Abstract: Implementations for a user interface for MapReduce job are disclosed. A method of the disclosure includes receiving a request to submit a MapReduce job to a cluster and providing a first graphical user interface (GUI) to input a first set of arguments and first argument values for a first portion of the arguments. The method may include receiving, from the first GUI, the first argument values and a second portion of the arguments, associated with a configuration of the cluster and without a corresponding argument value. The method may provide a second GUI that is to display the arguments, first argument values, and allow for a selection of second argument values for the second portion of the arguments. In response to a selection, the method may execute the MapReduce job in view of the first argument values from the first GUI and the second argument values from the second GUI.

Abstract: One embodiment of the present invention provides a dynamic overlay tunnel orchestration system. During operation, the system detects the appearance of a first virtual machine running on a hypervisor of a first host machine coupled to a first switch in a network. The system identifies a first virtual local area network (VLAN) associated with the first virtual machine and determines whether an overlay tunnel exists between the first switch and a second switch coupling a second virtual machine belonging to the first VLAN. If no such overlay tunnel exists, the system generates an instruction for the first and second switches to establish a first overlay tunnel between the first and second switches.

Abstract: Techniques for placing containers in a cloud (e.g., into virtual machines (“VMs”)) based on container policies. The container policies may specify compute-related qualities, storage-related quality, and/or network-related qualities that are to be met by the underlying software and hardware that supports execution of the virtual machines. A cloud director or other entity receives requests to place containers in a particular virtual machine based on the container policies and directs placement of the virtual machine based on the policies. The cloud director may migrate and/or reconfigure VMs, virtual machine disk files, and/or virtual network interface controllers to satisfy the container placement policy. After placement, the cloud director may disable migration to maintain the VM in a desired state.

Abstract: A method begins by a processing module of a dispersed storage and task (DST) execution unit receiving a data request for execution by the DST execution unit, the data request including an execution deadline. The method continues with the processing module comparing the execution deadline to a current time. When the execution deadline compares unfavorably to the current time the method continues with the processing module generating an error response. When the execution deadline compares favorably to the current time the method continues with the processing module determining a priority level based on the deadline and executing the data request in accordance with the priority level.

Abstract: Aspects of the present invention provide systems and methods that ascertain appropriate time windows for when a task or tasks are best suited to be performed and also for appropriately sizing the number of tasks so that they can be completed within a window or windows. In embodiments, a system for estimating a system resource comprises a component selection system configured to receive a selection of one or more components and/or one or more metrics to be monitored. In embodiments, an analyzer uses at least some of the gathered data to determine one or more resource capacity windows for performing a task or tasks and determines an appropriate job size or sizes for scheduling the tasks to be performed within the one or more resource capacity windows.

Abstract: A system and method is provided for migrating a virtual machine between servers to access physical resources. An exemplary method includes intercepting a request from a client device to access a physical resource; identifying a virtual machine on a first server that is currently hosting a remote application for the client device; identifying a second server associated with the physical resource; and migrating the virtual machine from the first server to the second server, such that the client device can access the physical resource using the second server.

Abstract: A plurality of requests are received for computing processing. At least some of the plurality of requests are replicated. The requests are replicated based on a fractional replication factor. Each received request and each replicated request are transmitted to a computer resource for processing. At least some embodiments provide the capability for meeting tail latency targets with improved performance and reduced cost.

Abstract: In a multi-tenant environment, separate virtual machines can be used for configuring and operating different subsets of programmable integrated circuits, such as a Field Programmable Gate Array (FPGA). The programmable integrated circuits can communicate directly with each other within a subset, but cannot communicate between subsets. Generally, all of the subsets of programmable ICs are within a same host server computer within the multi-tenant environment, and are sandboxed or otherwise isolated from each other so that multiple customers can share the resources of the host server computer without knowledge or interference with other customers.

Abstract: In an embodiment, a method includes receiving time-series data related to individual utilization rates of a plurality of central processing unit (CPU) cores. The method also includes, responsive to a trigger, transforming at least a portion of the individual utilization rates into a compressed value format, the transforming comprising: determining, from at least a portion of the time-series data, CPU-core utilization values of the plurality of CPU cores for a time constraint; evaluating the CPU-core utilization values against a preconfigured set of non-overlapping value ranges; responsive to the evaluating, determining a quantity of the CPU-core utilization values that correspond to each non-overlapping value range of the preconfigured set of non-overlapping value ranges; and generating a CPU-core utilization snapshot comprising the determined quantity of each non-overlapping value range of the preconfigured set of non-overlapping value ranges.

Abstract: Profiling operating efficiency deviations of a computing system includes: generating a profile of expected operating efficiency for a computing system in an ideal configuration; for each of a plurality of alternative configurations of the computing system, wherein each of the alternative configurations includes a variation of the ideal configuration that introduces a deviation in operating efficiency of the computing system, said variation comprising a root cause of the deviation: monitoring operating efficiency of the computing system identifying, from the monitored operating efficiency, a deviation of operating efficiency from the expected operating efficiency; and recording, in a data structure, an association of the deviation and the root cause of the deviation.

Abstract: Disclosed herein are an apparatus and method for in-memory-based virtual desktop service. The apparatus for in-memory-based virtual desktop service includes a connection broker for performing a task for coordinating a delivery protocol that is used between at least one user terminal that uses virtual desktop service and multiple servers that provide the virtual desktop service, a resource pool for providing software resources including an Operating System (OS) for the virtual desktop service; and virtual machine infrastructure for supporting hardware resources, and dynamically allocating software stored in the software resources to the hardware resources.

Abstract: A controller coordinates execution of a set of related processes executed by respective devices in the virtual network, wherein coordinating comprises causing the respective devices to execute the set of related processes; receiving a data set for the set of related processes from the respective devices, comprising receiving operational states of the related processes from the respective devices; reading a previous data set comprising previous operational states of the related processes from the respective devices; processing an update to the previous operational states from the received operational states of the received data set; and aggregating the received operational states of the data set with the previous operational states of the related processes to form aggregated data of updated operational states.

Abstract: Various embodiments proactively balance workloads between a plurality of processing units of a multi-processor computing device by making work-stealing determinations based on operating state data. An embodiment method includes obtaining static characteristics data associated with each of a victim processor and one or more of a plurality of processing units that are ready to steal work items from the victim processor (work-ready processors), obtaining dynamic characteristics data for each of the processors, calculating priority values for each of the processors based on the obtained data, and transferring a number of work items assigned to the victim processor to a winning work-ready processor based on the calculated priority values. In some embodiments, the method may include acquiring control over a probabilistic lock for a shared data structure and updating the shared data structure to indicate the number of work items transferred to the winning work-ready processor.

Abstract: A virtual runtime module that omits an internal functional implementation of an associated executable module and that includes a runtime-resolvable public interface of the associated executable module is obtained using a processor within a module-based system. The virtual runtime module within the module-based system is resolved, using the runtime-resolvable public interface of the virtual runtime module, to satisfy dependencies associated with the executable module within the module-based system. At least a portion of the internal functional implementation of the associated executable module within the module-based system is installed during runtime using the resolved virtual runtime module.

Abstract: Data processing workloads are selectively assigned within a data center and/or among data centers based on non-data processing overhead within the data center and/or among the data centers. Power consumption of a rack including servers is predicted based on data processing demands that are placed on the servers for a given data processing workload, and power consumed by the rack is measured when the servers are performing the given data processing workload. A metric of power consumed by the rack for non-data processing overhead is derived based on a difference between results of the predicting and the measuring. A future data processing workload is selectively assigned to the rack based on the metric of power of power consumed by the rack for the non-data processing overhead. Assignment may also take place at an aisle and/or data center level based on these metrics.

Abstract: In the various aspects, virtualization techniques may be used to improve performance and reduce the amount of power consumed by selectively enabling a hypervisor operating on a computing device during sandbox sessions. In the various aspects, a high-level operating system may allocate memory such that its intermediate physical addresses are equal to the physical addresses. When the hypervisor is disabled, the hypervisor may suspend second stage translations from intermediate physical addresses to physical addresses. During a sandbox session, the hypervisor may be enabled and resume performing second stage translations.

Abstract: Embodiments of an invention for virtualization exceptions are disclosed. In one embodiment, a processor includes instruction hardware, control logic, and execution hardware. The instruction hardware is to receive a plurality of instructions, including an instruction to enter a virtual machine. The control logic is to determine, in response to a privileged event occurring within the virtual machine, whether to generate a virtualization exception. The execution hardware is to generate a virtualization exception in response to the control logic determining to generate a virtualization exception.

Abstract: A determination is made as to whether a plurality of processing entities in a processor complex exceeds a predetermined threshold number. In response to determining that the plurality of processing entities exceeds the predetermined threshold number, a processing entity of the plurality of processing entities is reserved for exclusive execution of critical tasks. In response to determining that the plurality of processing entities does not exceed the predetermined threshold number, and in response to receiving a task that is a critical task for execution, a determination is made as to which processing entity of the plurality of processing entities has a least amount of processing remaining to be performed for currently scheduled tasks. In response to moving tasks queued on the determined processing entity to other processing entities, the critical task is scheduled for execution on the determined processing entity.

Abstract: The technology disclosed relates to providing strong ordering in multi-stage processing of near real-time (NRT) data streams. In particular, it relates to maintaining current batch-stage information for a batch at a grid-scheduler in communication with a grid-coordinator that controls dispatch of batch-units to the physical threads for a batch-stage. This includes operating a computing grid, and queuing data from the NRT data streams as batches in pipelines for processing over multiple stages in the computing grid.

Abstract: The technology described herein manages the deployment of a group of machines from a staged state to a production state, while maintaining both the production and staged machines behind a single virtual internet protocol (VIP) address. The machines may be deployed within one or more data centers. Requests for service addressed to the VIP can be sent by a load balancer to machines within a staged pool or a production pool. The load balancer can evaluate characteristics of the request against a policy to determine whether to communicate the request to a machine in the first or second pool.