Multifaceted networks such as Telcos, ISPs, and advanced enterprise networks are rearchitecting their solutions with the advent of NFV, 5G networks, edge computing, and IoT devices. Each of these technologies brings a significant increase in the total number of connected devices, increased bandwidth available per device, and a significant increase in cloud service load.
The cloud-based network business case is about achieving an optimal level of automation and ability to manage the highly complex operational environment. Intent based orchestration, that automates the full lifecycle management of services and their components, without having to manually script or create complex workflows, is key to achieve these targets and run commercially viable NFV services.
StratOSS™ is a cloud native product, designed to deliver web-scale levels of operational automation for the cloud-based networking world – both to telco and enterprise service providers. A new OSS “in the control plane”, StratOSS™ automates the ongoing operational lifecycle of virtual network services from onboarding to ongoing maintenance.
Leveraging movements such as DevOps and tools such as behaviour driven testing from the enterprise IT community, combined with Accanto’s unique intent-based orchestration engine, we have proven that we can achieve 30 times less operational effort for ongoing maintenance of virtual network services.
StratOSS™ offers this brutal automation in three stages:
Intent Driven Automation
StratOSS™ provides intent based orchestration tools through its Intent Engine which is one of the core functions in the product. The main purpose of the Intent Engine is to auto-generate and execute the most efficient steps to get a service into its intended state without the need for the manual programming of scripts or workflows. In this sense it operates like a satellite navigation system. In a satnav, you enter the destination (intended state) and the car knows where you are (based on GPS), it knows your intended location, and then figures out the most efficient way to get there. You don’t need to program each step manually, you don’t need to program any left/right turns in the system, the system does that for you. And if there is a roadblock or accident, it re-routes you in real-time to get you to the intended location.
The Intent Engine works in the same way. All you need is to have the VNFs wrapped in a standardized lifecycle model, then to design the service so that you describe the relationships between the VNFs in the service, ie., (“VNF1 needs to be installed first, after that it needs to tell its IP address to VNF2, which can then be installed”). This then allows the Intent Engine to auto-generate the lifecycle for the service, and always keeps it in an optimal state when something happens.
The key enabling factor to make intent based operations work is to standardize the way individual VNFCs are managed. To facilitate this, StratOSS™ comes with a definition of a standardised lifecycle model specifying allowed lifecycle transactions and operational states that each managed software component must follow.
While all the VNFs follow the same lifecycle you only need to define the basic lifecycle transitions for each VNF, including Install, Start, Configure, Integrity, Stop, and Uninstall. SLM’s Intent Engine is automatically resolving the execution plans for advanced operations like heal, scale, or upgrade without any manual definition of workflows. Even more importantly SLM auto-resolves the execution plans for all service level operations based on the defined service model, current state of the service, and the desired target state.
During the operational lifecycle of services Stratoss monitors the health of the underlying virtual infrastructure, deployed VNFs, and the service itself, and if needed can automatically trigger the appropriate automated tasks like healing, scaling or diagnostics.
Deploying StratOSS™ and its embedded Intent Engine helps you to manage the complexity of NFV based service operations in multiple ways:
- Ease and speed up onboarding and testing of VNFs,
- Facilitate and automate service design and verification
- Automating of operations during service production lifecycle.
Behaviour Driven Testing
Before activating new services for end customers, everything needs to be working correctly. In a virtualized world, achieving optimal quality of service requires the continuous testing of the service itself and its underlying resources. Testing should be performed in all phases during the service lifecycle including: VNF onboarding, service design, and eventually service deployment. A high level of automation is required in order to reduce manual resource requirements.
StratOSS™ contains an integrated test framework module facilitating automated service testing. With StratOSS™ it’s possible to define applicable test scenarios within the VNF packages and service models. The StratOSS™ test framework utilises the defined scenarios and automates their execution as an integral part of each step during service lifecycle.
Every time a new VNF or a new version of an existing VNF in on-boarded to StratOSS™ it is automatically verified against the defined test scenarios to secure it is correctly working.
Similarly, testing and verification is an integral step in the service design process. Each time a new service design is made available it is tested to be structurally accurate and that it behaves as expected. Behavioural testing is performed in pre-production environment before accepting new service models to production. Performing behavioural tests in a non-production environment enables a set of test resources, like traffic generators and probes, to be used. The StratOSS™ test framework automatically triggers the defined scenarios including spinning up test environment, instantiating the service in test mode, running defined test sequences, and finally reporting the results.
When a service is accepted to production environment and instantiated to an end customer, its accessibility and quality should be verified before the final activation. With StratOSS™ this ready for service testing is triggered automatically as a step of service activation process. This ensures that the service instance is working with customer specific parameters and in customer-specified locations before sending the service live.
StratOSS’s™ integrated test framework and its automated behavioural testing capabilities offers:
- A better quality of service offered to end customers
- Less time and effort spent on testing
- Less time and effort spent on troubleshooting and fixing issues.
AI Powered Diagnostics
While services are live in production it is important to secure the quality of service offered to end customers.
For VNF based services, this requires continuous monitoring of service quality and health of the related resources. Also, when issues occur there needs to be a solution to resolve the issues in timely manner.
StratOSS™ provides diagnostics tools that you can use both proactively to monitor service behaviour and react to potential capacity or quality issues, and reactively to diagnose occurred incidents.
Proactive monitoring of services and resources should be done on many layers. This includes real-time status of VNF health, failures within underlying infrastructure, and service quality determined based on data collected from multitude of sources.
Different layers of monitoring and diagnostics require different methods of data collection, diagnosing root causes as well as to resolve the issues.
In terms of real time monitoring of VNFs and their operational health StratOSS™ provides mechanisms for direct closed loop monitoring and automated healing. VNFs are continuously monitored to understand their health and load. Based on the collected measures and policies defined in service models SLM’s intent engine resolves the required corrective actions and executes them. This can mean healing of VNFs or scaling them up or down to adjust deployed capacity to offered load.
Monitoring loops can also involve external systems to collect resource and service quality related data, and perform analysis to determine appropriate actions. This can mean for example fault management systems receiving alarms from data center infrastructure or assurance tools analysing reasons for service quality degradations. While VNF based services become more and more complex direct correlations between service quality and underlying resources becomes harder and harder to find. Artificial Intelligence based solutions will have an increasing role in understanding the dependencies and deciding on what is the best corrective action to take.
In cases where external assurance and AI platforms are used as a part of monitoring loops, StratOSS™ provides them with APIs to both run diagnostic scenarios on a service to find reasons for issues, and eventually to request corrective actions to be executed on the service.
StratOSS’s™ integrated diagnostics capabilities are also available for operational personnel to be run reactively as a part troubleshooting process.