If there’s one thing the digital age has shown us, it’s that business is moving faster. Digital providers are laser-focused and fully automated, and they’re deploying new apps and services as quickly as they can, sometimes in just minutes. No doubt, this has upped the ante for telecommunications companies. As such, downtime is more costly than ever, and reliability, availability and serviceability have to happen at cloud scale. At the recent Red Hat Summit 2016, held June 27-30 in San Francisco, I gave a presentation on what telco operators are looking for in cloud computing and how a Telco NFV Service Delivery Framework can meet their needs.
The networks of the future need to adapt: imagine if they were fully elastic and fully automated. Fortunately, with the recent efforts in software defined networking (SDN) and network functions virtualization (NFV), and the evolution of all network elements, completely programmable and elastic telco networks—from compute, to network, to storage, mobile video and even optical networks—are a reality.
In the next-generation, cloud-based telco networks, all services are mobile.
French multinational telecommunications company Orange is a great example of a telco that is moving away from its traditional infrastructure and evolving toward a service oriented architecture, a cloud and platform infrastructure and DevOps. The transformation includes project organization and new skills and tools. In his presentation right after mine, Cyrille Puget, PaaS product manager with Orange’s Innovation, Marketing and Technology Division, walked attendees through Orange’s Platform as a Service (PaaS) initiative. Puget outlined how Orange moved from a traditional infrastructure to virtualization and cloud that is mainly VMWare based but includes new cloud infrastructure with OpenStack.
Orange’s PaaS initiative was designed to demonstrate that PaaS can meet its needs, and was designed to prove that in six months, Puget said. Orange chose Red Hat OpenShift and deployed a beta platform. Puget shared the deployment timeline below with summit attendees.
Here are some examples of applications that Orange runs on OpenShift, as outlined by Puget:
- Web application used in Orange shops
- Rich Communication Framework: infrastructure and client framework to ease development of communication and real-time services
- Backend of TV remote control mobile application providing an enriched experience by delivering information of on-going programs
- IoT data aggregation services targeting end users and service providers
- Ticketing tool used by some operations team
- Many backends for mobile applications
Because service providers manage their business to EBITDA (Earnings Before Interest, Taxes, Depreciation and Amortization and an approximate measure of a company’s operating cash flow), they try to get more out of their infrastructures and do more with less. Managing costs is key, but so is agility. A Telco NFV Service Delivery Framework can help: it is multi-vendor; built on community collaboration; consists of open, interoperable application programming interfaces (APIs); and helps to automate everything.
Now let’s take a look at the key requirements that need to be balanced to enable service delivery at cloud scale in a cost-effective manner. It’s about more than just the Network Functions Virtualization Infrastructure (NFVI) platforms. Validation has to occur across more than a single instance, for both scale and geo-redundancy. There has to be a balance of reliability, availability and serviceability at cloud scale, that’s cost-effective. It’s about all those factors combined, with diversity and flexible redundancy for the physical/virtual data plane, so that the infrastructure is easy to operationalize and still delivers secure, high performance and personalized services.
Red Hat’s solution for open telecommunications is designed to address and align with telcos’ key priorities. Our portfolio of products and services can help telcos improve their ability to deliver on service level agreements (SLAs), improve their customers’ experience, automate, minimize outages and more—all at cost-effective scale.
I also wanted to share with you some of the insights from Dave Ward, CTO of engineering and chief architect at Cisco Systems, who spoke about virtualized service provider networking. His presentation outlined a network built on an IP fabric for NFV that creates a scalable foundation for virtualized services. Such a foundation sets the stage for a Services Orchestration Platform—a platform that incorporates the policy-driven orchestration capabilities of a complete services life-cycle manager for implementing service workflow topologies that span the network and data centers.
The business impact of services orchestration, for service providers, are many, Ward said in his presentation. He outlined the impact in the graphic below.
You can see both David’s and Cyrille’s presentations at the same Youtube link. And following our three presentations, Ward, Puget, myself and my colleague Pierre Olivier Mathis, Director of Telco Global Solutions at Red Hat, went onstage for a panel discussion. The conversation was great (you can hear the panel starting at 1:06:30). Each shared some of the progress their companies are making toward PaaS and the need for MANO.
As Mathis said, “One of the misnomers is that as we as we move to this transformation to the cloud, we have a tendency to think everything will be virtualized on VMs. But the reality is, in the telco world, you are seeing a lot of different types of workloads. You will have bare metals, you will have VMs, you will also have containers – and it poses a challenge, the complexity of not only looking at individual workloads but across workloads. I use this term polymorphic infrastructure, and that is the way we need to start looking at this. And having the ability to cut across all these different layers in a single pane of glass, that is really the direction we really need to start taking into consideration.”