A self-adaptive mechanism for resource monitoring in cloud computing

Cloud computing is a distributed system with thousands of servers sharing computing power, network, storage and other resources through virtualization technologies and resource management technologies. Resource monitoring is one of the key modules of Clouds, which provides the resources information used by job scheduling, load balance, billing system and other modules. In this paper, we proposed a self-adaptive push model (SAPM), which is based on the popular monitoring methods in Grid computing. This model uses a transportation window to store the collected metrics before being delivered to the monitoring servers. And we design the WAIMD algorithm to control the metrics´ push. The experimental result shows that the SAPM model decreases the load on a network, and achieves a better performance in keeping data coherency between hosts and monitoring servers.