February 16, 2017 | Written by: Steven Teitzel
Categorized: Network Agility
Share this post:
“To unify or not to unify OSS….. is there an option?”
Reinventing networks in a virtual age requires rethinking the operation support system (OSS). It requires a unification of OSS. The keystone driving this unification is the focus on the life of a network service or the service lifecycle and the management of it throughout its life or ‘service lifecycle management’ (SLM). What is service lifecycle management? And what is core to service lifecycle management that is enabling this unification of OSS?
Cloud for the IT world ushered in a whole new approach to application development. Similarly in networks, a new approach to a network service is possible in how it is conceived, developed, deployed and run throughout its life. Thinking through the lifecycle of a network service forces the breakdown of the prior silos of network development, deployment and operations and with it, causes the unifying of OSS. Service Lifecycle management unifies OSS — bringing with it the opportunity for operations optimization and speed of innovation.
Across standards specification and open source groups, different views of the life cycle have been surfaced. Some focus on the run-side of the service with deployment and scaling within the life or run-time of the virtual function or the virtual service. Another view brings in the design-side of the virtual service, starting with the on-boarding of the functions, development into the service, testing and deployment into operations that is then combined with the scaling scenarios. Another view starts with the inception of the production of the function to include the supply chain side of the function through the design and run scenarios along with the customer consumption. Each view is valid and in aggregate fit together to form the parts of the view needed for service lifecycle management. We find a key focal point of all perspectives; in each of these a focus in the run-side of the lifecycle deployment is the scaling and orchestration of the virtual service. The scaling and orchestration is based upon not only what is going on within the virtual elements but the overall hybrid service (physical and virtual) to the consumer, an automation of the management and continuous monitoring to enable this.
There is a critical need within SLM to have an ability for continuous orchestration of the service to handle multiple scaling scenarios that occur during the life of the virtual function and virtual network service. The focus of enabling this continuous lifecycle orchestration from the point of inception to retirement in a ‘DevOps’ approach and all the iterations along the way brings together OSS and also makes this a key linkage point within SLM. This continuous orchestration needs to be closed loop, making sure the actions required to be taken are executed and the results of achieving performance needed are received. This needs to occur not only within the virtual elements of the function or the functions chained as a virtual network service but also based upon the service performance end-to-end, provided by both the physical and virtual pieces, the expected performance of that service, the consumer of that service and perhaps (as needed) the content of that service.
The power of cloud technology enables this closed loop, continuous orchestration. This orchestration also requires the capabilities of cognitive analytics of the service based upon policies defined at the definition of the service to scale at the speed of the cloud. As virtual network functions move from appliances being ported to a virtual environment to cloud native or broken down into micro-services, the ability of managing the performance of these micro-services together increases the need for this process to be highly scalable. The analytics of the performance need to be enabled with foresight – not just insight – to proactively scale. Though there are many with intelligence to manually decipher the timing needed to scale proactively, it is with the capabilities of cognitive that we can augment this intelligence to be continuously watchful for the patterns of performance to scale. The power of cloud and cognitive come together then to enable the unification of OSS needed for cloud-based networking with closed loop continuous orchestration.
Service lifecycle management enables the unification of OSS or uOSS. And it is the power of cloud innovation and cognitive that enable SLM. IBM has been actively involved in the work to define service lifecycle management, engaging with TM Forum and PoCs through catalyst projects for Recover First, Resolve Next: Closed Loop Control for Managing Hybrid Networks and A Well Enabled VNF Package. If you would like to see more, come see us in our private lab environment at Mobile World Congress and discuss how we can help you enable closed loop continuous orchestration to drive service lifecycle management to enable unifying OSS.