On this episode of Capital GEEK, we geek out with my good friend and network automation GEEK Chris Wade. We discuss trends in the network automation industry, best practices, and how to integrate NetOps with your cloud ops and SRE teams.
Chris co-founded Itential in 2014 to simplify and accelerate the adoption of network automation and to transform network operations practices. Using a model-based approach, Chris leads the innovation and development of the company’s flagship portfolio of network automation products.
For more information on today's episode, visit https://capitalgeek.io/
Connect with us:
Leave Some Feedback:
Did you enjoy this episode? If so, please leave a short review.
Links Mentioned In Today's Episode:
Website: www.itential.com
Josh Stephens 0:03
Hello everyone and welcome from here in the capital city of Austin, Texas. This is Capital GEEK, a podcast dedicated to the founders and operators that create the products we love and turn them into fabulous companies with meaningful exits, whether you're raising your first round of capital or racing toward an IPO. This is where we deep dive on the lessons learned from seasoned industry veterans, geeks of all types, the experts leading product and engineering teams operations and finance or sales and marketing, and will both learn from their mistakes and celebrate their successes while providing a roadmap for you to accelerate your own journey towards success. My name is Josh Stephens, CTO at elsewhere, partners, and I am the Capital Geek.
On this episode of Capital GEEK, we detail with my good friend and network automation geek Chris Wade. We just address trends in the network automation industry, best practices and how to integrate net ops with your cloud ops and SRE teams. This one's super geeky, so hold on tight. And let's get started.
Chris Wade, Hi. Good morning, sir. How the hell are you?
Chris Wade 1:13
appreciate it. Thanks for having me.
Josh Stephens 1:16
It's always a pleasure to hang out with you and chat. It's a little bit of an interesting time right here. A few days before the inauguration of Biden Hmm.
Chris Wade 1:27
Oh, man, I didn't know you were gonna go there to start.
Josh Stephens 1:30
How can you start anywhere else right now? I mean, the whole world is freaking out about it. How are things down in Atlanta?
Chris Wade 1:37
Good. I mean, as far as Georgia goes, we were kind of in the middle of this lately. So it's, like, you know, my kids on TV can't see anything other than political advertisements, I think we're gonna get back to some regular regular TV.
Josh Stephens 1:51
So that would be a really nice change for a while. But I'll tell you, it's a beautiful day here in Austin, the sunshine and the birds are singing. So it's still a great day to be American in my book.
Chris Wade 2:01
Absolutely.
Josh Stephens 2:02
Network automation. When did you first realize that that was where you wanted to focus for the next stage of your career.
Chris Wade 2:12
For sure. So basically, how we've been managing networks has been the same for a long time. And that's really been CLI based. So it was a bunch of, you know, highly trained individuals on top of complex networking equipment. And we've been doing speeds and feeds, we've been improving networks, as far as I can remember. But what really happened about five years ago was we started talking about network becoming more like software. And, that that kind of virtualization and programmability of the networking to the rest of infrastructure is very normal, like compute and storage has gone through this. And what we saw is basically, networking, which has been highly vertically integrated. For the past, you know, 30 years, is now starting to become more like software and becoming programmable. So the idea is really, that we need a new ecosystem to kind of take advantage of this programmability. And how we've been managing networks is not going to transition to how we manage software based kind of cloud based networking infrastructure.
Josh Stephens 3:11
So I think I see your point, you know, with compute with storage, and really, everything we do today in a modern DevOps or a cloud engineering team. It's all done as code, it's all automatable. It's all in the culture is so different for teams that are using automation tools, that I could see why you would have seen this early and thought, wow, this is how do we make this same reality true for the network space?
Chris Wade 3:45
That's right. And if we think about how we manage networks, we're always talking about risk mitigation. We have this, you know, network infrastructure, that, you know, we can only change in the middle of the night, we have to have highly trained individuals, we have to go through change management boards, if I go talk to my application team who's building code here, and I say, All right, we're only going to change code in the middle of the night, we got to make sure we don't break anything. Like it's just, it's kind of flipped on its head. It's like how do we make the system redundant? And how do we how do we apply DevOps principles to how we manage network infrastructure? Now, obviously, you know, there's a fiber optic cable coming out of our office here, like I can't, you know, I can't like make that redundant without extreme costs. But, you know, as we start moving networking into software, we start seeing innovation again. So I think that'd be my my big point for the day is, you know, networking has been boxes and wires for a long time. But once we start turning it into software, we're kind of creating this whole ecosystem. How do I get network as a service? How does multicloud networking work? How does network meshes and Kubernetes pods work? Like all you know, the OSI model still prevails? and networking is still networking, but it's cake. It's taking in new shapes and morphing into into how do I solve new problems. And there's this whole ecosystem built around how we've managed networks historically. And the question is, how much is that translates? And how do we have to have new practices to kind of deal with it?
Josh Stephens 5:07
It definitely seems as if it's, it's do sort of a generational leap forward. And I love what you said about the culture on those teams being all about managing risk.
You know, it's sort of like managing, you know, an insurance company or something where you're thinking about loss as opposed to gain. And what we see in modern DevOps teams is that it's all about the gain, it's about how can we push the technology forward faster, to have a bigger impact on the business? Whereas you're right in the networking world, for as long as I've been a network engineer, which is a really, really long time. You know, it's all been about just don't cause an outage, you know, don't cause a performance degradation. Don't Don't rock the boat. And and I love to see things moving forward rapidly. You said about five years ago, you had that realization, and you set down a path to change it. Tell tell us about the path that you've taken from that time to today?
Chris Wade 6:02
So, yeah, I guess it seems like ancient history at this point. But there was, you know, the idea was, there was a lot of Google and Stanford people talking about how we were going to basically turn networking in the software and the receives open source initiatives around controllers and orchestrators. And the early idea was, we were going to remove the control plane from the router. So the router is a big vertically integrated, you know, box. And if we can remove the brain from it, then we can just make these cheaper, faster data plane processing boxes, and we can use some of the benefits of software control and centralize that. So you know, routing protocols can be centrally managed, and all that kind of stuff. So the early days, the thought was, we were going to have a couple of open source controllers. So like an Open Daylight, which is now under Linux Foundation, the big guys came together and said, Hey, we're going to build a new operating system for how we build software controlling the networks. What in reality has happened is that the innovation is almost going so fast that nobody could wait on the group, the group thing to get that down. So just like in the software defined whitespace, you have 50 vendors all with their bespoke controllers. So I, you know, I really think that from a networking standpoint, we've kind of unleashed the beast, which is awesome. But we're gonna get innovation cycles so fast now, that is, it's really not normal, kind of in the networking space, like a new line card with new speeds and feeds with maybe a new routing protocol was normal. But the idea that, that the level of innovation on different types of networking technologies is is, you know, accelerating faster than I've ever seen it. I mean, I don't know what you think.
Josh Stephens 7:39
I think it reminds me of what we saw in the early to mid 90s. You know, where things were changing so fast that you could not wait upon a standards body or, you know, the I triple E or an RFC committee to come up with some sort of a standards, you just ran as fast as you could. And whoever grabbed the most market share the way they did, it became the standard. I can remember when NetFlow first came out, it was a game changer. And there was there were multiple specifications. But there was an RFC spec, there was Cisco's version. And, you know, there was also s, small s flow and IP fix. And eventually, the majority of the market just said, I will just use Cisco's model and just kind of go from there. I can remember at SolarWinds, when I was building the NetFlow product. We wanted to go support Juniper. And they use something called j flow. And so we started researching it. And finally, as a test, we just turned it on to see what would happen. And it worked flawlessly. Because it was the same exact thing as Cisco's NetFlow same structure in the packets. And so it was easy to add support for that. It bills sort of like that. But on steroids, it feels like that, but much, much faster. And what I love about it is that we've we've sort of unstyled the network team. You know, for years, the network team had this elite status among the rest of the infrastructure team, because again, it was about managing risks to be took the network down. You're just out of luck. A good buddy of mine, Dan Schoenbaum was just sending me a slack message. And our they live in Portland. They're having a horrific windstorm today. And internet service is down and phone service is spotty. And you know his wife's in general counsel for a tech firm. Dan works with me on several different projects we're working on together now with companies we've invested in. And they're, you know, they're dead. There's nothing they can do. They can't get out to the internet. They can't use zoom. They can't make a phone call it can't use slack because the network's down and the network's down, everything's down.
Chris Wade 9:51
I mean, my kids, we had power go out and they said that's all right. I'll just get on my iPad. what he's talking about. He's like, has a battery in it. I'm like, well Like, is going through that router that's plugged into the wall, which has no blinky lights on it. So there.
Unknown Speaker 10:05
Yeah.
Josh Stephens 10:06
Yeah, it's really interesting.
Chris Wade 10:08
Let me ask you about the silo conversation. Because the other part of this is that as we start to adopt, like cloud based networking with a VPC in AWS, which is basically just Amazon's router that you're consuming, you know, a lot of CIOs are splitting the net ops team from that cloud, you know, SRE DevOps team. And even though we might not be siloing, with an infrastructure, the question we have, that we talked about is how, how do CIOs view their networking net ops team and their SRE cloud based team because I think a lot of CIOs have been very comfortable splitting those teams and saying lift and shift and application from the data center into the cloud. And it goes from, you know, my networking team to my cloud team. And I think with AWS, sometimes called the AWS effect, when first people went first into the cloud networking was maybe viewed as less important.
Josh Stephens 11:02
Yes.
Chris Wade 11:03
Because configuring my cloth fabric in my data center might be viewed as complicated and, doing a transit gateway into VPC might be, you know, thought of is simpler. So, you know, people thought, I'm just going to keep moving things to the cloud and networking, maybe you know, from from an enterprise CIO perspective, might be less important. But now what we talk about is, as we've started to get outside of just AWS, and you start, you know, I have stuff going through equinix, or I'm connecting the SAS platforms like Salesforce service. Now, I have a hybrid multi cloud story, the way we talked about it is that your network exploded on the internet. So it went from like, inside the four walls of my data center to people swipe credit cards to nominate us to now like, what is your network? Like? What Where's it at?
Josh Stephens 11:48
It's a great question. And I think for very large companies that have a lot of own infrastructure, it probably makes sense to have a net ops team that may be separate. But you know, I mean, more and more companies today, that don't really even have an IT team separate from the rest of the team, because all of their infrastructure is in the cloud. And I think that one of the problems we're facing right now is we haven't done a good job of teaching our cloud engineers or SRE teams about networking. And if they really understand networking, and an advanced level, you know, that improves security of your cloud operations includes performance of your cloud operations.
But it's, it's dangerous, I think, to consider networking as Oh, I only need to know the fundamentals, I only need to know enough to get by. Because if you're not good at segmenting those networks, and providing the right safeguards in between them, you're leaving an opening for a hacker to get in.
and make a
Chris Wade 12:48
Yeah, I mean, the OSI model still prevails. I mean, just because it's a VPC or transit gateway, or, you know, an equinix, cross Connect ECX service. Like it's still networking principles still apply. And to your point, you know, it's when it's just a couple VPCs and a couple AWS regions, you kind of have that that simplification?
Josh Stephens 13:10
Yes.
Chris Wade 13:11
I would argue once you explode your network on the internet, and you're using SAS platforms, and people are sending their firewall stuff to z scalar. I'm using ns one for DNS, you know, it's not just going to a website, like you're routing traffic through critical network application infrastructure, across Wham Linux potentially encapsulated over your your internet connection. You know, what used to be simple now. Now you see those SRE teams, in enterprises, you're talking about saying, Let's invite the networking guys in explain something to us, and then send them back because I'm going to code it in a terraform plan or some other thing. And ultimately, I think that, you know, successful organizations are going to, you know, that that that wall will be built into a road as those teams kind of work together. Now, they facilitate that, how much of your infrastructure is on the cloud versus on prem? All that all that stuff is gonna drive individual decisions. But ultimately, you know, these two worlds are important to work together.
Josh Stephens 14:06
I completely agree. You know, I've been involved recently with some nonprofits doing mentoring and resume reviews and mock interviews for some students that are learning to code. And what I've noticed, even with students that are juniors and seniors in high school is in many cases, they're very proficient software developers by that age. I'm working with a kid right now. I'm mentoring him. He's a first generation American of Russian descent. Brilliant software developer, he codes in seven or eight languages. He's already been teaching coding to younger kids. But now that we're having this conversation, I'm realizing that, you know, I don't think these kids are learning very much about network engineering early on. And I think we're moving to a world where, just as I think software development, or at least basic coding skills, should probably be a prerequisite for 80% of the knowledge workers out there, regardless of your field. I think network engineering steals, at least the basic understanding of when you're over your head. So you can bring in some help, is also something we need to, you know, spread around the team. I've just seen that it's, it's really hard to find SRE teams that are really well rounded. And so I think we have to do something to change that.
Chris Wade 15:30
Yeah, and, you know, back to your coding example, for proficient software developers a question I asked, and we run the danger of sending, like, angry old networking guys here.
But, uh, you know, the system administration skills, just have infrastructure, we used to just consuming containers, like environments, like you have an environment as a service almost in this cloud environment. You know, the idea that there's, you know, hard disks attached, and you have to concern yourself with with different pieces of infrastructure is almost kind of the first wave before, people are really concerned with networking. The flip side is, if you believe in the Mac and edge world, if you believe in the super macro effect that, you know, we centralized for the last 15 years, and might, it might swing back the other way, where a lot of your workloads, and a lot of your applications are going to take advantage of some sort of edge infrastructure, you would start to think that the networking component of that continues to kind of increase in importance of how you think of actually building things. Because right now, it's, you know, there's, there's some edge applications, go talk to a manufacturer, they might have, you know, Ford Motor Company, I'm assuming has some stuff in the manufacturing plant, because they don't want to lose connection to a cloud far away. But there's gonna be all sorts of edge opportunities between the public clouds pushing out service providers, providing neck edges, and I think as you build applications that take advantage of edge, you know, networking considerations are going to become more and more important, I just see lots of tail winds towards, you know, networking, whether it's the innovation on networking products, or taking advantage of highly distributed workloads to invent new ways of achieving things.
Josh Stephens 17:12
Chris, if I would imagine some of our listeners out there aren't old angry network guys like you and I, and might be wondering what, you know, edge services, or maybe you could spend just a moment and describe that for people.
Chris Wade 17:28
Sure. So, um, you know, if we kind of go back to, you know, thin clients back with the Sun Microsystems used to have thin clients where most of the compute processing was off, and then we went to a lot of just, you know, PCs, I'm on a Mac Pro today and the processing, and this thing's crazy. And then obviously, with public cloud, that was a an extension of centralized workflows, whether it's, you know, mainframes and other things. So we put all the compute the big building, and we all connect to it. And, you know, the most common nomenclature is, you know, you use your cell phone, it doesn't use as much battery, but there's a refrigerator on the other end in some public cloud producing those maps and emails for you. So we've really pushed a lot of the computer to that public cloud infrastructure, which is unleashed all sorts of innovation. But there's a thought that as wireless networks become faster, as the fiber optic cables were connected to become faster that we can start to do some low latency applications. So whereas 4g, a lot of people would credit that kind of innovation, allowing something like the iPhone to exist. You know, there's a thought that, you know, whether it's, you know, you hear crazy things like, you know, autonomous vehicles using the MEK sites, but how do I push the compute closer to the edge closer to the end user, so that we can innovate on some new things? You know, there's a lot of discussion, I think CES was this week. Yeah, it was, I think they were talking about a lot of the stadiums producing, you know, local, high rez feedback from replays on the field to your phones. So they need to move that compute. We don't want to round trip time, you know, to Virginia every time we're gonna do something. So, you know, there's, there's, there's a big debate, I was at a trade show where somebody said, you know, one of the use cases is remote surgery. And then they asked the room who wants to raise their hand to be the first you know, cadre of people going to enter the remote surgery game? But you know, I think there's there's a lot of thoughts around this unleashing kind of a set of applications that weren't possible before. So it's hard for us to almost think about what these can be. But there's a there's a tremendous amount of, innovation being thought of on the edge. I mean, think about with with Coronavirus, we've been on zoom calls, people are having a little bit of, you know, zoom overload.
Josh Stephens 19:45
Yeah.
Chris Wade 19:45
But you know, if I could put all the if I could put all the codecs closer to the edge if I could integrate it with my unified communications it you know, if I can push some of that real time stuff to the edge, then maybe maybe we get more of a real time experience. Maybe it becomes more more more normalized, maybe my kids Oculus starts to play some new games that weren't possible before. I don't know.
Josh Stephens 20:09
It's really fascinating that, you know, edge computing and SD LAN are all over the news right now. And it's gratifying for me, because, you know, these technologies have been around for a really long time. But it's taken a while for us to innovate business use cases, that made sense to people where they could really sort of take advantage of them in a way where they saw ROI, and it was measurable. And so for me, at least, it feels like for the last six to 12 months, the market has just dramatically accelerated, it seems like we waited longer than we should have. And now and now everyone's trying to run faster, keep up and catch up to where they should have been.
Chris Wade 20:53
Yeah, and the byproduct of this is, it's some point, we have to figure out how to take advantage of it. You know, there's, there's a chicken and egg, you know, discussion around, does this explosion of bandwidth and might come with 5g, and this this ubiquity of compute. Does that. Does that facilitate application developers to innovate something new? Or do we need to invent something new that drives the demand for increased bandwidth? And, you know, I think it's a it's a, it's a great debate, I think time will tell, but at the end of the day, you know, how much bandwidth do you need? How much compute we seem to figure out pretty good ways to, to keep to keep the consumption model going?
Josh Stephens 21:36
Oh, yeah, for sure. It's, it's really fascinating to me to think about, for you as well, how our entire world has evolved around the evolution of the network, and bandwidth availability. You know, when I started my professional career, in the military, I think I built the very first website for the Air Force, and including the one for my base. Because it was it was brand new, I mean, hardly anyone used the internet back in the early 90s. And I can remember, when a 14, four modem was just screaming, and if the whole office, the whole building shared a T-1 at 1.54 megabits per second, like you were just fat and happy. And we would go to the office at night to download big files, because it was so much faster. And now we have gigabit to our homes and 200 megabit to our mobile phones in our pockets. I mean, it's, it's really fascinating. And it's the amount of innovation that it's allowed. And quite frankly, for the last year, I think it's allowed many of us to continue to work and stay in business, because without viable bandwidth, we wouldn't be working from home via zoom.
Chris Wade 22:51
Yeah, I mean, it's terrible is this is all been it's, it's, it's hard to imagine what would have happened maybe 15 years ago, as far as, you know, economically, or or, you know, to our to our personal lives without, without, you know, some of this technology.
Josh Stephens 23:07
Can you imagine what the lines would have been like at blockbuster. If we were all locked in our homes, you know, back in, you know, the mid 90s. And we're waiting for the newest release to come out on DVD or VHS and we're all in line standing six feet apart. You know, it would look like a breadline from, you know, from World War Two or something or the depression. And we're all waiting to exchange that DVD.
Chris Wade 23:35
I always remember the block, but you would go up and there would be a picture of it. And you'd have to pick it up and see if there was a movie behind it, you know? So it's like, it's like, Is there one back there? And then you'd go to you go to the return bucket and say like, Can I browse? What just got returned to see if I can steal it before it gets back?
Josh Stephens 23:50
Oh, I've done that so many times. So that's the person behind the counter. Hey, could you look and see if this movie might have just come in? Yeah, it's a, it's an amazing opportunity for us here. And what I hope happens next, and what I what I've seen already begin to happen is that while the last year has been horrible, there's no doubt about it. We have made a lot of progress as a country in terms of leaping forward, I would say a generation in terms of how we use technology and how we think about work. And now we have the opportunity to do things like maybe ask your city to work from home as much as they can certain days and accelerate road construction. You know, that's always been a huge problem for us in these cities like Austin and Atlanta, anywhere that's growing fastest. You know, how do you upgrade your infrastructure without having such a negative impact on local business but we've proven now that most of us can work from home and a lot of these jobs if if we need to or want to. And you know, for me, it's it as I walk around downtown Austin now. And I think about my friends and the way they're working, it feels sort of European to me, there are more bicycles out than ever before, people are taking walks, they're working a little, a little less, they're spending more time with their families. It feels like from a culture perspective, we might end up in a really good place when this settles down and vaccine has been has been distributed and the crisis is over. And now we're trying to, to leave that that section of our history.
Chris Wade 25:33
I mean, we've made a lot of big promises around digital transformation, you know, and that's, that's been an umbrella term that I think, probably been used quite a bit. And I think it's been, you know, one of those things spoken about and not lived for a lot of organizations and cities to your point. And in the last, in the last 12 months, I think people would agree we've, we've got five or seven years worth of digital transformation out of the year, have you had a necessity? And the question is whether the benefit of that has been, you know, identified and understood enough so that we continue at a similar pace, you know, on the digital Because ultimately, if I can not if I can get my, my decal on my car without going to the DMV, and if I can, you know, if I can, if I can do all those things, then that's more time to ride bicycles in Austin. And, yeah, that's great. Like you said.
Josh Stephens 26:31
I mean, I, I had a doctor's appointment this morning from this very room, you know, over zoom with my my physician. And if you rewind a year, mostly what you saw were offshore telemedicine companies, which offered an alternative to join into that office. What do you do go see your local physician, but there weren't a lot of local physicians who were also doing telemedicine. And that's, that's all changed. Now, I think for the better. And I think we're going to invent a lots of new ways to be more productive and happier. One of the companies that I worked with, they're called Lit Lingo, I'll give them a shout out now. And they make an AI based application that provides analytics and also real time correction, when you're saying something you shouldn't say over slack email, teams, Zendesk, or ServiceNow, whether it's a employee contractor type data, or di type data, or you know, anything that you would say that would cause corporate risk. And it's, it's going just really, really well. And so long story short, we had a holiday party for the company. And it was done over zoom. And it was a small company, you know, so I put a big monitor in the kitchen. And they use the company called trupple shuffle. And they sent everybody a box of ingredients for a couple of cocktails and a dinner. And French chef came on the zoom and taught everyone how to make the dish. So as a group, we all cooked in our own kitchens at home, made the cocktails enjoyed them cooked and ate dinner together. But all done from the safety of our homes, all via this remote chef experience. And it was fantastic. It was one of the best Christmas parties I've ever been to, or holiday parties.
Chris Wade 28:19
Very interesting. As much as a lot of these things have been good. The flip side, obviously to talk about the other side is the level of collaboration, I still think we have a long way to go before some of the zoom calls could be as collaborative as we are, you know, in the, in a software development shop, you know, nothing replaces the whiteboard and the marker and the big brains in the room. I don't know how we overcome that. But there's, it's really about the balance between all of these things. I know. I like to get out more, I'm actually in our office. We have a lot of space here. So we get to spread out I get to escape. Too many kids being allowed at home but the you know, the somehow we have to bring that balance of collaboration in the workspace and augmented by by the benefits we talked about.
Josh Stephens 29:09
I agree. I mean, so many times for the last year, I've said, you know, I've been chasing this problem for weeks, but it feels like if I could get into a room with a whiteboard for a couple of hours with the two or three people I need to collaborate with we could solve the problem. And you know, Lucid Spark recent release, it's pretty cool for a whiteboarding tool to use from home but you're right it doesn't replace the in person whiteboard experience. It's it's, at least for me, it's the hardest thing that I do to try to do remote and I haven't I don't think anyone's found the right solution. Although my brother Zack who works at SolarWinds, shout out to use that. You know, when he moved recently and he didn't have room in the rental house that he's staying in to his house that he's building is finished for his like 80 inch 4k TV. So he hung it in his office, and they set up like a telepresence. And for him, it is very much like a life size meeting. And maybe there's a solution that somewhere in there around how we do whiteboarding, if it was big enough, but that's an example, I think of an area that we really haven't advanced. I mean, I don't think I've bought anything new from my home office since the pandemic started. So I like haven't gone out and bought a giant screen or telepresence unit or some something to make, you know, doing zoom calls all day long, easier. And we could we could extend the experience at home to make it more productive yet. So far, what most of us have done is try to get by with whatever we were using before for the, you know, the odd late night meeting or some time maybe when you were sick. But I think there's a lot of innovation to be done for how do we equip home offices? Now that we're working there maybe, maybe half the time forever? Or maybe more than that?
Chris Wade 31:02
Yeah, and if you think about the networking side of this, you know, we had wayins, and we had office buildings. And if you wanted to gain access to your email, you had to VPN to, basically VPN into the land so that it looked like you were sitting in your office? And, you know, I'm sure you've had a guest on for zero trust, I'm sure you've talked about, you know, quite a bit, you know, people are everywhere, you know, how do I how do I securely perform whatever I'm doing, from anywhere I want to, with, the expectation of quality and security and, everything else that goes into it.
Josh Stephens 31:41
So what we've really talked about is, over the last year, there's been a gigantic amount of change for the attack surface for any company in the world, really, you've got people working from home, maybe shared offices, on a, continuous basis. What is the role of network automation, with all of this change and how it's evolving?
Chris Wade 32:12
So if we go back to something we said earlier, there's, you know, the the network device, the traditional network devices, a vertically integrated box, control plane data plane, it routes packets, it does what it does, as we start to turn that into software, we have new ways to operate that and we start to have these software controllers on top. So instead of configuring networks, box by box, so if I have 10 boxes in a data center, and I want to change a path, I might need to do some configuration on different devices. Now we have like this intelligence being built into each of these domains. And instead of an application being in a single data center, you know, the zoom call we're on is is traversing lots of services, some of them cloud services, some installed software, some routing things. So the question is, how do I make changes to applications that, you know, back to have exploded on the internet, it's not just, it's not just installing software on my laptop, but I have to coordinate control across a variety of technologies that are all, you know, inner, inner, inner locked in, in some way. So from a network automation standpoint, we think it's very important that people think about kind of the end to end service. We think that they need to continue to increase kind of the modern networking concepts. So you can take advantage of all these things. And then the question is, how do you how do you operationalize it? So network automation starts to look a lot more like programming infrastructure than then in traditional network CLI. So how do we how do we achieve that end to end? And how do you interact with a lot of IT systems, because if we look at changes to the network, when we look at it, usually like eight to 12%, of somebody's time is spent actually changing the network device back to the the risk mitigation in the middle of the night. We know, it's like a bunch of pre checks and post checks and copy paste and the tickets and changing it. There's all these kind of it integration. So it's not just automating, quote, unquote, the network. It's also kind of the entire end to end process. I mean, I don't know. I mean, what's the average number of SaaS applications that the companies you work with use?
Josh Stephens 34:26
Oh, my gosh, that's a great question. Well, I will tell you this. We invested in a company about almost two years ago, that had about 30 employees, 25-30 employees. And when we inventoried the SASS apps that they use as part of the business, you know, not things that people use personally. There were 84 of them.
Chris Wade 34:48
There you go.
Josh Stephens 34:50
So, you know, two and a half times the number of employees, and I don't know if that ratio scales, but what I do know is that we no longer look to our company to provide us with technology. Where, you know, in the old days, the IT team were the only technologists in the company, and you kind of looked at them to provide solutions. But now, you know, if you're working in, in a tech company, especially, but really, in many different specialty fields, everyone in the company is some sort of technologist. And we're all proficient with, with deploying cloud apps and using SAS apps and things like that. So, you know, we've democratized the use and acquisition of technology to such an extent that it's hard to put your arms around, you know, and I think people struggle with it.
Chris Wade 35:43
The other way I would describe it also is that, you know, when the CIO looks at a challenge, and looking at how to optimize the entire, the entire company, right, that's, that's, that's the goal. But we've allowed business units to optimize themselves, which I think like to democratizing at the business unit level. So if you think of like, I'm, let's say, I'm responsible for I am, I'm responsible for like authentication, I might optimize myself and say, I'm going to use ad is a service from Microsoft as your like, I'm going to no longer have my server. And then the Unified Communication person says, I'm no longer going to host an Exchange Server, I'm going to use Office 365, the firewall team says, we're going to do zero trust, I'm going to use z scaler, the DNS team says I'm going to use ns one. And that's those teams, trying to use the most modern techniques and technologies to achieve the best outcome for that business unit. But when you step back, and you look at the total, now I need to route route the traffic from the land securely to the Z scalar. Pop, to route back to the data center with low latency to my AWS application. And all of a sudden networking goes from, you know, this ubiquitous bandwidth that is just always there to now being a critical, or maybe the most critical component of networking, my company, which has now expanded across, you know, the Internet, and it shows no sign of slowing, you know, in my job.
Josh Stephens 37:09
Yeah, you know, I invested in a company and now work with them called wt fast or what the fast, and it's an network acceleration software for gaming networks. And I'd never given much thought to the challenges of gaming networks, but in many ways, it is a great example of sort of edge computing, and, and how to build very low latency optimized network routes for those users. Because if you're a competitive gamer, you know, half a second, or you know, two or 300 milliseconds of delay, from the time you hit the button to kill your enemy, to when that packet gets there, it can change your career, it's a really big deal. And, you know, I hadn't, I hadn't really worked with applications in a while that were that latency sensitive at that scale. And so it really is for them about, you know, some sort of a way to optimize and, and tunnel that traffic to a gaming server on the edge of someone's network, and, and back and forth. So it's been fun working with the team there, what the fast and, Rob, if you listen to the podcast, hey, but I think you're right, I think it's, it's, it's really, really challenging for these organizations. And by the time if you're the, CIO, by the time you survey what everyone's using, and try to put together a cohesive strategy for the organization, it's probably changed, you know, it's changing so rapidly. I don't know how you get in front of it.
Chris Wade 38:40
And, you know, some people listening might say, Well, you know, we've read the stories that people high frequency trading desks, building their own optical network to shave, you know, latency out of trading applications, and using the high speed gaming and these, the sound barrier edge, right, the sound very, like not not normal for the greater population. But the question is, what sort of innovation will we imagine a couple years from now as being just part of part of our life? Right? Like, what's, what's one of those applications that are going to improve, you know, our ability to continue to digitally transform and use technology kind of ubiquitously.
Josh Stephens 39:19
It's, it's really amazing, you know? Yeah, I don't know where we're going from here. But I'm very excited about the level of innovation and the level of change. And for any senior technologist or anyone who's geeky, like, like you and I, this amount of change tends to lead to lots of innovations that are hard to predict. And, to me, that makes it a very exciting time to be alive. Maybe we could talk just for a few minutes about Itential. You and Ian started the company. Tell us a little bit about the history and and what you guys are doing in 2021.
Chris Wade 39:54
Sure. So we really looked at where we believe networking was going We've really felt like this is one of the biggest transformations in networking. You know, we talk about speeds and feeds. But this is really kind of how we consume networking and how we're going to use this technology to really, you know, enable enterprises and service providers. So, you know, we made the, you know, the the bat to kind of move out towards where networks, we're going to be more kind of API centric and be more modern. And we've we've, seen how it's evolved with, you know, different things like SD, LAN multicloud, those types of things. And we're really focused on providing kind of that end to end automation concept to your networking and cloud teams. And what that means is that we've kind of alluded to this a few times in the podcast is that we have strong networking talent within the industry, we we have growing software development talent, in the industry. And the question is, how do we facilitate people participating in automation, kind of where automations come from is that somebody who understands infrastructure, or the network might describe what they want, you know, and if somebody's being fancy, talking agile, then there will be a user story. And it goes over to a developer to instrument and it kind of, we call it DevOps A lot of times, but but really, we need SRE, that are really thinking about, you know, reducing the toil doing it themselves. Everybody's participating in automation. And the question is, how do you facilitate that. So we've been working on this low code approach, to get the most people to participate in automation, because we think what is thought of as automation today would just be networking tomorrow, like this concept that you do networking, and then you might want to automate it? You know, networking will be automation.
Josh Stephens 41:42
I agree completely.
Chris Wade 41:44
We will all be working on the machine that automates the network, all of the complexity we just talked about today, you can't linearly grow your team with the complexity of these modern networking infrastructure. So, you know, I like to talk about all the stuff under the hood. But at the end of the day, we're trying to facilitate these teams so that we can have the most people participate in automation, with diverse sets of skills to achieve, you know, the outcome, which is ultimately supporting application owners and moving at the speed of the enterprise.
Josh Stephens 42:14
You know, Itentals offerings have evolved a lot last few years from, you know, a very robust on prem product that some of the largest companies and service providers in the world use with your complimentary services, obviously, now, with a SASS version out there for anyone to use. What what are some of the, you know, push backs that you hear from people in terms of why they might avoid investing in a network automation strategy? And what would you tell those people in terms of why they should think about this for 2021?
Chris Wade 42:50
Sure, so when we're talking to folks like this is the concept of automating is, not foreign to most people, I would say four or five years ago, we first started to be talking to people, we had to start with the value of automation, most people are using Ansible playbooks, they're writing Python scripts, if they're a cloud SRA, they're using terraform. And if you think about the maturity of automation, it starts off by, I'm trying to make myself more efficient in my job. So I might write a script, I might write a playbook so that I can do my job, you know, more efficiently. The question is, how do we start sharing that intellectual property across the team? How do I share it with the person next to me, ultimately, how do I make that self serve for the application owner? So it's really about, it's not about, you know, don't use the technologies you like, you know, from an Itental standpoint, we're talking about, like, we want to do all of the above strategy, automation is good in some capacity. And a lot of our discussions are really around, bring your own automation, whatever you have, let's bring it let's put an API on top of it. And let's start instrumenting it together. Because ultimately, we say we're in the trust and confidence business. And that means, you know, back to the risk adverse network engineer, you know, we have to build trust that we can have enough operational control around these things, that if you're making changes to critical infrastructure, you're authenticated, we know who you are, we're auditing it, and we're making sure you're not breaking stuff. So it's really about how do we bring kind of individual automation to the larger organization? And how do we facilitate that in such a way that we can use the tools and a lot of the processes that you already have today?
Josh Stephens 44:30
I love Itental's story. Um, because to me, the way you guys built the product and have evolved is like, incredibly difficult, you know, we built SolarWinds, and when I built Orion, I can remember, you know, selling the first you know, a few 100 copies. And we were really constrained to, you know, about 1000 network elements total. And it only takes about three bait Cisco switches to have 1000 elements. And I can remember, you know, over time we improve the product and made it more scalable and more reliable. And eventually, we were able to sell to larger environments. With automation, it has to be so reliable and bulletproof compared to a monitoring tool, like a SolarWinds, that you've got to make sure that it works 100% of the time, and you guys started with the largest networks in the world, bulletproof the solution. And now you're coming down market to other, you know, enterprises, and even middle market companies, which to me is, is really interesting, I think with automation, with infrastructure, especially like the network. And with security, you kind of have to do it that way. Because you want to be able to be sure that you can satisfy the hardest use cases, so that you can make sure that it'll handle what's needed down below. It's not okay to fail. Whereas with a, you know, if you're collecting monitoring data, and you miss a polling cycle, it averages out in the end, and who really cares, it's very different. So I've been really, really impressed with the way that you guys have built a product and matured it over time.
Chris Wade 46:13
I appreciate that. I mean, at the end of the day, when you think about automation, there's the thought of, you know, the complexity of your network is somehow correlated to your desire to automate. That's, that's, that's how people traditionally, if you think of RPA, if you think about other parallel automation domains, it's really about like, it takes X amount of time, and then we're going to do this, the reality is, we're moving into a self serve model for everything, like there's an expectation, you talked about the democratization earlier, like, I have an expectation that I can turn on Netflix, and the show will show up our application teams, I mean, we like to think of them as the user of this stuff, they've come to be their expectations have been changed through SAS and cloud to the extent that they expect to be able to do it themselves. So it goes from one where, you know, put a ticket in and wait five days, and then it will be turned up to one of like, how do I enable those teams? How do I provide governance around it to provide to provide self serve? And that's, that's ultimately kind of where, you know, it's easy to say, or harder to implement. But the idea is, how do we make it simple enough so that our developer friends, our application owners, which are driving, you know, a lot of the digital transformation? Like how do we enable them to consume it infrastructure in a self serve, kind of cloud governance model, independent of your technology choices, and independent of your networking vendors. And that's, that's, that's, that's what really resonates, kind of with that handshake between the business units within enterprises being enabled, working, working with their CIO partners.
Josh Stephens 47:54
That's fascinating. And that, I think it's so important because what you're doing is you're building in a system of trust. So people can validate, you know, the changes that are supposed to be made. And I think in order to do that, you have to first have a brand around your product that people really trust and rely upon. And you guys have built that which is really just outstanding. One of the other things I love about Itental, and I don't even know if you guys are still doing it, but you had a really fantastic intern program. You guys still doing that? Is it something you're still doing? And so how's it going?
Chris Wade 48:29
Yeah, so the the handful of us that were here at the beginning, a lot of us were in Atlanta in Midtown. So we're across the street from Georgia Tech. So we started I was I was a co op from Georgia Tech when I was there. So we're a couple of of the early developers here. So for one of the first things we did was reach out to start the co op program. It's just, it's it's really amazing how much smarter the kids are coming out now than then than when we were coming out. But the but yeah, we always have kind of a whole cohort of Co Op on rotation. Obviously, we'd like them to stay as full time developers, but it's it's really, it's, it's interesting. It's really unbelievable what they can achieve within the time period. They're here, how much they become part of the peer group. But also from a culture perspective. It's been it's been really fascinating. It's really enjoyable, because you really create that whole environment where people are, you're allowed to not understand something, you're allowed to ask questions. When you have you have kind of that Co Op rotation. It's a little bit different than interns because they do come three or four times. So by the time by the time they leave, they'd have you know, a year and a half, two years worth of full time experience, which which I think's invaluable. While you're going through school, you have a different appreciation of what you're learning. Sometimes, you know, academically if you have spent more time understanding how it how it relates to solving real world challenges.
Josh Stephens 49:55
I spend a lot of time thinking about diversity, equity and inclusion And how to integrate some of those concepts into a startup or a scale up company. Because, you know, you start out with two buddies, you want to start a company. And so you start building something and you add three or four more people. And before you know it, you've got 50 people around, and maybe you all look alike and think alike, who knows. But one thing I love about your co op program is, you know, I can guarantee you that two old angry guys like you and I have very diverse thought patterns from, you know, kids in college still. And so to me, one of the things that this program does, it adds a great amount of diversity to what you're doing there and the company, if you are going to advise other companies in my portfolio or out in the world who are of similar size. As to some advice, as they're thinking about starting their own Co Op program or college internship program. Can you share two or three tips on things I should maybe think about? And keep in mind?
Chris Wade 51:00
Sure. Yeah. I mean, even before I answer that, one of my favorite things, when I drive in, dropping kids off at school is coming over the bridge, and somebody skateboarding from the school over to keep me keep me young. So but I think, you know, obviously, being physically near makes it easy. You know, if people have to have to relocate every three months, that makes it difficult, so that, I don't know, if everybody necessarily has the option of where they're located. I think the most important thing is to partner with the school, and really invest time and energy. It's not just, it's not just Hey, you know, we're looking for some co ops to help us test stuff, you know, we're looking for giving real world experience to, you know, consumer, you know, computer science graduates, and we're gonna, we're gonna offer them every opportunity that everybody you know, there's nothing that you can ask a co op to do that you can't ask any anybody around here to do. So it's really important that they're treated, you know, the same as they get tickets, just like everybody else, they get the opportunity to make the same impact as everybody else. And, you know, I think, by the second rotation or third rotation we might have, we might hire a new person, and they might not even know that person, a co op, they were they going, Oh, they're going back to school. Some people would be like, I didn't even realize I mean, I've been working with that person for two months. I didn't realize, you know, that was
Josh Stephens 52:22
Yeah, I mean, I've been blown away by the skill level of some of these kids I've been talking to, again, juniors in high school, through seniors in college. And several of them very young, you know, junior senior in high school, junior in college, would be very qualified to work part time today as a developer. And I've challenged several those kids. I'm like, Why are so many students working as bartenders, and waiters, and waitresses, especially now with COVID, like, maybe you should be writing code four hours a day, you know, throughout your college career, that that may be a way we avoid debt in the future is some sort of work program with local, you know, companies and startups, especially now that we've learned to work remotely. You know, the first hackathon I was ever a part of Was that your company Chris. And I had, I don't even know that ever heard that concept before. This was, you know, three years ago or something. And since then, I've just become enamored by the whole concept, maybe tell the listeners a little bit about your hackathons and what you've learned from those?
Chris Wade 53:34
Sure, I think traditionally, you know, when people think about hackathons, it's kind of over the weekend, you know, pizza on the floor, you know, do build whatever you want, the coolest thing wins, I think what we found is most useful is having some sort of guidance on what we're trying to achieve. And obviously, you want to leave it as loose as possible. But generally speaking, we force for strong word, we generally we request that you join with people that you don't work with every day, right? So so it's a great opportunity for collaboration. We do think it's a mixture of normal work day, and you can work at night, but the thought of like, you know, Friday at 6pm, till Sunday at 8pm is probably not the best idea. So, and people are so excited to participate if they know the outcome of the hackathon could influence the product roadmap could be rolled out on website could could do something beneficial. So I think, you know, in the traditional sense, like in a university setting, no hackathon might be just the coolest thing wins. But you but there's, if you talk to folks, they always have the I've been asking that we do this or I think this is a good idea. This is the opportunity to I don't I say prove that the idea it should have gone to the top of the list, but this is the opportunity for you you to do what you think's best, you know, with some with some guidance, and it's been you'll see some very small stuff like an improvement to the usability of the product that everybody was like. Wow, you know, we're gonna put that we're gonna we're gonna merge that in next release to entire concepts was a gentleman who in two different hackathons built the same thing, two variants of it was an app store for automation platform. And, now its core, its core to what we do, we call them pre builds. And it's the idea is that after I build an automation, I want to be able to share it with everybody else. So now we have this entire developer site, we have all of our pre builds out on Gitlab, and not only can you go download them, but now there's this app store concept in the platform that connects to the Gitlab. So if something's promoted tomorrow night, it all of our users can have access to that to download and use that automation. So it's a way to facilitate sharing of, of different concepts. And, you know, that goes back to maybe we thought that was a good idea. But having having three people super passionately build that present it to the team, and then everybody say what a good ideas. I mean, really, you know, it's not just a, you know, a thing to mess around, it's really, it really facilitates a lot of culture and facilitates a lot of good ideas, frankly,
Josh Stephens 56:07
You know, I love what Itential to does, and I think that when anyone works in a field for a decade or something, and it goes out and builds a product or company that's built specifically to help people who were doing the job, they just left do it more efficiently and better. There's something very special authentic about that process. And that's what I love about what you guys have done, because you took your time in as a network engineer, and working to other companies, you saw a way to help, you know, people who were like you, you know, and, and you built this product to do that. It's really fantastic. For people who don't know the company, we're talking about Itential, i t e n t i a l . com. And that's Chris Wade's company. Chris, we're kind of running out of time here. Are there anything, any other things you'd like to mention or things you want to talk about before we wrap up?
Chris Wade 57:07
I appreciate the time. And, obviously, you know, reach out for any sort of discussion on the topic, whether it's market related or Itential related. We do we have our commercial website at Itential.com, but we also have our developer site, which this audience maybe would find more fascinating with all sorts of, you know, different tooling, and documentation and other things. That's a developer.Itential.io. And I'm assuming my contact information will be attached to the podcast if anybody wants to reach out.
Josh Stephens 57:38
Yeah we'll put all that stuff on their social links and whatnot. Well, Chris, this has been a lot of fun. I really appreciate it. It's always great to see you. And you know, for anyone out there who's on a net ops team or who's thinking about ways to make changes faster in your network. Itential is a great solution. You should check it out.
Thanks a lot, Chris.
Chris Wade 57:58
Thanks, Josh.
Josh Stephens 58:00
That's it, everyone. And thank you for joining capital geek, subscribe via Apple, Stitcher or any platform where you usually find fantastic podcast. Tune in again soon for another great episode of capital GEEK.
Transcribed by https://otter.ai