US-12625738-B2 - Systems and methods for hypergraph edge resource demand load representation

Abstract

Managing the resource demand load for edge systems is significantly more complex than for other systems, such as cloud environments. A time period in which an application or task is operating based on initial demand resource load values that are provided by a customer may be inaccurate, which may expose sub-standard execution. Embodiments herein seek to significantly mitigate the potential of sub-standard execution. Embodiments collect a repository of resource demand load usage data over a time period that can be used to accurately determine the statistical moments of uncertain resource demand load. In one or more embodiments, a repository of hypervector and/or hyperspace representations may be generated and used to help with resource demand load estimation.

Inventors

• William Jeffery White
• Said Tabet

Assignees

• DELL PRODUCTS L.P.

Dates

Publication Date

20260512

Application Date

20230807

Claims (20)

1. 1 . A processor-implemented method comprising: for a set of instances of a task, collecting resource-related data associated with handling the task; using a dataset comprising the resource-related data associated with handling the task to determine resource statistics for one or more edge resources for the task; determining, for the task, one or more resource demand values for the one or more edge resources using one or more of the resource statistics; forming an edge hypervector representation that is associated with the one or more resource demand values for the one or more edge resources for handling the task; and storing the edge hypervector representation in a hypergraph.
2. 2 . The processor-implemented method of claim 1 wherein the step of using a dataset comprising the resource-related data associated with handling the task to determine resource statistics for one or more edge resources for the task comprises: using the dataset comprising the resource-related data associated with handling the task and a M-PCM-OFFD (Multivariate Probabilistic Collocation Method-Orthogonal Fractional Factorial Design) methodology to obtain the resource statistics for the one or more edge resources for the task.
3. 3 . The processor-implemented method of claim 2 wherein the dataset comprises resource-related data associated with handling the task collected from a plurality of edge systems handling the task over an evaluation time.
4. 4 . The processor-implemented method of claim 1 wherein the edge hypervector representation further comprises a lower control limit for each of the one or more edge resources, one or more input characteristics for the task, and one or more output performance metrics associated with handling the task.
5. 5 . The processor-implemented method of claim 1 further comprising: receiving a request for a new task to be performed; and checking the hypergraph to ascertain whether the hypergraph contains an edge hypervector representation that can be used to obtain one or more resource statistics for the new task for resource demand load estimation.
6. 6 . The processor-implemented method of claim 5 wherein the step of checking the hypergraph to ascertain whether the hypergraph contains an edge hypervector representation that can be used to obtain one or more resource statistics for the new task for resource demand load estimation comprises: responsive to determining that the hypergraph contains an edge hypervector representation corresponding to the new task, using one or more resource statistics from the edge hypervector representation in the hypergraph for the new task for resource demand load estimation.
7. 7 . The processor-implemented method of claim 5 wherein the step of checking the hypergraph to ascertain whether the hypergraph contains an edge hypervector representation that can be used to obtain one or more resource statistics for the new task for resource demand load estimation comprises: responsive to determining that the hypergraph does not contain an edge hypervector representation corresponding to the new task: dispatching the new task to an edge system using initial demand resource load values; collecting resource-related data associated with handling the new task for a time period; using a dataset comprising the resource-related data associated with handling the new task to determine resource statistics for one or more edge resources for the new task; determining, for the new task, one or more resource demand values for the one or more edge resources using one or more of the resource statistics; forming a query edge hypervector representation that is associated with the one or more resource demand values for the one or more edge resources for handling the new task; using the query edge hypervector representation to search the hypergraph to obtain a set of candidate edge hypervector representations; using a divergence measure to evaluate similarity of one or more probability distribution functions of the query edge hypervector representation and each of the candidate edge hypervector representations; and responsive to determining that one of the candidate edge hypervector representations is a sufficient match with the query edge hypervector representation: selecting the candidate hypervector representation as a proxy edge hypervector representation for the new task; and using one or more resource statistics associated with the proxy edge hypervector representation for the new task for resource demand load estimation.
8. 8 . The processor-implemented method of claim 7 wherein the step of responsive to determining that one of the candidate edge hypervector representations is a sufficient match with the query edge hypervector representation further comprises: collecting resource-related data associated with handling the new task; using a dataset comprising the resource-related data associated with handling the new task to determine resource statistics for one or more edge resources for the new task; determining, for the new task, one or more resource demand values for the one or more edge resources using one or more of the resource statistics; forming an edge hypervector representation that is associated with the one or more resource demand values for the one or more edge resources for handling the new task; and storing the edge hypervector representation for the new task in the hypergraph.
9. 9 . The processor-implemented method of claim 7 wherein the step of checking the hypergraph to ascertain whether the hypergraph contains an edge hypervector representation that can be used to obtain one or more resource statistics for the new task for resource demand load estimation comprises: responsive to determining that none of the candidate edge hypervector representations is a sufficient match with the query edge hypervector representation: adding the query edge hypervector representation to the hypergraph as a temporary edge hypervector representation; collecting resource-related data associated with handling the new task; using a dataset comprising the resource-related data associated with handling the new task to determine resource statistics for one or more edge resources for the new task; determining, for the new task, one or more resource demand values for the one or more edge resources using one or more of the resource statistics; forming an edge hypervector representation that is associated with the one or more resource demand values for the one or more edge resources for handling the new task; and storing the edge hypervector representation for the new task in the hypergraph, which replaces the temporary edge hypervector representation.
10. 10 . A processor-implemented method comprising: receiving a request for a task to be performed; responsive to determining that a hypergraph contains an edge hypervector representation for the task, using one or more resource demand values associated with the edge hypervector representation for the task for resource demand load estimation for dispatching the task to an edge system; and responsive to determining that the hypergraph does not contain an edge hypervector representation for the task: dispatching the task to an edge system using initial demand resource load values; collecting resource-related data associated with handling the task for a time period; using a dataset comprising the resource-related data associated with handling the task to determine resource statistics for one or more edge resources for the task; determining, for the task, one or more resource demand values for the one or more edge resources using one or more of the resource statistics; forming a query edge hypervector representation that is associated with the one or more resource demand values for the one or more edge resources for handling the task; using the query edge hypervector representation to search the hypergraph to obtain a set of candidate edge hypervector representations; and using a divergence measure to evaluate similarity of one or more probability distribution functions of the query edge hypervector representation and each of the candidate edge hypervector representations to determine whether one of the candidate edge hypervector representations is a sufficient match with the query edge hypervector representation to act as a proxy edge hypervector representation for the task.
11. 11 . The processor-implemented method of claim 10 further comprising: responsive to determining that one of the candidate edge hypervector representations is a sufficient match with the query edge hypervector representation: selecting the candidate hypervector representation as the proxy edge hypervector representation for the task; and using one or more resource demand values associated with the proxy edge hypervector representation for the task for resource demand load estimation.
12. 12 . The processor-implemented method of claim 11 wherein the step of responsive to determining that one of the candidate edge hypervector representations is a sufficient match with the query edge hypervector representation further comprises: collecting resource-related data associated with handling the task; using a dataset comprising the resource-related data associated with handling the task to determine resource statistics for one or more edge resources for the task; determining, for the task, one or more resource demand values for the one or more edge resources using one or more of the resource statistics; forming an edge hypervector representation that is associated with the one or more resource demand values for the one or more edge resources for handling the task; and storing the edge hypervector representation for the task in the hypergraph.
13. 13 . The processor-implemented method of claim 10 further comprising: responsive to determining that none of the candidate edge hypervector representations is a sufficient match with the query edge hypervector representation: adding the query edge hypervector representation to the hypergraph as a temporary edge hypervector representation; collecting resource-related data associated with handling a new task; using a dataset comprising the resource-related data associated with handling the new task to determine resource statistics for one or more edge resources for the new task; determining, for the new task, one or more resource demand values for the one or more edge resources using one or more of the resource statistics; forming an edge hypervector representation that is associated with the one or more resource demand values for the one or more edge resources for handling the new task; and storing the edge hypervector representation for the new task in the hypergraph, which replaces the temporary edge hypervector representation.
14. 14 . The processor-implemented method of claim 10 wherein the step of using a dataset comprising the resource-related data associated with handling the task to determine resource statistics for one or more edge resources for the task comprises: using the dataset comprising the resource-related data associated with handling the task and a M-PCM-OFFD (Multivariate Probabilistic Collocation Method-Orthogonal Fractional Factorial Design) methodology to obtain the resource statistics for the one or more edge resources for the task.
15. 15 . 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, causes steps to be performed comprising: receiving a request for a task to be performed; responsive to determining that a hypergraph contains an edge hypervector representation for the task, using one or more resource demand values associated with the edge hypervector representation for resource demand load estimation for dispatching the task to an edge system; and responsive to determining that the hypergraph does not contain an edge hypervector representation for the task: dispatching the task to an edge system using initial demand resource load values; collecting resource-related data associated with handling the task for a time period; using a dataset comprising the resource-related data associated with handling the task to determine resource statistics for one or more edge resources for the task; determining, for the task, one or more resource demand values for the one or more edge resources using one or more of the resource statistics; forming a query edge hypervector representation that is associated with the one or more resource demand values for the one or more edge resources for handling the task; using the query edge hypervector representation to search the hypergraph to obtain a set of candidate edge hypervector representations; and using a divergence measure to evaluate similarity of one or more probability distribution functions of the query edge hypervector representation and each of the candidate edge hypervector representations to determine whether one of the candidate edge hypervector representations is a sufficient match with the query edge hypervector representation to act as a proxy edge hypervector representation for the task.
16. 16 . The one or more information handling systems of claim 15 wherein the one or more non-transitory computer-readable medium or media further comprise one or more sets of instructions which, when executed by at least one of the one or more processors, causes steps to be performed comprising: responsive to determining that one of the candidate edge hypervector representations is a sufficient match with the query edge hypervector representation: selecting the candidate hypervector representation as the proxy edge hypervector representation for the task; and using one or more resource demand values associated with the proxy edge hypervector representation for resource demand load estimation for the task.
17. 17 . The one or more information handling systems of claim 16 wherein the step of responsive to determining that one of the candidate edge hypervector representations is a sufficient match with the query edge hypervector representation further comprises: collecting resource-related data associated with handling the task; using a dataset comprising the resource-related data associated with handling the task to determine resource statistics for one or more edge resources for the task; determining, for the task, one or more resource demand values for the one or more edge resources using one or more of the resource statistics; forming an edge hypervector representation that is associated with the one or more resource demand values for the one or more edge resources for handling the task; and storing the edge hypervector representation for the task in the hypergraph.
18. 18 . The one or more information handling systems of claim 15 wherein the one or more non-transitory computer-readable medium or media further comprise one or more sets of instructions which, when executed by at least one of the one or more processors, causes steps to be performed comprising: responsive to determining that none of the candidate edge hypervector representations is a sufficient match with the query edge hypervector representation: adding the query edge hypervector representation to the hypergraph as a temporary edge hypervector representation; collecting resource-related data associated with handling a new task; using a dataset comprising the resource-related data associated with handling the new task to determine resource statistics for one or more edge resources for the new task; determining, for the new task, one or more resource demand values for the one or more edge resources using one or more of the resource statistics; forming an edge hypervector representation comprising the resource statistics for the one or more edge resources for handling the new task; and storing the edge hypervector representation for the new task in the hypergraph, which replaces the temporary edge hypervector representation.
19. 19 . The one or more information handling systems of claim 15 wherein the step of using the dataset comprising the resource-related data associated with handling the task to determine resource statistics for one or more edge resources for the task comprises: using the dataset comprising the resource-related data associated with handling the task and a M-PCM-OFFD (Multivariate Probabilistic Collocation Method-Orthogonal Fractional Factorial Design) methodology to obtain the resource statistics for the one or more edge resources for the task.
20. 20 . The one or more information handling systems of claim 15 wherein the initial demand resource load values are user specified and are obtained from input received in association with the request for the task.

Description

CROSS-REFERENCE TO RELATED APPLICATIONS This patent application is related to and claims priority benefit under 35 USC § 119(e) to commonly-owned U.S. Pat. App. No. 63/450,237, filed on 6 Mar. 2023, entitled “EDGE RESOURCE UTILIZATION,” and listing William Jeffery White and Said Tabet as inventors, which patent document is incorporated by reference herein in its entirety and for all purposes. This patent application is a continuation-in-part of and claims priority benefit under 35 USC § 120 to co-pending and commonly-owned U.S. patent application Ser. No. 18/355,351, filed on 19 Jul. 2023, entitled “EDGE DOMAIN-SPECIFIC ACCELERATOR VIRTUALIZATION AND SCHEDULING,” and listing William Jeffery White and Said Tabet as inventors, which patent document is incorporated by reference herein in its entirety and for all purposes. This patent application is also related to the following commonly-owned patent documents: U.S. patent application Ser. No. 18/366,461, filed on 7 Aug. 2023, entitled “SYSTEMS AND METHODS FOR EDGE SYSTEM RESOURCE CAPACITY PERFORMANCE PREDICTION,” and listing William Jeffery White and Said Tabet as inventors, which patent document is incorporated by reference herein in its entirety and for all purposes; U.S. patent application Ser. No. 18/366,490, filed on 7 Aug. 2023, entitled “SYSTEMS AND METHODS FOR EDGE SYSTEM RESOURCE CAPACITY DYNAMIC POLICY PLANNING FRAMEWORK,” and listing William Jeffery White and Said Tabet as inventors, which patent document is incorporated by reference herein in its entirety and for all purposes; U.S. patent application Ser. No. 18/366,520, filed on 7 Aug. 2023, entitled “SYSTEMS AND METHODS FOR HYPERGRAPH EDGE RESOURCE DEMAND KNOWLEDGE MANAGEMENT,” and listing William Jeffery White and Said Tabet as inventors, which patent document is incorporated by reference herein in its entirety and for all purposes; U.S. patent application Ser. No. 18/366,538, filed on 7 Aug. 2023, entitled “SYSTEMS AND METHODS FOR CONTINUED EDGE RESOURCE DEMAND LOAD ESTIMATION,” and listing William Jeffery White and Said Tabet as inventors, which patent document is incorporated by reference herein in its entirety and for all purposes; U.S. patent application Ser. No. 18/366,549, filed on 7 Aug. 2023, entitled “SYSTEMS AND METHODS FOR EDGE RESOURCE DEMAND LOAD ESTIMATION,” and listing William Jeffery White and Said Tabet as inventors, which patent document is incorporated by reference herein in its entirety and for all purposes; and U.S. patent application Ser. No. 18/366,555, filed on 7 Aug. 2023, entitled “SYSTEMS AND METHODS FOR EDGE RESOURCE DEMAND LOAD SCHEDULING,” and listing William Jeffery White and Said Tabet as inventors, which patent document is incorporated by reference herein in its entirety and for all purposes. BACKGROUND A. Technical Field The present disclosure relates generally to information handling systems. More particularly, the present disclosure relates to edge platforms. B. Background The subject matter discussed in the background section shall not be assumed to be prior art merely as a result of its mention in this background section. Similarly, a problem mentioned in the background section or associated with the subject matter of the background section should not be assumed to have been previously recognized in the prior art. The subject matter in the background section merely represents different approaches, which in and of themselves may also be inventions. As the value and use of information continues to increase, individuals and businesses seek additional ways to process and store information. One option available to users is information handling systems. An information handling system generally processes, compiles, stores, and/or communicates information or data for business, personal, or other purposes thereby allowing users to take advantage of the value of the information. Because technology and information handling needs and requirements vary between different users or applications, information handling systems may also vary regarding what information is handled, how the information is handled, how much information is processed, stored, or communicated, and how quickly and efficiently the information may be processed, stored, or communicated. The variations in information handling systems allow for information handling systems to be general or configured for a specific user or specific use, such as financial transaction processing, airline reservations, enterprise data storage, or global communications. In addition, information handling systems may include a variety of hardware and software components that may be configured to process, store, and communicate information and may include one or more computer systems, data storage systems, and networking systems. Multi-cloud edge platforms are large-scale distributed systems that enable organizations to manage and optimize their computing resources across multiple cloud environments and edge devices. Typically, t