Job scheduling management

摘要

Resource utilization data for a set of system components of a computing system is collected. The resource utilization data includes performance records for a set of jobs. By analyzing the collected resource utilization data for the set of system components, a resource allocation is identified for a particular job of the set of jobs. A first execution time for the particular job is determined based on the resource allocation for the particular job and the resource utilization data for the set of system components. A location at which to execute the particular job is determined based on how the particular job has been executed at the location previously. The first execution time may be a time when the computer system achieves a resource availability threshold with respect to the resource allocation. Aspects are also directed toward performing the particular job at the first execution time.

主权项

1. A computer-implemented method for managing jobs in a computing system, the method being executable by a processor performing operations comprising:
collecting, using a set of performance monitors, resource utilization data for a set of system components of the computing system, the resource utilization data including performance records for a set of jobs and indicating the degree of usage in relation to a system architecture limitations of a respective system component; analyzing the resource utilization data for the set of system components by: identifying, based on a current resource utilization data of the resource utilization data for the set of system components, a set of address spaces for job execution, wherein the set of address spaces has been used previously for a particular job of the set of jobs, wherein the particular job is part of a batch job process that is executed by the computer system at predetermined regular intervals and includes available performance records; generating, using a historical resource utilization data of the resource utilization data from the available performance records, for the set of system components, a set of aggregate summary records for the set of address spaces relating to previous executions of the particular job at the set of address spaces; and selecting, from the set of address spaces, based on the set of aggregate summary records, a first subset of address spaces for execution of the particular job; identifying, by using the analysis of the resource utilization data, a resource allocation for the particular job by; locating, in the resource utilization data for the set of system components, a set of summary records corresponding to a set of execution instances of the particular job, wherein summary records are reports that are generated upon the completion of the particular job each time the particular job runs on the computer system, wherein the summary records include a job execution status and historical resource usage data for the set of execution instances of the particular job; determining a subset of the set of summary records having a job execution status indicating successful execution of the particular job by analyzing each summary record and removing both summary records that indicate abnormal ending and summary records that include a condition recognized as unsatisfactory based on a condition code; computing, based on a subset of the historical resource usage data for the set of execution instances of the particular job which correlates to the subset of the set of summary records, the resource allocation for the particular job; determining, based on the resource allocation for the particular job and the resource utilization data for the set of system components, a first execution time for the particular job, the first execution time being a time when the computing system achieves a resource availability threshold with respect to the resource allocation; and performing, at the first execution time, at first subset of address spaces, the particular job.