31 January 2022 in ACM Computing Surveys
Containerization is a lightweight application virtualization technology, providing high environmental consistency, operating system distribution portability, and resource isolation. Existing mainstream cloud service providers have prevalently adopted container technologies in their distributed system infrastructures for automated application management. To handle the automation of deployment, maintenance, autoscaling, and networking of containerized applications, container orchestration is proposed as an essential research problem. However, the highly dynamic and diverse feature of cloud workloads and environments considerably raises the complexity of orchestration mechanisms. Machine learning algorithms are accordingly employed by container orchestration systems for behavior modeling and prediction of multi-dimensional performance metrics. Such insights could further improve the quality of resource provisioning decisions in response to the changing workloads under complex environments. In this article, we present a comprehensive literature review of existing machine learning-based container orchestration approaches. Detailed taxonomies are proposed to classify the current researches by their common features. Moreover, the evolution of machine learning-based container orchestration technologies from the year 2016 to 2021 has been designed based on objectives and metrics. A comparative analysis of the reviewed techniques is conducted according to the proposed taxonomies, with emphasis on their key characteristics. Finally, various open research challenges and potential future directions are highlighted.
17 April 2020 in ACM Transactions on Internet Technology
Containers, as a lightweight application virtualization technology, have recently gained immense popularity in mainstream cluster management systems like Google Borg and Kubernetes. Prevalently adopted by these systems for task deployments of diverse workloads such as big data, web services, and IoT, they support agile application deployment, environmental consistency, OS distribution portability, application-centric management, and resource isolation. Although most of these systems are mature with advanced features, their optimization strategies are still tailored to the assumption of a static cluster. Elastic compute resources would enable heterogeneous resource management strategies in response to the dynamic business volume for various types of workloads. Hence, we propose a heterogeneous task allocation strategy for cost-efficient container orchestration through resource utilization optimization and elastic instance pricing with three main features. The first one is to support heterogeneous job configurations to optimize the initial placement of containers into existing resources by task packing. The second one is cluster size adjustment to meet the changing workload through autoscaling algorithms. The third one is a rescheduling mechanism to shut down underutilized VM instances for cost saving and reallocate the relevant jobs without losing task progress. We evaluate our approach in terms of cost and performance on the Australian National Cloud Infrastructure (Nectar). Our experiments demonstrate that the proposed strategy could reduce the overall cost by 23% to 32% for different types of cloud workload patterns when compared to the default Kubernetes framework.