CT-T: A Laboratory Workbench for Security Research

NSF 0524096 CT-T：安全研究的实验室指南Jay Lepreau这项研究开发了直接改善安全相关软件系统研究的新技术和技术。 研究可信软件和恶意软件的主要障碍，特别是在计算机网络上运行的软件，是提供现实但受控的环境。 这些环境不仅必须提供强有力的遏制保证，而且还必须提供与外部世界的强有力的连接，既要现实，又要能够分析和控制被测系统。第一种是允许分布式系统暂停、设置检查点以及在时间上前后移动的技术。 这项研究导致了网络系统中混合真实的和虚拟时间的进步。 第二种方式是实验工作流程的工作台，允许人们在实验过程中向前和向后移动。 这在许多领域都取得了进展，包括分布式系统控制，科学工作流管理，大规模数据存储，数据分析和可视化。 第三个“及时”的结果是定期部署这个基础设施，作为流行的基于Web的“Emulab”测试平台软件的一部分。 该项目的工具不断部署在在线网络测试平台中，并交付给研究，教育和安全社区使用和扩展。

NSF 0524096CT-T: A Laboratory Workbench for Security ResearchJay LepreauThis research develops new techniques and technologies that directly improve the study of security-related software systems. A primary barrier to studying both trustworthy and malicious software, especially software that runs on networks of computers, is providing realistic but controlled environments. The environments must provide not only strong containment guarantees, but also strong connections to the outside world, both for realism and to enable analysis and control of the systems under test.This research develops and deploys new technologies to make such environments publicly available, in three "timely" ways. The first is technology that allows distributed systems to be paused, checkpointed, and moved forward and backward in time. This research is leading to advancements in mixing real and virtual time in networked systems. The second way is a workbench for experiment workflows, allowing people to move forward and backward through the experimentation process. This is yielding advances in many areas including distributed system control, scientific workflow management, large-scale data storage, data analysis, and visualization. The third "timely" result is the regular deployment of this infrastructure as part of the popular Web-based "Emulab" testbed software. The project's tools are continuously deployed in online network testbeds, and delivered to the research, education, and security community for use and extension.