It's for on-prem edge We see that that is a necessity there combining with connectivity a lot of access Points in a lot of sim devices or different users conveyor belts for example a lot of robots and and not just the manufacturing robots, but of course also Automated gate guided vehicles for example Transporting stuff between them all of them require low latency and high reliability to to secure the service video surveillance and different applications operations technology applications running on top of that edge connecting and interacting with the The connectivity through a set of exposed services brings the key benefits around connectivity and applications, of course centralized management security data privacy and ultra low latency another one and rather different Use case is then that you actually do what to bring the technology out of the business So you want a very smooth and quick pop-up store established You don't want a lot of on-prem IT equipment and connectivity equipment So you want a simple connectivity and you want the operator to host that connectivity for you So having Then still low latency is important High bandwidth is definitely important as well security And to reduce the on-prem IT cost, etc So I think these are the two examples of use cases two examples of benefits that the combination of connectivity and applications edge That brings that to the enterprises and that your operators can really benefit from using and implementing If we move on then From this and talk a little bit more about the network as a platform, I think that will also be A word of 2021 and going forward where we see how we put things like network slicing in place Or we put things like edge Combining with connectivity and I think we we definitely and I think in the telco industry sometimes get a little bit too nerdy We are super fascinated about 3d pp and different APIs and and complex ways, but we also know that the application developers on top of this They are no they are not 3d pp experts They want a simple API that they can easily understand and utilize to connect to the network and build their To utilize the capabilities for example latency resiliency Reliability security other things that they can get out of the network and embed as features and Support for their application that they are building so they want simple APIs They want those APIs exposed into the development environment, but they also want the APIs for partner onboarding Consumer onboarding and the operator needs of course to provide Capabilities towards the users the enterprise for self-service portals, etc Below that API for external integration There is a set of automation and orchestration machinery to orchestrate the workloads In the network in the mobile network as well as the third-party applications to bring together a set of end-to-end services And below that of course the network components as well as the cloud platforms on edge and central Provided from a hybrid and multi-cloud perspective So here we definitely see the need of abstracting and presenting Reliable and simple and easy to use set of APIs as well as the capability of course to Orchestrate and expose the capabilities and the features and the network functions at the place that they should be Whether that's on-prem with the enterprise on the network edge or central to deliver certain types of KPIs and Maintain SLAs for the enterprise But it's of course then an open-source conference and I think I'm as I said in the beginning here Super excited to be here and we see definitely of course this ecosystem being built Through a set of standards whether it's 3dpp for the network functions or team forum for the APIs that are needed to interact and create a Multi-user and multi-vendor environment an easy integration environment We also see the industry alliances and that comes back to comes back to the need from a telco industry to listen to the Enterprises and the application builders that are going to use that that's so important to listen for example We work in 5G ACIA together with a set of manufacturing industry specialists. They give their requirements on the API This is the functionality. We want to access. This is the way we want to access it And this is what we need to know and this is what you need to abstract away So that is super important and then of course we see open source Helping us building the foundation building this in an open and collaborative way And of course Linus foundation edge and networking are key parts here as I said cloud native compute foundation as well it's also and of course a very integral and important part of this whole solution o-ran and own up are bringing different types of Capabilities and of course Ericsson sees the value of this sees the value of engaging and using open source Together with our our proprietary logic to create the the full end-to-end solution So finally then I when I spoke inside the Ericsson and Collaborated with a set of the people that are engaged in open source. I asked them Okay, what what kind of message what kind of call for actions? Do you see as important here? How do we make open source lead the way even more here to build the necessary foundation for for the operators and for the telco equipment vendors to Deliver on the requirements of enterprises to really make 5g the transformation on event and and We came with three three topics here I think we see with cloud native and definitely a need to move much faster and quicker towards the Cloud native way of working So integrating orchestration and management of the resources Together with the software lifecycle management and SUDP and here we see own up and CDF foundation That there is we we really want to continue to drive that collaboration and that integration to really maintain that and improve that collaboration and Bring additional features here to the industry Then I think we are working in a multivendor reality and we see kubernetes being our foundational element, but we also see that there is no Aligned view today. So Creating a common kubernetes profile from a CNCF to really drive simplification and efficiency In the ecosystem and industry when you're orchestrating different workloads across different Kubernetes alternatives that that will be an important way for open source to support the industry and then I think These collaboration not only inside a specific open source project, but between them and I think we have an example here of of a D map and secure an alignment there with kafkan and minio Instead of maintaining bespoke functionality inside on up. So really urging the The open source community to start even better. I know you are working on that already and we are working on that but really better even better Collaborate across to build that full end-to-end system in the best and the most efficient way So with that I I close my Talk here. Thank you very much for listening to me All right. Thank you Anders. This is very insightful. I think you're Your end-to-end view of pulling in, you know core edge access You know enterprise telcos cloud kind of bringing it all has really really You know resonated. I like the call to action also There there are a few questions that have come in. I think the first one I would ask is There's a project anuket where there's re2 that defines common common kubernetes profiles for telcos would you see You know cncf helping with that directly And and getting there or do you see something else on on that? In in reference to your kubernetes Yeah, yeah, so so that very very good and I think here we are engaged in in anuket And the the two different products that it came out of so we definitely see that as as a good A good base but also of course inside cncf and we have the the cnf work group there So definitely joining forces between anuket and cncf here would be a real really important way forward awesome very good, and again, I think I don't have to mention this but uh, you know you You and ericsson in general have really helped with the with the collaboration of standards alliances and open source as seen on your on your slide The the question is really around Do you see any missing pieces in terms of other standards organization? Or other alliances that we should be thinking about as a community when you look at vertical specifically I think to further work. I think one one of the challenges with The operator industry and or the telco industry is the compared to for example the development Developer environment on on hyperscale is is the federation or alignment of interfaces there I see a team forum and meh for example us as key Areas to facilitate that east west and also facilitate the multivendor Or integration around orchestration and business management Logic so those are two areas where we are putting a lot of effort into Very good. All right. Very good on this. Thank you very much for for taking the time really appreciate your your insights here Thank you very much All right, and I think there was another question that was directed probably at me from a linux Foundation perspective, which was you know one of the key objectives of lf is to maintain an ecosystem of multiple contributors in key markets Uh, you know, there's like three manufacturers for radio. You know, how do you see them? You know in each of the market you you see Uh, what what would happen if there's only two one? And I think I would probably be the first one to say That uh, first of all, you know, it's not three. It's more than five Uh, and more importantly The disaggregation of the entire architecture allows for a lot more startups and innovators to join And then change the thinking and provide the pieces of the puzzle. In fact, uh, to so so much as, uh, you know In which conference would you have? You know back-to-back presentation from what I would say, you know, which traditionally the the World would call competitors, but they are collaborators in our open world so with that said our next speaker is from nokia and You know, he is one of the most active executives in Linux foundation From an open source perspective, you know board strategy, etc He's the head of open source initiatives at nokia. Uh, please welcome yona sonanin Hello, so I'm um I'm uh, yona sonanin and I work for nokia. Uh, as harbors says, I'm the head of open source initiatives here And I will be talking to you a little bit about my approach and our approach to open source and the uh, the edge And then that opens our staff and the edge, but first of all, I have to say this is a great, um Event and it's been really interesting to see the great presentations And I have to say I'm very humbled to be part of this. Um, great lineup where Uh, where there has been so many good speakers and with so many interesting areas Now let's see if this works it does great. Um, so what does nokia do? Um, so what nokia does is it builds networks? We do we are a network vendor. We do network infrastructure We're one of the few vendors who have um, who are basically in all of the spaces of of networking And what we like to say that what we build is we build critical networks and technologies to bring together the world's intelligence and what that means is that critical networks are the networks, um That are critical for this world. So it's not just a special kind of networks But these are the networks that are critical for the world. So taking into account our service providers uh industry automation digitalization and all this that enables basically us to move on as an digital world And what is needed in the digital, uh in the digital transformation The where we see is that this is an intersection of the gary grade Resilience and performance and the web scale as the elasticity and flexibility And that is what we are aiming at and as, um We have been mentioned already during this week. We work with the um with the leading vendors like andre fudge from at&t mentioned That we're working together, but we're also then working together to provide our services on on our solutions on top of web scalers. So we are working like george from From google mentioned a little bit before in this in this forum as well What do we base this on so what we base this on plan in technology evolution that we think that matter that a technology evolution that we think is important That first of all, it's clear. It's cloud native software and the transformation to cloud native that is a key transition in our industry today We're proud to be leading that in our industry. This is something that we are we find extremely important And this is something that is important not only for us, but also for our customers and for our compatibility with other platforms This cannot be done without openness without having the best performance and having the leading Basically having about having technologies and having architectures that are open that provide the best performance and the lowest tco for a multivendor environment um, there is no possibility of Of a close environment succeeding anymore in this world And and therefore the openness is a big part of what we do It's not just about however about software. It's also about silicon. It's also about hardware and making sure that we have the best silicon for our customers to use And this is of course when we look at these things Many of these have aspects that we have discussed in this form this week Many of these have aspects Have aspects that are relevant to open source The however, we are not anymore in normal times the world has changed last year in the poor When the kovid set in and this has basically um, been a tremendous Strain and that depends challenge of course for people individually But also for our customers and for the industry as a whole What we have there Throughout the pandemic we have focused on making sure that our customers Can live through and can provide the best possible service during these times and the Three phases that we have a basically focused on Is secure that we keep the networks up and running Scale that we adapt to these demands and stimulate that we Continue to evolve business models and use cases in this changing world It is actually interesting to see that what we saw we have We have published a report in our by our deep field department and the About what is what how did we see that the world changed in the networks? When the kovid set in it is actually quite interesting to see that within the first weeks the growth of traffic Was about the same so um 30 to 40 percent what you usually see in one year The good thing is that um Basically the networks that was what they were meant to do and the networks survived without any clutches and basically the um Every everything worked This is what we do. This is what we live for is basically making sure that these critical networks work regardless what happens and Though this has been a very difficult time for everybody of us And the testament of that is that we're now Here remotely and we're not in the same place everybody um The basically the telecom networks and the telecom paradigm starts something that do um help us in uh do actually Allow us to live through this and allow us to continue to work together What I do still do want to talk about a little bit is About openness. What does that mean for us and what does that mean for the industry in our mind? we like to look at when we look at openness and as Basically a question of four pillars. We have open forums open interface open culture and open ecosystems it is quite clear that the Open source is big part of this openness um But and I'm going to talk a little bit about that how that fits in in there a little bit more In a just a minute, but also the thing is basically We need also other things of openness for this to succeed One of the key things that I would like to emphasize is the open culture You cannot be open in terms of open source Or in terms of open standards if you don't have an open culture inside your company to start with that brings together then Using the build block building blocks that open source open standards and so on create Opening interfaces and creating open ecosystems, but the culture is where everything starts from And this is something that the company needs to have Internally nurtured and it is basically something that the company needs to It needs to be built in rather than being something to add on otherwise, you know the Otherwise we cannot move forward and we cannot actually provide the other populars in in the support pillar system now if we look at the If we look at then at the open forums That is the area where we see open source, but we also see open standards and open architectures um the basically standards and open source is something that what we see as The sides are two sides of the same coin Standards are are something that is are is needed To create interoperability and to ensure compatibility between different kind of implementations Open source on the other hand Fits the best into something where you one implementation Provide so to say the standard over the world and of course, there is interaction between these basically the Open source sometimes implement standards and standards some sometimes References to open source We see this and nowadays more and more in the activities that we do that There is an interaction between these The two sides of the same coin The thing is that often is I have been asked previously So do these things compete and will open source win over standardization Or will standardization Be something that is actually more important and is open source something that Will fade away over time um We have had standardization for a very long time. Those will not go away. That is for sure The they are still needed and they bring a basis for everything where there needs to be compatibility with Between different kinds of implementations That is augmented by open source nowadays in In the industry and there are basically places where open source is more direct tool and more More adapt tool to actually bring the bring the benefits And we can see this especially in the areas where There is little differentiation between the players who are actually doing the open source together That they have an interest to cooperate and they have an interest to come to a common solution Standards on the other hand allow also more space for proprietary competition as well But when we look at then open source, what are the areas where we are as Nokia interested and where we are participating in is basically the Are the When we look at the Linux foundation space um, it is I mean like you've seen these logos most probably And most definitely throughout the The presentations here. So if we look at that the If we look at the open source projects that are important for us in the LF space and in the Linux foundation space, it's clearly over an software community LF networking LF AI and LF edge and we're part of those Of course something that is not on this slide Is that what a lot is now built on which is cnc app and the cloud native movement the These are areas where we are very much involved where we have Where we are having Board positions where we have intact positions where we are having also where we are active in the In the projects and if we look at that the kind of like couple of projects that are important for us In those spaces is for instance in orange software community. We are active. We are driving the Near real-time rick implementation In LF networking What we are we are working especially in on up and annu cat And so on and so on there And in addition to this there are of course other open source organizations where an open source projects big and small that Where we are participating in It's clear that open source basically Defines big part of that what telecommunication networking is today now looking at the call for action the Thing is and I think this was mentioned by somebody already before It's important to work together This is about open source. This is about community. We have different communities That are But the thing important thing is to bring a big open source community where we work together There is little room for competing Competing Competing technologies it is better to make sure that we work together rather than trying to create new ones on top of the old ones So this is very important and it's this means also that we have to work in the communities That are actually responsible for those Technologies and we have to have those at least as upstream from projects that are That where where there is a relationship needed and where there is that where the Community that owns the technology Is the place where we come to together? Even if we are discussing the technology somewhere else as well The other thing is that for open source. It's quite important that we actually focus on the value What is where is that open source brings extra value? um Where that we work together that where we it is not a point solution from one company, but it's actually something where we have Contributed together and we have compatibility and Reusability over the industry is important and that is I think that it's Clear that we should that's where open source is at its best Doing open source just to copy Activities or copy basically something that is already there That doesn't bring as much value as bring building something new and building something that where the community really has power The least but not the last one the last one, but not the least one. Sorry that way around Is of course then we should take security more seriously The world has changed security is something that is now paramount for our customers for and The kind of like the threats are have actually evolved as well So in that sense, I think that we what we need Is to focus more on security that we have done before and this is a quite an important step that we need to do um in To make sure that open source Is considered trustworthy also in the future Now that was What I considered to tell you I um, I'm very thankful and very honored to be part of this lineup I hope you found this at least somewhat useful um, if you have any questions Um, I think that there might be some time now But then the question is also that you can Contact me. I'm sure that you can find me on LinkedIn and so on And as any questions on that um Oh, that was that was great. Thank you, Yona And uh, I really appreciate your your insights here So let me introduce the next speaker He's from facebook obviously, but uh, brian barrett who's been a a thought leader in Uh, a project called magma who had just recently moved to the Linux foundation And we are really excited to have brian Talk about the ecosystem and How magma is a critical piece of an end-to-end 5g stack for the telecom mobile core. So please welcome brian Great. Thank you arpit. Hi everyone. My name is brian barrett. I'm an engineering manager at facebook connectivity So at facebook connectivity Our our mission is to bring the world online to a better and faster internet So let's start there with a bit of an overview of how we're doing so far The state of global connectivity There's some good news Global access is improving According to the international telecommunications unions internet bandwidth index in 2020 Total connectivity of the average household is is increasing worldwide by about 45 percent And 4g connections are increasing So we now have 65 of lower and middle income countries Have 4g coverage But there are some problems still One of them is that network performance is at risk 3.3 billion people in developing and emerging markets are at risk of a degraded network performance by 2023 and facebook Cares about connectivity because it's good for the world and good for facebook bringing connect being connected Opens up new opportunities to make yourself heard to connect with people and communities You care about and to build new businesses So at facebook connectivity one of the things we do is we look across the entire connectivity ecosystem and the value chain Try to understand the reasons Why people have suboptimal connectivity and what we could do to improve it One of the things we noticed is that in many markets You have enough population density to really warrant a cellular network existing But they don't exist in a lot of markets or the ones that do exist aren't affordable by the average people Because in markets where the average revenue per user is very low Often the existing telecom models of the network core With the subscription models for licensing and the equipment costs Are just too high to make it affordable For a company to provide cellular coverage in that area And so one of the things that we set out to do at facebook connectivity Is to enable the open source telco To catalyze an open source 5g movement And to dramatically reduce the total cost of ownership of cellular networks So we developed magma magma was open sourced last year and it just recently launched in february as a new linux foundation based project So magma is different than other network cores Uh, it's hyperscalable and also distributed and so by distributed I think it's fairly common now to see network cores where people have containerized their epc and able to run it on On a rack of servers at a point of presence or run it in the cloud Magma goes a step further you can run magma like that if you want to But magma is also designed to run Without a rack of servers anywhere you can run magma on very low end Far edge appliances at the far edge of the network. So imagine a small ruggedized x86 server for a couple hundred dollars Deploying one at each of your cell sites Letting them break out to ip Right at the far edge And having them all connect back Uh to a control plane in the cloud or on-prem On an on-prem cloud It's open source And completely licensed free. So this is licensed with a bsd Three clause licensed completely open source software free for others to use free stuff free for others to white label and sell It's cloud native both for central cloud and we have work for even edge cloud native With control plane user plane separation It's vendor and transport agnostic And it's also designed shown in this diagram here with an optional component Where if you want to use magma in a lower cost market But still connect to A legacy core used on a larger network You can federate with that larger core and inherit from the hss and pcrf policies From an existing core and just use magma in lower cost regions Where it might be advantageous for you And it's designed for all access conversions. So it's not just 4g. It's 4g 5g private lte and wi-fi All in this open source solution So at facebook we see the evolution of open source and telco Following a similar model as enterprise So 15 years ago open source and enterprise started with You know red hat shipping linux on cds and retail stores and hobbyists experimenting with linux in niche use cases You saw linux users groups enterprise started adopting it and Eventually it gained broad adoption in the enterprise telco is where telco open source We see following similar steps Right now you see niche deployments by cable satellite operators and developing countries places with low arpu And we see that growing in the same way Where there's select mno's they're starting to pay attention and follow in the footsteps of georackadin dish Looking at bringing more of these open solutions and open source implementations to their networks Before we think it will then eventually gain as broad of adoption as open source has an enterprise now And we think it'll be different between core and ran We think open ran will follow the journey of open compute Where there will be many implementations And we think open core will follow the journey of kubernetes or open stack Where there will be many attempts to make an open source telco stack But there will be one emergent de facto implementation that will become the standard And we hope that magma is the basis of that standard So in 2021 we established the magma core foundation magma has been open source licensed since 2020 But we've started hearing a lot of interest from large organizations that would like to contribute meaningful teams of software engineers to the project and a lot of mature organizations Like to know there's a very familiar and trusted governance model for the project So in 2021 magma core foundation was established under the linux foundation So that we'd now operate under linux foundation governance Completely neutral So facebook continues to contribute a large team of software engineers to help anchor the ecosystem But the ecosystems really flourishing we founded magma foundation with a number of large partners We have participants from across the industry participating in magma foundation Either as founding members on the governing board or or just participants in the open source movement And we also have ecosystem partnerships with open infrastructure foundation Open air interface alliance And the telecom info project are all part of magma And we think because of this it's really emerging as the de facto open source network core So it's our belief that it will be the one open core implementation That is open source converged cloud native And multi-tenant by design from the ground up So here's how we're doing We're really seeing strong a strong community growth in magma So this graph shows our growth in Advocates which are people that are working on magma all the time contributors people that contribute to the code base and participants which is people that Use magma and converse in our slack channels and ask questions and engage and comment on design And that's really grown radically lately in july of last year. We announced publicly that we had a partnership with open air interface telecom info project and open infrastructure foundation And in in february we announced the linux foundation announcement and you can see the large growth in the magma community over that time If you add that to that observer, so these are people who have started our repos or downloaded magma or experimenting with it and using it It's even larger We have now over a thousand people participating in the magma ecosystem So one of the things we're really trying to build a healthy robust self-sustaining ecosystem around magma. We do that in a few ways One is that we partner with industry groups that are interested in helping scale the number of developers driving developer adoption We partner with vendors and operators to help drive adoption of magma including people that want to participate in trials and deploy 5g or 4g networks in new areas Learn how to use magma We partner with academia because one of the things we're observing is that Communities that want to experiment with the latest 5g release 16 release 17 features Open source is the place those things appear first because if you're a scientist or a researcher You want to experiment with bleeding edge features in the core of the ram The easiest place for you to do that is to start with open source and add the code you need for your new feature When we partner with startups, there's a lot of people that are now building distros on magma Our partners at freedom fire one example And so through these engagements, we help drive adoption of magma future use case experimentation and innovation in the ecosystem So I wanted to share a couple really exciting things that partners startups and ecosystem members are doing with magma These are just two of many highlights from our developers summit in in february One of them is is our partnership with amazon So one of the things that amazon did is leveraged magma's distributed architecture To make it one click deployable On their far edge appliances So everyone's probably familiar with amazon central regions and maybe even amazon outposts where you can put a rack of servers on prem for amazon cloud But they also have the snowball and snow cone Far edge appliances that are shown here on the picture at left and so Working together we actually demonstrated at the developers conference the ability to one click deploy magma Where the magma control plane deploys to an aws region and the magma user plane Gets deployed on a number of snow cone appliances that are just about this big that have eating shipping labels And you can ship them out anywhere in the world and plug them in and the zero touch provision connect back into the central cloud Another really exciting use case our partners have done with magma Is is found with a partnership between freedom fire and helium, which is a crypto blockchain uh Organization So unlike normal cryptocurrencies where you burn electricity to mine bitcoins With helium you mine helium coins by connecting the unconnected So uh helium has operated previously with laura wan where you connect laura wan hot spots at your house And they all join together to make one large iot network But with a partnership with freedom fire, they're bringing that use case to 4g and 5g So that you connect a magma based appliance Connect a 4g or 5g radio It calls in and gets cbrs spectrum from a spectrum allocation service Makes a small 5g hot spot and they all join together to make one larger network and mine helium crypto coins By for providing coverage and allowing users to obtain internet access Through that hot spot. So it's a real people's network for connecting the unconnected Which is a really interesting and exciting use case behind magma So we'd like to invite everyone on this call to join us and build your connectivity Software as a service enterprise or consumer business faster with magma and help ride this 5g wave with us There's a few ways you can get involved If you're an industry group, uh, we'd like to work with you to bring your developer scale And evangelize magma to your users in your community And just help support and drive adoption and magma use cases in your community If you're a vendor or a systems integrator magma is really designed to be brought to market Through systems integrators and value-added resellers that want to build their own districts So we invite you to partner with us and create development teams to focus on magma development and integrate It is a key part of your open source strategy To help you build with magma And use it as a production grade commercial ready lego block for you to build a larger solution that you want to take to market And lean on the magma community to help leapfrog your competition in a post-covid world If you're an operator or a service provider We think magma is pretty compelling because it lets you take control of your 5g vision and destiny And say goodbye to vendor lock-in Help you get full visibility in your production networks right from the code level Including shift left for dev sec ops and security vulnerabilities because you can see them right in the source space and patch them very quickly And if you're a startup or investor you can leverage magma stack to build faster and accelerate your product development And time to market you can lean on our expert community to leapfrog your competition And explore adjacencies And lastly if you're an academia or in the public domain You can leverage the magma stack for your academic research and r&d projects Unlock future use cases and collaborate for advancing open source telco And 5g to next gene movements. So we have some Really exciting opportunities. We're already seeing with 5g release 16 and 17 features with integrated access and backhaul And non-terrestrial network connectivity with public sector opportunities From DARPA and the european space agency and the us do d lining up behind this Thank you for your time. We'd like to invite you to engage after this Um today magma release 1.4 is releasing today You can find that on magma core.org You can join us on github pull the repo start building start developing and join us on slack at magma core.slack.com We'd love to help you get started. Thanks Thank you, brian. Uh, that was insightful Uh, and without any further delays I'm going to move to the Next speaker here Sean Zandy who is the head of engineering at linkedin You know another one of those A very sharp thought leader technically Very insightful and we're glad to have you here. So shan take it away Thank you very much. It's great to be here. It's a pleasure to present at this event I'm here to talk about our software approaches uh being Infrastructure that we're running in linkedin about linkedin. What is linkedin linkedin is an online platform that connects sports professional I've been at linkedin during the time of incredible growth. Our growth in member engagement has generated Enormous need for data storage for infrastructure and operation at scale that forced us to Change the way that we are architect and operate our network Currently we have nearly 740 members in 200 countries with more than 57 million companies Have a page on linkedin COVID has increased our member engagement to a new scale during the COVID last year from march to december We saw six times increase in number of remote job posting in united states To demonstrate the scale and importance of our infrastructure We serve we serve close to 40 trillion cough glasses per day And our economic graph has grown to 200 billion graphages this demand from Our linkedin platform requires us to constantly grow our infrastructure and scale Our data center back when an edge network to serve our platform In terms of vision our vision for network and infrastructure engineering is to use the best methods and practices to provide a reliable infrastructure to linkedin We believe that infrastructure Network in general is there to enable our business and application And the network is definitely so important but its importance is enabling business and application capabilities And the more that becomes transparent it more The more value it can bring for us to keep up with the scale demands of the requirements and operations first mindset We put our focus on programmability of our infrastructure to deploy New pods seamlessly and data center Provide self-reliant capabilities Reduce operational toil to scale our application Out closer to to our members and improve the performance We started our journey to desegregation 2015 and our intent was to speed up in innovation velocity bring the features that we needed most faster Into production and by owning those features control our network architecture and destiny As our network group we realize that we should pivot our focus and our investment on the software and ideally a uniform code across different hardware that we operate in our network Open switch project under linux foundation was very important to us running linux on switches and routers enabled us to To to use the same toolset that we use on our servers and our compute And extend the same philosophy to our network devices Now with sonic over a dozen different sqs including Variety of different OEM ODM and open 19 platforms are supported in a uniform fashion that we can run same Sonic software across multiple different hardware with different chipset By using switch abstraction interface or side we utilize the latest silicon from 12.8 to 25 terabit and It enabled us to uplift our data center from 100 to 400 gig across the board in order to scale our routing and control plane be on boarded Free-range routing to sonic community or frr under linux foundation that enabled us to scale our eGP network in data center and improve eCMP properties to have better control over routing convergence And management of how routes get propagated across the network We also designed our own Data center routing protocol based on linked state properties to improve Better utilization of forwarding paths in a widely connected parallel fabric Which is open sourced and is available on free range routing as fabric B as we call it in an open fabric We've begun to control and management plane. I believe one easy step toward programmabilities to move network to linux We have linux as the default experience from Of course compute but top of racks switches to spine and leaf along with our management network This enabled us to design our network just like a distributed system as opposed to conventional box by box operation The next step for us was instead of using different management protocols That were difficult to customize such as SNMP sys like sflow We came up with a standard way of pulling and pushing data out of the network elements using a pop-up pipeline in a distributed manner They developed net cm Which is our common platform for network configuration management based on open config and yank bundles And its job is not only to perform provisioning and configuration tasks, but also To Compare configs and notify us about configuration derives amongst thousands of network elements that we operate in our data centers telemetry data is sent from each element to distributed setup process as a collector's process and Those collectors can convert patterns and make meaningful and context out of that data that we can feed into time series database Count the frequencies of individual or related events or perform data science To improve our data aggregation and also our self healing Properties that they're going to talk about The the other aspect of managing network as a distributed system is that if you store network state at a point of time Then you can perform data analytics by going back to previous distorted state for root cause analysis Or restore the last known good state when needed it Opens up many possibilities and make things quite interesting for example you You can replay how your routing system reacts and follow update propagation Across the distributed system and figure out inconsistencies and perhaps black holes or micro loops or longer convergence time and by Understanding that data you can you can use it as feedback loop to reiterate and improve on those finding and make your infrastructure better Moving on to forwarding plane. We talked about Properties of control panel management plane and this is something that comes along with control plane, but it's mostly focused on forwarding plane I want to talk about a project that we spent Past couple of years and we're happy to announce that we're working with Linux Foundation to open source it to the community I believe the philosophy of network monitoring or fundamental value is to to quickly identify if network is The root cause of a problem whenever a GCN happens regards of symptoms network is legitimates suspect and If it's not a network issue then we should be able to rule ourselves out as quickly as possible and Data is the ultimate authority for us Argos provides that data to know if there's a fault In the network and where it is Argos is our forwarding plane telemetry fault detection and fault isolation system operating a large-scale data center There are many components links optics network elements can go down Meantime between failure is directly proportional to the number of components in the system Argos creates a network between participating and host probing each other and connected through Kafka pipeline It measures performance of every single connection in a network from host to host perspective and creates a matrix that consists of sla such as latency loss G there for different connections Argos helped us in not only detecting network issues, but also pointing out what nodes or link across the network Have that problem It also provides an input to our self-healing system to steer traffic away from the fault change the ECMP grouping For better hashing and provide the best delivery to our application so Open fabric is the evolution of how we architect and operate our infrastructure at Meantime Moved us from box by box to a completely distributed system based on open source that was realized by shifting our investment to to software and code The intent is not solving all of the problems by centralizing or distributing them but to have a programmable network stack that can Enable seamless service provisioning scale on demand And self-healing capabilities to make this work as transparent as possible And to enable business to serve our members as for call fraction the value of Any open source project is in the community and participation behind it At the LinkedIn our engineering team has open source close to 100 projects spanning different categories including data framework system operation testing we believe that Open sourcing project makes our engineers better at what they do They grow in their craft by having their work shared with the community time reviewed this project in broad adoption that are no part of Apache software foundation and Linux foundation as well There are many tools that internal Web scale companies use internally that can bring tremendous value to the industry And I would like to invite everyone to join us, you know passers Thank you. All right. Thank you Thanks a lot Sean And I would like to sort of first of all say, you know, it has been such a change And and for those of you who would like to ask questions again I think I missed the opportunity to say that in the in the previous session with Brian and Yona, but You know, please do a q&a on the top, right? But we really welcome, you know, what a change from, you know, say five years ago when it was a proprietary telecom type Industry now we have the cloud. We have the enterprise and users And you know leaders like yourself and Facebook and others right like from a web 2.0 or a web scale company And and the thing that we really like Or the community likes is the fact that these tools that That you have deployed and use and that you're contributing back to the community through open source They're at scale. They have been proven and You just can we can all utilize the benefit And and make the networks much better So there's a couple of other other questions that that have also come in if you have a, you know, I think we have a couple of minutes. Yep, sure um, you know, how do you see The the enterprise data centers the massively scalable data centers like the you know, as yours of the world, etc Which have very custom workloads, but they're very simple In nature versus sort of the, you know, LinkedIn which also have, you know, one or two five unique workloads versus You know a telco data center or an edge data like how do you see the evolution of these? workloads Getting spun up between different types of data centers Yeah, excellent question enterprise networks are Back back in the day they were designed around this three hierarchical system of access to solution and core and we used this to the same set of switches being used the intelligence was in the core and access was purely connecting nodes together The shift that we're seeing now as opposed to bring the traffic through the network to the core and then process it there or send it to service modules like load balancing and firewall At scale people are trying to make network as simple as possible. Just, you know, provide massive bandwidth between nodes We see service mesh. We see distributed the way of Deploying application So network becomes just bunch of boxes that provide massive bandwidth and then services can be easily moved to the edge to be processed by For load balancing for firewall and different reasons I can see that the enterprise is following the footsteps of large scale in terms of simplifying the core and then bring the intelligence to the edge Gotcha very nice. Well, thank you for your leadership and thank you for contributing back and really appreciate Your your insights here Sean. Thank you Pleasure to be here. Thank you all right So with that, I think we have We're saving obviously A very important member Of of the global ecosystem and and some of you know, some of you may Most of you may know him for his vision vision and leadership, but everybody knows Deutsche telecom And you know, they're global footprint and you know, if you see the onyf and the types of speakers, you know, we are basically Representing, you know, almost 70 percent of the global Subscribers as covered by these carriers, right? You know DT being one of the top ones So without any further delay, please welcome, you know, Arash Asharia SVP Group technology innovation at the Deutsche telecom rush Yes, good morning. Good afternoon. Good evening wherever you are Thanks for having me what I like to do today is to After all these great talks from different industries Put a bit of an operator head and look look at this whole development from an operator view And from an operator view, obviously If you look at different aspects, we have Challenges on one side because There's a level of competition with hyperscalers on one side and cooperation on the other side there is There's a lot of change in customer needs and behavior We started to see that with the launch of 5g And particularly with the pandemic and the way The pandemic forced A hybrid way of working Top-notch connectivity anytime anywhere Is bringing new requirements Then you have Obviously from an operator's perspective It's very difficult to innovate So you always need to be very focused about, you know, where are you going to put your bets on? And We have As Deutsche telecom, you know, we always have this ambition not only to be the leading European telco, but In the innovation is our DNA and we have a set up Very clear strategy For infrastructure, which is basically Transform that infrastructure to new open distributed cloud based And eventually cloud native architecture with simpler more intelligent and automated production model end to end And Obviously the three ingredients to do that is desegregation both in mobile and fix. I will cover that It will be the use the notion of cloud and cloud nativeness orchestration And most importantly in order to monetize the asset open APIs This We have been sharing the left part of this picture with all of our partners around the globe, which is This is the foundation from the network differentiation program, which we have And it boils down to the three things I mentioned desegregation Orchestration application and open APIs now if you look in desegregation It's public that for us open run as a founding member of X run, which then turn into open run We clearly believe in in Open in open ecosystem And we are heavily driving that As within all the communities From from our alliance itself to tip to in the foundation to own up and to many others on f to drive that open ecosystem And we believe this is not something Um Individual operators should drive for their own we need to create scale And we need to create speed of demand and therefore we have lined up with all the leading european telcos and have published an mou on oran in order to To foster innovation and to drive the ecosystem development with high speed And we also have set up an Uh open tick lab in berlin Which is not a telecom exclusive. In fact, it's together with telefonica and both of them are competitors in europe And all our major suppliers were open like nokia Who are open mavin here? Samsung were open to um work with us in that space And the idea there is to jointly drive open standard leverage collaboration And create an environment where you can have different telcos Working and uh collaborating with the suppliers on the fixed side We have been pioneers on what we call excess 4.0 Which we did on the on f to drive desegregation on fix basically on OLT level and vng we have Tried in labs. We did pox and we are running customer friendly ties and we are actually live with Thousands Few thousands of customers and we are now looking for partnerships to scale that up and make sure that we we can Roll this out across our footprint starting in germany In terms of open apis the real promise of 5g Is obviously network slicing industry connectivity, but we also believe there is more there is um We have defined control points In terms of how do we want to collaborate with hyperscalers and what is strategic for us? So for us the network is a platform rather than just a pipe for connectivity and we are trying to Make use of common apis to drive monetization Even in hyperscaler partnerships in the future And then last but not least in terms of Leveraging the cloud for the telco We think finally with 5g s a core as already mentioned with some of the previous speakers We have the right technology at hand to as a foundation to drive drive cloud nativeness In an operator environment and for us at Deutsche Telecom 5g core s a which We did very recently an announcement is the first cloud native network function um at Deutsche Telecom And obviously we're going to build on that to drive network slicing and development of end to end capabilities for customers b2b and b2c And we have running trials with multiple leading suppliers um For those of you who are following Deutsche Telecom, uh, you know that the zero touch Automation has been a vision which we have been heavily vocal and promoting for many many years and We have been active in in multiple organizations in the industry Driving the economy as driving the community towards that vision and we we Obviously automation is a transformation which we started years ago And it was domain by domain project by project And I think we have achieved the point where for me if I look, um from now onwards is about Harmonization and scaling up so that you can reuse and reproduce those Those solutions across across your network and across different countries We really believe in engagement in open source and standard interfaces And to drive apis for successful integration and uh, I personally believe that this is key If you want to master the desegregation you need to master The apis and the interfaces and the integration part of that now We are a big player in the community. Um, as I mentioned earlier on, uh, um in oran as you know, um, um There have been a founding member in fact, um, Alex. Dr. Alex Choi is uh, the coo of the oran alliance. We, uh, we heavily focused on this agrational strategy for the run part on oran and For the first time at least to my knowledge We went from day one. We planned this to be not just another park or another lab Actually, we have developed the concept what we call oran town Which is which is a big program run by the best people in the company cross functional team between a group and people from Germany check and other places within our community Which is about to bring oran to life and the oran town is actually a life is a physical area In the life network in germany where we're going to build um an oran network 5g oran network and where we want to not only test and see In a real scenario the behavior of the network and the customer experience, but most importantly we want to have that network integrated into our operational procedures Which is one of the most difficult tasks and there where on up is coming to play So we have decided to develop a service management orchestration For for the oran town fully based on on up We have completely Upgeared the amount of resources And our engagement in this case to make sure that we learn we take the community But we are also able to adopt the the learnings And make sure that we are ready for the next modernization cycles with a full end-to-end open run solution Which actually can be operated for with a operator grade Class given, you know our our dna as the network leadership and quality leadership brand who we are And another thing I want to mention is that The engagements in all of this different for us Has enabled us to influence to participate and to contribute On the journey For the transformation, but it remains one major challenge, which I like to address and to actually Have a call for action And if we are honest to each other is All these initiatives over the last seven ten years have been The the the foundation of that was always the thought There is something We want to address or some bright people can be the idea to address And they believe that it cannot be done the standard way By the classical traditional standardization bodies for multiple reasons Speed coordination so they People are coming up with a new initiative And I have to say now looking back over the last couple of years We had one initiative after the other and they are growing like mushrooms Now all of them Makes sense. The problem is we are desegregating So we have I think one of the most and this is this is my plea To the whole forum and that's what our industry is We need to make use Of all this for us, but we need to be careful because we are risking in redoing work There's a bit of competition and from an Operator perspective and I'm sure for the suppliers is the same We have it's basically the same resources So you have a resource pool which you need to put on all these for us and try to drive and to to manage So my plea would be let's work more together Let's see where we can collaborate more And where and be very precise what topic we want to address where To make sure that we use those forum as best as possible and don't create any kind of overlap and And any kind of let's say duplication of work And number two is we have In terms of at the end of the day, whatever we do We we don't want this to be another track. We want one track of standardization So there needs to be also a feedback loop and I think that was mentioned again earlier by One of the presenters that we need to make sure that all these great initiatives all this innovation Will ultimately found its way to the main track and doesn't stay as something let's say An alternative Because what it does is it it creates diversion and we are missing again scale opportunity now if If I would have A couple of Let's say wishes to history and I think to all the listeners From a telco perspective, it's essential that we have to try hard To harmonize the past layer In order to low the effort of onboarding for all the different Variation and inclination of operator class So I think we we need to really think to go one level higher and think about The ways to try to harmonize at least limit the number of variation in On the past layer to To enable Faster adaptation why because if not, there is a very high cost of integration When if you put that cost in every project, you will realize that the business case won't work And a lot of time if you go for the cost you you're gonna Not be cloud native or you you basically are not using the latest technology. That's number one number two is with 5g Stand alone and again, that was that was covered previously from the from I think it was Ericsson But from the operator perspective Whatever we do in this space the cncf Compliancy has to be there We need to make sure that the the solutions are compliant so that We are all in this industry able to monetize the The infrastructure and expose apis and then obviously i've been following The development of on up and I think we came far Long journey, however My plea is we need much more engagement particularly from the operator community We need more engagement if we want this to be If we want to this to be a solution where The local adaptation is just part of the job, but not basically Needing to working out many many different aspects of the solution, which is still a bit in development We need more operator engagement to with all the experience from integration day to day operations all the pain points This needs to be fully covered and we are we are super committed We are putting a large number of people In all these initiatives and driving that and again, we are happy to work together I think if there is a biggest learning in the telco industry is you compete on customer experience And Wherever it makes sense and legally allowed you collaborate on technology development and technology innovation And with that I end my presentation. I'm happy to take questions. Thank you Oh very good rasha excellent excellent call to actions and uh, you know the The questions are coming in. I would like to make sort of three quick comments On yours as well. Uh, you know, it was just uh last on yes in you when dr. Alex joy actually presented Or launched the concept of or on town You know, it's just less than even six months. It's been Such a big success. You know, thanks to you and your leadership You know, particularly excited about The use of on app and the contributing community. So that's kind of very good Um, I like your reuse and reproduce statement You know as as as as you scale and as you kind of get there. Um on the pass. I do want to uh Bring up to the attendees as well. Um, a new project called xg villa has been uh incubated into lf networking It's kind of a contribution from several service providers globally That is looking at the past problem and and kind of standardizing it So your call to action there is absolutely important and then on the cloud native as well cncf and lfn are extremely tightly working on the whole definition testing workbench, etc So Absolutely great insights. There's one question That has come up, which is uh, would dt prefer if uh 5g ran? We're completely standardized around oran or can uh, you know There be a partial oran with a coexisting proprietary ran solution work also in the interim like like how do you envision the Full generality to oran based solutions from from where we are today Very good question. Thanks for that and to be honest If you if you ask what we vision and what we wish obviously It has to be for us as a carrier grade one Requirements it's essential that it's oran based and ultimately All the oran standards should be the whole specification should be adopted sooner or later with gpp so As I said early on you And we have seen this in the past in in many different areas you The the biggest success of the mobile industry has been that there were It's our ability across the world for different purposes Now what has happened is with that And it was great. And you know, I'm I'm someone who is very proud to spend 20 years Soon 21 years in the telecommunication and I've been following, you know, the hype before 3g You know and the big The big disappointment on 3g The great success of 4g and now with 5g. However, we have to admit the standardization of today is Dominated by suppliers who put a lot of efforts and yes, and so single run is a closed system. So with with With the oran alliance The idea is not only to open up interfaces to have an alternative But basically make that the de facto standard of the future along the way And that's what our plea and that's where we are working that together with our with our other operate Community as well as our partners Gotcha Now this is exciting very exciting and I'm we really appreciate all the leadership From from Deutsche telecom on on the open source world exactly. Absolutely. So thank you for taking the time very insightful You know, I'm sure the community gained a lot, especially your call to action. So thank you very much You're welcome Alrighty and so to wrap it up. I want to Present two quick slides as a summary. So if you want to sort of put that up It's all collaboration and open innovation and you must have seen through the the rest of the Three days that we focused a lot on that So if you go to the next slide I showed this as an architecture diagram and you heard the theme You know from disaggregation to re-aggregation, right? So from poc to production, right? Chris Rice said it. Well, Arash said it very nicely, right? How do you do collaborations between these projects so that they can be used in an end-to-end? Solution we heard from uh from our networks on how Proposal to do a 5g super blueprint could be done in in lf networking Uh, so I think the the The industry is moving from these individual projects To gluing them together for deployment. So not only are these projects deployed individually, but the integration is all becoming open So it's like open solutions. So if you go to the next slide, which I showed Um, this is why when I open the uh industry open on if I showed this slide and it may not have registered with everybody But you now can see that we had presenters From multiple verticals that are wanting to use open source in their industry They don't know everything But they do know that this is the way to go forward We heard from madland today on the broadcasting industry, right? Where on app and oran on app network slicing oran as cplay a large The building block we heard from the the government of uh and dod yesterday or sorry on on uh on wednesday We have heard from governments and other policy From from developing countries in ea pack, etc Right where there's a lot of global connectivity going on. You heard brian talk about that in in the facebook presentation, right? How do you drive it into the remote and the rural areas build on these in open source projects? And then most importantly the enterprise networking industry, right? Whether it's private fry g lte You heard from shan on all the end to end visibility and monitoring you heard from wal-mart, right? Trying to drive visibility across their workloads And then most importantly you heard from several of these verticals in the edge and iot space that are utilizing unique applications on cameras or sensors or manufacturing or retail so this is all coming together And we are really excited that you know, we have been An enabler to help facilitate you all coming together and contributing here So that's kind of my closing uh 2021 is going to be a great year. Uh, we're looking forward to seeing everybody um Hopefully in person. Uh, this is right now planned as a hybrid event in october 11th 12th la This will be uh a large You know co-located and a joint event, right? So so that whole week, you know, just just block it off, right? We have kubecon right after ones and you know, we're integrating everything seamlessly for you But you know, let's stay safe. Stay healthy. Thank you all for joining and um Thank you for all the questions. Hope you like the format of three days three hours three time zones Very quick No fluffy stuff live and thank you all for all the speakers Uh signing out now. Thank you