With data and virtualization accelerating, it is highly likely that the average network operations team is currently neck deep in water. It is time to throw them a life preserver. Both the Operations and Engineering need a common network automation and management tool that can simplify service design and operations.
This management tool should complement virtual infrastructure resource managers (e.g., OSM, ONAP) and other systems, such as SDN controllers, IT DevOps tools (Ansible, Docker, Kubernetes) and legacy EMS/OSS, which will form part of a hybrid physical-virtualized environment. In a distributed software environment, the entire lifecycle of VNFs and services should, ideally, be automated.
The scope of this new tool should include:
- Service design (ongoing to resolve any recurring performance issues)
- VNF and service onboarding (simplified with standardized lifecycle models)
- VNF and service testing – integration with test tools for automated testing (lo-cally, and in a preproduction environment). Service bundles of multivendor VNFs should be tested for interoperability.
- Release management of VNF software packages – third-party VNFs should be “wrapped” in a standardized way so they can be managed via APIs, avoiding the need for custom integration and potential errors in production.
- Deployment – automated programming of all the steps required to take a service from concept to reality. VNFs and services must be continually created, config-ured, updated, scaled, healed and migrated with no human intervention.
- Automated VNF and service testing, monitoring and diagnostics – services are continuously monitored and, where possible, automated corrective action is tak-en, removing the need for manual resolution handling. If customer-service issues pass undetected by the monitoring or test systems, then the diagnostic tool should present coherent information to operations staff to reduce time to resolve and obviate the need to escalate issues to higher-tier support.
- Optimization of the production environment to maximize performance
- Automated healing, scaling, upgrading and migration of production VNFs and in-frastructure
By removing the need for manual monitoring and resolution handling, operations staff can focus on more value-added activities, such as predictive maintenance. Damaged service components that are causing faults can be replaced without human interaction, or sent back to engineering for modification. Greater interaction between engineering and operations should benefit both teams, leading to better overall service delivery across design, test and deployment. And by hiding the complexity of the underlying multi-vendor infrastructure and VNFs, residing in multiple data centers, service designers can be freed from defining complex workflows for each component, and operations teams can focus on managing the system on a more holistic basis.