专利快速检索-快速检索全球专利，免费商用专利数据库-IPRDB

1. 发明专利

JP2000148512A METHOD FOR SCHEDULING JOB 失效
公开(公告)号：JP2000148512A
公开(公告)日：2000-05-30
申请号：JP31613398
申请日：1998-11-06
申请人： HITACHI LTD
发明人： SATO KAZUHIRO , IWAKURA YOSHIYUKI , AKIBA CHIEKO , ISHIOKA ICHIRO
IPC分类号： G06F9/46 , G06F9/48
摘要： PROBLEM TO BE SOLVED: To prevent a job from going down and to moothly use CPU resources by guaranteeing the allocation time of a CPU to respective jobs without changing the execution priority of these jobs. SOLUTION: A CPU allocation guarantee control part 12 counts CPU tirne used for a job within unit time, and when a job guaranteed at its CPU allocation does not use the CPU more than fixed time, a CPU schedular 15 preferentially allocates the CPU to the job independently of the execution priority of jobs. The preferential allocation of the CPU is appliced also to a job securing a shared resource. Even in an environment that a job having higher execution priority occupies the CPU, rninimum necessary CPU allocation is applied to respective jobs.

2. 发明专利

JPH09185496A METHOD FOR MAINTAINING RESIDENT PROGRAM LIBRARY IN PARALLEL SYSTEM 失效
公开(公告)号：JPH09185496A
公开(公告)日：1997-07-15
申请号：JP34223195
申请日：1995-12-28
申请人： HITACHI LTD
发明人： MATSUDA TATSUMI , AKIBA CHIEKO , IWAKURA YOSHIYUKI
IPC分类号： G06F9/06
摘要： PROBLEM TO BE SOLVED: To dynamically maintain a resident program library without successively monitoring the program library or inputting any command by an operator by managing the program library resident on each OS. SOLUTION: A program library 20 is made resident in a program library resident area 11 on each OS and a program library management processing part 12 manages it in a program library managing area 32 on a storage device 30 which can be accessed at high speed in an entire parallel system. When the program library 20 under managing is updated while using a program library updating means 13, program library update information is reported to the program library management processing part 12 on the OS 10 as an update object. At the program library management processing part 12, the exchange of program library resident area 11 is executed based on the reported program library update information.

3. 发明专利

JPH08314769A FAULT INFORMATION SAMPLING METHOD 失效
公开(公告)号：JPH08314769A
公开(公告)日：1996-11-29
申请号：JP11953095
申请日：1995-05-18
申请人： HITACHI LTD
发明人： MASUDA HIDEKI , IWAKURA YOSHIYUKI
IPC分类号： G06F11/34
摘要： PURPOSE: To provide a fault information sampling method which supports the analysis of a fault factor by providing a step to designate the inquiry of a fault content when fault information is sampled and a step to sample the fault information as holding the state of an address space. CONSTITUTION: When abnormality occurs in an operating system provided with multiplex virtual storage function, an input means 2 informs the fault acceptance means 1 of the fault information managing program of an OS of it. The fault acceptance means 1 performs the confirmation and inquiry of the fault content from the input means 2 via the input means 2, and delivers the fault content to a fault judging means 3. The fault judging means 3 judges an inputted fault content and the state of the system, and decides a fault sampling method, and delivers it to a fault information sampling process 4. The fault information sampling process 4 outputs the fault information to a fault information output device 5, and uses it as the input information of an analyzing means 6 for fault factor. In this way, when the abnormality such as the one in which despatch is disabled after a job is executed, etc., occurs, the fault information is sampled as holding job information.

4. 发明专利

JP2000122918A ADDRESS CONVERSION DEVICE AND ITS METHOD 失效
公开(公告)号：JP2000122918A
公开(公告)日：2000-04-28
申请号：JP29717498
申请日：1998-10-19
申请人： HITACHI LTD
发明人： HIRAIWA YURI , NAGASUGA HIROFUMI , AKIBA CHIEKO , IWAKURA YOSHIYUKI
IPC分类号： G06F12/02 , G06F12/06
摘要： PROBLEM TO BE SOLVED: To reduce a use impossible area generated in real storage areas divided into plural sections. SOLUTION: An addition circuit 108 adds a point value (indicating the head of a section information table 102) stored in a control register 101 and SCD (section identifier) 201 of a real address 200 and extracts section information (section start address information 103 and section length information 104) corresponding to SCD from the section information table 102. A comparison operation circuit 106 compares section length information with the page address 203 of the real address. When the latter is larger than the former, an access exception processing is executed. When the latter is smaller than the former, an addition circuit 107 adds section start address information 103 and the page address, obtains target SCD 301 and a target page address 303, connects it with a displacement 304 obtained by copying the displacement 204 of the real address and generates a target address 300.

5. 发明专利

JPH11353284A JOB RE-EXECUTING METHOD 失效
公开(公告)号：JPH11353284A
公开(公告)日：1999-12-24
申请号：JP16180898
申请日：1998-06-10
申请人： HITACHI LTD
发明人： HONDO YURI , NAGASUGA HIROFUMI , AKIBA CHIEKO , IWAKURA YOSHIYUKI
IPC分类号： G06F11/14 , G06F11/16 , G06F15/16
摘要： PROBLEM TO BE SOLVED: To attain the efficient operation of a cluster type computer system wherein the influence of computer faults is suppressed by finding a job to be re-executed at a local computer node among jobs which were executed on a 1st computer and re-executing the found job on the local computer node. SOLUTION: Computers 106 connected by a high-speed coupling device 101 consisting of a CPU 102 and a memory 103 are each composed of a CPU 107 and a memory 108. In a memory 108 of each computer 106, a jog start program 109 which starts a job in response to a report from a jog reloading part 600 and a job execution program 110 which executes the job are set. A computer which re-executes the jog is determined according to information of the job and load information from a node. When it is decided that the determined local computer suitably re-executes the job, the job is loaded onto the local computer and re-executed.

6. 发明专利

JPH10283330A LOAD DECENTRALIZATION CONTROL METHOD FOR PARALLEL COMPUTER 失效
公开(公告)号：JPH10283330A
公开(公告)日：1998-10-23
申请号：JP8628497
申请日：1997-04-04
申请人： HITACHI LTD
发明人： MORITA YOSHINORI , SUDA TAKEHIKO , SHIMIZU HIDENORI , IWAKURA YOSHIYUKI
IPC分类号： G06F15/16 , G06F9/46 , G06F15/177
摘要： PROBLEM TO BE SOLVED: To improve the performance of parallel computers by setting the upper-limit value of the number of processes that can be performed without losing load balance and updating load information on a subsystem into the latest load information when the number of processes in an update interval of the load information exceeds the upper-limit value. SOLUTION: Each time a process request is accepted in the update interval of the load information, a load decentralization control mechanism 12 allocates the process to a subsystem with a light load according to the load information 121 and makes a process number counter 122 count up. When the count value of the process number counter 122 is less than the upper-limit value 123 of the number of successive processes, nothing is done and the process is ended. When the process number counter 122 exceeds the upper-limit value 123, the load information 121 of all subsystems is updated into the latest load information and then the process number counter 122 is reset. Even when time-out information is received, the load information of all the subsystems is updated into the latest load information and then the process number counter 122 is reset.

7. 发明专利

JPH103395A METHOD FOR COLLECTING RESOURCE USING INFORMATION OF COMPUTER 失效
公开(公告)号：JPH103395A
公开(公告)日：1998-01-06
申请号：JP15409896
申请日：1996-06-14
申请人： HITACHI LTD
发明人： OCHI FUTOSHI , IWAKURA YOSHIYUKI , OKUDA TORU , MATSUDA TATSUMI
IPC分类号： G06F11/34 , G06F9/46
摘要： PROBLEM TO BE SOLVED: To set resource using information in which resource using state is accurately reflected without using transient resource using information when resource using information of a computer is collected, and to reduce overhead of the computer spent in collecting resource using information when load of the computer is relatively high. SOLUTION: Each resource using rate collecting part 11, 21 and 31 provided in respective processors 1, 2 and 3 collects a using value of computer resource at intervals (t), and sets the mean value of successive resource using values for (n) times when an integer >=2 is defined as (n), is set as the resource using value at intervals (t)×(n). Then the latest resource using value is compared with the last resource using value. If the resource using value is increased, the collecting interval (t) is increased, while if reduced, the interval (t) is reduced.

8. 发明专利

JPH09167141A LOAD DISTRIBUTION CONTROL METHOD 失效
公开(公告)号：JPH09167141A
公开(公告)日：1997-06-24
申请号：JP32844295
申请日：1995-12-18
申请人： HITACHI LTD
发明人： OKUDA TORU , IWAKURA YOSHIYUKI , NAGASUGA HIROFUMI
IPC分类号： G06F15/16 , G06F9/50 , G06F15/177
摘要： PROBLEM TO BE SOLVED: To provide a technology for load-distributing plural kinds of jobs to plural computers constituting a computer group. SOLUTION: In the method, a parallel computers system provided with the computer group constituted by the plural computers 2-4 and with a high speed connection mechanism 1 being a computer managing the operation of the computer group load-distributes the plural kinds of jobs. The use target values of resources for the respective computers 2-4 constituting the computer group are set and the use condition of the resources for the respective jobs of the computers 2-4 are collected. Then, they are reported to the high speed connection mechanism 1. The high speed connection mechanism 1 selects the computers 2-4 to which starting requests are newly given and which execute the jobs among the plural computers 2-4 of the computer group from a difference between the use target values of the resources for the respective jobs of the computers 2-4 and the reported use condition of the resources for the respective jobs. The selected computers 2-4 execute the jobs to which the starting requests are newly given.

9. 发明专利

JPH0944268A TIME CHANGING METHOD FOR COMPUTER SYSTEM 失效
公开(公告)号：JPH0944268A
公开(公告)日：1997-02-14
申请号：JP19285195
申请日：1995-07-28
申请人： HITACHI LTD
发明人： IWAKURA YOSHIYUKI , MASUDA HIDEKI
IPC分类号： G06F1/14 , G06F15/16
摘要： PROBLEM TO BE SOLVED: To simultaneously change the time of all computer systems without stopping any system by changing the time of plural computer systems through a shared device. SOLUTION: Time update request receiving processing 6 for updating a working area 2 and an indicator 3 on a shared device 1 when the present computer system receives a time change request and indicator state monitoring/time setting processing 7 for monitoring the state of the indicator 3 on the shared device 1 and setting the time of the present computer system when that state is changed are arranged in respective computer systems 8a and 8b. In the computer system 8a that receives the change request of time, the contents of change are stored in the working area 2 and the indicator 3 is updated later. The respective computer systems 8a and 8b monitor the state of the indicator 3 by the monitoring processing 7 and when the state change of the indicator 3 is detected, time 9 of the present computer system is reflected with the value of time stored in the working area 2 on the shared device 1.

你已经成功收藏专利！

检索式保存成功!

IPRDB

热门服务

关于我们

友情链接

联系方式