Thanks to all who joined us for the first of The Modern Telco is Open Series, The Open Telco Framework. During the webinar, we saw how Red Hat’s open 5G architecture foundation helps modern telcos deploy new services (MEC apps, vRAN, IoT) – faster, economically and at massive scale. After the webinar, we had a live Q&A with the audience. Unfortunately, we ran out of time before we could get to all the questions. See the full Open Telco Framework Q&A below.
Part 2 of the Modern Telco is Open Series: Virtual Central Office (vCO) in an Open Telco World is now available!
What are you seeing around the world in terms of this evolution to an Open Telco?
Telcos are already in production with their common virtualization platforms and offering business and mobile services now – some are starting to transform their consumer
services and we are starting to see media services up next. However, while the common infrastructure platforms are available now, we are still working on distributing that architecture so we can place any function anywhere and also working on operational aspects — telemetry , service assurance, monitoring and thus much of the activity at OPNFV, ONAP, etc is currently focused on these in parallels with the 3GPP/5G/MEC efforts to standardize the architectures.
Is the journey moving as fast as it needs to? What is holding it back?
A couple of things have been happening along the way…
Firstly, the common hardware platforms are not delivering the performance to match the proprietary appliances / network processors — thus we have to be economically pragmatic about the functions virtualize as well as where we can place them — thus the architecture will be a hybrid physical virtual distributed one that leverages the right technologies to optimize the creation of business value for the telcos.
Secondly, the network applications from many vendors — remember , telcos don’t develop these , they consume them from multiple vendors — these have not really been refactored into small micro services pieces leveraging containers that can be scaled independently with a standardized software mechanism to communicate between them. API managers can help with that communications, but we need an open interoperable and scalable approach, and we have to decide which apps stay on bare metal (not refactored) which ones in VMs — those that are data intense, and those ones that are containerized micro-services that we can place anywhere at different service scale points.
Which open source projects were picked up by Red Hat?
First off, all of Red Hat products are open source today, based on multiple open source projects that are integrated, secured, and hardened for production deployment — Openstack, Kubernetes, Origin, Ansible, Drools etc… just to name a few. Red Hat also leads and influences many other open source projects and communities to accelerate innovation across the Telco ecosystem – eg ONAP, ODL, FD.io, MEC, CNCF, Open Air Interface, and many more.
What features distinguish your platform from other open-source platform, e.g. ONAP? Any unique features only Red Hat can offer?
ONAP is an end-to-end orchestration and automation platform and framework, and the majority of its elements have been open sourced. Red Hat supports the ONAP community and many Red Hat elements are already included in various parts of the framework – Ansible, Drools, etc. Key features that distinguish Red Hat software platforms – Openstack, Openshift, JBOSS, AMQ-P etc is that they are secure, hardened, completely open and interoperable and they are never customized for a single customer requirement. In our “upstream first” model, all features and updates made to the code is delivered first to the community. When the community innovates, it adds to the code that is already available, and does not need to be merged, nor any need for “telco version” to be integrated, which accelerates innovation and differentiation. In essence, Red Hat and the open source community enables key capabilities for our customers: choice of suppliers, innovation at the pace of the industry, more control over their destiny, and accelerates their ability to deliver their services.
Are network platforms build to deliver data in real-time?
For quite some time, network platforms and appliances were built with specific hardware to support delivery of voice/data/video in real-time. In the virtualized environment, our linux operating system (RHEL) and the hypervisor (RT-KVM) software integrated into our Openstack platform (RHOSP) support real-time capabilities that enable consistent low latency performance characteristics required for many of the emerging 5G/MEC applications.
How do service providers view integration, since that is a big problem with common infrastructure?
SPs have always dealt with a need to integrate technologies into their environment as they consume them from multiple vendors and they have chose traditional systems integrators or large equipment suppliers that they can negotiate guaranteed service level agreements with them. In essence, SPs view it as a necessity that has grown even more important as open source has enabled them to select from a broader set of vendors/partners to create business value from their service offerings.
Along those lines, does this view change with the Tier 2/ Tier 3 operators?
All SPs are being challenged in the marketplace to deliver services while keeping up with growing customer demands and optimizing both CAPEX and OPEX spend curves. With Tier 2/Tier 3 operators, their reliance on trusted partners and systems integrators is even higher as they have less internal resources they can apply to integrate technologies into their environments.
How are vendors/operators monetizing solutions in this Open Telco framework?
The key to the Open Telco framework is a set of three agile open platforms – network infrastructure, data management, and application development from which the operators can rapidly develop and deploy innovative services for their end customers. The first benefit from this framework is the ability to lead the market with new services that generate revenue faster than their existing services. The operators can also leverage the large amounts of data generated by their infrastructure to optimize and operationalize their environment and may also create services for their end customers from analytics and machine-learning (AI) technologies.
What is the relevance of new skills sets that telcos would have to structure to face the challenges of the new architectures?
Increased use of open source and cloud-native technologies drives the need to adapt or infuse additional personnel with many new skill sets, as well as a open and collaborative mindset that can help them face the challenges of our digital economy. Telcos need to empower small teams with executive support to successfully leverage these skill sets in small projects that can be expend over time across their entire organization.
With these open platforms, what are the advantages to deploy all common infrastructure than using branded or other infrastructure? And where is the position of Red Hat on those infrastructures?
The key advantages are vendor choice combined with agility of service delivery and a faster pace of innovation than with proprietary branded solutions. Agility is the only sustainable advantage today. Using common platforms across the entire life-cylce of the telco business, not just for infrastructure ,but for data management and development enables the SPs to leverage common operational tooling, software and data assets to create business value across a broader set of service offerings.
How do you see the “edge cloud footprint” (how small shall it be?)
The edge cloud “footprint” size will vary depending on the types of services being deployed and how many subscribers or customers are being service from a location. This is no different from how one looks at networking today with existing technologies, the size of the edge nodes is based on throughput (data / sessions) required to support a certain number of customers. If we look at the mobile vRAN use case, that edge node will be using compute technology right at the foot of the radio towers – so very small — call it a micro-node / micro CPE. If we look at the Enterprise edge business services or the IOT gateway use cases, these medium sized nodes will serve one/more enterprise locations and thus mileage will vary upwards in numbers of compute nodes and storage. As we move closer in to the operator POPs, the edge cloud “footprint” would continue to grow – depending on the volume of services / subscribers aggregated at these virtual central offices (VCO).
Who are the early adopters of the Open Telco at the moment?
There are many early adopters of the Open Telco approach in every geography ( Telco/Mobile/Media delivery providers) today as they started their virtualization journeys quite some time back and already have mobile, business, content, and IOT services deployed today using vertical solution approaches. Many of these operators are contributing to open source projects from ONAP, OPNFV, Openstack, CNCF and are adopting cloud native infrastructure, common operational tooling, billing, analytics, and agile development platforms to converge their services and application delivery across their entire ecosystem.
Where are we seeing success and why?
We are seeing success in different regions of the world – Asia Pacific, Europe , as well as in Canada and USA. Successes have already come from a few areas:
- focus on specific services or problems
- participation and contribution to open communities to accelerate innovation
- building strategic business partnerships with their vendors, integrators and each other
- executive support to disrupt their businesses
And as said by Mario Andretti — if you are not feeling discomfort you are not going fast enough.