Unleash your Cisco Nexus Network potential

Hello again, I'm Shannon Phoeby, the DCN.00:10Program manager and I'll be your host for today's Nexus one webinar. I'm pleased to introduce Robert Burns, the product develop development manager for Nexus dashboard. Before I hand over to Mike I have a few housekeeping items about this presentation. Today's webinar will be recorded and available on.00:15In our DCN community. I'll provide the details at the end of the session. We'd love to hear from you today during the presentation. If you have a question.00:35Please feel free to send it through the Q and A panel. It's near the bottom window. Without further ado, I'd like to kick things off by handing it to Robert Burns. Awesome. Thank you Shannon. Very happy to be here today. Thank you for everyone for dialing in and joining today. As Shannon said I'm a product manager for Nexus dashboard.00:45I've been working at Cisco for about 18 plus years, so I've been around the block in various roles, and really excited to be in this position today to speak to you guys around Nexus dashboard. We really want to focus today. This is gonna be a session for folks that are either net new to Nexus dashboard have never heard about it or for those who have maybe thought about.01:05Even demo it in their lab. But really our goal here is to kind of open up the eyes and see what Nexus Dashboard can offer to you in your environment.01:25So again, why, why are we here? This is, you know, we all want to understand what's the focus here, what's driving this adoption of the Nexus dashboard? Obviously there's a lot around costs and everything from gasoline to electricity has gone skyrocketing today. So every customer we speak with daily is really focused on getting control around costs. How can they do more with less?01:36How can they be more, again, intelligent with their investments and get their return on investments? And we're trying to do this at the same time, everyone's trying to grow. You know, we've seen a lot of AI explosion over the last, you know, year here. Companies like Nvidia and AMD are just taking the market by storm, and if we're gonna be doing that, we know there's large customers doing it, but also we're starting to see a lot of small.01:56Our customers also investing in AI infrastructure. So how do we scale that? How do we grow by again being more efficient so we can again balance what the business wants to focus on? Now, everything in data center, even though we're I'm primarily data center focused, security is at top of mind for everybody. There's been a lot of obviously vulnerabilities that have been released in the world.02:16Build and especially with the explosion of AI even more relevant than ever. So we really want to make sure that as we're operating our networks, we're doing so very securely as much as we can, but doing so safely as well, so we can accommodate the advancements we want to go towards with AI, but also again do it in a very structured approach. And I'm really proud to work with Cisco because we.02:36We've taken a very security heavy approach towards AI. We understand the value, we know it's here to stay, but we're also being very careful on how we implement controls that we can better manage and and ensure that everyone remains safe around, you know, the things that do come along with the AI investments.02:56So as operators, you guys might be running networks or running systems in your environment. Surely there's a ton of different operational challenges here, but just a few to kind of highlight just to set the stage, obviously things that could really slow you down from a day to day perspective could be things like manual workflows. Are you configuring your network switches, you know, by CLI or are you leveraging.03:13In some sort of automation tool like Ansible or terraform or network as code. A lot of ways we could really help optimize there.03:33Other ones we have as well are gonna be things like skills gap. There's obviously a lot going on with technology today with ACI NXOS, there's VXLAN, so there's a wealth of new technologies at our disposal. So there could be a bit of a skills gap there as new recruits are coming up to speed in your organization, and that's where having a controller like Nexus.03:40Dashboards really gonna help expedite that learning curve to make sure we're doing things correctly and making sure that they're done to best practices.04:00Now we do that on a lot of different ways. We obviously have best practice recommendations, we validate your config, so as you guys are building your fabrics and doing your configurations, we're gonna make sure you're doing it in the best, most Cisco validated way. Now, in terms of managing that as well, we're also using tools like AI assistance that we're building in our system.04:08Right now. So in the coming year here, you're gonna see some really cool enhancements on how we're gonna leverage things like AI assisted remediation to better predict how things are happening in our network and how we can really target our operations to be more efficient in that matter.04:28Obviously, escalating mean time to recovery is very important as well. So how can we shorten the amount of time that operators like yourselves are spending debugging devices? I came from TAC many years ago and it was really a hop by hop basis back in that day. You had to start with one box, you did your troubleshooting and that would usually point to your next step.04:43So how can we reduce that overall manual effort and let the system like Nexus Dashboard help us with our debugging and troubleshooting?05:03Alright, let me build this slide out here. So Nexus dashboard, what is it in its entirety? So Nexus Dashboard is a platform that delivers a whole suite of capabilities from provisioning your network from day zero to secure.05:17Bringing it making sure it's always in real in exact lines with your intent.05:30To doing your management, expanding it, growing it, obviously you're gonna start off small probably and then branch out. So from a single fabric deployment to a multi global site deployment, Nexus dashboard's there to help you make sure everything's connected nice and securely. And the big area that I like to focus in on is around the telemetry piece. And again, this is all part of the.05:35Nexus dashboard umbrella, is it provides that telemetry to give you visibility.05:55Looking at things like anomalies, what's behaving poorly, what's behaving healthy? Are there any flows that are being dropped for any number of reasons? So again, we've really wanted to help expedite and turn your operations team into powerhouses. And the Nexus dashboard is not just for land. It has a very very big focus on land and sand environments, but it is also used.06:00For storage networks as well. So if you do happen to have, you know, a storage area network using Cisco MDS, you're also going to be using Nexus dashboard to manage that storage fabric as well.06:20Now in the center here, we kind of have Nexus at the forefront, but again, we just want to make sure we understand that it's not just Nexus. MDS, there's also catalyst devices and iOS XE XR devices that can also be managed by Nexus dashboard. So whether you're running, you know, a Nexus classic LAN or ACI or VXLAN evp.06:30Said, Nexus dashboard's gonna give you that solution to configure, manage, provision, and analyze.06:50So, there's been a new buzzword going around Cisco and this is around Nexus one. So I want to just spend a quick minute here to talk about this because Nexus one is not a new product. This is something we've been developing as an architecture over the past few years. So we've obviously had things like ACI, a very application centric solution.07:00That was designed with, you know, very fine tuned controls on security and then we had our classic Nexus. Nexus, which can do all your standards, EVPN based protocols, your classic vpcs, all that is there. Now nexus one is kind of the overarching architecture that sits across all these environments regardless of using.07:20ACI or XOS. We know there's customers that have a use case for either or and our goal is to give the customers choice. So you decide what type of operational model you want.07:40If you want to have a, you know, controller based or a SaaS based offering, we've got, you know, controllers in a, in a management style that is gonna be best suited for your environment. So Nexus hyper Fabric is our SaaS offering. Again, it still falls underneath our Nexus one umbrella, but that is obviously a SaaS managed solution.07:50Still using the same Nexus 9000 series switches. So again, doesn't matter how you wanna deploy your environment, it really comes down to your operational model. Do you want to have an on prem device controller you own or do you want to do things through a SaaS managed offering as well? So depending on which you like, that is completely up to the users.08:08Now all these foundations are really centered on four very important pillars. The silicon, right? Cisco's very big on producing our own custom ASICs. We can do more with them, we can develop them faster, and they have them far more rich capability than you would get off the shelf. So we've got cloud scale, which has been around for a number of years, extremely.08:26Powerful and very feature rich. We've also got our next generation silicon one asics. These are gonna help us achieve very dense 800 gig and 1.6 T fabrics that we need to serve all these AI workloads on.08:47And the colors, we also partner with Nvidia to work with them in their spectrum excellent of switches where we can provide again their their ASICs inside of our switches as well. Now system wise, all these regardless of whatever operational model you like, they're all using Cisco N 9000 series switches. So at the end of the day, you decide the operational model, the hardware is gonna re.08:59Remain the same across all these architectures.09:19Optics is one area we don't want to overshadow as well. Optics are not just, you know, vanilla. There's a lot of advancements that we're gonna be introducing into our optics to give better visibility, better troubleshooting because we own the ASICs, we own the switches, we want to extend the level of visibility down to our optics. So we can do that by using Cisco leveraged optics that we have.09:22Some really cool capabilities coming down the pipe.09:41Last but not least, we've got software to finish it off. Cisco NXOS most tried and trusted NXOS operating system in the industry, very stable, very robust and extremely feature rich. We also have ACI. ACI again has been purpose built, really good around segmentation, has a very structured object.09:44Model, and then now we're also supporting song. So if you want an open, an open shared environment, you could use that as well. Open source has been top of mind for a lot of customers. So again, we're giving you choice. What type of operating system do you want to run on your devices? And our N 9000 series can run all those solutions on top.10:04Let's talk about two really important areas when we talk about AI because there's there's two pieces to this. There's the networking for AI, which is how we're building our AI fabrics through front end and backend inferencing. And then we also have AI for networking, and I'm going to differentiate the two and kind of talk to how we leverage both of these with Nexus dashboard.10:25So let's start off 1st by talking about networking for AI. So we've seen obviously a huge investment by customers who are looking to build out their own AI infrastructure to better serve their needs of their business. A lot of customers may not be doing so much as the training, but definitely from an inferencing perspective, a lot of customers are looking to build.10:45Build their own networks to do inferencing and their own llmiss on prem. So again, Nexus Dashboard has been tuned and can build those AI fabrics for you. We've delivered a lot of really powerful features that are specific for looking at AI functionality, looking at job completion time, looking at GPU health. So a lot of these things really help Nexus dashboard.11:06Give you that edge when it comes to AI monitoring and networking.11:26So taking a quick look in here, this is just a quick little sneak peek of if we're inside of our AI fabrics, which is one of the many fabric templates we can deploy with Nexus dashboard. You're gonna see, again, this is a pretty hefty network we can see in the background here, but our goal here is not only to give you topology, topological information, but also overlay that with things.11:33Is like job completion awareness. So that is the most critical piece on an AI fabric's performance. How quickly are we performing jobs or how quickly are we providing that 1st token time to 1st token? These are all very critical metrics that we need to be have closely monitored by our operational teams.11:53Now from a troubleshooting perspective as well, we have ways where we can monitor link health, we can also talk with the next down to the GPU. So we go right down to the hosts and we can actually share information to Nexus dashboard to relays some of those operational health metrics that we want the operators to be aware of. And then as an operator, we can decide if we have to move workloads around.12:11Or make any capacity adjustments.12:31Now, on the flip side, we talk about AI for networking. This is where we've really invested heavily into Nexus dashboard. So there's a number of ways that we've actually gone down this path, and this has been a journey we've been on for a long period of time. So these are some of the things that we've done and that we have in the solution today in terms of.12:35AI capabilities. So things like conformance, looking at your environments, making sure we're modeling it correctly, making sure you're running the right versions across the board and that also everything is compatible. But from a conformance perspective, we can also make sure as well that everything's compatible, you're not running out of life, you've got enough runway left in your environment. So when you.12:55You're cycling out old switches or you're finding out when's the best time, the life cycle of the old switches, we give some prediction on here's how much time you have left remaining for both a hardware and a software perspective, and we can even forecast that into the future. So you can say Nexus dashboard, give me my 18th month outlook on what of my devices are gonna be going last day.13:15To support or when they're gonna go, you know, end of life, and again, as an operator, that's very valuable information so I can plan from a budgeting in a, in a, in a operational perspective.13:35We also have things like insights for SAN. So again, Nexus Dashboard does both land and SAN. So the AI piece on there, we leverage it to look at how are things operating and we can make recommendations to you on how you can better improve and tune your network from a performance perspective.13:45So one of the really powerful things that we can do with Nexus Dashboard is we can model your network. We know where you're running at and how your baselines are, and then as we see fluctuations or deviations from your known working good state, we can raise that to the operator, so you can be aware that, hey, we're seeing some different behavior that we've typically seen. We want you to be aware of being the operator.14:01So it may not necessarily be a bad thing or a problem yet, but at least we're giving you some heads up that we're seeing a difference in behavior than we typically modeled against.14:22Now, another thing you're gonna hear a lot about at Cisco Live, so if you're going, we'll look forward to seeing it there, and if not, we're gonna also see lots of recordings available on some of the great announcements we've got coming. And one of the big announcements we're gonna be making at Cisco Live is all around Cisco Cloud control. And why am I bringing this up when we're here to talk about Nexus dashboard? Well.14:32Cloud control is gonna be kind of Cisco's solution that's gonna go across a number of domains. We're gonna be able to look across your security portfolio, your campus, your data center, all across these to give you more actionable insights, and we're gonna bring all those informations together. Now, where Nexus dashboard fits in here is it's one of.14:53For integrated products that we're gonna be going forward with on day one. You'll see more and more Cisco products added to this list as we go forward here.15:12And one of the really powerful things that cloud control is gonna bring us, especially as being operations, is this thing called AI canvas. So think of AI canvas as being kind of a generated dashboard of all the things that are really important. So if you're a network operator and you have some problems with an application.15:20AI canvas can generate a real time contextual dashboard bringing in things like the healthier fabric or the healthier WAN or your links, all these pieces come together again come up and bring you a very purpose built dashboard that you can interface with. So AI assistant's gonna be working on top of that as well. So imagine using like a chat gp.15:39Key style of interfacing saying, hey, I'm noticing application XYZ is slow, is anything affecting that? And then cloud control can go to all the various elements and determine are there any known WAN issues that thousand eyes might have identified? Are there any security changes in the network that have happened that might be blocking something or any problems with any of my links in my fabric.15:59And we can bring all that information back to you from the operator's perspective. So again, this is just a very sneak peek of what's to come, but we're really excited about cloud control coming out here in the near future.16:19So just double click on AI canvas. This is kind of what it'll look like. These are, you know, very early early screenshots of what it's gonna look like. We're gonna be showcasing a lot of really cool demos at Cisco live. But you'll see you're gonna have a generative UI interface. You'll be able to ask real time real questions in kind of natural language, and then we're gonna leverage interfaces with.16:31With our various solutions through an MCP, that's how we're gonna interface rather than going through kind of a legacy API. The MCP is gonna give you more kind of directed capabilities of what information I can extract from Nexus dashboard or from campus environments or from or from security.16:51So our goal here, especially with AI canvas, is to break down some of those silos. So while I may not be able to access or manage other environments outside of my responsibility, it is gonna bring that, you know, maybe read only view to see what's happening in my adjacent silos. So that way I've got a full view of what might be impacting or causing issues.17:08As an operator and information's always king in any type of operations role.17:28All right, let's focus in on Nexus Dashboard now. So we kind of set the stage of why we're here and where we're going. Let's zero in now and go drill down a little deeper. So Nexus Dashboard has been out for a number of years. We've recently done a merge of individual services. So Nexus Dashboard prior to four.one.17:34It was separate instances. So you installed your Nexus dashboard, think of it as being kind of a hosting platform, and then on top you would install services. So if you wanted to do NXOS automation, you would install the fabric controller. And if you wanted to do orchestration for ACI, you would install orchestrator. And if you wanted visibility like telemetry.17:54You would install the insights piece.18:14And these were all separate services that sat on top of Nexus dashboard. So from version four.one and later, we've merged these all just into the base Nexus dashboard. And now we've got a single unified solution that just provides all these capabilities. So we don't really talk about fabric controller or insights anymore. We just say Nexus dashboard.18:16And that's gonna give us all those capabilities at our fingertips. Well, how do we deliver Nexus dashboard? Let's start there. We can deliver dashboard as either a virtual or a physical appliance, and depending on what your preference is, you know, if you prefer having virtual appliances and you've got the compute to run it, we can do up to 200.18:36Switches in a virtual cluster, and you can scale that down as well. So if you, you know, have less than 100 switches, three nodes will give you up to a hundred switches, and if you want to go to the top end, six nodes will get us up to 200 switches fully managed, meaning I can manage those switches, I get all the telemetry, all the other capabilities are available to me.18:56Now, if you need to go larger than that or you just wanna have an appliance that the networking team is gonna own, that's where the physical appliance are a perfect fit. So our physical appliances are built on UCSC series. They are just a purpose built appliance, so they're very hardened. They've got again enough resources to scale.19:16Forward very, very well, and in three nodes, we can get up to a thousand switches, so we can scale very large from a form factor perspective.19:36Now, if I don't have enough scale and you need to go even larger, we could do a multi cluster connectivity between multiple MD clusters. And a lot of customers will do this if they have, say, an east Coast and the west Coast deployment. They may have a cluster in the west Coast and east coast, and they can merge those together from a multi cluster federation.19:45Perspective. And then as an operator, you can log in and you can access your entire suite of all your fabrics that are under both of those clusters. Now we have all the R back in place as well, so if you need to read and write to one cluster and then only read only from the other, we can do that, absolutely. So you can divide out whatever level of access is.20:05The most appropriate for you.20:25So software wise, let's talk a bit about software. So we are coming out with a slightly updated version of our release cadence for Nexus dashboard. So historically, we've released a new major version every six months, and it's major version, major version, major version. And.20:29As customers we've talked to, they really like having a maintenance idea. They like having a maintenance bill that they can rely on. They know there's not any features in it, and really it's going to be stable. So ACI and NXOS have adopted this approach where they'll do three feature releases, and then the 4th release in that major train becomes the maintenance release.20:46And then that maintenance release becomes the recommended release that you'll see on Cisco.com. So we're gonna take a similar approach starting at the end of this year with Nexus dashboard. So we will release a feature release six months later, we'll release 2nd feature release of that train.21:06And then three months after that, we'll enter the maintenance period. So we'll have the maintenance release, and that maintenance release will be our recommended release on cisco.com as soon as we release it. So customers who want the most amount of stability can wait for that 3rd release in every major train and they can stay there or upgrade from there. So what this gives you is NO new.21:23Features will go in a maintenance release. It'll purely be built on stability and resolving any defects that have been found in earlier releases.21:43Yeah, go ahead, quick question. So at this time in time, what is our recommended release or our golden release? Yeah, good question. So right now we are our recommended release is three.22.21:54So that's been out for quite some time, so we've had a long gap between our next upcoming recommended release. Our next recommended release in the four.x train will be four.three, which is due out early July. And when that comes out, then we'll make that our next recommended release, and you can take confidence that you could always upgrade from one recommended release to another.22:10Perfect. And I remember posted that let me know if you have additional details and let me go ahead and use our Q and A panel, that's the easiest way. Thanks. Perfect. Thank you. So again, just to kind of reiterate our four.three, which is our next targeted recommended release is slated to come out.22:32End of June early July, so we're just around the corner from it. And then this new cadence is gonna start in the next version after that. So towards December of 2026 this year, we'll enter this new cadence and you'll have that December, June release, and then the follow on maintenance release.22:51Okay, so let's put outside the fabric software and all the hardware pieces and let's talk a little bit more about Nexus dashboard. So one of the things that we can do with Nexus Dashboard or it's highly used for is around fabric management. So to build a fabric, to manage your devices. Now this applies even if you don't want to do automation.23:10If you just want to do monitoring and do telemetry only, we still create a fabric kind of construct.23:30Four devices. So we've had customers who aren't using a controller today, they just have a bunch of standalone devices, but they want to get some level of telemetry and monitoring for those devices. So we handle that in Nexus dashboard by creating a very classic fabric and you can onboard those devices. Now, I don't have to manage the.23:36If you don't want to. Absolutely you can, but it's not required that you do so. This is something you can decide if you want to manage the full config and telemetry or maybe I just want to use as a monitoring only fabric and we can decide how to, how to do that.23:56But if you do decide to go fabric manage or managed fabric, that means that we can help you manage all those policies, all the configuration, and ensure you're running per Cisco best practice. So we'll make sure that you have policies that align with your devices. And the great thing about having a controller is if something is misallowed.24:10Find with your intent, we can flag it. So let's imagine you have a golden config that you've decided these are the policies that I want deployed on my devices. Well nexus dashboard has that intent at all times, and we're constantly checking your devices.24:30Well, if somebody should log onto a switch directly and they make a change, we call that kind of an out of band change, then Nexus dashboard will raise that as a config drift. We'll say your config is here, but this switch over here has a different configuration actually running. And then you can view that and look at it side by side and see what exactly has changed. And then.24:44As the operator, you get to decide. Do I wanna override the config and push back what my original intent was? Or do I wanna keep that and ingest that as being part of my new intent based config? So again, really important to make sure your devices stay within the the intended spec.25:04Now, one area I'm really passionate about, the whole I'm passionate about Nexus dashboard, of course, but one area I'm especially passionate about is all around visibility. Again, coming from TAC and working in the industry for a number of years, the more I can see, the more information available to me, the more powerful I can be in my role. So visibility is top of my.25:27For me especially. And one area that Cisco does excel, and I can absolutely say with whole confidence that we blow away the competition is around telemetry. So we've got two ways we look at telemetry.25:47We have software telemetry that we can pull from the switch. We can look at things like the control plane, the data plane, environmental protocol information, all those things are available, and those are really available to anybody really. We get that information and we ingest that. Now where we have more secret sauce is around the hardware based telemetry. So as I was saying.25:59From the beginning, Cisco's developed our own ASICs, we've got cloud scale, we've got silicon one, but because we develop these ASICs, we build in specific functionality around telemetry. So we can actually stream ASIC level telemetry right from the hardware and we can ingest that and along with the software telemetry to really give you accurate information.26:19On flows, on your healthier environment. Now the great thing about having that hardware based telemetry is it comes from the ASIC, which means I don't have to burden my switch CPU with additional load. I can pull that off the ASIC, it gets streamed out of the port, and then we hit Nexus dashboard where we do all the processing.26:39So telemetry wise, it doesn't matter if you're running ACI or NXOS or even a campus VXLAN fabric, we can ingest all this telemetry. So we ingest it, we take it in the Nexus dashboard, and this is where the magic happens. So this is where we will do data enrichment. We can assign things like security groups or policies to it.27:00Better understand it. We can also augment it with.27:20DNS information, host information, even hypervisor information. So if I can integrate with vCenter or Kubernetes or openshift, I can apply additional metadata to these fabrics endpoints. So now I can correlate an IP address with a VM name or a MAC address. I can bring all that together to give me more con.27:23Actual information. The other really powerful thing that Nexus Dashboard is gonna do is correlation. So when you have a problem.27:43It's gonna be yes, NO. I have a fault or something's up or down, but data correlation is really something that nexus dashboard does really well. We can look at it as a whole system and say, if I had a route that went away and I lost an adjacency and LLDP seems to be flapping, I can correlate all these events and trace it back.27:51Back to the source problem, and that's one thing that Nexus dashboard will do. So we have the concept of anomalies or these problems we see, and our goal is we want to trace you back to what the source of that problem is. And we can actually show you in a little tree structure.28:11If you resolve this root correlated anomaly, you're gonna result resolve all these child anomalies that have been generated as part of that. And that really helps our operators really focus on what matters the most. Now when we get these insights and again, we call them insights or anomalies or actions we can respond to, they're very easy to read. I'm gonna show you.28:24An example in about a minute here. And then we can ingest those either by logging into Nexus Dashboards UI, if that's your preference or we can export this information out to you. And we can go via webhooks, we can integrate with ServiceNow, send it via syslog or we can even stream it via Kafka. So whatever you like to ingest it, even email if you want daily digest and how you.28:44Your network is operating, we can send these informations to you and then you can decide how you want to respond to it.29:04So what are some of the answers that we could help you answer or questions we could help you answer with Nexus dashboard? So one thing that we get, you know, asked a lot is, how do we know if a change is gonna impact my environment? Well, a lot of us who've been in the networking industry for a while, we typically would have a lab fabric or a lab environments. We test our change there and then if it is.29:12Successful, we roll it out to production.29:32Well, the only problem with that is that a lot of our changes that we do in our lab don't have the same inputs, right? We have different systems that are connected, there's different factors that we can't replicate in a lab environment all the time. So one of the really cool things we can do in Nexus dashboard is we can do a pre change analysis and we use your running.29:34Fabric.29:54To model what that change would do without actually touching it. So we ingest the config, we ingest the running state of that fabric, you take your change how you want to model it, and then we will tell you if you deploy this change, here's what's gonna happen. You know, if it was a problem or we're gonna exceed TCAM capacity or raise an anomaly because of it because we violated compliance rule.29:55Whatever it might be, we give you that pre check before you actually roll it out. And then if everything looks good and there's NO problems.30:15Boom, you can go ahead and deploy that change out, and now we get more confidence that your change will not have some undesired, some unknown foreseen issues that might be risen. Another question that we get is security group A should never talk to B So this is a compliance rule and compliance is a really big piece of next dashboard as well.30:22So especially around ACI where it's very security centric and we have security groups and endpoint groups, you can define rules saying I wanna make sure that this group never talks with this group of servers, right? They should never have an interaction. So you can define these, these compliance rules in Nexus dashboard, and then if somebody accidentally or.30:42Maliciously happens to violate that rule and let something communicate.31:03Will be made aware. You get known as the user, hey, this compliance rule is being violated. And I'll talk a bit more about that in a little bit. Lastly, from a reactive perspective, we have a really cool capability called explore and connectivity analysis. And these two features work hand in hand. So this is where, if I'm an operator, the 1st information I'm.31:07To get is, hey, I can't reach my web server. Okay, great. What's your host name? What's your VM name? What's your IP address? So we gather some information being operators, and then the 1st thing I can do is I can query Nexus dashboard. Hey, can A talk to B? And I can.31:27Run that way through the system and then it's gonna validate if there's policies and ACLs and all the routings there that would allow A to talk to B Well, that's great. I can test that in real time. The other huge value is I can roll that back to last week. I can go back up to 30 days.31:42So let's imagine I had a problem on Friday night, you know, everything was working fine. I come in Monday morning, now it's broken. So you can actually say, could A talk to B on Friday night, and you can, you know, query the system based on what it knew about the state of the environment at that time. Really, really powerful.31:58Now connectivity analysis takes it a step further. So explore looks from a policy perspective, is the routing there? Does everything look good? Now what connectivity analysis does is actually trace the path throughout the the environment. So I can say, here's my source, here's my destination, can A reach B And then we can actually.32:16Be set.32:36An Elam trigger through every hop of your fabric. So if you've not heard of VLAN, it's embedded logic analyzer module and it's almost like a virtual tripwire to see if we see certain packets going through our network. And this is something you can do on the CLI, but it's very tedious to configure. On Nexus dashboard, it's one checkbox. So it really just.32:37Points the path and then it will see if that IP match can hit the destination on any hop, and then that helps me understand where along my path is it being dropped or being prevented? Could be on a leaf, could be on a spine or anywhere in between.32:57So some other operations that you're gonna do day to day, obviously the 1st thing everyone wants to do being a networking team is we want to prove that the network is not the problem. So mean time to innocence, how quickly can I tell the operations folks, it's not me, go look at your applications. So this is something again that's really gonna help you expedite that process.33:13Maintenance is also something that we took very, very seriously as well. So we want to make sure that one, when you're doing upgrades, that you do them confidently, that you know everything's compatible, and that nothing's gonna cause a problem. We can also do a software upgrade analysis saying, here's all the potential vulnerabilities that are affecting your existing version.33:33And then you can upgrade and say, alright, I'm addressing these bugs.33:53What could I expect in the new version? Is there any vulnerabilities that I want to be aware of? So again, we can look at that and give you all that information without you have to go through release notes or a bug toolkit to understand what bugs are applicable and might be impacting your environment. Lastly, resource planning.33:57How do I know when I should, you know, be adding more links or increasing bandwidth from a hundred gig to 400 gig? So we can monitor your your environment and any time we see certain thresholds get crossed from a capacity perspective, we can make you aware. And again, we're gonna look at trending and that trending that I talked about in the beginning. If we're running fine, but once we cross our.34:13Thresholds, we're gonna predict and say, hey, you know, by the next month you're gonna see an oversaturation of some of your links and you may want to consider increasing your avenue that shouldn't capacity. So hopefully we can do that before these are causing any type of operational impact.34:33All right. I want to double click on anomalies. So anomalies are really letting Nexus dashboard do the homework for you. So if you think of a fault, it's a very binary yes, NO, and usually gives you a severity level, but anomalies, again, could be a deviation from a last NO good state. So we observe these through things like machine learning.34:49We know how you're operating and how you're functioning normally. Are you changing from that intended behavior?35:09We also have thresholds, so again, we've got different levels as well. So as certain things perhaps my, my V six table is getting, you know, very close to being exhausted. Well, we're gonna raise obviously different thresholds as those increments, so we'll give you a warning at a lower level. As you keep incrementing, we'll give you a major, and as you get close to the.35:1590 %, we're gonna make it very well known that you need to do something about the capacity there.35:35Now, not just the amount of change, but it's how quickly are we changing as well. So if we happen to see, say, a giant number of IPV four learns in the course of a very small period of time and that's not normal, we're gonna raise an anomaly. So that might be perfectly fine and it might be within capacity, but again, it's not normal, it's not.35:41Expected for this environment based on how we, we baseline it. So we look at things not only crossing thresholds, but how quickly is our rated change as well.36:01And then of course we have our forwarding and data plane anomalies we look at as well. Are we looking to seeing congestion? Are we seeing things like latency? And these are all mechanisms that we'll also use to raise anomalies. So all eyes are on anomalies, and this is what, again, we really focus in from Nexus dashboards analytics perspective.36:09So here's our anomaly dashboard inside Nexus dashboard. So you'll notice a couple of things. We 1st break it down by anomaly level, so if I want to zero in and just look at the critical ones, I can click on critical and it just shows me the ones that really need to see my, my attention. And then they're also categorized. So we have different categories of anomalies.36:29Some that are based on configuration, some that are based on, you know, connectivity and some are based on hardware. Perhaps I had a, you know, a hardware fan that blew out in one of my switches. Well, you'll get an anomaly for that happening as well.36:49And these allow you to not only group them together, but again, that correlation as well happens in the background. So as I have these issues that might be happening, they, you know, have a domino effect, they are gonna be correlated together to give the operator where to focus. So these anomalies, we can assign them to users as well. So if I'm an operator and I have a tier one and tier two.37:02New team, I can say, alright, I want to send all these hardware based anomalies to my operations, my ops team who's in the data center. And again, that way you can have management and dividing them up and have that full life cycle management of your anomalies.37:22So here's just a quick example of what a root anomaly would look like. So again, back to the domino effect, you're gonna see multiple issues in syslog. If I'm just looking at syslog events, I'm gonna see things like LDP flapping adjacency down, lost route, OSPF neighbor gone. All these things might point back and I might have an interface problem or a transceive.37:37Your issue. So we'll trace it out for you. You're gonna focus on just the root event or root anomaly.37:57And that's gonna show you all the child anomalies that are contributing to that impact. Now the other kind of alerting mechanism we have is something called advisories and advisories are all about bringing information to the user. So things like P Cert and bug notifications. So rather than you having.38:03Go check your, your version of software and your hardware for any known bugs, we bring them to you.38:23And we will tell you that your bugs are either susceptible, which means you're running the right version, but maybe you're not experiencing the bug yet or they can be flagged as active. Active means we've actually confirmed that bug is happening in your environment. And we do that by looking at the software version, hardware version, logging into the switch and looking at.38:29Logs and services. And if we see a signature, we can identify, we can accurately say yes, you're absolutely hitting this bug. It is an active bug that we want you to handle.38:49We also bring things to your attention like your advisories of when things go end of life. So if you have a large fleet of a bunch of different devices, you may not be aware when the end of life for that hardware is gonna happen. We bring those advisories to you as well. Same along the lines, we also support field notices. So if we had a bad hard drive or a firmware needed to be updated.38:58For a field notice that again would get brought to the attention of an advisory.39:18Last but not least, two really important pieces is we have best practice advisories. So if there's a way we could further secure your environment, maybe it's disabling insecure protocols that we found you had open using things like, you know, Telnet, which isn't secure. We can.39:23Bring up these best practice advisories and say these particular devices, we recommend you to harden them or, you know, do X Y, Z but we bring that information to you. All these advisories get updated every two weeks. So we produce a metadata file. If you have Nexus dashboard, we highly encourage you to tether it to the cloud, which means it talks to our inter site.39:40And that way we can automatically ingest and update our metadata file every couple of weeks.40:00Now, don't worry if you're not able to connect your Nexus dashboard, we also have an airgrapped air graph process as well. So you could grab the metadata file separately and then sideloaded it if you're in a, you know, air gapped environment or a secure facility that you don't have WAN access.40:05Upgrading. Upgrading is another big important event. And again, even if you're not doing anything fancy, you really just want to have control over your inventory of devices and then you want to manage your upgrades. Nexus dashboard's gonna help you do that reliably. So we'll do all your pre checks for you, make sure everything's compatible, and then we do something really, really critical.40:25So we do a bunch of checks before and after. And if you've ever done upgrades, I've always been in the data center when they're happening and I typically take note of my routing tables and my endpoint counts and whatever else, and then after the upgrade, I go back through that list again.40:44Do I have all the routes I expected or same number of endpoints still reachable, right? Well, now we can automate all that for you. So next dashboard will run these reports for you when you do your upgrade, give you that delta and if anything's missing or different, you're easily flagged and you can understand where and what has been changed before and after.41:00Now, when we get into optimizing your network, a very powerful feature inside Nexus dashboard is called traffic analytics. So this does a whole bunch of really cool things. So this is applicable for ACI and NXOS, and what we do is we look at all the traffic being sent across your environment. So again from source.41:20Offs the destination, we track any communication that happens between a pair of endpoints and we summarize it in what we call conversation. So if I have a web server and an application server and they're talking back and forth.41:40We're gonna categorize that as a conversation and then we measured on three really important metrics. We look at congestion, latency, and drops. And then based on that, we have a calculation that we will give you a traffic analytics score, and the leader be healthy or critical or any lower level if we need to give you that awareness. So helps you as the user to understand.41:53And what traffic in my environment is healthy or which one might be experiencing some type of congestion?42:13Now we do categorization automatically of your traffic. So we know about well known ports for TCP and UDP. So if I see TCP port 443I know that belongs in the web category, and if I see UDP 503I know that's gonna be DNS traffic. So we automatically break your traffic down into well known.42:20Categories.42:40We can also support custom categories. So if you have a purpose built or a custom application, and you can understand it has a unique TCP or UDP port and protocol, we can also create a custom category as well. So what I can do with that information is look in my environment, what's consuming the most amount of bandwidth, which ones might be consuming or.42:41You're experiencing the most amount of congestion and you get a lot of really powerful insights with this information.43:01Okay, couple other quick highlights that we recently introduced in our most current version, and I'm just gonna touch on a couple quick ones and I do want to save a couple minutes for a quick demo. Multi tenancy. So if you have a shared infrastructure for NXOS and you want to have multiple groups work on that environment, we now have the.43:09Ability to segregate your environment from a resource perspective. So even though they might share the underlying switches and devices, we can say these VLANs, these networks, these births, they belong to customer or tenant A.43:30And then these ones over here belong to tenant B So if you did wanna share your environment, you have a way to logically divide it up where you can keep resources completely isolated from the other. Now this is the concept if you've worked with ACI, we've had this concept for a while. Like I said, under that Nexus one umbrella, we're bringing everything together. So this concept that.43:42Has been around for a long time. We're just now extending it across both ACI and NXOS.44:03Live protect. This is another really exciting capability we've just introduced. So think of this as being a way to protect and mitigate your environment from a vulnerability without having to go through a full software update. So we recently acquired Isobilin and there's a Tetragon agent that we've deployed in all of our switches.44:10Which allows us to leverage embedded Berkeley packet filtering on the switches. So if there was a vulnerability, again, a Linux vulnerability, we can actually mitigate that vulnerability through our packet filtering on the switch. So that saves me from having to go do a emergency patch or upgrade on that device.44:30Nexus dashboard can deploy a shield, and that shield provides a set of countermeasures to mitigate that vulnerability and pretty much nullify. Now, that's not a permanent fix. It's a mitigation, but that's gonna buy me time until I get to my next planned maintenance window at which time I do a full blown upgrade and then that's gone away. But think of this as being a really quick.44:48Way to issue a very quick hotfix without having to reboot the switch, without having to do a full blown upgrade, and especially without having to do a revalidation of the software version because nothing else will change.45:08So this is coming to NXOS 1st. It came out with NXOS six one or 1061, sorry. And this will follow on as well come to ACI. Now ACI being a little more complex usually lags a little bit of behind NXOS, but we're planning to bring this out across our fleet of devices.45:20Oh, away. Another cool thing we've added in four.two is the support for doing catalyst telemetry. So in Nexus dashboard, one of the fabric templates we have is called campus VXLAN fabric. And this is something that we have, you know, had a bit of a gap for because Nexus, NO problem. We know them very well, but we're seeing more and more.45:41For catalyst.46:01Creep up in the, in the data center fabric space. So we wanted a way to be able to bring telemetry and ingest that from catalyst. So Catalyst is the next platform we're looking at ingesting telemetry from. We are also looking at additional platforms like ASR, NCS, and then eventually we want to go down the path to also ingest 3rd party telemetry as well. So making.46:02Access dashboard really be that true single pane of glass from an operations perspective.46:22Alright, I'm gonna go through a couple more here and we're almost at time here. Okay, another really cool thing I have to highlight. A lot of customers have invested in Splunk, right? Cisco owns Splunk, and this is a very powerful reporting and metrics collector, and one of the big advantages that Splunk gives us is it can collect data.46:29From multiple sources and vendors. So we can aggregate things across and then do cross solution and cross vendor correlation. So one of the cool things we've done is we've always supported sending data to Splunk, so if you had Splunk enterprise, you can ingest data from Nexus dashboard. We would export it.46:48Splunk Enterprise would ingest it, but with Splunk enterprise, you pay on how much data you ingest, and that's where the cost can really creep up. So what we've done with Nexus Dashboard is we've embedded a Splunk instance on Nexus dashboard. So now we've extended the data lake into your Splunk enterprise. So now if I wanted.47:06Do correlation or do reporting, I can do that without the expensive mechanism of having to export all that data from one place to another.47:26Now we can just access it from Nexus dashboard through Splunk enterprise. So that's what it's gonna look like inside of the Nexus dashboard. You'll have a Splunk widget there, and again that'll allow you to launch into the nexus dashboard and then right across launch into your enterprise instance, and again build very powerful dashboards and reports based on all the data we can now.47:34Ingest from Nexus dashboard.47:54Okay, I've got a little bit of time left here and I'm having too much fun here, but I do want to jump into a quick demo. I'm just gonna do something very simple. I'm gonna create a new fabric, bring in some switches just to show you how quick and easy we can do this here. So let me just kind of shift gears, and as I'm doing that, Shannon, is there any questions that we need to attend to?47:57I'll keep that as a NO. Sorry, I was speaking on mute. Could you revisit a little bit about the nash Nexus dashboard support for catalyst nine K in the data center in the.48:20What that means? Yeah, absolutely. So this is a Nexus dashboard instance I have running in my lab here, and I'm gonna go to my fabric view, and these are all the fabrics that I have currently built. Now, if I was building a new fabric, I'm gonna go here and create a new fabric, so you'll see a wizard here that we bring.48:36Up.48:56What I see here, the 1st question is it wants to know am I creating a new fabric or am I onboarding existing? So I decide, is it greenfield or is it gonna be brownfield? So in my case, I'm going to do a brownfield and you'll see here which devices we support for both NXOS, XE, XR, ok? So I can actually onboard Greenfield and brownfield.48:56Scale with those catalyst devices on both.49:17And this is something that Nexus Dashboard has done for quite some time. The new piece that we've added is the telemetry capability. So that traffic analytics that I talked about, that's now been extended to catalyst devices, which it wasn't previously. So we could always do the fabric management, we could onboard catalyst devices, we could manage the firmware, the software versions, but the telemetry.49:19The piece was something that we couldn't do before. So I'm gonna go ahead and create a fabric here. I'm gonna onboard existing. And again, we have all of our existing templates. Is it gonna be a VXLAN fabric? Is it gonna be classic land, your typical ag access? Is it external fabric which can also support 3rd party devices?49:39Or IP for fabric media. So I'm gonna do an existing VXLAN environment and I'm gonna use the enhanced data center cause all my virtual devices I'm gonna onboard happen to be 9000 devices. So I can see here that I can support those here. Now when I get into this part here, it's gonna ask me for a bunch of criteria. What do I want to call it? So let's give it a.49:57Fabric name, and this is just a friendly logical name.50:17And location, I'm gonna put mine in Toronto where I live, give it a BGP ASN, and then I have licensing. So if I wanted to do some of the higher tiered telemetry features, those do belong to the premier tier or if I want to just do very basic monitoring and very basic management, I can click keep it as essentials.50:22And that's what I'll do in my my example here. And then that's all it's gonna take, right? This is all the preview config. If I go back a quick step, these were the only fields that I was prompted for, not very many. If I do want to expose more of those what we call nerd knobs, I can go to an advanced view and then I have a lot more.50:42Capabilities and a lot more things I can I can tweak tinker with. I got things like fabric settings, underlay overlays, subnets, you name it. So if we don't want all that, we really just want to have something very basic, very simple. We can keep it simple and keep our default method here. Okay, so I'm ok with this. I'm gonna go ahead and submit. So now Nexus dashboard is not.51:02Pushing my devices yet, it's really just creating that placeholder or that logical structure for my fabric. So even if these devices were standalone devices or whatever purpose, could be out of band management or could be a manually built VXLAN environment.51:22We're just creating the construct or the placeholder for them. So now that it's there, I'm gonna go ahead and do the next step. I'm gonna add some switches to it. Well, how do I add switches to my fabric now? Well, I can either explicitly put in the management IPs because these are brownfield, they're already existing and I don't wanna, you know, change them. I can put in the.51:37IP addresses, you can use a range or you can use something called a seed IP. So imagine I had hundreds of switches that I wanted to onboard. What I could do is I can put in the seed IP and what that seed IP does is it discovers that switch and then it crawls outward using CDP and LLDP from there. So we'll go out.51:57And we'll see how many switches deep you want to explore. So the max hop would be how many switches away from my seed IP do I want to crawl out?52:17Now in my case here, I'm explicitly saying.fifty one and.fifty four. So I don't really need to have any hops because I'm putting in exactly the four switches that I want. I'll need some of my credentials here, so I'll put in my password and I'm gonna use these credentials for managing my devices.52:25Okay, now really importantly here, I have preserve config, so this means I've already got configuration on my switches, I don't wanna touch it. I want to leave it as is. Okay, so we're gonna go ahead and go discover. Now Nexus dashboard's gonna go out, call to those switches, try to hit those IPs, try to hit the username and password.52:45If I did everything correctly, hopefully I did. These will come back and saying yes, they are manageable. They are, you know, the credentials are good. I can manage these devices.53:05So I can pick which of those devices I want. I'm gonna pick all these guys here and you can see I've already got host names assigned to them. I can see the IPs, serial numbers, even the version of code that's running on them. Okay, now I'm clicking add switches, so now we're actually onboarding those switches. And again, we're just gonna pull some information from them. We're not.53:14Not gonna change anything and messed up the operations so this is not disruptive.53:34And then once it's done, I'll now have all the policies from those devices in Nexus dashboard. So we talked about taking the intent and then now making that a part of my golden config. So all the policies on my interfaces or OSPF or whatever other protocols I have configured, those now become Nexus dashboard policy.53:39And then I have a one to one mapping and I know that everything that is running there today is what my golden intent is.53:59So this will take a minute here cause again dashboard is creating all those policies. I'll let that finish off. Now while that's happening, I'm going to leave that finish off, and I'm gonna poke around a couple other spots really quickly. I talked about explore inside of Nexus dashboard. This is again is a great starting place if I'm doing any type of troubleshooting.54:08So let's imagine I'm looking for an endpoint and seeing if it's in my environment.54:28Well, I can start typing in an IP address, ten.and it's gonna go through and look through all the fabrics of all the devices that are onboarded, all the endpoints that it's learned are all now known to Nexus dashboard. So you can see here I'm getting a match here of all the devices. So if I.54:32Further drill down into it.54:52It's gonna narrow that down and right below it, it might be a little hard to see, but right below it, you're gonna see which fabric it learned that endpoint, which leaf it's on, which interface, and then which VLAN it's in. So right away I can say alright, well, show me this particular endpoint. That's the one I was looking for. What do you know about it? So here now it's brought up all this information on this.54:55Point, where it's connected, how long it's been up.55:15And if there were any anomalies that were affecting it, I could see them right here. So I'd be able to look really quickly and are there any problems with this endpoint, right? Endpoint history. Now this particular endpoint I'm looking at is an ESX host itself, it's not a VM, but Vmiss move around. We know the V motion and they bounce around. So if this was a VM, I would have.55:19But it's complete endpoint history here of where it moved and when. So if you're, you know, I had a problem with this at 02:00, I can roll back the clock and say, Yeah, show me the last 2 h or the last week of all its endpoint activity and where it moved around.55:39So very, very powerful. The other cool thing we can do, if I'm gonna go to one of my fabrics here, let me go to an ACI fabric, and remember I said we can do that can query, can A talk to B So now that I'm scoped just the one fabric, I wanna know.55:52Can, and here's where I can type in normal pro norm normal english. Can endpoints, and again I can either type it in or I can select it, and it's building out that query. So the two endpoints that it knows about on this fabric, there's not many, are these two here. I've got one oh one and.fifteen.56:09So I'm gonna say, can this endpoint talk to anyone? Can he talk to a specific VLAN? Can he talk to a specific endpoint group? Or maybe I wanna see, can he talk to another endpoint? So now I wanna see if he can talk to, let's go.56:25Don't have too many exciting endpoints, we'll do this guy here. So now it's gonna do a query in saying, can I talk to B? And if they can, we get a green check mark. So right away I've proven yes, these two endpoints can talk to each other from a policy perspective. And if I scroll down, I can see which verse or tenants they belong to.56:43Where they're connected, what the topology looks like, and then how they talk, I can see again destination EPG information, if they had any forwarding rules or what verfs or subnets are being allowed to communicate, all that information comes to me as the user. So super super powerful.57:02All right, I'm just gonna bounce back really quickly to my fabric to show you it's all built. So now here's all some information on my, on my fabric. There's not much happening cause it's pretty, pretty empty, but I'll see my switches will be listed here. And again.57:20I can take a look at them, I can go deeper and explore them, look at the interfaces, look at any policies, but all that information is now brought to, to my, my, my console here. Okay, I've gotta switch gears back really quickly to close this off here in time.57:38I'm gonna talk really quickly on licensing. Last couple slides here. So when you have Nexus dashboard, there's NO additional cost. And again, outside of the physical hardware, if I had, you know, a physical appliance, you obviously have to purchase the appliance, but from a licensing perspective, all your switches are already licensed. If they have an essential.57:55Tools advantage or premier license on the switch, that entitles them to be used on Nexus dashboard. So when we license our our fabrics, we will set it according to whatever switch license is currently deployed on those devices. Now this is all soft enforced, it's not hardware enforced yet. So again, if you only had essentials and you want.58:15To play with some of the premier features, you can turn out to Premiere without any problems.58:35So again, second last slide here, just be aware you're licensing, if you're purchasing a subscription, you're already entitled. So you're paying for nexus dashboard. We absolutely want you to deploy it, play within the lab, make sure you see value, and then if you're happy, obviously you want to deploy it. And our goal is to have you look at all the features. Try out connectivity.58:41Try out traffic analytics. Make sure you see that value. And if you're happy and you need to bump your licenses, then, you know, we're even happier. But even if you're using essentials only, there's a lot of things we can do for you. You can fully manage your devices and will give you a base level of telemetry. So we absolutely want you to take advantage of your investment that you've already made on your.59:02The Nexus 9000 switches.59:21All right, with that, there are a few other resources here that, you know, we will share with you at the end of the session and I'm sure Shannon might have a couple of closing remarks, so with that, I'll say thank you very much everybody, and I'll turn it back over to Shannon. Great. Thank you very much, Robert. That was excellent. I did post some links.59:25Into the chat, feel free to grab those. This session will be posted on our community page, but you also have the link. If you guys happen to be at Cisco Live, I would definitely join any of our sessions. You are one of our best. So with that, have a wonderful day.59:44And we will connect with you guys later. We will be planning more webinars like this. Thank you. All right, thanks everyone.01:00:04