Rakuten Symphony’s cloud unit eyes up distributed network edge role

Robin, the power under the hood of Symphony's cloud platform, says public cloud partnerships on the way, as it lays out how telcos can move to a true cloud ops model.

Partha Seetala, President of Rakuten Symphony’s Unified Cloud Business Unit, has said that the company is soon to announce partnerships with public coud players that will enable telcos to manage network functions running as distributed cloud native workloads.

Speaking to TMN in a lengthy interview published in its TMN Quarterly magazine, the former Robin.io CEO Seetala described the company as a “distributed edge cloud platform with capabilities to do bare metal to service orchestration, on top of providing the rich cloud capabilities.”

He added that as well as Robin acting as an independent business unit, selling to its own customers and forming its own partnerships, part of his role within Rakuten Symphony, which bought Robin.io earlier this year, is “to lead what we call the unified cloud business unit within Rakuten.”

“Our goal is to become a very important and big revenue source for Symphony, selling this distributed edge cloud platform, as well as the automation platforms that are coming,” he added.

As telcos explore the distributed edge cloud, we have seen a variety of approaches, from working with public cloud providers and/or working with providers such as Red Hat and VMWare to provide the cloud host and management environment. So how would Robin.io fit into that landscape?

“Yeah, so that’s interesting, right? So you mentioned Red Hat and VMware, and I would say, you know, our products are fundamentally different. All of us are built on top of Kubernetes, as a base building block. Both Red Hat and VMware approach the problem from the perspective of taking Kubernetes and creating a commercial distribution out of it. And the ability to support strong storage-centric or network-centric workloads, is almost like an afterthought.

I’m under NDA, so I really cannot share the names, but pretty soon you’ll start to see some joint marketing activities

“We actually started from a very different perspective, we started from the get-go to bring storage and network intensive workloads to Kubernetes. And Kubernetes is the base building block for most of these edge workloads, right? As a result of that, from a product perspective we have a leaner, more performant and more agile stack.

“So in terms of capabilities, because we started from an architectural starting point of storage – and network-centric workloads, we have built a lot of capabilities around automating how a network function, for example, would look at the underlying infrastructure. So the network function does not really have to do any work and bind itself to the underlying hardware infrastructure, which is lacking in both these products.

“But having said that, you also mentioned hyperscalers. So we do work very closely with hyperscalers. I’m under NDA, so I really cannot share the names, but pretty soon you’ll start to see some joint marketing activities.

“For example, our storage stack. It’s the industry’s first and the most powerful application for storage – why is it important? The reason is because when you look at distributed edge, storage is a big challenge. Unlike a centralised data centre where you can
hire storage administrators for managing your storage, that’s not possible in the distributed edge because you have so many thousands of sites and multiple petabytes of capacity, but dispersed across so many different Edge locations.

“So you have to make that automated from the get-go, from discovering the underlying storage infrastructure, to making it programmably accessible to these edge applications, to ensuring that there are performance guarantees when multiple apps are sharing the same edge infrastructure. So all of that has to be automated, right?

“We are the first to actually approach this entire problem from an application centric view that actually appeals to hyperscalers.”

Seetala added, “If you think about the hyperscalers, not all of them have a rich portfolio of telco grade applications yet? Maybe they’ll get there at some point. But I think it’s very important in this world to be very open and support the entire ecosystem that comes with it.”

In the full interview, Seetala explores Robin.io’s role within Rakuten Symphony, how Robin enables network functions to operate on its cloud platform without developers needing to re-architect or cange their code, and how operators can gain the “belief” to move to cloud ops models, plus how Kubernetes’ “opinionated” stack will change the game.

He also explains how Rakuten Symphony will approach the market, operating its Symworld platform.

<<< To read the full interview with Seethala, as well as interviews with Vodafone, Netcracker, Itential and others, subscribe to receive TMN Quarterly’s latest issue on the Cloud Native telco. >>>