Auto-scaling cloud-based computing clusters dynamically using multiple scaling decision makers
Systems, methods, and non-transitory computer-readable media are disclosed for determining an accurate and efficient auto-scaling action for a cloud-based computing cluster based on multiple proposed auto-scaling actions from multiple scaling models. For example, the disclosed systems can determine an auto-scaling action to perform on a cloud-based computing cluster by weighing multiple proposed auto-scaling actions from multiple scaling models based on confidence scores associated with the proposed auto-scaling actions. Moreover, the disclosed systems can modify the cloud-based computing cluster using the determined auto-scaling action (e.g., to accurately and efficiently provision computing resources for a cloud-based computing system).
1. A non-transitory computer-readable medium storing instructions that, when executed by at least one processor, cause a computer device to:
determine a first proposed auto-scaling action using a first scaling model based on information associated with one or more cloud-based computing clusters;
determine a first confidence score that corresponds to an accuracy of the first scaling model generating auto-scaling actions in relation to the information associated with the one or more cloud-based computing clusters;
determine a second proposed auto-scaling action using a second scaling model based on the information associated with the one or more cloud-based computing clusters;
determine a second confidence score that corresponds to an accuracy of the second scaling model generating auto-scaling actions in relation to the information associated with the one or more cloud-based computing clusters;
determine an auto-scaling action to perform based on the first proposed auto-scaling action and the second proposed auto-scaling action using the first confidence score and the second confidence score; and
modify the one or more cloud-based computing clusters based on the determined auto-scaling action.
2. The non-transitory computer-readable medium of claim 1 , further comprising instructions, that when executed by the at least one processor, cause the computer device to determine the auto-scaling action to perform by selecting between the first proposed auto-scaling action and the second proposed auto-scaling action based on a comparison of the first confidence score and the second confidence score.
3. The non-transitory computer-readable medium of claim 1 , further comprising instructions that, when executed by the at least one processor, cause the computer device to:
determine a third proposed auto-scaling action using a third scaling model; and
determine the auto-scaling action to perform by selecting the third proposed auto-scaling action based on determining that the first confidence score and the second confidence score do not satisfy a threshold confidence score.
4. The non-transitory computer-readable medium of claim 1 , wherein the information associated with the one or more cloud-based computing clusters comprises at least one of CPU utilization information, memory usage information, or network usage information.
5. The non-transitory computer-readable medium of claim 1 , further comprising instructions, that when executed by the at least one processor, cause the computer device to modify the one or more cloud-based computing clusters by modifying a computing resource allocation associated with the one or more cloud-based computing clusters based on the determined auto-scaling action.
6. The non-transitory computer-readable medium of claim 1 , further comprising instructions, that when executed by the at least one processor, cause the computer device to:
determine the first proposed auto-scaling action using one of a reactive auto-scaling model, a forecasting auto-scaling model, or a reinforcement auto-scaling model; and
determine the second proposed auto-scaling action using another of the reactive auto-scaling model, the forecasting auto-scaling model, or the reinforcement auto-scaling model.
7. The non-transitory computer-readable medium of claim 1 , wherein the one or more cloud-based computing clusters comprise a first set of computing resources from a first cloud-based service provider and a second set of computing resources from a second cloud-based service provider and further comprising instructions, that when executed by the at least one processor, cause the computer device to modify the one or more cloud-based computing clusters based on the determined auto-scaling action by allocating a computing task from the first set of computing resources to the second set of computing resources.
8. The non-transitory computer-readable medium of claim 1 , wherein the first proposed auto-scaling action comprises instructions to increase or decrease computing resource utilization on the one or more cloud-based computing clusters.
9. The non-transitory computer-readable medium of claim 1 , further comprising instructions, that when executed by the at least one processor, cause the computer device to:
determine the first proposed auto-scaling action by receiving the first proposed auto-scaling action from a first cloud-based service provider; and
determine the second proposed auto-scaling action by receiving the second proposed auto-scaling action from a second cloud-based service provider.
10. A system comprising:
one or more memory devices comprising:
a forecasting auto-scaling model;
a reinforcement auto-scaling model;
a reactive auto-scaling model; and
information associated with one or more cloud-based computing clusters; and
one or more server devices that cause the system to:
determine a first proposed auto-scaling action and a first confidence score associated with the first proposed auto-scaling action using the forecasting auto-scaling model based on the information associated with the one or more cloud-based computing clusters, the first confidence score corresponding to an accuracy of the forecasting auto-scaling model;
determine a second proposed auto-scaling action and a second confidence score associated with the second proposed auto-scaling action using the reinforcement auto-scaling model based on the information associated with the one or more cloud-based computing clusters, the second confidence score corresponding to an accuracy of the reinforcement auto-scaling model;
determine a third proposed auto-scaling action using the reactive auto-scaling model based on the information associated with the one or more cloud-based computing clusters;
determine an auto-scaling action to perform by selecting between the first proposed auto-scaling action, the second proposed auto-scaling action, and the third proposed auto-scaling action based on the first confidence score and the second confidence score; and
modify the one or more cloud-based computing clusters by modifying a computing resource allocation associated with the one or more cloud-based computing clusters based on the determined auto-scaling action.
11. The system of claim 10 , wherein:
the one or more cloud-based computing clusters comprise a first set of computing resources from a first cloud-based service provider and a second set of computing resources from a second cloud-based service provider and
the one or more server devices cause the system to modify the one or more cloud-based computing clusters by allocating a computing task from the first set of computing resources to the second set of computing resources.
12. The system of claim 10 , wherein the one or more server devices cause the system to determine the auto-scaling action to perform by selecting the third proposed auto-scaling action based on determining that the first confidence score and the second confidence score do not satisfy a threshold confidence score.
13. The system of claim 10 , wherein the information associated with the one or more cloud-based computing clusters comprises at least one of CPU utilization information, memory usage information, or network usage information.
14. The system of claim 10 , wherein:
the first proposed auto-scaling action comprises instructions to increase computing resource utilization on the one or more cloud-based computing clusters; and
the second proposed auto-scaling action comprises instructions to decrease computing resource utilization on the one or more cloud-based computing clusters.
15. The system of claim 14 , wherein the one or more server devices cause the system to determine the auto-scaling action to perform by selecting the second proposed auto-scaling action based on a comparison of the first confidence score and the second confidence score.
16. The system of claim 10 , wherein the one or more server devices cause the system to:
determine the first proposed auto-scaling action by receiving the first proposed auto-scaling action from a first cloud-based service provider; and
determine the second proposed auto-scaling action by receiving the second proposed auto-scaling action from a second cloud-based service provider.
17. A computer-implemented method for dynamically scaling cloud-based computing resources, the computer-implemented method comprising:
determining a first proposed auto-scaling action using a first scaling model;
determining a second proposed auto-scaling action using a second scaling model;
performing a step for determining an auto-scaling action to perform based on the first proposed auto-scaling action and the second proposed auto-scaling action; and
modifying one or more cloud-based computing clusters based on the determined auto-scaling action.
18. The computer-implemented method of claim 17 , wherein modifying the one or more cloud-based computing clusters comprises allocating computing resources between multiple cloud-based service providers.
19. The computer-implemented method of claim 17 , wherein modifying the one or more cloud-based computing clusters comprises modifying a computing resource allocation associated with the one or more cloud-based computing clusters based on the determined auto-scaling action.
20. The computer-implemented method of claim 17 , wherein:
determining the first proposed auto-scaling action comprises receiving the first proposed auto-scaling action from a first cloud-based service provider; and
determining the second proposed auto-scaling action comprises receiving the second proposed auto-scaling action from a second cloud-based service provider.