An Efficient Online Placement Scheme for Cloud Container Clusters

Containers represent an agile alternative to virtual machines (VMs), for providing cloud computing services. Containers are more flexible and lightweight, and can be easily instrumented. Enterprise users often create clusters of inter-connected containers to provision complex services. Compared to traditional cloud services, key challenges in container cluster (CC) provisioning lie in the optimal placement of containers while considering inter-container traffic in a CC. The challenge further escalates, when CCs are provisioned in an online fashion. We propose an online algorithm to address the above challenges, aiming to maximize the aggregate value of all served clusters. We first study a one-shot CC placement problem. Leveraging techniques of exhaustive sampling and ST rounding, we design an efficient one-shot algorithm to determine the placement scheme of a given CC. We then propose a primal-dual online placement scheme that employs the one-shot algorithm as a building block to make decisions upon the arrival of each CC request. Through both theoretical analysis and trace-driven simulations, we verify that the online placement algorithm is computationally efficient and achieves a good competitive ratio.