Abstract: A technique for processing instructions in an electronic system is provided. In one embodiment, a processor of the electronic system may submit a unit of work to a queue accessible by a coprocessor, such as a graphics processing unit. The coprocessor may process work from the queue, and write a completion record into a memory accessible by the processor. The electronic system may be configured to switch between a polling mode and an interrupt mode based on progress made by the coprocessor in processing the work. In one embodiment, the processor may switch from an interrupt mode to a polling mode upon completion of a threshold amount of work by the coprocessor. Various additional methods, systems, and computer program products are also provided.

Abstract: Systems, methods and computer program products for provisioning a virtual machine are disclosed. A request to provision a virtual machine is received. Resources are identified and utilization of the resources determined for each cluster of the plurality of virtual machine clusters, the resources comprising internet protocol (IP) addresses, memory and CPUs. The availability of clusters is analyzed based on the determined resource utilization for each cluster. The analysis includes assigning dynamic weights to resources of each cluster and calculating an availability of each cluster. A cluster availability report is output indicating an availability status for each virtual machine cluster. A system for provisioning a virtual machine includes a provision manager, a resource determiner, an availability analyzer and an availability reporter.

Abstract: A method includes receiving a deployment template at a host system, partitioning a networking device of the host system with a virtual network partition based upon the deployment template, receiving attribute data that includes a first attribute associated with the first virtual network partition, loading the networking device with the attribute data, launching a virtual machine manager on the host system, reading the attribute from the networking device, and assigning a virtual network interface of a virtual switch associated with the virtual machine manager to the virtual network partition in response to reading the attribute.

Abstract: Different types of hardware processing resources in a system are identified (102). In response to a request to perform a task, a respective one of the different types of hardware processing resources is selected (104) to perform the task.

Abstract: Method for avoiding Guest I/O timeout, including generating a request to an I/O device in a Guest Operating System (GOS) running inside a VM; setting a VM timeout deadline; in a hypervisor, setting a supervisory timeout deadline for the I/O device to respond that is shorter than the VM timeout deadline; if no response is received before expiration of the supervisory timeout deadline, pausing the GOS; freezing system time of the GOS; activating a handler for the I/O device response in the hypervisor; upon receipt of the response prior to the supervisory timeout deadline, forwarding the response to the GOS, unpausing the GOS, unfreezing the system time of the GOS; upon receipt of the response after the supervisory timeout deadline, and after the GOS is unpaused by the hypervisor, the Guest OS receives the response from the hardware device; and if no response is received, the GOS is kept paused.

Abstract: A single workload scheduler schedules sessions and tasks having a tree structure to resources, wherein the single workload scheduler has scheduling control of the resources and the tasks of the parent-child workload sessions and tasks. The single workload scheduler receives a request to schedule a child session created by a scheduled parent task that when executed results in a child task; the scheduled parent task is dependent on a result of the child task. The single workload scheduler receives a message from the scheduled parent task yielding a resource based on the resource not being used by the scheduled parent task, schedules tasks to backfill the resource, and returns the resource yielded by the scheduled parent task to the scheduled parent task based on receiving a resume request from the scheduled parent task or determining dependencies of the scheduled parent task have been met.

Abstract: Techniques, systems, and articles of manufacture for automatically propagating updates in a data center. A method includes dividing multiple virtual machines in a data center into one or more equivalence classes, wherein each equivalence class is based on a signature corresponding to an offline, online and/or manual update, automatically creating an offline and/or online virtual machine manifest for a first virtual machine of the multiple virtual machines based on one or more file system changes during the offline, online and/or manual update for the first virtual machine, wherein said manifest is applicable to each additional virtual machine in the same equivalence class as the first virtual machine, and applying the offline and/or online virtual machine manifest for the first virtual machine to the remaining multiple virtual machines in the same equivalence class to automatically update the remaining multiple virtual machines in the same equivalence class.

Abstract: A set of jobs to be scheduled is identified (402) in a system including a processing pipeline having plural processing stages that apply corresponding different processing to a data update to allow the data update to be stored. The set of jobs is based on one or both of the data update and a query that is to access data in the system. The set of jobs is scheduled (404) by assigning resources to perform the set of jobs, where assigning the resources is subject to at least one constraint selected from at least one constraint associated with the data update and at least one constraint associated with the query.

Abstract: A computer-readable recording medium stores an execution control program that causes a computer to execute a process that includes receiving an execution request for a given operation for a system; detecting number of operations that are of a type identical to that of the given operation and are under execution by a computing device that is in the system and involved in the execution of the given operation for which the execution request is received; comparing the number of operations detected at the detecting and the number of operations that are of the type and simultaneously executable by the computing device such that the execution of the given operation is completed within a given period by the computing device; and assigning the given operation to the computing device, based on a result of comparison at the comparing.

Abstract: A “just in time” feedback-driven strategy for provisioning storage in a cloud computing environment is described. An agent miming in a guest virtual machine can monitor the storage needs of the guest virtual machine. When a specified threshold for the guest virtual machine is reached, the host can be notified of the changing storage needs. The host can respond by re-sizing the storage device and notifying the guest virtual machine of the changed storage capacity.

Abstract: A method includes a workload management (WLM) server that receives a first CHECK WORKLOAD command for a workload in a queue of the WLM server. It may be determined whether the workload is ready to run on a WLM client. If the workload is not ready to run, a wait time for the workload with the WLM server is dynamically estimated. The wait time is sent to the WLM client. If the workload is ready to run, then a response is sent to the WLM client that workload is ready to run.

Abstract: A method of processing a job is presented. A packet selector determines a candidate job list including an ordered listing of candidate jobs. Each candidate job in the ordered listing belongs to a communication stream. One or more shared resources required for execution of a first job in the candidate job list are identified. Whether the first job is eligible for execution is determined by determining an availability of the one or more shared resources required for the first job, and, when the one or more shared resource required for the first job are unavailable and no jobs executing within the data processor are from the same communication stream as the first job, determining that the first job is not eligible for execution.

Abstract: A system and method for providing quality of service (QoS) for clients running on host computers to access a common resource uses a resource pool module and a local scheduler in at least one of the host computers. The resource pool module operates to compute an entitlement of each client for the common resource based on a current capacity for the common resource and demands of the clients for the common resource. In addition, the resource pool module operates to assign a portion of the computed current capacity for the common resource to a particular host computer using the computed entitlement of each client running on the particular host computer. The local scheduler operates to allocate the portion of the computed current capacity among the clients running on the particular host computer.

Abstract: A multicore system includes multiple processor cores; a scheduler in each of the processor cores and allocating a process to the processor cores when having a master authority that is an authority to assign processes; and a master controller performing control to repeat until a process to be executed no longer exists, a cycle in which the schedulers transfer the master authority to another processor core after receiving the master authority and before assigning a process to the processor cores, discards the master authority after assigning the process to the processor cores, and enters a state of waiting to receive the master authority.

Abstract: Determining collective barrier operation skew in a parallel computer that includes a number of compute nodes organized into an operational group includes: for each of the nodes until each node has been selected as a delayed node: selecting one of the nodes as a delayed node; entering, by each node other than the delayed node, a collective barrier operation; entering, after a delay by the delayed node, the collective barrier operation; receiving an exit signal from a root of the collective barrier operation; and measuring, for the delayed node, a barrier completion time. The barrier operation skew is calculated by: identifying, from the compute nodes' barrier completion times, a maximum barrier completion time and a minimum barrier completion time and calculating the barrier operation skew as the difference of the maximum and the minimum barrier completion time.

Abstract: Determining collective barrier operation skew in a parallel computer that includes a number of compute nodes organized into an operational group includes: for each of the nodes until each node has been selected as a delayed node: selecting one of the nodes as a delayed node; entering, by each node other than the delayed node, a collective barrier operation; entering, after a delay by the delayed node, the collective barrier operation; receiving an exit signal from a root of the collective barrier operation; and measuring, for the delayed node, a barrier completion time. The barrier operation skew is calculated by: identifying, from the compute nodes' barrier completion times, a maximum barrier completion time and a minimum barrier completion time and calculating the barrier operation skew as the difference of the maximum and the minimum barrier completion time.

Abstract: Computer-implemented methods, computer-program products, and systems instantiate a resource comsumption optimizer framework (RCOF) for a plurality of sub-process associated with a process. A plurality of sub-processes are loaded into a memory in accordance with a calculated optimized resource comsumption pattern. Each sub-process is associated with an agent that communicates with the RCOF. A particular sub-process is executed and the execution start is gated by an associated agent based upon at least a determined buffer value. The RCOF is notified of the particular sub-process execution completion.

Abstract: Method for avoiding Guest I/O timeout, including generating a request to an I/O device in a Guest Operating System (GOS) running inside a VM; setting a VM timeout deadline; in a hypervisor, setting a supervisory timeout deadline for the I/O device to respond that is shorter than the VM timeout deadline; if no response is received before expiration of the supervisory timeout deadline, pausing the GOS; freezing system time of the GOS; activating a handler for the I/O device response in the hypervisor; upon receipt of the response prior to the supervisory timeout deadline, forwarding the response to the GOS, unpausing the GOS, unfreezing the system time of the GOS; upon receipt of the response after the supervisory timeout deadline, and after the GOS is unpaused by the hypervisor, the Guest OS receives the response from the hardware device; and if no response is received, the GOS is kept paused.

Abstract: A technique for identifying conflicting sub-processes easily in a computer system that processes a plurality of transactions in parallel is provided. In a conflicting sub-process identification method according to the present invention, a comparison is made between processing time periods for which transactions executed in parallel have conflicting parts and for which transactions executed in parallel do not have conflicting parts. This comparison is made for a plurality of transaction processing programs each comprising a plurality of sub-processes. Then, on the basis of correspondence relationships between performance indices obtained from the comparison result, conflicting combination of sub-processes competing for a resource are identified.

Abstract: In one embodiment, a method includes tracking at a network device in communication with a plurality of virtual machines, virtual machine movement based on a device identifier in a packet received from one of the virtual machines and storing location information for the virtual machine in a virtual machine move list at the network device. The location information from the virtual machine move list is inserted into a forwarding information base for use in optimized forwarding of packets destined for the virtual machine. An apparatus and logic are also disclosed herein.