Automatic configuration of software systems for optimal management and performance using machine learning

Patent metadata

Field	Value
Publication number	US-11200139-B2
Application number	US-202016744523-A
Country	US
Kind code	B2
Filing date	Jan 16, 2020
Priority date	Jan 16, 2020
Publication date	Dec 14, 2021
Grant date	Dec 14, 2021

Abstract

In one embodiment, information (workload, performance, and configuration) is obtained about identified sub-systems (a target component plus other components that influence its performance). The identified sub-systems are clustered into workload clusters and also into performance clusters, where identified sub-systems of particular workload clusters have similar workload measurements, and identified sub-systems of particular performance clusters have similar performance metrics. The techniques herein then determine a given mapped performance cluster for a given workload cluster that corresponds to a best set of performance metrics from among all performance clusters mapped to the given workload cluster. A configuration change recommendation is then generated for a given identified sub-system of the given workload cluster that is not within the given mapped performance cluster corresponding to the best set of performance metrics based on configuration information about each identified sub-system within the given mapped performance cluster that corresponds to the best set of performance metrics.

First claim

What is claimed is:

1. A method, comprising:
obtaining, by a process, information about a plurality of identified sub-systems of a plurality of instances of an end-to-end software solution system, wherein the information comprises workload measurements, performance metrics, and configuration information about each of the plurality of identified sub-systems, and wherein each of the plurality of identified sub-systems comprise a target component plus one or more other components of the system that influence performance of the target component;
clustering, by the process, the plurality of identified sub-systems into a plurality of workload clusters such that each identified sub-system of a particular workload cluster has similar workload measurements;
clustering, by the process, the plurality of identified sub-systems into a plurality of performance clusters such that each identified sub-system of a particular performance cluster has similar performance metrics;
mapping, by the process, a given workload cluster of the plurality of workload clusters to one or more mapped performance clusters of the plurality of performance clusters such that each identified sub-system within the given workload cluster is also within one of the one or more mapped performance clusters;
determining, by the process, a given mapped performance cluster of the one or more mapped performance clusters for the given workload cluster that corresponds to a best set of performance metrics from among the one or more mapped performance clusters for the given workload cluster;
determining, by the process, the configuration information about each identified sub-system within the given mapped performance cluster that corresponds to the best set of performance metrics; and
generating, by the process, a configuration change recommendation for a given identified sub-system of the given workload cluster that is not within the given mapped performance cluster that corresponds to the best set of performance metrics based on the configuration information about each identified sub-system within the given mapped performance cluster that corresponds to the best set of performance metrics.

2. The method as in claim 1 , further comprising:
performing, by the process, the configuration change recommendation to the given identified sub-system.

3. The method as in claim 1 , further comprising:
obtaining information about a newly identified sub-system, wherein the information comprises at least workload measurements and configuration information about the newly identified sub-system;
clustering the newly identified sub-system into the given workload cluster; and
generating a specific configuration change recommendation for the newly identified sub-system based on the configuration information about each identified sub-system within the given mapped performance cluster that corresponds to the best set of performance metrics.

4. The method as in claim 3 , wherein the newly identified sub-system is operating in real-time.

5. The method as in claim 3 , wherein the newly identified sub-system is a simulation.

6. The method as in claim 1 , wherein the configuration change recommendation for the given identified sub-system comprises a configuration change to one or both of the target component and the one or more other components of the system that influence performance of the target component.

7. The method as in claim 1 , further comprising:
identifying at least a portion of the plurality of identified sub-systems through auto-discovery by the process.

8. The method as in claim 1 , further comprising:
identifying at least a portion of the plurality of identified sub-systems through identification received via a user interface.

9. The method as in claim 1 , further comprising:
reducing the information from a larger raw data set based on one or more of filtering, aggregation, and dimensionality reduction of the raw data set.

10. The method as in claim 1 , wherein the performance metrics comprise a feature set of maximum, minimum, average, and standard deviation values derived from a larger raw data set.

11. The method as in claim 1 , wherein the performance metrics comprise one or more of resource utilization, process timing, operational cost, and process throughput of the target component.

12. The method as in claim 1 , wherein the configuration change recommendation for the given identified sub-system comprises a configuration change to a topology of the given identified sub-system.

13. A tangible, non-transitory, computer-readable medium having computer-executable instructions stored thereon that, when executed by a processor on a computer, cause the computer to perform a method comprising:
obtaining information about a plurality of identified sub-systems of a plurality of instances of an end-to-end software solution system, wherein the information comprises workload measurements, performance metrics, and configuration information about each of the plurality of identified sub-systems, and wherein each of the plurality of identified sub-systems comprise a target component plus one or more other components of the system that influence performance of the target component;
clustering the plurality of identified sub-systems into a plurality of workload clusters such that each identified sub-system of a particular workload cluster has similar workload measurements;
clustering the plurality of identified sub-systems into a plurality of performance clusters such that each identified sub-system of a particular performance cluster has similar performance metrics;
mapping a given workload cluster of the plurality of workload clusters to one or more mapped performance clusters of the plurality of performance clusters such that each identified sub-system within the given workload cluster is also within one of the one or more mapped performance clusters;
determining a given mapped performance cluster of the one or more mapped performance clusters for the given workload cluster that corresponds to a best set of performance metrics from among the one or more mapped performance clusters for the given workload cluster;
determining the configuration information about each identified sub-system within the given mapped performance cluster that corresponds to the best set of performance metrics; and
generating a configuration change recommendation for a given identified sub-system of the given workload cluster that is not within the given mapped performance cluster that corresponds to the best set of performance metrics based on the configuration information about each identified sub-system within the given mapped performance cluster that corresponds to the best set of performance metrics.

14. The computer-readable medium as in claim 13 , wherein the method further comprises:
performing the configuration change recommendation to the given identified sub-system.

15. The computer-readable medium as in claim 13 , wherein the method further comprises:
obtaining information about a newly identified sub-system, wherein the information comprises at least workload measurements and configuration information about the newly identified sub-system;
clustering the newly identified sub-system into the given workload cluster; and
generating a specific configuration change recommendation for the newly identified sub-system based on the configuration information about each identified sub-system within the given mapped performance cluster that corresponds to the best set of performance metrics.

16. The computer-readable medium as in claim 13 , wherein the configurat

Assignees

• Cisco Tech Inc

Inventors

• Singh Rishabh
• Mehta Saket
• Singhal Prerana

Classifications

• G06F11/3006

  where the computing system is distributed, e.g. networked systems, clusters, multiprocessor systems (multiprogramming arrangements G06F9/46; allocation of resources G06F9/50) · CPC title

• G06F2201/865

  Monitoring of software · CPC title

• G06F11/3433 Primary

  for load management (allocation of a server based on load conditions G06F9/505; load rebalancing G06F9/5083; redistributing the load in a network by a load balancer H04L67/1029) · CPC title

• G06N5/043

  Distributed expert systems; Blackboards · CPC title

• G06F9/453

  Help systems · CPC title