Abstract: Systems and methods are provided for the deterministic simulation of distributed systems, such as vehicle-based processing systems. A distributed system may be represented as a plurality of subsystems or “nodelets” executing with a single process of a computing device during a simulation. The nodelets may communicate using in-process communication. A task scheduler can schedule the nodelets to execute separately in serially-occurring frames. A simulated clock may be used to mitigate the variability in timestamped data that may be caused by latency or jitter.

Abstract: Techniques for an optimization service of a service provider network to help optimize the selection, configuration, and utilization, of virtual machine (VM) instance types to support workloads on behalf of users. The optimization service may implement the techniques described herein at various stages in a life cycle of a workload to help optimize the performance of the workload, and reduce underutilization of computing resources. For example, the optimization service may perform techniques to help new users select an optimized VM instance type on which to initially launch their workload. Further, the optimization service may monitor a workload for the life of the workload, and determine new VM instance types, and/or configuration modifications, that optimize the performance of the workload. The optimization service may provide recommendations to users that help improve performance of their workloads, and that also increase the aggregate utilization of computing resources of the service provider network.

Abstract: A process deployment controller creates an updated image for an intermediary engine in order to execute one or more applications on a host infrastructure. The process deployment controller generates a partial image by executing source code from a template repository. The partial image provides a structure used to create an intermediary engine used with a container, which includes an application, as well as binaries and libraries required to execute the application in an infrastructure via the intermediary engine. The process deployment controller transmits an identifier of the infrastructure to a component registry; receives a component description of the infrastructure from the component registry; and uses the component description to create an updated image of the partial image. The process deployment controller, upon receiving a request for the application to run on the infrastructure, utilizes the updated image and intermediary engine to execute the application on the infrastructure.

Abstract: Architectures for computing clusters with enhanced scalability and fault isolation capabilities are described. The enhanced capabilities are enabled by partitioning hosts of the clusters into multiple storage domains of the cluster. Each of the hosts includes processing and storage resources. The resources are logically partitioned into a processing layer and a storage layer of the cluster. The processing layer includes a virtualized processing resource for each of the storage domains. The storage layer includes a virtualized datastore for each of the storage domains. The processing layer provides users with access to virtualized machines (VMs). The storage layer provides the VMs with access to virtualized datastores of the associated storage domain. The aggregated resources of the host computing devices within the cluster may be employed to instantiate the processing layer and the storage layer of the cluster.

Abstract: The present disclosure relates to a method for a computer system comprising a plurality of processor cores, including a first processor core and a second processor core, wherein a cached data item is assigned to a first processor core, of the plurality of processor cores, for exclusively executing an atomic primitive. The method includes receiving, from a second processor core at a cache controller, a request for accessing the data item, and in response to determining that the execution of the atomic primitive is not completed by the first processor core, returning a rejection message to the second processor core.

Abstract: Implementing an application using a plurality of data processing engines (DPEs) can include, in a first pass, mapping, using computer hardware, a data flow graph onto an array of DPEs by minimizing direct memory access (DMA) circuit usage and memory conflicts in the array of DPEs and, in response to determining that a mapping solution generated by the first pass requires an additional DMA circuit not specified by the data flow graph, inserting, using the computer hardware, additional buffers into the data flow graph. In a second pass, the additional buffers can be mapped, using the computer hardware, onto the array of DPEs by minimizing the memory conflicts in the array of DPEs.

Abstract: A data management and storage (DMS) cluster of peer DMS nodes manages data of a tenant of a multi-tenant compute infrastructure. The compute infrastructure includes an envoy connecting the DMS cluster to virtual machines of the tenant executing on the compute infrastructure. The envoy provides the DMS cluster with access to the virtual tenant network and the virtual machines of the tenant connected via the virtual tenant network for DMS services such as data fetch jobs to generate snapshots of the virtual machines. The envoy sends the snapshot from the virtual machine to a peer DMS node via the connection for storage within the DMS cluster. The envoy provides the DMS cluster with secure access to authorized tenants of the compute infrastructure while maintaining data isolation of tenants within the compute infrastructure.

Abstract: A device and a method for operating a computer system, a job to be processed by the computer system being assignable to a task from a plurality of tasks for processing, the job to be processed being assigned as a function of a result of a comparison, a first value being compared to a second value in the comparison, the first value characterizing a first computing expenditure, which is to be expected in the computer system in the processing of the job to be processed in a first task of the plurality of tasks, the second value characterizing a second computing expenditure, which is to be expected in the computer system in the processing of the job to be processed in a second task of the plurality of tasks.

Abstract: In an approach for managing physical processor usage of a shared memory buffer, a processor receives a request for memory. A processor receives a request for memory from a process running on a physical processor. A processor determines whether the request for memory is less than or equal to a pre-determined threshold, wherein the pre-determined threshold is based on characteristics of a server on which the physical processor resides, needs of the server, and a frequency of requests of each memory size. Responsive to determining the request for memory is greater than the pre-determined threshold, a processor identifies a node on which the physical processor resides. A processor identifies a memory buffer of a set of memory buffers allocated to the node on which the physical processor resides. A processor allocates the memory buffer.

Abstract: In an approach for isolating physical processors during optimization of virtual machine placement, a server is provided comprising a plurality of containers and a plurality of physical processors. A processor builds a set of bit masks for each type of physical processor required for a logical partition. A processor builds a set of solution spaces based on the plurality of containers and an amount of each type of container of the plurality of containers. A processor completes a combinatorial search of the set of bitmasks and the set of solution spaces. A processor identifies a solution space of the set of solution spaces for the logical partition. The physical and logical configuration of the server is changed based on the solution space for the first logical partition.

Abstract: A system may detect a quit operation corresponding to a child application. In response to detection the quit operation, the system may store, in a memory, a mapping between the child application identifier and the child application task, and suspend display of the child application. The system may generate a foreground and background switching entry corresponding to the child application, the foreground and background switching entry associated with the child application identifier. The system may display the foreground and background switching entry in a visible region of the graphical user interface generated based on a parent application. The system may detect a selection operation indicative of the foreground and background switching entry. In response to the selection operation, the system may obtain the child application task from the memory according to the child application identifier, and resume display of the child application based on the child application task.

Abstract: Systems and methods are disclosed for managing resources associated with cluster-based resource pool(s). According to illustrative implementations, innovations herein may include or involve one or more of best fit algorithms, infrastructure based service provision, tolerance and/or ghost processing features, dynamic management service having monitoring and/or decision process features, as well as virtual machine and resource distribution features.

Abstract: Aspects of the present invention provide an approach that evaluates a locally running image (e.g., such as that for a virtual machine (VM)) and determines if that image could run more efficiently and/or more effectively in an alternate computing environment (e.g., a cloud computing environment). Specifically, embodiments of the present invention evaluate the local (existing/target) image's actual and perceived performance, as well as the anticipated/potential performance if the image were to be migrated to an alternate environment. The anticipated/potential performance can be measured based on another image that is similar to the existing/target image but where that image is running in a different computing environment. Regardless, the system would display a recommendation to the end user if it were determined that the image could perform better in the alternate environment (or vice versa). It is understood that performance is just one illustrative metric for which the system would perform a comparison.

Abstract: Reconstituting a machine image separates constituent parts of a machine image, and for each part, determines whether an exact version of the part is available on the target machine. If an exact version of the part is not available on the target machine, an inexact part is looked for on the target machine. Whether an inexact part is found may be determined based on attribute policy specification and similarity computation. For the inexact part found on the target machine, any dependencies may be identified and processed as a part to be reconstituted for the machine image. If no exact part and no inexact part are found on the target machine, the part is transferred from a source machine to the target machine. A machine image is created based on parts.

Abstract: A computing system may be in communication with client computing devices. The computing system may include a cloud infrastructure, an offline cache, and a VDA configured to concurrently have a first registration with the cloud infrastructure, and a second registration with the offline cache, and provide corresponding virtual desktop instances for the client computing devices based upon either the first registration or the second registration. The offline cache may be configured to broker local resources for the virtual desktop instances when the cloud infrastructure is unavailable. The VDA may be configured to transition to the offline cache using the second registration when the cloud infrastructure is unavailable.

Abstract: The present application discloses a task allocation method and task allocation apparatus for distributed data computing. The task allocation method includes: receiving storage parameters for target data to be computed in distributed data; mapping data slices of the target data to a resilient distributed dataset based on the storage parameters, each data slice corresponding respectively to a partition in the resilient distributed dataset; assigning each partition to a storage node to generate a computing task and perform the computing tasks. By using data storage information in a distributed database to allocate computing tasks to storage nodes corresponding to the data, Only data in local memories need to be called during the computing process, IO redundancy and time consumed due to repeated data forwarding are reduced.

Abstract: A plurality of processing entities are maintained. A plurality of task control block (TCB) groups are generated, wherein each of the plurality of TCB groups are restricted to one or more different processing entities of the plurality of processing entities. A TCB is assigned to one of the plurality of TCB groups, at TCB creation time.

Abstract: Described herein are systems and methods for providing data sets from a constantly changing database to a streaming machine learning component. In one embodiment, a data streaming sub-system receives multiple incoming streams of data sets, in which each stream is generated in real-time by one of multiple data sources. The streaming sub-system sends data sets, on-the-fly as they are received, to storage in the memory of a database, in which there is a linkage between the storage and the time of arrival or the time of storage, of the data sets. The database receives, from a machine learning component, a request to receive data sets according to a particular time or time period. In response to such request, the database identifies such data sets according to the particular time or time period and sends them to the machine learning component.

Abstract: Techniques promote monitoring of hypervisor systems by presenting dynamic representations of hypervisor architectures that include performance indicators. A reviewer can interact with the representation to progressively view select lower-level performance indicators. Higher level performance indicators can be determined based on lower level state assessments. A reviewer can also view historical performance metrics and indicators, which can aid in understanding which configuration changes or system usages may have led to sub-optimal performance.

Abstract: A set of resources required to process a data integration job is determined. In response to determining that the set of resources is not available, queue occupation, for each queue in the computing environment, is predicted. Queue occupation is a workload of queue resources for a future time based on a previous workload. A best queue is selected based on the predicted queue occupation. The best queue is the queue or queues in the computing environment available to be assigned to process the data integration job without preemption. The data integration job is processed using the best queue. It is determined whether a preemption event occurred causing the removal of resources from the best queue. A checkpoint is created in response to determining that a preemption event occurred. The checkpoint indicates the last successful operation completed and provides a point where processing can resume when resources become available.