Method, device and computer program product for job management

Patent metadata

Field	Value
Publication number	US-11409558-B2
Application number	US-202016802502-A
Country	US
Kind code	B2
Filing date	Feb 26, 2020
Priority date	Oct 18, 2019
Publication date	Aug 9, 2022
Grant date	Aug 9, 2022

Abstract

According to embodiments of the present disclosure, a method, device and computer program product for job management are proposed. The method comprises: receiving, at a first device, status information of a first job sent by a second device at a first sending time, the first job running on the second device, the status information indicating a completion status of the first job; in accordance with a determination that the status information indicates the first job has not been completed, determining a send distribution period related to at least one second job different from the first job, the send distribution period covering respective sending times when devices running the at least one second job send status information of the at least one second job to the first device; and providing the second device an indication on a second sending time in the send distribution period, to instruct the second device to send further status information of the first job to the first device at the second sending time. Therefore, the present solution can achieve flexible self-adaptive job status information transmission.

First claim

We claim:

1. A method for job management, comprising:
receiving, at a first device, status information of a first job sent by a second device at a first sending time, the first job running on the second device, the status information indicating a completion status of the first job;
in accordance with the status information that indicates the first job has not been completed, determining a send distribution period related to at least one second job different from the first job, the send distribution period covering respective sending times when devices running the at least one second job send status information of the at least one second job to the first device;
wherein determining the send distribution period comprises:
determining a current time as a start time of the send distribution period, determining an expected length of the send distribution period; and
determining, based on the expected length of the send distribution period, an end time of the send distribution period;

wherein determining the expected length comprises:
obtaining a resource usage of the first device and the respective sending time of the status information of the at least one second job; and
determining the expected length based on the resource usage and the respective sending times of the status information of the at least one second job; and

providing the second device an indication on a second sending time in the send distribution period, to instruct the second device to send further status information of the first job to the first device at the second sending time.

2. The method of claim 1 , wherein obtaining the resource usage comprises obtaining at least one of the following:
processing resource usage of the first device;
storage resource usage of the first device; and
input/output usage of the first device.

3. The method of claim 1 , wherein determining the expected length based on the resource usage and the respective sending times of the status information of the at least one second job, comprises:
computing a set of time intervals based on the respective sending times of the status information of the at least one second job, each time interval in the set of time intervals being a difference of two adjacent respective sending times;
determining a candidate period associated with the expected length;
determining, based on the resource usage and the set of time intervals, expected resource usage of the first device within the candidate period; and
in accordance with a determination that the expected resource usage is lower than a predetermined usage threshold and the candidate period has a time length greater than a predetermined time length, determining the time length as the expected length.

4. The method of claim 1 , wherein determining the end time of the send distribution period comprises:
obtaining a sending time of status information of a last job in the at least one second job; and
in accordance with a determination that the expected length exceeds an interval length between the sending time of the status information of the last job and the start time of the send distribution period, adding the start time and the expected length, to determine the end time of the send distribution period; and
in accordance with a determination that the expected length is less than the interval length, dividing the expected length by a number of the at least one second job, to obtain a divided result, and adding the divided result and the sending time of the status information of the last job, to determine the end time of the send distribution period.

5. The method of claim 1 , wherein providing the second device the indication on the second sending time in the send distribution period comprises:
determining two adjacent times having a maximum interval therebetween in the send distribution period;
determining a candidate time falling between the two adjacent times as the second sending time; and
providing the second device the indication on the second sending time.

6. The method of claim 1 , wherein the resource usage of the first device is predicted using a neural network model.

7. The method of claim 1 , wherein the expected length of the send distribution period is changed dynamically with a load of the first device.

8. A device for job management, comprising:
at least one processing unit; and
at least one memory coupled to the at least one processing unit and storing instructions executed by the at least one processing unit, the instructions, when executed by the at least one processing unit, causing the device to execute acts comprising:
receiving, at a first device, status information of a first job sent by a second device at a first sending time, the first job running on the second device, the status information indicating a completion status of the first job;
in accordance with the status information that indicates the first job has not been completed, determining a send distribution period related to at least one second job different from the first job, the send distribution period covering respective sending times when devices running the at least one second job send status information of the at least one second job to the first device;

wherein determining the send distribution period comprises:
determining a current time as a start time of the send distribution period, determining an expected length of the send distribution period; and
determining, based on the expected length of the send distribution period, an end time of the send distribution period;

wherein determining the expected length comprises:
obtaining a resource usage of the first device and the respective sending time of the status information of the at least one second job; and
determining the expected length based on the resource usage and the respective sending times of the status information of the at least one second job; and
providing the second device an indication on a second sending time in the send distribution period, to instruct the second device to send further status information of the first job to the first device at the second sending time.

9. The device of claim 8 , wherein obtaining the resource usage comprises obtaining at least one of the following:
processing resource usage of the first device;
storage resource usage of the first device; and
input/output usage of the first device.

10. The device of claim 8 , wherein determining the expected length based on the resource usage and the respective sending times of the status information of the at least one second job, comprises:
computing a set of time intervals based on the respective sending times of the status information of the at least one second job, each time interval in the set of time intervals being a difference of two adjacent respective sending times;
determining a candidate period associated with the expected length;
determining, based on the resource usage and the set of time intervals, expected resource usage of the first device within the candidate period; and
in accordance with a determination that the expected resource usage is lower than a predetermined usage threshold and the candidate period has a time length greater than a predetermined time length, determining the time length as the expected length.

11. The device of claim 8 , wherein determining the end time of the send distribution period comprises:
obtaining a sending time of status information of a last job in the at least one second job; and
in accordance with a determination that the expected length exceeds an interval length between the sending time of the status information of the last job and the start time of the send distribution period, ad

Assignees

• Emc Ip Holding Co Llc

Inventors

• Wen Jian
• Wang Yi
• Min Xing

Classifications

• G06F9/4881 Primary

  Scheduling strategies for dispatcher, e.g. round robin, multi-level priority queues · CPC title

• G06F9/4856 Primary

  resumption being on a different machine, e.g. task migration, virtual machine migration (G06F9/5088 takes precedence) · CPC title