Granted Patent B2

US 12,645,501 · App. 18/366,555 · Granted Jun 2, 2026

Inventors: William Jeffery White (Plano, TX); Said Tabet (Austin, TX)

Assignee: DELL PRODUCTS L.P.

G06F9/5038G06F9/4881G06F9/505

Abstract

Claims (85)

1 . A processor-implemented method for orchestrating edge resource demand load scheduling, the method comprising:

given a set of application job requests comprising one or more application job requests, in which each application job comprises a respective set of tasks:

for each application job of the set of application job requests, generating a respective ordered list of tasks by sorting the respective set of tasks based on total resource demand load size for each task of the respective set of tasks; and

for each particular task of each respective ordered list of tasks for each application job of the set of application job requests:

identifying a respective critical resource demand load size for the particular task;

generating a respective ordered list of candidate systems by sorting a respective set of candidate systems associated with the particular task based on total assignable resource capacity for each candidate system of the respective set of candidate systems;

attempting to identify a target candidate system from the respective ordered list of candidate systems that comprises assignable critical resource capacity that satisfies the critical resource demand load size for the particular task;

if the target candidate system is identified, determining whether the target candidate system comprises assignable non-critical resource capacity that satisfies non-critical resource demand load size for the particular task; and

if the target candidate system comprises assignable non-critical resource capacity that satisfies the non-critical resource demand load size for the particular task, assigning the particular task to the target candidate system.

2 . The processor-implemented method of claim 1 , wherein the total resource demand load size, the respective critical resource demand load size, the non-critical resource demand load size, the total assignable resource capacity, the assignable critical resource capacity, and the assignable non-critical resource capacity are represented in a normalized resource vector framework in which resource demand load and assignable resource capacity are normalized based on reference values for different resource types.

3 . The processor-implemented method of claim 1 , wherein the respective set of candidate systems is determined by:

defining an initial set of candidate systems using a set of constraints associated with the particular task; and

defining the respective set of candidate systems based on an assessment of resource utilization trajectory for one or more candidate systems of the initial set of candidate systems.

4 . The processor-implemented method of claim 1 , wherein the total assignable resource capacity for each candidate system of the respective set of candidate systems is based at least in part on a selectively modifiable slack parameter.

5 . The processor-implemented method of claim 1 , wherein the respective ordered list of tasks is sorted in descending order based on the total resource demand load size for each task of the respective set of tasks, and wherein the respective ordered list of candidate systems is sorted in ascending order based on the total assignable resource capacity for each candidate system of the respective set of candidate systems.

6 . The processor-implemented method of claim 1 , further comprising:

if the target candidate system is not identified, assigning the particular task to a next backlog for processing in a next iteration; or

if the target candidate system fails to comprise assignable non-critical resource capacity that satisfies the non-critical resource demand load size for the particular task, determining whether any subsequent candidate system in the respective ordered list of candidate systems comprises assignable non-critical resource capacity that satisfies the non-critical resource demand load size for the particular task, and (i) in response to determining that a subsequent candidate system comprises assignable non-critical resource capacity that satisfies the non-critical resource demand load size for the particular task, assigning the particular task to the subsequent candidate system, or (ii) in response to determining that no subsequent candidate system comprises assignable non-critical resource capacity that satisfies the non-critical resource demand load size for the particular task, assigning the particular task to a next backlog for processing in a next iteration.

7 . The processor-implemented method of claim 6 , further comprising:

in response to assigning the particular task to the target candidate system or the subsequent candidate system, updating the total assignable resource capacity for the target candidate system or the subsequent candidate system.

8 . The processor-implemented method of claim 1 , further comprising:

given a set of additional tasks comprising one or more individual tasks or one or more backlog tasks associated with a previous iteration:

generating an ordered list of additional tasks by sorting the set of additional tasks based on critical resource demand load size for each additional task of the set of additional tasks; and

for each particular additional task of the ordered list of additional tasks:

generating a respective ordered list of additional candidate systems by sorting a respective set of additional candidate systems associated with the particular additional task based on total assignable resource capacity for each additional candidate system of the respective set of additional candidate systems;

attempting to identify an additional target candidate system from the respective ordered list of additional candidate systems that comprises assignable critical resource capacity that satisfies the critical resource demand load size for the particular additional task;

if the additional target candidate system is identified, determining whether the additional target candidate system comprises assignable non-critical resource capacity that satisfies non-critical resource demand load size for the particular additional task;

if the additional target candidate system comprises assignable non-critical resource capacity that satisfies the non-critical resource demand load size for the particular additional task, assigning the particular additional task to the additional target candidate system;

if the additional target candidate system is not identified, assigning the particular additional task to a next backlog for processing in a next iteration; and

if the additional target candidate system fails to comprise assignable non-critical resource capacity that satisfies the non-critical resource demand load size for the particular additional task, determining whether any additional subsequent candidate system in the respective ordered list of additional candidate systems comprises assignable non-critical resource capacity that satisfies the non-critical resource demand load size for the particular additional task, and (i) in response to determining that an additional subsequent candidate system comprises assignable non-critical resource capacity that satisfies the non-critical resource demand load size for the particular additional task, assigning the particular additional task to the additional subsequent candidate system, or (ii) in response to determining that no additional subsequent candidate system comprises assignable non-critical resource capacity that satisfies the non-critical resource demand load size for the particular additional task, assigning the particular additional task to a next backlog for processing in a next iteration.

9 . The processor-implemented method of claim 1 , wherein the respective ordered list of tasks omits one or more low-priority tasks associated with the application job, and wherein the processor-implemented method further comprises:

for each particular low-priority task of the one or more low-priority tasks:

selecting a subset of candidate systems from a respective set of candidate systems associated with the particular low-priority task, wherein each candidate system of the subset of candidate systems comprises assignable resource capacity;

if a particular candidate system of the subset of candidate systems comprises assignable resource capacity that satisfies a resource demand load associated with the particular low-priority task, assigning the particular low-priority task to the particular candidate system;

if no particular candidate system of the subset of candidate systems comprises assignable resource capacity that satisfies the resource demand load associated with the particular low-priority task, selecting a new subset of candidate systems from the respective set of candidate systems to obtain a particular candidate system that comprises assignable resource capacity that satisfies the resource demand load associated with the particular low-priority task; and

if no particular candidate system of the new subset of candidate systems comprises assignable resource capacity that satisfies the resource demand load associated with the particular low-priority task, assigning the particular low-priority task to a next backlog for processing in a next iteration.

10 . One or more information handling systems collectively comprising:

one or more processors; and

one or more non-transitory computer-readable medium or media comprising one or more sets of instructions which, when executed by at least one of the one or more processors, cause steps to be performed comprising:

given a set of application job requests comprising one or more application job requests, in which each application job comprises a respective set of tasks:

for each application job of the set of application job requests, generating a respective ordered list of tasks by sorting the respective set of tasks based on total resource demand load size for each task of the respective set of tasks; and

for each particular task of each respective ordered list of tasks for each application job of the set of application job requests:

identifying a respective critical resource demand load size for the particular task;

generating a respective ordered list of candidate systems by sorting a respective set of candidate systems associated with the particular task based on total assignable resource capacity for each candidate system of the respective set of candidate systems;

attempting to identify a target candidate system from the respective ordered list of candidate systems that comprises assignable critical resource capacity that satisfies the critical resource demand load size for the particular task;

if the target candidate system is identified, determining whether the target candidate system comprises assignable non-critical resource capacity that satisfies non-critical resource demand load size for the particular task; and

if the target candidate system comprises assignable non-critical resource capacity that satisfies the non-critical resource demand load size for the particular task, assigning the particular task to the target candidate system.

11 . The one or more information handling systems of claim 10 , wherein the total resource demand load size, the respective critical resource demand load size, the non-critical resource demand load size, the total assignable resource capacity, the assignable critical resource capacity, and the assignable non-critical resource capacity are represented in a normalized resource vector framework in which resource demand load and assignable resource capacity are normalized based on reference values for different resource types.

12 . The one or more information handling systems of claim 10 , wherein the respective set of candidate systems is determined by:

defining an initial set of candidate systems using a set of constraints associated with the particular task; and

defining the respective set of candidate systems based on an assessment of resource utilization trajectory for one or more candidate systems of the initial set of candidate systems.

13 . The one or more information handling systems of claim 10 , wherein the respective ordered list of tasks is sorted in descending order based on the total resource demand load size for each task of the respective set of tasks, and wherein the respective ordered list of candidate systems is sorted in ascending order based on the total assignable resource capacity for each candidate system of the respective set of candidate systems.

14 . The one or more information handling systems of claim 10 , wherein the steps to be performed further comprise:

if the target candidate system is not identified, assigning the particular task to a next backlog for processing in a next iteration; or

if the target candidate system fails to comprise assignable non-critical resource capacity that satisfies the non-critical resource demand load size for the particular task, determining whether any subsequent candidate system in the respective ordered list of candidate systems comprises assignable non-critical resource capacity that satisfies the non-critical resource demand load size for the particular task, and (i) in response to determining that a subsequent candidate system comprises assignable non-critical resource capacity that satisfies the non-critical resource demand load size for the particular task, assigning the particular task to the subsequent candidate system, or (ii) in response to determining that no subsequent candidate system comprises assignable non-critical resource capacity that satisfies the non-critical resource demand load size for the particular task, assigning the particular task to a next backlog for processing in a next iteration.

15 . The one or more information handling systems of claim 10 , wherein the steps to be performed further comprise:

given a set of additional tasks comprising one or more individual tasks or one or more backlog tasks associated with a previous iteration:

generating an ordered list of additional tasks by sorting the set of additional tasks based on critical resource demand load size for each additional task of the set of additional tasks; and

for each particular additional task of the ordered list of additional tasks:

generating a respective ordered list of additional candidate systems by sorting a respective set of additional candidate systems associated with the particular additional task based on total assignable resource capacity for each additional candidate system of the respective set of additional candidate systems;

attempting to identify an additional target candidate system from the respective ordered list of additional candidate systems that comprises assignable critical resource capacity that satisfies the critical resource demand load size for the particular additional task;

if the additional target candidate system is identified, determining whether the additional target candidate system comprises assignable non-critical resource capacity that satisfies non-critical resource demand load size for the particular additional task;

if the additional target candidate system comprises assignable non-critical resource capacity that satisfies the non-critical resource demand load size for the particular additional task, assigning the particular additional task to the additional target candidate system;

if the additional target candidate system is not identified, assigning the particular additional task to a next backlog for processing in a next iteration; and

if the additional target candidate system fails to comprise assignable non-critical resource capacity that satisfies the non-critical resource demand load size for the particular additional task, determining whether any additional subsequent candidate system in the respective ordered list of additional candidate systems comprises assignable non-critical resource capacity that satisfies the non-critical resource demand load size for the particular additional task, and (i) in response to determining that an additional subsequent candidate system comprises assignable non-critical resource capacity that satisfies the non-critical resource demand load size for the particular additional task, assigning the particular additional task to the additional subsequent candidate system, or (ii) in response to determining that no additional subsequent candidate system comprises assignable non-critical resource capacity that satisfies the non-critical resource demand load size for the particular additional task, assigning the particular additional task to a next backlog for processing in a next iteration.

16 . The one or more information handling systems of claim 10 , wherein the respective ordered list of tasks omits one or more low-priority tasks associated with the application job, and wherein the steps to be performed further comprise:

for each particular low-priority task of the one or more low-priority tasks:

selecting a subset of candidate systems from a respective set of candidate systems associated with the particular low-priority task, wherein each candidate system of the subset of candidate systems comprises assignable resource capacity;

if a particular candidate system of the subset of candidate systems comprises assignable resource capacity that satisfies a resource demand load associated with the particular low-priority task, assigning the particular low-priority task to the particular candidate system;

if no particular candidate system of the subset of candidate systems comprises assignable resource capacity that satisfies the resource demand load associated with the particular low-priority task, selecting a new subset of candidate systems from the respective set of candidate systems to obtain a particular candidate system that comprises assignable resource capacity that satisfies the resource demand load associated with the particular low-priority task; and

if no particular candidate system of the new subset of candidate systems comprises assignable resource capacity that satisfies the resource demand load associated with the particular low-priority task, assigning the particular low-priority task to a next backlog for processing in a next iteration.

17 . A non-transitory computer-readable medium or media comprising one or more sequences of instructions which, when executed by at least one processor, cause steps to be performed comprising:

given a set of application job requests comprising one or more application job requests, in which each application job comprises a respective set of tasks:

for each application job of the set of application job requests, generating a respective ordered list of tasks by sorting the respective set of tasks based on total resource demand load size for each task of the respective set of tasks;

for each particular task of each respective ordered list of tasks for each application job of the set of application job requests:

identifying a respective critical resource demand load size for the particular task;

generating a respective ordered list of candidate systems by sorting a respective set of candidate systems associated with the particular task based on total assignable resource capacity for each candidate system of the respective set of candidate systems;

attempting to identify a target candidate system from the respective ordered list of candidate systems that comprises assignable critical resource capacity that satisfies the critical resource demand load size for the particular task;

if the target candidate system is identified, determining whether the target candidate system comprises assignable non-critical resource capacity that satisfies non-critical resource demand load size for the particular task; and

if the target candidate system comprises assignable non-critical resource capacity that satisfies the non-critical resource demand load size for the particular task, assigning the particular task to the target candidate system.

18 . The non-transitory computer-readable medium or media of claim 17 , wherein the total resource demand load size, the respective critical resource demand load size, the non-critical resource demand load size, the total assignable resource capacity, the assignable critical resource capacity, and the assignable non-critical resource capacity are represented in a normalized resource vector framework in which resource demand load and assignable resource capacity are normalized based on reference values for different resource types.

19 . The non-transitory computer-readable medium or media of claim 17 , wherein the respective set of candidate systems is determined by:

defining an initial set of candidate systems using a set of constraints associated with the particular task; and

defining the respective set of candidate systems based on an assessment of resource utilization trajectory for one or more candidate systems of the initial set of candidate systems.

20 . The non-transitory computer-readable medium or media of claim 17 , wherein the respective ordered list of tasks is sorted in descending order based on the total resource demand load size for each task of the respective set of tasks, and wherein the respective ordered list of candidate systems is sorted in ascending order based on the total assignable resource capacity for each candidate system of the respective set of candidate systems.

Assignments (1)

Continuity (3)

References Cited (62)