Hello, and welcome to our Strategy Fabric podcast for this month. My name is Darren Fulwell. I'm product evangelist for IP Fabric. I'm joined today by Graham, who'll be helping me break down how network assurance helps manage service providers improve the service that they offer to their customers. Graeme, can you give us a quick introduction?

Yeah. Certainly. Hi, everyone. Yeah. So as Darren says, my name is Graeme Hogsworth.

I am IP Fabric's director of strategic partner development. Now what that means is, IP Fabric is typical of a a a youngish growing company that you go through this life cycle where your founders discover they can solve a problem, and we'll get back to that because that's very pertinent to this conversation. Very much. They they discover they can solve a problem. You create a product.

You start selling a product. You start selling that product through partners. And at some point, it reaches a stage where you bring someone in with kind of the experience that I have working with partners in in different environments to mature how you work with partners and start looking at different models. Now getting involved in the podcast today is it's something I'm particularly pleased to do because from the moment I first saw IP Fabric prior to my first interview with the company, the immediate sort of flashing light in my head was this has got managed services written all over it. Yeah.

Now I'm not surprised that we haven't done it up until now. It it it takes a, you know, it takes a leap away from that original business model. For a lot of organizations, it can involve commercial changes, product changes, for us less so, which is good. But, you know, it's it's really really good to be able to talk about this now because it feels like this is the first time we're we're we're vocalizing this publicly, but it's something that's certainly been in my head for the last 8 or 9 months. And I know, Darren, you and I have had had a lot of conversations, and we started talking to some MSPs about this.

Yeah. I was just gonna say. So so I guess the beauty of where we are right now is those conversations are happening, and we're and we're speaking with all manner of different managed service providers. Right? We're not just talking about, the traditional Internet service providers and those those people who who deliver connectivity, but obviously the ones that get really interested are almost the enterprise folk who are who are delivering management of of the the end to end networks, I suppose, for their customers.

So, you know, what kind of insights is that giving us, Graham? Where do we start? Yes. Well, the the the first thing I learned is that I was ridiculously overthinking this, which which, frankly, is not the first time in my career in life I've heard this. So started speaking to a few managed service providers with fantastic ideas for and you'll notice you you've been involved in the demos and and you've looped this light in your career.

That's true enough. True enough. Remediate response was like, woah. Woah. Woah.

Woah. Woah. Woah. Woah. Woah.

Backpedal. Let's start with the very basics. When we take on a new customer, just discovering what it is we're gonna be managing is a feat in itself. And we've had responses from everything from well, we just kind of negotiate with the customer and sign up to to an agreed guess at what the network looks like, or we spend months months months scrabbling around the network pulling on pieces of cable to discover how it hangs together. Now very quickly, to go back to the founder story of the company, we were founded by 3 network consultants who were doing that very task.

They had to go out and manually map the network. It was after a couple of days presumably in a bar, I certainly would be by then, that they had the surely you could automate this conversation. And, you know, here we are 8 years later as IP Fabric doing just that. So, you know, you have the experience in it in this just discovering what it is you're taking on with a customer. How bad a task is this?

You see, this this is something that you have with every network you walk into. Right? And there's and I'm sure, you know, if there are network engineers listening to this, they will be sitting here nodding in in agreement. That that you you step into a new an organization where you've never seen their infrastructure before. You go to documentation.

That documentation is always patchy at best. Right? What you'll have is you'll have certain parts of the network where you'll you'll have great documentation because it's maybe just been implemented or maybe, and a systems integrator has come come in and built you some new stuff or whatever. But the bits over over the other side where you've got best tribal knowledge of of certain aspects of what's going on, and the rest is just a mystery. These are the ones that you run into problems with straightaway because you can't get that oversight of what why the network looks the way it is because you don't even know what it looks like.

And, yeah, the very first step when you walk in and any consultancy engagement from my perspective was tell me what's there. Let's let me see what's there. Let me let me get a feel for how it's put together. I'm guessing as well for a from a a managed service perspective, the very first thing you're trying to do is onboard it. Right?

Onboard that network. And without that baseline, you're not gonna have a clue what that looks like. Yeah. I I was speaking with a a small UK based managed service provider who it would be great to get on one of these conversations soon. And I was saying to them, well, you know, are the networks that you're taking on to manage, are they that poorly documented before you take them on?

And his answer which really struck a chord with me was, if they were accurately accurately documented, they wouldn't need us. Yeah. Yeah. You know, if you've got high control of your network. From a from a small network perspective, absolutely.

And I think this is this is the point. Right? You know, you've got small networks that are badly documented, and that's difficult enough. What about the big ones? Right?

I mean, there's you know, you you've got multiple sites connected with a WAN. You've got, you know, data centers. You've got cloud infrastructure, all of these things that you're trying to piece together. When I first started networking, it was simple. Right?

We had we had a load of machines in a room that were all cabled together, and that was your network. And we used to use here's a story. Right? So we used to use 3 tools to to to actually deliver the that map management. We'd have spreadsheets, right, in order to track which ports were connected to what and which IP addresses were used and those kinds of things.

We would use, a diagramming tool to show us how things were connected together, and we'd use, like, an SNMP monitoring tool to do real simple kind of just making sure things were up. Right? Today's networks, we have, as I said, we've got multiple campus environments. You've got wireless, you've got SD WAN connecting over your Internet, but also your your private WAN. You've got data centers here, there, and everywhere.

You've got cloud connectivity, all those things. What tools do we use, do you reckon, Graham, to to manage that? Should I tell you? Go on. Spreadsheets to track IP addresses and ports and the like.

Diagramming tools, physio, SNMP monitoring, same tools we used back in the 19 nineties on these really simple networks. Different and better way of doing it. So that's why we have to look a different and better way of doing it. How better to do that than through a managed service provider? Yeah.

And for the managed service provider, I guess it comes back to, you know, way way back in the midst of time, you know, Henry Ford identified that if you're gonna do anything at scale, It's gotta be process led, and you've got to automate automate everything you possibly can. It feels like it's still a very, very manual process for people. And I'm guessing it's a manual process that you're having to use very let's be honest, expensive, well qualified people for. Yeah. I mean, that's really, really important.

The the the discovery itself is not necessarily particularly difficult. It's you still gotta have an understanding and and whatever. And and it is you know, you talked about the the consultant going in and and looking at a network for the first time. That consultant needs a a certain level of of network understanding in order to be able to to to get that that view of what the network is doing. I guess this is the point at which we introduce what IP fabric's doing here.

Right? Because it's from a network assurance perspective, what we what we're doing is taking that same knowledge and understanding and using it to automate the process of discovery, but then not just do it once, but but do it over and over creating these snapshots of the network so you can see how things are changing over time as well. And and it's and it's really like you say. It's it's using that same knowledge and understanding, but codifying it and automating the process. That that means that we discover the network in the same way that that highly skilled individual would would go ahead and do it, who might be being you you might be better charging them out a 1,000 quid a day to actually go do some some proper consultancy rather than the grunt work effectively of of the discovery and the assessment.

And I you know, that that's a very, very good point because throughout my career, whenever I've spoke to anybody about anything automation, there are always people that will throw their hands up and kinda go, does that not mean half my workforce are gonna be made redundant? Well, I think in this case, from what I'm seeing is actually these highly qualified, highly skilled, brilliant network engineers. They love it because they can actually get on and do more interesting tasks. Yeah. Exactly.

Exactly. No one wants to to be doing that grunt work. No one because it is that. It is it's oh, man. The number of times I've sat in front of a in front of a console with a pen and paper and and logged into a device, found the information about that device I need, then logged into the next device, and to the next one, and to the next one.

And had to, you know, may I could tell you stories for days on that. So, if if you know If you're listening to this rather than watching, you just missed the face of a scarred man who does this party mix up. One of the things he was saying a few minutes ago was we work on this basis of you take snapshot of the network and that tells you what your network looks like. And you mentioned that you take snapshots and you take snapshots again and again and again and again. Yeah.

Yeah. So once you've gone beyond this, okay, initial discovery, we know what the network looks like, our customer knows what the network looks like, We're taking over management of that network now. Yeah. What are the benefits of just keep snapshotting over and over and over again? Well, for one thing, having the accurate view, time after time when changes are being made.

Right? So, again, here's that scarred face again. There's if you ask network engineers about documentation, it's a necessity. It's a necessity, not even a necessity Because, of course, how can you manage an infrastructure without knowing what's there? Right?

But keeping up to date is probably the biggest pain in the backside that any network engineer faces from from day to day. There's the the the manual work that's involved in that is time consuming. Right? What does the one thing that a network engineer doesn't have, especially when they're in a change window in the middle of the night doing the work that they need to to maintain the network. They don't have time.

And so it always gets put to the back of the queue of things that that that need to be done and very often just completely slips off the radar. So being able to to automatically create something that replaces the need to document, because I I wouldn't describe it as documentation. What what what you've got here is effectively a database. It's much more detailed than than than normal documentation would be. You've got that understanding of how the network hangs together, but you also got a lot of information about the state of the network at that point in time that you wouldn't normally have in documentation.

So you but you are creating this this this ability to, to to have this this dataset that's always kept up to date, which means that then when you're doing any activity on the network, you've always got something that you can trust represents the state of the network as it was working at the point when that snapshot was created, which is invaluable. You you if you're making a change, you you got a known good state before that change was made. You can then create another snapshot after the change and do a comparison and say, right. Actually, is everything that was working before working afterwards? That that kind of thing.

You know? It's it's a complete revelation. Yeah. Now I'm I'm probably oversimplifying up to this point because in my mind, I'm still thinking customer, manage service provider, manage service provider, manages the customer's network. It's not that simple as networks get larger.

You have you have different telco carriers, you have wide area network providers, you have cloud providers. I've seen it where you can have, you know, multiple managed service providers managing different bits of the network. How does IP Fabric help when you have this sort of mixed management, everyone responsible for a different bit, but they all need to play nicely together. Well, it's that's a really good one. I mean, they and it's not just different parts of the network, but sometimes different vendors of of the network equipment.

Right? You might have, you know, subject matter experts in particular, vendors, environments, that kind of thing. You might have a security overlay that you need to be concerned about as well. But the point with what IP fabric is trying to achieve is is an oversight of the whole thing, right, to to give you the view of everything from that. That user is connecting out at the wireless access point right the way through to the application workload that's sitting in the cloud and everything in between.

It's where the real value comes in it. And and so what we're able to do is pull together data from it, all of those environments to give you that oversight. It can be used in all kinds of ways, but but, you know, let's say you want to validate that a service provider is giving you the the service you were expecting. You know, the how how do you prove that? Will you prove that by pulling together that data with everything that's around it and making sure that all of the interactions between those different networks are as you need them to be?

And I think from a from an MSP perspective, what that allows them to do is is almost share the the management of the of the network with whoever else, but also give them the oversight to understand how they're interacting with everyone else who supports the network. That's got to help. The the I hate to serve customer stickiness, but I'm gonna use it anyway. That has to help with the customer stickiness. If you if you're the person that's providing that that single source of truth or single source of observed truth that everybody is relying on, You know, you you've gone you've gone from being useful to essential, which is always a good place to be.

Totally. And and, you know, the the the customer can choose to take, to move from one vendor to another for a certain part of the network or choose to go from one provider to another for a different part of the network, doesn't matter. So long as you've got that same access to the to the network, IP fabric will just read through from that and and just swap out one bar of the network for another because of the way the discovery works. We can we can track that change as it happens. So I guess we we've gone from, hey.

You know, the my my conversation with the managed service provider, which was, woah. Woah. Woah. Woah. Why this back start at the beginning?

We've kind of gone through the gathering of information, so we've got a a common view of what it is we're gonna be managing when we take this network on. We've kind of gone through it's a great way of of kind of saving, as you say, a lot of the manual grunt workers changes are made and tracking the changes and managing multiple vendors. You're starting to talk about security and and compliance and things like that. Where do we how can we help if you've got a managed service provider who's saying, well, it's great. I'm providing these managed network services.

It would be great if I could do more, differentiate myself, provide additional value add services, maybe approach the market or customer from a different way. Yeah. Where else can we play with this sort of No. I mean, this is this is this is perfect. Right?

So so and and I guess this is where, IP fabric really comes into its own. Right? Because because by being able to effectively replace the manual documentation process, what you've done is you've just dropped in, a a new level of data and insight into existing process. So the things that the the service provider already does, incident management, you know, adding new capability into network, you know, making ads moves changes, that kind of stuff. You've you've just improved those process sim by simply swapping out, an existing capability for for an improved one.

So which which means there's there's minimal impact into the way that they you as a service provider delivers their service because they they haven't had to change the way they they actually deliver into the customer. Network that that they never had before. Right? So they then and you and I'll we'll talk about this in stages. You mentioned compliance there.

Right? There's the ability now to start measuring, how closely the network matches with an intent, which which, of course, unless you've got some sort of automated process, you you wouldn't have that capability. But by being able to do that, you can now start to do effectively audit. Instead of having to to do it year on year for your PCI compliance or whatever, you can do it day on day. You can understand exactly what the network's doing day to day and say, where's my, you know, where am I not compliant with either my regulatory standards or even my own operational standards?

You know, where's my consistency? Being able to see that day to day means that remediation activity might be a half hour change rather than a 6 month remediation project. So it completely changes that that understanding of what the network is doing at a point in time. But what that then leads you to is saying, well, I've got all this data, I've got all this insight, and I've got all this reporting. This is fantastic stuff.

How can I now adapt the way that I deliver, deliver that service so that I'm using much more succinct processes, things that don't need human interaction now perhaps? So it might automate the whole thing, start to look at providing more self-service, start to look at those kinds of ideas because you have a means by which you can validate the outcome of those things across the whole network. And here's the thing. Right? So a lot of people talk about network automation.

They talk about network automation as being automated config change. Well, that's great. You can you can push a config change to a device, and you can check that that change has happened. That's all you can do. With something like IP fabric there, with a network assurance capability, what you're doing is you're looking at, has that change impacted the behavior of the network as a whole?

Is that actually delivering the outcome I'm looking for? Not is that device now configured to do the thing it needs to do. A network is a distributed system, and it needs to be treated as such. And so that's where this this then introduces a whole new level of capability. So this is recognizing the difference between kind of working and working how I want it to work, which are 2 different things.

Completely. What what do you mean by working? Right? So this is where you you think of traditionally what a monitoring platform does. What it what it will do is effectively ping a device and go, are are you responding?

It's not saying, is that device doing the thing I want it to do? It's literally probing at certain pieces of data and saying, is it does it look like it's possibly doing the thing that I want it to do? Whereas if you look at the PIC network as a whole, you can actually identify, yes, that device as part of this particular application flow, let's say, is forwarding traffic the way I want it to, is is configured correctly, is behaving the way I need it to, and that's that's the difference. So there's a tremendous amount of capability there, And, obviously, every customer network is different. There's gonna be different vendors and, different providers.

Oh, yes. If you turn around a 180 degrees and look back into the managed service provider, they want it to be quite rightly completely the opposite. Everything standard, everything running on a standard platform, as few screens in front of every human being as possible. So there's obviously a lot of integration needed into reporting systems, network monitoring systems, ticketing, all of those things that the service provider uses Yeah. To streamline their business without getting too far down the technical work hole Yeah.

Yeah. At this point. How easy is that done and how is it done? Questions, is that right? Yeah.

You're good at this, Graham. We we should do this No worries. Get yourself. I've done it before. No.

It's it's it's a great question. I mean, the the beauty of of the way that I IP fabric's been put together as a tool, it, is API first. Right? So and and and this is this is always it's a bit of a cop out answer in some ways, but but we'll we'll run with it anyway because because the idea here is that the data that sits in IP fabric and the insight that that the engine at the back end is delivering is presented through an API. The the the user interface for for the product actually uses the API in order to get to that data.

Great. Mhmm. But that same API is accessible externally. Right? So we can build integration of that data into pretty much anything.

Our solution architects are brilliant at this because what they've what they've done is they've created a a a software development kit that allows access to all of that data through, using the Python programming language and using Ansible, which is a very commonly used automation tool, in order to to use IP Fabric data elsewhere. We're working a lot with a lot of other people who sit in our ecosystem, in that sort of automation ecosystem in order to build standardized approaches to integrating that data. So if you want, for example, to make sure that your ServiceNow CMDB is kept up to date or your NetBox implementation if you're if you're doing automation, If you've got a ticketing platform that you wanna be able to insert data from IP Fabric, relatively straightforward to then take API calls to to retrieve the data from IP Fabric and insert it into the places where it needs to go. Relatively straightforward, says the very highly skilled and qualified network engineer. You bet.

So kind of pulling this together because I'm I'm aware we've gone from, hey, can we work with managed service providers right through to the art of the possible doing everything? I think it's fair to say that, you know, we are talking about evolution rather than revolution. So, you know, I I think if you were to look at this step by step, it would follow that. Well, start by just automating the manual tasks that you are doing already. You know, don't don't try and boil the ocean.

Don't completely rework the way you run your business, write your processes. Just start with accurately discovering networks Yeah. Something more enjoyable. And from there, you can start to develop the the additional services, the additional capabilities that you need around it. Exact exactly that.

Yeah. I mean I mean, we we all as operational folk, you know, we all go through a process where we have to do grunt work in order to to deliver an outcome. So eliminate the grunt work where you can. Still retain the same approach, but just just take out those those those elements that are causing you pain. But then once you've got to that stage, that's when the door opens fully because you can see then, here's all this data, here's all this insight that you can you can build new capability on.

One of the things I love about doing this job and particularly with a product like IP Fabric is every time I speak to a service provider, there's always something that they will throw in that they wanna do. You can go, actually, I hadn't thought about that. Yeah. Yeah. And, generally, we can where we are API first and there is a lot of flexibility.

It's true. Once you've once you've got that network data, what you choose to use it for and how you choose to use it is really incredibly flexible. Yep. And that's where a lot of the strength lies. Yeah.

As and and I suppose from from that point of view, it's not just the the the network operations team then that can use that data, and and you realize just how how widely, disseminate that data can be and and how how many people it can actually help. We've got that there's a great use case where we were we were bought by a particular organization to to audit the network in terms of looking at support contracts. Right? So we had to help help procurement effectively, range of different ways you can use that data. And as you know, I'm I'm gonna I'm gonna share this story.

We were, we were visiting a managed service provider, a couple of weeks ago, and one of their engineers walked in the room and sort of pointed to it. You guys are IP fabric, aren't you? And I said, yeah. Anyway, oh, you've changed my life. I love you.

Like, we were long lost brothers. It was it was fantastic. So It gets gets you right in the feels, that one. It does. It does.

It it kinda it kinda makes it all worthwhile. And, yeah, as as as you know, we can talk about this stuff all day. True enough. I think probably a good point to think about just summarizing kind of where we are, where we're going. And I think a lot of that is we are still discovering how far we can go with managed service provision and delivery of consulting services using using IP fabric as a tool rather than a product to sell to the customer, and we're still learning this with every, you know, with every meeting, every conversation we have with service providers.

I think that this podcast today is very much the start of a conversation. Exactly. You know, it will be great to involve some of the service providers we're working with to get their input on where managed service provision is going so, you know, we can use that as part of part of our thought and development cycle going forwards. Yeah. But it's exciting.

It it's it's truly exciting to see this see this stuff being used in anger. Totally. As you as you as you said before, I mean, to to actually sit and have the conversations with people and have your own eyes open. I mean, hey. Look.

You know, we've been working in this industry long enough to to to to to see what this means to people. But, I mean, the the amount of time and energy I personally have had to expend in in actually delivering these kinds of services, this would have revolutionized and changed the way that I approached what I did day in, day out. So to be able to to give people that opportunity and and to see what they make of it, you know, is is, like you say, super exciting because it's not it's not just about us. You know? This is this is a conversation that will bring our experience, but it's as much about about, the service providers themselves, you know, having having their understanding, their perspective as to what they believe that we, that we can bring, and then working with them on on developing the product if needs be into a bit certain direction to move things along.

You know, this is, you know, this is all part of a trajectory, Right? We're trying to get to a place where we can make things easier for everybody. Definitely. Which feels like kind of a nice nice point to wrap it up for today. Obviously, if anything in this conversation has has peaked your interest, you wanna continue the conversation just with us, you know, feel free to reach out.

You can reach us through the website. You know, the usual methods of getting hold of people. It would be great to have the conversation, continue the conversation. Like I say, yeah, we we learn more every time, which is which is fantastic. Nope.

Looking forward to to hearing, hearing some feedback. Darren, thank you very much for your your your knowledge, experience, and expertise in this. Pleasure my friend as always. Thank you.