Abstract: In one embodiment, a computer system stores entries for one or more instances of an application with keys generated for the instances in storage. The instances of the application are instantiated on the computer system. The computer system receives a request from the application with a current key for a current instance and parses the storage to determine if the current key is stored in the keys associated with the application. The computer system returns a response to the application with an indication whether the current key is stored as an entry in the one or more entries. The application uses the response to determine a redirection action to one of the one or more instances of the application when the current key is associated with an instance other than the current instance.

Abstract: In one embodiment, a computer system stores entries for one or more instances of an application with keys generated for the instances in storage. The instances of the application are instantiated on the computer system. The computer system receives a request from the application with a current key for a current instance and parses the storage to determine if the current key is stored in the keys associated with the application. The computer system returns a response to the application with an indication whether the current key is stored as an entry in the one or more entries. The application uses the response to determine a redirection action to one of the one or more instances of the application when the current key is associated with an instance other than the current instance.

Abstract: Resources used by a foreground process associated with an application as well as one or more background tasks associated with the application that are running as separate processes on the computing device are tracked. Historical data relating to resource usage by the application is utilized to predict a resource usage amount for the application which is then stored. A subsequent launch of the application can be conditioned upon whether the predicted resource usage amount is available. If the predicted resource usage amount is available, it can be allocated to the application and the application can be launched. If the application then exceeds the predicted resource usage amount, other processes executing concurrently therewith may be terminated to free up the resource. Regardless of how the application performs, new resource usage data is obtained for the application, which is then used to adjust the predicted resource usage amount.

Abstract: Resources used by a foreground process associated with an application as well as one or more background tasks associated with the application that are running as separate processes on the computing device are tracked. Historical data relating to resource usage by the application is utilized to predict a resource usage amount for the application which is then stored. A subsequent launch of the application can be conditioned upon whether the predicted resource usage amount is available. If the predicted resource usage amount is available, it can be allocated to the application and the application can be launched. If the application then exceeds the predicted resource usage amount, other processes executing concurrently therewith may be terminated to free up the resource. Regardless of how the application performs, new resource usage data is obtained for the application, which is then used to adjust the predicted resource usage amount.

Abstract: Resources used by a foreground process associated with an application as well as one or more background tasks associated with the application that are running as separate processes on the computing device are tracked. Historical data relating to resource usage by the application is utilized to predict a resource usage amount for the application which is then stored. A subsequent launch of the application can be conditioned upon whether the predicted resource usage amount is available. If the predicted resource usage amount is available, it can be allocated to the application and the application can be launched. If the application then exceeds the predicted resource usage amount, other processes executing concurrently therewith may be terminated to free up the resource. Regardless of how the application performs, new resource usage data is obtained for the application, which is then used to adjust the predicted resource usage amount.

Abstract: Techniques are described that enable an application executing on a computing device to dynamically control how an allocated portion of a resource of the computing device is to be divided among multiple processes associated with the application. For example, an application executing on the computing device may interact with an operating system of the computing device to dynamically control how an allocated portion of a resource of the computing device is to be divided among a foreground process associated with the application and one or more background processes associated with the application.

Abstract: Techniques are described that enable an application executing on a computing device to dynamically control how an allocated portion of a resource of the computing device is to be divided among multiple processes associated with the application. For example, an application executing on the computing device may interact with an operating system of the computing device to dynamically control how an allocated portion of a resource of the computing device is to be divided among a foreground process associated with the application and one or more background processes associated with the application.

Abstract: Resources used by a foreground process associated with an application as well as one or more background tasks associated with the application that are running as separate processes on the computing device are tracked. Historical data relating to resource usage by the application is utilized to predict a resource usage amount for the application which is then stored. A subsequent launch of the application can be conditioned upon whether the predicted resource usage amount is available. If the predicted resource usage amount is available, it can be allocated to the application and the application can be launched. If the application then exceeds the predicted resource usage amount, other processes executing concurrently therewith may be terminated to free up the resource. Regardless of how the application performs, new resource usage data is obtained for the application, which is then used to adjust the predicted resource usage amount.