System and Method of Strategy Driven Optimization of Computer Resource Configurations in a Cloud Environment

摘要

A computer-implemented method of optimizing a computer resource configuration in a cloud environment to provision an information technology service includes determining performance objectives for a provisioned information technology service. Constraints for the provisioned information technology service based on the determined performance objectives are determined. A desired computer resource configuration to achieve the determined performance objectives for the provisioned information technology service and to meet the determined constraints for the provisioned information technology service is determined. Then a computer resource configuration is compared with the desired resource configuration to determine configuration actions wherein the configuration actions comprise transitioning a computer resource in the computer resource configuration to a new computer resource configuration. Available computer resources in the cloud environment are configured as well as the computer resource configuration in the cloud environment to realize the determined configuration actions. A current performance of the configured available computer resources in the cloud environment is evaluated as well as a current performance of the configured computer resource configuration in the cloud environment compared to the performance objective. Then, the method determines if the desired computer resource configuration has been achieved based on the evaluated current performance of the configured available computer resources in the cloud environment and the evaluated current performance of the configured computer resource configuration in the cloud environment. Then a history is generated of performance using the evaluated current performance of the configured available computer resources in the cloud environment and the evaluated current performance of the configured computer resource configuration in the cloud environment. A new desired computer resource configuration to achieve the determined performance objectives is then determined using the generated history of performance.