Dynamic Policy Management System for High Availability in a Multi-site Cloud

Edge computing promises to improve the network operational cost and improve the quality of service by pushing computation resources to the network edge. Each edge cloud consists of heterogeneous servers. The Service Function Chain deployed in Edge sites needs to be flexible to meet the QoS requirement. The service function's location and policy can vary based on latency, so stringent latency constraints are applied to a service. Although there are solutions to monitor systems flexibly, the policy for service function when migrated to a new edge site is still a challenge. In particular, Network operators may need to quickly change the network service's policy to adapt to the new underlying infrastructure. In this paper, we propose a Dynamic Policy Management, which monitors the whole infrastructure and network virtualization environments to detect the anomaly and dynamically update maintenance action to guarantee the high availability of network services, such as auto-scaling and migration.