使用系统调用感知调度保护基于容器的云

Michael V. Le, Salman Ahmed, Dan Williams, H. Jamjoom
{"title":"使用系统调用感知调度保护基于容器的云","authors":"Michael V. Le, Salman Ahmed, Dan Williams, H. Jamjoom","doi":"10.1145/3579856.3582835","DOIUrl":null,"url":null,"abstract":"Container-based clouds—in which containers are the basic unit of isolation—face security concerns because, unlike Virtual Machines, containers directly interface with the underlying highly privileged kernel through the wide and vulnerable system call interface. Regardless of whether a container itself requires dangerous system calls, a compromised or malicious container sharing the host (a bad neighbor) can compromise the host kernel using a vulnerable syscall, thereby compromising all other containers sharing the host. In this paper, rather than attempting to eliminate host compromise, we limit the effectiveness of attacks by bad neighbors to a subset of the cluster. To do this, we propose a new metric dubbed Extraneous System call Exposure (ExS). Scheduling containers to minimize ExS reduces the number of nodes that expose a vulnerable system call and as a result the number of affected containers in the cluster. Experimenting with 42 popular containers on SySched, our greedy scheduler implementation in Kubernetes, we demonstrate that SySched can reduce up to 46% more victim nodes and up to 48% more victim containers compared to the Kubernetes default scheduling while also reducing overall host attack surface by 20%.","PeriodicalId":156082,"journal":{"name":"Proceedings of the 2023 ACM Asia Conference on Computer and Communications Security","volume":"6 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2023-07-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Securing Container-based Clouds with Syscall-aware Scheduling\",\"authors\":\"Michael V. Le, Salman Ahmed, Dan Williams, H. Jamjoom\",\"doi\":\"10.1145/3579856.3582835\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Container-based clouds—in which containers are the basic unit of isolation—face security concerns because, unlike Virtual Machines, containers directly interface with the underlying highly privileged kernel through the wide and vulnerable system call interface. Regardless of whether a container itself requires dangerous system calls, a compromised or malicious container sharing the host (a bad neighbor) can compromise the host kernel using a vulnerable syscall, thereby compromising all other containers sharing the host. In this paper, rather than attempting to eliminate host compromise, we limit the effectiveness of attacks by bad neighbors to a subset of the cluster. To do this, we propose a new metric dubbed Extraneous System call Exposure (ExS). Scheduling containers to minimize ExS reduces the number of nodes that expose a vulnerable system call and as a result the number of affected containers in the cluster. Experimenting with 42 popular containers on SySched, our greedy scheduler implementation in Kubernetes, we demonstrate that SySched can reduce up to 46% more victim nodes and up to 48% more victim containers compared to the Kubernetes default scheduling while also reducing overall host attack surface by 20%.\",\"PeriodicalId\":156082,\"journal\":{\"name\":\"Proceedings of the 2023 ACM Asia Conference on Computer and Communications Security\",\"volume\":\"6 1\",\"pages\":\"0\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2023-07-10\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Proceedings of the 2023 ACM Asia Conference on Computer and Communications Security\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1145/3579856.3582835\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Proceedings of the 2023 ACM Asia Conference on Computer and Communications Security","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1145/3579856.3582835","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0

摘要

基于容器的云(其中容器是隔离的基本单元)面临安全问题,因为与虚拟机不同,容器通过广泛且易受攻击的系统调用接口直接与底层高度特权的内核进行接口。无论容器本身是否需要危险的系统调用,共享主机(坏邻居)的受损或恶意容器都可以使用易受攻击的系统调用危及主机内核,从而危及共享主机的所有其他容器。在本文中,我们不是试图消除主机妥协,而是将恶意邻居攻击的有效性限制在集群的一个子集中。为了做到这一点,我们提出了一个新的度量,称为外部系统调用暴露(ExS)。调度容器以最小化ExS可以减少暴露易受攻击的系统调用的节点数量,从而减少集群中受影响的容器数量。在SySched(我们在Kubernetes中的贪婪调度器实现)上对42个流行的容器进行了实验,我们证明,与Kubernetes的默认调度相比,SySched可以减少多达46%的受害节点和多达48%的受害容器,同时还将总体主机攻击面减少了20%。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
Securing Container-based Clouds with Syscall-aware Scheduling
Container-based clouds—in which containers are the basic unit of isolation—face security concerns because, unlike Virtual Machines, containers directly interface with the underlying highly privileged kernel through the wide and vulnerable system call interface. Regardless of whether a container itself requires dangerous system calls, a compromised or malicious container sharing the host (a bad neighbor) can compromise the host kernel using a vulnerable syscall, thereby compromising all other containers sharing the host. In this paper, rather than attempting to eliminate host compromise, we limit the effectiveness of attacks by bad neighbors to a subset of the cluster. To do this, we propose a new metric dubbed Extraneous System call Exposure (ExS). Scheduling containers to minimize ExS reduces the number of nodes that expose a vulnerable system call and as a result the number of affected containers in the cluster. Experimenting with 42 popular containers on SySched, our greedy scheduler implementation in Kubernetes, we demonstrate that SySched can reduce up to 46% more victim nodes and up to 48% more victim containers compared to the Kubernetes default scheduling while also reducing overall host attack surface by 20%.
求助全文
通过发布文献求助,成功后即可免费获取论文全文。 去求助
来源期刊
自引率
0.00%
发文量
0
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
确定
请完成安全验证×
copy
已复制链接
快去分享给好友吧!
我知道了
右上角分享
点击右上角分享
0
联系我们:info@booksci.cn Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。 Copyright © 2023 布克学术 All rights reserved.
京ICP备2023020795号-1
ghs 京公网安备 11010802042870号
Book学术文献互助
Book学术文献互助群
群 号:604180095
Book学术官方微信