遇见数据集

The Paradox of No Logging Being Slower Than Logging - Data

收藏
Zenodo2023-08-02 更新2026-05-26 收录
数据链接:
官方服务:

资源简介:

This dataset contains the result data of the paper "More is Less in MooBench? The Paradox of No Logging Being Slower Than Logging". In particular, it contains: <strong>reproducability.tar.xz</strong>: The data for reproducing the no-logging-faster-than-logging effect on i7-4770, i7-6700 and Raspberry Pi 4 for Java 11 and 17 (Section 3.1), and the data of increasing the artificial overhead (Section 3.3) <strong>results.tar.xz</strong>: The plotted results from the resproducibility execution <strong>graphs_100_withPerf.tar.xz</strong>: The resulting data for the analysis with perf (Section 3.2) <strong>dumpqueue.tar.xz</strong>: The data for using the DumpQueue, which was freshly introduced because of the findings of this paper All results are reproducible using the commands from the following gist: https://gist.github.com/DaGeRe/c7159297bad4bd13fb556dcf696d0699 In case the used repositories do not exist any longer, this dataset contains <strong>REPRODUCTION.md</strong>, which describes the execution <strong>moobench-2023-08-02.tar.xz</strong>, which is the git repository in the state that was used to execute the experiments

本数据集收录了论文《MooBench中“多即是少”?无日志记录反而比有日志记录更快的悖论》(原文标题:More is Less in MooBench? The Paradox of No Logging Being Slower Than Logging)的相关实验结果数据。具体包含以下内容: <strong>reproducability.tar.xz</strong>:用于复现实验的数据集,包含在i7-4770、i7-6700及树莓派4(Raspberry Pi 4)设备上,针对Java 11与Java 17环境复现“无日志记录比有日志记录更快”效应的相关数据(对应论文3.1章节),以及人工开销递增实验的相关数据(对应论文3.3章节); <strong>results.tar.xz</strong>:复现实验执行后得到的可视化结果数据; <strong>graphs_100_withPerf.tar.xz</strong>:借助perf性能分析工具开展分析所得到的实验结果数据(对应论文3.2章节); <strong>dumpqueue.tar.xz</strong>:针对DumpQueue的相关实验数据——该组件是基于本论文的研究发现全新引入的。 所有实验结果均可通过下述Gist链接中的命令进行复现:https://gist.github.com/DaGeRe/c7159297bad4bd13fb556dcf696d0699。 若原使用的代码仓库已无法访问,本数据集包含<strong>REPRODUCTION.md</strong>文档,其中详细说明了实验执行流程;同时包含<strong>moobench-2023-08-02.tar.xz</strong>归档文件,该文件为执行本实验时所用Git仓库的对应快照版本。

提供机构:
Zenodo
创建时间:
2023-08-02
二维码
社区交流群
二维码
科研交流群
商业服务