hive-24hr

24 hours of logs generated from a Hive stack consisting of Hive, YARN, MapReduce, and HDFS driven by HiBench workloads that repeatedly create, query, and delete tables. This dataset was first used in the evaluation of "Non-Intrusive Performance Profiling for Entire Software Stacks Based on the Flow Reconstruction Principle."

该数据集包含由HiBench工作负载驱动的Hive技术栈生成的24小时日志，该技术栈涵盖Hive、YARN、MapReduce与HDFS，其所运行的工作负载会反复执行表的创建、查询与删除操作。该数据集首次被应用于《基于流重构原理的完整软件栈非侵入式性能剖析》一文的评估工作。