Service providers demand to avoid unpredicted failures of allocating cloud resources to their services under limited resources of their private cloud, when they planned to scale their service capacity such as video streaming service and voice communication service against predicted congestion due to big events like concert and sport games.
OPNFV Promise project was initiated to achieve this requirement and is collaborating with Blazar project to address this issue by introducing resource reservation for future use. We tried to apply Blazar to our NFV platform to ensure on-demand service provisioning successfully.
In this session, we will explain how we solve the problem facing NFV by Blazar. It includes the concrete problem statements we hit in our NFV as well as the solution solved by Blazar. We also show a demo of planed resource scaling with Blazar in situations where cloud resource is limited and the demand of service changes with hours.
This session covers the issues we hit when we deployed VNFs on OpenStack cloud as well as the architecture of our NFV cloud. The attendances will learn following things:
- Concrete issues with VNF deployment on OpenStack Cloud
- The solution to solve the issues by OpenStack/OPNFV
- Demo of how the OPNFV framework works