Overcoming Legacy Challenges: The Path to Cloud-Based Network Consolidation – The Last Great Cloud Transformation EP6

Hey everyone, it's Alan Shimel of Techstrong, and welcome to the last great Cloud transformation. For those of you who may be new to this series, this is a, uh, every other week series that we do, or twice a month on. Just that, the last Great Cloud transformation and, uh, what is the last great cloud transformation?

Well, we, we've been exploring this now for a couple of months. We're really proud to partner with our friends at Cloud Flare on this. We do these episodes in two flavors.

What you're watching here today is the prerecorded panel, uh, flavor of it. But every other show or every two shows, we do a live round table, uh, flavor of this show where we ask the audience to get involved. You get a chance, do check out on our website.

com, um, when the next live, uh, event is because it's really a lot of fun if you're into these kinds of topics and you wanna participate with our panel of experts and our audience talking about some of these very, very important issues. Today's episode is consolidating your network architecture, right? Key strategic considerations for network consolidation.

You know, and as the cloud has evolved, so has our network consolidation strategies evolved. But, you know, for a while, well, we'll get into it. Let me introduce you to our panel and then we can get into it.

First of all, joining us from CloudFlare is Mike Ripley. Hey, Mike, how are you? Why don't you say a little bit about yourself?

Hey, Alan, thank you very much. Nice to meet you. I'm a product and work product marketing at CloudFlare, and specifically in our network services area, and I'm glad to join the conversation today.

All right, Mike, nice to have you on here. Our next guest comes to us from a recommendation by my good friend John Willis, and anyone that John Willis would recommend is good enough for me. Uh, he was also a business partner with John Willis, though, and a company maybe he'll mention and tell us what he's doing now.

It's Brent Salisbury. Hey, Brent, how are you? Hey, yeah, nice to meet everybody.

Um, yeah, so I came from, uh, I've been in networking for like 25 years. Uh, been from, you know, started out in ops and then I've transitioned, you know, transitioned over time and development. Uh, did a startup with John called Cycl Plane.

We got acquired by Docker back in the day. Uh, right now I'm at Red Hat, uh, just writing code, loving life, and, uh, enjoying, you know, this transition to ai. So, uh, last year spent a lot of time writing a VPN service, so I love the topic.

It's gonna be exciting. Absolutely. I could think of a better person and to have us here.

Thank you, Brent, for joining us. And then lastly is my cohost for the last great cloud transformation series. He's FU and VP analyst for, uh, DevOps in cyber.

Mitch Ashley. Hey, Mitchell, how are you? Hey, Alan.

Always good. Always good. You know, I'm just gonna drop the, I started in the ATM re relay days, so we go back a little ways in this whole cloud journey, so it's amazing how far we've come.

Absolutely. ATM frame. Yeah, I remember those two.

Um, you know, I look just a full confession. I first got involved in networking and data centers in the late nineties. Uh, we, I had sold my hosting company to what we called an A SP back then.

So these are the days where everyone resold the same T one line. We were big shots. We had T three lights, right?

Crazy. And, and we were buying our bandwidth from this little old company down in Texas called Enron. And, um, yeah, I could see if we have beers one day over the show.

I'll tell you about my experiences that the building with the big E outside and, and their arbitrage bandwidth, uh, yeah, it was, it was crazy. But that being said, the, the, the networking capabilities of data centers at that time were infantile, right? And, but I've seen so many things, so much water flow under this bridge, like bits through a fiber wire, a fiber, uh, piece of fiber.

So it's gonna be an interesting talk. Um, you know, I'll, I'll start it off with this 2005, 2006. I first, you know, this whole cloud thing burst on the scene, right?

Everybody's, we're moving everything to the cloud. IBM's running commercials with grandma's pointing up saying, uh, my pictures are in the cloud, right? And she's pointing at the clouds.

I think she really thought they were up there like in heaven. Um, but for many of us, we thought, wow, this might be the end of networking as we know it, right? We no longer have to worry about that land, land.

We never, we no longer have to, you know, we're gonna move away. The de parameterization of security, the, the moat and castle approach to how, you know, traffic comes in at ingress and egress, it's all in the cloud. And aren't these cloud guys going to do something about making networking easier, famous less words.

Brett, you were there for that, right? You, you lived this. Give us kinda your, your take on those early days.

You talk about AI hallucinations, I think we're all hallucinating, but get, yeah, right? Uh, I mean, for unfortunately, I get a chance to buy, uh, buy t horns from Enron, though. That's, that's an amazing story.

Oh, might I keep stories? I need, need a lawyer to hear anything else though. Yeah, we, we pretty much did.

Um, yeah, Yeah, yeah. To Point we'll into MCI, anyway, go ahead. Yeah, Yeah.

So I probably ripped out more ATM than I installed, but, uh, yeah, that's, uh, that's some fun stuff right there. So, so the, the problem has always been, so back in the day, right, like, there was always a manager on the hook, there was a lot of responsibility because you're managing, you know, every layer of the stack, all the way up the cloud comes along. And in my mind, pe people don't want to carry pagers.

People don't want the hassle. People don't want the responsibility. So they say, let's ship this off to the cloud and everything's gonna get fixed, right?

Turns out that it's expensive to do that too, and need lease lines and new try to, to providers. So, you know, nothing is easy, nothing is simple and nothing is fast. Um, but what, what the fundamental change that came about, obviously AWS pioneered a lot of this was the user experience.

So that user experience is, is in incredible from, from all the service Cloud player does an amazing job on it, you know, do a little shilling there, but they, you know, they do like every, every, every page runs through that. So, so it allows the, you know, this, this, you know, network manager to like, not carry, not, not lu the pager around and not, you know, have to carry the, the burden of the responsibility. Uh, the problem is you still have clients and you still have on-prem stuff, and how do I manage on-prem, private cloud to public cloud regulation, PCI, hipaa, every, everything that goes along with it, it gets really tricky.

So at the end of the day, I think we have salike services that, that are amazing, and, and those just need to be decoupled from the underlying infrastructure. So you mentioned endpoints, like we, we start creating these overlays, uh, and, and we, we do create this kind of ubiquitous environment where we have microsegmentation for security. We've got, and, you know, analytics, we've got visibility and everything.

Problem is you get this complexity stack that's huge, right? So when everything's abstracted, you have no idea what's going on, or, and I think that's, I think we're running into a wall on that. Like, that's, that's a fundamental problem right now is it's so complex to manage networks because you need, everything's so intermingled with the application now because application developers came along and said, Hey, I don't need a network.

I've got this host on the endpoint. That's not my router, that's not my switch, that's not my load balancer. So now you have what used to be a dozen, you know, load balancers, you've got thousands of load balancers because every node is not, so it's just, it's a massive management problem.

It really is. Mike sounds like a, a good, uh, setup for why you need something called, you know, the connectivity cloud, if you will. Yeah, I mean, again, like, just kind of talking, what Brent mentioned is that's a pretty common thing we hear when talking to different organizations is just the fact that, you know, as customers of transitioned from sort of dealing with on-premise, moving into the cloud, networking in the cloud, you know, what does that look like?

And being able to have not only the, you know, sort of the connectivity, but also the visibility into what's going on, or if there's a problem, where's the problem happening? Is it between my user on their device? Is it between the device and somewhere in the middle mile, right?

Is it in the cloud? Is it through the ISP or is it even somewhere at the end destination? Is it a specific application server?

You know, something else that's happening even beyond that, right? Because to your point, Brent, right, there's a load balancer, but that load balancer is just front-ending a lot of, you know, application servers, databases, and other things that you may not have full visibility into. So again, being able to have that sort of seamless connectivity, but also be able to understand like what's happening, you know, at each specific step in the sort of the path is, is something that we hear a lot from customers and something, again, it's a, a pretty common problem that everyone's trying to solve today.

You know, I think part of the transition that we're making is really, we've continued to do kind of the overlay approach, right? You know, speeds and feeds is we used in protocols, what we talked about in networking all the time. And then point to point, how does all this interconnection web of whatever that star top, you know, topology, whatever it looks like, uh, rings, you know, all these things that, that were, that are part of networking, but putting more and more of that on that kind of thinking in our, into our networks.

And cloud not only presents another layer on top of that where we do VPCs to interconnect everything and kind of do our same sort of network again, just in a different medium to now we're pushing things to the cloud that are networking, right? We're doing edge processing inside the cloud. We're doing managing content distribution and workers for our applications in the cloud.

Um, we have Kubernetes moving into the cloud at the edge, at the core. Um, so, so the cloud's changing in and of itself, but we're also changing how we're thinking about networking. I, to me, that's what's fundamental about this is let's not think again of an yet another overlay.

And I'm not criticizing what we've done that's, that's gotten us here to where we are. But the, those overlays are also part of the, this complexity that we're talking about. And what are services that we can use that help us that are multi-cloud, cross-cloud on-prem and cloud that help us not just abstract it, but deal with the complexity that's underneath the abstraction.

Agreed. You know, what kinda gets me going though? You know, I, I get the, uh, it's, you know, it's chaos principles, I guess, but every time we do something that we think is gonna make things simpler, it gets more complex, right?

I don't know if we just don't think it through, or that's just the nature of the beast, right? As the entropy kind of curve, right? Right.

As you go forward, things go more complex, not least less complex, but certainly cloud consolidation and networking has grown, grown, complex, complex. Um, let's talk about this animal called cloud-based networking, though. 'cause I think, you know, I think we understand what it is we're, we mean when we say that I think we do, but let, can we put our thumb on that?

You know, Mike, I'm gonna ask you your CloudFlare here. So this is on you, man. What do you, what do we mean when we say cloud-based networking?

So, uh, what exactly are you talking? Yeah, I, I think, and this is something that, uh, to your point, I think if you talk to five different people, you might get five different answers in terms of what they mean for cloud-based networking. So our sort of view on this is essentially just saying as a, you know, as an organization, you are trying to connect sort of your corporate network to services, you know, applications and things that exist in the cloud, and being able to do that in a very seamless way so that, you know, if it's users connecting to this, is it resources?

You know, again, it's taking that, again, cloud-based networking concept. But again, and to Mitch's point as well as you start to think is like, we're taking a lot of this sort of similar constructs that we've had on premise, and now we're moving 'em into the cloud and then trying to, you know, deal with it that way. And it's created this unfortunate, again, to Brent Point as well, is just this complex mash of services meshes, um, you know, different services that are all trying to, you know, a customer has to understand, an organization has to deal with.

And so every cloud provider has their sort of, um, their type of way they deal with it, whether it's, you know, networking VPCs or if it's offering some type of service on top of it to be able to manage it. And of course, you start to think about networking. We haven't even started talking about security, but it's funny because as you start to talk about security, it's like there's still the underlying network that sort of, you know, brings it all together.

But again, it's something that with cloud-based networking is how we connect sort of users and resources to applications and services that are existing in the cloud. That's, that's our view on it. So, so Mike, I, I, I think you are right there.

Um, the, the thing about it for me is, you know, when you live in a cloud native kind of world, there are so many different aspects to what we talk about cloud-based networking. There is the, to Brent's point, the end user endpoint guy person, right? And, and how is he connected in and accessing?

Then there's the edge stuff that, you know, and so, but that's kind of normal, right? You go from core to edge to endpoint, then, you know, we, we, the world turns inward in this container based world, and we look at container to container networking, right? Within let's say a Kubernetes system or something.

And then how does that attach to storage, you know, data, um, uh, identity and access, and how does that play into this? I mean, there's just man, it, it gets, it gets complex really, really quickly. And I, I don't, you know, kudos to CloudFlare for tackling it rent.

I know, you know, the folks at Red Hat working with AWS and a lot of the stuff that goes on tackle this, but it, this just screams to me as like, we gotta call out the calvary, right? We need, I don't think any one company tackles this. Th this is a, an ecosystem kind of issue.

Yeah. I mean, In my mind, Yeah, absolutely. Maybe you think about it, there's entire market segments that have popped up to, you know, sort of address these types of issues, whether it's, you know, application performance monitoring, network performance monitoring, you know, end user monitoring.

And, you know, it's just, again, it's, it just to your point, goes to show you how difficult and how there's not one approach that's solved all these problems. It's different markets and different companies and organizations trying to solve, you know, different problems. 'cause again, it's just the complexity of it all.

I mean, I love, I mean, maybe this is, you know, I don't wanna get ahead of myself, but maybe something AI could help with potentially just with, you know, being able to read information and be able to do a lot of the sort of, um, I'll call 'em, you know, trivial tasks that maybe a network person has to do when trying to troubleshoot. But, you know, instead of, you know, maybe augmenting some of that work, but just, yeah, it's, uh, again, it's just something that's very complex that I don't think any one company is gonna be able to solve anytime soon. At least not any Part of, part of it, Mike, is we went 15 minutes, and so we mentioned ai.

That's pretty damn good. Bing, bing frazzled. Yeah.

The, the duck should drop from the ceiling, right? There's an old reference. Um, so to, to your point, Mike, it's, it's also that it's complex, but also the scale of it is very large, right?

I mean, I'm talking for a small business, but you know, for any medium and large enterprise, of course, it, it, you can't throw people at the problem anymore. It isn't, you know, how many people can respond to how many alerts, how fast, right? It's now how do you diagno diagnostically determine what's going on?

Uh, which I think to your point about AI is, you know, that, uh, holds hopefully a lot of promise to helping this, helping this situation. I'm curious, Brett, you mentioned you're, say you're spending a lot of time working, doing development, working with ai. Is it networking related or anything that kind of helps with this problem at all?

Yeah, I mean, it's, It's everything related. It's all domains, right? So I, I, I think there, there really isn't any, any technological domain that it doesn't impact.

So networking is obviously on the forefront. I mean, if we think about, let's just take network data and logging alone, there's probably more logging and then the network than there is in your entire ERP. So it, these, this is like generally unstructured data depending on what it's coming from.

So it, it's just not fit for human consumption. So if we can figure out how to classify this to some degree, it helps, but it really is like, I, I, I think the future is probably a pipeline that's fine tuned into, into your logs, and then that output in that model gets thrown into, into an analytics pipeline. So you also mentioned like security, security is networking now.

Like there is no like standalone, like we're not gonna Innovate. Well, this photo was always networking. Yeah, because when I came in, Mitchell and I came into security, it was, it was all network security.

That was security, security, man. But I digressed, I'm sorry. Go ahead, Brett.

No, it absolutely is. And so, so what cloud-based networking has done is provided tooling around security that can be automated. So instead of having this, so we used to have north south traffic flows, right?

95% of your traffic, this is an old, uh, Facebook numbers, 95% of the traffic stayed, left the data center and 5% stayed east west. So 95 north south, 5% east, west. That flipped on the TED probably about a decade ago.

So it was like 95 is east, west, 5% is north south. So when you, when you get into those tradition, the traditional security on per miles, you have, you have these big iron stack of million, you know, a million dollar stack of firewalls. Now all your traffic between your security zones gets pipelined through those.

So, so you wound up with these, this traffic trombone that's completely inefficient. It's expensive to do. So then we said, okay, well let's just turn every edge endpoint into a switch.

Let's turn every edge endpoint into a router, into a security device. And so now with that, unfortunately Linux gives APIs to do that pretty easily. So like that, that's like Linux is fundamentally like the disruptor, in my opinion.

A lot of this in the early, like 10, 15 years ago. 'cause it gave, it gave APIs. You can go in and, and do the same things that your networks, routers, switches, firewalls, load balancers were doing.

You can do that programmatically. The other oss have come a long windows. It's pretty good there.

Uh, OSX is terrible, uh, but it, it, it, there there are at least APIs. So how do we take that? So you've got these endpoint users and got servers.

Ideally you've got some ubiquity across that where everything can efficiently get to the right endpoint without creating, you know, this tromboning effect. We've also got data gravity that comes with the mix. So, you know, you've got tons of data on the edge of your network.

Do you really wanna ship that to the cloud because, you know, getting anything out of the cloud costs money. Uh, so, so thinking about traffic patterns that, that might heat on us is important. And then like fundamentally, it's just time and space too.

We're never gonna get faster than light. That's just what it is. Because maybe someday it'll quantum obviously, but that's probably dead by then.

Uh, so all of those complexities create this, this natural leaning towards, uh, an overlay of endpoints that, that can talk directly to each other in a secure fashion in a v VPN like fashion. If you were to ask me, like, what is the panacea? That's probably it.

And that gives you the identity. Like, and it used to be, you'd never know, like the only way you knew who the user was, was like looking it up, active directory. Like, oh, here's IP for that.

Man, it's terrible. Like everything has to be identifiable real time. But, and I, I think we're trying, we're trending toward that.

And a lot of it is like kudos to like system administrators that got into networking because they, they took it from another lens. And the long term is, the problem with networking is it always has to be up. You can't take down networks and experiment, right?

Because now you just lost your job because you create an outage. So it, so, uh, if you say, if you, you know, everything moves slow until you know, it eventually changes. Problem is you, you just can't go in and change networks without like massive disruption.

See, there's tons of CapEx that goes into that redundant there. That was a lot of, that was a lot of topics, but it's like you said, it, it's, it's a big basket. Well, Mike, you're, you're, you're making, I'm sorry, Brent, you're making me long for the days where our biggest problem was we're gonna run out of IP addresses and how do we get to P six, Right?

And, and now you just how My things have changed. Yeah. And now you just throw 10,000 nodes into the same sub that there is no broadcast of main problem.

Exactly. On the cloud, you do that OnPrem and you just melt it down your network until you've gone and pulled every redone fiber link in your data center out to break the loops. So, yeah.

Oh my, you know what? So be thankful we're coming. It's Thanksgiving.

Be thankful. At least we, we've solved that issue. Um, so we, we, we talk about network and, and to me it's almost oxymoronic, right?

That at the same time we talk about consolidation, we're constantly expanding. You know, it's, it's this, this thing going on here. And, um, but for folks out there saying, Hey, we're moving to the cloud, the cloud is where we've gotta be.

Yes, there are the repatriation folks who are trying to move back, you know, and finops and all that good stuff, cost of cloud. But overall, I, I still say there's more people who are say, you know, go west young man, we're moving to the cloud. And when they get there, they realize, well, it's not just a cloud, it's multi-cloud, it's everywhere cloud.

Oh. And then they say, oh, wait a second, we better do a cloud consolidation kind of project here. Right?

That's step two. Um, what would you, how would you advise people out here? What, how do you even prepare for consolidation projects?

What, you know, what are the, what are the telltale signs? What are the steps? Any advices appreciated?

Brett, you are, you're the, you're the, you're the, you know, seasoned vet here on this. What do you recommend to people other than, you know, go hire someone. No, I love it.

It's the exact, it's the exact right question. Uh, and so my answer is, it starts with application. You have to make applications portable.

Like that's the fundamental thing that you, you should be able to tear down your application and spin it up somewhere else within seconds. And so until you get to that point, you're, you're just, you're, you're trying to like do this giant lift and shift with this traditional legacy monolithic app that's just gonna be a disaster migration. So like, you start breaking things down into composable pieces.

You introduce complexity with microservices, but you introduce the ability to move to, to have that portability to where it's not tied all together. Like, once you have that, now you can say, well, if I can move my application from one side of the data center to the other, to a completely different IP address, why can't I move it anywhere? Now I can start leveraging, okay, start leveraging the cloud.

The next next natural transition that you mentioned is exactly right. Like, okay, well we might be spending more here, but now what if we start leveraging multi-cloud, basically spot markets for clouds. You've got that ability because your application is not portable.

Your, your security policies are portable. Your visibility is portable. So like, you can start doing this inside of your data center today.

Once you're prepped for it, then you start moving it off to, you know, shipping it off to clouds. And then you can even get into multi-cloud and start leveraging, uh, you know, prices and, and negotiating better prices on your, your monthly spend. So yeah, I, I personally, that's, that's like the, the, the sweet spot is getting started with that, whether it's microservices, whether it's VMs, whatever, as long as it's portable.

Like you, you're on the right track, Mitchell, you've did, you've done some transformations and consolidations in your day. What do you think on that? Anything you could add to it?

Brent? What out there? Well, I, I go back to the thinking that got you here is now the thinking that's gonna take you go forward, right?

So what's changed and what do we need to kind of either reconsider or, you know, things we might have done in a small way we wanna do a big way? Um, you, you mentioned Brent portability rather than networks being all things to all people, which is what we do. Like we'll take any traffic and put it anywhere it needs to go, which is a really tough problem to solve.

And you, we've been saying that, um, the idea of workload portability, you know, I think of it as workload, Mike, you mentioned it could be v it could be, you know, microservices could be VPNs, or excuse me, VMs, lots of things. But I think, I think we have to think of networks as put the workload where it's best to have that workload perform. Now that's, you know, a bit of a panacea to, to say it's that easy, but it's not.

But that's, I think, the error that we're moving into. And maybe that's part of what AI will do, is help us with determining where the best place to put that workload is. So maybe I'm pushing it into, into the cloud itself to happen as opposed to a, a cloud provider or like a hyperscaler or OnPrem, wherever that might make sense, because, you know, in a separate domain in this conversation, the world is going to Kubernetes everywhere.

And that's, that is becoming the workload, kind of back plane of networks, of software, of infrastructure, of all things. And it might be a new thing to networking folks, and that's fine. I'm not saying you need to go out and become a Kubernetes expert, but I think you need to start looking at it and understand.

'cause that is kind of our workload portability platform of today, and maybe it'll evolve to be in the future. So I think it's, you know, more than software defined, it's software run and managed, you know, through platforms like Kubernetes. I, I, I do think that the whole cloud native architecture has made it easier at some level to, to do this consolidation, right?

But you've gotta be, I, I think as Brent laid out, you gotta be committed to the stack, right? You're going to, you're gonna go micro, uh, you know, microservices, you're going to go distributed with Kubernetes and containers. You're going to, you know, take advantage of the, of the topography, if you will, to, to run your best app rather than this, you know, monolith that, quite frankly, why would you, if you're gonna run it like that, why put it in the cloud at some level?

Almost, you know, Alan, one of the things that, that two cloud native has going forward is networking. People already understand it. They just don't realize that yet, or maybe not had a chance to look at it.

We, we talk about microservices and containers and things like that. What cloud native really is, is smaller pieces of code talking over a network, whether it's on a cluster in a Kubernetes cluster, it's in a data center, it's across data centers. There's security of those communications, you know, over APIs, the same kind of, you know, GraphQL and, and, uh, you know, restful, uh, interfaces that, that we use on our networks today.

So it, it's, it's not an unfamiliar approach, it's just taking it to a whole new granular level. And if you think about it that way, it takes a lot of the mystery out of cloud native, I think. Greed, greed.

Um, guys, we only have a, I believe it or not, a few minutes left. It's amazing how quick the time goes here, but Mike, I wanted to ask you specifically, you know, we didn't really get into connectivity cloud as much in this episode. We've covered it in previous episodes, but really, how does connectivity cloud here help us with these consolidation projects?

Yeah, absolutely. I mean, I guess the best way to put it, as you start to think about the way, again, we view this is that we really put it into like three different buckets. It's the connectivity cloud is sort of what stitches everything together.

But inside of that, the big things we hear customers asking about is, help me modernize my network, help me modernize my security and help me modernize my application. So depending on sort of what part of the journey the customer is on, whether it's like you said, network consolidation, or maybe they're looking to move to the cloud, and how can CloudFlare help in that with applications, or if they're starting to explore things like zero trust and VPN replacement or secure access, that's really where we can kind of come in and help them. And again, there's lots of little pieces that we do and lots of different solutions, but again, it's really about just creating that seamless connectivity between users and applications or applications to applications.

So that's, uh, my quick sort of, uh, elevator pitch on the connectivity cloud, Brent, are those the three things we focus on? Security application? Go ahead.

Yeah, And so, and so I totally agree. I mean, ZTN is is, It is a crucial piece To this. You know, everything has to be filterable, everything has to be visible, everything has to have an identity attached to it.

Like things just can't talk will anymore to each other anymore. And we've got, we've got mechanisms to do it. Crew is obviously just a bedrock of that at this point.

Managing applications without it, it can be painful getting everything converted into a declarative configuration, but once it's there, it's there. It can be redeployed at any time. It can be redeployed everywhere.

It's, and you've got tools there for, for know, taking a peek into what the application's doing. I think the, one of the things like the, the practical on the ground for, for NetOps, we'll start automating everything. Learn automation, just, uh, uh, treat your network as code, implement ci, because if not, you're gonna, you know, you, you've turned yourself into your code into a chaos monkey that's gonna go take down your network.

So, I mean, it's a dangerous tool, but it's important one, uh, so, you know, like no more manual changes. Everything needs to be automated. Work that into your change controls.

And I, I think once you're there, then, then the rest starts to fall into place because it's just another API call, it's, okay, well now I've got these APIs at this cloud, now I'm got with this cloud, and I can start, uh, transitioning into that. So like, thinking about network as code is probably a really good first step too. Yeah.

Mitch, I'm gonna give you the last word we got maybe a minute or so too. Yeah, I think we, we've explored a lot of territory on this, and I think the, the main takeaway for me is there's a lot of people working on this trying to figure out how to best not only do it today, but position ourselves so that we can evolve our networks. And maybe the network isn't a static thing, it's an, it is just constantly changing.

Kinda like our software is, it's gone from a static thing that we could point to it and say, that's version X, and it's been that way for three months or 30 days. We're kind of moving. The same thing in our, in our networks, in our cloud, is that, you know, it isn't, it isn't the same way.

Matter of fact, software is changing it for us. Maybe someday AI is changing for us, but the automation is a big part of it too. So sometimes just taking those constraints away helps us think about, okay, now if that's gonna, that's going to be true, or maybe it already is, how should we solve this problem?

Fair. I, I'll end it with this. You know, when we first entered the cloud era, it was the era of lift and shift.

I took my application, I was running in my data center, and I, you know, ran in a vm ba basically in the cloud and didn't work so great to tell you the truth. I mean, you know, we, we, we soon discovered not a, not a great use of the cloud. And, and so we embarked on this era of transformation, and, and along the way we discovered the need for consolidation.

And I think we can't lose sight that these two things go adjoined at the hip, right? If you're going to go through a, a, a, a cloud consolidation, a cloud networking consolidation, that's probably the time to do your transformation, to take advantage of the medium, right? If, if you're, because otherwise, I, I think you're doing that chacha one step forward, two step back kind of thing, and, and it's just not synced.

And you're going to, you're only setting yourself up for more heartache and work, you know, going forward. So my advice is, you know, preparing for consolidation, product projects, they're transformation projects. And that may be the time to go to a, a, uh, a, a cloud native kind of, uh, a stack, you know, maybe time to change that monolith into multithread right in, into, uh, microservices.

Um, right? Because otherwise it's kinda like the technical debt issue. You just, you just putting off to do something tomorrow that you probably need to do today if you really wanna do the consolidation and transformation, right?

So that being said, hey Brent, you know, John Willis said you were a genius about this stuff. He wasn't lying. Thank you very much.

I appreciate you coming on and, and joining us today. Good luck in your role there at Red Hat. Mike, Mike rip leads, great to have you on here, man.

I appreciate you coming in and, and lending you some expertise. Mitchell, as always, pleasure on behalf of, uh, CloudFlare and Techron, you've just watched another episode of the Last Great Cloud transformation and we hope to see you soon, especially maybe at the live round table where we can hear from you. Until then, though, this is Alan Shimmel.

Have a happy Thanksgiving. Happy holidays everyone. We'll see you soon.