Automatic compute environment scheduling using machine learning

Patent metadata

Field	Value
Publication number	US-12417128-B2
Application number	US-202217705516-A
Country	US
Kind code	B2
Filing date	Mar 28, 2022
Priority date	Mar 28, 2022
Publication date	Sep 16, 2025
Grant date	Sep 16, 2025

Abstract

Various embodiments provide techniques for automatic compute environment scheduling using machine learning (ML). This includes identifying a first compute resource among a plurality of compute resources operating in a compute infrastructure, where the first compute resource is in a first operational state. It further includes determining, based on comparing a first time with a compute resources schedule generated using an ML model, that the first compute resource should be placed in a second operational state different from the first operational state. It further includes determining whether the compute resources schedule should be disregarded, and either (1) in response to determining that the compute resources schedule should not be disregarded, placing the first compute resource in the second operational state, or (2) in response to determining that the compute resources schedule should be disregarded, allowing the first compute resource to remain in the first operational state.

First claim

What is claimed is:

1. A method, comprising:
identifying a first compute resource among a plurality of compute resources operating in a compute infrastructure, wherein the first compute resource is in a first operational state;
determining, based on comparing a first time with a compute resources schedule generated using a machine learning (ML) model, that the first compute resource should be placed in a second operational state different from the first operational state;
determining whether the compute resources schedule should be disregarded; and
(1) in response to determining that the compute resources schedule should not be disregarded, placing the first compute resource in the second operational state and automatically modifying a disabled state of a first portion of a user interface to enable overriding of the first compute resource being placed in the second operational state while maintaining a disabled state of a second portion of the user interface that enables overriding associated with other compute resources, or
(2) in response to determining that the compute resources schedule should be disregarded, allowing the first compute resource to remain in the first operational state and maintaining the disabled state of the first portion of the user interface and the second portion of the user interface.

2. The method of claim 1 , further comprising:
identifying that the first compute resource is in the second operational state;
determining, based on comparing a second time with the compute resources schedule generated using the ML model, that the first compute resource should be placed in the first operational state, wherein the second time is after the first time; and
placing the first compute resource in the first operational state.

3. The method of claim 1 ,
wherein the first operational state is activated and the second operational state is deactivated, or
wherein the first operational state is deactivated and the second operational state is activated.

4. The method of claim 1 , wherein determining that the schedule should be disregarded comprises:
determining at least one of:
occurrence of user activity for the first compute resource within a first threshold time period before the first time; or
reception of a user override within a second threshold time period before the first time.

5. The method of claim 4 , further comprising:
updating the ML model based on the at least one of the user activity or the user override, wherein updating the ML model comprises:
training the ML model using additional training data including the at least one of the user activity or the user override; and
generating a second compute resources schedule using the updated ML model.

6. The method of claim 1 , further comprising:
identifying that the first compute resource is in the second operational state;
receiving an indication of a user override relating to the first compute resource; and
restoring the first compute resource to the first operational state, based on the received indication.

7. The method of claim 6 , further comprising:
disabling the first portion of the user interface for the user override relating to the first compute resource, based on the restoring the first compute resource to the first operational state, wherein the indication of the user override is received using the user interface.

8. The method of claim 1 , further comprising:
identifying, based on the first compute resource and a dependency graph, a second compute resource to be placed in the second operational state.

9. The method of claim 8 , wherein identifying, based on the first compute resource and the dependency graph, the second compute resource to be placed in the second operation state, further comprises:
determining, based on the dependency graph, that a first software module associated with the first compute resource uses a second software module associated with the second compute resource; and
determining, based on the dependency graph, that no other software modules that use the second software module remain in the first operational state.

10. The method of claim 1 , wherein the plurality of compute resources operating in the compute infrastructure comprises at least one of: (i) a plurality of cloud computing resources, or (ii) a plurality of on-premises computing resources.

11. A non-transitory computer-readable medium containing computer program code that, when executed by operation of one or more computer processors, performs operations comprising:
identifying a first compute resource among a plurality of compute resources operating in a compute infrastructure, wherein the first compute resource is in a first operational state;
determining, based on comparing a first time with a compute resources schedule generated using a machine learning (ML) model, that the first compute resource should be placed in a second operational state different from the first operational state;
determining whether the compute resources schedule should be disregarded; and
(1) in response to determining that the compute resources schedule should not be disregarded, placing the first compute resource in the second operational state and automatically modifying a disabled state of a first portion of a user interface to enable overriding of the first compute resource being placed in the second operational state while maintaining a disabled state of a second portion of the user interface that enables overriding associated with other compute resources, or
(2) in response to determining that the compute resources schedule should be disregarded, allowing the first compute resource to remain in the first operational state and maintaining the disabled state of the first portion of the user interface and the second portion of the user interface.

12. The non-transitory computer-readable medium of claim 11 , the operations further comprising:
identifying that the first compute resource is in the second operational state;
determining, based on comparing a second time with the compute resources schedule generated using the ML model, that the first compute resource should be placed in the first operational state, wherein the second time is after the first time; and
placing the first compute resource in the first operational state.

13. The non-transitory computer-readable medium of claim 11 ,
wherein the first operational state is activated and the second operational state is deactivated, or
wherein the first operational state is deactivated and the second operational state is activated.

14. The non-transitory computer-readable medium of claim 11 , wherein determining that the schedule should be disregarded comprises:
determining at least one of:
occurrence of user activity for the first compute resource within a first threshold time period before the first time; or
reception of a user override within a second threshold time period before the first time.

15. The non-transitory computer-readable medium of claim 14 , the operations further comprising:
updating the ML model based on the at least one of the user activity or the user override, wherein updating the ML model comprises:
training the ML model using additional training data including the at least one of the user activity or the user override; and
generating a second compute resources schedule using the updated ML model.

16. The non-transitory computer-readable medium of claim 11 , the operations further comprising:
identifying, based on the first compute resource and a dependency graph, a second compute resource to be placed in the

Assignees

• Disney Entpr Inc

Inventors

• Vinson Katherine S
• Cox Jason Alexander
• Arredondo Adam
• Scott Brian Lawrence
• O'Connor Patrick
• Tatum James H
• Wagner Steven

Classifications

• G06N20/00

  Machine learning · CPC title

• G06F9/4881

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

• G06F9/5094

  where the allocation takes into account power or heat criteria (power management in computers in general G06F1/3203; thermal management in computers in general G06F1/206) · CPC title

• G06F9/5055 Primary

  considering software capabilities, i.e. software resources associated or available to the machine · CPC title

• G06F9/4893 Primary

  taking into account power or heat criteria (power management in computers in general G06F1/3203; thermal management in computers in general G06F1/206) · CPC title