How open source can accelerate 5G, edge deployment for NSPs: A Q&A with Chad Foster, Red Hat

This interview has been edited for length and clarity.

Q: Could you touch on some of the trends around 5g and edge computing? I’d be curious to hear what you’ve been hearing from your customers.

CF: You know, right now it’s a very interesting time that we live in. 5G was on the radar for some time [but now] all of this acceleration is taking place. It’s more immediate, more urgent now than it was before. And a lot of this has to do with COVID-19 and the traffic demands that has put on networks. We’ve seen customers with traffic spikes as much as 25 to 60 or 70% this year.

And so you’ve got people working from home more in this digital economy, you’ve got this push towards the digital transformation – things that were going to take much longer to develop. Think about telemedicine, as an example, that didn’t really go anywhere for so many years, now all of a sudden out of necessity we’re seeing more and more emphasis because it’s the only way that we can do that safely, unfortunately. And we’re seeing more video content. All of that combined with the needs of applications to have low latency, and high-performance carrier-grade applications, whether we’re talking about using network slicing for e-health or emergency services and that is going to place demands on the network.

Q: How is that impacting network service provider (NSP) as they try and build out new services where the business model is also new and untested?  

CF: A big part of the push to 5G is obviously getting outside of the traditional (NSP) model and into moving from a model that was redundant in hardware to a model that is resilient in software. Using software defined networks to create the resiliency in the network at a lower cost, so that they don’t have as much vertical workload orientation and moving to a more horizontal model.

They (NSPs) can write applications once and deploy anywhere so they don’t have to write and rewrite; they can be more efficient with the developer resources that they have.

We’ve seen some scenarios where customers are interested in revenue sharing models (with Red Hat) to help out with that. There has been, an increased focus on revenue sharing models to where customers need to figure out “How do I advance my future while at the same time not really knowing what sort of monetization I’m going to have in the future?” Those are things that we’re seeing get more and more emphasis within certain customers that that is probably the most common that I’ve seen thus far.

It’s still pretty early innings on this transition. A lot of our larger customers have started the journey but it is a lot of work just to get the 5G core built and then out into the RAN. We’ve got some early, early adopters who are definitely pushing that quite a bit. The current economic climate puts a strain on customers but they’re able to save money in a horizontal model versus a vertical model and using software that is resilient versus hardware that is redundant. We’re able to generate some savings for them so that they can continue to fund the innovation that they’re going to need for tomorrow.

Q: Is your work with companies in the telco space largely a North America equation or do you have discussions with folks in Europe and Asia Pacific as well? Are there any differences in what’s happening across the different markets?

CF: We’re in active discussions both in Asia as well as in Europe and in North America right now, and in Latin America for that matter. So I would say that, in general, what the industry is contending with right now is how do they get their 5G core ready.

We have customers that are driving towards a universal cloud type environment where they can have essentially the same technology sitting underneath IT and network workloads. There is a lot of the use cases that they need to support and this ties in to edge computing.

That’s all part of our horizontal strategy with OpenShift and its why IBM was so excited to partner with us and acquire us last year. It’s because you can take the workload and run it, whether that’s in a private data center or the telco [data center], or whether it’s in a customer’s private data center, or whether that is out on a near-edge or far-edge data center, or it’s on any public cloud.

Q: So on the one hand we have cloud infrastructure from Amazon and Microsoft. They are moving further away from those core data centers with their services and in some cases, there are some partnerships they are developing with telcos.

What’s the overlap with products like OpenStack and OpenShift, and how you see Red Hat supporting telco efforts to stay relevant in the cloud wars and offer edge, compute services, along with their network services?

CF: I’m sure you saw recently where Microsoft acquired Metaswitch. Things are changing in the space: you’ve got large cloud providers who are getting closer and closer to those VNF (Virtualized Network Function) workloads. This is where we can really help customers, whether it’s an AT&T or Verizon or any large telco. We’re convinced everything in the future will be multi-hybrid cloud. Not everything will be in a Microsoft cloud. Not everything will be in an Amazon cloud. Not everything will be in a private cloud.

If you think about things like smart manufacturing, smart farming, automotive, telemedicine, all the use cases that we know that are on the horizon, all of these are going to require a blend of environments. We want CIOs and CTOs to have an optionality of choice of where they run their workloads. And by us focusing on the workloads and making sure that we enable workload compatibility within OpenShift product across multiple cloud environments, we feel like that’s going to be a winning formula because there’s always been some form of lock-in for customers [historically speaking].

We want to provide a stable foundation for innovation to take place so that customers can continue to have a choice of where they want to run their workloads. We feel by playing, you know, Switzerland, if you will, allowing the customer to have more control over the control plane and where they choose to run their workloads, based on technical requirements – whether that be latency, availability, privacy, pricing and otherwise – by giving them that flexibility. They could determine which cloud environment will meet their needs and give them the most value.

Q:  There are many industrial production scenarios where there’s still a lot of proprietary equipment with proprietary protocols for communicating. Those systems have stayed the same way for a long time because, frankly, they work, and it is mission-critical. Bringing change into that kind of an environment is a very big challenge. How do address digital transformation in these cases?

CF: That change that people are facing when you get into manufacturing, whether it’s safety or data and analytics and that’s another thing with edge computing if you think about [the] massive amount of data on the manufacturing floor, you’ve got these instruments and sensors collecting all this data that needs to be processed. Obviously, you don’t want to take up this huge pile of data and drive it up to the cloud to do the processing. Because it’s a lot of data, you want to process it locally. It makes sense to have more edge computing take place there and there needs to be some synergy with the centralized data center, but how do you glean the insights from that and have this have the balance of local computing and data, combined with the overarching need that you need for your overall architecture?

This is something that we’re seeing customers push for and I think that the main thing as it relates to digital transformation and yes, I did mention [Covid-19] has accelerated, a lot of this because of the need to create more digital experiences, as opposed to physical experiences but it’s really about business agility.

Whereas before the whole lifecycle of deploying an application and getting the hardware and all of that was just, an unacceptable length of time that it would take, now getting an application to market with OpenShift can be done in a matter of minutes.

Being able to move more quickly is critical, especially post-digital transformation, where applications are a large part of how customers want to monetize [their business]. It’s not just the IT function now, digital transformation is actually affecting the top line too.

More about the subject of our interview

Prior to his role as an executive at Red Hat, Foster became the first blind graduate of the Harvard Business School leadership program and is a public speaker on the topics of resilience and leadership. Foster is the author of an upcoming book about his life experiences.

Article Topics

5G  |  AWS  |  edge computing  |  Google  |  IBM  |  IoT  |  Microsoft  |  NSP  |  public edge cloud  |  Red Hat  |  telco