GoCJ: Google Cloud Jobs Dataset

The GoCJ dataset is comprised of different files. Each file contains the sizes of a specified number of jobs in terms of Million Instructions (MI) derived from the workload behaviors witnessed in google cluster traces (surveys published). The name of the file determines the number of jobs within the file i.e.; GoCJ_Dataset_100 has the sizes of 100 jobs. The GoCJ dataset is accepted for publication in MDPI Data journal, the citation of the article will be provied very soon. Moreover, the GoCJ dataset (i..e named as Google-like realistic dataset) is used in the following research work of the author. Hussain, A., Aleem, M., Khan, A. et al. Cluster Comput (2018). https://doi.org/10.1007/s10586-018-2414-6GoCJ数据集由多个不同文件构成。每个文件存储指定数量作业的指令规模数据，单位为百万指令（Million Instructions，MI），数据源自已公开的谷歌集群追踪调研中的实际工作负载行为。文件名直接对应文件内包含的作业总数，例如GoCJ_Dataset_100即包含100个作业的指令规模数据。本GoCJ数据集已被MDPI Data期刊收录待刊，该文章的引用信息将于近期公布。此外，本GoCJ数据集（又名类谷歌真实数据集）已被应用于作者的如下研究成果：Hussain, A., Aleem, M., Khan, A. 等. 《Cluster Comput》(2018). https://doi.org/10.1007/s10586-018-2414-6

搜集汇总

背景与挑战

背景概述

GoCJ数据集包含多个文件，每个文件记录了不同数量作业的大小(以百万指令MI为单位)，数据来源于Google集群跟踪的工作负载行为。数据集按作业数量分类(如100,150,200等)，并已用于云计算相关研究。