Abstract: A method by a cloud orchestrator to provide subscription-based throttling of virtual applications in a cloud. The method includes determining that a physical resource in the cloud is being overutilized, identifying a virtual application running in the cloud that utilizes the physical resource, determining that the virtual application is to be throttled based on a determination that the virtual application is subscribed to a throttling subscription registry, where the throttling subscription registry specifies a throttling mechanism by which to throttle the virtual application, and executing the throttling mechanism to throttle the virtual application in response to a determination that the virtual application is to be throttled.

Abstract: A method by a cloud orchestrator to provide subscription-based throttling of virtual applications in a cloud. The method includes determining that a physical resource in the cloud is being overutilized, identifying a virtual application running in the cloud that utilizes the physical resource, determining that the virtual application is to be throttled based on a determination that the virtual application is subscribed to a throttling subscription registry, where the throttling subscription registry specifies a throttling mechanism by which to throttle the virtual application, and executing the throttling mechanism to throttle the virtual application in response to a determination that the virtual application is to be throttled.

Abstract: A method by a cloud orchestrator to provide subscription-based throttling of virtual applications in a cloud. The method includes determining that a physical resource in the cloud is being overutilized, identifying a virtual application running in the cloud that utilizes the physical resource, determining that the virtual application is to be throttled based on a determination that the virtual application is subscribed to a throttling subscription registry, where the throttling subscription registry specifies a throttling mechanism by which to throttle the virtual application, and executing the throttling mechanism to throttle the virtual application in response to a determination that the virtual application is to be throttled.

Abstract: A method by a cloud orchestrator to provide subscription-based throttling of virtual applications in a cloud. The method includes determining that a physical resource in the cloud is being overutilized, identifying a virtual application running in the cloud that utilizes the physical resource, determining that the virtual application is to be throttled based on a determination that the virtual application is subscribed to a throttling subscription registry, where the throttling subscription registry specifies a throttling mechanism by which to throttle the virtual application, and executing the throttling mechanism to throttle the virtual application in response to a determination that the virtual application is to be throttled.

Abstract: According to the teachings herein, provisioning operations carried out via electronic processing in a communication network (60) benefit from the use of reordered workflows (20) having task orderings that are at least partly optimized with respect to task failure probabilities and/or resource-blocking penalties. The reordered workflows (20) are obtained by optimizing predefined provisioning workflows (10). Each predefined workflow (10) comprises two or more tasks (12) ordered along one or more task execution paths (14) in a task tree (16), according to a default task ordering that reflects any required inter-task dependencies but, in general, does not reflect any optimization in terms of failure penalties, resource blocking, etc. Among the several advantages provided by the teachings herein, carrying out provisioning operations in accordance with reordered workflows (20) wastes fewer compute cycles and reduces the needless blocking of network resources in the event of provisioning task failures.

Abstract: According to the teachings herein, provisioning operations carried out via electronic processing in a communication network (60) benefit from the use of reordered workflows (20) having task orderings that are at least partly optimized with respect to task failure probabilities and/or resource-blocking penalties. The reordered workflows (20) are obtained by optimizing predefined provisioning workflows (10). Each predefined workflow (10) comprises two or more tasks (12) ordered along one or more task execution paths (14) in a task tree (16), according to a default task ordering that reflects any required inter-task dependencies but, in general, does not reflect any optimization in terms of failure penalties, resource blocking, etc. Among the several advantages provided by the teachings herein, carrying out provisioning operations in accordance with reordered workflows (20) wastes fewer compute cycles and reduces the needless blocking of network resources in the event of provisioning task failures.