Data driven converged infrastructure components evaluation

• US 10,102,055 B1
• Filed: 03/22/2016
• Issued: 10/16/2018
• Est. Priority Date: 03/22/2016
• Status: Active Grant

• Current Assignee: Emc IP Holding Company LLC (Dell Technologies Inc.)
• Original Assignee: Emc IP Holding Company LLC (Dell Technologies Inc.)
• Inventors: Gaber, Shiri, Ben-Harush, Oshry, Grubshtein, Alon J., Tenenboim-Chekina, Lena, Cohen, Raphael
• Primary Examiner(s): Lottich, Joshua P
• Application Number:
  US15/077,372
• Time in Patent Office: 938 Days
• CPC Class Codes: G06F 11/0712 in a virtual computing platform, e.g. logically partitioned systems
• CPC Class Codes: G06F 11/0754 by exceeding limits
• CPC Class Codes: G06F 11/0784 Routing of error reports, e.g. with a specific transmission path or data flow
• CPC Class Codes: G06F 11/079 Root cause analysis, i.e. error or fault diagnosis in a hardware test environment G06F11/22; in a software test environment G06F11/36

Abstract

An apparatus comprises a processing platform configured to implement an analytic engine for evaluation of at least one of a converged infrastructure environment and one or more components of the converged infrastructure environment. The analytic engine comprises an extraction module configured to extract one or more features corresponding to the converged infrastructure environment, a learning and modeling module configured to predict an expected quantitative performance value of at least one of the converged infrastructure environment and the one or more components of the converged infrastructure environment based on the extracted one or more features, and comparison and ranking modules. The comparison module is configured to calculate a difference between an actual quantitative performance value of at least one of the converged infrastructure environment and the one or more components of the converged infrastructure environment and the expected quantitative performance value. The ranking module determines anomalies based on the difference.

Claims

1. An apparatus comprising:

a processing platform configured to implement an analytic engine for evaluation of at least one of a plurality of converged infrastructure environments and one or more components of the plurality of converged infrastructure environments;

wherein the one or more components of the plurality of converged infrastructure environments comprise at least a virtualization component implemented using a virtualization platform comprising at least one of one or more virtual machines and one or more containers;

wherein the plurality of converged infrastructure environments each implement different versions of the virtualization component;

wherein the plurality of converged infrastructure environments are connected to the analytic engine over at least one network;

wherein the analytic engine comprises;

an extraction module configured to extract one or more features corresponding to the plurality of converged infrastructure environments, wherein the extracted one or more features comprise at least the different versions of the virtualization component;

a learning and modeling module configured to predict an expected quantitative performance value of at least one of the plurality of converged infrastructure environments and the one or more components of the converged infrastructure environment based on the extracted one or more features; and

a comparison module configured to calculate a difference between an actual quantitative performance value of at least one of the plurality of converged infrastructure environments and the one or more components of the plurality of converged infrastructure environments and the expected quantitative performance value; and

a ranking module configured to determine whether at least one of the plurality of converged infrastructure environments and the one or more components is anomalous based on whether the difference between the actual quantitative performance value and the expected quantitative performance value exceeds a statistically derived threshold;

wherein the ranking module is further configured to determine whether a version of the virtualization component is anomalous using statistically derived values computed based on the plurality of converged infrastructure environments implementing the different versions of the virtualization component; and

an output module connected to the analytic engine over at least one network wherein the output module is configured to;

report the determined anomalies to at least one user via a visualization interface so as to permit the at least one user to take an action with respect to the determined anomalies; and

generate in the visualization interface a plot view visualization that displays actual quantitative performance values as a function of at least the different versions of the virtualization component of the plurality of converged infrastructure environments;

wherein the processing platform comprises one or more processing devices each comprising a processor coupled to a memory.

2. The apparatus of claim 1 wherein the ranking module is further configured to rank at least one of the plurality of converged infrastructure environments and the one or more components based on the difference between the actual quantitative performance value and the expected quantitative performance value.

3. The apparatus of claim 1 wherein the predicting is performed using a multivariate predictive model in which the extracted one or more features are input variables.

4. The apparatus of claim 3 wherein the learning and modeling module is further configured to implement machine learning processing to learn one or more normalization coefficients to be applied to the extracted one or more features when using the multivariate predictive model to predict the expected quantitative performance value of at least one of the plurality of converged infrastructure environments and the one or more components.

5. The apparatus of claim 4 wherein the learning and modeling module is further configured to apply a learned normalization coefficient to a corresponding one of the extracted one or more features to establish baseline expected quantitative performance values for at least one of the plurality of converged infrastructure environments and the one or more components.

6. The apparatus of claim 3 wherein the multivariate predictive model is a regression model.

7. The apparatus of claim 1 wherein the extracted one or more features further comprise at least one of product family, number of drives, capacity, number of power supplies and a time period a product has been in service.

8. The apparatus of claim 1 wherein the expected and actual quantitative performance values comprise one of support time periods, maintenance time periods, monetary costs, and operational time periods.

9. The apparatus of claim 1 wherein the plot view visualization further displays baseline expected quantitative performance values for at least one of the plurality of converged infrastructure environments and the one or more components.

10. The apparatus of claim 1 wherein the one or more components of the converged infrastructure environment further comprise at least one of a management component, a compute component, a storage component and a network component.

11. An information processing system comprising the apparatus of claim 1.

12. The apparatus of claim 1 wherein:

the plurality of converged infrastructure environments each implement different versions of the virtualization platform; and

the extracted one or more features further comprise the different versions of the virtualization platform.

13. A method comprising:

extracting one or more features corresponding to a plurality of converged infrastructure environments;

predicting an expected quantitative performance value of at least one of the plurality of converged infrastructure environments and one or more components of the plurality of converged infrastructure environments based on the extracted one or more features;

calculating a difference between an actual quantitative performance value of at least one of the plurality of converged infrastructure environments and the one or more components of the plurality of converged infrastructure environments and the expected quantitative performance value; and

determining whether at least one of the plurality of converged infrastructure environments and the one or more components is anomalous based on whether the difference between the actual quantitative performance value and the expected quantitative performance value exceeds a statistically derived threshold;

wherein the one or more components of the plurality of converged infrastructure environments comprise at least a virtualization component implemented using a virtualization platform comprising at least one of one or more virtual machines and one or more containers;

wherein the plurality of converged infrastructure environments each implement different versions of the virtualization component;

wherein the extracted one or more features comprise at least the different versions of the virtualization component;

wherein the method further comprises;

determining whether a version of the virtualization component is anomalous using statistically derived values computed based on the plurality of converged infrastructure environments implementing the different versions of the virtualization component;

reporting the determined anomalies to at least one user via a visualization interface so as to permit the at least one user to take an action with respect to the determined anomalies; and

generating in the visualization interface a plot view visualization that displays actual quantitative performance values as a function of at least the different versions of the virtualization component of the plurality of converged infrastructure environments; and

wherein the steps of the method are implemented using at least one processing device comprising a processor coupled to a memory.

14. The method of claim 13 wherein the predicting is performed using a multivariate predictive model in which the extracted one or more features are input variables.

15. The method of claim 14 further comprising implementing machine learning processing to learn one or more normalization coefficients to be applied to the extracted one or more features when using the multivariate predictive model to predict the expected quantitative performance value of at least one of the plurality of converged infrastructure environments and the one or more components.

16. The method of claim 15 wherein a learned normalization coefficient is applied to a corresponding one of the extracted one or more features to establish baseline expected quantitative performance values for at least one of the plurality of converged infrastructure environments and the one or more components.

17. The method of claim 13 wherein:

the plurality of converged infrastructure environments each implement different versions of the virtualization platform; and

the extracted one or more features further comprise the different versions of the virtualization platform.

18. A computer program product comprising a non-transitory processor-readable storage medium having stored therein program code of one or more software programs, wherein the program code when executed by at least one processing device causes said processing device:

to extract one or more features corresponding to a plurality of converged infrastructure environments;

to predict an expected quantitative performance value of at least one of the plurality of converged infrastructure environments and one or more components of the plurality of converged infrastructure environments based on the extracted one or more features;

to calculate a difference between an actual quantitative performance value of at least one of the plurality of converged infrastructure environments and the one or more components of the plurality of converged infrastructure environments and the expected quantitative performance value; and

to determine whether at least one of the plurality of converged infrastructure environments and the one or more components is anomalous based on whether the difference between the actual quantitative performance value and the expected quantitative performance value exceeds a statistically derived threshold;

wherein the one or more components of the plurality of converged infrastructure environments comprise at least a virtualization component implemented using a virtualization platform comprising at least one of one or more virtual machines and one or more containers;

wherein the plurality of converged infrastructure environments each implement different versions of the virtualization component;

wherein the extracted one or more features comprise at least the different versions of the virtualization component;

wherein the program code when executed by at least one processing device further causes said processing device to;

determine whether a version of the virtualization component is anomalous using statistically derived values computed based on the plurality of converged infrastructure environments implementing the different versions of the virtualization component;

report the determined anomalies to at least one user via a visualization interface so as to permit the at least one user to take an action with respect to the determined anomalies; and

generate in the visualization interface a plot view visualization that displays actual quantitative performance values as a function of at least the different versions of the virtualization component of the plurality of converged infrastructure environments.

19. The computer program product of claim 18 wherein the predicting is performed using a multivariate predictive model in which the extracted one or more features are input variables.

20. The computer program product of claim 18 wherein:

the plurality of converged infrastructure environments each implement different versions of the virtualization platform; and

the extracted one or more features further comprise the different versions of the virtualization platform.