JP2006202220A - Program for causing computer to execute processing of confirming consistency of handover information, and method thereof - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To switch a current system to a standby system with a consistency of handover information in a configuration having the current system and a plurality of standby systems. <P>SOLUTION: A computer executable program for confirming the consistency of handover information when switching the current system to a standby system in the configuration having the current system and the plurality of standby systems, reads out standby discrimination information of the standby system to be switched to, from an external shared storage area shared among the current system and the plurality of standby systems and stores the read information in an internal storage area and confirms the consistency by coincidence between standby discrimination information stored in the internal storage area and its own discrimination information and uses handover information held in the internal storage area to execute switching so that the standby system starts operation as the current system. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to a computer-executable program and method for switching to a standby system with consistency of takeover information in a configuration having an operation system and a plurality of standby systems.

  In recent years, with the 24 hours of services such as information provision via networks and the globalization of business, the area of information systems important for companies and society is rapidly expanding. In the unlikely event that an abnormality occurs in these important information systems and the system is stopped for a long time, it may develop into a social problem, and countermeasures against factors that hinder the continuous operation of the information system are an important issue.

  Under such circumstances, an operation system that is normally used for business and a standby system that can be switched when the operation system is stopped due to maintenance or some problem is provided for a predetermined business. It has been. In such a configuration, a technique has been proposed that enables smooth switching from an operation system to a standby system as described below.

  For example, the setting information storage means stores the settings to be activated as either the operation system or the standby system, the priority order, and the standby state. When the standby system is activated, the setting information storage means is referred to Migration is performed (for example, Patent Document 1). This technique reduces the burden on the operator and enables reliable and quick start-up even when a failure occurs in a predetermined information processing device.

  Further, when switching from the operation system to the standby system occurs, the location management file of the hot standby standby system is updated to specify the connection path (for example, Patent Document 2). With this technology, it has become possible to operate a flexible hot standby standby system in three or more host computers and to perform high-speed switching when switching occurs.

  Furthermore, the standby system has a priority table in the operation system and switches to a standby system having a higher priority in the case of a failure, and the switched standby system similarly has a priority table and switches in the event of a failure (for example, Patent Document 3). This technology has made it possible to improve the reliability of a standby computer system in a hot standby system comprising a computer system for operation and standby in a loosely coupled multiprocessor system.

In addition, an active processing function that performs online data processing is associated with a standby processing function that continues the processing in the event of an abnormality, and a standby processing function that does not have an active processing function is selected when a failure occurs (for example, Patent Document 4). With this technology, when there is no standby processing function associated with the active processing function due to a failure of the computer, a standby processing function that can be newly associated without an operator's operation is selected. It is now possible to restore reliability by creating a hot standby relationship automatically by making a correspondence.
Japanese Patent No. 329883 Patent No. 2768267 Patent No. 3203935 JP-A-9-81409

  The proposed technology as described above has made it possible to smoothly switch from the operation system to the standby system. However, in a system configured to have two or more standby systems for one operational system, the priority order for switching is usually determined for a plurality of standby systems as described above, Depending on the timing of the occurrence, there may occur a case where the data of the operation system and the standby system cannot be consistent, and such a situation cannot be solved by the above-described conventional technology. For example, as shown in FIG.

  In the operation standby system configured by the operation system shown in FIG. 1 and the standby systems having the switching orders 1 and 2, the operation system and the standby system in the order 2 are activated, and the standby system in the order 1 is stopped. If it is (state 1), the operation system establishes a connection with the standby system of rank 2 and reflects the data being operated on the standby system of rank 2 according to a predetermined procedure. From this state, when the standby system of rank 1 that is switched over with priority over the standby system of rank 2 is activated (state 2), the operation system normally purges the data held by the standby system of rank 2 Then, the connection is switched to the standby system of rank 1 so that the failure occurring in the operation system can be dealt with.

  However, if the operation system goes down before the data stored in the standby system of rank 2 is purged (state 3), the system is switched to the standby system of rank 1 that is activated. The standby system of rank 1 starts to operate as an operation system (state 4). Thus, when the switching to the standby system of rank 1 is completed and the operation is continuously started, new data held in the operational system (standby system of rank 1) at this time and Thus, the old data (takeover information) that remains in the standby system of rank 2 exists inconsistently.

  In this state, if the operation system (standby system of rank 1) goes down again, the operation is switched to the stand-by system of rank 2, and the operation is resumed with old takeover information, causing a problem (state 5).

  This is because inconsistent takeover information exists in the active system.

  The present invention has been made in view of the above-described problems, and in a configuration having an operation system and a plurality of standby systems, a computer-executable program for switching to a standby system with consistency of takeover information The purpose is to provide.

  In order to solve the above-described problem, the present invention provides a takeover information that is taken over when an operation standby system configured by an operation system and a plurality of standby systems is switched from the operation system to one of the plurality of standby systems. A program for causing a computer to execute processing for checking the consistency of data stored in the external shared storage area from the external shared storage area shared by the operation system and the plurality of standby systems. The standby identification information reading procedure for reading the standby identification information for identifying the switching standby system and storing it in the internal storage area, and the match between the standby identification information stored in the internal storage area and the self-identification information A consistency confirmation procedure for confirming the consistency of the information, and the standby identification information and the self-identification information by the consistency confirmation procedure. When the consistency of the takeover information is confirmed by the match, the operation switching procedure for starting the operation as the operation system is executed using the takeover information held in the internal storage area. .

  In a computer in which such a program is installed, the standby identification information (for example, IP address) of the standby system that is the transmission destination of takeover information is stored in an external shared storage area (for example, a shared disk device). The standby system can determine whether or not the takeover information stored therein is valid information when switching to the operation system.

  Further, when the consistency of the takeover information held inside is confirmed, it is possible to perform high-speed switching to continue the operation without reading the data (database data) in the external shared storage area.

  As means for solving the above problems, the present invention operates as a takeover information consistency confirmation method for executing processing in the active standby system, or as a standby system or operating system realized by executing the program. It can also be a node device.

  According to the present invention, in an operation standby system that constitutes a plurality of standby systems, by storing the IP address of the standby system to which the takeover information is transmitted in a shared disk device, each standby system can be switched when switching to the operation system. It is possible to determine whether or not the takeover information held inside is valid information.

  Also, when the takeover information held inside is valid information, it is possible to perform high-speed switching that continues operation without reading the DB data of the shared disk device.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings.

  FIG. 2 is a diagram showing a network configuration of an operation standby system according to an embodiment of the present invention. In FIG. 2, an operation standby system 100 includes nodes 0, 1, 2, and 3 as server computers and a shared disk device 10 as a secondary medium, and is connected to each other via a network 2. Nodes 0, 1, 2, and 3 function as server computers of the relational database management system.

  The shared disk device 10 has DB data 16 and 18 used in one or more business operations, and, for example, a business A shared file 12 shared by an operation system and a standby system corresponding to the business A, and a business B The business B shared file 14 is shared by the operation system and standby system corresponding to.

  The business A shared file 12 stores node identification information for identifying a node as a standby system that holds the takeover information set by the operation system. The business B shared file 14 stores node identification information for identifying a node as a standby system that holds takeover information by the operation system. For example, an IP address is set as the node identification information.

  The takeover information is a plurality of DB pages indicating a log for recovery of the DB data 16 and a user definition body constituting the DB data 16 actually used for the job A.

  In the operational standby configuration for business A (business standby configuration for business A), node 0 is set as an operational system, and nodes 2 and 3 are set as standby systems for business A. Priorities for switching preferentially when the operation system goes down are set in advance in the nodes 2 and 3 as the standby system for the business A. For example, node 2 is set with priority 1, node 3 is set with priority 2, and so on. In the operation standby configuration for business A, each terminal (not shown) that actually performs the business is connected to, for example, the node 0 operating as the operation system via the network 2 or another network. When the operation system is switched, the connection with each terminal that actually performs business is switched in the order of the node 2 and the node 3 with priority.

  In such a business A operation standby configuration, the node 0 operating as the operation system holds the DB data 16 of the shared disk device 10 used for the business A in a storage area in the node 0 and performs a predetermined update. The DB data 16 of the shared disk device 10 is updated according to the processing, and is sequentially transferred as takeover information to the node 2 having the highest priority (for example, priority 1) among the active server computers. When the node 0 transfers the takeover information to the node 2, the node identification information of the node 2 is stored in the business A shared file 12 in advance. The DB data 16 held in the node 0 is all or part of the DB data 16 of the shared disk device 10.

  In the operation / standby configuration for business A, when the operation system is switched to the node 2 as the standby system, the node 2 indicates that the node identifier stored in the business A shared file 12 indicates the node 2. Starts immediately using the takeover information already stored in the storage area in the node 2 as the DB data 16. On the other hand, if the node identifier does not indicate node 2, the operation is started after the DB data 16 of the shared disk 10 is read.

  The same applies to the operation standby configuration for business B. In this case, the DB data 18 for the business B is transferred to the node 3 as takeover information, and the IP address of the node 3 is stored in the business system B shared file 14 by the node 1 as the operation system.

  For convenience of explanation, the DB data 16 and 18 are shown for each business. However, the DB data 16 and 18 may be configured for one business, or may be shared by a plurality of business.

  FIG. 3 is a diagram illustrating a hardware configuration of each node. In FIG. 3, each of nodes 0 to 3 (hereinafter simply referred to as nodes) includes a CPU (Central Processing Unit) 51, a memory unit 52, a display unit 53, an output unit 54, an input unit 55, and a communication unit. 56, a storage device 57, and a driver 58, which are connected to the system bus B.

  The CPU 51 executes processing according to a program stored in the memory unit 52. The memory unit 52 includes a RAM (Random Access Memory), a ROM (Read-Only Memory), and the like, and is obtained by a program executed by the CPU 51, data necessary for processing by the CPU 51, and processing by the CPU 51. Stored data. Further, a partial area of the memory unit 52 is allocated as a work area used for processing by the CPU 51.

  The display unit 53 displays various information required under the control of the CPU 51. The output unit 54 has a printer or the like, and is used for outputting various types of information in accordance with instructions from the user. The input unit 55 includes a mouse, a keyboard, and the like, and is used for inputting various information necessary for the administrator of the operation standby system. The communication unit 56 is a device that performs inter-node communication and communication control when accessing the shared disk device 10 via, for example, the Internet or a LAN (Local Area Network). The storage device 57 is composed of, for example, a hard disk unit, and stores data such as programs for executing various processes, DB data 16 or 14, takeover information received by inter-node communication, and the like.

  A program for realizing processing performed by the node is provided to the node by a storage medium 59 such as a CD-ROM (Compact Disc Read-Only Memory). That is, when the storage medium 59 storing the program is set in the driver 58, the driver 58 reads the program from the storage medium 59, and the read program is installed in the storage device 57 via the system bus B. . When the program is activated, the CPU 51 starts its processing according to the program installed in the storage device 57. The medium for storing the program is not limited to a CD-ROM, and any medium that can be read by a computer may be used. The program for realizing the processing according to the present invention may be downloaded via the network by the communication unit 56 and installed in the storage device 57. The program installed in this way is a program for causing a node as a computer to execute each process described later.

  FIG. 4 is a diagram for explaining the functional configuration of each node. In FIG. 4, each of the nodes 0 to 3 (hereinafter simply referred to as a node) has an OS (operating system) 61, a cluster management unit 62, and a database management unit 63, and according to these processing units 61 to 63. Are installed as processing units 61 to 63 in the node.

  The OS 61 is a UNIX (registered trademark) OS such as, for example, UNIX (registered trademark), Solaris (registered trademark), or Linux (registered trademark). In addition, the cluster management unit 62 performs life / death monitoring of processes by each node and the database management unit 63 on each node mainly by inter-node communication. Further, the database management unit 63 is an application program, and in cluster linkage, DB data 16 or 18 and data stored in a node such as a database management list and a log file, and a shared disk device Management is performed so as to maintain consistency between the data held at 10 and the data held at the nodes of the standby system.

  The cluster management unit 62 uses the shared disk device 10 and a storage area in the node (a storage area assigned to the cluster management unit 62 in the memory unit 52 or the storage device 57) and includes a system list including the switching priority order. 122 is managed. The system list 122 is determined based on an environment definition set in advance by an administrator of the standby system 100.

  The database management unit 63 acquires the system list 122 from the cluster management unit 62 every time the database management unit 63 itself is activated, and stores the storage area in the node (the database management unit in the memory unit 52 or the storage device 57). And the storage area allocated to 63). Further, when the business A is in operation and the node itself is the operation system, the node list information for identifying the node operating as the standby system is referred to as the connection destination IP address 124 with reference to the system list 122, and the node And the shared file 12 for business A of the shared disk device 10 are set.

  The system list 122 and the connection destination IP address 124 need only have a mechanism to be stored in at least the shared disk device 10, and are limited to being configured in a shared file corresponding to a business as shown in FIG. Is not to be done.

  FIG. 5 is a diagram illustrating an example of the switching priority order included in the system list. In FIG. 5, in the system list 122, entries indicated by pairs of node names and IP addresses are arranged in descending order of switching priority.

  Connection and disconnection processing (operation / standby connection / disconnection processing) between the active system and the standby system is the entire system startup in which all nodes 0 to 3 are started, the entire system stop in which all nodes 0 to 3 are stopped, and the standby system Stand-alone stand-alone system in which a node that is operating as a stand-alone system is stopped (for example, normal stop by maintenance work, down due to failure), stand-alone system start-up in which a node that should operate as a stand-alone system is booted independently (for example, maintenance) Triggered by the restart at the end of the work, the restart due to failure recovery), the standby system switchover (operation transfer from the active system to the standby system, or change of the standby system to a higher priority node) It will be executed. The active / standby connection / disconnection process is a process executed mainly by a node operating as an operating system, and the node operating as a standby system does not take the initiative.

  FIG. 6 is a diagram showing a processing flow for starting the entire system. In FIG. 6, when the database management unit 63 of the node operating as the operation system is activated, it acquires the system list 122 from the cluster management unit 62 (step S101). Further, invalidity is set to the connection destination IP address 124 of the shared disk device 10 (step S102). For example, all zero (hereinafter referred to as zero clear) is set.

  Subsequently, the database management unit 63 of the node operating as the operation system performs connection trial processing in order from the top of the entries for all the entries of the switching priority included in the acquired system list 122. A standby connection process (described later) is executed (step S103). The node identifier of the node that has been connected during standby is stored in the shared disk device 10 as the connection destination IP address 124 (step S104).

  On the other hand, when activated, the database management unit 63 of the node operating as the standby system acquires the system list 122 from the cluster management unit 62 as in the operation system (step S201). However, the system list 122 is not used while operating as a standby system. Used after switching to the operation system. When the database management unit 63 of the node operating as the standby system receives a connection request from the node operating as the operation system, the database management unit 63 returns a response and executes the connection process during operation standby (step S202).

  FIG. 7 is a diagram showing a processing flow for stopping the entire system. In FIG. 7, the database management unit 63 of the node operating as the operation system, when the connection destination IP address 124 of the shared disk device 10 is not invalidated (set to zero), the connection destination IP address 124 (operates as a standby system). The operation standby disconnection process is executed for the node that is operating (step S121). In this case, the connection destination IP address 124 of the shared disk device 10 is cleared to zero.

  On the other hand, when receiving the disconnection request from the node operating as the operation system, the database management unit 63 of the node operating as the standby system executes the disconnection process during operation standby (step S221).

  FIG. 8 is a diagram showing a processing flow of stand-by system single stop. In FIG. 8, when receiving the notification of the stopped node name from the cluster management unit 62 (step S131), the database management unit 63 of the node operating as the operation system uses the system list 122 to identify the stopped node name. Conversion into an address (step S132). When the connection destination IP address 124 matches the IP address of the stopped node, the database management unit 63 executes an operation standby disconnection process (described later) (step S133). In this case, the connection destination IP address 124 of the shared disk device 10 is invalidated (zero cleared).

  Subsequently, when the connection destination IP address 124 is invalidated, the database management unit 63 of the node operating as the operation system is connected to all nodes in the system list 122 in order to connect to the node operable as the standby system. Then, the connection trial process is executed in order from the top of the entry, and when the connection is established, an operation standby connection process (described later) is executed (step S134). The node identifier of the node that has been connected during standby is stored in the shared disk device 10 as the connection destination IP address 124 (step S135).

  In this case, no special processing is performed on the node operating as the standby system.

  FIG. 9 is a diagram showing a processing flow of stand-by system independent activation. In FIG. 9, when the database management unit 63 of the node operating as the operation system receives notification of the activated node name from the cluster managing unit 62 (step S141), the activated node name is set to IP using the system list 122. Conversion to an address (step S142). The database management unit 63 uses the system list 122, and when the switching priority is higher in the IP address of the activated node than in the connection destination IP address 124, the connection destination IP address 124 is disconnected during operation standby. Processing (described later) is executed (step S143). In this case, the connection destination IP address 124 of the shared disk device 10 is invalidated (zero cleared).

  Subsequently, when the connection destination IP address 124 is invalidated, the database management unit 63 of the node operating as the operation system is connected to all nodes in the system list 122 in order to connect to the node operable as the standby system. Then, the connection trial process is executed in order from the head of the entry, and when the connection is established, the connection process between operation standby (described later) is executed (step S144). The node identifier of the node that has been connected during standby is stored in the shared disk device 10 as the connection destination IP address 124 (step S145).

  On the other hand, when activated, the database management unit 63 of the node operating as the standby system acquires the system list 122 from the cluster management unit 62 as in the operation system (step S241). However, the system list 122 is not used while operating as a standby system. Used after switching to the operation system. When the database management unit 63 of the node that operates as the standby system receives a connection request from the node that operates as the operation system, the database management unit 63 returns a response and executes connection processing during operation standby (step S242).

  FIG. 10 is a diagram showing an operation standby connection processing flow. In FIG. 10, the database management unit 63 of the node operating as the operation system performs a connection trial process for all the entries of the switching priority included in the system list 122 in order from the top of the entry, so that there is a response. The connected node is connected as a standby system (step S301).

  When receiving the connection request from the operation system, the database management unit 63 of the node as the standby system executes a connection process (step S401) and purges an internal storage area for storing the takeover information (step S401). S402), a response to the connection request is made to the operation system.

  The database management unit 63 of the node operating as the operation system writes (writes) the IP address of the node that has responded to the connection request to the shared disk device 10 as the connection destination IP address 124 (step S302). Then, the operation standby transfer information transfer processing is executed to the node that has responded (step S303). In this case, all the handover information held by the operation system is transferred to the standby system (all pages batch transfer).

  FIG. 11 is a diagram showing a flow of a disconnection process during operation standby. In FIG. 11, the database management unit 63 of the node operating as the operation system sets invalidity to the connection destination IP address 124 of the shared disk device (step S311). For example, all zeros (hereinafter simply referred to as “0” (zero)) are set.

  Then, the database management unit 63 of the node operating as the operation system stops transmission of takeover information (step S312) and disconnects from the standby system (step S313). The database management unit 63 of the node that operates as the standby system disconnects the connection with the operation system (step S411).

  FIG. 12 is a diagram showing a processing flow for transferring information during operation standby. In FIG. 12, the database management unit 63 of the node operating as the operation system determines whether or not the takeover information transmission is started (step S501). That is, it is confirmed whether or not the standby system is connected. If it is not in the transmission start state, the operation standby information transfer process is terminated. On the other hand, when it is in the transmission start state, n pages (n is an integer of 1 or more) are read from the top of the DB data stored in the internal storage area (the storage area of the memory unit 52 or the storage device 57), and the shared disk 10 and also to a node operating as a standby system (step S502).

  The database management unit 63 of the node operating as the standby system receives n pages, stores them in the internal storage area (the storage area of the memory unit 52 or the storage device 57) (step S601), and receives all the n pages. A completion notification indicating this is transmitted to the operation system (step S602).

  When receiving the completion notification from the standby system, the database management unit 63 of the node operating as the operation system determines whether or not transmission has been completed (step S503). If all transmissions have not been completed, step S502 is executed again. On the other hand, when all the transmissions are completed, the operation standby handover information transfer process is terminated.

  In the active / standby handover information transfer process shown in FIG. 12, page n indicates all DB pages of the DB data stored in the internal storage area at the time of activation, and is a batch transmission of all pages. In addition, when the takeover information is updated, the updated DB page of the DB data stored in the internal storage area is targeted, and single page transmission or multiple page transmission is performed.

  Next, the switching process will be described. FIG. 13 is a diagram showing a switching process flow.

  In FIG. 13, the database management unit 63 confirms that the operation standby mode indicates standby (step S601), and reads the connection destination IP address 124 from the shared disk device 10 (step S602). It is determined whether or not the connection destination IP address 124 acquired from the shared disk device 10 is equal to the own IP address (step S603). That is, the consistency of the takeover information (DB data 16 or 18) held inside is determined by comparing the IP addresses.

  When the connection destination IP address 124 and the own IP address do not match (that is, when the takeover information is inconsistent), the internal takeover information is purged, and the DB data 16 corresponding to the job is shared from the shared disk device 10. Or 18 is read (step S604).

  The database management unit 63 performs a switching process so that the business can be continued with the takeover information held therein (step S605), and sets the operation to the operation standby mode (step S606). Then, the database management unit 63 sets invalidity to the connection destination IP address 124 of the shared disk device 10 (step S606).

  Next, the database management unit 63 attempts to connect to a node that operates as a standby system. That is, connection trial processing is executed in order from the top of the entries for all the entries of the switching priority included in the system list 122, and when connected, connection processing during standby (described above) is executed (step S608). The node identifier of the node that has been connected during standby is stored in the shared disk device 10 as the connection destination IP address 124 (step S609).

  As described above, in steps S602 and S603, the connection destination IP address 124 is read from the shared disk device 10 and compared with the own IP address, thereby determining whether the takeover information held inside is valid information. Can do.

  Further, by rewriting the internal takeover information with the DB data of the shared disk device 10 in step S604, it becomes possible to prevent inconsistency of the DB data due to the old takeover information.

  Furthermore, when the connection IP address and the own IP address match, the internal takeover information is used as it is, so that the switching process can be performed at high speed.

  In other words, in the processing flow described above, the connection destination IP address 124 is managed by the shared disk device 10 in order to prevent the use of old takeover information at the time of switching from the operation system to the standby system, and the connection destination is triggered at the following timing. The IP address 124 is manipulated to purge the takeover information.

  [Trigger 1] is a trigger immediately after the completion of the connection process during standby and before the transmission of takeover information. At [Trigger 1], the operation system updates the connection destination IP address in the shared disk device 10 (step S302 in FIG. 10), and the standby system purges the takeover information (step S402 in FIG. 10). ).

  [Trigger 2] is a trigger immediately before the disconnection process during operation standby. At this [Trigger 2], the operation system initializes the connection destination IP address 124 of the shared disk device 10 (step S311 in FIG. 11), so that the IP addresses of all standby systems do not match the connection destination IP address 124. To prevent old takeover information from being used continuously.

  [Timing 3] is when switching is started. In [Trigger 3], processing according to the comparison result between the connection destination IP address 124 of the shared disk device 10 and the own IP address is executed. If they match, the internal handover information is used without being purged (high-speed switching) (step S605 in FIG. 13). On the other hand, if they do not match, the handover information is purged and the DB data is read from the shared disk device 10 (switching by re-resident) (steps S604 and S605 in FIG. 13). In any of the comparison results, the connection destination IP address 124 of the shared disk device 10 is initialized (step S607 in FIG. 13), so that the IP addresses of all the standby systems and the connection destination IP address 124 are made inconsistent. The old takeover information is not used continuously.

  FIG. 14 illustrates a process for changing a standby system to a node having a higher priority according to the above-described processing flow when a node serving as a standby system having a higher switching priority is activated. FIG. 14 is a diagram for explaining a case where a standby system with a high switching priority is activated.

  In FIG. 14, the node 0 operating as the operation system has the node 1 with the switching priority 1 stopped, so that the node 2 with the priority 2 is connected as a standby system of the node 0, and the shared disk device The IP address of the node 2 is set to the 10 connection destination IP addresses 124 (state A1). Then, the node 0 operating as the operation system transfers the takeover information updated as needed to the node 2.

  Thereafter, when the node 1 with the priority 1 is activated (state A2), the node 0 operating as the operation system recognizes that the node 1 with the priority 1 has been activated by the inter-node communication, and connects the shared disk device 10 The destination IP address 124 is initialized ([Trigger 3]). The node 0 executes the operation standby disconnection process to disconnect the operation standby communication path (state A3). In this case, since the node 0 interrupts transmission of the takeover information to the node 2, the takeover information in the node 2 becomes old information.

  The node 0 operating as the operation system executes connection processing during standby for the high priority node 1, establishes a communication path during operation standby with the node 1, and updates the connection destination IP address 124. (Status A4) ([Trigger 1]). In this case, since the node 1 does not hold the takeover information immediately after the activation, the takeover information is not purged. Further, the IP address of the node 1 is set as the connection destination IP address 124 of the shared disk device 10.

  The node 0 operating as the operation system starts communication during operation standby with the node 1 (state A5), and the updated takeover information is transferred to the node 1.

  When node 0 operating as the operation system goes down in this state, the operation system is switched (state A6) ([Trigger 3]). Usually, the operation system is preferentially switched to the node 1. In this case, the IP address of the node 1 matches the connection destination IP address 124 of the shared disk device 10. Since the node 1 holds the latest takeover information inside, the node 1 is switched at high speed. Further, when the node 1 is stopped for some reason, the operation system is switched to the node 2. In this case, the IP address of the node 2 and the connection destination IP address 124 of the shared disk device 10 do not match. Therefore, the node 2 reads the DB data from the shared disk device 10 and makes it resident as takeover information.

  Therefore, even if new and old takeover information exists in the state A5, the operation can be continued with the latest takeover information by referring to the connection destination IP address 124 of the shared disk device 10.

  Next, FIG. 15 illustrates a process for changing a standby system to a node having the next highest priority according to the above-described processing flow when a node as a standby system having a higher priority for switching is stopped. FIG. 15 is a diagram for explaining a case where a standby system with a high priority of switching is stopped.

  In FIG. 15, the node 0 operating as the operation system is in a state (state B1) in which the takeover information is transferred using the node 1 having the switching priority 1 as the standby system, and the node having the priority 1 is transmitted by inter-node communication. Is recognized (state B2), the connection destination IP address of the shared disk device 10 is initialized ([Trigger 1]), and the operation standby communication path is executed to disconnect the operation standby communication path (state B3). ).

  Then, the node 0 operating as the operation system executes connection processing during standby for the node 2 with the second highest priority, and establishes a communication path during operation standby with the node 2. The connection destination IP address 124 is updated (state B4) ([Trigger 1]). In this case, since there is a possibility that the node 2 holds the old takeover information as it is, the node 2 purges the internal takeover information. Further, the IP address of the node 2 is set as the connection destination IP address 124 of the shared disk device 10.

  The node 0 operating as the operation system starts communication between the standby and the node 2 (state B5), the updated takeover information is transferred to the node 2, and the node 2 holds the latest takeover information inside. It will be.

  When node 0 operating as the operation system goes down in this state, the operation system is switched (state B6) ([Trigger 3]). Usually, the operation system tries to switch to node 1 preferentially, but since node 1 is in a stopped state, switching to node 1 does not occur. Therefore, the operation system is switched to the node 2 having the priority order 2. Here, the IP address of the node 2 matches the connection destination IP address 124 of the shared disk device 10. Since the node 2 holds the latest takeover information inside, the node 2 is switched at high speed.

  As described above, in the operation standby system 100 that constitutes a plurality of standby systems, each standby system is held internally by storing in the shared disk device 10 the IP address of the standby system to which transfer information is transferred. It is possible to determine whether the inherited information is valid information.

  In addition, when the takeover information held inside is valid information, it is possible to perform high-speed switching that continues operation without reading the DB data of the shared disk device 10.

  In the present embodiment, a plurality of standby systems are mainly exemplified for one operation system. However, as shown in FIG. 2, operation standby in which a plurality of standby systems are configured for two or more operation systems. Applicable in system 100.

Regarding the above description, the following items are further disclosed.
(Appendix 1)
In an active / standby system comprised of an active system and a plurality of standby systems, the computer executes processing for checking the consistency of the takeover information that is taken over when switching from the active system to one of the standby systems. A program that causes the computer to
Standby for reading standby identification information for identifying a switching standby system stored in the external shared storage area from the external shared storage area shared by the operation system and the plurality of standby systems and storing the standby identification information in the internal storage area An identification information reading procedure;
A consistency check procedure for checking the consistency of the takeover information by matching the standby identification information stored in the internal storage area with the self-identification information;
When the consistency confirmation procedure confirms the consistency of the takeover information by matching the standby identification information with the self-identification information, the operation system uses the takeover information held in the internal storage area. A computer-executable program for executing an operation switching procedure for starting an operation as
(Appendix 2)
In the computer,
Takeover information erasure procedure for erasing the takeover information held in the internal storage area when the consistency confirmation procedure confirms inconsistency of the takeover information due to a mismatch between the standby identification information and the self identification information When,
The computer-executable program as set forth in claim 1, wherein a takeover information reading procedure for reading the takeover information from the external shared storage area is executed.
(Appendix 3)
In the computer,
When the computer operates as the operation system, it is connected to one of the standby systems in accordance with the switching priority order, and the standby identification information that identifies the connected standby system is written to the external shared storage area. Procedure and
In accordance with the update of the takeover information, the takeover information is written to the external shared storage area, and a takeover information update procedure for transferring the takeover information to the connected standby system is executed. Computer executable program.
(Appendix 4)
The computer-implemented operation according to claim 3, wherein the takeover information update procedure is transmitted to the standby system connected to the external shared storage area in accordance with a predetermined information unit constituting all of the takeover information or the takeover information. Possible program.
(Appendix 5)
In an active / standby system comprised of an active system and a plurality of standby systems, the computer executes processing for checking the consistency of the takeover information that is taken over when switching from the active system to one of the standby systems. A takeover information consistency check method, wherein the computer
Standby identification information that reads standby identification information for identifying the switching standby system stored in the external shared storage area is read from the external shared storage area shared by the operation system and the plurality of standby systems. Information reading procedure;
A consistency check procedure for checking the consistency of the takeover information by matching the standby identification information stored in the internal storage area with the self-identification information;
When the consistency confirmation procedure confirms the consistency of the takeover information by matching the standby identification information with the self-identification information, the operation system uses the takeover information held in the internal storage area. A method for confirming the takeover information consistency, comprising performing an operation switching procedure for starting the operation as
(Appendix 6)
A node device that operates as the standby system for confirming the consistency of takeover information that is taken over when switching from the operation system to one of the plurality of standby systems in a configuration system of an operation system and a plurality of standby systems; There,
Standby identification information that reads standby identification information for identifying the switching standby system stored in the external shared storage area is read from the external shared storage area shared by the operation system and the plurality of standby systems. Information reading means;
Consistency confirmation means for confirming the consistency of the takeover information based on a match between the standby identification information stored in the internal storage area and the self identification information;
When the consistency confirmation unit confirms the consistency of the takeover information by matching the standby identification information with the self identification information, the operation system uses the takeover information held in the internal storage area. A node device characterized by executing operation switching means for starting operation as
(Appendix 7)
A node device that operates as an operation system for confirming the consistency of takeover information that is taken over when switching from the operation system to one of the plurality of standby systems in a configuration system composed of an operation system and a plurality of standby systems. And
Standby identification information writing means for connecting to one of the standby systems according to the priority of switching, and writing the standby identification information for identifying the connected standby system in the external shared storage area;
In response to the update of the takeover information, the node device writes the takeover information to the external shared storage area, and causes the connected standby system to execute takeover information update means for transferring the takeover information.

  The present invention is not limited to the specifically disclosed embodiments, and various modifications and changes can be made without departing from the scope of the claims.

It is a figure for demonstrating the case where old and new takeover information exists in the conventional switching method. It is a figure which shows the network structure of the operation standby system which concerns on one Example of this invention. It is a figure which shows the hardware constitutions of each node. It is a figure for demonstrating the function structure of each node. It is a figure which shows the example of the switching priority contained in a system list. It is a figure which shows the processing flow of a whole system starting. It is a figure which shows the processing flow of a whole system stop. It is a figure which shows the processing flow of stand-by system independent stop. It is a figure which shows the processing flow of stand-by system independent starting. It is a figure which shows the connection processing flow between operation standby. It is a figure which shows a disconnection process flow during operation standby. It is a figure which shows the taking over information transfer process flow between operation standby. It is a figure which shows the switching process flow. It is a figure for demonstrating the case where the standby system with a high priority of switching starts. It is a figure for demonstrating the case where the standby system with a high priority of switching stops.

Explanation of symbols

Nodes 0 to 3 Operation system or standby system 2 Network 10 Shared disk device 12 Shared file for business A 14 Shared file for business B 16 DB data 18 DB data 51 CPU
52 Memory Unit 53 Display Unit 54 Output Unit 55 Input Unit 56 Communication Unit 57 Storage Device 58 Driver 59 Storage Medium 100 Operation Standby System

Claims (5)

In an active / standby system comprised of an active system and a plurality of standby systems, the computer executes processing for checking the consistency of the takeover information that is taken over when switching from the active system to one of the standby systems. A program that causes the computer to
Standby for reading standby identification information for identifying a switching standby system stored in the external shared storage area from the external shared storage area shared by the operation system and the plurality of standby systems and storing the standby identification information in the internal storage area An identification information reading procedure;
A consistency check procedure for checking the consistency of the takeover information by matching the standby identification information stored in the internal storage area with the self-identification information;
When the consistency confirmation procedure confirms the consistency of the takeover information by matching the standby identification information with the self-identification information, the operation system uses the takeover information held in the internal storage area. A computer-executable program for executing an operation switching procedure for starting an operation as In the computer,
Takeover information erasure procedure for erasing the takeover information held in the internal storage area when the consistency confirmation procedure confirms inconsistency of the takeover information due to a mismatch between the standby identification information and the self identification information When,
The computer-executable program according to claim 1, wherein a takeover information reading procedure for reading the takeover information from the external shared storage area is executed. In the computer,
When the computer operates as the operation system, it is connected to one of the standby systems in accordance with the switching priority order, and the standby identification information that identifies the connected standby system is written to the external shared storage area. Procedure and
2. The takeover information update procedure for writing the takeover information to the external shared storage area and transferring the takeover information to the connected standby system in response to the update of the takeover information. Or the computer-executable program of 2 description.   4. The computer according to claim 3, wherein the takeover information update procedure is transmitted to the standby system connected to the external shared storage area in accordance with all of the takeover information or a predetermined information unit constituting the takeover information. An executable program. In an active / standby system comprised of an active system and a plurality of standby systems, the computer executes processing for checking the consistency of the takeover information that is taken over when switching from the active system to one of the standby systems. A takeover information consistency check method, wherein the computer
Standby identification information that reads standby identification information for identifying the switching standby system stored in the external shared storage area is read from the external shared storage area shared by the operation system and the plurality of standby systems. Information reading procedure;
A consistency check procedure for checking the consistency of the takeover information by matching the standby identification information stored in the internal storage area with the self-identification information;
When the consistency confirmation procedure confirms the consistency of the takeover information by matching the standby identification information with the self-identification information, the operation system uses the takeover information held in the internal storage area. A method for confirming the takeover information consistency, comprising performing an operation switching procedure for starting the operation as

Images