JPH0778037A - Decentralized processing system - Google Patents

Abstract

PURPOSE:To exerts influence on none of other computers as much as possible when one computer is stopped and restarted by managing the stop scheduled time and restart scheduled time of the computer. CONSTITUTION:The decentralized processing system wherein plural computers 111-115 are connected to a network 12 has at least one management program 14 which manages the stop scheduled time and restart scheduled time of the respective computers 111-115 in the decentralized processing system. Consequently, when the time from the stop to the restart of a computer is within a relatively short scheduled period, this computer seems not to stop to other computers.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a distributed processing system in which a plurality of computers are connected to a network and execute a plurality of programs in a coordinated manner. In particular, the invention relates to maintenance and expansion of computers in the distributed processing system. The present invention relates to a distributed processing system having means for managing stop / resume.

[0002]

2. Description of the Related Art In a distributed processing system having a plurality of computers, programs on each computer cooperate to execute processing. In such a distributed processing system, there is a possibility that individual computers will be stopped / restarted due to maintenance / expansion, etc. It is important to prevent this from affecting the entire distributed processing system. By simply stopping the computer individually, for example, the program communicating with the program operating on the computer cannot obtain the reply. Therefore, the suspension of one computer affects other computers in the distributed processing system.

Conventionally, in such a distributed processing system, there is, for example, Japanese Patent Application Laid-Open No. 62-292564 as a device for preventing the stoppage of one computer from affecting the other. This improves the fact that when a certain device in the distributed processing system stopped, the failure detection function for all devices in the system was stopped during the stop period, and the stop time zone can be recognized. In this way, only the failure detection function for that place is stopped, and the failure detection function for other devices is activated.

[0004]

However, in the above-mentioned prior art, the maintenance management system (fault detection hardware) is simply used.
Its only purpose is to continue fault detection for equipment that is not down. However, among the computers in the distributed processing system, the computers may be stopped from a predetermined time to the time due to regular maintenance / version upgrade / main power supply work, etc. In such a situation, it is necessary to notify other computers of the scheduled stop and the scheduled restart time in advance before stopping the computer.
If the computer can be stopped and restarted in a relatively short period of time, the program running on the computer to be stopped may be temporarily stopped and restarted after the computer is restarted. It is desirable to make it appear to other computers as if not, but such a situation cannot be dealt with by the conventional technology.

An object of the present invention is to estimate a computer's scheduled stop time /
It is to manage the scheduled restart time so that one computer can be stopped / restarted so as not to affect other computers as much as possible, and especially, it is a relatively short predetermined period from the stop to the restart of the computer. In the case where it is completed, the other computer should be able to appear as if the computer is not stopped.

[0006]

In order to achieve the above object, the present invention has a management program for managing scheduled stop times and restart scheduled times of each computer in a distributed processing system. In addition, the management program includes management of program startup in the distributed processing system, management of suspension / restart of programs in the distributed processing system, management of communication between programs in the distributed processing system, computers in the distributed processing system. At least one management such as management of processing performed between the two is performed. The management program has a schedule management table,
The scheduled stop time and the scheduled restart time of each computer are stored.

In order to achieve the above-mentioned object, the present invention has, as another method, a computer management program for managing scheduled stop times and scheduled restart times of each computer on all computers in the distributed processing system. Further, the computer management program manages the start-up of programs in the distributed processing system and the temporary stop / restart of programs in the distributed processing system. The computer management program has a schedule management table and stores the scheduled stop time and the scheduled restart time of each computer.

[0008]

According to the present invention, the management program manages the start-up of the program in the distributed processing system, does not start the program on the computer whose scheduled stop time is approaching, and the scheduled restart time is The start of the program operable by the approaching computer is delayed until the scheduled restart time. The management program also manages suspension / restart of programs in the distributed processing system,
The program running on each computer is temporarily stopped before the scheduled stop time, and the program is restarted after the scheduled restart time. Furthermore, the management program manages communication between programs in the distributed processing system, and delays communication with a program running on a stopped computer until after the scheduled restart time of the computer. The management program is
Furthermore, the processing performed between the computers in the distributed processing system is managed, and the processing for the stopped computer, the program on the computer, or the operating system is delayed until after the scheduled restart time of the computer.

Further, the computer management program manages the start-up of the program in the distributed processing system, does not start the program on the computer whose scheduled stop time is approaching, and does not start the program on the computer whose scheduled stop time is approaching. The activation of the operable program is delayed until the scheduled restart time. The computer management program also manages suspension / restart of the program on the computer, suspends the program running on the computer before the scheduled stop time, and restarts the program after the scheduled restart time. Start up.

[0010] With this, when the computer is stopped and restarted in a relatively short predetermined period, the computer can be made to appear to other computers as if they were not stopped.

[0011]

An embodiment of the present invention will be described with reference to the drawings.
FIG. 13 shows an outline of a distributed processing system which is an embodiment of the present invention. A plurality of computers are connected to a network such as a communication bus, and programs on each computer communicate with each other to operate. Reference numerals 111 to 115 are computers, and a plurality of programs operate in cooperation on the plurality of computers. Although a system configuration in which five computers are connected is shown in the figure, it is possible to construct a system with more or less computers. Reference numeral 12 is a network such as a communication bus, an optical fiber, and a crossbar communication network. The programs on each computer communicate with each other via the network 12. Reference numerals 131 to 137 are user programs, which execute processing independently or in cooperation with each other. Reference numeral 14 denotes a management program, which has a role of centrally managing stop / restart of computers in the distributed processing system. Further, the management program 14 manages program activation and communication in the distributed processing system. Therefore, when each user program wants to perform processing such as program activation and communication, it does not directly perform the processing but requests the management program for the processing. The management program acts on behalf of these processes.

FIG. 1 shows a block diagram of a distributed processing system in which the inside of a computer is described in more detail. In FIG. 1, computer 11
1-115, network 12, user program 13
1 to 137 and the management program 14 are the same as those in FIG. Each computer is composed of a central processing unit 15, a main storage unit 16, an input / output control unit 17, etc., and exchanges data using an internal bus 18. The input / output control device 17 is usually
It is connected to an external storage device 171, such as a disk, and a terminal 172, and exchanges data with the outside. Main memory 16
User programs 131 to 137 and a management program 14 are present on the top and are processed by the central processing unit 15. Generally, an operating system 161 is present to manage each program on the computer.

The types of processing performed by the management program 14 that manages the distributed processing system include at least the following four types. (1) Managing scheduled stop time / restart scheduled time of each computer (including the computer in which it exists): It manages by having the schedule management table 21 shown in FIG. (2) Manage program start / end on each computer: Start / end the program on each computer according to the user's or program's request. At this time, the computer whose scheduled stop time is approaching does not start the program. Further, if there is a computer near the scheduled restart time, the program is started after waiting for the scheduled time. (3) Managing program suspension / restart on each computer: When there is a computer that is scheduled to be stopped, the execution status of the program running on that computer is checked before the scheduled stop time. The program is restarted by using the execution status stored after being stored in the external storage device of the computer and temporarily stopped, and after the scheduled restart time. (4) Managing the communication to the program on each computer: Delay the communication to the program existing in the stopped computer until the scheduled restart time of the computer.

FIG. 2 shows an embodiment of a data structure (schedule management table 21) for the management program 14 to manage suspension / resumption of computers in the distributed processing system. In this example, when the suspension / restart is performed at a predetermined time such as maintenance, version upgrade, backup, etc., the next scheduled stop time / restart scheduled time of each computer is stored in the schedule management table. In the schedule management table, the column 211 includes all computers existing in the distributed processing system. Column 21
3, 214 indicates the next scheduled stop time and restart scheduled time of each computer. The stop / restart time is not necessarily set in advance for all computers, and if the scheduled time is indeterminate, nothing is stored.

The scheduled time stored here is not exactly the time when the computer is stopped or the time when the computer is restarted. As described above, before the computer is stopped, the process of suspending the program running on the computer must be performed. In addition, it is necessary to consider the time difference between the computers in the distributed processing system. The time required for the suspension process is T
s, the maximum estimated time difference between computers is Td, the following value T ′ calculated from the accurate scheduled stop time T
Is stored in the column 213. T ′ = T−Ts−Td In the case of the scheduled restart time, only the time difference between the computers in the distributed processing system becomes a problem. Therefore, the following value T ′ is stored in the column 214 with respect to the accurate scheduled restart time T. T ′ = T + Td The scheduled time is set by the administrator of the entire system directly entering the management program 14 each time the scheduled time is changed, or the individual computer (administrator) manages it via the network 12. Notify program 14 and manage program 1
4 is used for updating. The column 212 is data indicating the state of whether each computer is stopped or operating. If the state of the computer is in operation, the computer is next stopped at the time shown in the scheduled stop time column 213 and then restarted at the time shown in the scheduled restart time column 214. If the computer is in the stopped state, the computer is restarted at the next time indicated in the scheduled restart time column 214 and then stopped at the time indicated in the scheduled stop time column 213.

The management program 14 further has a program table 22 in the schedule management table 21 in order to manage the start / end of the program on each computer. Each entry in the program table has a list of programs running on the corresponding computer (computer-specific program list). In the case of the program allocation shown in FIG. 13, the computer-specific program list structure shown in FIG. 2 is obtained. Each entry in the computer program list is the program name 2
21 and state 222. A state 222 indicates whether each program is running or is temporarily stopped. When the scheduled stop time of the computer approaches, the execution status of the program running on the computer is stored in the external storage device and temporarily suspended. At this time, the state in the corresponding entry is changed to the suspended state. . Conversely, when the computer is restarted, the execution status of the program in the suspended state is read from the external storage device and restarted. At this time, the state in the corresponding entry is changed to the operating state.

Furthermore, the management program 14 waits for processing in order to delay the start-up of a program on a computer whose scheduled stop time is approaching or the computer that is stopped and the communication to the program on the computer. Matrix 2
Have three. The process queue 23 is a process for a computer whose scheduled stop time is approaching or for a computer that has stopped, and which is delayed until the scheduled restart time of the computer is a queue (computer-specific process queue). This is a list to register. Each entry 231 of the computer-specific processing queue is classified into a communication mediating process and a program starting process. In case of mediation processing of communication (Fig. 2
In the middle message), the entry indicates the program of the communication partner and the communication content. In the case of program startup processing (Fig. 2
Middle exec), the entry indicates the program to be started.

Reference numeral 24 is data indicating the current time. In FIG. 2, the current time 24 is placed under the control of the management program 14, but it may be placed under the control of the operating system and inquired at a necessary time. In the example of FIG.
The current time is close to the scheduled stop time of the computer 114,
Further, the scheduled restart times of the computers 112 and 114 are also close to each other, and therefore the processing performed on these computers, that is, the verify message communication to the program 132, the activation of the program 138, the program 13
Processing such as print message communication to 5 and activation of the program 139 is stored in the processing queue and delayed until the scheduled restart time.

A process flow 31 of the management program 14 for performing these processes is shown in FIG. However, this processing flow cannot manage stop / restart of the computer in which the management program itself is operating. The stop / restart of the computer on which the management program is operating will be introduced in the processing flows of FIGS. 7, 8 and 9. The management program has a flow of being released from the waiting state due to the occurrence of an event (process 314) and transitioning to the waiting state again after the process is completed (process 313). The event to release the waiting state is (a) a scheduled time change notification from another computer,
(B) Scheduled time input from the administrator of the entire system,
(C) The arrival of the nearest scheduled time, and (d) notification of a message from the user program. These are determined in processing 315, and the processing mode is entered. Less than,
Details of each of these event processes will be described with reference to FIGS. 4, 5 and 6. Note that the management program prepares a variable Tn representing time. The earliest scheduled time in the schedule management table 21 is selected (process 311), and Tn is selected.
The selected scheduled time is substituted in (step 312). The arrival of time Tn means the occurrence of the event (c).

(A) Scheduled time notification from another computer (FIG. 4): According to the notified scheduled time, the data in the schedule management table 21 of the corresponding computer is changed (process 3).
21). As a result, the closest time among the scheduled stop / restart times of each computer may change. For this reason,
The time Tn stored as the closest time among the scheduled times until now is compared with the changed scheduled time (Process 32).
2). If the stored time Tn is later than the changed scheduled time, the changed scheduled time is newly stored as the closest time Tn (process 323). If the current time is designated as the scheduled stop time, the event (c) will subsequently occur, and it can be notified that the computer will stop immediately.

(B) Scheduled time input from the administrator of the entire system (FIG. 4): The requested scheduled time in the schedule management table 21 is changed (process 331).

As in the case of (a), this may change the closest time among the scheduled stop / restart times of each computer. However, since multiple scheduled times may be changed, the time Tn that has been stored as the closest time among the scheduled times until now, and the latest scheduled time from the current time among the changed multiple scheduled times. A close one is selected (process 332). This scheduled time is newly stored as the closest time Tn (process 333).

(C) Arrival of the closest time among scheduled times (FIG. 5): Time T stored in the above (a) and (b).
When n arrives, the management program is released from the event waiting state. First, it is investigated whether this is the scheduled stop time or the scheduled restart time of the computer (process 341). At the scheduled stop time, the execution status of the program running on the computer is stored in the external storage device of the computer (process 34
2) Pause (process 343) and schedule management table 2
It is noted in 1 that the computer has stopped (process 344). At the scheduled restart time, the execution status of the program suspended in the computer is restored (process 345) and restarted (process 346), and the schedule management table indicates that the computer has restarted (process 347). Further, the program activation process and the communication mediation process registered in the process queue 23 used to delay the process until the scheduled restart time are performed (process 348). Then, the arrived scheduled time Tn is deleted from the schedule management table (process 349), and the scheduled time closest to the current time in the schedule management table is selected (process 350). Finally, the selected time is set to Tn (process 351), and the next event occurrence time is set.

(D) Message notification from user program (FIG. 6): management program starts / starts program
In order to manage termination and communication, a program start / termination request and a communication request to another program are issued to the management program. When these request messages arrive, the type of message is determined and the process is executed according to the type (process 361). In this embodiment, the types of messages to be received are (α) program start request, (β) communication request,
(Γ) Other three types. Hereinafter, these processes will be described.

(Α) In the case of a program activation request, first, it is checked whether the computer to be activated is stopped or the scheduled stop time is approaching (process 362). If this condition is not satisfied, the program may be started as usual (process 363). If the condition checked in the process 362 is satisfied, the time to the scheduled restart time is checked (process 364). If this is short, a program start request is registered in the process queue and the process is delayed until the scheduled restart time ( Process 3
65). If the time up to the scheduled restart time is long, an error is returned to the requester user or the program to notify that the program cannot be started until the scheduled restart time (step 366).

(Β) Even when the message is a communication request,
Similarly, it is checked whether the communication partner computer is stopped or the scheduled stop time is approaching (process 367).
If this condition is not established, communication is mediated as usual (process 368). When the condition checked in the process 367 is satisfied, the communication request is registered in the process queue and the process is delayed until the scheduled restart time (process 369). Also in this case, a method of checking the length of time until the scheduled restart time and notifying an error based on this may be adopted.

If the (γ) message has other contents, the process is performed regardless of the state of the schedule management table 21 (process 370).

Next, with reference to FIGS. 7, 8 and 9, a method for enabling the suspension / resumption of the computer on which the management program is operating will be described. 7 and 8 show a processing flow 31 of the management program for enabling suspension / resumption of the computer on which the management program is operating. 7 and 8, the processing 311 to 315, the processing 321 to 323, the processing 331 to 333, the processing 341 to 351 and the processing 361 to 370 are the same as those in FIG. 3, FIG. 4, FIG. 5 and FIG. Pause process 41, end process 42, restart process 43
Has been added. Pause process 41 and end process 4
2 is the closest time Tn in the scheduled time, and
If the time is the scheduled stop time of the computer on which the management program is operating, it is performed. This condition determination is performed in process 411.

At this time, the execution status of the program operating on the computer is stored in the external storage device of the computer (step 412) and temporarily stopped (step 413). Thereafter, the arrived scheduled time Tn is deleted from the schedule management table 21 (process 414), and the scheduled time closest to the current time in the schedule management table is selected (process 415). Next, it is determined whether or not the selected time is the scheduled restart time of the computer on which the management program 14 is operating (process 41).
6). If this condition is satisfied, there is no need to perform processing using the schedule management table until the scheduled restart time, so the management program itself is suspended. After notifying each program in the distributed processing system of the suspending and resuming scheduled times of the computer on which the management program is operating (process 417), the execution status of the management program (including the contents of the schedule management table) is stored in the external storage device. Stored in (process 41
8), it suspends itself (process 419). If the condition of process 416 is not satisfied, the process of the management program is delegated to another computer. First, it is noted in the schedule management table that the computer on which it is operating is stopped (process 42).
1). Next, the schedule management table is transferred to another computer and the management program is activated on that computer (process 42).
2), forcibly terminates itself (process 423).

The management program 14 in the suspended state executes the restart process 43 at the same time as the computer is restarted. For this, it is sufficient to restore the execution status of the schedule management table 21 and the like stored in the external storage device (process 431) and thereafter shift to the same process as the normal start-up process (process 311 to 311).
5). The restart of the user program stopped at the same time as the temporary stop of the management program is performed in the processes 345 to 346.

In order to enable the suspension / resumption of the management program 14, the user program must also be modified so that it can handle this. When the user program notifies the management program of a message, Notification is made in the processing flow 51 as shown in FIG. First, it is investigated whether or not the computer on which the management program is operating is stopped (process 511). When it is determined that the process is stopped, the process waits until the scheduled restart time notified in process 417 (process 512), and then notifies the management program of a message (process 513).

Next, FIG. 14 shows a block diagram for distributedly managing stop / restart of computers in the distributed processing system. In the figure, computers 111 to 115, a network 12,
The programs 131 to 137 are the same as those in FIG. A computer management program 64 exists on each computer and manages suspension / resumption of individual computers. Furthermore, the computer management programs communicate with each other and cooperate with each other to manage the suspension / resumption of each computer in the distributed processing system. further,
The computer management program manages program activation and communication in the distributed processing system. For this reason, when each user program wants to perform processing such as program startup and communication, it does not directly perform the processing but requests the computer management program on the same computer to perform the processing. The computer management program acts as a proxy for these processes. Each computer management program has the role of the management program 14 as a whole.

FIG. 10 shows a block diagram of a distributed processing system in which the inside of the computer is described in more detail. 10, the computers 111 to 115, the network 12, the user programs 131 to 137, and the computer management program 64 are the same as those in FIG. Central processing unit 15, main storage device 16, input / output control device 17, internal bus 18, external storage device 171,
Terminal 172 and operating system 161
Is the same as in FIG.

Next, FIG. 11 shows an example of implementation of a data structure (schedule management table 71) for each computer management program 64 to manage suspension / resumption of computers. FIG. 11 shows a schedule management table on the computer 111. In FIG.
Computer name column 211, status column 212, scheduled stop time column 2
13, the scheduled restart time column 214, the processing queue 23 (entry 231), and the current time column 24 are the same as the schedule management table 21 for centralized management shown in FIG. A change for managing suspension / resumption of computers in a distributed manner is a program table 22 having information on programs on all computers.
Is a program table 72 having information on only the programs on the managed computer. Calculator 11
Since the programs 131 and 132 exist in 1, the program list is composed of only two entries. In addition, the operation of the program in each entry 721 /
The status column indicating suspension is not required.

Process flow 8 of computer management program 64
1 is shown in FIG. Similar to the management program 14, the computer management program has a flow of being released from the waiting state due to the occurrence of an event (process 812), and transitioning to the waiting state again after the process is completed (process 811). In addition,
When managing suspension / resumption of computers in a distributed manner, the suspension / resumption of other computers is not an event. Therefore, stop / restart of another computer may occur between the event occurrence of a computer and the event occurrence. Therefore, when an event occurs, the state of the computer is changed from the current time and the scheduled stop time / restart scheduled time of each of the other computers (process 813). The state is determined as follows.

(1) Conditions for setting the computer status to operating: current time <scheduled stop time <scheduled restart time, scheduled restart time <current time <scheduled stop time, scheduled stop time <scheduled restart time <current time ( 2) Conditions for setting the computer status to stop: current time <
Scheduled restart time <Scheduled stop time, Scheduled stop time <Current time <
Scheduled restart time, Scheduled restart time <Scheduled stop time <Current time If the scheduled stop time / scheduled restart time is not set, it is assumed that the time after an infinite time is set, and the context of the above time is calculated. To do. Further, the scheduled time before the current time is deleted from the schedule management table 71.

The event for releasing the waiting state is (a) a notice of change of scheduled time from another computer, (b) input of scheduled time from the administrator of the own computer, (c) arrival of the stop time of the own computer, (D) It is assumed that the computer has a non-empty processing queue, the scheduled restart time has arrived, and (e) a message is notified from the user program. These are determined in process 814,
Shift to each processing mode. Hereinafter, FIG. 12 will be described for each of these event processes.

(A) Notification of scheduled time from another computer: According to the notified scheduled time, the data in the schedule management table 71 of the corresponding computer is changed (process 821).

(B) Scheduled time input from the administrator of his computer: The requested scheduled time in the schedule management table 71 is changed (process 831). Further, the changed scheduled time is notified to the computer management program 64 of another computer by broadcasting it via the network 12 (process 832).

(C) Arrival of stop time of own computer: When the scheduled stop time of the own computer stored in (b) above comes, the computer management program 64 is released from the event waiting state. At this time, the temporary stop processing 84 of the computer management program is performed. First, the execution status of a program running on the computer is stored in an external storage device of the computer (process 841) and temporarily stopped (process 8).
42). Thereafter, the scheduled stop time that has arrived is deleted from the schedule management table 71 (process 843). Finally, the execution status of the computer management program (including the contents of the schedule management table) is stored in the external storage device (process 844), and the computer temporarily stops (process 845). The computer management program 64, which is suspended, performs a restart process 86 at the same time as the computer restarts. First, the execution status of the schedule management table 71 and the like stored in the external storage device is restored (process 861), and the scheduled restart time of the computer is deleted from the schedule management table (process 862). After that, the execution status of the user program stopped at the same time as the temporary stop of the computer management program is restored (process 863) and restarted (process 864). At this point, the restart process ends and the process shifts to the next event waiting state (process 811).

(D) The processing queue is not empty The arrival of the scheduled restart time of a computer: When the processing queue for a computer is not empty, it means that there is a program waiting for the scheduled restart time of the computer. Therefore, when this event occurs, the program activation process and the communication mediation process registered in the process queue corresponding to the computer are performed (process 851).

(E) Message notification from user program: The same processing as the centralized processing is performed (processing 361 to 361).
70).

[0043]

As described above, according to the present invention, when a computer is stopped and restarted in a relatively short predetermined period, it is possible to show the computer to another computer as if it was not stopped. Yes, stop one computer /
As long as the restart is possible, it is possible to avoid affecting other computers.

[Brief description of drawings]

FIG. 1 is an overall configuration diagram of a distributed processing system that is a first embodiment of the present invention.

FIG. 2 is a diagram showing a configuration of a schedule management table according to the first embodiment of the present invention.

FIG. 3 is an overall view of a processing flow of a management program according to the first embodiment of the present invention.

FIG. 4 is a partial detailed view of FIG.

5 is a partial detailed view of FIG. 3. FIG.

FIG. 6 is a partial detailed view of FIG.

FIG. 7 is a diagram showing a processing flow of a management program including temporary stop / restart.

FIG. 8 is a diagram showing a processing flow of a management program including suspension / restart.

FIG. 9 is a diagram showing a message notification processing flow to a management program including temporary stop / restart.

FIG. 10 is an overall configuration diagram of a distributed processing system that is a second embodiment of the present invention.

FIG. 11 is a diagram showing a configuration of a schedule management table according to a second embodiment of the present invention.

FIG. 12 is an overall processing flow of a computer management program according to the second embodiment of the present invention.

FIG. 13 is a schematic diagram of a distributed processing system that is a first embodiment of the present invention.

FIG. 14 is a schematic diagram of a distributed processing system that is a second embodiment of the present invention.

[Explanation of symbols]

111-115 ... Calculator, 12 ... Network, 131
~ 137 ... Program, 14 ... Management program, 15 ...
Central processing unit, 16 ... Main storage device, 161 ... Operating system, 17 ... Input / output control device, 171 ... External storage device, 172 ... Terminal, 18 ... Internal bus, 21, 71
... schedule management table, 22, 72 ... program table, 23 ... processing queue, 31 ... management program processing flow, 64 ... computer management program, 81 ... computer management program processing flow.

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Satoru Murata 7-1-1, Omika-cho, Hitachi-shi, Ibaraki Hitachi Ltd. Hitachi Research Laboratory (72) Inventor Tomoaki Nakamura 5-2, Omika-cho, Hitachi-shi, Ibaraki No. 1 Stock company Hitachi Ltd. Omika factory

Claims (10)

[Claims] 1. In a distributed processing system in which a plurality of computers are connected to a network, a scheduled stop time and a scheduled restart time of each computer in the distributed processing system are managed, and a program in the distributed processing system is managed. A function that manages start-up and prevents the program from starting on a computer whose scheduled stop time is approaching, and a function that delays the start of a program that can run on a computer whose scheduled restart time is approaching until the scheduled restart time. Among them, a distributed processing system comprising at least one management program having one or more functions. 2. In a distributed processing system in which a plurality of computers are connected to a network, a scheduled stop time and a scheduled restart time of each computer in the distributed processing system are managed, and a program in the distributed processing system is managed. It has a function of managing the start-up, notifying each computer of the next scheduled restart time before the computer is stopped, and delaying the startup of the program operable on the computer whose scheduled restart time is approaching until the scheduled restart time. A distributed processing system comprising at least one management program. 3. In a distributed processing system in which a plurality of computers are connected to a network, a scheduled stop time and a scheduled restart time of each computer in the distributed processing system are managed, and a program in the distributed processing system is managed. At least one management program having a function of managing suspension / restart, suspending the program running on each computer before the scheduled stop time, and restarting the program after the scheduled restart time is provided. A distributed processing system characterized by the above. 4. In a distributed processing system in which a plurality of computers are connected to a network, a scheduled stop time and a scheduled restart time of each computer in the distributed processing system are managed, and a program in the distributed processing system is managed. Manages suspension / restart, notifies the next resuming scheduled time from each computer before the computer is stopped, and suspends the program running on the computer when the computer is notified of stop, A distributed processing system comprising at least one management program having a function of restarting the program after a scheduled restart time. 5. The distributed processing system according to claim 3 or 4, wherein the management program manages communication between programs in the distributed processing system, and communicates with a program running on a stopped computer. A distributed processing system having a function of delaying until a computer is scheduled to restart. 6. In a distributed processing system in which a plurality of computers are connected to a network, a scheduled stop time and a scheduled restart time of each computer in the distributed processing system are managed, and between the computers in the distributed processing system. And a management program having a function of delaying processing of a computer that is stopped or a program on the computer, and processing for the operating system until after the scheduled restart time of the computer. Distributed processing system. 7. The distributed processing system according to claim 1, claim 2, claim 3, claim 4, claim 5, or claim 6, wherein the management program is before the stop of a computer in which the management program is present. Has a function of notifying all programs in the distributed processing system of the scheduled restart time, and each program in the distributed processing system has a function of delaying the processing to the management program until the scheduled restart time. A distributed processing system characterized by the above. 8. In a distributed processing system in which a plurality of computers are connected to a network, scheduled stop times and scheduled restart times of each computer in the distributed processing system are managed on all the computers in the distributed processing system. Function that manages the start of the program in the distributed processing system and does not start the program on the computer whose scheduled stop time is approaching, and the program that can run on the computer whose scheduled restart time is approaching A distributed processing system comprising a computer management program having any one or more of the functions of delaying the execution time to the scheduled restart time. 9. In a distributed processing system in which a plurality of computers are connected to a network, scheduled stop times and scheduled restart times of each computer in the distributed processing system are managed on all the computers in the distributed processing system. And suspend / run programs running in the distributed processing system.
A computer management program having a function of managing restart, suspending the program running on each computer before the scheduled stop time, and restarting the program after the scheduled restart time is provided. Distributed processing system. 10. In a distributed processing system in which a plurality of computers are connected to a network, scheduled stop times and scheduled restart times of each computer in the distributed processing system are managed on all the computers in the distributed processing system. In addition, the distributed processing system is provided with a computer management program having the following function a. a. The computer management program has a function of notifying the computer of the other computer in the distributed processing system of the scheduled restart time before the scheduled stop time of the computer in which the computer itself exists. The computer management program that has received the notification has a function of delaying the processing performed for the computer management program that has notified the scheduled restart time until the scheduled restart time.