Abstract: An online system predicts a number of interruption events within a time period and identifies anomalous numbers of interruption events using an interruption prediction model. The online concierge system maintains application state data that describes a state of an application workflow for a client application. The online concierge system identifies interruption events that represent interruptions to the application workflow and logs interruption events in an interruption log, wherein each entry of the interruption log describes an interruption event and its corresponding state. The online concierge system predicts a number of interruption events that will occur within a time period based on an interruption prediction model. The online concierge system computes an actual number of interruption events that occurred during the time period and computes a difference between the actual number and the predicted number.

Abstract: There is provided a method performed by an entity for handling execution of functions in a function-as-a-service (FaaS) system. The FaaS system includes one or more nodes on which functions are executable. The method includes grouping functions into a group of functions to be executed on a single one of the one or more nodes. The grouping is based on information from the one or more nodes. The information is indicative of which functions access which data during execution of the functions and each function of the group accesses the same data as at least one other function of the group.

Abstract: An example method may include generating a block list comprising a plurality of list items, wherein each list item identifies a respective block of a source virtual machine image, and the list items are ordered in the block list according to a timestamp of each respective block, wherein the timestamp indicates a time of a last access of the respective block, sending the block list to a destination computing device, receiving, from the destination computing device, one or more candidate blocks, determining whether the one or more candidate blocks are included in the source virtual machine image, and sending, to the destination computing device, a result indicating whether the one or more candidate blocks are included in the source virtual machine image.

Abstract: Provided are techniques for reducing a total amount of memory usage of an edge device by swapping application logics. A first application is executed in a foreground on an application framework, where the first application has a shown status. In response to receiving a request to execute a swap application, it is determined that the first application and the swap application use a same screen type and a same non-shareable resource. An application logic of the first application is stored in a suspended application logic pool, where the first application changes from the shown status to a hidden-suspended status. The swap application is executed in the foreground on the application framework, where the swap application has the shown status.

Abstract: Systems and methods for secure saving sand restoration of virtualized states is disclosed herein, wherein a guest and/or a virtual machine is aware of being executed from a restored snapshot. An example system comprises a hypervisor; an application running in an encrypted virtual machine; a memory; a processor in communication with the memory executing an application, wherein the application is configured to modify, by an application running in an encrypted VM, a shadow identifier (Shadow ID) saved in private memory; and save a state of the encrypted VM, as a snapshot.

Abstract: A module optimizes active and inactive components and resources of a hybrid computing system to optimize a combination of private and public resources to minimize cost and maximize performance of the hybrid computing system. A learning model may analyze past usage and present usage metrics of one or more components with respect to performance criteria or cost criteria. Cost factors associated with components of the private system may be based on wear—the higher the wear the less desirable a component's use becomes due to lower reliability and higher warranty costs. When activating a component of the private system, a deactivation of a higher wear component may be delayed allowing time for a recently activated component to be integrated with the private system. A resource of a public system may be used while deactivation of a high wear component is delayed.

Abstract: A fair and efficient guest to hypervisor virtual machine socket protocol may be provided by: in response to a host determining to reject a message received from a guest that was previously accepted for processing by the host, transmitting a rejection to the guest; in response to receiving, at the guest, the rejection, adding the message to a processing request queue on the guest; in response to determining that resources to handle the message have become available for the guest, transmitting an indication to the guest that the host is able to reaccept the message; in response to receiving, at the guest, the indication, retransmitting the message to the host according to the processing request queue; and in response to receiving the message from the guest a second time, accepting the message in an execution queue in a virtual memory of the host.