Nautobot 3.0, 3.1, and Beyond

Hello, everyone! Welcome! We are really excited to have you all joining us here today, wherever you happen to be in the world. As you’ll notice, there’s a poll popping up there on the screen for you, so please do give us an answer there on that poll here. We’re going to wait about 2 minutes for everyone to get joined in. It always does take a minute to get the room filled up. And, we are super excited to see so many of you out there. We have the largest attendance we have ever had on a webinar, so we’re really excited about that.

It’s pretty cool to see, like, literally hundreds of you piling in, so amazing. Super cool. As we wait here, we’re just going to wait a couple more minutes. For those of you that have just piled in, because they’re really ticking in fast here, please do answer that webinar poll for us. We’ll get back to that, and we’ll look at the results later on during the webinar as we move through. So with me today, we’re going to have some exciting information, and I’ve got my colleague and friend. You’ll get a little sneak peek for those of you that are here early.

I’ve got Brian O’Sullivan with me today, and one of the things I’m kind of curious about, and I get this question a lot, which is, where did the name Nautobot come from? And I’m not sure how many of you know the answer to that, or even if Brian knows the answer to that, but it was actually a project that Jason started way back before Nautobot as we know it today existed, and what it stands for is Network Automation Bot. Or Network Automation Robot, right? That’s where the name came from, it was consolidated into Nautobot, and when Nautobot became a thing, that name was available, and we had the domain already, so that’s how we got Nautobot.

That’s where it came from. A little bit of trivia for those of you that maybe didn’t know that history from the name, because we get it a lot. Where’d that name come from? Why do we call it Nautobot? There you go. Now you know. All right, let’s see. We’ve got quite a few people here in the room, amazing, love it. And I think it looks like it’s stabilized, so we’re going to go ahead and get into our webinar for today.

So let’s go ahead and get started. We’re gonna move into our beginnings here, so Nautobot 3.0 and 3.1, the future of Nautobot, how we can unlock new capabilities to empower your network automation journey. So, with me today, as I mentioned earlier, I do have a colleague and friend. I’ll let him introduce himself.

Hey, thank you, Tim. This is Brian O’Sullivan. I’m responsible for product management here at Network to Code, and I did know about the naming of Nautobot, just for the formal record. Yes, although I confess, I did think it was, like, a nautical thing. for a few days until I figured out what it stood for. But yeah, happy to be here, really excited to, to talk about 3.0 and 3.1 with folks.

Amazing. And for those of you that may not know me, my name is Tim Schreyack, and I am the Director of Strategic Transformation here at Network to Code. So, for our agenda today, we’re going to look at the Nautobot 3.0 features, then we’re going to roll into those Nautobot 3.1 features, we’ll do a demo of Nautobot 3.0 stuff. Of course, always time for a Q&A, and then we’ll do some wrap-up. Quick housekeeping items while we’re here. One, you’ll notice that the chat in the webinar is disabled, so please do use the Q&A box. That’s the best way that we can make sure that we get your questions answered, and we can share them with the group as well.

And sometimes, especially on a large webinar like this, we’re probably going to get so many questions, we’re not going to have time to answer them all. But don’t worry, just fire those questions into the Q&A, we will take note of them, and we will get you an answer after the webinar if we don’t get to you while we’re here live on the webinar. But we do our best to get through as many as we can.

But I’m pretty sure we’re gonna have so many, we’re not gonna get to them all. One of the questions we get all the time on every single webinar is, will this be recorded, and will it be available later? The answer is yes, it will be. So, don’t worry about that, you’ll absolutely be able to watch this later. Give it to your friends, your colleagues, whoever. One more thing. Down on the bottom of the screen, wherever your taskbar is here in Zoom, you’ll notice there’s a reactions button. That is really helpful for us if you use that. If there are things that you’re lighting up on, go ahead and react.

Give us a smiley face, whatever it is, because that can be really helpful for us to understand what’s interesting to you, what you’re enjoying, what’s of value to you as we go through the webinar here today. So please do use that as we move through. Okay, so let’s just do a quick Nautobot overview as we think about what the platform is. For maybe those of you that are new here to Nautobot, or, you know, you’re just beginning to understand what it does.

There are two major functional areas within Nautobot, and it is an open source platform, so at its core, we have that source of truth, specifically network source of truth, and an automation platform. It gives you a structured way to model intended network state, so automation operates from trusted data. You can execute automation directly from Nautobot, or even integrate it with your existing stack, and Nautobot apps provide you a fast on-ramp for common automation workflows. That’s the foundation that we’re enabling so that we’ll have newer capabilities that we’re going to walk through today as we look at the rest of the platform.

So those are the basic concepts as we think of Nautobot itself. Now, Nautobot deployment options. This always comes up. How do I operate Nautobot beyond just what it does? So, Nautobot can be deployed in two primary ways. On-premises, or you can consume it through our SaaS platform, Nautobot Cloud.

And while many organizations are still choosing the on-prem to meet specific requirements, Nautobot Cloud really simplifies adoption by removing that platform maintenance, enabling streamlined deployments, and making integration with other NTC services, such as NautobotGPT, much easier. So now, let’s start as we think of Nautobot 3.0, right? We’re going to move into what is coming next here as we look at these features and things that were released in the Nautobot 3.0 feature set. But before we do that. Let’s take a quick look at those poll results that you all had back there at the beginning.

Really curious to see what they were. And this is always kind of interesting for us, because depending on the audience that we get here, you know, we get… we see different things here. So we’ve got a large group of you that are running Nautobot in production today, so welcome to all of our, already fans, if you will. And then we’ve got actually a pretty significant group that’s either exploring or testing, as well as new here. So, for all of the newcomers here, also, welcome into the group.

I hope that you’re able to get some really great insights into all things Nautobot and where we’re going with the platform. So absolutely, great information there. Thanks, everyone, for taking a minute to answer that poll for us. Okay, and with that, we’re going to move over into Nautobot 3.0, which I’m going to hand off to Brian for him to tell us all about all these great features that came with it.

All right, thank you, sir. And yeah, absolutely. I think one of the main things that, we’ve kind of worked on, and are working on in, In Nautobot is working on the user experience, and improving what’s there, and making sure that is thematically one of the main areas that we’re focusing on within the system. The most obvious thing from this for 3.0 is that the change in the menus themselves. So, in 2.x, we had what was called an accordion navigation. We’ve moved to a fly-out menu, so the things will kind of move out from the right.

This means that they’re always in the exact same position. on the left, within the menu bar, so it just… it just really makes it nice and clean to be able to click exactly where you want in the system. You can scroll a lot less, you can get much deeper in the menu. move faster, move fewer clicks, have a streamlined workflow as you’re viewing the entire system. So this is probably the most obvious change that we made with 3.0. That’ll jump out at anybody who’s used to the system.

Although, I’ll say, as Tim pointed out, there’s a lot of folks on this who maybe aren’t as used to the system, so I guess welcome to the only menu system you’ll have known, which is the 3.0 system. But this is a main area we kind of added within this solution. as I mentioned, this is the top strategic and focal point that we’re working on, is improving that user experience, so it’s not just the menu itself. There’s a lot of other areas that we have added. Some of them are kind of listed here. Up on the top left, we, we improved the search. It’s streamlined, improved.

It’s always there up at the top, so you can always see search. It’s always available. It has, it has, keywords for you to be able to use, like, in devices there, so you can scope for precisely where you’re searching. You can use CTRL-K or CMD-K to be able to bring it up, depending on your system, and that’ll instantly bring it up there. If you want to, to have that at your fingertips throughout the, the product, no matter where you are. Filtering is also an area.

So, a lot of times when people talk search, they talk filter immediately afterwards. That was certainly an area we looked to improve. Part of that was to say, okay, how can we quickly see context? That’s also a flyout. In this case, it’s a fly-out from the right. Quickly see context, move through it fast, ensure that we can apply filters at will, and get it really, really quickly and easily, so you can determine precisely what it is you want to see. We’ve added the ability to collapse panels, shown in the bottom left. That gives you control over real estate.

If there’s issue… things you don’t want to see in panels, they’re not germane to what you’re looking at, you can completely customize that, collapse them as you will. Which brings them kind of out of view of your immediate view. But of course, they’re always there. They’re just collapsed. You can instantly bring them back. And then finally, at the end, kind of show the old and the new, is breadcrumb navigation.

Really just want to make it just pop as to what was there, and what was the most important bits that you’re searching for and focused on. In this case. In the example, we’re clicking through to a device, and it’s really obvious what that device name is. It’s in bold, it’s much bigger, much clearer than it was in the older system. It’s still got all the breadcrumbing, it’s still got all the hyperlinks there. If you want to dance back to EMEA, you can absolutely do so, but just to really make it prettier and cleaner. As to, as to what was there.

In terms of modeling, this is something, Tim had this on one of the earlier slides, we have a ton of models, we’ll continue to add models to the system, and this is part of Nautobot Core. One of the first things we added was, in 3.0, was the ability to model VPNs, specifically GRE and IPSec. For any of these kind of point-to-point VPNs. So, it is dealing with, transport mode, tunnel mode, you can do up and spoke. But this is a comprehensive view of what is there, and a comprehensive ability to model those components. I mentioned IPSec.

What that means is you can do, both Ike and IPSec. So I said IPSec because, of course, it’s just normally called that, but includes that initial Ike phase, as well as the IPSec phase. All of that can be modeled. You can model the endpoints, you can model whatever the secrets are within it. And that just allows you to ensure that we are modeling whatever customers have in their real-life networks. And make sure that we’re keeping abreast of what is there in their environments. All right, from a load balancer’s perspective, again, continuing along, it’s a new model, adding that in. the focus here is to have the structure for what is needed in order to correctly model a load balancer.

So it’s the bits listed here on the left. From virtual servers, to members of the pod, to pools, to certificates, all of that is available so that you can model that correctly. You can also monitor and do health checks on this. So, I have a pool set up. is everything responding within that pool?

So this is a feature supported by load balancers themselves as part of their pooling, and we made sure to model that and ensure that we have a consistent way to get this across whatever kind of vendors you may have. F5, A10, Citrix, what have you. All of these can be modeled within the system and make sure there is that kind of standards-based modeling of, again, your network and your network behavior.

I think this is one of the really cool, powerful features that came in Nautobot 3.0. Particularly, we would get this a lot from our clients of how can I model my load balancers and my VIPs? Now you have a vendor-neutral way in which you can manage them, monitor them, set them up, configure them, so whether you’re using a mix of load balancers in your environment, which probably a lot of you are if you’re operating large environments, you probably have more than one vendor in the mix. Now you have just one easy way to configure, to design, and have a simple way in which you now manage all of your VIPs from one place. It’s a really, really powerful feature.

Yeah, and you know, that’s… it’s a great point, and I think this is one reason why we do try to reach out to customers and ensure that we know exactly what… where are the issues they’re seeing in their network, and what we can model. And that’s why it’s really crucial to get that feedback from customers and users, and as Tim mentioned, we heard that loud and clear with load balancers and made sure that was within the system. All right, so for approval workflows, This is a way to have a multi-staged review before the actual execution goes on within the system.

So, the way it works, and we’ve got kind of the personas listed here, whether it’s a workflow architect, or the operations engineer, or the approvers, whoever they might be, they each have a different role to play in a modern network environment. They can set up what the… they can have different roles within these, I should say. The workflow that goes through, the architect defines what is the definition, what is the workflow I want to do? Operations engineer can then execute that, and then it goes on to an approver.

And the approver is… can be one person, can be a group of people, but you can define precisely who is able to approve this before it actually executes within the system itself. So that means you can segregate who is allowed to do these things, you can have Maybe someone’s a little more junior, actually set up your operations, be ready to execute, but when they hit go, it doesn’t actually immediately go out to the network, it’ll go into an approver queue. And again, this is if you choose to set it up this way, if you want to take advantage of the approval workflow system itself.

And this is supported for both, scheduled, which is as we dance into right here, this is supported for scheduled as well as immediate jobs. So as we go through this, you are able to say, hey, I want to support this for… You know, I’m executing this now, but it’s something that will happen a day later. That’s absolutely fine. It goes into the queue. We kind of listed out some of the functionality here, and some examples of what would be there.

I can define in my definition what is the name, the content, I can do constraints on it, I can do a weight to determine its relative priority, among other things. And then I can have stages listed on the right. So as this goes through, it’ll go through those stages, could get to a denial message, could get to the approver, but depending on what goes, it will go through these steps before it is actually executed within the network itself.

Hopefully you’re seeing the, really, the power of this feature as it comes into the platform. Nautobot jobs in and of themselves are remarkably powerful, and one of the downsides to that is they’re remarkably powerful. It means you could easily do something wrong. Now, with the approval workflow in there, we can really put some good business practice around that amazing capability the job gives you, and now you can have that security of knowing that something isn’t going to access accidentally kick off without going through all the proper approval chains.

So this is a really amazing feature, I think, from an operational perspective, really allowing you to take the power of Nautobot and make sure that it conforms to your business process.

Yeah, and again, just when you talk about the business process, again, this is why it is great to hear from customers and hear what there is, because that’s how we are able to model this correctly and ensure that we are matching your business process. But, so this is showing a screenshot of what’s there, so you can see the workflow itself, you can see the scope to a type, the constraints, all of this is the bits I had talked about, actually sort of shown here, and all of this is completely customizable and definable, however would meet your needs within the network and your operations team’s needs.

If you think this is cool, hopefully you do. We are going to dive into this further in the demo, so please, we’ll go more there and fire away questions at us. We’ve got a few folks there handling the Q&A, so anything you’ve got, just put them in the Q&A, we’ll get those answers for you. I do see a couple hands up in the attendees list as well. If you have a question, just go ahead and put it in the Q&A, we’ll get you an answer.

Alright, so talking about, applications, Ansible automation. This is something, again, can’t stress enough how this is built upon, you know, feedback from customers and users of the system. What this allows you to do is execute those automations from within Nautobot. So what we’re not doing is we’re not saying, okay, we’re going to be taking over Ansible’s things. We’re calling out into Ansible. We’re discovering what templates are there, what jobs are there. This could be from AAP, the Ansible Automation Platform. It could be from AWX.

Whatever system you have set up in your environment, we’re going to be able to reach out to that via APIs. surface that within Nautobot itself. The thinking behind this and the purpose behind this is just to simplify, again, the user experience. We want to focus on ensuring that it’s easy and simple for you to see these things. execute them as you need. This way, you’re not swivel-chairing between the two products. So, if you happen to be using Ansible in whatever kind of form, we feel that it is a really useful thing there.

It’ll understand the data models within Nautobot, understanding the relationships, inventory, etc. That means you can ensure that you’re inputting the right data to get into those Ansible automations so that everything is executed seamlessly inside your network itself. What’s kind of running here? This is the example again. Two systems, we’re not kind of taking over what it is they’re doing, but using APIs to call into it. So, we would have credentials for the user, in this case, AWX, go through, determine what the permissions are to run it.

We are using the APIs, as I mentioned, that’s secure communication. as part of that integration with HTTPS, And then going out and executing that. So it’ll be dispatched by the job template itself. Ansible Controller, the application within our system, is going to be able to read that, and then as it flows through, it will eventually execute within Ansible, in this example, AWX.

The cool features, I think, that Nautobot has always brought to the table was this layer add. or value add, if you will, to Ansible. And we’re just taking that one step further and really removing that friction to using Ansible and your source of truth. So now you can just feed that source of truth data into your Ansible playbooks really seamlessly and easily, just really further enhancing the contribution between these two different platforms.

They really do serve different purposes. But together, they’re really, powerful together, as well as really just making the life of your operators easier, that we don’t have to do copy-paste or move back and forth between these screens, as Brian mentioned. So, amazing to be able to do these things all now in one place.

Salute, buddy. Yeah, so here’s the screenshot of it, showing what’s there, and again, it’s exactly as Tim was just saying, we want to ensure that we can make it easier and simpler, improve the user experience by allowing you to do that all in one area here. So I can see what the templates are that are integrated, now they’re here, it’s Nautobot jobs, I can then execute them within Ansible itself, and there’s a play so I can just sort of Execute it immediately right there. Keep it all simple and easy right within the one screen.

Okay, with that, we have, wrapped up our Nautobot 3.0 presentation portion of the talk here today. We’re gonna move into the Nautobot 3.1 sneak peek, things that are coming as we build on that foundation built by Nautobot 3.0. And don’t worry, we do have that demo coming up for you. We’re going to talk about all of those… we’re going to show some of those 3.0 features in that demo. But before we got to that, just wanted to make sure you could see what’s coming next as we look to 3.1, which will be released later this year. So, Brian, I’ll let you continue talking about that.

All right. Yeah, we are very excited about 3.0. It’s absolutely out now, but we’re also really excited about what’s coming, to Tim’s point, and we will talk about some of the things that are, on deck for 3.1. The first one is operational compliance. So what this is, is allowing you to validate outcome of network changes. And it’s doing that by looking at device state. In this case, when we say operational commands, we’re referring to really show commands within the device itself. So you can compare and contrast between results before and after a change has been made.

And that’s kind of shown here. You can run pre-checks. Those pre-checks, again, could be however you define your operational commands to be. You can go through and execute a change, and then run your post checks afterwards. See at a quick glance, you know, because you’ve been able to set things, and we’re showing down here. what the values are pre and post. You can have a comparison, and there’s a bunch of things listed.

We listed a handful here. the one that’s yelling at you no, and the third one down is an exact match, because for whatever reason, it hasn’t been an exact match, but there’s a bunch of things you can do. You can do a max, you can do a tolerance, you can make sure parameters are matching, so just ensure that Your state has remained either exactly the same or close enough, whatever close enough means. In these two pre- and post-checks. again, and I harp on this quite a bit, but it’s intended to simplify operations and to make it easier and simpler for you to use these things.

So, previously, if you had to do it, and this is going to dovetail to our next talk, you make some significant change You want to ensure that that change has worked properly, and this is a fantastic way, we feel, to be able to do that.

Yeah, I mean, I think this feature, I wish I had had this when I was a network engineer. This is really letting you codify your business process and just make it easy. Because, I mean, we all know the one thing that network engineers love doing is repetitive tasks, right? Like, we just love doing that thing over and over again. It’s absolutely not, right? What ends up happening is we’re supposed to do these pre-checks and these post-checks when we do our changes, right? But we all know we don’t do it. Because it’s hard, it takes time, it gets in the way, we’re trying to get the project done, we’ve got a deadline we need to meet.

Now, there’s just an easy button to getting and making sure we follow a proper process. Because, I mean, I think pretty much every network engineer has this story of, like, hey, I was making this change, I should have run that post-check afterwards, and I, yeah, I skipped it, somebody called me, I forgot about it, whatever, I didn’t do it, and then I got bit by it later.

Probably I got paged because I was the one on call, and, you know, then I had a problem to deal with as a result of that. This is just… you can avoid those problems, I can make sure my changes go as they should, I can capture things before they become a problem. So, amazing feature. Start using this when we release it, it’s going to be amazing. We’re going to show you a cool way in which we are leveraging it here in the platform, but you can leverage this any way you want in your workflow.

So, think about that, how you might do it beyond what we’re enabling for you here in the platform as well. I’m gonna let Brian talk about the next one, which I also really think is amazing, but we’ll move into that next here with OS Upgrades.

Yeah, I think it really does dovetail well with this, you know, as again, we talked about changes and what changes you are enacting in your network environment. Well, certainly one of the big ones is upgrading an operating system. And what OS Upgrades allows you to do, OS Upgrades the application, allows you to, again, quickly, simply, easily upgrade your operating systems across your network devices. This is… I mean, again, I almost certainly not have to tell this group what, what headaches it can cause if there are issues with network upgrades.

Their estimations, I think. There was a study from Cisco, it said 80% of network infrastructure could be running older systems. You have to ensure that everything is up to date. it’s absolutely a problem that everyone has to ensure that everything is running correctly and running the latest, or the proper operating systems they should be. So that’s what… exactly what we’re trying to do here with OS Upgrades. Now, if we talk about it in relation to operational compliance, again, this is one of the main changes you’re making.

It’s even better when you can use operational compliance in conjunction with OS upgrades to ensure that you’re checking state both before and after, and have that pre- and post-check With, with OS upgrades, and that’s… Probably one of the main areas we would expect people, to, to utilize this, but as Tim said, use your imagination with operational compliance. There’s plenty of areas you can utilize it, but I would certainly expect OS upgrades would be an obvious one, where people would say, boy, I just made a massive change, I upgraded the entire operating system, let me ensure that everything is really going, or if I’m in some kind of weird state.

I’m personally really a huge fan of this application. In fact, it exists in large part because of my personal efforts to bring it to life and push for and promote it within Network to Code that we should invest in this. having a vendor-neutral way in which you can manage your upgrades, you’re not locked into any given vendor, into their ecosystem, for their upgrade process. You have a centralized way to do it, you can do it with good practice. We have clients, a lot of clients at NTC are extremely large enterprise, and when I say extremely large, I mean they have hundreds of thousands of network devices on their networks that they manage through Nautobot.

And they are using this application to manage their upgrades today. So this is a rock-solid app that we have actually already in deployment, so definitely, when it comes out and available to the general public, you’re going to want to check it out. It’s one of our best new apps, I think, and I think that it will become one of our most popular, right up there with Golden Config, which is, of course, our number one most popular app.

But I’ll hand back to Brian here for carrying on before I get off my cheerleader soapbox.

No, it’s exactly correct, and I think, you know, just building what he’s saying, you know, with this slide here, if you have 5 devices, you can do things manually, right? That’s not a concern. We are dealing in Nautobot, and our customers are dealing with significantly large networks. And I think this is one of the reasons, and I’ve been talking a lot about user experience, and I’ll certainly continue to talk about user experience, but one of the reasons I haven’t really focused on scale as of yet is that’s been built into Nautobot for years.

And so we have customers who are utilizing the system at scale, and I think this is an example here where, you know, showing, okay, I’ve only got a handful of engineers, I have to do manually, I have to do it across a thousand devices, and believe me, we have customers with a lot more than a thousand devices. You know, then if you look at, okay, what do I get from automated? and click on to the next one there, where I can show how it’s much better from an automated perspective. That just slams things down, right?

Because we can ensure that I don’t have to have 5 users working on it, 5 engineers working on it, I can have a smaller staff doing it, I can do it in a much compressed timeline. I can still do things. We’re not saying please go out and upgrade 1,000 devices, you know, on a Tuesday. I actually don’t recommend that if you only have 1,000, but don’t upgrade everything in one go, but you can still do things where parcel it out into regions, into sections, into zones, so you can ensure that you have backup, and you have capability, but it’s still vastly faster.

And again, harping on it, but with the pre- and post-checks, you can ensure that it’s worked properly, and that everything has gone well within the upgrades themselves on the individual devices. Alright, so, We’ve been talking a lot about what to do with the devices within Nautobot, but how do you know what devices are there to be put into Nautobot? And that’s one of the things we’re kind of adding with Network Discovery. This is a complementary thing to device onboarding.

So device onboarding is there, it’s been there, you’ll absolutely be able to get the detailed device onboarding. That, that is already within Nautobot. What this is doing is saying, okay. what is out there? How can I scan the network and ensure that I’m looking at it? When I say scan, I’m talking SNMP, I’m talking SSH. Determine what devices are out there in what order. Are any of them missing from Nautobot, or any ones that are out there, whether they’re perhaps a rogue device or something that is unmanaged or unexpected that you don’t know is there.

But running this and having the ability to run this allows you to ensure that, again, you are as closely in sync that you, you, should be, again, not about us acting as a source of truth. We want to ensure that we are able to keep up with whatever is in the network and bring that in under management. and onboard those devices. So it allows you to do this. It retains metadata to be able to see when a device was first seen, when it comes up, how often it comes up.

Does it come up intermittently, which might be a security issue if something is popping in and out, for example. So you can investigate this, and then, again, hopefully the answer is it’s a thing that’s been added, we need to onboard into Nautobot, that’s the best case scenario. But otherwise, you’re still able to see what’s out there, do that scanning, and determine. what to do with it next. So, at the beginning, Tim was talking about the kind of deployment options of Nautobot, and whether it is self-managed, or whether it is using Nautobot Cloud.

This is a functionality for Nautobot Cloud, so this is not technically a Nautobot thing in and of itself. This is a, we’re calling Cloud Secure Proxy. That is part of 3.1. What it is, is simplifying connectivity in a secure fashion with the networks to, again, manage those devices, whether it’s onboarding, whether it’s managing them day-to-day, to ensure that you have the connectivity between the on-premise networks and Nautobot Cloud. You can absolutely do that today. Of course, Nautobot Cloud supports connecting to it.

What this is doing is adding the ability to proxy securely from within the network. And as to why you would do this, there’s a handful of reasons. One of them is you may, want to have fewer holes in your firewall. I shouldn’t say holes, but, you know, connections in your firewall, you say, look, I know the proxy is connecting through this SSL tunnel, that’s connecting back, therefore I only have that one connection to that network. Itself, and there could be 10,000 devices in the network, and you’re still connecting it with one proxy communication.

Another, and kind of shown here on the right, is there’s IP overlap for some reason. We have some customers who have acquired companies, they happen to be using the exact same IP addressing. For whatever reason, you may have something where Nautobot needs to talk to device A and device B, And both of them have exactly the same IP address, but they’re in different networks. Secure Proxy helps with that immensely, because it’s actually connecting to the network securely.

Nautobot says, aha, I’m going to device A, I need to go through this tunnel to get back to it. I’m going to device B, it’s a different tunnel. So even though it’s the exact same IP, it’s still able to handle that. perfectly. And once that tunnel’s up, and that’s a call home from Secure Proxy, which is living with inside the customer network, back to Nautobot Cloud, that tunnel is up. It’ll remain up, it’s a bi-directional tunnel for communication, so you can ensure that that is always on.

Yeah, I think I’ve spoken to a lot of you that have expressed interest in Nautobot Cloud, but the one hurdle was that, hey, how do I connect my environment into the Nautobot Cloud infrastructure? And the answer was, I had to use an IPSec tunnel, or maybe I used Direct Connect via AWS, and you would say, hey, that’s not really gonna work for me. Well, now we’ve got an answer. Right now, you don’t need to worry about any of that. We install this nice little appliance, and off we are, up and running, able to make that connectivity and deal with a lot of problems as well, that are a result of having sort of these disconnected networks.

So, this is the easy button to using Nautobot Cloud. If this was a problem for you, maybe what held you back from using Nautobot Cloud, we’ve got you covered. This is going to come and answer that problem, so definitely look for this Becoming here very soon with our new release of 3.1.

That’s right.

And Brian’s gonna wrap us up. A few more other cool things. Sorry about that, Brian.

Yeah, no, no, I stepped all over you, my bad. There is a little bit more we’ll talk about from an AI perspective. This is something we released NautobotGPT in May of last year, I believe, and we’re continuing to extend that out and expand what we’re doing from an AI solution. So a couple things listed here, just really quickly. The first thing listed there, that note about data integration, Right now, NautobotGPT answers great questions about Nautobot itself. But not about your Nautobot. So if you want to ask questions about how do I create a job, or what’s the documentation for this, or how to create a user, all that’s fine.

What you can’t say is, how many devices do I have in location X? Because currently, NautobotGPT does not talk to your Nautobot. Those first two bullets there are answering that question. One is using an agentic component, we talked about the data integration, to be able to have an agent sync into your Nautobot, and then have that communication, and know about things that are going on in your Nautobot environment.

For some customers who don’t want to do that, we’re also adding an MCP server. This is, model context protocol, allow you to, to communicate with it. It’s an API layer for AI. That allows you to communicate… it means you’re using your own inference engine, not our backend. But it still allows you to take advantage of the AI solution, number one. And number two, get integration into your Nautobot. And to ask, and I just picked how many devices in a location, but that’s… the tip of the iceberg as to what you can ask.

Again, trying to make things easier and simpler for you to use, and focusing on user experience, ensuring that you have A way to use natural language to get any kind of querying that you look to do within the system.

And the final bit there is IDE integration. So, we’ve talked about VS Code here, the vast majority of folks we talk to are using VS Code, but ensuring that if you’re doing job creation and you’re utilizing this, you can have this already within your IDE.

So connect your IDE VS Code in this example, back to NautobotGPT, have that power at your fingertips to be able to do it as you’re writing These, these jobs and, and, and generating any of this code that you would be doing.

Brian, while we’re here on this slide, we’ve got a couple of questions. They’re all related around the same thing, so before we leave this slide, I’ll just see if we can get those answered real quick, which is, first of all, MCP Server, will that be available for our on-prem clients? And then, as well as, in general, these features, will they be available as part of the open source, or some other way?

Yeah, great question. So the first one is, absolutely, MCP Server is available for on-premise customers. One of the things right now, and that’s why I talked about the inference engine, we have, with NautobotGPT, our own inference engine on the back end, the data is going to us. With the MCP server, absolutely, if you’re an on-premises customer, you’re able to utilize that, utilize that with your own inference engine. The data never flows to us at all. It’s completely controlled by you. We’re giving you the MCP server to allow that integration with it.

Building on that, and that kind of second question, it is part of the commercial offering. So, in general, this is part of our commercial offerings, this is not an open source component. And is, you know, something not only for Nautobot Cloud, but also, as I mentioned, for on-premises commercial customers.

Yeah, so as you think about using Nautobot as you move forward, definitely keep that in mind. These features will become available via subscription and or your Nautobot support contracts, so definitely look at that, speak to your account reps as far as if you’re interested in taking advantage of these cool new features that are coming. And if you don’t have Nautobot support, or are signed up for that, maybe it’s time to do that. So, that’s the… but that’ll be the end of my sales talk here, don’t worry about that. Let’s move off into the demo portion of the webinar here today.

We’re going to take a quick little peek at what we can do using Nautobot 3.0, and as I mentioned, we’re going to look at that workflow definition component that we saw earlier. So let’s just kind of take a look at what that looks like here in Nautobot Now, right? So we now have this idea of approvals and the workflow definitions, which, if you’ve been playing around in the Nautobot 3.0 sandbox, or maybe you have it in your environment, you may have seen and wondered, what is that? What does it do?

Well, here you go. So we can see here, as we look at this, scheduled, job approval workflow that we built here in the demo environment, what we’re looking at is, you know, the definition of this workflow. So we have its name, you know, what does it do, so we understand what it does. It’s model content type, and this is kind of interesting, right? This is what it’s attached to, essentially, right? This is where we are applying this particular workflow approval, and what it’s going to reflect against. So, this is where we can now tie the approval to basically any object or thing, if you will, in Nautobot.

So, in this example, we said, hey, we’re going to apply it to the definition of a scheduled job. Some constraints. Here we can define, whether it should apply to a particular action, right? So we want to say, hey, we want to use this particular workflow and only apply it to this particular action. So maybe it doesn’t apply to all the things. And then finally, really important as well, is this idea of weight.

This is how we’re saying, hey, if there are multiple definitions that might happen to apply, this is how you build that tree of the order in which they are applied, right? So, highest is going to be the prioritized one, right? So the higher, the higher priority. So that would be how you would organize your workflow definitions as you think of how they work. Now, over here on stages, these represent the groups that have to approve the request before it can be carried out, right? So, essentially, the number of approval levels, if you will, right?

In this case, we set up a single approval level. In a lot of production environments, you probably have two approval levels. You know, you would have your peer or, you know, peer reviewer, and then maybe a manager or, you know, senior level reviewer, for example. Really common scenarios that you might do. So, probably in production, you might have more than one than one. Or maybe it’s depending on the type of workflow.

I think that’s also really cool here, this flexibility, where you can say, hey, these are pretty simple changes, we just allow for one level of reviews, and these more complex changes, maybe we say we have to have two levels of reviews, for example. So you have a lot of flexibility how you might use that there. Now, let’s go back and actually just kind of take a look at those definitions again, and let’s create one, and do it here, and see how it actually functions here in the demo. So we’re going to start off with an example, and let’s just create a webinar test approval workflow.

And to make things easy, and so I can avoid making silly typing mistakes, I am going to copy-paste, so as we go through there, right? We’re going to choose that, we’re going to apply it to our scheduled jobs, and then we’re going to put in our constraints. Now, we’re doing something pretty similar to the one that you saw already in there, here, for the demo purposes, but again, we’re just applying this particular workflow and constraining it to, this particular job class, right?

We’re going to give it a weight of 100, because you might have noticed that other one was set to 50, so we want this one to take priority. And we’re going to set, you know, our approval levels, right? So here, let’s just have 1. We’ll set our approval to approver 1, we’re going to say we need a minimum of 1, and our denial message will be that we failed the approval. Finally, the group, here that we’re going to apply to, and in this case, we’re going to choose the scheduled job approver. So that’s pretty much it, that we need to do in order to create that definition.

And as with all things in Nautobot, right? Of course, you’ve got your notes and your tags, which you can use as you wish, depending on what makes sense for your environment. So there we go. We’ve created that workflow definition, and now we have it available for us for consumption. So, let’s go actually give it a try. So what we did was, under the jobs section here, we created a scheduled jobs, and we’re going to use that to actually give this a run.

So we’re going to go down to this export jobs object list, which is just a simple job right here, just for purposes of demonstration. That maybe someone needs to go run, and they’re going to run it against our, circuits provider, right, for example, perhaps. Since that is what, I think what we selected, when we were, back on the prior screen. So we’re going to say, hey, we want to run this job, we’re going to run it against our circuits providers, and let me see, I think that is all that I need to put in this screen, and we’re going to say go ahead and run the job now.

What we’re going to see as we click that run there is, it was successfully submitted for approval, right? So, if you’re familiar with Nautobot Jobs, what you would have expected to have happen is the job kicks off, because I had that run immediately. But now, instead of it kicking off, we’re going to send it off for approvals. So if we go back into our approval dashboard now, we’ll be able to see those approvals, and they’re going to be sitting over here under My Requests, because I sent it in as a request, right?

So here we can see, again, you know, what I have out there for reviews. I can maybe take a look at, you know, what I need to do, or what’s actually in that approval. You know, here’s an example as we drill in, right? So now we can see what is it, who asked for it, what do I need to do? You know, in this case, I’m logged in as admin, so, you know, you’re seeing admin here, but of course, you know, you would see, you know, whoever actually submitted this in a normal production environment. And we go through, like, you know, hey, when is this gonna run?

What does it do? And of course, we can even drill further in, right? So you’re saying, hey, there’s the job that it’s going to do, but I’m not really familiar with that job, I don’t exactly know what it does, no worries, just go ahead and click on that job, it’ll take you right over into the job infrastructure, right? And then we can see, you know, what is that… what is a particular job actually going to do? So I can approve or not approve with some knowledge, right?

So let’s just go back into our approval dashboard, now that we’ve maybe investigated and said, hey, this is cool, I like it, it’s going to be good to go, right? So we’ll go back into that screen there, and now we’re able to say, hit, okay, I’m one of the approvals here, I need to approve this, so we’re just going to hit that little green checkbox over there. I’m going to put in any notes here, and I can make whatever comments I might wish to do of, like, hey, this is approved, or what have you, right? It’s going to log that I did it. So, you know, you’ll get all that information there as well, so you don’t need to worry about putting your name or anything like that.

We’ll have that as there as part of the record for this change from now in our historical records, right? Once we hit that approve button, it’s going to then fire the job once all appropriate approvals have been given. So, if there was a second level of approval, it would then move on to that next queue and wait for approval in there before finally moving into the execution stage.

But since we only had one level of approval, if we go into our jobs now, we will see that that… oh, I clicked on jobs instead of job results. If we go into our job results, we will then see that that export job list just kicked off and went through all of its regular process. Of course, you can even look in and see, like, you know, what was that flow, right? So then we can see they have that historical record as well, as far as what happened. So we have that record there, and it will continue to reside under our approval dashboard as well, where we can see that historical record in there as well, as we think of, you know, things that have been approved or not approved.

We have that history there as well for us. So, great way to enforce some process around how your changes occur. This allows you to potentially offer up things that you might have been hesitant to let maybe your Tier 1 or your Help Desk do, because you’re like, I’m not sure if they’ll do it correctly. Well, now you can enable them, and now we can just flow right through to appropriate approvals for changes.

So, just making your life easier and better as you move forward into business process. A really key element to not only being able to have your source of truth and your network automation platform, but how do I integrate that into my process so that it works for my organization? So, great features as we have here. All right. Now, we intentionally left a few minutes here at the end of the session because we knew that there were going to be a ton of questions. And as expected, there are. So, we are going to start working our way through these. We’re gonna take a peek at them.

We’ve got about 15 minutes here to go. So, let’s see here what we’re gonna, fire off with. Brian’s gonna come off mute. I’m gonna pitch questions at him, and we’ll see how he does. It’s gonna be like watching him try to play Fruit Ninja, I think. We’re gonna see how many he can slice and dice. So, alright, first coming up from Brandon here, this one, I get it all the time, so we’re just gonna fire right away, which is, I heard later this year, within regards to 3.1, what does that mean?

Yes, first off, I’d like to thank Tim for not giving a precise date, just because, you know, I want to make sure we hit it properly. No, it’s imminent, so we’re looking at releasing this in April, so we’ll have this out the door in April. And and yeah, you’ll be able to take advantage of OS upgrades and operational compliance and a ton of other stuff that is coming in 3.1, and… I guess wait for the 3-1 webinar. We’ll go into a lot of details then and give a demo at that point. But yeah, coming in just a few months.

Yeah, really soon. And for those of you, maybe, that are thinking of questions as we go through here, please do fire them into the Q&A. We do have some time here, so we’ll try to get to as many of them as we can. Next up, from Jason, what version of AAP is supported with the app? So, thinking of that Ansible integration, what… is there a particular minimum version that we have to use in order to, take advantage of that?

Yeah, it’s a great question, and honestly, I don’t know that off the top of my head. There are particular versions that we have tested with, and I would have to say we’d have to kind of follow up with you on the specifics of it. Certainly, we’ll look to ensure that’s published in the documentation, I’ll say that much.

Absolutely. When it’s released, you will have that public information, and Jason, of course, always reach out to your account team. We can get you direct answers as well. We’ll reach back into the engineering team and get you an answer. Let’s see, is there a plan to have a real job runner which does require access to Redis and Postgres? I’m actually not sure where this question’s headed. Do you have an idea there, Brian?

I… I… well, it could be in a couple things. It could be from a job orchestration perspective, and dealing with that, and, you know, multiple bits, that you could stitch together, and multiple jobs you could stitch together into a single thing.

But it might be something different. I would say, perhaps, if the… if the person who did that could… could elaborate a little more. Yeah, you could just give us… Yeah, if you could give us a follow-up there on what you meant. I think perhaps where you were going with is not about going to become an orchestration engine, and I think the short answer to that is no, right? Like, there’s no reason for us to go build an orchestration engine. There are lots of good ones out there. AAP is one of them. Why would we reinvent the wheel of something that’s a well-established protocol, right?

So, the answer to that would probably be a solid no. But we do… are enabling the soft features, right? Like job chaining. these capabilities that you have within the platform to make your life easier. So for simple orchestration, you can do it within Nautobot. But for complex orchestration. Use an orchestrator, that’s what they’re built for.

Yeah, and that’s also why we built out the Ansible Automation app, to ensure we could integrate with those orchestration components as well.

There’s a question here about custom fields. So, I have a lot of custom fields I need to see in places like the interface list screen, etc. And this pushes This makes it difficult, basically, to see. Is it possible to move those buttons around, on the screen? And so, yeah, this is something from a UX perspective, I think, and I know we are doing a lot of work there. Brian, feel free.

Yeah, I think… so the short answer is there’s no changes in 3-1. I’ll put it that way. This is something we’re looking at, quite actively right now. I had two calls this morning on, with our UX designer and working on things. So it is an area I think we can look at, and I’m couching it that way because I’m really talking, you know, end of year, next year kind of timing, as we’re kind of looking at some additional UX improvements. But it is a valid question, and I think one of the things, one of the things I’ll say from a filter perspective, in 3.0, the apply filter is now always at the bottom.

And that was something in 2.x, it wasn’t. It would have to require you to scroll off-screen. I know you’re describing custom fields and not filters, but it was the same kind of problem in that you had to go to it. So that is a thing we’re looking at doing and making sure we can say, okay, how can we make this simpler and easier to do?

Not something in 3.1, definitely, but it is something we kind of look at a little further on, and certainly those kinds of user experience things are right at the forefront of improvements we’d like to make.

Yeah. Next question is from Todd about the OS upgrade application, and his example is, let’s say, suppose I have to do 100 devices. Are we gonna do those upgrades concurrently, in series, linearly? Because he’s kind of curious about, like, hey, I’ve got a limited maintenance window, how many can I get done in that maintenance window?

Yeah, so we’re trying to leave it as much up to the user as we can, so in… to do it… to have the ability to do it serially, or sorry, parallel… in parallel, as opposed to necessarily just serially. But yes, that is something we’re looking to kind of build in. We’re still not exactly GA with this yet, so it is still ongoing development, but it is something we’re looking to ensure we’ll get in. Whether it’s going to be in 3-1 or later is not exactly 100% set as of yet.

Now here’s one. Is there a version of 3.x that has been identified as long-term support, or LTS? Given that we are currently at 3.0, I would say probably no, but I’ll leave that to Brian to give the correct answer.

No, you’re absolutely right. There is actually a webpage we put up on this, kind of with our terms and, you know, what we mean by GA versus what we mean by GA LTM. For folks who are maybe newer to Nautobot, that means long-term maintenance, so it’s a thing that we will have engineering support for much longer periods of time. That’s a minimum of 2 years support, for, for engineering fixes, and right now, on the 2X version, it is 2.4.

So we’re continuously rolling out, fixes on the 2.4 branch. We haven’t determined it yet for 3.0, and we quite frankly won’t until we get much, much closer to 4.0. It’ll be the last release of 3.x, will be the LTM. what that one is, we honestly don’t know. It’s 100% not 3-3, or earlier. Is it going to be 3-4, 3-5, 3-6?

We don’t actually know yet, but we’re continuing to come out with the releases, and one of the things we’ve We’ve done… Is ensure there is this regular cadence of maintenance releases, excuse me, of minor releases, to ensure that we continuously have you know, branches that are updated and stable and continuously updated with bug fixes and security fixes. So 3.1 will come out, and then 3.2, and then 3.3 on that regular cadence, and then we’ll get to the LTM a little bit further down the pike.

Another one here about the MCP server. Like, how is it actually going to be delivered to our clients? Is it a black box package? Are they free to tweak it or make adjustments to it? You know, understanding it’s not open source, right? But, like, how is it actually delivered, and can I see the code?

Yeah, we are looking to have it as simple as we can, and there’s Jeff purposely, wonderfully sharing this software lifecycle policy, which is published. Thank you, sir. We are looking to make it as simple and easy to deploy as we can, and therefore would be more of a, you know, a single box that would be deployed. I’m a little hesitant on the, can we modify it? I think maybe let’s have a conversation about that, just because then it’s hard to ensure that everything is working as planned, if it has been modified, especially for the API layer.

But, so it should just be just a simple drop-in that we’re looking to do in the 3.1 timeframe. You can drop it into your network, and then… have at it. You register what your inference engine is, you register the components, and then it handles the rest. But mostly black box, I would say. I mean, let’s maybe have a conversation with the team about what you’re looking to do from a customization perspective.

Yeah, absolutely. And as I mentioned several times, as with all things, for further detailed questions, please reach out to your account team. We can always get you in touch with the right folks to get you answers there.

Yeah, and you know, one thing, and I hadn’t actually talked about this, is the MCP server is designed to be, to allow it to expand. So, what I mean by that is if, you know, we’ll come out with more data models in 3.1, we’ll come out with more data models in 3.2, I’m just picking data models as a thing, but you won’t necessarily… MCP, as a deployed system is going into Nautobot, querying its APIs, and then utilizing that So, as you upgrade, it will continue to support everything there.

So, that is certainly comprehensive. For custom apps, I think that’s an area where certainly we can have a conversation on how those APIs are exposed. But yes, Tim is exactly right. Talk with the team, and we can get you more detailed information.

Yeah. So now we’ve got one that’s kind of an interesting question, and it really starts to show some of the gray area, if you will, between these tools as we think to the future, between, say, Nautobot, the source of truth, and then maybe my NMS. Can Nautobot track port state last change? So, does not have that awareness of, you know, the state… the port was up or the port was down? So that it and the NMS are in sync. You know, an example might be, say, hey, we have an edge port, it’s in a down-shut state, but we want to track this, you know, across our environment.

Now, this is one of those challenging spaces, right? Because Nautobot as its intention is where we define how the network should look, not what it actually looks like, right? It’s our authoritative source of truth. But there is this practical application to some things, for example, host-level IPs, so slash 32s, as well as port state. These are interesting spaces where maybe there’s some gray area, and we’ll see if Brian has any thoughts here. If not, this will definitely be a follow-up conversation with the team.

Yeah, I mean, honestly, Tim is being modest. He’s much more of an expert on precisely how this would be done than I am on this, but yeah, it is a challenge and a question, and I’ll be honest, one of the things from a product perspective is you do want to be careful as to the ephemerality of the data you’re storing. And when you’re talking about port state, that can be extremely ephemeral. Those things can pop in and out of existence quite quickly, as opposed to You know, devices are a little bit more… more stable. Yeah, I would say, I would say that to have a further conversation on that, as to how best to implement that within the solution.

Yeah, absolutely. And always keeping in mind, Nautobot’s in place and purpose in life is to describe what that port should look like. So Nautobot says the port should be shut, not that it actually is, right? That would be an enforcement level. You can do enforcement with Nautobot, but Nautobot’s purpose in life is to say what the environment should look like, not necessarily what it actually looks like. That’s for your network monitoring system, right? Yep. I’ve got a few more here, I’m gonna try to roll through some of them.

Let’s see. Can we constrain groups within the approval workflows? So maybe we want to say, hey, some things trigger certain workflows, others don’t. Short answer, I believe, here is the platform is extremely flexible in how you might apply these things, and how you might choose, you know, where you use a workflow versus where you don’t, how it triggers versus doesn’t. You saw there, you know, I was able to do constraints, for example, so you could use those constraints to limit what it can and can’t do.

Or, you know, when it’s applied versus when it’s not applied, so you have a lot of flexibility. This is… Nautobot, in a nutshell, is when you ask a question, can it do… name the thing? We will say, yes, it can do that, right? Because it’s really designed to be super flexible for you. So, yeah, super easy one there. Nautobot Cloud. Will we ever get FedRAMP compliance? Huh, great question.

The answer is not this year. And therefore, let’s have a conversation and talk about that. I’ve done FedRAMP at a previous company with a previous product. It is a fairly involved solution, and so I want to ensure that we have enough customer demand to really make that requirement stick. So currently, there are no plans. It’s not a thing that’s going to be live. 2026. It is a thing we’ve talked about and bandied about.

I have gotten this from a handful of customers, but again, let’s have a conversation, talk with the account team, and let’s see what that would entail, because we do want to make sure we have enough customers to drive that demand there. It is pretty comprehensive to be able to do that.

We had a question from Ionis here about, hey, kind of seeing Nautobot and realizing that Nautobot’s role in life is really all about operations, and how do I take data and use it to drive automation in my environment? And I think the question really is leading into, like, what happened to that? You know, we just used to be the documentation purpose in life. And I’ll just give you a quick synopsis on that. When have you ever found a source of truth that its sole purpose in life was documentation, that it was accurate? And I would argue, never.

And the reason is, we don’t do documentation as network engineers. It just, flat out, we don’t do it. So therefore, Nautobot, what it does is solves that problem by changing the way you think of data. Data is no longer documentation. Data is the way you instantiate change. And that is a key philosophy behind Nautobot and the way that we build it. So, you think of Nautobot as the way that I go to do things, not the place that I go to document things, necessarily.

Now, what happens when you think this way, hey, I go to do things in Nautobot, all of a sudden, documentation becomes a natural fallout from my natural work. So now, I don’t need to worry about documentation, because I just do it naturally as part of my process. That’s really a philosophical change, and that’s a core component of the way Nautobot operates, and all of the things you see us building, they’re operational first, because that’s what we do in the real world. We don’t do documentation, so let’s stop trying to pretend that we do. So that’s really the core of it.

Hopefully that gives you a good answer there, Ionis. That’s one of the things I love to talk about. If you ever hear me at a conference talk, that’s probably what I’m going to talk about, because it is the number one thing that we do, right? And it’s really core to how we do it. So thanks for that question. You got me a chance to put on my conference talk hat and get on my soapbox and yell at you all about how you should do it. So yes, thank you for that one. Alright, we’ve got one minute left.

I’m going to try to hit one more question here that we’ve got in the queue, Oh, so this is a great one here. So, thinking long-term, is Django going to stick around? Are we moving away from Django?

We are not. So, yes, the short answer is yes, we are doing it. There is actually, we haven’t discussed this, we’ll talk about this a lot more once 3.1 is live. There is an upgrade coming to Django 5.2 in 3.1. Nautobot, and that is, is it within Nautobot Core? So yes, the short answer is absolutely, there’s no change to change the infro with that, and we’ll continue to upgrade it, continue to support it, and make sure that it, it continues along with the system.

And with that, we have reached the top of the hour. Thank you, everyone, for joining us on this really cool webinar, and also for being part of the largest webinar that we’ve ever had here at NTC. Thank you for helping us make that milestone. Look forward to seeing you on the next one.