Techniques for workload toxic mapping
View Patent ↗Techniques for toxic workload mapping are provided. A state of a target workload is recorded along with a configuration and state of an environment that is processing the workload. Micro valuations are taken, via statistical sampling, for metrics associated with the workload and for different combinations of resources within the environment. The sampling taken at micro second intervals. The valuations are aggregated to form an index representing a toxic mapping for the workload within the environment. The toxic mapping is mined, in view of policy, to provide conditions and scenarios that may be deemed problematic within the workload and/or environment.
1. A method implemented in a non-transitory machine-readable storage medium and processed by one or more processors configured to perform the method, comprising:
acquiring a workload state for an executing workload and a processing environment state for a processing environment that the workload is executing within;
taking micro measurements at micro second intervals for the workload, the processing environment, resources within the processing environment, and combinations of: the workload, the processing environment, and the resources, and wherein taking further includes determining a status for the processing environment using an Application Programming Interface (API) processing command and wherein taking further includes collecting at least some micro measurements as a particular time that a request to an external service takes to be answered;
mapping the micro measurements in a toxic map, wherein the toxic map is represented as one of: a table, a database, and a user-defined data structure associated with the workload, the workload state, the processing environment, the processing environment state, the resources, and the combinations;
the toxic map is mined, in view of policy and based on decision rules, to provide conditions and scenarios deemed problematic within one or more of: the workload and the processing environment.
2. The method of claim 1 , wherein acquiring further includes obtaining a unique resource state for each of the resources.
3. The method of claim 2 , wherein taking further includes generating some of the micro measurements as measurements for each of the unique resource states.
4. The method of claim 3 , wherein mapping further includes mapping within the toxic map the measurements for each of the unique resource states.
5. The method of claim 1 , wherein taking further includes performing a statistical sampling operation to identify the combinations.
6. The method of claim 1 , wherein taking further includes generating a quantifier value for each of: the workload state, the workload, the processing environment state, the processing environment, each of the resources, and each of the combinations, wherein each quantifier value is a summation of particular micro measurements taken during a configured interval of time.
7. The method of claim 6 , wherein mapping further includes using each quantifier value as a particular micro measurement that is mapped within the toxic map.
8. The method of claim 6 , wherein taking further includes iterating the taking of the micro measurements at micro second intervals for a configured interval of time.
9. The method of claim 1 , wherein taking further includes obtaining the micro measurements based on dynamic evaluation of rules and/or policies.
10. The method of claim 9 , wherein obtaining further includes acquiring the rules and/or policies from a principal via the Application Programming Interface (API).
11. The method of claim 1 further comprising, evaluating the toxic map based on decision rules to automatically draw one or more conclusion and each conclusion triggering one or more actions that are automatically processed.
12. A method implemented in a non-transitory machine-readable storage medium and processed by one or more processors configured to perform the method, comprising:
accessing a toxic map represented as one of: a table, a database, and a user-defined data structure having micro valuations taken at micro second intervals for a workload, a workload state for the workload, a processing environment, a processing environment state for the processing environment, resources used on the processing environment, and combinations of: the workload, the workload state, the processing environment, the processing environment state, and the resources;
evaluating the micro valuations and relationships to the micro valuations as defined in the toxic map to produce results representing decisions, wherein an Application Programming Interface (API) processing command is used to determine a status for the processing environment and wherein evaluating further includes collecting at least some micro measurements as a particular time that a request to an external service takes to be answered;
the toxic map is mined, in view of policy and based on decision rules, to provide conditions and scenarios deemed problematic within one or more of: the workload and the processing environment;
packaging the results for delivery to a target principal.
13. The method of claim 12 , wherein evaluating further includes modifying the micro valuations and the relationships.
14. The method of claim 13 , wherein modifying further includes augmenting the micro valuations by producing weighted micro valuations based on evaluation of rules and/or policies.
15. The method of claim 14 , wherein augmenting further includes producing the results by processing a Bayesian algorithm, processing a neural network algorithm, and/or processing a fuzzy logic algorithm.
16. The method of claim 12 , wherein packaging further includes formatting the results into a target format that is automatically processed by the target principal to take one or more automated actions based on the decisions represented by the results, the target principal an automated service.
17. The method of claim 12 further comprising, producing a baseline for the results based on evaluation of the micro valuations.
18. The method of claim 17 further comprising, comparing the results to previous baselined results and using policy to take one or more automated actions.
19. A system, comprising:
a processor configured with a toxic mapper represented as one of: a table, a database, and a user-defined data structure that resides and is implemented within a non-transitory computer-readable storage medium and that executes on the processor; and
the processor or another processor of a network configured with a results packager that resides and is implemented within a non-transitory computer-readable storage medium that executes on the processor or the other processor;
wherein the toxic mapper is configured to take micro valuations at micro second intervals of: a workload, a workload state for the workload, a processing environment, a processing environment state for the processing environment, resources within the processing environment, and combinations thereof, and wherein the results packager is configured to modify the micro valuations and derive automated decisions based on evaluations of the modified micro valuations defined in policies, the results packager further configured to package the decisions and deliver the decisions to a target principal for automated action by the target principal, and wherein a determination on a status for the processing environment is made using a processing command from an Application Programming Interface (API) command and wherein taking further includes collecting at least some micro measurements as a particular time that a request to an external service takes to be answered;
the toxic map is mined, in view of policy and based on decision rules, to provide conditions and scenarios deemed problematic within one or more of: the workload and the processing environment.
20. The system of claim 19 , wherein the results packager is further configured to use the policies in view of previous automated decisions associated with previously modified and recorded micro valuations.