Methods and Systems of Dynamic Management of Resources in a Virtualized Environment

摘要

The methods and systems described herein centralize simulation resources and effectively delivering training and simulation services to a broad set of distributed users at both the enterprise and operational levels. The cloud-based delivery of simulation applications described herein enables on-demand network access to a shared pool of configurable computing resources (e.g., networks, servers, storage, applications, and services) that can be rapidly provisioned and released with minimal management effort or service provider interaction. In exemplary systems, users may provision computing capabilities, such as server time and network storage, as needed, automatically without requiring human interaction.

主权项

1. A process for dynamically managing a pool of virtual and physical resources accessible by multiple tenants running multiple instances of a distributable simulation application, the method comprising:
monitoring, by at least one processing server in the pool, a processing load of each of the multiple instances of the distributable simulation application running within one or more virtual machines in the pool, the one or more virtual machines each comprising a plurality of nodes; determining, by the at least one processing server, that one or more nodes of the one or more virtual machines is overloaded or underloaded by a status of each of the multiple instance of the distributable simulation application; if overloaded then starting, by the at least one processing server, at least one new virtual machine; and instructing, by the at least one processing server, the one or more multiple instances of the application causing the overload to transfer a portion of its processing load from the one or more overloaded nodes to the at least one new virtual machine; if underloaded then instructing, by the at least one processing server, the one or more multiple instances of the application currently running on one or more underloaded nodes to transfer all of its processing load from the one or more underloaded nodes to one or more alternate nodes; and stopping, by the at least one processing server, the one or more of the underloaded nodes.