OSS as a service, cloud-native—what is next?

OSS as a Service is continuing its evolution, as the cloud-native approach takes an even firmer hold across all Operations Support Systems (OSS) layers. What role do microservices and containers play?

Sr.Director Product Management Ericsson Inc.

Sr.Director Product Management Ericsson Inc.

OSS - Fundamental to SDN & NFV transformation

Before I get in to microservices and containers, let me explain what generally constitutes an OSS system. OSS in general includes management functions such as service fulfillment including network inventory, provisioning and activation, service assurance, network management, fault management, monitoring and customer care. According to Telecommunications Management Network (TMN), there are four layers that constitute OSS, namely Business Management Level (BML), Service Management Level (SML), Network Management Level (NML), and Element Management Level (EML).

OSS systems are some of the most complex and expensive systems for operators to manage. Even though OSS systems are managing the network, they are usually managed by an IT team within the operator’s network. Many operators realize that OSS evolution is fundamental to their SDN and NFV transformation because without nimble and modern OSS systems, operators will not be able to achieve the desired automation, time to market, innovation, and scalability.

Orchestration, automation, and analytics will play a major role within OSS evolution. Any changes to be done in OSS systems are complex, take time, and are also expensive. When managed services came along, several vendors were providing OSS systems as part of their managed services offering, and this really helped operators to focus on their actual business and leave the management of OSS systems and the network to the managed services vendors. The cost moved from capex to opex. Managed services vendors took care of software upgrades, hardware related upgrades and maintenance of OSS systems.


The evolution of OSS systems

In my opinion, from a telecom network management perspective, managed services laid the foundation for "as a service" offerings. Then came the cloud. Virtualization provided several key capabilities such as agility, automation, scalability, and faster time to market, which were very difficult to achieve in legacy OSS systems. In addition to virtualization, here comes microservices and containers.

Microservices and containers will play role in the evolution of OSS systems, as in other areas of the telecom infrastructure. Like any other applications, microservices and containers will provide scalability, agility, and acceleration to OSS systems. In addition to OSS systems themselves moving to cloud-native, networks and applications will start to become cloud native and this poses new operational requirements to current legacy OSS systems. In their current form, OSS systems are complicated, but the advent of microservices-based networks and applications will put more demands on OSS systems. Certainly there were number of moving parts in the legacy network such as mobile subscribers, services, VPN, and so on, but nothing like what is going to happen in microservices-based networks. The management complexity that will be created by moving applications to micro services and containers will be profound compared to simple virtualization of network functions on a VM.


The microservices and container era

In the microservices and container era, there will be more automation, more dynamic workloads that are moving around, and containers that are constantly instantiated and removed in different parts of the clusters. This requires highly sophisticated monitoring, correlation, and management capabilities. Not only applications and infrastructure need to be monitored and managed, but also intercommunication among microservices/containers, their status, and parts of applications running inside the containers need to be monitored and correlated. Machine learning will play a key role in the management of microservices-based networks. Virtualization, microservices, and containers create new demands for securing the network, and security is even more critical to OSS systems.

Telecom and enterprise networks will continue to evolve and new technologies such as Function as a Service (FaaS), and serverless computing will provide more capabilities. More features and capabilities also bring complexity to manage various hybrid networks and legacy systems. Not only will OSS systems need to manage legacy systems, VMs, and container-based networks, but they also will be moving to VMs and containers. Depending on their use cases, service providers might evolve their networks and management systems in a step-wise manner.

Some of the management functions within the larger OSS systems may take a lead in the cloud-native journey. For example, it could be network management, cloud management, or an orchestrator. In the same manner, some of the OSS systems, for example, an NMS or an inventory system, might take a lead in the SaaS delivery model. From the user perspective, users are provided with an OSS portal; meanwhile, the OSS systems are hosted and deployed in many ways such as over bare metal, private cloud, or public cloud.


New capabilities in OSS systems

To manage the microservices and containers, new capabilities are needed in OSS systems. One possibility is to leverage the new tools that are coming out of the container world. Cloud Native Computing Foundation (CNCF) projects such as Prometheus, OpenTracing, Jaeger, and Fluentd provide capabilities around tracing, monitoring, and observability. Security is key in the microservices area, and it is highly relevant for microservices-based OSS systems as well. New projects in CNCF such as notary and TUF will play a role in the area of security. As container technologies evolve, OSS systems will need to evolve further by absorbing new technologies in different layers of OSS.

Below, I have provided two links. It is worth reading them to learn about the Prometheus project from CNCF (see below link and diagram) and also about a network manager that is offered as a SaaS model.

Read more about Ericsson Network Manager 'software as a service' launch:

Read more about CNFC's Prometheus project:

The Ericsson Blog

Like what you’re reading? Please sign up for email updates on your favorite topics.

Subscribe now

At the Ericsson Blog, we provide insight to make complex ideas on technology, innovation and business simple.