CSR-PDOS : Safe at Any Speed: Safe and Fast Distributed Applications

该项目包括用于开发快速、正确的分布式应用程序的技术和工具。有两个方面是新颖的。首先，使用这些技术开发的应用程序模块在并行执行时被静态地分析潜在的冲突。可能发生冲突的模块被序列化以避免错误。其次，在运行时透明地监视应用程序，以显示保守序列化在哪些地方过度地减慢了应用程序的速度。构建并发系统的传统方法是编写模块的串行版本，然后考虑并发执行可能导致的错误。在应用程序可以安全地执行并发请求之前，每个潜在的问题必须适当地同步。该项目分析代码以保守地排除危险的交互，并在并发环境中安全运行。所得到的安全应用程序可以使用开发人员现在使用的相同技术来确保安全性。例如，锁可以用来使两段代码对并发执行是安全的。尽管静态分析允许开发人员了解模块可能发生冲突的地方，但该项目的第二部分，对并发应用程序的透明监控，帮助开发人员将注意力集中在实践中发生的争用领域。该项目的技术将允许开发大型并发应用程序，而不会出现通常由并行计算引入的错误。随着我们使用的软件越来越多地在中心站点上运行，为成千上万的用户进行多路复用，这些担忧正在增长。我们的技术将作为论文在系统和语言会议上传播，并作为展示其实用性的软件工件传播。

This project consists of techniques and tools for developing fast,correct distributed applications. Two aspects are novel. First,application modules developed with these techniques are staticallyanalyzed for potential conflicts when executed in parallel. Modulesthat may conflict are serialized to avoid errors. Second,applications are transparently monitored at run-time to show whereconservative serialization has unduly slowed the application.The traditional approach to building concurrent systems is to writethe serial version of a module and then consider how concurrentexecutions might cause errors. Each potential problem must besynchronized appropriately before the application may be executedsafely on concurrent requests. This project analyzes code toconservatively preclude dangerous interactions and run it safely in aconcurrent environment. The resulting safe application can be madefaster using the same techniques that developers use today to ensuresafety. For example, locks can be employed to make two segments ofcode safe for concurrent execution. Although static analysis allowsdevelopers to learn where modules may conflict, the second piece ofthis project, transparent monitoring of concurrent applications, helpsdevelopers focus their attention on areas of contention that occur inpractice.This project's techniques will allow large, concurrent applications tobe developed without the bugs normally introduced by parallelcomputation. As more of the software we use runs on central sites,multiplexed for thousands of users, these concerns are growing. Ourtechniques will be disseminated as papers at systems and languagesconferences, and as software artifacts that demonstrate their utility.