Red Hat's Jordan-Smith: Digital Transformation Happens at the Mobile Edge

When it comes to digital transformation, Red Hat plays something of a matchmaker -- to hear the way the open source company's vice president of global information and communications technology, Darrell Jordan-Smith, tells it.

"For an example, we might bring a bank to an AT&T, in order for AT&T to offer them a software-defined WAN solution," Jordan-Smith told Telco Transformation, describing how Red Hat Inc. (NYSE: RHT) "helps 'telco X' reach out to specific vertical market elements or business customers."

As any matchmaker knows, however, there is more to a relationship than the introduction. Red Hat plays coach and facilitator as well -- and Jordan-Smith seemed happy enough to tell Telco Transformation all about it in the first part of this Q&A series (lightly edited for length and clarity.) Jordan-Smith talked about digital transformation by proxy by describing some of the strategies, technical practicalities, and business-case considerations that go into enabling those transformations for telcos' enterprise customers as well as other network operators.

In part two, Jordan-Smith delves deeper into the subject of digital transformation by proxy as he discusses Red Hat's close partnership with Orange (NYSE: FTE).

Telco Transformation: How is Red Hat helping telcos' with their own digital transformation and, in turn, digitization by service provider customers?

Darrell Jordan-Smith: A lot of the work we've been doing at the moment is really helping the telecommunications companies, for want of a better word, digitize their own network infrastructure -- such as building a lot of software definition around it to switch on or switch off services in minutes versus months. This enables them in turn to offer differentiated business services to their business customers as well as services to their customers. We're beginning the journey right now. For example, in the Internet of Things space, enabling a lot of the users of software technology and open source in general to do some intelligent things at the edge of the network.

You see a lot of discussion at the moment around mobile edge, and what that actually means, and where the telcos are looking to differentiate at the edge. Their strategy typically is that they want to focus on providing the best possible network infrastructure to enable things at the edge -- whether they own the edge property themselves in terms of network infrastructure, or they partner with, say, a Microsoft or a Google or an Amazon to provide mobile-edge type services and manage the efficient backhaul and networking elements needed to facilitate some of those applications that are running at the edge of the network. The reason why a lot of companies are looking at edge-based mobile or compute-based services is because there are certain applications that require low latency and/or are more cost-effective running at the edge of the network versus having to transport everything back through the core of the network out to a data center and then back again.

TT: Do you have any examples or particular use cases you can discuss?

DJS: Examples of that are difficult because there are many of them, but we're doing some interesting things with some service providers in healthcare, such as Telus, in Canada.

In healthcare, we're doing remote monitoring. So instead of having someone in the hospital, we can have them at home because typically people repair and/or recover better in their home environment than they do in a hospital-based environment. Plus there's saved money for the insurance and all the hospitals that are involved in that … And we can provide all the real-time low-latency monitoring information back to the hospital. So that's one application that we've been actively deploying at the edge of the network. We have also been enabling mobile workforces, as well as getting involved in autonomous-car trials in different cities -- those types of networking-based applications that need to be dynamic and agile by definition -- with the telcos, using open source software to enable them. A lot of that is sort of early days and in new areas. We're seeing lots of edge-based transmission technologies coming along soon, but at the moment a lot of the trial-based services are happening over existing infrastructures. So it's a particularly interesting time looking at that.

A lot of the telcos also see companies like Red Hat having a big presence in traditional corporate America, in banking and other areas, and there are vertical opportunities in terms of providing software-defined wide-area networks (SD-WAN), and providing mobile VPN services as a service -- all that with software. So telcos, in terms of a route to market, can go offer those types of new services to those enterprise-based customers. We're doing a lot of things in that space as these telcos become agile in the way that they deliver network-based services.

TT: To speak of the boom in demand for mobile-edge services, we're starting to see a lot more data and reports about how over the next few years more and more TV and video content will be viewed on mobile devices. How are you seeing the increasing mobile video demand of both consumers and enterprises impacting what your own network-operator customers want?

DJS: All of that is driving a significant chunk of data. I think you can read almost any statistic; anybody will tell you it will exponentially grow. The cost of storing that data is quite expensive from a technological perspective. So open source technologies such as Ceph and Gluster, which are open software storage technologies in a hyperconverged environment -- you can scale CPU storage and networking independent of each other at the edge of the network -- are what we're seeing being built and deployed at the edge. Now what we're doing with the software with some of the data that we're storing at the edge is significantly lowering cost. It becomes more affordable to store the data at the edge, cache the data at the edge and also collect the telemetry at the edge.

A lot of the work that we are doing is enabling that by lowering the cost of being able to deliver those types of services at the edge for the service providers. That then has interesting applications that developers can write, both for consumers and businesses, at edge of the network. That's just one aspect, on the storage side.

TT: If it's going be a lot more affordable and more accessible to store data at the edge instead of in core data centers, and do other typical core data center activities at the edge, will the core data center as we know it become obsolete? How do you see these factors impacting cloud and data center demand?

DJS: The core data centers are still going to exist because it doesn't make sense to operate the core applications -- such as billing and operation support and all the elements that you need to deliver a quality of service across a network infrastructure -- at the edge. Those are big applications servicing many millions of customers. So we see those being more efficiently deployed in a cloud-based solution and a cloud-based environment -- normally virtualized. Certain network functions are better operated in the data center at the core of the network than they would be at the edge. So we think that the data center in the core of the network will still be there. It will be "cloudified" -- for want of a better word, and become much larger.

But a lot of the intelligence doesn't need to always come back to the core data center; a lot of that can be kept at the edge of the network. Actually, it's more cost efficient to do that at the edge of the network because you don't have to transport all of those bits back into the core and then back out again. And it frees up networking resources. Therefore, the cost per bit is dramatically reduced.

— Joe Stanganelli, Contributing Writer, Telco Transformation