ETSI OSM Comes Close to Commercial Reality, Bringing the Benefits of Open Source in the Complex MANO Area

ETSI OSM recently announced Release Four of Open Source Management and Orchestration (OSM MANO), an architectural advancement that constitutes progress in terms of functionality, user experience, and overall maturity of the platform. The first notable enhancement in Release Four is the OSM’s increased independence from the Operational Support System (OSS) layer, building on its innate modular and model-driven architecture. The new release also offers a cloud-native model, a setup that draws on cloud capabilities for resource management and deployment. Equally important in the new release is the discussion around a new Virtual Infrastructure Manager (VIM) plugin model, which mitigates the existence of VIM incompatibilities due to a lack of standard interfaces. This has been a significant and fundamental pain point in early Network Function Virtualization (NFV) deployments, where vendor cloud hardware has been incompatible with third-party VIMs, thus causing compatibility issues in the hardware layer. This has kept early adopters from advancing to higher layers (e.g., Virtual Network Functions [VNF] and NFV Orchestration [NFVO]), and is a reason for the delay in NFV advancement.

Building on Telefónica’s OpenMANO initiative, OSM enjoys the support of a number of notable telcos and vendors, namely BT, Verizon, Red Hat, and Intel among many others. The OSM community is keen to remind the industry that it is making inroads toward an open telco-ready, production-quality, VIM-independent MANO stack. Release Four promotes functional modularity and independent interfaces between the OSM stack and the rest of the OSS modules. This approach points to a clean architecture free of unpredictable interdependencies across components of the value chain, but it raises a support and development control debate that ought to be considered by telcos and vendors alike. Furthermore, OSM adoption on a wide scale presents vendors with different market dynamics; therefore, they should fully explore its strengths and opportunities but remain cautious of the challenges that lie ahead.

An open source MANO stack provides telcos with the agility and innovations associated with the Internet world. It also propels the collective industry vision toward new business models and new revenue streams. However, an aspect that warrants a discussion is the issue of software support. Telcos will require a highly reliable MANO platform to deliver their services, so the cost of testing, supporting, and updating the underlying code is significant. It is therefore imperative that telcos consider how an open source MANO platform will be supported before deploying it. This is particularly important because a specialized and tightly controlled telco infrastructure will need to be supported over a longer period of time—potentially several decades—when contrasted with fluid Internet technologies. These architectural differences carry implications in terms of the ability to support an open source MANO stack, and in the control over its development. Telcos and open source communities should, therefore, seek a sustainable means of software support over these timescales, regardless of who is going to provide the support and who controls its development.

At the same time, vendors are debating the merits and drawbacks of open source business models, and are finding that copying what works in the Web domain may not necessarily succeed in the telco domain due to carrier-grade reliability requirements.

There is also the issue of compatibility and OSM’s adherence to mainstream standards, in this case ETSI OSM’s NFV framework. The prospect of OSM, or other open source MANOs such as ONAP, becoming mainstream can potentially change vendors’ basis of competition, pointing toward a gradual evolution in product architecture away from proprietary designs toward modular architecture. This enhances flexibility but, due to adherence to a tight standard specification, it can potentially stifle innovation and hinder technological advancements from an engineering perspective. Inevitably, the openness of the OSM platform will mean that vendors compete “in the market” rather than “for the market.” Vendors should be ready to face a scenario where competition between multiple proprietary platforms is supplanted by competition on an open platform. They should match their product architecture to their competitive circumstances to capture growth in different layers of added value to telcos.

OSM, just like ONAP, includes a NFV MANO stack, but the latter covers a broader scope that extends to BSS, OSS, and inventory management. Therefore, though different in size, there are complementary capabilities that telcos can compatibly implement in their networks. The momentum behind open source MANOs, however, raises a discussion that centers on the role that telcos and vendors will play in driving this initiative forward. ABI Research believes that telcos will be the main driving force until the initiative reaches a tipping point. When that materializes, vendors need to think of a robust and flexible model that builds on open source MANOs, and encourage innovation and continued investment in software development. Adding an element of their own that is both substantial and differentiated is going to be crucial for vendors to offer products and services that deliver added value to telcos.

The modular nature of OSM alters the industry structure because it opens new opportunities for independent, nonintegrated vendors to sell, buy, and assemble solutions. Ultimately, ETSI OSM’s OSM framework may coalesce as an industry standard. When that happens, telcos can mix and match components from best-of-breed vendors to respond conveniently to diverse use cases. Vendor competition will revolve around two themes: price and product features. Vendors need to think strategically and focus on a subset of (compatible) components aimed at how to best build a rich digital ecosystem using open source and independent interfaces, and cloud-native technologies.

The undertaking to run and upgrade today’s telco infrastructure is not a trivial matter; therefore, serviceability and operability is a key facet of OSM’s commercial prospects. Historically, there tends to be a huge lag between the time an open source community announces a new release and the point at which when the release is ready for the operational network. Telcos and vendors should work together to optimize the full value chain with an eye to switch from a sequential to a fully parallel release model. For example, vendors can work with the open source community in a synchronous fashion to test pertinent code. An arrangement whereby vendors develop code in parallel with the upstream community, and release their code upstream at the same pace, should aid the maintenance part.

At the end of the day, the telco industry is still undecided regarding which deployment model to follow: open source coupled with professional services (e.g., Amdocs and ONAP), or strong product-driven deployment (e.g., Netcracker). OSM Release FOUR and ONAP Beijing will provide further operational insight into real-life deployments, but it may be a long time before open source becomes the established deployment standard for MANO in telco networks, especially since there is competition between OSM and ONAP.