Graph embeddings for drift identification

Patent metadata

Field	Value
Publication number	US-12566686-B2
Application number	US-202418764067-A
Country	US
Kind code	B2
Filing date	Jul 3, 2024
Priority date	Jul 3, 2024
Publication date	Mar 3, 2026
Grant date	Mar 3, 2026

Abstract

A system can create a first graph that represents first values of current configurations of a computer hardware as first nodes and first dependencies of the current configurations as first links. The system can create first embedding vectors based on the first graph. The system can input the first embedding vectors to a graph isomorphism network to produce modified first embedding vectors. The system can identify specified configurations for the computer hardware. The system can create a second graph that represents second values of the specified configurations. The system can create second embedding vectors based on the second graph. The system can input the second embedding vectors to the graph isomorphism network to produce modified second embedding vectors. The system can perform a graph isomorphism evaluation on the modified first embedding vectors and the modified second embedding vectors to determine an amount of difference.

First claim

What is claimed is:

1 . A system, comprising:
at least one processor; and at least one memory that stores executable instructions that, when executed by the at least one processor, facilitate performance of operations, comprising:
identifying current configurations of a computer hardware system;
creating a first graph that represents first values of the current configurations as first nodes and first dependencies of the current configurations as first links;
creating first embedding vectors based on the first graph;
processing the first embedding vectors comprising inputting the first embedding vectors to a graph isomorphism network to produce modified first embedding vectors;
identifying specified configurations for the computer hardware system;
creating a second graph that represents second values of the specified configurations as second nodes and second dependencies of the current configurations as second links;
creating second embedding vectors based on the second graph;
processing the second embedding vectors comprising inputting the second embedding vectors to the graph isomorphism network to produce modified second embedding vectors;
performing a graph isomorphism evaluation on the modified first embedding vectors and the modified second embedding vectors to determine an amount of difference between the current configurations and the specified configurations; and
in response to the amount of difference being determined to satisfy a similarity criterion, storing an indication that the current configurations satisfy a configuration criterion.

2 . The system of claim 1 , wherein the computer hardware system comprises a bare metal as a service system, wherein a first entity associated with the system performs management of physical information technology infrastructure of the computer hardware system, and wherein a second entity is a user of the computer hardware system and performs active management of the computer hardware system.

3 . The system of claim 1 , wherein the current configurations of the computer hardware system comprise configuration variables for a product of the computer hardware system or a release of the computer hardware system.

4 . The system of claim 3 , wherein the current configurations of the computer hardware system comprise identifications of relationships between the configuration variables.

5 . The system of claim 1 , wherein the specified configurations of the computer hardware system comprise recommended configuration variables for a product of the computer hardware system or a release of the computer hardware system.

6 . The system of claim 5 , wherein the specified configurations of the computer hardware system comprise identifications of relationships between the recommended configuration variables.

7 . The system of claim 5 , wherein a user account that is configured to utilize resources of the computer hardware system is configured to edit the recommended configuration variables to form recommended values to user-specified values.

8 . A method, comprising:
creating, by a system comprising at least one processor, a first graph that represents a current configuration of computer hardware; creating, by the system, first embedding vectors based on the first graph; inputting, by the system, the first embedding vectors to a graph isomorphism network process to produce modified first embedding vectors; creating, by the system, a second graph that represents a potential configuration of the computer hardware; creating, by the system, second embedding vectors based on the second graph; inputting, by the system, the second embedding vectors with the graph isomorphism network process to produce modified second embedding vectors; performing, by the system, a graph isomorphism evaluation on the modified first embedding vectors and the modified second embedding vectors to determine an amount of difference between the current configuration and the potential configuration; and based on the amount of difference being determined to satisfy a similarity criterion, determining, by the system, that the current configuration satisfies at least a defined level of configuration.

9 . The method of claim 8 , wherein at least part of the amount of difference between the current configuration and the potential configuration is based on a product or a release that has been updated relative to a first version of the product or a second version of the release identified by the potential configuration.

10 . The method of claim 8 , wherein a first number of nodes of the first graph differs from a second number of nodes of the second graph.

11 . The method of claim 8 , wherein a first ordering of nodes of the first graph differs from a second ordering of nodes of the second graph.

12 . The method of claim 8 , wherein the first graph comprises variables of the current configuration that are modeled as a configuration hierarchy and generalized in graph form.

13 . The method of claim 8 , wherein the amount of difference between the current configuration and the potential configuration is identified as no difference, and wherein the first graph comprises a subgraph of the second graph.

14 . The method of claim 8 , wherein the graph isomorphism evaluation comprises an evaluation based on a Weisfeiler Leman graph isomorphism test.

15 . A non-transitory computer-readable medium comprising instructions that, in response to execution, cause a system comprising at least one processor to perform operations, comprising:
creating first embedding vectors based on a first graph that represents a current configuration of computer hardware; processing the first embedding vectors according to a graph isomorphism network to produce modified first embedding vectors; creating second embedding vectors based on a second graph that represents a different configuration of the computer hardware; processing the second embedding vectors according to the graph isomorphism network to produce modified second embedding vectors; applying a graph isomorphism evaluation to the modified first embedding vectors and the modified second embedding vectors comprising determining an amount of difference between the current configuration and the different configuration; and based on the amount of difference being determined to satisfy a similarity criterion, determining that the current configuration satisfies a configuration condition.

16 . The non-transitory computer-readable medium of claim 15 , wherein processing the first embedding vectors or the second embedding vectors according to the graph isomorphism network comprises aggregating respective embeddings that correspond to respective neighbor nodes of the first graph or the second graph, respectively.

17 . The non-transitory computer-readable medium of claim 15 , wherein processing the first embedding vectors or the second embedding vectors according to the graph isomorphism network technique utilizing respective neural networks that correspond to respective nodes of the first graph or the second graph, respectively.

18 . The non-transitory computer-readable medium of claim 15 , wherein processing the first embedding vectors or the second embedding vectors according to the graph isomorphism network comprises modeling an injective hash function using a neural network to define colors for the first graph or the second graph, respectively, and wherein applying the graph isomorphism evaluation comprises applying the graph isomorphism evaluation based on the colors.

Assignees

• Dell Products Lp

Inventors

• Sawal Vinay
• Liu Jason
• Atkinson Chris
• Frey Kirk
• Jreij Elie Antoun
• Savir Amihai
• Hutcheson Jacob R

Classifications

• G06F16/9024 Primary

  Graphs; Linked lists (G06F16/9027 takes precedence) · CPC title

• G06F11/3051 Primary

  Monitoring arrangements for monitoring the configuration of the computing system or of the computing system component, e.g. monitoring the presence of processing resources, peripherals, I/O links, software programs (verification or detection of system hardware configuration G06F11/2247) · CPC title