Architecting Scalable Multi-Cloud Environments for Telecom OSS with Advanced Infrastructure Automation Tools

Telecom companies need to rapidly deploy new digital services. The evolving multi-cloud landscape containing public and private clouds and edge computing resources can assist telecom companies in scaling up IT operations. However, public clouds have limited service assurance. To make the best use of the public cloud while ensuring service quality, telecommunication companies need to build their own multi-cloud environments, using multiple public clouds together with edge and private clouds. It is not easy to build a multi-cloud environment as a good mix of several clouds is quite complex but those environments are essential. This approach comprises four components. The first component does the service modeling and design to create a service model containing cloud resource requirements, deployment and scaling policies. The second component finds suitable clouds for the service model. The third one selects the cloud instances of the clouds, and all deployment and scaling policies, making sure that constraints against service execution are not violated. The last component periodically monitors the execution status of the service and triggers the deployment and scaling policies when needed. While the first three components have been studied and implemented in earlier works, this last step was not addressed before. Large-scale experiments on the footprint of a telecom OSS workload at publicly available clouds demonstrate that the solution operates quickly and correctly, and the assurance of the telecom OSS service on public clouds is well satisfied.