Abstract: In the computer apparatus which has a processing unit, a power consumption measuring unit, and a power counter, the power consumption of running programs on the processing unit is measured at arbitrary constant period, wherein the measuring value is integrated to the power counter. When the power counter overflows, the processing unit is interrupted for sampling information required for analysis. Then the processing unit which received the interruption executes a sampling of the power consumption base. So, power consumption based sampling and profiling becomes to be enabled.

Abstract: A communication performance analyzing program, a communication performance analyzing apparatus and a communication performance analyzing method make it possible to highly reliably grasp the communication performance of a computer system by automatically analyzing the communication performance. The communication performance analyzing program that causes a computer to analyze a data tendency of communication performance of a plurality of execution periods of a computer system formed by connecting a plurality of computers by a network, the program comprises a communication time acquisition step S2 that acquires communication time data among the computers of the computer system and a statistical analysis step S3 that analyzes the tendency of communication performance data of each execution period of the computer system, using the communication time among the computers, by statistically analyzing the communication time data acquired by the communication time acquisition step.

Abstract: A disclosed priority control program recorded in a computer-readable medium causes a computer to execute, in job allocation for computational resources, a first step of lowering a job allocation priority of a user based on an estimated utilization amount of a job associated with the user, the job allocation priority indicating a degree of priority of the user in obtaining an allocation of the computational resource, and the estimated utilization amount being an amount of the computational resources estimated to be used for the job and being submitted to and recorded in a memory device on a job-to-job basis; and a second step of increasing the job allocation priority over time at a restoration rate which corresponds to a user-specific amount of the computational resources available for the user per unit time, the user-specific amount being recorded in the memory device on a user-to-user basis.

Abstract: A performance monitoring device has an interrupt detection unit that detects generation of an interrupt process to be executed by a processor in accordance with TLB entry invalidation executed in an operating system. A counter value acquisition unit acquires a counter value of a predetermined event counted by the processor when the interrupt process is detected by the interrupt detection unit. A process information acquisition unit acquires identification information for identifying a process executed on the processor from the operating system immediately before the interrupt process is detected by the interrupt detection unit. An associating unit associates the counter value acquired by the counter value acquisition unit during the interrupt process with the identification information acquired by the process information acquisition unit immediately before the interrupt process.

Abstract: When a data transfer request is received, a transmitting-side NIC generates an RDMA packet from transfer data and speculatively transmits it without inquiring the transfer destination about presence of reception permission. If reception of a receiving area is unpermitted, the RDMA packet is retransmitted when a retransmission request is received from the transfer destination. When transfer unpermission is determined by referencing transfer area control information upon packet reception, a receiving-side NIC discards the received packet and, if transfer permission is determined thereafter, transmits a retransmission request, thereby transferring it. If reception is unpermitted, that may be notified to the transmission source, thereby ceasing the transfer.

Abstract: The system provides a technique for detecting changes in behavior of a computer system. An address data storing unit obtains an address being accessed by a CPU at prescribed sampling intervals, stores address data comprising the address and its acquisition time in an address record table in time series. Then an executed process detector detects the name of the process that was executed when the address data was obtained, and sets the process name for the address data in the address record table. When an analysis request specifying an analysis time period is made, an analyzer counts up each of the process names included in the address data being stored in the address record table for each specified analysis time period, and analyzes the breakdown of executed processes for each analysis time period. Then a display unit displays the analysis result of the analyzer.

Abstract: A computer system including input/output devices that transfer data and a computer that controls a process using a virtual storage and inputs data to and outputs data from a medium, wherein the input/output devices and the computer include address conversion tables for converting a virtual address into an actual address, the computer requests all the input/output devices to permit page operation such as page-out processing, processing for deleting a virtual address space following termination of a process, and processing for changing a page, the input/output devices perform operation for the address conversion tables in response to a request from the computer and notifies the computer of permission of the page operation, and, when the permission for the page operation is obtained from all the input/output devices, the computer performs operation for the address conversion table and performs page operation.

Abstract: A system analysis apparatus acquires time series data representing information on a series of processes executed in a computer system in a time-serial manner. A run table shows the ratio of the number of runs of each process to the total number of processes executed in a predetermined unit time in a time-serial manner. A first graph is displayed representing ratios of a series of processes which has been executed in a time-serial manner by collecting the ratio of each process stored in the run table. A moving average of the ratio of each process is calculated by referring to the run table when a parameter change for the first graph is accepted. A second graph is displayed representing the ratios of the series of processes executed by collecting the moving average of each process.

Abstract: A computer has a plurality of processors with a cache memory. When a spinwait detecting unit provided to a processor detects execution of a spinwait command, it instructs monitoring of a variable value as a spinwait end condition and changes an operating state of a processor. A value change detecting unit provided to the cache memory monitors the variable value specified by the spinwait detecting unit, and when it detects that the variable value is changed, it posts the value change to the processor so as to return the operating state into its original state.

Abstract: A communication performance analyzing program, a communication performance analyzing apparatus and a communication performance analyzing method make it possible to highly reliably grasp the communication performance of a computer system by automatically analyzing the communication performance. The communication performance analyzing program that causes a computer to analyze a data tendency of communication performance of a plurality of execution periods of a computer system formed by connecting a plurality of computers by a network, the program comprises a communication time acquisition step S2 that acquires communication time data among the computers of the computer system and a statistical analysis step S3 that analyzes the tendency of communication performance data of each execution period of the computer system, using the communication time among the computers, by statistically analyzing the communication time data acquired by the communication time acquisition step.

Abstract: Information on an execution status of an information-collection target program is collected with an interrupt handler that is activated by an interrupt generated when a predetermined condition is satisfied. A target range for collecting the information with the interrupt handler is specified, and information to be collected by the interrupt handler when the interrupt is generated within the specified target range is set.

Abstract: A cluster control apparatus controls an allocation of a job to each node in a cluster system formed by mutually connecting a plurality of nodes by a high-speed network. A node selecting unit selects an allocation node for allocating the job and a similar node that is similar to the allocation node. A data-redundancy command unit issues a command to store data created from the job executed by the allocation node in the allocation node and the similar node.

Abstract: A computer-readable recording medium storing a system analysis program capable of detecting changes in behavior occurring in a very short time without fail. An address data storing unit obtains an address being accessed by a CPU at prescribed sampling intervals, stores address data comprising the address and its acquisition time in an address record table in time series. Then an executed process detector detects the name of the process that was executed when the address data was obtained, and sets the process name for the address data in the address record table. When an analysis request specifying an analysis time period is made, an analyzer counts up each of the process names included in the address data being stored in the address record table for each specified analysis time period, and analyzes the breakdown of executed processes for each analysis time period. Then a display unit displays the analysis result of the analyzer.

Abstract: A recording medium which is readable by a computer stores a performance analyzing program for searching for a node that is peculiar in performance in a cluster system, as well as unknown problems. The performance analyzing program enables the computer to function as various functional units. A performance data analyzing unit collects performance data of nodes which make up the cluster system from performance data storage unit for storing a plurality of types of performance data of the nodes, and analyzes performance values of the nodes based on the collected performance data. A classifying unit classifies the nodes into a plurality of groups by statistically processing the performance data collected by the performance data analyzing unit according to a predetermined classifying condition.

Abstract: When a data transfer request is received, a transmitting-side NIC generates an RDMA packet from transfer data and speculatively transmits it without inquiring the transfer destination about presence of reception permission. If reception of a receiving area is unpermitted, the RDMA packet is retransmitted when a retransmission request is received from the transfer destination. When transfer unpermission is determined by referencing transfer area control information upon packet reception, a receiving-side NIC discards the received packet and, if transfer permission is determined thereafter, transmits a retransmission request, thereby transferring it. If reception is unpermitted, that may be notified to the transmission source, thereby ceasing the transfer.

Abstract: In an accounting apparatus, a conflict determination unit determines whether or not the accounting mode is in a conflict state where a process is executing in another logical CPU and stores the determination result in an accounting information storage unit, when a process of the user starts to be executed in a logical CPU of an SMT processor. And a CPU use time acquisition unit collects the CPU use time of the process in the conflict state or the non-conflict state distinctively and stores it in an accounting information storage unit. Thereafter, a CPU use time conversion unit converts the CPU use time in the conflict state, with a predetermined weighting, based on the CPU use time in the conflict state and the non-conflict state, after the end of executing the process, and an accounting calculation unit calculates the accounting amount for the process from an effective use time.

Abstract: An input/output device stores host status information about the status of a host, and controls the input/output of data. By referring to the input/output information and the host status information, the device performs optimization control of selecting notification either by an interrupt process or a non-interrupt process. When a notification by the interrupt process is selected, data transfer control is performed to the host, and the data is transferred to the host by an interrupt. When a notification by the non-interrupt process is selected, data transfer control is performed to the host, and notification control is performed to transfer the data to the host by polling.

Abstract: An apparatus which can collect information for a long time or collect information on an entire memory space with a small hardware physical quantity while avoiding unnecessary overhead. The apparatus comprises a counting unit for counting the number of events having occurred in a processor or a computer system, a retaining unit for retaining a count value obtained by the counting unit, and a control unit for controlling writing of a count value into the retaining unit, wherein the control unit controls the writing so as to write a new count value obtained by the counting unit in said retaining unit while leaving a count value of a high degree of significance in the retaining unit. The apparatus is used to measure the number of events having occurred in the processor when the performance of the processor or the a computer system is measured or tuned.

Abstract: A task scheduling apparatus of a distributed processing system having a plurality of processing units for processing a plurality of distributed tasks is provided. As a first task scheduling method, the task scheduling apparatus allocates a task to a processing unit having the lowest temperature. As a second task scheduling method, the task scheduling apparatus selects a task based on both temperature of each processing unit and characteristic values of tasks related to degree of temperature rise or consumption power increase caused by execution, and allocates the selected task to the object processing unit. For example, as the second task scheduling method, a task producing a large degree of temperature rise (for example, a task having a number of instructions to be processed per unit time) is allocated to a processing unit having a low temperature. With such a scheduling method, uniform temperature of each processing unit can be obtained.

Abstract: A computer system including input/output devices that transfer data and a computer that controls a process using a virtual storage and inputs data to and outputs data from a medium, wherein the input/output devices and the computer include address conversion tables for converting a virtual address into an actual address, the computer requests all the input/output devices to permit page operation such as page-out processing, processing for deleting a virtual address space following termination of a process, and processing for changing a page, the input/output devices perform operation for the address conversion tables in response to a request from the computer and notifies the computer of permission of the page operation, and, when the permission for the page operation is obtained from all the input/output devices, the computer performs operation for the address conversion table and performs page operation.