Truly Cloud Native SBCs and voice services with Mark Cobbold at Ribbon

Mark Cobbold, Vice President of Product Management at Ribbon Communications, explores the evolution of telecom networks from hardware to cloud-native solutions.

• The benefits of cloud-native deployments include reduced operational costs, improved software velocity and robust network security
• The variety of cloud deployment models and how they cater for different business needs
• Using real-time insights into network performance to help pinpoint and resolve issues efficiently
• Ribbon's innovative approach to cloud architecture, enabling flexibility, security and operational efficiency

Thanks to Ribbon, this episode's sponsor, for their continued support.

Transcript

Copy Transcript Copied

Mark Cobbold: This is an architectural change in the way software is built and developed and maintained. We have Teams now that owned a microservice. It's a reusable piece of software that can be used in a different cloud native function. It might be a logging utility or it might be a call processing utility. And then it has hardened interfaces.

It can be security patched, upgraded independently of others. And possibly the most important is it can be scaled out independently.

Tom Arbuthnot: Welcome back to the Teams Insider Podcast. This week we have Mark Cobbold who's Vice President of Product Management at Ribbon. Mark has a history, a long history in our space, at Nortel, then Genband, and now Ribbon. And he takes us through what Ribbon are doing with the new approach to cloud architectures and cloud native software in telco.

So that might be on public cloud, that might be on private cloud. You might know, I know Ribbon as a large physical SBC provider, but they've really been innovating in the cloud space. And Mark talks us through what's happening in telcos and also in enterprise. So many thanks to Mark for taking the time to jump on the podcast and many thanks to Ribbon for all their support as well.

On with the show, Hi everybody, welcome back to the podcast. Excited for this one. Uh, this has been a few months in the making. Mark and I have been at various shows together over the last six months or so, um, and excited to get into his role at Ribbon as vice president of product management and also what Ribbon are doing with kind of their approach to cloud architecture.

Um, first off, Mark, welcome to the show. 

Mark Cobbold: Yeah, thanks, Tom. Thanks for having me. 

Tom Arbuthnot: Awesome, thanks for coming on. So, um, I'm really keen to get into what you're doing at the moment, but first off, you've got quite an illustrious history in our space. You've got Nortel, you've got Genband, and now Ribbon. Can you just give us a little bit of the journey stories to how we got here?

Mark Cobbold: Yeah, yeah, um, sure. Way back I started in customer support and then I moved into R& D at Bell Northern Research years and years ago in wireless. I did wireless software development and then I moved to Europe. I ran our R& D group in Europe for about four years, lived out of the UK, had two kids while I was living there out of the four. And then, you know, Moved back into, uh, into our IP or an optical group in, uh, in Ottawa, Canada.

And we, uh, we, we are, um, I ran R& D for our, some of our optical transmission products. Then I moved into our cloud team, voice cloud team, and, and took on product management and business development. And, and basically landed here now, uh, running our portfolio of what we call Cloud and Edge and our voice products, security products for voice networks. And that's that takes you up to the present day. 

Tom Arbuthnot: Awesome. 

And, and loads of people will know Ribbon for, you know, like traditional physical gateways from, from, you know, from edge physical gateways that I've deployed quite a lot of into the big chunky carrier stuff. Um, but talk us through this whole kind of cloud native approach and cloud native journey.

What's that mean? 

Yeah, so, so, uh, you know, telecoms networks were where time division multiplex or TDM networks typically be the enterprise or service provider. PBX networks connected in the circuit networks for voice. And as things have evolved, networks have gone from physical network deployments where they have gateways and soft switches and PBXs to virtualized software.

This is various virtualization platforms and now evolving from there is really a more automated approach with a fully cloud native operations model. 

And this seems to be a big focus of rooms at the moment. You were quite early on with things like, you know, Kubernetes, Docker, microservices, like you've been doing this for a while now.

Mark Cobbold: Yeah, we, we've been investing quite heavily in, uh, cloud native software to help our customers drive down operations costs, uh, improve software velocity, um, and improve the security of the networks too. Very important, um, growing trend in networks today is, uh, you need to secure your networks. There are a lot of, uh, security threats out there, and, uh, and there are, and it's a growing problem as time marches on too.

So we've got a strong emphasis on innovation and automation of continuous integration and continuous deployment pipelines. So we've got, we've got working in deployment or in production networks now, both, uh, uh, Mostly in service provider and now the enterprises are really transitioning. 

Tom Arbuthnot: And it's when we talk about these things like containerization and cloud, like my mind and I'm guessing a lot of people's mind immediately jump to large scale public cloud to your Azure, your AWS, your GCPs.

How does that play with this and how does that play with the carriers? Are they running it on public cloud? Are they running it on private cloud? How does that work? 

Mark Cobbold: Yeah, I know a lot of, a lot of, uh, the connotation is it's public cloud, but really cloud means more than just public cloud. Many carriers and enterprises run private clouds, have their own infrastructure, uh, and so a cloud native deployment can mean in a private network, a private data center.

or a public, public, uh, cloud network, um, or a hybrid. And, and it really, it really depends on the situation and the business needs. You may want to expand internationally. It may be easier to expand in a public cloud to get started. Then you may want to consolidate. There may be, there may be regulatory reasons you need to build a private cloud as well, or, uh, you know, solve some of those challenges in a public cloud.

It really depends on that business need, how rapidly you're expanding, maybe outside your own data center footprint, or, you know, then consolidating maybe. 

Tom Arbuthnot: And do you see customers doing a mix of all of the above, depending on their use cases, or do they tend to stick to one type or the other type? 

Mark Cobbold: Um, you know, different, different organizations have different, you know, business drivers to select, you know, which is a public cloud deployment.

And, and often what we've seen with many of our customers, they might need to expand rapidly in multiple, multiple countries. And that can often be easier in a public cloud deployment. So we've built our products so they're flexible and they can, they can be, uh, repatriated back into a private cloud later or, uh, or, or a hybrid scenario where there's a bit of both in the network.

You might have a core network in a, in a core country and then expand in a public cloud and, uh, in a satellite country. 

Tom Arbuthnot: And what does, What does manageability look like there? Is that they run the same management? They're different? They're isolated? What does that look like? Say I've got some on prem and some or some private cloud, some public cloud.

What does that look like? 

Mark Cobbold: Yeah, we try to make that whole experience look as common as possible. In that CICD or that, uh, con continuous integration. Continuous deployment pipeline. So you, you use the various automation tools. We have test automation tools. Uh, these are common tools that work, work with, um, with, uh, the private cloud solution or the public cloud solution.

And, and really the, these, these tool sets are really coming from the IT domain. And we're, we're adapting them into the telecom domain. So there's, there's a, there's a skill set that's being built up in the, on the IT um, side of the industry. And, and that same skill set's portable onto telecom's networks now.

Tom Arbuthnot: And, and that's really interesting 'cause obviously I kind of crossed the boundaries of those worlds, like things like CICD and, um, you know, Kubernetes and containerization. I, I feel like, uh. Dare I throw a slur at my, my telephony friends, like traditionally, like continuous, uh, you know, patching and development and, and keeping things up to date.

Often it was like, we've got it working now we're going to patch it once a year, but that's not like you mentioned this, it's not really the world anymore, is it? 

Mark Cobbold: No, no, and there's many drivers, um, you know, the complexity of the networks is growing as time marches on. It becomes harder and harder to upgrade the software.

And the longer you're on the old software, sometimes it feels like it might be easier for a while. But you're, you're really creating a deficit in the future that gets very hard to dig your, dig your way out of. And, and the security threats, um, NIST, the, uh, the, uh, the organization that publishes known security, uh, vulnerabilities.

These are, there are 30, 000 a year that are coming out on various, uh, uh, across the industry. And, and you gotta, you gotta keep up to date with these or you can have serious, uh. You know, critical security vulnerabilities in your network. 

Tom Arbuthnot: And how does that play down to, like, the, the enterprise customer?

Does this, does this apply to them, even if they're deploying, you know, I mean, a few thousand seats potentially at some of your, you know, I know you scale to all directions, but does this same technology apply there? 

Mark Cobbold: Yes, it does. Um, many, many, uh, of our products and, and, um, The industry standards are using Red Hat Linux, uh, you know, uh, operating systems that are public domain, um, and these have, uh, security vulnerabilities that need to be addressed or these can be attack points in your network, uh, for, for, uh, you know, hackers to, to enter your network and then maybe get it more valuable pieces or get into user, user data and, and, and the ability to, uh, to crack into critical systems.

Tom Arbuthnot: So if, uh, let's say somebody listening might be, um, in a, you know, mid sized enterprise, traditionally they've had physical SBCs, maybe, you know, gateway slash SBCs. What, what does the journey look like for them to get into this kind of new world of more cloud native virtual containerized? What does it, what's the journey look like?

Mark Cobbold: You know, you know, often it comes from looking at. Uh, their overall picture with their IT assets and workloads that they have deployed, along with their telecoms, network assets, doing an assessment, seeing what, what technologies work best, uh, be it public or private cloud, some of the, the platforms available out there, we, we support numerous platforms, you know, Red Hat OpenShift, for example, um, AWS, others.

Now, now then, then, then we, uh, we typically then work with customers to show them some of the tool sets, like our test automation. We have a capability that can capture call flows and generate automated, in an automated fashion, a completely automated test suite. And, and with that, you can fill, you can, uh, build that into a, a, a pipeline and we can start trials and, and get.

You know, get, build confidence with the teams working on these technologies and, and, and helping them, you know, support their users in a transition. 

Tom Arbuthnot: That's awesome. And, uh, I'm guessing they again could run a, a mix of these technologies as, as they transition to more of a cloud approach. 

Mark Cobbold: Yeah, yeah, absolutely.

Um, you know, many of the networks would interoperate. We would, you know, our products will interoperate with each other, be they the physical, uh, SBCs or, or, uh, call servers or application servers, um, and then with, with the virtual or the cloud versions. And then we would provide, you know, an assessment with that automated testing on whether it's.

It's really going to work when you migrate. 

Tom Arbuthnot: Yeah, the testing, we did actually a separate podcast on the whole testing platform and you can kind of copy and test theoretical config changes, can't you, and prove they're going to work or prove what the impact is. It's really quite impressive how accurate the testing can get.

Mark Cobbold: Yeah, and the challenge there is really, uh, In making your lab configuration look as much like your production configuration. And we have a, a, a digital twinning technology to do that and map the configuration between lab and production and therefore you, you can build a, you know, a test suite that it actually extracts from the pod the, uh, the, the production network real call flows and then uses machine learning techniques to narrow it down into the.

Unique set of call flows, and therefore, you know, the right set of test cases that gives you full coverage. 

Tom Arbuthnot: That's awesome. And can you tell us a little bit more about the analytics and visibility? I know that when we were at the last Insights event, there was a whole session on that. And it seems to be a combination of, you know, platform optimization and security is driving this real need to really understand what's going on.

How does that work and what does that look like? 

Mark Cobbold: Yeah, we, you know, very complimentary to our cloud native, uh, offering is, is the observability piece. So this, there's, there's, there's gonna, I I, I think of it two ways. There's one where you have the cloud native, the Prometheus backbone, you know, pulling in the, the various fault reports and things, uh, from a common microservices architecture where all your cloud native applications are, are architected to, to feed into that, that, uh, that back plane if we will.

The software back plane. The other one is, uh, where you take a lot of information, a bunch of different pieces of information. It could be call detail records, it could be call flow information, it could be fault reports. Bring that all into a common data lake. That's our Ribbon Analytics. Uh, platform. And then from there, you can, you can take various, uh, observability applications.

We have one called Most Probable Cause, for example. It can kind of take a, uh, a number of different pieces of data, correlate them, and come back and give you a, you think the problem is here. And I'll tell you, it's, it's this, this network element that's causing the problem today. Uh, you should go investigate that a little further.

Tom Arbuthnot: Awesome. And, uh, the, the question that nobody likes to answer, but our, our audience want to know about, how, how's this look commercially? Obviously not exactly, but is it like, because in telco sometimes it's session based, sometimes it's device based, um, uh, how does it scale? How's that work, Mark? 

Mark Cobbold: The cloud solution really frees up capital uh, by moving your investment into an annualized OPEX model, and it really is in line with, for example, Microsoft 365 model. Uh, you know, we have many of our customers saying they, they want to reserve their capital for more, uh, for projects that really generate new revenues for them. in, in their business, not, not their telecoms network that they use to run their business.

And so, so, so really it's, uh, it's an OpEx model, gets you into, uh, you know, a continuously supported, uh, rapid software delivery, new functionality as it becomes available can be rapidly deployed, and, and it breaks it up into smaller pieces, lower risk. As you, uh, deploy it in your network. 

Tom Arbuthnot: Yeah. And that, that's kind of what I was interested in actually, is the, the ongoing supportability and update.

So essentially you are much like modern software these days. You are, you are buying in and you're getting the continuous updates, feature improvements, and, and rolling those out hopefully more quickly because you have this flexible architecture and kind of the ability to do the, the CIs, uh, CD and the containers.

Mark Cobbold: Yeah, yeah, absolutely. Many of our customers, some of our key, key customers are looking at it. I can get a new feature out, rolled out into my entire network in three months. I can get a software fix in a week. Or no more graveyard shifts to do maintenance and upgrades and migrations. 

Tom Arbuthnot: Yeah, if it's a security issue, then you can't, you can't have three months anymore, can you?

Because that's not, that's not okay anymore. 

Mark Cobbold: That's right. You need to get that security fix out in days into your network. If it's a critical vulnerability. 

Tom Arbuthnot: Yeah. 

Mark Cobbold: You might have a business critical problem where calls are failing or, you know, things, things aren't working for your customers in a call center, for example, that you really need to get.

You know, repair it very quickly. And, and then there's the, you know, it's, it's, it's nice for my staff not to have to work on weekends or graveyard shifts to do maintenance or upgrades. Um, yeah, those are some of the objectives and things you can achieve with cloud native. 

Tom Arbuthnot: And how are you finding it with the telcos?

I know you work with a lot of big telcos. Like, is it, is it a big culture change? Are they already on that journey to cloud with their architectures? 

Mark Cobbold: I'd say it's mixed right now. Um, that there are, uh, there are customers we have who are, you know, who are very leading edge and have moved sometimes it's, it's because they invested in a physical network and want to skip the virtual network and go straight to cloud.

Right. Um, some of them have kind of eased their way in with virtual deployments and now are finding some of those platforms are expensive and aging and they need to do lifecycle changes on them. Um, and therefore. Uh, go to cloud and, and then, you know, some, some of them are very, uh, you know, very enthusiastic about the savings they can see they can achieve with the cloud native model.

Tom Arbuthnot: Yeah, I've noticed, um, working with some of the Operator Connect telcos that there's, it's interesting in telco, we've got increasing regulation need to have local in country, uh, kit or in country connectivity. And it seems like the, the idea that you can virtualize. SBCs in public cloud is working really well there because they're not shipping physical kit, but they are in region in the appropriate public cloud, and they have that flexibility of scale.

So they might be starting off with a handful of customers and planning to grow. They can scale the architecture. 

Mark Cobbold: Oh yeah. The, the, it, it, yeah. The, the, the, uh, the, uh, idea of setting up a new point of presence or, you know, setting up a new interconnect, that these are things that are, uh, that, that can be very rapidly deployed with a cloud native approach.

And you, you know, you bring along all the, the tool sets with that, the, the automation tools, the CICD pipeline, then that. That can basically piggyback on your core network, let's say, if you're expanding, right, and it becomes a very common operational model for you. 

Tom Arbuthnot: Awesome. And you have a concept, just terminology wise, you have a concept of cloud native network functions.

So can you explain what that means? 

Mark Cobbold: Yeah, so a cloud native network function is a single function, but it's been broken up into what we call microservices. This is an architectural change in the way software is built and developed and maintained. We have Teams now that owned a microservice. It's a reusable piece of software that can be used in a different cloud native function.

It might be a logging utility or it might be a call processing utility. And then it has hardened interfaces. It can be security patched, upgraded independently of others, and possibly the most important is it can be scaled out independently. So that function requires more capacity, or it requires a variable capacity through the course of a day.

You can, it can be quite friendly, the overall application, the overall solution can be quite, um, Um, friendly on the usage of resources on your network, freeing them up and other times for other workloads on your network. 

Tom Arbuthnot: So it gives you that flexibility of they're available when they're needed, but not, not sitting there doing nothing when they're not.

Mark Cobbold: That's right. Yeah. And Kubernetes really is the, the way we're seeing that achieved, um, in virtual. The orchestrators never really got that all to work. Um, there was a lot of, um. A lot of angst in getting an orchestrator to scale out a VNF and scale it back and, and, and so many, many of our customers.

Tom Arbuthnot: Well, I remember our first gen of doing this stuff virtually.

Basically, we fixed the, the cause to the, to the devices. And it was like, it was technically virtual, but like, uh, it's the same with all the Skype servers the first time, well, Lync servers, we did that. Like, it was virtualized, but it was one to one hardware essentially at the start. 

Mark Cobbold: Yeah, bingo. And, uh. And the orchestration with the various VNFs, or as we call them, the virtual network functions, it never really, never really achieved that scale in, scale out, scalability, flexibility, and life cycle management efficiency that Kubernetes does.

It seems to be just a, just a, natively good at doing that life cycle management and the scaling functions. 

Tom Arbuthnot: That's awesome. And you guys obviously offer this as software, but you also run cloud service for some of your customers. Are you using the virtualized technology inside of your own service offering?

Mark Cobbold: Yeah, we, like, as an example, we have our call trust solution. It, it runs in a AWS and it uses, uh, all cloud native components. It, we have a CICD pipeline and we deploy, fix updates, security updates, everything into our cloud solution. We, we, we, we eat our, we, we, uh, we, we, we eat our own, uh, uh, or we use our own technology for that.

Tom Arbuthnot: Yeah. It's either dog food or champagne, depending on what you're, 

Mark Cobbold: Yeah 

Tom Arbuthnot: That's awesome. I appreciate you giving us an insight there. Again, like it's a, there's a whole, I got a lot from the sessions at the Ribbon events of how much of this has moved on to, you know, again, some of the more modern approaches to cloud technology and cloud architecture and containerization.

It's really exciting to see the kind of progression in that space. 

Mark Cobbold: Yeah, we're, uh, we're, we're, we're doing the same thing again for, uh, call it Ribbon Connect Plus, an SBC as a service offering, too, that be available in public cloud, um, or private cloud, and it'll be, uh, available to customers to, to, to purchase where they won't need, they won't need to build their own cloud infrastructure.

Tom Arbuthnot: Awesome. So, um, if you want to find out more, Mark, what's the best thing to do to find out more? 

Mark Cobbold: Best thing is to, uh, is to check out our ribbon.com website. Um, you can leave your details there. We'll get back to you. We've got all kinds of information on our solutions there as well. Um, or, or your local sales rep or your Ribbon sales rep or your, or our partners who, uh, who work with us.

Tom Arbuthnot: Awesome. Well, thanks, Mark. I appreciate all the support and offline conversations and, uh, thanks for coming on the show. 

Mark Cobbold: All right. Thank you, Tom. Appreciate the opportunity.