A scalable and adaptive cloud-based message brokering service

Recently, message brokering services that are based on publish and subscribe paradigm are widely adopted in many areas such as remote facilities monitoring. Typically, they support a large scale distributed network of exchanging messages between services and clients. In this paper, we introduce a cloud-based message brokering service which allows better scalability and dynamic load balancing between brokers. Our system framework monitors each virtual machine (VM)'s resources and in order to meet the service level agreement (SLA), a VM is rented or recollected automatically by the system. The framework also provides the best available broker that a client can use for connections. Our experimental results show that our system framework can achieve up to 58.60% of scalability improvement at the maximum number of subscriber connections.