Lock scheduling using machine learning

• US 10,613,899 B1
• Filed: 11/09/2018
• Issued: 04/07/2020
• Est. Priority Date: 11/09/2018
• Status: Active Grant

• Current Assignee: ServiceNow Incorporated
• Original Assignee: ServiceNow Incorporated
• Inventors: Saha, Prabodh, Mall, Amit Kumar, Shende, Manojkumar Haridas
• Primary Examiner(s): Kessler, Gregory A
• Application Number:
  US16/185,815
• Time in Patent Office: 515 Days
• Field of Search: None
• CPC Class Codes: G06F 2209/5014 Reservation
• CPC Class Codes: G06F 2209/5019 Workload prediction
• CPC Class Codes: G06F 9/4881 Scheduling strategies for dispatcher, e.g. round robin, multi-level priority queues
• CPC Class Codes: G06F 9/5005 to service a request
• CPC Class Codes: G06N 20/00 Machine learning
• CPC Class Codes: G06N 20/20 Ensemble learning
• CPC Class Codes: G06N 5/04 Inference or reasoning models

Abstract

The present approach relates to systems and methods for facilitating run time predictions for cloud-computing automated tasks (e.g., automated tasks), and using the predicted run time to schedule resource locking. A predictive model may predict the automated task run time based on historical run time to completion, and the run time may be updated using machine learning. Resource lock schedules may be determined for a queue of automated tasks utilizing the resource based on the predicted run time for the various types of automated tasks. The predicted run time may be used to reserve a resource for the given duration, such that the resource is not available for use for another task.

Claims

1. A system, comprising:

one or more client instances hosted by a platform, wherein the one or more client instances are accessible by one or more remote client networks;

a memory configured to store instructions;

a processor configured to execute instructions, wherein the instructions, when executed, cause the processor to perform operations comprising;

receiving a request to perform an automated task, wherein the automated task impacts at least one resource utilized by the system;

in response to the request, determining a predicted run time for the automated task using a statistical predictive model, wherein the automated task is associated with one or both of a title or a description to characterize the automated task as a task type for use by the statistical predictive model in generating a predicted run time for the task type;

locking a respective resource based on at least the predicted run time; and

executing the automated task.

2. The system of claim 1, wherein the system is further configured to:

update a historical database used to parameterize the statistical predictive model based on an actual run time of the automated task.

3. The system of claim 1, wherein the statistical predictive model performs linear regression, random forest, or a combination thereof, to determine the predicted run time.

4. The system of claim 1, wherein the respective resource is reserved for the predicted run time plus a buffer time.

5. The system of claim 2, wherein the statistical predictive model uses the update to predict subsequent runtimes for automated tasks.

6. The system of claim 5, wherein a resource lock schedule is determined for automated tasks, wherein the resource lock schedule locks the respective resource for a duration based on one or more predicted run times for the automated tasks.

7. The system of claim 6, wherein a plurality of automated tasks utilizing the respective resource are in a queue based on the resource lock schedule.

8. The system of claim 7, wherein one or more subsequent automated tasks wait in queue past a respective predicted runtime when an initial automated task is not completed within the predicted run time.

9. The system of claim 8, wherein the respective resource is locked past the predicted run time of the initial automated task when the initial automated task is not completed.

10. A method for determining resource lock duration for an automated task performed in a computing system comprising one or more client instances hosted by a platform, wherein the one or more client instances are accessible by one or more remote client networks, the method comprising:

receiving a request to perform an automated task, wherein the automated task impacts at least one resource utilized by the system;

in response to the request, determining a predicted run time for the automated task using a statistical predictive model, wherein the automated task is associated with one or both of a title or a description to characterize the automated task as a task type for use by the statistical predictive model in generating a predicted run time for the task type;

locking a respective resource based on at least the predicted run time; and

executing the automated task.

11. The method of claim 10, wherein the predicted run time for the automated task is based on a historical database of data, records, or a combination thereof, corresponding to prior automated tasks.

12. The method of claim 11, wherein the method updates the historical database used to parameterize the statistical predictive model based on an actual run time of the automated task.

13. The method of claim 10, wherein the statistical predictive model performs linear regression, random forest, or a combination thereof, to determine the predicted run time.

14. The method of claim 10, wherein the respective resource is reserved for the predicted run time plus a buffer time.

15. The method of claim 12, wherein the statistical predictive model uses the updated historical database to predict subsequent runtimes for automated tasks.

16. A non-transitory computer-readable storage medium storing executable instructions that, when executed by a processor, cause operations to be performed comprising:

receiving a request by a computing system to perform an automated task, wherein the automated task impacts at least one resource utilized by the computing system;

in response to the request, determining a predicted run time for the automated task using a statistical predictive model, wherein the automated task is associated with one or both of a title or a description to characterize the automated task as a task type for use by the statistical predictive model in generating a predicted run time for the task type;

locking a respective resource based on at least the predicted run time; and

executing the automated task.

17. The non-transitory computer-readable storage medium of claim 16, wherein a resource lock schedule is determined for automated tasks, wherein the resource lock schedule locks the respective resource for a duration based on one or more predicted run times for the automated tasks.

18. The non-transitory computer-readable storage medium of claim 17, wherein a plurality of automated tasks utilizing the respective resource are in a queue based on the resource lock schedule.

19. The non-transitory computer-readable storage medium of claim 18, wherein one or more subsequent automated tasks wait in queue past a respective predicted runtime when an initial automated task is not completed within the predicted run time.