Optimized diagnostics plan for an information handling system
A diagnostics optimization platform employs cloud-based resources, including a diagnostics repository that accumulates health data from managed endpoints, and machine learning (ML) resources that generate endpoint-specific diagnostic plans based on the accumulated health data. The ML resources may be configured to generate diagnostic plans that prioritize any appropriate diagnostic testing parameter or objective including, as a non-limiting example, a reduction in diagnostic testing execution time and/or diagnostic testing frequency. The ML resources may maintain a continually updated training database derived from the collected health data to develop endpoint-specific data collection and diagnostic testing models. The ML resources may include a diagnostics optimization module to develop diagnostic testing models and provide corresponding endpoint-specific diagnostic plans to each endpoint. The ML resources may further include a data collection module to develop data collection models and generate endpoint-specific data collection plans for each of the managed endpoints.
1. A method for managing diagnostic testing of one or more information handling systems, the method comprising:
determining an endpoint-specific data collection plan for each of one or more endpoints;
collecting health data for the one or more endpoints in accordance with the endpoint-specific data collection plan and storing the health data in a diagnostics repository;
determining a health weight factor (HWF) for one or more endpoints, wherein the health weight factor for an endpoint is determined, at least in part, in accordance with usage information for the endpoint wherein the usage information includes anticipated usage information indicative of an anticipated use of the endpoint; and
creating, based on the health data and the health weight factor for each endpoint, an optimized diagnostic plan for each endpoint, wherein the optimized diagnostic plan for each endpoint includes distinct plans for each of two or more operational contexts, the two or more operational contexts including a preboot context and a host operating system context.
2. The method of claim 1 , wherein each data collection plan indicates one or more trigger points and, for each trigger point, one or more data collection parameters.
3. The method of claim 1 , wherein the optimized diagnostic plan for a particular endpoint includes a diagnostic plan component for each of one or more subsystems of the endpoint.
4. The method of claim 1 , wherein the anticipated usage information is indicative of an anticipated use of a particular subsystem during a particular future interval.
5. The method of claim 3 , wherein the optimized diagnostic plan component for a particular subsystem of the particular endpoint indicates an endpoint-specific group of one or more diagnostic algorithms to perform.
6. The method of claim 5 , wherein the group of one or more diagnostic algorithms to perform includes less than all diagnostic algorithms associated with the particular subsystem.
7. The method of claim 5 , wherein the diagnostic plan component for the particular subsystem indicates an endpoint-specific sequence of the group of one or more diagnostic algorithms.
8. The method of claim 1 , wherein the two or more operational contexts further include a service operating system context.
9. The method of claim 1 , wherein creating the optimized diagnostic plan comprises creating the optimized diagnostic plan in accordance with an administrative policy, wherein the administrative policy prioritizes at least one prioritized objective and influences the optimized diagnostic plan.
10. The method of claim 9 , wherein the at least one prioritized objective is selected from a first prioritized objective prioritizing an average duration of the optimized diagnostic plan and a second prioritized objective prioritizing a reduction in a frequency of diagnostic testing.
11. A diagnostic optimization platform for managing diagnostic testing of one or more managed endpoints, wherein each managed endpoint comprises an information handling system, the platform comprising:
a diagnostics repository to receive health data generated by the endpoints; and
machine learning resources, communicatively coupled to the diagnostics repository and the managed endpoints, to perform diagnostic optimization operations, wherein the diagnostic operations include:
determining an endpoint-specific data collection plan for each of one or more endpoints;
collecting health data for the one or more endpoints in accordance with the endpoint-specific data collection plan and storing the health data in the diagnostics repository;
determining health weight factors for each of the one or more endpoints, wherein the health weight factor for an endpoint is determined, at least in part, in accordance with usage information for the endpoint wherein the usage information includes anticipated usage information indicative of an anticipated use of the endpoint; and
creating, based on the health data and the health weight factor for each endpoint, an optimized diagnostic plan for each endpoint, wherein the optimized diagnostic plan specifies an endpoint-specific set of diagnostic testing algorithms, wherein the optimized diagnostic plan for each endpoint includes distinct plans for each of two or more operational contexts, the two or more operational contexts including a preboot context and a host operating system context.
12. The platform of claim 11 , wherein each data collection plan indicates one or more trigger points and, for each trigger point, one or more data collection parameters.
13. The platform of claim 11 , wherein the optimized diagnostic plan for a particular endpoint includes a diagnostic plan component for each of one or more subsystems of the endpoint.
14. The platform of claim 11 , wherein the anticipated usage information is indicative of an anticipated use of a particular subsystem during a particular future interval.
15. The platform of claim 13 , wherein the optimized diagnostic plan component for a particular subsystem of the particular endpoint indicates an endpoint-specific group of one or more diagnostic algorithms to perform.
16. The platform of claim 15 , wherein the group of one or more diagnostic algorithms to perform includes less than all diagnostic algorithms associated with the particular subsystem.
17. The platform of claim 15 , wherein the diagnostic plan component for the particular subsystem indicates an endpoint-specific sequence of the group of one or more diagnostic algorithms.
18. The platform of claim 15 , wherein the two or more operating system contexts further include a service operating system (OS) context.
19. The platform of claim 11 , wherein creating the optimized diagnostic plan comprises creating the optimized diagnostic plan in accordance with an administrative policy, wherein the administrative policy prioritizes at least one prioritized objective and influences the optimized diagnostic plan.
20. The platform of claim 19 , wherein the at least one prioritized objective is selected from a first prioritized objective prioritizing an average duration of the optimized diagnostic plan and a second prioritized objective prioritizing a reduction in a frequency of diagnostic testing.