JPH08185330A - Method for switching redundant computer system - Google Patents

Abstract

PURPOSE: To provide a redundant computer system switching method capable of more economically attaining system reliability equivalent to or more than conventional one. CONSTITUTION: Primary stand-by computers 102, 112, 122, 132 fixedly allocated to current computers 101, 111, 121, 131 to execute high speed backup are connected to a transmission line 160 and secondary stand-by computers 141, 142 for improving the reliability of the system are also connected to the transmission line 160 so as to be shared by all the current computers 102, 112, 122, 132 and the primary stand-by computers 101, 111, 121, 131. In order to secure the reliability of the stand-by computers, all the computers are periodically driven as current computers by a centralized computer managing device 150 at the time of normal operation so as to quickly detect abnormality, and when abnormality is generated in the current computers, the primary stand-by computers 102, 112, 122, 132 are switched to current computers to execute high speed backup and the shared secondary stand-by computers 141, 142 are allocated to the new primary stand-by computers.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a plurality of active computers, a primary spare computer which is a spare of the active computers, a small shared secondary spare computer, and a centralized computer which centrally controls switching between computers. The present invention relates to a method of switching a redundant computer system configured with a management device.

[0002]

2. Description of the Related Art Conventionally, the following method has been adopted as a backup method when a computer malfunctions. (1) n-redundant configuration: a method in which a spare computer is fixedly assigned for the active computer and the spare computer takes over the processing when the active computer fails. Also, if higher reliability is required, add the above spare computer,
This is a method of increasing redundancy. On the other hand, the normality of the spare computer has been confirmed by executing a maintenance test program at the time of regular maintenance. (2) k out of n configuration: In a network distributed system of LCMP, when k computers are required, n computers are prepared, and when a computer is abnormal, another computer backs up processing. (3) High-speed preliminary switching method of DIPS (described in NTT Research Practical Report Vol. 36, No. 8, P1021 to 1057):
The checkpoint data file is shared between the two active computers and one standby computer, and the standby computer keeps the two active computers and the programs and data operable as a load sharing system.
This is a method of backing up a checkpoint data file to a spare computer under the control of a system control processing unit that centrally manages these computers when the current computer is abnormal.

[0003]

The above-mentioned conventional techniques have the following problems. (1) In the n-duplex configuration, the spare computer is fixedly allocated for the active computer, and the status of the spare computer is updated in synchronization with the status change of the active computer, so that high-speed backup can be performed when the active computer fails. It can be realized. However, in order to secure the reliability of the double redundant configuration or more, it is necessary to add a spare (hereinafter referred to as a secondary spare) to the spare (hereinafter referred to as a primary spare). The cost was about (n-1) / n, which was uneconomical.
Further, particularly when n-1 is a spare computer, if the spare computer has already failed due to an abnormality due to aged deterioration or the like, there is a problem that a system failure occurs due to multiple failures when the working computer is abnormal. . (2) The k out of n configuration system composed of k active computers and (n−k) standby computers and the high-speed standby switching method of DIPS are shared by the backup computer when the active computer fails. It is necessary to set the state of the spare computer to the state immediately before the failure of the failed computer, but since it is shared, there is a problem that it takes time to bring it to the state before the failure. Also, since the shared spare computer is used when the active computer is abnormal, if the shared spare computer has failed due to aging, etc., multiple failures will occur in the case where the active computer fails, and switching will take time and reliability will be high. However, there was a problem in that An object of the present invention is to connect a primary spare computer, which is fixedly assigned to an active computer and realizes high-speed backup, to a transmission line, and also a secondary spare computer for achieving high system reliability as a transmission line. Connected, this secondary spare computer is shared by all working and primary spare computers, and in order to ensure reliability of the spare computer, at normal times, all computers are made to operate periodically as an active computer and abnormalities are detected. It can be detected quickly, and when the working computer is abnormal, the primary spare computer is switched to the working one for high-speed backup, and the shared secondary spare computer is assigned to this new primary spare computer. A redundant computer system switching method for realizing the above system reliability is provided. In the door.

[0004]

In order to achieve the above object, an operation mode of all computers (active, primary spare, secondary spare, abnormal), a state management table for managing node addresses (231 in FIG. 2), A switching control processing unit (240 in FIG. 2) having a timer (241 in FIG. 2) for notifying the operation mode change timing for all computers and performing switching control between computers according to the flows shown in FIGS. , A computer management unit (230 in FIG. 2) for reading / updating the status management table according to the flows shown in FIGS. 6 and 7, and an abnormal computer notified from the switching control processing unit according to the processing flow shown in FIG. 9. The node address information is notified to the network administrator, or the repair completion notification is sent from the maintenance person who repaired the abnormal computer. A computer to which an element (1) of an administrator interface control unit (250 in FIG. 2) for performing interface control with a terminal used by the network administrator for updating the computer state, etc. is added to the central computer management device (FIG. 1).
150), and a. An address table for storing the node address of the computer and the node address of the computer which is in a current or primary spare relationship with the computer (43 in FIG. 10).
1), b. A switching processing unit (45 in FIG. 10) that performs switching processing between computers according to the flows shown in FIGS.
0), c. The first timer (441 in FIG. 10) and the second timer (FIG. 1)
No. 442 of FIG. 0) and monitors the abnormality of the active computer according to the flow shown in FIG. 18 (4 in FIG. 10).
40), d. An address management unit (430 in FIG. 10) that updates or reads the address table in accordance with instructions from the monitoring processing unit and the switching processing unit and the centralized computer management device according to the flow shown in FIG. 16, e. Elements a to e of the hot standby processing unit (460 in FIG. 10) that operates according to the operation flow shown in FIG. 17 are added to the working primary backup computer and secondary backup computer, and all of these computers are connected in the transmission path. Connecting.

[0005]

In the present invention, by adding the above means, the centralized computer management device manages all the computers, and these computers are sequentially operated at regular time intervals.
The holding information is transmitted from the next spare computer to the second spare computer, and the operation mode is changed so that all the computers operate as the current computer. When the primary spare computer detects an abnormality in the active computer, the primary spare computer immediately becomes the active computer for high-speed backup and notifies the central computer management device of the detected abnormality of the computer. The centralized computer management device notified of the abnormality of the computer interrupts the periodical operation mode change control, assigns an arbitrary secondary spare computer as the primary spare of the computer which is newly in use and notified of the abnormality, and newly allocates it. By transferring the retained information from the currently used computer to the newly assigned primary spare computer to become the primary spare computer, rapid detection and switching can be performed with an economical configuration.

[0006]

An embodiment of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram of a computer system according to an embodiment of the present invention, which comprises a plurality of computers, a central computer management device 150 and a transmission line 160.
101, 111, 121 and 131 are active computers, and 102, 112, 122 and 132 are primary spare computers, 101 and 102, 111 and 11 respectively.
Reference numerals 2, 121 and 122, 131 and 132 indicate the correspondence between the active computer and the primary spare computer. 141
And 142 are shared secondary spare computers. In addition,
Dotted lines indicate additional parts within each computer.

FIG. 2 shows an example of the configuration of the centralized computer management apparatus in one embodiment of the present invention. The central computer management device includes a communication control processing unit 210 and a message processing unit 2.
20, a switching control processing unit 240, a timer 241, a computer management unit 230, a state management table 231, and an administrator interface control unit 250. The timer 241 notifies the switching control processing unit 240 of operation mode switching instructions at preset time intervals. In addition, 151 shows an additional part.

Here, a processing flow of the switching control processing unit 240 is shown in FIGS. In this embodiment, when a preliminary request message is received from the primary spare computer that has detected an abnormality in the active computer (step 3011), a response message indicating completion of reception of the abnormality notification / preliminary request message is transferred to the primary spare computer (step 3011). Thirty
12). Alternatively, when a preliminary request message is received from the active computer that has detected an abnormality in the primary spare computer (step 3013), a response message indicating the completion of the abnormality notification / preliminary request message reception is transmitted to the active computer (step 3014). After that, the periodic switching timer is stopped to interrupt the periodic switching (step 3015). Next, the node address of the working computer in which the abnormality is detected and the node address of its primary spare computer are extracted from the received message (step 3
016), the node address of the abnormal computer is notified to the administrator interface control unit (step 3017), and the node address of the abnormal computer and a request for allocating a secondary spare computer that is newly a primary spare candidate are sent to the computer management unit. Notify (step 301)
8). On the other hand, when the node address of the secondary spare computer that is a primary spare candidate is received from the computer management unit (step 3021), the secondary spare computer
The node address notification message of the newly used computer is transferred (step 3022).

When a node address setting completion message is received from the secondary spare computer to which the node address notification message has been transferred (step 303)
(YES at 1, 3032), the role change message including the node address of the secondary spare computer is transferred to the newly active computer (step 3033).
Alternatively (NO in step 3032), the administrator interface control unit is notified of the abnormality (step 3034), and the computer management unit is notified of the allocation request for the secondary spare computer of the new primary spare candidate (step 303).
5). On the other hand, when the role change processing completion report message is received from the active computer to which the role change message is transferred (YES in steps 3041 and 3042),
A request for updating the operation mode of the secondary spare computer whose role has been changed in the state management table to the primary spare in the state management table is transmitted to the computer management unit (step 3043), and the timer is set for periodic switching (step 3043). Step 30
44). When the role change process completion report message is received (NO in steps 3041 and 3042), the administrator interface control unit is notified of the abnormality (step 3).
034) and notifies the computer management unit of a request for allocation of a new secondary spare computer as a primary spare candidate (step 3035). When the interruption of the periodic switching timer occurs (step 3051), the node address of the switching source and the switching destination is requested to the computer management unit (step 3052). When the node addresses of the switching source and the switching destination are received from the computer management unit (step 3061), the node address notification message of the switching source is transferred to the switching destination (step 30).
62).

The processing flow of the computer management section 230 is shown in FIGS. 6 and 7. In this embodiment, when the node address of the abnormal computer and the notification of the secondary spare computer search request are received from the switching control processing unit (step 5081), the operation mode of the abnormal computer is updated to abnormal, and the secondary spare computer is updated. The node address is searched (step 5082) and the switching control processing unit is notified (step 5083). On the other hand, when a request for the node addresses of the switching source and switching destination computers is received from the switching control processing unit (step 5).
091), the relationship between the switching source and the switching destination is set as the primary spare and the secondary spare, and the working and the primary spare are sequentially searched for respective node addresses from the state management table (step 5).
092), and notifies the switching control processing unit (step 5).
093). When receiving a request for updating the state management table from the switching control processing unit (step 5101), the update is executed (step 5102) and the execution result is notified to the administrator interface control unit (step 5103).
Further, when the update / search notification of the status management table is received from the administrator interface control unit (step 511).
1), update / search is executed (step 5112), and the execution result is notified to the administrator interface control unit (step 5113).

The configuration example of the state management table 231 is as shown in FIG. 8, and the node address column 301 holding the node address information of each computer, the operation mode column 302, the active / primary spare node address column 303
Consists of Further, FIG. 9 shows a processing flow of the administrator interface control unit 250. In this embodiment, when the display information is received from the computer management unit or the switching control processing unit, the notification information is edited and displayed (step 700).
1, 7002). Also, it receives the input information and notifies the computer management unit of various requests (step 701).
1).

FIG. 10 is a block diagram of a computer according to an embodiment of the present invention. In this embodiment, a computer including a message transfer unit 420, a communication control processing unit 410, etc. is provided with an address management unit 430 and an address table 43.
1, a monitoring processing unit 440, a first timer 441, a second timer 442, a switching processing unit 450, and a hot standby processing unit 460 are added. In addition, 170 shows an additional part.
The processing flow of this switching processing unit 450 is shown in FIGS.
5 shows. In this embodiment, as shown in FIGS. 11 and 12, when a report of abnormality detection of the active computer is received from the monitoring processing unit (step 9011), the respective units 430, 440, 4 are transmitted.
The operation modes of 50 and 460 are updated to the current one (step 9
012), transmits an abnormality notification / preliminary request message to the central computer management device (steps 9013, 901).
4). On the other hand, the abnormality notification from the central computer management device
When a response message for receiving the preliminary request message is received (step 9041), the process normally ends (step 90).
42, YES), and transfers the processing completion message to the central computer management device (NO in step 9042, 90).
36). When the role change message is received from the central computer management device (step 9021), the retained information is read out and transmitted as an information transfer message to the computer of the primary preliminary candidate (step 9022). When the information transfer completion report message is received from the secondary spare computer (YE in steps 9031 and 9032).
S), the operation mode of the computer is checked (step 9033), and if it is a primary spare, each unit 430, 44
The operation mode of 0, 450, 460 is updated to the secondary spare (step 9034), and if it is in use, 1 is sent to the address management unit.
The node address setting request of the next spare computer is notified (step 9035), and the processing completion message is transferred to the central computer management device (step 9036).

As shown in FIG. 13, when the retained information transfer message of the computer is received from the switching source (step 9051), the transfer source node address of the message is compared with the switching source node address (step 9052). If they do not match, a transfer source node address mismatch information transfer completion report message is transmitted to the transfer source (step 9053). If they match, the received information is stored (step 9054), and each unit 430, 44 is stored.
The operation modes of 0, 450, and 460 are updated to primary spare (step 9055), and an information transfer completion report message is transmitted to the transfer source (step 9056).

In the case of performing periodic switching as shown in FIG. 14, when a role change message is received from the central computer management unit (step 9061),
The operation mode of the corresponding computer is checked (step 9062), and if it is the primary spare, the retained information is read,
It is transmitted to the switching destination as an information transfer message (step 9063). If it is in use, an information transfer message including an instruction to switch to the active mode is transmitted to the primary spare computer (step 9064). On the other hand, when the information transfer completion report message is received from the switching destination (YES in steps 9071 and 9072), the operation mode of the transfer source is checked (step 9073), and if it is in use, the operation modes of the respective units 430, 440, 450 and 460 are checked. Update to the primary spare (step 9074), if it is the primary spare, update the operation mode of each unit 430, 440, 450, 460 to the secondary spare (step 9075), and transfer the processing completion message to the central computer management device. (Step 9076).

Further, in the case of performing the periodic switching, as shown in FIG. 15, when the retained information transfer message of the computer is received from the switching source (step 90).
81), and compares the transfer source node address of the message with the switching source node address (step 908).
2) If they do not match, the information transfer completion message of the transfer source node address mismatch is transmitted to the transfer source (step 9).
083). If they match, it is further checked whether or not only mode switching is instructed (step 9084), and if so, the operation mode of each unit 430, 440, 450, 460 is updated to the current one (step 9085). If not, the received information is stored (step 9086), the operation mode of each unit 430, 440, 450, 460 is updated to primary reserve (step 9087), and the information transfer completion report message is transmitted to the transfer source. Yes (step 908)
8).

The processing flow of the address management unit 430 is shown in FIG. In this embodiment, when the working computer node address notification message is received from the central computer management device, the working node address is set in the address table (step 1301). FIG. 17 is a computer address table 4 according to an embodiment of the present invention.
It is a figure which shows the structure of 31. The address management unit 430
As shown in FIG. 17, two node addresses are set according to the operation mode. The node address is set to the node address of the computer when the computer is in the active mode, and is set to the node address of the active computer corresponding to the computer when the computer is in the primary spare mode. In the case of the secondary spare and abnormal mode, only the node address of the computer is set.

FIG. 18 shows a processing flow of the hot standby processing unit 460. In the present embodiment, when updating the information held in the active computer, the update information accumulated in the memory or the like is transmitted to the primary spare computer as an information message (step 1401). If an information transfer completion report message is received from the primary spare computer and there is an abnormality (NO in step 1421), an abnormality notification / preliminary request message is transmitted to the central computer management device (step 1422). When the information transfer message is received from the active computer, the information stored in the memory or the like is updated (step 1411) and the information transfer message / completion report message is transmitted to the active computer (step 1412).

The processing flow of the monitoring processing unit 440 is shown in FIG. In this embodiment, when a timeout is received from the second timer or an abnormality is detected by receiving a monitoring response message from the active computer corresponding to the primary spare computer, the switching processing unit is notified of the abnormality ( Step 1501). If the response is normal, the second timer is stopped (step 1512). When the interrupt from the first timer is received, the monitoring message from the active computer corresponding to the primary spare computer is transferred (step 1521), and the second timer for the monitoring message is set (step 152).
2).

Next, an embodiment of communication between computers of the present invention will be described with reference to FIGS. 20 and 21
FIG. 6 is a flowchart showing a concept of switching when a computer is abnormal in one embodiment of the present invention. In this embodiment, in the centralized computer management device, when an abnormality report is received from each computer, the periodical switching operation is stopped (step 1701), and the computer operation state management means (230, 231 etc. in FIG. 2) causes the secondary operation. The node address of the spare computer is selected, and the active computer is instructed to take a standby dual configuration with the secondary spare (step 1702). In response to this instruction, the working / primary spare computer transmits its own holding information and an instruction to change to the primary spare mode to the designated secondary spare computer (step 1711). When the completion report is received from the destination (step 1712), it is transferred to the central computer management device (step 1713). The central computer management device receives the completion report from the active / primary spare computer (step 170).
3) The timer for periodic switching is restarted (step 1704). Upon receipt of the operation mode change instruction, the secondary spare computer stores the holding information transmitted at the same time (step 1721), updates the operation mode of its own machine to the primary spare (step 1722), and holds the holding information. A completion report is transmitted to the information transmission source (step 1723).

Next, the case of the active computer abnormality will be described in more detail. FIG. 22 is a diagram showing communication between computers (case of abnormal computer in use) according to an embodiment of the present invention. In this embodiment, when the primary spare computer 112 detects an abnormality in the active computer based on the monitoring message reception time interval and the content of the reception monitoring response message (a), the operation mode of the primary spare computer is changed to the active mode (b). Then, the abnormality notification / preliminary request message is transferred to the central computer management device 150 (c). Upon receiving the abnormality notification / preliminary request message, the central computer management device 150 transfers (d) the reception response message to the primary spare computer 112, and interrupts the periodical switching control of the operation mode between the computers. In addition, a node address notification message that transfers the node address notification message that searches the state management table, reads the node address of the secondary spare computer 142 that backs up the computer 112, and notifies the secondary spare computer 142 of the node address of the primary spare computer 112 (E). Upon receiving this node address notification message, the secondary spare computer 142 stores the notified node address in the address table and transfers the setting completion message to the central computer management device 150 (f). Further, the central computer management device 150
Upon receiving the setting completion message, transfers a role change message for changing the role with the secondary spare computer 142 to the new active computer 112 (g). Upon receiving the role change message, the active computer 112 reads out the possession information such as the memory information / information storage / storage information of the computer and transfers it to the secondary spare computer 142 as an information transfer message (h). The secondary spare computer receiving this information transfer message is a message from the new working computer corresponding to the already received node address.
After receiving and storing in the storage location corresponding to the received information and after the processing is completed, the operation mode of the computer is set to the primary standby mode and the completion report message is sent to the working computer 112.
(I). Then, the active computer 112 that has received this completion report message updates the operation mode of the new primary spare computer to primary spare, and transfers the role change processing completion report message to the central computer management device 150 (j). . Upon receiving the completion report message, the centralized computer management device 150 changes the operation mode of the secondary spare computer 142 to the primary spare and restarts the periodical operation mode switching process. On the other hand, the abnormal working computer is used as an unused computer after restoration.

Next, the concept of periodic switching in this embodiment will be described. 23 and 24 are flowcharts showing the concept of periodical switching of computers in the embodiment of the present invention. In the present embodiment, in the centralized computer management device, after restarting the periodic switching timer (step 1901), the node addresses of the primary spare and secondary spare computers are selected from the computer operating state management means, and between them, The operation mode change is instructed (step 1902). Working / 1 which received this instruction
The next spare computer transmits its own holding information and an instruction to change the operation mode to the designated secondary spare computer (step 1911), and when it receives the completion report from the destination (step 1912), The operating state is changed to secondary standby (step 1913), and a completion report is sent to the central computer management device (step 1914).
The centralized computer management device receives the completion report when the replacement instruction transmission destination is the primary spare (step 190).
3) Next, the node address of the active computer and the secondary spare computer is selected from the computer operation state management means, and the operation mode change between them is instructed (step 1904). When the completion report is received when the replacement instruction transmission destination is the current one (step 1905), the process returns to step 1901 to repeat the series of processes. Upon receipt of the operation mode change instruction, the secondary spare computer stores the holding information transmitted from the active / primary spare computer (step 1921) and changes itself to the operating state of the holding information transmitting source. (Step 1922), the completion report is transmitted to the transmission source (step 1923).

Next, the periodic switching case will be described in more detail. FIG. 25 is a diagram showing communication between computers (periodic switching case) in the embodiment of the present invention. The central computer management device 150 searches the computer information held in the state management table 231, first searches for the primary spare computer 112 and the secondary spare computer 142, and then the secondary spare computer 142 to the primary spare computer 112. The node address notification message for node address notification of is transferred (A). Upon receiving the node address notification message, the secondary spare computer 142 stores the node address in the address table in the computer,
Set completion message is sent to central computer management device 150
Transfer to (a). The centralized computer management device 150, to which the setting completion message is transferred, transfers to the primary spare computer 112 a role change message for changing roles with the secondary spare computer 142 (c). Upon receiving the role change message, the primary spare computer 112 reads out the memory information of the computer and the stored information such as the information storage device storage information and transfers it to the secondary spare computer 142 as an information transfer message (d). The secondary spare computer receiving this information transfer message compares the node address of the received message with the node address of the primary spare computer received in the node address notification message, fetches when they match, and stores the information in the received message correspondingly. Store in place.
After the processing is completed, the operation mode of the computer is set to the primary spare computer, and the completion report message is transferred to the primary spare computer 112 (e). Then, the primary spare computer 11 to which this completion report message is transferred
2 is a central computer management apparatus 1 that changes the operation mode to the secondary standby mode and sends a completion report message of the role change processing.
Transfer to 50 (f). Upon receiving this completion report message, the centralized computer management device 150 changes the operation mode of the computer 142 on the status management table to primary reserve and the operation mode of the computer 112 to secondary reserve. Next, the operation mode switching process from the active to the secondary spare is continued, the active computer 111 is searched, the secondary spare computer 112 is searched, and the secondary spare computer 112 is notified of the node address of the active computer 111. Forward the address notification message (G). Upon receiving this node address notification message, the secondary spare computer 112 stores it in the address table in the computer and transfers the setting completion message to the central computer management device 150 (h). Upon receiving the setting completion message, the centralized computer management device 150, to which the setting completion message is transferred, transfers a role changing message for changing the role with the secondary spare computer 112 to the active computer 111 (K). . Upon receiving the role change message, the active computer 111 reads out the memory information of the computer and the retained information such as the information storage device storage information,
It is transferred to the secondary spare computer 112 as an information transfer message (K). The secondary spare computer receiving this information transfer message compares the node address of the received message with the node address of the active computer transferred in the node address notification message, fetches when they match, and stores the information in the received message at the corresponding storage location. After completion of the processing, the operation mode of the computer is set to the active mode and the completion report message is transferred to the active computer 111 (S). Then, the active computer 111 to which the completion report message is transferred changes the operation mode to the secondary standby mode and transfers the completion report message of the role change processing to the central computer management device 150 (S). Upon receiving the completion report message, the centralized computer management device 150 changes the operation mode of the new active computer 112 on the status management table to the active mode and changes the operation mode of the new secondary spare computer 111 to the secondary spare. Then, the above operation mode switching process is repeated.

[0023]

According to the present invention, the following effects can be obtained. (1) Even in a normal state, the failure of the primary spare computer and the secondary spare computer can be promptly detected by periodically switching the operation mode between the active primary spare computer and the shared secondary spare computer. Furthermore, the operation of regular maintenance becomes unnecessary, and maintenance costs can be reduced. (2) With a small number of shared secondary spare computers, when a computer malfunctions, any shared secondary spare computer can be quickly assigned, and this shared secondary spare computer can immediately form a standby duplex redundant configuration. It is possible to build a reliable and economical redundant system.

[Brief description of drawings]

FIG. 1 is a configuration diagram of a computer system according to an embodiment of the present invention.

FIG. 2 is a configuration diagram of a centralized computer management apparatus according to an embodiment of the present invention.

FIG. 3 is a part of a processing flowchart of a switching control processing unit according to an embodiment of the present invention.

FIG. 4 is a part of a processing flowchart of a switching control processing unit according to an embodiment of the present invention.

FIG. 5 is a part of a processing flowchart of a switching control processing unit according to an embodiment of the present invention.

FIG. 6 is a part of a processing flowchart of a computer management unit according to an embodiment of the present invention.

FIG. 7 is a part of a processing flowchart of a computer management unit according to an embodiment of the present invention.

FIG. 8 is a diagram showing a configuration of a state management table according to an embodiment of the present invention.

FIG. 9 is a processing flowchart of an administrator interface control unit according to an embodiment of the present invention.

FIG. 10 is a configuration diagram of a computer according to an embodiment of the present invention.

FIG. 11 is a part of a processing flowchart of a switching processing unit according to an embodiment of the present invention.

FIG. 12 is a part of a processing flowchart of a switching processing unit according to an embodiment of the present invention.

FIG. 13 is a part of a processing flowchart of a switching processing unit according to an embodiment of the present invention.

FIG. 14 is a part of a processing flowchart of a switching processing unit according to an embodiment of the present invention.

FIG. 15 is a part of a processing flowchart of a switching processing unit according to an embodiment of the present invention.

FIG. 16 is a processing flowchart of an address management unit according to an embodiment of the present invention.

FIG. 17 is a diagram showing a configuration of an address table of a computer according to an embodiment of the present invention.

FIG. 18 is a processing flowchart of a hot standby processing unit according to an embodiment of the present invention.

FIG. 19 is a processing flowchart of a monitoring processing unit according to an embodiment of the present invention.

FIG. 20 is a part of a flowchart showing a concept of switching when a computer is abnormal in one embodiment of the present invention.

FIG. 21 is a part of a flowchart showing the concept of switching when a computer is abnormal in one embodiment of the present invention.

FIG. 22 is a diagram showing communication between computers (case of abnormal computer in use) according to an embodiment of the present invention.

FIG. 23 is a part of a flowchart showing the concept of periodic switching of computers in an embodiment of the present invention.

FIG. 24 is a part of a flowchart showing the concept of periodic switching of computers in an embodiment of the present invention.

FIG. 25 is a diagram showing communication between computers (periodic switching case) according to an embodiment of the present invention.

[Explanation of symbols]

101, 111, 121, 131: active computer,
102, 112, 122, 132: primary spare computer, 141, 142: shared secondary spare computer, 15
0: central computer management device, 160: transmission line.

Claims (3)

[Claims] 1. A system comprising a plurality of computers having a plurality of operation modes, wherein a primary spare computer assigned to an active computer and a secondary spare computer shared by the active and primary spare computers are connected on a transmission line. Then, when an abnormality of the active computer is detected, the primary spare computer is switched to the active computer, and all the retained information is transferred from the active computer to the primary spare computer,
A redundant computer system switching method, characterized in that any of the secondary spare computers is assigned as a new primary spare computer. 2. All the holding information is periodically transmitted from the primary spare computer to the secondary spare computer and from the active computer to the primary spare computer to change the operation mode to the plurality of computers, 2. The redundant computer system switching method according to claim 1, wherein the computer is operated as an active computer. 3. A plurality of computers having a plurality of operation modes, and a centralized computer management device which centrally manages the operation states of the computers and changes and controls the operation modes of the computers when a computer is abnormal Connected via the above-mentioned computer, the computer is assigned an active computer in an operation mode for executing a processing request, and a primary spare computer that operates in synchronization with the active computer when the active computer is abnormal. In a system used as a secondary spare computer shared by a primary spare computer, a) when the centralized computer management device receives a notification from a timer that notifies the opportunity to periodically switch the operation mode of each computer, The process of restarting the timer and b) centralized computer The data management device selects the node addresses of the primary spare computer and the secondary spare computer from the computer operating state management means, and instructs the primary spare computer to change the operating mode between the secondary spare computer and the secondary spare computer. C) A step in which the primary spare computer, to which the operation mode change instruction is transmitted, transmits the information held by the computer and the operation mode change instruction to the primary spare computer to the designated secondary spare computer; Retained information sent from computer 2
The next computer stores the information, changes the operation mode to the primary spare computer, and sends the completion report to the primary spare computer that is the source of the retained information, and (e) receives the completion report from the secondary spare computer. The step of the primary spare computer changing the operation mode to the secondary spare and transmitting the completion report to the central computer management device; (f) the central computer management device receiving the completion report from the primary spare computer The step of receiving the notification, and g) the central computer management device selects the node address of the active computer and the secondary spare computer from the computer operating state management means, and the active computer is allowed to change the operation mode between the secondary spare computer. The step of instructing, and h) the active computer that has received the operation mode change instruction is A step of transmitting the information held by the computer and an instruction to change the operating mode to the active computer to the standby computer, and (b) the secondary standby computer to which the retained information is sent from the active computer stores the information and sets the operating mode to the active mode. Change and send the completion report to the active computer that is the sender of the retained information, and n) The active computer that received the completion report from the newly active computer changes the operation mode to secondary standby and concentrates. A step of transmitting a completion report to the computer management device; a step of receiving the notification from the timer by the central computer management device which has received the completion report from the newly secondary spare computer; When the computer error report is received, the timer is stopped, the operation mode change is interrupted, and the computer operation status management The step of selecting the node address of the secondary spare computer from the stage and instructing the active computer to generate a standby dual redundant configuration with the secondary spare computer; A step of transmitting the holding information of the computer and an instruction to change the operation mode to the working to the designated secondary spare computer, and (f) the secondary spare computer to which the holding information is sent from the working computer stores the information. The process of changing the operation mode to active and transmitting the completion report to the active computer that is the transmission source of the retained information, and y) The active computer that received the completion report from the newly active computer secondary reserves the operational mode. And sending a completion report to the central computer management device, and The centralized computer management device that has received the completion report receives the notification from the timer, and) when the backup of the abnormal computer is completed, restarting the timer and restarting the operation mode change. Redundant computer system switching method.