JavaScript is disabled for your browser. Some features of this site may not work without it.
X-Containers: Breaking Down Barriers to Improve Performance and Isolation of Cloud-Native Containers

Author
Shen, Zhiming; Sun, Zhen; Sela, Gur-Eyal; Bagdasaryan, Eugene; Delimitrou, Christina; Van Renesse, Robbert; Weatherspoon, Hakim
Abstract
“Cloud-native” container platforms, such as Kubernetes, have become an integral part of production cloud environments. One of the principles in designing cloud-native applications is called “Single Concern Principle”, which suggests that each container should handle a single responsibility well. Due to the resulting change in the threat model, process isolation within the container becomes redundant in most single-concerned containers, and inter-container isolation becomes increasingly important. In this paper, we propose a new exokernel-inspired architecture called X-Containers that improves both the security and the performance of cloud-native containers. We show that, through relatively minor modifications, the Xen hypervisor can serve as an exokernel, and Linux can be turned into a LibOS. Doing so results in a highly secure and efficient LibOS platform that, unlike other available LibOSes, supports binary compatibility and multicore processing. X-Containers have up to 27× higher raw system call throughput compared to Docker containers, while also significantly outperforming recent container platforms such as Google’s gVisor, Intel’s Clear Containers, as well as Library OSes like Unikernel and Graphene on web benchmarks.
Sponsorship
This work was partially supported by the National Science Foundation (NSF) CSR-1422544, NSF NeTS CSR-1704742, the U.S. Department of Commerce, National Institute of Standards and Technology (NIST) award number 70NANB17H181, Cisco, and Huawei.
Date Issued
2018-08-29Subject
X-Container; Cloud Computing; Cloud-Native; Containers; Operating System; Virtualization
Rights
Attribution-NonCommercial-NoDerivatives 4.0 International
Type
technical report
The following license files are associated with this item:
Except where otherwise noted, this item's license is described as Attribution-NonCommercial-NoDerivatives 4.0 International