CSR-SMA: Language Support for Data-Centric Systems Monitoring

必须监控复杂的分布式系统，以发现问题、记录/存档系统健康状况、监督系统运行、检测恶意活动以及执行无数其他任务。 为了提高系统监控器的可靠性、安全性、性能、易构建性和易用性，本研究开发了一种高级的、特定于领域的语言来指定监控系统积累、归档和呈现给用户的数据。 使用分布式数据源的高级规范，我们的编译器生成可靠，安全和高性能的监控工具，并发地获取分布式数据，归档（自我描述的）数据用于以后的分析，查询数据以解决问题，显示统计数据摘要以监视实时系统健康。为了产生广泛的影响，PI将开发用于监视分布式系统的工具，包括集群和广域平台，如PlanetLab，这是一个全球网络研究测试平台，具有400-450个节点和200 -250个在任何给定时间运行的网络实验。 我们将使这项技术开源，使其他人能够将监控系统用于他们自己的项目。此外，我们还将与工业合作伙伴合作，将我们的监测技术转移到工业中。最后，正在构建的语言和编译器系统将在网络社区之外产生广泛的影响，因为监测系统中发现的广告数据也出现在自然和社会科学中，包括生物学，化学，因此，将有可能使用规范语言来描述科学数据集的格式，并生成查询和可视化工具，以帮助提高广泛的生产力。计算机科学家

Complex distributed systems must be monitored to find problems,record/archive system health, oversee system operation, detectmalicious activity, and perform a myriad of other tasks. To improvethe reliability, security, performance, ease-of-construction andmaintaince of system monitors, this research develops a high-level,domain-specific language to specify the data that monitoring systemsaccumulate, archive and present to users. Using high-levelspecifications of distributed data sources, our compiler generatesreliable, secure, and high-performance monitoring tools thatconcurrently fetch distributed data, archive (self-describing) datafor later analysis, query data to troubleshoot problems, and displaystatistical data summaries to monitor real-time system health.To make a broad impact, the PIs will develop tools for monitoringdistributed systems, including clusters and wide-area platforms suchas PlanetLab, a global network research testbed with 400-450 nodes and200-250 network experiments running at any given time. We will makethe technology open-source, enabling others to use the monitoringsystem for their own projects. Additionally, we will also work withindustrial partners to transfer our monitoring technology to industry.Finally, the language and compiler system being built will make abroad impact outside the networking community as the kind of ad hocdata found in monitoring systems also appears across the natural andsocial sciences, including biology, chemistry, physics and economics.Hence it will be possible to use the specification language todescribe the formats of scientific data sets and to generate queryingand visualization tools that help improve the productivity of a widerange of computational scientists.