JPWO2013108351A1 - Computer system and logical storage area management method - Google Patents

Abstract

An object of the present invention is to provide a computer system capable of constructing a shared storage having a fault tolerance function using a general computer. A computer system including a plurality of computers, wherein the computer uses a storage area provided by each storage medium of the plurality of computers to generate a logical storage area, and the service using the logical storage area. The redundancy processing unit generates a logical storage area using the storage areas provided by the storage medium of the primary computer and the secondary computer, the logical storage area, and the logical storage area The redundant information including the correspondence with the storage medium of the primary computer and the secondary computer constituting the system is generated, the redundant information is written in the storage medium of the primary computer and the secondary computer, and the logical storage area is provided as a service providing unit. When the access request is received, the logical storage area is accessed with reference to the redundancy information.

Description

  The present invention relates to a distributed shared file system constructed using a plurality of computers. In particular, the present invention relates to a computer system capable of constructing a shared storage having a fault tolerance function and a shared storage management method.

  A computer system composed of a plurality of computers is required to have a fault-tolerant function capable of continuing operation as a whole system even when a failure occurs in one computer.

  As a fault-tolerant function, there is a method that prepares a primary computer that is normally used and a secondary computer that stands by in preparation for a failure of the primary computer, and when the primary computer fails, the secondary computer takes over the operation. Taken. In this case, the main computer and the sub computer have a shared storage, and the sub computer can take over the contents of the storage changed by the main computer.

  However, in the configuration as described above, there is a possibility that the data is destroyed when the primary computer unintentionally accesses the shared storage after the secondary computer takes over the operation. In order to eliminate the risk of data destruction as described above, exclusive control of access to the shared storage is necessary.

  For example, in Patent Document 1, when a function stoppage is detected by the function correctness monitoring unit, a plurality of nodes issue commands to instruct the shared disk to occupy the shared disk, and acquire the occupation right A method is disclosed in which only a node can control a shared disk.

JP 2002-185478 A

  Generally, the shared storage as described above is configured using a dedicated storage system. The storage system includes a function for providing a storage area that can be shared and used by a plurality of computers as described above (hereinafter referred to as a shared function) and an exclusive control function. Therefore, there is a problem that the shared storage cannot be configured only with a general computer.

  For example, Patent Document 1 prepares a dedicated shared storage device outside the computer, but does not constitute a shared storage using a storage device included in the computer.

  Also, there is no description on how to control the shared storage when a failure occurs in one computer when the shared storage is configured using a storage device included in the computer.

  For example, when a shared storage configured using storage devices included in a plurality of computers is used, if a failure occurs in one computer, a failure occurs in the shared storage. Therefore, the primary computer or the secondary computer cannot continue the service using the shared storage. Patent Document 1 does not describe a configuration for continuing the service without recognizing the failure of the shared storage.

  The present invention has been made in view of the above problems, and an object of the present invention is to provide a computer system that is inexpensive and has a fault-tolerant function using a computer that does not have a sharing function and an exclusive control function. To do.

  A typical example of the invention disclosed in the present application is as follows. That is, a computer system including a plurality of computers, each of the plurality of computers including a processor, a memory connected to the processor, a storage medium connected to the processor, and another device via a network A plurality of computers, each of which includes one main computer that provides a service and one or more secondary computers that take over the service when a failure occurs in the main computer. The main computer generates a logical storage area using a storage area provided by the storage medium of each of the plurality of computers, and manages access to the generated logical storage area. A first redundancy processing unit, and a first service providing unit that provides the service using the logical storage area, and the subordinate computer includes: A fault control unit that monitors a system computer and detects a failure of the main computer, and executes a process for taking over the service; and a second service provision that provides the service using the logical storage area And the first redundancy processing unit generates a first logical storage area using a storage area provided by the storage medium of the primary computer and the storage medium of the secondary computer, Generating first redundancy information including a correspondence relationship between the first logical storage area and the storage medium of the primary computer and the storage medium of the secondary computer constituting the first logical storage area; The first redundancy information is written in the storage medium of the primary computer and the storage medium of the secondary computer constituting the first logical storage area, and the first logical storage area is stored in the first service. Providing to the provision department, the first When a service providing unit is instructed to start the service, and an access request is received from the first service providing unit, the first logical storage area is accessed with reference to the first redundancy information. The failure control unit activates the second service providing unit when detecting a failure of the primary computer, and the second service providing unit is stored in the first logical storage area. The service is continued using information.

  According to one embodiment of the present invention, it is possible to construct a logical storage area (shared storage) having a fault tolerance function using a general computer. In addition, when a failure occurs in the main computer, the business processing unit can continue the service without being aware of the failure in the logical storage area.

It is a block diagram which shows the structure of the computer system of 1st embodiment of this invention. It is a flowchart explaining the process which the structure setting part (main system) in 1st embodiment of this invention performs. It is a flowchart explaining the process which the sub system selection part in 1st embodiment of this invention performs. It is a flowchart explaining the process which the redundancy process part (primary system) in 1st embodiment of this invention performs. It is a flowchart explaining the process which the structure setting part (secondary system) in 1st embodiment of this invention performs. It is a flowchart explaining the process which the failure control part in 1st embodiment of this invention performs. It is a flowchart explaining the process which the structure setting part (secondary system) in 1st embodiment of this invention performs when a failure generate | occur | produces. It is a flowchart explaining the process which the redundancy process part (subsystem) in 1st embodiment of this invention performs. It is a flowchart explaining the process which the structure recovery part (primary system) in 1st embodiment of this invention performs. It is a flowchart explaining the process which the structure recovery part (secondary system) in 1st embodiment of this invention performs. It is the block diagram which showed the structure of the computer system of 2nd embodiment of this invention.

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

  (First embodiment)

  FIG. 1 is a block diagram showing the configuration of the computer system according to the first embodiment of this invention.

  The computer system according to this embodiment includes a computer 101A and a computer 101B. The computer 101A and the computer 101B are connected to each other via the network 180. For example, the network 180 may be a LAN (Local Area Network). However, the present invention is not limited to the connection form of the network 180. Hereinafter, when the computer 101A and the computer 101B are not distinguished, they are described as the computer 101.

  In the present embodiment, the shared storage is configured using a storage medium included in the computer 101A and the computer 101B. Further, it is assumed that the computer 101A operates as a primary computer that provides a service using a shared storage, and the computer 101B operates as a secondary computer that continues the service when a failure occurs in the computer 101A. In FIG. 1, there are two computers 101, but there may be three or more computers.

  First, the hardware configuration of the computer 101 will be described.

  The computer 101A includes a processor 102A, a memory 103A, a storage interface 104A, a disk device 105A, and a network interface 106A.

  The processor 102A executes a program stored in the memory 103A. The function of the computer 101A can be realized by the processor 102A executing the program. Hereinafter, when processing is described with the program as the subject, it indicates that the program is being executed by the processor 102A.

  The memory 103A stores a program executed by the processor 102A and data necessary for executing the program. As the memory 103A, for example, a semiconductor memory such as a DRAM can be considered, and the memory 103A can be accessed at a higher speed than the disk device 105A. The program and data stored in the memory 103A will be described later.

  The storage interface 104A is an interface for connecting to a disk device 105A capable of storing a large amount of data.

  The disk device 105A stores information (for example, file data) necessary for a predetermined service. As the disk device 105A, for example, an HDD (Hard Disk Drive) can be considered.

  In addition, it may be a storage medium such as an SSD (Solid State Drive) other than the disk device 105. Further, there may be a plurality of disk devices 105A. Further, the disk device 105A may be in the form of being externally attached to the computer 101A.

  The network interface 106A is an interface for connecting to other devices via the network 180.

  In addition to the memory 103A and the disk device 105A, the computer 101A may include a storage device that stores information such as an OS (Operating System).

  Since the hardware configuration of the computer 101B is the same as that of the computer 101A, description thereof is omitted.

  Next, the software configuration of the computer 101 will be described.

  In the memory 103A of the computer 101A, a program for realizing the service providing unit 150, the redundancy processing unit 151A, the disk driver unit 152A, the network disk driver unit 153A, the sub system selection unit 154, the configuration setting unit 155A, and the configuration recovery unit 156A. Stored.

  The service providing unit 150 provides a predetermined service using the shared storage. The service providing unit 150 outputs an access request for the shared storage when providing the service.

  The redundancy processing unit 151A receives an access request for the shared storage, and executes an access process corresponding to the access request. Specifically, the redundancy processing unit 151A outputs an access request to the disk driver unit 152A and the network disk driver unit 153B. Details of the processing executed by the redundancy processing unit 151A will be described later with reference to FIG. The redundancy processing unit 151A can be realized by a software RAID function provided in the OS, for example.

  The disk driver unit 152A accesses the disk device 105A constituting the shared storage based on the access request output from the redundancy processing unit 151A.

  The network disk driver unit 153A accesses the disk device 105B constituting the shared storage via the network 180. Specifically, the network disk driver unit 153A transmits an access request to the network disk driver unit 153B.

  The sub system selection unit 154 selects a computer 101 that is a sub system computer from a plurality of computers 101 included in the computer system. Details of processing executed by the sub system selection unit 154 will be described later with reference to FIG.

  The configuration setting unit 155A sets information necessary for operating as a main computer. Details of the processing executed by the configuration setting unit 155A will be described later with reference to FIG.

  When a failure occurs in the computer 101A, the configuration recovery unit 156A executes a process for recovering the failure. Details of the processing executed by the configuration recovery unit 156A will be described later with reference to FIG.

  The program stored in the memory 103A may be stored in the disk device 105A or an external device (not shown). In this case, each program is read from the disk device 105A, or each program is read from an external device via the network 180 and stored in the memory 103A.

  In the memory 103B of the computer 101B, a program for realizing the failure control unit 170, the alternative service providing unit 171, the redundancy processing unit 151B, the disk driver unit 152B, the network disk driver unit 153B, the configuration setting unit 155B, and the configuration recovery unit 156B Is stored.

  The failure control unit 170 monitors the operation of the computer 101A that operates as the primary computer, and when the failure of the computer 101A is detected, the computer 101B executes processing for continuing the service. Details of processing executed by the failure control unit 170 will be described later with reference to FIG.

  The alternative service providing unit 171 provides a service instead of the computer 101A in which a failure has occurred.

  The redundancy processing unit 151B is the same as the redundancy processing unit 151A, receives the access request output from the alternative service providing unit 171 and executes access processing for the shared storage. The disk driver unit 152B is the same as the disk driver unit 152B. Details of the processing executed by the redundancy processing unit 151B will be described later with reference to FIG.

  The network disk driver unit 153B outputs an access request to the disk driver unit 152B based on the access request received from the network disk driver unit 153A. Thereby, access to the disk device 105B constituting the shared storage can be realized. Further, the network disk driver unit 153B accesses the disk device 105A constituting the shared storage via the network 180.

  The configuration setting unit 155B sets information necessary for operating as a secondary computer. Details of processing executed by the configuration setting unit 155B will be described later with reference to FIGS.

  The configuration recovery unit 156B executes processing for operating as a secondary computer again after the failure of the primary computer is recovered. Details of the processing executed by the configuration recovery unit 156B will be described later with reference to FIG.

  The program stored in the memory 103B may be stored in the disk device 105B or an external device (not shown). In this case, each program is read from the disk device 105B, or each program is read from an external device via the network 180 and stored in the memory 103B.

  The configuration setting unit 155A and the configuration setting unit 155B are programs that provide the same function, and can execute the setting process of the main computer and the setting process of the sub computer. The configuration recovery unit 156A and the configuration recovery unit 156B are programs that provide the same function, and can execute recovery processing for the main computer and recovery processing for the sub computer.

  Hereinafter, when the disk driver unit 152A and the disk driver unit 152B are not distinguished from each other, the disk driver unit 152 is referred to as the disk driver unit 152. When the configuration setting unit 155A and the configuration setting unit 155B are not distinguished from each other, the configuration setting unit 155 is described. When the configuration recovery unit 156A and the configuration recovery unit 156B are not distinguished from each other, they are described as the configuration recovery unit 156.

  Hereinafter, processing of each configuration will be described. First, the setting method of the main computer and the sub computer will be described with reference to FIGS.

  FIG. 2 is a flowchart illustrating processing executed by the configuration setting unit 155A (main system) in the first embodiment of the present invention.

  The configuration setting unit 155A starts processing upon receiving an instruction from the administrator of the computer system (step S201). At this time, the instruction includes the total number of computers 101 in the computer system and the identification number of the computer 101A on which the configuration setting unit 155A is executed.

  For example, in the example of FIG. 1, the total number of computers 101 is “2”, the identification number of the computer 101A is “1”, and the identification number of the computer 101B is “2”.

  The configuration setting unit 155A calls the sub system selection unit 154 (step S202). The called sub-system selection unit 154 executes a process described later (see FIG. 3), so that the computer 101 to be a sub-system computer can be determined. At this time, the configuration setting unit 155A acquires the identification information of the computer 101 that is the sub computer from the sub system selection unit 154.

  The identification information of the computer 101 only needs to be information that can uniquely identify the computer 101 in the computer system. For example, the computer name, MAC address, and IP address of the computer 101 can be considered.

  The configuration setting unit 155A instructs the network disk driver unit 153A to connect to the network disk driver unit 153 included in the secondary computer (step S203). In the example shown in FIG. 1, the network disk driver unit 153A is instructed to connect to the network disk driver unit 153B. Accordingly, the network disk driver unit 153A can transmit an access request to the network disk driver unit 153B.

  The configuration setting unit 155A instructs the redundancy processing unit 151A to generate a logical device (shared storage) (step S204). Here, the logical device is a logical disk device that the service providing unit 150 can recognize as one disk device. In the present invention, a logical device generated from the storage area of the disk device 105 of the plurality of computers 101 is used as the shared storage.

  As a result, logical devices are generated from the storage areas of the disk device 105A and the disk device 105B. The same write request is output to each of the disk driver unit 152A and the network disk driver unit 153A. Therefore, in the writing process to the logical device, data is written to each of the disk device 105A and the disk device 105B. Since the data is stored in the different disk devices 105 as described above, even if a failure occurs in the main computer, the service can be continued using the data stored in the disk device 105 of the secondary computer.

  A distributed shared file system can be constructed on the logical device. All the computers 101 can access the files arranged on the shared file system in the same way.

  The configuration setting unit 155A instructs the service providing unit 150 to access the logical device, and ends the process (steps S205 and S206). Specifically, the configuration setting unit 155A notifies the identification information of the logical device as a storage area used for the service. As a result, the service providing unit 150 outputs an access request to the logical device.

  FIG. 3 is a flowchart illustrating processing executed by the sub system selection unit 154 according to the first embodiment of this invention.

  When called from the configuration setting unit 155A, the sub system selection unit 154 starts processing (step S301). When called, the secondary system selection unit 154 receives from the configuration setting unit 155A the total number of computers 101 constituting the computer system and the identification number of the computer 101 that executes the configuration setting unit 155A.

  The sub system selection unit 154 determines whether or not its own identification number is the maximum identification number among the identification numbers assigned to the computers 101 constituting the computer system (step S302). For example, the sub system selection unit 154 determines whether or not its own identification number is the same as the total number of computers 101. If its own identification number is the same as the total number of computers 101, it is determined that its own identification number is the maximum identification number.

  When it is determined that its own identification number is the largest, the sub system selection unit 154 selects the computer 101 having the smallest identification number (for example, the identification number is “1”) as the sub system computer, and the process proceeds to step S305. (Step S303).

  When it is determined that its own identification number is not the maximum, the sub system selection unit 154 selects the computer 101 having the identification number obtained by adding “1” to its own identification number as the sub system computer, and proceeds to step S305 (step S305). S304).

  The sub system selection unit 154 acquires the identification information of the computer 101 selected as the sub system computer, and ends the process (steps S305 and S306).

  For example, a method in which data in which an identification number is associated with identification information is prepared in advance, and the subsystem selection unit 154 obtains the identification information by referring to the data based on the identification number of the subsystem computer Can be considered.

  The selection method of the subordinate computer is not limited to that shown in FIG. 3, and may be a method of selecting based on the resource amount, usage rate, etc. of the computer 101. When two or more secondary computers are selected, the process shown in FIG. 3 may be repeatedly executed with the identification number of the computer 101 selected as the secondary computer as a new input.

  FIG. 4 is a flowchart for explaining processing executed by the redundancy processing unit 151A (main system) in the first embodiment of the present invention.

  The redundancy processing unit 151A starts processing upon receiving a logical device generation instruction from the configuration setting unit 155A (step S401). The redundancy processing unit 151A acquires the identification information of the secondary computer (step S402).

  The redundancy processing unit 151A generates a logical device by integrating the disk device 105 of the primary computer and the disk device 105 of the secondary computer (step S403). In the example shown in FIG. 1, a logical device in which the disk device 105A and the disk device 105B are integrated is generated.

  The redundancy processing unit 151A generates redundancy information in which the logical device is associated with each disk device 105 (step S404). The redundancy information includes at least logical device identification information and disk device 105 identification information. The redundancy information may include other information such as identification information of the computer 101.

  The redundancy processing unit 151A writes the redundancy information to each disk device 105, and ends the processing (steps S405 and S406). In the example shown in FIG. 1, the redundancy processing unit 151A outputs a request for writing redundancy information to each of the disk driver unit 152A and the network disk driver unit 153A. As a result, the redundancy information is stored in the shared storage, that is, the disk device 105A and the disk device 105B constituting the logical device.

  FIG. 5 is a flowchart for explaining processing executed by the configuration setting unit 155B (secondary system) according to the first embodiment of the present invention.

  The configuration setting unit 155B starts processing upon receiving an instruction from the administrator of the computer system (step S501). Note that the processing of the configuration setting unit 155B is executed before the processing of the configuration setting unit 155A.

  The configuration setting unit 155B sets the network disk driver unit 153B to permit access from the other computer 101 to its own disk device 105, and ends the processing (steps S502 and S503). In the example shown in FIG. 1, access from the computer 101A to the disk device 105B is permitted. By the processing executed by the configuration setting unit 155B, the computer 101A can access the disk device 105B and can provide a storage area constituting the logical device.

  The above is the processing executed when the shared storage is configured. After the processes in FIGS. 2 to 5 are completed, the service providing unit 150 provides a predetermined service using the logical device.

  At this time, when receiving the write request to the logical device from the service providing unit 150, the redundancy processing unit 151 reads the redundancy information from the disk device 105A, and based on the redundancy information, the disk device 105A and the disk device Data is written to each of 105B. Note that the read redundancy information is temporarily stored in the memory 103A. This can reduce the occurrence of I / O to the disk device 105 when accessing the logical device.

  Further, upon receiving a read request from the service providing unit 150 to the logical device, the redundancy processing unit 151 reads data from the disk device 105A with reference to the redundancy information.

  Next, processing executed when a failure occurs in the computer 101A operating as the main computer will be described with reference to FIGS.

  FIG. 6 is a flowchart illustrating processing executed by the failure control unit 170 according to the first embodiment of this invention.

  The failure control unit 170 periodically monitors the main computer, and starts processing when a failure of the main computer is detected (step S601). In the example illustrated in FIG. 1, the failure control unit 170 monitors the computer 101A that operates as the primary computer, and starts processing when it detects a failure in the computer 101A.

  As a monitoring method, a method of monitoring communication via the network 180 can be considered. However, the present invention is not limited to the failure detection method of the main computer.

  If there are a plurality of secondary computers, priorities may be given to the secondary computers in advance, and the secondary computer having a higher priority may be configured to execute the process.

  The failure control unit 170 instructs the network disk driver unit 153B to prohibit access to the disk device 105 from another computer 101 configuring the logical device (step S602). When there are two or more secondary computers, access not only from the primary computer but also from other secondary computers is prohibited. In the example shown in FIG. 1, access from the computer 101A to the disk device 105B is prohibited.

  The failure control unit 170 calls the configuration setting unit 155B (step S603). The failure control unit 170 waits for notification of processing completion from the configuration setting unit 155B. Details of the processing executed by the configuration setting unit 155B will be described later with reference to FIG.

  Thereafter, the failure control unit 170 instructs the alternative service providing unit 171 to start processing, and ends the processing (steps S604 and S605).

  Through the above processing, the alternative service providing unit 171 can continue the service in place of the service providing unit 150.

  In step S602, access from the computer 101A to the disk device 105B is prohibited. This is to avoid the risk of data loss due to collision between accesses from the alternative service providing unit 171 and the service providing unit 150 to the disk device 105B when the computer 101A starts again unexpectedly.

  FIG. 7 is a flowchart for explaining processing executed by the configuration setting unit 155B (secondary system) in the first embodiment of the present invention when a failure occurs.

  The configuration setting unit 155B starts processing when read from the failure control unit 170 (step S701). The configuration setting unit 155B instructs the network disk driver unit 153B to connect to the network disk driver unit 153B included in another secondary computer (step S702). As a result, the network disk driver unit 153B can transmit an access request to the network disk driver unit 153 of the other secondary computer.

  Since the secondary computer has already been selected, the processing corresponding to step S202 is omitted. In addition, the configuration setting unit 155B can specify another sub computer by referring to the redundancy information.

  The configuration setting unit 155B instructs the redundancy processing unit 151B to create a logical device (shared storage) (step S703). In the example illustrated in FIG. 1, a logical device is generated from the disk device 105B included in the computer 101B.

  The configuration setting unit 155B instructs the alternative service providing unit 171 to access the newly created logical device, and ends the process (steps S704 and S705).

  FIG. 8 is a flowchart for explaining processing executed by the redundancy processing unit 151B (secondary system) according to the first embodiment of this invention.

  The redundancy processing unit 151B starts processing when read from the configuration setting unit 155B (step S801). The redundancy processing unit 151B acquires the identification information of the secondary computer (step S802). The process of step S802 is the same process as step S402.

  The redundancy processing unit 151B generates a logical device by integrating the disk devices 105 of the secondary computers (step S803).

  The processing in step S803 differs from step S403 in that a logical device is generated from the disk device 105 excluding the disk device 105 of the main computer. That is, a logical device is generated only from the disk device 105 of the secondary computer.

  In the example shown in FIG. 1, a logical device is generated only from the disk device 105B.

  The redundancy processing unit 151B generates redundancy information in which the logical device is associated with each disk device 105 (step S804). The process of step S804 is the same process as step S404.

  The redundancy processing unit 151B writes the redundancy information to each disk device 105 and ends the processing (steps S805 and S806). In the example shown in FIG. 1, the redundancy processing unit 151B outputs a request for writing redundancy information to the network disk driver unit 153B. As a result, the redundancy information is stored in the disk device 105B constituting the logical device.

  The newly generated logical device identification information is set to be the same as the first generated logical device identification information. As a result, the service can be continued under the same operating environment as before the occurrence of the failure. However, identification information of different logical devices may be set.

  In the present embodiment, there is an effect of improving the fault tolerance by the processing described above. Normally, when the redundancy processing unit 151B receives a write request, the redundancy processing unit 151B reads the redundancy information stored in the disk device 105 and identifies the disk device 105 constituting the logical device. Further, the redundancy processing unit 151B writes the same data to all the specified disk devices 105.

  However, when a failure occurs in the computer 101 that operates as the primary computer, the disk device 105 of the primary computer constituting the logical device cannot be used. Therefore, the redundancy processing unit 151B determines that an error has occurred in the logical device and the logical device cannot be used. Therefore, the redundancy processing unit 151B cannot write data to the plurality of disk devices 105 constituting the logical device. That is, data redundancy cannot be realized.

  Therefore, in this embodiment, the redundancy processing unit 151B called from the failure control unit 170 constructs a new logical device using the other disk devices 105 other than the disk device 105 of the primary computer. As a result, the redundancy processing unit 151B can write data to the plurality of disk devices 105 when receiving a write request from the alternative service providing unit 171.

  In addition, as shown in FIG. 1, even if there is only one disk device 105 constituting a logical device, the alternative service providing unit 171 does not recognize that a failure has occurred in the logical device, There is an effect that it can be continued.

  FIG. 9 is a flowchart for explaining processing executed by the configuration recovery unit 156A (main system) in the first embodiment of the present invention.

  The configuration recovery unit 156A starts processing upon receiving an instruction to start processing from the administrator of the computer system after the computer 101A in which the failure has been recovered restarts (step S901). The configuration recovery unit 156A calls the configuration recovery unit 156B of the secondary computer (step S902). As a result, processing for taking over the service from the alternative service providing unit 171 to the service providing unit 150 (see FIG. 10) is executed.

  The configuration recovery unit 156A instructs the network disk driver unit 153A to connect to the network disk driver unit 153B included in the secondary computer (step S903). The process of step S903 is the same process as step S203. By the processing in step S903, the computer 101A can access the logical device.

  The configuration recovery unit 156A instructs the redundancy processing unit 151A to acquire redundancy information from the secondary computer (step S904). The redundancy processing unit 151A that has received the instruction outputs an access request for acquiring redundancy information from the secondary computer to the network disk driver unit 153A.

  In step S904, the redundant configuration of the new logical device generated in step 803 is read. As a result, the service can be continued while maintaining the redundant configuration.

  In the example shown in FIG. 1, data cannot be duplicated because there is only one disk device 105B constituting the logical device, but the service can be continued without recognizing that a failure has occurred in the logical device. There is an effect.

  The configuration recovery unit 156A instructs the service providing unit 150 to access the logical device, and further instructs the start of the service (step S905). The process of step S905 is the same process as step S205.

  The configuration recovery unit 156A instructs the redundancy processing unit 151A to place the disk device 105A under management, that is, instructs the reconfiguration of the logical device (step S906).

  The redundancy processing unit 151A that has received the instruction executes processing similar to the processing illustrated in FIG. Specifically, in step S403, the redundancy processing unit 151A generates a logical disk using all of the disk device 105 of the primary computer and the disk device 105 of the secondary computer. That is, a logical device having the same configuration as that before the failure occurs is generated. In step S405, the redundancy processing unit 151A writes the redundancy information to each of the disk device 105 of the primary computer and the disk device 105 of the secondary computer. At this point, data is not reflected on the disk device 105 of the main computer constituting the logical device.

  With the above processing, the configuration of the logical device before the failure can be recovered. In the example shown in FIG. 1, the data can be double-written by the processing in step S706.

  After the logical device is generated, the configuration recovery unit 156A instructs the redundancy processing unit 151A to copy the data stored in the disk device 105 of the secondary computer to the disk device 105 of the primary computer. Is finished (step S907, step S908).

  The redundancy processing unit 151A that has received the instruction outputs an access request (read request) to the network disk driver unit 153A. As a result, data can be acquired from the disk device 105 of the secondary computer. Also, the redundancy processing unit 151A outputs an access request (write request) for the acquired data to the disk driver unit 152A. As a result, data is written to the disk device 105 of the main computer.

  By the processing in step S907, the same data is reflected in all the disk devices constituting the logical device. That is, the data written to the logical device by the alternative service providing unit 171 while the main computer is stopped is written to the disk device 105 of the main computer.

  FIG. 10 is a flowchart illustrating processing executed by the configuration recovery unit 156B (secondary system) according to the first embodiment of this invention.

  The configuration recovery unit 156B starts processing when called from the configuration recovery unit 156A (step S1001). The configuration recovery unit 156B stops the alternative service providing unit 171 (step S1002).

  The configuration recovery unit 156B sets the network disk driver unit 153B to permit access from the computer 101A to the disk device 105B, and ends the processing (steps S1003 and S1004).

  With the above processing, access from the other computer 101 to the network disk driver unit 153B is resumed, and access from the computer 101A becomes possible.

  When a failure occurs in the computer 101B, the configuration setting unit 155B is started after the computer 101B is restarted. As a result, access from the network disk driver unit 153A to the network disk driver unit 153B can be recovered. Thereafter, the redundancy processing unit 151A automatically copies data from the disk devices 105A to 105B.

  In the first embodiment, it has been described that the computer 101A is a primary computer that provides a service, and the computer 101B is a secondary computer. However, this is an example, and each computer 101 may have the same configuration. . As a result, a configuration is possible in which each other can be the other subsystem computer.

  According to the first embodiment, when a shared storage is configured using disk devices of a plurality of computers, even if a failure occurs in any computer, the alternative service providing unit 171 of the secondary computer Services can be continued using storage. That is, the alternative service providing unit 171 of the secondary computer can recognize that the redundant configuration of the shared storage is maintained.

  As a result, it is possible to construct a shared storage using a general computer without using a dedicated storage system, and it is possible to ensure resistance to failures.

  (Second embodiment)

  A second embodiment will be described. In the second embodiment, an actual system construction example will be described using a specific device configuration.

  Hereinafter, the difference from the first embodiment will be mainly described.

  FIG. 11 is a block diagram showing a configuration of a computer system according to the second embodiment of this invention.

  Since the configuration of the computer system is the same as that of the first embodiment, description thereof is omitted.

  In the second embodiment, the hardware configuration of the computer 101 is partially different. Specifically, in the second embodiment, the network interfaces 106A and 106B are Ethernet controllers 1101A and 1101B (Ethernet is a registered trademark, the same applies hereinafter). Since other hardware configurations are the same as those of the first embodiment, the description thereof is omitted.

  In the second embodiment, the software configuration of each computer 101 is different. Specifically, in the second embodiment, the redundancy processing units 151A and 151B are the software RAID function units 1111A and 1111B included in the OS, the network disk driver unit 153A is the iSCSI initiator 1112, and the network disk driver unit 153B is the iSCSI. It becomes the target 1113. Further, the computer 101B includes a network filter unit 1114 as a configuration for controlling access from the computer 101 via the network 180.

  The processing of the configuration setting unit 155A is different in the following points. In step S202, the configuration setting unit 155A acquires the name of the secondary computer. In step S203, the configuration setting unit 155A converts the acquired name of the secondary computer into an IPv4 address or an IPv6 address, and instructs to connect to the iSCSI target 1113 using the converted address. Other processes are the same as those in the first embodiment.

  The processing of the failure control unit 170 is different in the following points.

  In step S602, the failure control unit 170 stops the iSCSI target 1113 or instructs the network filter unit 1114 to discard the TCP / IP protocol packet transmitted to the iSCSI target 1113. Other processes are the same as those in the first embodiment.

  The processing of the configuration recovery unit 156B is different in the following points.

  In step S803, the configuration recovery unit 156B instructs the iSCSI target 1113 to restart, or instructs the network filter unit 1114 to cancel discarding of the TCP / IP protocol packet transmitted to the iSCSI target 1113. Other processes are the same as those in the first embodiment.

  According to the second embodiment, a shared storage can be constructed at low cost by using software RAID, iSCSI target, iSCSI initiator, and Ethernet.

  In addition, this invention is not limited to embodiment mentioned above, Various modifications are included. For example, the above-described embodiments have been described in detail for easy understanding of the present invention, and are not necessarily limited to those having all the configurations described. Further, a part of the configuration of an embodiment can be replaced with the configuration of another embodiment, and the configuration of another embodiment can be added to the configuration of an embodiment. Moreover, it is possible to add, delete, and replace other configurations for a part of the configuration of each embodiment.

  Further, each of the above-described configurations, functions, processing units, processing means, and the like may be realized using hardware by designing a part or all of them with, for example, an integrated circuit. In addition, each of the above-described configurations, functions, and the like may be realized using software by the processor interpreting and executing a program that realizes each function. Information such as programs, tables, and files for realizing each function can be stored in a recording device such as a memory, HDD, or SSD, or a recording medium such as an IC card, SD card, or DVD. Further, the control lines and information lines indicate what is considered necessary for the explanation, and not all the control lines and information lines on the product are necessarily shown. Actually, it may be considered that almost all the components are connected to each other.

Claims (12)

A computer system comprising a plurality of computers,
Each of the plurality of computers has a processor, a memory connected to the processor, a storage medium connected to the processor, and a network interface for connecting to another device via a network,
The plurality of computers includes one main computer that provides a service, and one or more sub computers that take over the service when a failure occurs in the main computer,
The main computer is
A first redundancy processing unit that generates a logical storage area using a storage area provided by the storage medium of each of the plurality of computers, and manages access to the generated logical storage area;
A first service providing unit that provides the service using the logical storage area;
The subsystem computer is
A failure control unit that monitors the main computer and executes a process for taking over the service when a failure of the main computer is detected;
A second service providing unit that provides the service using the logical storage area;
The first redundancy processing unit includes:
Generating a first logical storage area using a storage area provided by the storage medium of the main computer and the storage medium of the sub computer;
Generating first redundancy information including a correspondence relationship between the first logical storage area and the storage medium of the primary computer and the storage medium of the secondary computer constituting the first logical storage area;
Writing the first redundancy information to the storage medium of the primary computer and the storage medium of the secondary computer constituting the first logical storage area;
Providing the first logical storage area to the first service providing unit and instructing the first service providing unit to start the service;
When an access request is received from the first service providing unit, referring to the first redundancy information, accessing the first logical storage area,
The failure control unit activates the second service providing unit when detecting a failure of the primary computer,
The computer system according to claim 2, wherein the second service providing unit continues the service using data stored in the first logical storage area. The computer system according to claim 1,
The secondary computer has a second redundancy processing unit that generates the logical storage area and manages access to the generated logical storage area,
When the failure control unit detects a failure of the main computer, the failure control unit calls the second redundancy processing unit,
The second redundancy processing unit includes:
A new second logical storage area is generated using only the storage medium of the secondary computer, excluding the storage medium of the primary computer,
Generating second redundancy information including a correspondence relationship between the second logical storage area and the storage medium of the secondary computer constituting the second logical storage area;
Writing the second redundancy information to a storage medium of the secondary computer constituting the second logical storage area;
The failure control unit provides the second logical storage area to the second service providing unit, and instructs the second service providing unit to start the service,
When the access request is received from the second service providing unit, the computer system is configured to access the second logical storage area with reference to the second redundancy information. The computer system according to claim 2,
The failure control unit prohibits access from the main computer before calling the second redundancy processing unit. The computer system according to claim 2,
The first redundancy processing unit includes:
When a write request for the first logical recording area is received, the same data is written to the storage medium of the primary computer and the storage medium of the secondary computer that constitute the first logical storage area,
When a read request for the first logical storage area is received, data is read from the storage medium of the main computer constituting the first logical storage area;
The second redundancy processing unit includes:
When a write request for the second logical storage area is received, the same data is written to the storage medium of the subsystem computer that constitutes the second logical storage area,
When a read request to the second logical storage area is received, data is stored from the storage medium of the subordinate computer that constitutes the second logical storage area and that executes the second service providing unit. A computer system characterized by reading. The computer system according to claim 2,
The main computer has a recovery unit for recovering from a failure of the main computer,
The recovery unit is
Stopping the second service providing unit;
Obtaining the second redundancy information from a storage medium of the secondary computer constituting the second logical storage area;
Providing the second logical storage area to the first service providing unit and instructing the first service providing unit to start the service;
Calling the first redundancy processing unit;
The first redundancy processing unit includes:
A third logical storage area is generated using a storage area provided by the storage medium of the primary computer and the storage medium of the secondary computer,
Generating third redundancy information including a correspondence relationship between the third logical storage area and the storage medium of the primary computer and the storage medium of the secondary computer constituting the third logical storage area;
The third redundancy information is written to the storage medium of the primary computer and the storage medium of the secondary computer constituting the third logical storage area,
The recovery unit copies the data stored in the storage medium of the secondary computer constituting the third logical storage area to the storage medium of the primary computer constituting the third logical storage area. A computer system characterized by The computer system according to claim 1,
The main computer has a selection unit for selecting the sub computer,
The selection unit includes:
Selecting one or more secondary computers from the plurality of computers included in the computer system;
A computer system characterized by establishing a connection with the selected sub computer. A logical storage area management method in a computer system comprising a plurality of computers,
Each of the plurality of computers has a processor, a memory connected to the processor, a storage medium connected to the processor, and a network interface for connecting to another device via a network,
The plurality of computers includes one main computer that provides a service, and one or more sub computers that take over the service when a failure occurs in the main computer,
The method
A first step in which the main computer generates a first logical storage area using a storage area provided by the storage medium of the main computer and the storage medium of the sub computer;
The primary computer includes a first redundancy including a correspondence relationship between the first logical storage area and the storage medium of the primary computer and the storage medium of the secondary computer constituting the first logical storage area. A second step of generating the conversion information;
A third step in which the main computer writes the first redundancy information to a storage medium of the main computer and a storage medium of the sub computer constituting the first logical storage area;
A fourth step in which the main computer starts the service using the first logical storage area;
A fifth step of accessing the first logical storage area with reference to the first redundancy information when the main computer receives an access request to the first logical storage area; ,
And a sixth step of continuing the service using the data stored in the first logical storage area when the secondary computer detects a failure of the primary computer. Logical storage area management method. The logical storage area management method according to claim 7,
The sixth step includes
A seventh step in which the secondary computer generates a new second logical storage area using only the storage medium of the secondary computer, excluding the storage medium of the primary computer;
The sub system computer generates second redundancy information including a correspondence relationship between the second logical storage area and the storage medium of the sub system composing the second logical storage area. Steps,
A ninth step in which the secondary computer writes the second redundancy information to a storage medium of the secondary computer constituting the second logical storage area;
A tenth step in which the secondary computer starts the service using the second logical storage area;
The method further comprises:
An eleventh step of accessing the second logical storage area with reference to the second redundancy information when the secondary computer receives an access request to the second logical storage area; A logical storage area management method comprising: The logical storage area management method according to claim 8, comprising:
The seventh step includes a step of prohibiting access from the primary computer before the secondary computer generates the second logical storage region. The logical storage area management method according to claim 8, comprising:
In the fifth step, when the main computer receives a write request for the first logical recording area, the storage medium of the main computer and the secondary computer constituting the first logical storage area When the same data is written to the storage medium of the system computer, and when a read request for the first logical storage area is received, the data is read from the storage medium of the main computer constituting the first logical storage area,
In the eleventh step, when the secondary computer receives a write request for the second logical storage area, it is the same as the storage medium of the secondary computer that constitutes the second logical storage area. When data is written and a read request for the second logical storage area is received, the second logical storage area is configured and the data is read from the storage medium of the secondary computer that provides the service A logical storage area management method. The logical storage area management method according to claim 8, comprising:
The method further comprises:
The primary computer stopping the service executed in the secondary computer;
The primary computer obtaining the second redundancy information from a storage medium of the secondary computer constituting the second logical storage area;
The main computer starts the service using the second logical storage area; and
The main computer generates a third logical storage area using a storage area provided by the storage medium of the main computer and the storage medium of the sub computer; and
The main computer includes a third redundancy including a correspondence relationship between the third logical storage area and the storage medium of the main computer and the storage medium of the sub computer constituting the third logical storage area. Generating archiving information; and
The main computer writes the third redundancy information to a storage medium of the main computer and a storage medium of the sub computer constituting the third logical storage area;
The primary computer copies the data stored in the storage medium of the secondary computer constituting the third logical storage area to the storage medium of the primary computer constituting the third logical storage area And a logical storage area management method comprising the steps of: The logical storage area management method according to claim 7,
The first step includes
The primary computer selecting one or more secondary computers from the plurality of computers included in the computer system;
And a step of establishing a connection between the main computer and the selected sub computer.

Images