利索-全球专利检索

  1. 登录
  2. 注册

搜索结果

66 条记录 (耗时 0.025 秒)

    Job scheduling analysis method and system using historical job execution data
    27.
    发明授权
    Job scheduling analysis method and system using historical job execution data 失效
    作业调度分析方法和系统使用历史作业执行数据

    公开(公告)号：US06889243B1

    公开(公告)日：2005-05-03

    申请号：US09249765

    申请日：1999-02-16

    申请人： Yuri Hondou ,  Hirofumi Nagasuka ,  Tadashi Yamagishi ,  Shunji Tanaka ,  Toshiyuki Kinoshita

    发明人： Yuri Hondou ,  Hirofumi Nagasuka ,  Tadashi Yamagishi ,  Shunji Tanaka ,  Toshiyuki Kinoshita

    IPC分类号： G06F9/48 ,  G06F9/00

    CPC分类号： G06F9/4881

    摘要： A job scheduling analysis method and system are disclosed in which a job schedule is analyzed by use of historical job execution data in a computer system in which a plurality of jobs are executed in parallel. Historical execution data of a plurality of jobs and the file names of files accessed by the jobs are collected. The maximum multiplicity of jobs capable of operating in parallel on the computer system is inputted. When the file name of a file accessed by one job and the file name of a file accessed by another other job coincide with each other, an execution start condition of the plurality of jobs are determined to execute the one job and the other job at the earliest instants within the maximum job multiplicity so that the sequence of execution of processings by the one job and the other job is maintained and the execution time of the one job and the execution time of said other job do not overlap each other. Thereby, it is possible to simulate the influence of a change in system construction and to search for an effective batch processing with a reduced number of idle spaces.

    摘要翻译： 公开了一种作业调度分析方法和系统，其中通过在并行执行多个作业的计算机系统中使用历史作业执行数据来分析作业调度。 收集多个作业的历史执行数据和由作业访问的文件的文件名。 输入能够在计算机系统上并行操作的最大作业多样性。 当由一个作业访问的文件的文件名和由另一个作业访问的文件的文件名彼此一致时，确定多个作业的执行开始条件以执行一个作业，并且另一个作业在 在最大工作多样性内的最早时刻，使得维护一个作业和另一个作业的处理执行顺序，并且一个作业的执行时间和所述其他作业的执行时间彼此不重叠。 因此，可以模拟系统构造的变化的影响，并且以空闲空间减少的方式搜索有效的批处理。