June 23, 2020

Ludovic Rubin

Director of Product Marketing

5 Reasons Why You Should Include Disaggregated Router in Your Next Network RFP

For many Communications Service Providers (CSP), it’s already 2021. With network traffic stabilizing at a 20-30% increase over pre-pandemic days and worldwide traffic patterns shifting from enterprise networks to broadband, CSPs are starting to relook at their network’s scale and cost structure.

5 Reasons Why You Should Include Disaggregated Router in Your Next Network RFP

The good news is that throughout this period and the challenges imposed by the Coronavirus, CSPs continued to successfully provide services, upgrading their network infrastructure for what would have been planned for the rest of 2020 in a matter of weeks. The not-so-good news is that to support the traffic growth and the new opportunities with webscalers at the Telco Edge, CSPs need to address long-term network issues, such as:

  • The network architecture is operationally complex and under-utilized,
  • Lack of scaling flexibility
  • Declining profitability due to legacy models and vendor lock

Now is the time for CSPs to rethink their network architecture and align it with modern hyperscale clouds. Bank of America issued a primer paper encouraging CSPs to look at white box-based routers and new disaggregated network models.  These open software-based networking solutions running over white boxes offer substantial flexibility and efficiency over proprietary, hardware-centric incumbent solutions. Just like hyperscale cloud infrastructures, these networks are more flexible, operationally efficient and cost-effective than legacy networks. Disaggregated networks are already gaining momentum in 5G network architectures. Several Tier-1 operators have adopted them for their core and aggregation. If you plan to issue an IP network RFI or RFP, make sure to include the newer disaggregated router vendors, preferably DriveNets, which the Bank of America research notes as leading in the space. Here is why:

Reason 1: It scales more efficiently. Ask the hyperscalers

Using server virtualization, low-cost white boxes, cloud-native software on top and a simpler scale-out architecture (CLOS), hyperscalers solved compute and storage scaling in the cloud. Similarly, disaggregated networks are based on open standardized design, cloud-native networking software (using containers and microservices) and standard networking white boxes from a choice of ODMs. DriveNets Network Cloud is implemented based on the Distributed Disaggregated Chassis (DDC) design submitted to the Open Compute Project (OCP) by AT&T. It can support a single white box-based router of 4 Tbps and scale up to 768 Tbps in a single router entity based on a cluster of white boxes, all with only two white box building blocks, covering access, edge and core routing in the same solution.

Reason 2: It is open and avoids vendor lock and price lock

With cloud-native software, standard networking white boxes and flexible operational and commercial models, network disaggregation change network economics, allowing it to scale capacity and services much faster. By enabling Service Providers to choose white box vendors independently of the networking software and growing the network by adding more white boxes over time, disaggregated network models break vendor lock and, more importantly, port pricing lock. Openness means open architectures with well-defined interfaces between the different components and open designs through standards – like the OCP DDC model for SP routing, the Open RAN alliance for the 5G radio network, and open source software. But the ultimate transformation that disaggregation brings is an open ecosystem of high-skilled players, each of them focusing on their own domain of expertise. This fosters competition, innovation at each component level and overall, choice between the different vendors as well as a fresh approach from new entrants, like DriveNets. 5-Reasons-A-Disaggregated-Network-Vendor-Should-Be-in-Your-next-IP-network-Infrastructure-RFI-RFP

Reason 3: Simpler operational model, lower OpEx

Disaggregated network solutions are much easier to operate and manage. This turns the different embedded hardware-centric functions of a traditional router into separate manageable software-driven planes: the data plane (packet forwarding), the control plane (router intelligence) and the management plane. With software-driven modular architecture and cloud-native technologies, the hardware is composed of standard white boxes or x86 servers. In DriveNets’ case, we only need two types of white boxes to build any size router for the core, edge or access. While the most scalable solutions use a distributed architecture and hundreds of white boxes, they are still managed like one router. Software is the same regardless of the hardware type on which it runs – single box or multiple boxes – simplifying deployment. And, it is automated. To ensure that a distributed disaggregated chassis acts as a single-vendor integrated chassis, it automates the provisioning and management of the router as an integrated solution, with zero-touch provisioning and life cycle management for the hardware and software. This ensures fewer manual errors, faster time-to-revenue and higher visibility.

Reason 4: It’s telco-grade but more flexible than traditional chassis

DriveNets Network Cloud offers a complete router solution. Just like an incumbent router vendor, DriveNets delivers the full testing, certification, operations and support of the solution while working with the ecosystem of vendors (ODMs, silicon vendors, optics). This happens behind the scenes – aligning roadmaps, bug fixes, lab and pre-production testing. The software vendor ensures the telco-grade performance of the entire solution. Like traditional router deployments, a Value-Added Reseller (VAR) plays a part in the ecosystem, usually involved in the deployment of the full system, making sure that it is deployed and maintained according to customer requirements.

Reason 5: 50% lower TCO

Disaggregated network solutions are built to deliver the lowest TCO, when compared to traditional router vendors and their high margins. The white box hardware is a simpler form factor, identical across network layers (access, edge, core, peering), sold on a cost+ model, while the software license model enables the decoupling of the network growth and costs. The virtualization of the hardware – the use of any service on any port, enables a much higher overall network utilization (more than double) and therefore better value out of the network’s physical resources. On the operational side, network simplification and automation reduce network engineering, software upgrade and service assurance expenses. According to ACG Research, DriveNets Network Cloud offers a 50% reduction over legacy solutions.

5 year TCO comparison

It’s time to move to disaggregated networks. Other Tier-1 operators have.

Several network operators globally have tried to develop their own disaggregated network solutions – developing or buying software that runs over standard white boxes. Many of them realized the benefits of it, but also the advantages of using commercially available software, tried and tested by other high-scale operators. DriveNets is now in many tier-1 operators globally at various deployment and testing phases. It’s worth considering it in your next RFI/RFP.

AvidThink White Paper

Network, Meet Cloud - Scaling Disaggregated Network Infrastructure using Cloud Principles

Read more