>> Announcer: Live from San Diego, California. It's the Cube. Covering Cisco Live US 2019. Brought to you by Cisco and its ecosystem partners. >> Welcome back. We're here in San Diego Convention Center for Cisco Live 2019. I'm Stu Miniman and my co-hosts Dave volante. Lisa Martin's also been here for our three days, wall to wall coverage with about 28 to 30,000 here in attendance for the 30th anniversary of the user and partner show for Cisco. Happy to welcome to the program our first time guest, Jeff Scheaffer, who's Vice President of Product Management Strategy inside of Cisco. Thanks so much for joining us. >> Thank you Stu, glad to be here. >> Alright, so Jeff DNA center, some of the environments that you have. There were some announcements on Monday, why don't we start there? Walk us through the updates to the product line. >> Thanks Stu. So we're really excited, right. >> So as you know, with Cisco, we're really building out the intent based network, in support of digital transformation for all of our customers. And one of the key aspects of the intent based network is that we have incredible programability in the network, all supported through the DNA see controller. And this week, we're really excited. We announced two new innovations in the controller. One was incorporating a new set of analytics and machine learning capabilities as part of our assurance package so that we can more quickly troubleshoot network issues. And the third is being able to connect together the multi dev environments. So how do we stitch together, the software defined access to the Software defined WAN, to the software defined data center so that we can lay digital services across the entire network. >> Stu: Yeah, it's a story we've been watching the last couple of years at Cisco is that this move to software in many ways as a unifying factor. >> Yes. >> Used to be I had all these product lines, and I need to learn the interfaces, as my friends that come to the show for many years, when you said single pane of glass, there's like, come on that's spelt P-A-I-N. Today's world, it's an API economy. And what's been really interesting to watch the last... Recently, a lot of times, it's that ml and AI underneath that spans and helps automate a lot of those pieces underneath the covers now. >> Yes, absolutely, it does. So the thing we're excited AI is a broad topic as you know. And underneath that umbrella we have built in new capabilities around the machine learning, the ability to do deep learning, as we look at anonymized data sets in the cloud on deduce patterns that people don't know yet. And then thirdly, we're looking at machine reasoning. So how do we take that the analytics to pinpoint or identify anomalies in the fabric of the network in these new IBM fabrics and then be able to couple that with a set of orchestrated automation so that we can we can emulate the behavior that a network engineer would normally do in order to troubleshoot and diagnose problems and so we're that much faster at identifying them, pinpointing the root cause and then actually being able to give recommendations if not automatically fixing it, the recommendations to resolve the issues. >> So the programability aspect, of course, we're here in the DevNet zone. So, can we dig into that a little bit? I mean I'm imagining by that, it allows me to provision, bandwidth performance, certain levels, all through API calls and it's through software. And I can set thresholds, I can talk more about what that does for me as a customer. >> Sure. So in general, within DNA center, we have a kind of very rich programability capability. It's very much an API first developed controller. So everything that we can do within the controller is a setup of published and curated API's. And those API's come in different categories. So we have API's that are around the automatic configuration and the provisioning of the network infrastructure. Very much, Dave as you're suggesting that you can run as a headless entity inside of a if somebody like a lot of service providers and partners are using this to offer a service to their end customers. And so they can automatically provision out the network and they can do that in support of new applications. We're really actually excited with that in our partnership with Red Hat, we just introduced a new set of answerable plugins as an example, to support a DevOps process by which software developers build new applications as part of that DevOps code pipeline. By using the answerable plugins, we can actually drive automation into the network to provision the networking in support of those apps. So a lot of capabilities there. >> So is the uniqueness of that you can actually do that or is the uniqueness that you're Cisco. >> I know. So to a large part of it... In terms of the way things have been done historically, and this is that most of the time, if you were to look at how do you configure a network, it was device by device and it was through command line interfaces. >> Right. >> And it's error prone, it's complex, it's cumbersome, it takes a lot of time. And so what we've done with the network controllers, is by moving above the top of the intent based network, we're actually able to automatically configure and provision either policy through policies, either QoS or segmentation for security and do that in a very automated way. And to be able to do it in a scalable way. So that is an amusing new kind of configuration mechanisms we're very proud of the work we're doing around NetComp and YANG Models and that's something that's very unique in the way we're approaching that in the market today. >> Yeah, Jeff, historically, one of the concerns was the network. It's tied to the application wasn't as tight there. It's like you talked about networking people, it's just the water that runs through the pipes that we watched with first with ACI now with intent based networking. We're getting closer and closer to that application, especially you talk about multi domain what's going on at ties right with the microservices architectures that are coming on there. So as the applications get more complicated, the network needs to be able to understand what's happening there and respond to what's needed and give back the services that they need. >> Yeah. So I think building on where you're taking that conversation, one of the really kind of key or anchor points of the whole idea around the multi domain integrations that we provided, when we bring together Software Defined access, the software defined when and the software defined data center. We're using the the rich API's that exists in all of our controllers, whether its DNA see for software defined access, whether it's vManaged for the software defined WAN, or whether it's ACI around the software defined data center. And we're integrating all three of these controllers so that they can do a couple of really important things for customers. So in the context of the application. So the very first thing is how do we provide for segmentation? So segmentation is about how do we provide access policy and drive that into the network so that the right users have access to the right applications and other people don't. And then being able to use the programability. If the user moves from say the campus to the branch, we can automatically have the policy follow the user. If the application moves from the data center to the cloud, We can have the policy automatically follow the application and that way we always ensure that the right people have access to the right applications at the right time. The other use case around the around the API's and the multi domain segmentation is that to the extent an application needs quality of service. We think about an 82% of the US we move forward in time or roughly 82% of the traffic on the internet is starting to become video on the networks and like an 8k video is 7.29 terabits per second. So how do you ensure that there's enough bandwidth and enough quality of service to ensure the latency of like virtual reality or augmented reality where latency matters. And so the other thing we're doing with policy is provisioning up the Qos, so that as you configure it in the data center, it's honored in the software defined WAN and it's honored it all the way to the end user in the software defined access. >> Stu you were saying earlier how traditional is like don't touch my network, right, okay. So Jeff, my question is if you look over the last 10 years, as you bring in the software design defined data center and infrastructure programmable and infrastructure, infrastructures code. How is that change? And is it changing the relationship between infrastructure pros and application developers and application development heads? >> Yeah, absolutely, absolutely. So kind of comes from two different angles. So number one, from the extent that every software development organization becomes more and more of a DevOps organization and they want to be able to deliver value, the experience, the velocity around the applications to production, as they go through those code pipelines through the DevOps practices, they need to be able to automatically provision out test environments, they need to be able to automatically provision out the pre stage and then automatically promote the applications into production environments. And that's what why it's so important, for example, as we as we mentioned earlier about the answerable plugins that allow the configuration of the network in support of application DevOps teams. The second thing that we find with the network engineers themselves, to the extent we've defined everything as software defined. We have this rich set of digital services that we're starting to manage in these infrastructures and through the controller that programability, the controller, network engineers are having to evolve their capability. And so, you know, we're really excited that the show that Susie Wee, who leads our DevNet program introduced a whole new category of offerings in support of how you start to become both not only a network engineer, but also development aware through for example, the DevNet automation exchange, by making DevNet certifications available through Cisco certifications. And by hosting these curated communities, where we're able to contribute new applications that run on top of our controllers. We're able to contribute new integrations, new automation, logic, new AI, ml logic, everything that's possible. And by curating it, it means it's curated code, but then it's stored in very common repositories like GitHub. It's still really excited with everything she's doing. If Cisco is going to prime the pump with initially right with apps and code started-- >> We've been priming the pump. But we've actually over the last little bit, we've developed a number of partnerships. So we have over 25 partners that have built value added applications and integrations for example, DNA center, we have a number of partners that have actually taken DNA center and wrapped it with a new kind of a new service offering a solution offering to the network. I'll give you a great example. So one of our partner that Tall tale. So what they've done is they've taken DNA Center at the core and our ability to quickly provision wireless infrastructure. And they've wrapped it with a hook telling solution for Convention Center. So think about a convention center, you have a large space, multiple concurrent shows, every one of them setting up their own wireless network with their own SSID. And so the ability to set these up, tear these down, resolve problems quickly in these high density spaces. And they built out completely as a solution on top of the programability of DNA center and our IBM infrastructure. So it's a very exciting time. >> Alright, so Jeff, some new things announced but I also want you to touch on I believe, about a year ago, API and programability was launched. What lessons learned? What feedback from the customers? Give us the update. >> Yeah, so we launched the programability. And as we mentioned before, it's a kind of a very broad ecosystem of capability. Now, the first thing about, programability is really meant to do a couple of key things. One, and probably most important, is to enable all of our customers to be able to make sure that everything's fit for purpose. So how do we integrate into their existing enterprise ecosystems. And so we know we're API first, we have SDK sample apps, we have a bunch of out of the box integrations that use all these SDK for example, integrations with service now and others. And so that's new. The second thing that we're doing with all of the program abilities, we're using it to enable the community. So the challenges we all know, in technology spaces, is that the rate of innovation continues to expand. It's like innovation is on the sphere. You know, it's the surface area of the sphere. And as every year progresses, that surface area gets bigger. The rate of innovation gets bigger and the ability to keep up with that, exceeds the capacity of most organizations. So the best way to harness that is a community based approach. And if you don't have communities, if you don't have programability, if you don't take that strategic approach, it's very hard to stay current and relevant to the market of the future. So that's another big thing that we're really excited by is being able to track that community and then an ecosystem of partners in development. >> Great, Jeff, want to just give you the final word, a lot going on in your space, any final customer takeaways or things you want people to walk away from your team from Cisco Live 2019. >> Yeah, so the present thing that I'm most excited by is that as we look at the transformation of networks, from kind of the device centricity to how do we manage segmentation in these multi domain fabrics. And do that in support of ensuring quality of experience, so the bandwidth and the latency and support of the applications. Ensuring the segmentation and the secure, being able to minimize the attack surface on these new infrastructures by driving segmentation through the fabric. It really takes kind of a systems approach to this, which is how we bring the intent based network together with, for example, the new Cat9K family from Cisco and then bringing DNA centered the controller together. By bringing those two things together. We're really helping to change the entire architecture of the industry for the next 20 years. So very excited to be here. Thank you guys. >> Excellent. >> Dave: You are welcome. >> Well, Jeff Schafer, really appreciate all the updates. Congratulations on what's going there. For Dave volante, Dave, you have a final word to say. >> Well go Bruins is the only thing I can think about right now Stu. >> That's what I was expecting you to say. So Dave Volante I'm Stu Miniman. We back, getting towards the end of three days wall to wall coverage here from San Diego, for Cisco Live 2019. Thanks for watching the Cube (techy music)