MCloud: Efficient Monitoring Framework for Cloud Computing Platforms

Abstract

Cloud computing platforms have the characteristics of large scale, complex interaction and dynamic environment, which cause it more and more serious to software and system security, and the monitoring technology is one of the key technologies to ensure the reliability. However, the spatial topology of a cloud computing platform often changes, so static monitoring can bring huge network overhead when collecting multi-node and multi-level monitoring data. To address the above problems, this paper proposes MCloud, an efficient monitoring framework for cloud computing platform. We first propose an organization model for monitoring objects with a tree-linked list. Then, we optimize the indexing mechanism for the efficient retrieval of monitoring data. Finally, we design and implement a monitoring framework MCloud integrated in cloud computing platforms, and the experimental results show that MCloud can effectively reduce the monitoring performance of the cloud platform by more than 20%.