Determining a work distribution model between a client device and a cloud for an application deployed on the cloud

A computer-implemented system comprising:

• one or more processors;
• and
• one or more computer storage media storing computer-usable instructions that, when used by the one or more processors, cause the one or more processors to:
• detect a request from a client device to start a computing application;
• on a virtual instance of a machine, provision an application container for executing the computing application, in response to the request from the client device:
• monitor, including while executing the computing application, network parameters and utilization of a computing resource on the client device, the network parameters including an amount of available communication bandwidth between the client device and the virtual instance of the machine;
• generate upon execution of the computing application a work distribution model for the computing application, the work distribution model indicating a first portion of the computing application to be executed on the client device and a second portion of the computing application to be executed on the virtual instance of the machine, the work distribution model being based on the utilization of the computing resource on the client device, the amount of available communication bandwidth between the client device and the virtual instance of the machine, and a permissible latency associated with the computing application;
• detect a fluctuation in at least one of the monitored network parameters and utilization of the computing resource on the client device while executing the computing application;
• and
• in response to detecting a fluctuation, dynamically adjust the work distribution model for the computing application.

A non-transitory computer storage medium storing instruction data that, when used by one or more computing devices, cause the one or more computing devices to perform operations comprising:

• detecting, at a server computer, a request from a client device to start a computing application;
• provisioning, on a virtual instance of a machine, an application container for executing the computing application, in response to the request from the client device;
• monitoring, including while executing the computing application, network parameters and utilization of a computing resource on the client device, the network parameters including an amount of available communication bandwidth between the client device and the virtual instance of the machine;
• generating, while executing the computing application a work distribution model for the computing application, the work distribution model indicating a first portion of the computing application to be executed on the client device and a second portion of the computing application to be executed on the virtual instance of the machine, the work distribution model being based on the utilization of the computing resource on the client device, the amount of available communication bandwidth between the client device and the virtual instance of the machine, and a permissible latency associated with the computing application;
• detecting a fluctuation in at least one of the monitored network parameters and utilization of the computing resource on the client device while executing the computing application;
• and
• in response to the detected fluctuation, dynamically adjusting the work distribution model for the computing application.