JP2014137622A - Computer system and automatic detection method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To automatically detect a shared disk and reduce problems such as setting errors, in a computer system including the shared disk.SOLUTION: A detection instruction unit 131 of a management computer 101 instructs a server computer 102 within node list information 141 to perform a detection operation for a shared disk. A shared disk detection unit 161 of a plurality of server computers 102 including a disk management table 183 storing a result of the detection operation writes a unique symbol to each shared disk according to the instruction by the detection instruction unit 131, reads the written unique symbol from each shared disk, and stored it in the disk management table 183. Each of the plurality of server computers 102 transmits the stored content to the management computer 101, and the management computer 101 stores the content of the disk management table 183 in a disk management master table 145 for each unique symbol.

Description

  The present invention relates to a storage system having a possibility of being shared by a plurality of computers, and particularly relates to an automatic detection technique for automatically detecting a shared storage.

  There is a technology for making a single storage device or a single virtual disk configured in the storage device accessible from a plurality of computers using a network classified as a storage area network. In order to maintain the consistency of recorded data for disks shared by multiple computers in this way, multiple computers should not read and write to the same part of the same disk at the same time. Need to control. Patent Document 1 is an example of such control. When a file open operation is performed by a client computer, the client computer 1 uses a channel switch to record a disk via a channel switch based on actual file information acquired by the client computer. By directly accessing the device to perform input / output processing related to file operations, and managing the disk device with a child server computer, a plurality of client computers 1 can handle different real files on the same virtual file. A method is disclosed that allows I / O processing to be performed simultaneously.

JP 2007-293634 A

  In order to enable the control as described above, it is necessary to correctly set information about which computer shares which disk with respect to a group of computers sharing the disk. Using the information set in this way, for example, when the disk 1 of the computer A and the disk 2 of the computer B are set as a shared physical single disk, these disks are used. The software performs exclusive control so that computers A and B do not write simultaneously to the same part of disks 1 and 2, respectively.

  However, when the number of computers and shared disks increases, manually setting information on shared disks is a labor-intensive operation, and there is an inconvenience that setting errors are likely to be caused.

  An object of the present invention is to provide a storage system and a shared disk automatic detection method capable of solving the problems in the case of using the above-described prior art and allowing components in the system using the disk to automatically find the shared disk. is there.

  To achieve the above object, according to the present invention, a computer system includes a plurality of computers and a storage that can be shared by the plurality of computers, and each of the plurality of computers is connected to the computer. A computer system configured to include a detection symbol operation unit that records a unique symbol and outputs the recorded unique symbol from storage.

  In order to achieve the above object, in the present invention, a computer system shared storage automatic detection method, the computer system includes a plurality of server computers, a shared storage that can be shared by a plurality of server computers, A management computer connected to a plurality of server computers via a network, and each of the plurality of server computers records a unique symbol in the shared storage connected to the server computer, and shares the recorded unique symbol Provide an automatic detection method of shared storage output from storage.

  The present invention eliminates the need to manually set the existence of shared storage in an environment having storage shared with a large number of computers, and reduces the complexity of settings for programs that control shared storage and the possibility of setting errors. be able to.

It is a block diagram which shows one structure of the computer system of a 1st Example. It is a flowchart explaining an example of a user's operation in a 1st Example. It is a flowchart explaining an example of the process which a detection instruction | indication part performs in a 1st Example. It is the flowchart explaining an example of the process which the object disk detection part in a 1st Example performs. It is the flowchart explaining an example in the process which the detection symbol operation part in a 1st Example performs. It is the flowchart explaining an example of the process which the management server report part in a 1st Example performs. It is the flowchart explaining an example of the process which the disc information receiving part in a 1st Example performs. It is a figure which shows the structural example of the disk management table in a 1st Example. It is a figure which shows the structural example of the disk management master table in a 1st Example. It is a block diagram which shows one structure of the computer system of a 2nd Example. It is the flowchart explaining an example of the process of the file system client part in a 2nd Example. It is the flowchart explaining an example of the process of the metadata management part in a 2nd Example.

  Hereinafter, various embodiments of the present invention will be described in detail with reference to the drawings. A preferred embodiment of the present invention is a computer system including a plurality of computers and a shared disk, and the computer system includes: A detection instruction that includes one management computer, one or more server computers, and one or more shared disks shared between the server computers, and the management computer instructs each server computer to perform a shared disk detection operation A disk information receiving unit that receives a disk information detection result from each server computer and clarifies the existence of a shared disk, node list information that holds a list of server computers as an input from the administrator, and management Disk attribute information that holds the attribute information of the target disk for automatic detection, which is input from the administrator, and the disk management manager that holds the result of shared disk automatic detection Each server computer has a target disk detection unit for creating a list of disks to be automatically detected, and each server computer records a symbol unique to the disk connected to its own computer. Each server computer outputs a unique symbol recorded on a disk connected to its own computer, a management server reporting unit for notifying the management computer of the processing result of the detected symbol operation unit, and a management A disk attribute condition that is a copy of the disk attribute information input to the computer, a node number uniquely assigned to each server computer notified from the management computer, and a list of all disks accessible from each server computer Each server computer has certain disk information and a disk management table that records the processing progress of the detection symbol operation unit. A unique symbol is written to each shared disk by an instruction from the instruction unit, and each server computer reads the unique symbol from each shared disk by an instruction from the detection instruction unit and holds it in the disk management table. The computer system is configured to transmit the contents of the management table to the management computer, and the management computer holds the contents of the disk management table in the disk management master table for each unique symbol.

FIG. 1 is a block diagram showing an example of the configuration of a computer system including the storage system of the first embodiment.
As shown in the figure, the computer system of this embodiment comprises a management computer 101, server computers 102A and 102B, and disk devices 103A and 103B. The management computer 101 and the server computers 102A and 102B are connected using the network 104. As the network 104, for example, a local area network (LAN) is conceivable. However, in this embodiment, the type and connection form of the network 104 are not limited to LAN. Server computers 102A and 102B are connected to disk devices 103A and 103B using a network 105. The network 105 is, for example, a SAN (Storage Area Network). However, as described above, the connection format of the network 105 is not limited thereto. Hereinafter, when the server computers 102A and 102B are not distinguished from each other, they are referred to as server computers 102. In FIG. 1, there are two server computers and two disk devices, but there may be three or more server computers and disk devices. Both the management computer 101 and the server computer 102 have a general-purpose computer configuration, and include a keyboard, a mouse, an input / output unit of a display device, etc., which are not shown.

First, the hardware configuration of the management computer 101 in FIG. 1 will be described.
The management computer 101 includes a memory 110, a network interface 111, a processor 112, and a storage 113. The processor 112 executes a program stored in the memory 110. The function of the management computer 101 can be realized by the processor 112 executing the program. Hereinafter, when the processing of the management computer is described with the program as the subject, it indicates that the program is being executed by the processor 112. The memory 110 stores a program executed by the processor 112 and data necessary for executing the program. The memory 110 may be a semiconductor memory such as a DRAM. The program and data stored in the memory 110 will be described later. The network interface 111 is a device for communicating with the server computers 102A and 102B. The storage 113 is a storage device that stores information such as an operating system.

Next, the hardware configuration of the server computer 102 in FIG. 1 will be described.
The server computer 102A includes a memory 161A, a network interface 162A, a processor 163A, and a storage interface 164A. The processor 163A executes a program stored in the memory 161A. The function of the server computer 102 can be realized by the processor 163A executing the program. Hereinafter, when the processing of the server computer 102A is described using the program as the subject, it indicates that the program is being executed by the processor 163A. The memory 161A stores a program executed by the processor 163A and data necessary for executing the program. The memory 161A may be a semiconductor memory such as a DRAM. The program and data stored in the memory 161A will be described later. The network interface 162A is a device for communicating with the management computer 101 and the server computer 102B.

The storage interface 164A is a device for communicating with the disk devices 103A and 103B. Although not shown, the server computer 102A may be provided with a storage device for storing information such as an operating system. Since the configuration of the server computer 102B is the same as that of the server computer 102A, description thereof is omitted.
The disk devices 103A and 103B store detection symbols 190A and 190B, which will be described later, and attribute information 191A and 191B.

Next, the software configuration of the management computer 110 will be described.
The management computer 110 has a disk detection management unit 120 and basic software 121 in the memory 110. The disk detection management unit 120 instructs the server computer 102 to execute a procedure for automatically detecting a shared disk and collects data, so that a detection instruction unit 131, a disk information reception unit 132, node list information 141, disk attribute information 142 are collected. And a disk management master table 145. The detection instructing unit 131 outputs a request for instructing the server computers 102A and 102B to start detecting a disk, which will be described in detail later with reference to FIG. The disk information receiving unit collects disk information detected by the server computers 102A and 102B and detects the presence of a shared disk, which will be described in detail later with reference to FIG. Note that the detection instruction unit 131 and the disk information receiving unit 132 of the disk detection management unit 120 are stored in the memory 110 and configured as program software that is executed by the processor 112 that is a processing unit. It can also be said that the instruction unit 131 and the disk information receiving unit 132 are functional configuration blocks of the processor 112 which is a processing unit.

The node list information 141 is data input from the user, and holds a list of computers that are targets for automatic disk detection. The disk attribute information 142 is data input from the user, and holds the attribute of the disk to be detected. The disk management master table 145 is data for managing a list of disks detected by each server computer.
The basic software 121 is a known operating system. Although not shown, the management computer 101 may execute other software simultaneously.

Next, the software configuration of the server computer 102A will be described.
The server computer 102A has a shared disk detector 170A and basic software 189A in the memory 161A. The shared disk detection unit 170A performs a part of the procedure for detecting the shared disk from the disks connected to the server computer 102A. The target disk detection unit 171A, the detection symbol operation unit 172A, the management server report It includes a section 173A, a disk attribute condition 180A, a node number 181A, total disk information 182A, and a disk management table 183A. The target disk detection unit 171A selects a disk to be automatically detected and will be described in detail later with reference to FIG. The detection symbol operation unit 172A performs disk input / output for detecting the shared disk, and will be described in detail later with reference to FIG. The management server reporting unit 173A reports the result of the detected symbol operation unit 172A to the management server which is the management computer 101, and will be described in detail later with reference to FIG.

  The target disk detection unit 171A, the detection symbol operation unit 172A, and the management server reporting unit 173A in the shared disk detection unit 161A are configured as program software stored in the memory 102A and executed by the processor 163A that is a processing unit. Therefore, it can be said that the target disk detection unit 171A, the detection symbol operation unit 172A, and the management server reporting unit 173A are functional configuration blocks included in the processor 163A as a processing unit.

The disk attribute condition 180A is a copy of the disk attribute information 142 of the management computer 101. The node number 181A is information storing a unique number notified from the management computer 101 to each server computer. The total disk information 182A is temporary data stored by the target disk detector 171A during the operation. The disk management table is created by the detection symbol operation unit 172A and stores data to be reported to the management computer 101 by the management server reporting unit 173A. An example thereof is shown in FIG. The basic software 189A is a known operating system. Although not shown, the server computer 102A may execute other software simultaneously.
Since the software configuration of the server computer 102B is the same as that of the server computer 102A, description thereof is omitted.

  Hereinafter, processing of each component block of the computer system according to the first embodiment will be described with reference to FIGS.

FIG. 2 is a flowchart for explaining a user's operation using the system of the first embodiment.
The user of the system of the first embodiment first inputs the node list information 141 of the disk detection management unit 120 in the management computer 101 using the input / output unit and the network 104 (step S201). Next, in step S202, the disk attribute information 142 is similarly input in the management computer 101. Steps S201 and S202 can be realized by, for example, a method of preparing a file in which a node list is written in a file system provided by the operating system. In step S203, the detection instruction unit 131 is activated. For example, when steps S201 and S202 prepare files, the detection instruction unit 131 is activated while designating these two files as inputs in step S203.

FIG. 3 is a flowchart illustrating processing executed by the detection instruction unit 131 of the disk detection management unit 120 in the first embodiment.
The detection instruction unit 131 starts the flow process of FIG. 3 by a user operation (step S301). At this time, the instruction includes the node list information 141 and the disk attribute condition 142 described above.

  The detection instruction unit 131 confirms whether or not an unprocessed node is included in the node list information 141 (step S302). Here, an unprocessed node means a server computer 102 that has not been processed as a target of the subsequent steps S303 to S305. For example, when the node list information 141 is a file in which the name of the server computer is written on each line, the detection instruction unit 131 records the number of lines already processed, and can determine the condition in step S302 by comparing with the number of lines in the file It is.

  If the condition determination in step S302 is true, the detection instruction unit 131 selects one unprocessed node from the node list information 141 (step S303). For example, in the situation where the node list information 141 is a file in which the name of the server computer is written on each line, and the processing of N lines has already been completed, the server computer written on the N + 1 line is selected.

  The detection instruction unit 131 transmits the disk attribute condition 142, the node number, and the detection command to the server computer selected in step S303 (step S304). Here, the node number is an integer uniquely given to each server computer described in the node list information 141. For example, when the node list information 141 is a file in which the name of the server computer is written in each line, a unique number can be given if the node number is the number of the line in which the name of the server computer is written.

The detection instruction unit 131 stores the server computer selected in step S302 as processed (step S305). For example, when the node list information 141 is a file in which the name of the server computer is written in each line, this step can be realized by adding 1 to the number of processed lines.
The detection instruction unit 131 waits for a write completion notification from the server computer selected in step S302 (step S306).

As described above, if steps S303 to S306 are performed for all server computers described in the node list information 141, the condition determination in step S302 is false.
In response to this, the detection instruction unit 131 sends a disk information report command to all the server computers described in the node list information 141 (step S307). The operation of the server computer having received this will be described later with reference to FIGS.
The detection instruction unit 131 activates the disc information receiving unit 132 (step S308). The operation of the disc information receiving unit will be described later with reference to FIG.

  Through step S307 and step S308, each server computer 102 reports the name of the disk in each server computer and the detection symbol written as disk data for the disk accessible from the respective computer. When a disk with the same detection symbol is reported from a plurality of server computers, it can be detected that the disk is a shared disk. The contents of this detection signal will be described later.

  FIG. 4 is a flowchart for explaining processing executed by the target disk detection units 171A and 171B in the shared disk detection units 161A and 161B of the server computers 102A and 103B in the first embodiment. Since the operations of the target disk detection units 171A and 171B are the same, they are collectively referred to as the target disk detection unit 171. In the following, similarly, A or B is added in FIG. 1, and those without A or B in the text indicate that A and B are performing the same processing.

The target disk detection unit 171 starts processing when receiving a detection command from the management computer 101 (step S401). This detection command includes a disk attribute condition and a node number.
The target disk detection unit 171 stores the received disk attribute condition as a disk attribute condition 180 that is information on the memory (step S402).
The target disk detection unit 171 stores the received node number as the node number 181 that is information on the memory (step S403).

  Next, the target disk detection unit 171 uses the function of the basic software 189 to obtain a list of all disks accessible from the computer 102 and their attributes. The list of acquired disks and their attributes are recorded in all disk information 182 (step S404). Here, the disk list is the file name of the disk device disk used when accessing from software, and the attribute is the name of the disk controller of the disk device. The all disc information 182 has a flag (value of 0 or 1) for recording whether or not the processing of steps S406 to S409 is completed for each detected disc. Hereinafter, this flag is referred to as a “processed flag”. At the end of step S404, this processed flag is 0 for any disk.

The target disk detection unit 171 determines whether or not there is a disk whose processed flag is zero in the all disk information 182 (step S405).
If the determination in step S405 is true, the target disk detection unit 171 selects one disk whose processed flag is 0 from all the disk information (step S406).

  The target disk detection unit 171 compares the attribute of the disk selected in step S406 with the attribute condition 180, and determines whether the condition is satisfied (step S407). For example, in the attribute condition 180, the condition “the manufacturer of the disk controller is company A” is written, and if the attribute of the disk selected in step S406 (obtained in step S404) is company A, the determination in step S407 is true. It is false if it is other than A company.

  If the determination in step S407 is true, the target disk detection unit 171 adds the name of the disk to the disk management table 182 (step S408). If the determination in step S407 is true, the target disk detection unit 171 skips this process.

  The target disk detection unit 171 sets the processed flag of the disk selected in step S406 to 1, and returns to step S405 (step S409). A series of processing from step S405 to S409 is processing for listing all disks connected to the server computer 102 that satisfy the disk attribute condition 180 and recording them in the disk management table 182. When the attribute survey for all the disks is completed, the determination in step S405 becomes false.

If the determination in step S405 is false, the target disk detection unit 171 activates the detection symbol operation unit 172 (step S410).
The process of the target disk detection unit 171 ends here (step S411).

  FIG. 8 shows a configuration example of the disk management table 183 described above. The disk management table 183 has a structure that can record an arbitrary number of lines including three items of a disk name, a write detection symbol, and a read detection symbol. At the time when the processing flow of the target disk detection unit 171 is completed, a line in which only the disk name is recorded is recorded, and the number of lines is equal to the number of disks that satisfy the condition of the disk attribute condition 180. The writing detection symbol and the reading detection symbol will be described later.

FIG. 5 is a flowchart illustrating processing executed by the detection symbol operation unit 172 in the shared disk detection unit 161 of the server computer 102 in the first embodiment.
The detection symbol operation unit 172 starts processing in step S410 of the target disk detection unit (step S501).

The detection symbol operation unit 172 determines whether or not there is a disk in which the write detection symbol is not written in the disk management table 183 (step S502). Step S502 means that all the disks entered in the disk management table 183 described in FIG. 8 are repeated for each disk, and the processes of steps S503 to S507 are performed.
The detection symbol operation unit 172 selects one disk in which the write detection symbol is not entered from the disk management table 183 (step S503).

  The detection symbol operation unit 172 creates a detection symbol using the node number 181 and the name of the selected disk (step S504). The detection symbol is a symbol that guarantees that the detection symbol is always different if the disk name or node number is different. For example, if the node number is represented by a character string indicating a decimal number, the disk name is also represented by a character string, and the concatenation of these two character strings is used as a detection symbol, the above detection symbol is generated. Is possible. In this example, for example, if the disk name existing in the server computer 102 with the node number of 1 is the disk with the name sda, the detection symbol is “001 sda” on the assumption that the node number is aligned with 3 digits.

  The detection symbol operation unit 172 writes the detection symbol generated in step S504 in a predetermined area of the disc selected in step S504 (step S506). For example, if the disk selected in step S504 is the disk device 103A in FIG. 1 and the name of this disk is sda and the node number 181A of the server computer 102A is 1 in the server computer 102A, detection is written to the disk device 103A. The symbol 190A becomes “001sda”. The predetermined position where the detection symbol is written is an arbitrary place agreed in the entire system, and for example, the head of the disk can be used.

  The detection symbol operation unit 172 writes the detection symbol generated in step S504 in the write detection symbol column for the disk selected in step S503 in the disk management table 183 (step S507).

The detection symbol operation unit 172 returns to step S502, and performs the processing of steps S503 to S507 for all the disks in the disk management table.
When the determination in step S502 is false, the detected symbol operation unit 172 transmits a write completion notification to the management computer 101 (step S508).

  The target disk detection unit 171 of the shared disk detection unit 161 of the server computer 102 and the steps S501 to S507 of the detection symbol operation unit 172 operate in response to the instruction of the management computer 101 in step S304. The write completion notification in step S508 corresponds to step S306 in the detection instruction unit 131 of the management computer 101.

  After the management computer 101 instructs all server computers 102 to execute the steps S501 to S508 of the target disk detection unit 171 and the detection symbol operation unit 172, the step S307 of the detection instruction unit 131 causes all the server computers 102 to Issue a report order. The detected symbol operation unit 172 waits for reception of this disk report command after step S508 (step S509).

  The detection symbol operation unit 172 reads the detection symbols for all the disks in the disk management table 183 and records them in the read detection symbols of the disk management table 183. For example, in FIG. 1, the disk accessed by the server computer 102A with the name sda is the disk device 103A, and this disk is also shared with the server computer 102B, and the server computer 102B can access the disk device 103A with the name sdb. Suppose that the network 105 is configured as described above.

  In this case, the server computer 102A that has received the instruction from the detection instruction unit of the management computer 101 writes “001sda”, which is the detection symbol, in the detection symbol 190A of the disk device 103A. Thereafter, the server computer 102B that has received an instruction from the detection instruction unit of the management computer 101 overwrites the detection symbol “002sdb” on the detection symbol 190A of the same disk device 103A. As a result, the server computer 102A reads in step S510 a detection symbol different from the detection symbol written in the disk sda in step S506.

  Such a situation is illustrated in the first row 801 of FIG. Note that, as in the second line 802 in FIG. 8, the written symbol may not be overwritten and the written symbol may be read as it is.

In the server computer 102, the detected symbol operation unit 172 activates the management server reporting unit 173 (step S511).
This completes the processing of the detected symbol operation unit 172 of the server computer 102 (step S512).

FIG. 6 is a flowchart illustrating processing executed by the management server reporting unit 173 in the shared disk detection unit 161 of the server computer 102 in the first embodiment.
The management server reporting unit 173 starts processing in response to step S511 of the detection symbol operation unit 172 (step S601).

The management server reporting unit 173 transmits the node number 181 to the management computer 101 (step S602).
The management server reporting unit 173 transmits a set of a disk name and a read detection symbol for all the disks in the disk management table 183 (step S603).
These steps S602 and S603 correspond to steps S703 and S704 of the processing of the disk information receiving unit 132 of the disk detection management unit 120 of the management computer 101 described next with reference to FIG.

FIG. 7 is a flowchart for explaining processing executed by the disc information receiving unit 132 in the first embodiment.
The disk information receiving unit 132 is activated in response to step S308 of the detection instruction unit 131.

The disk information receiving unit 132 determines whether there is an unreceived node among the nodes in the node list information 141 (step S702). This step means that the processing from step S703 to step 710 is repeated for all server computers 102.
The disk information receiving unit 132 receives the node number from the server computer 102 (step S703). This is transmitted in step S602 of the management server report unit 173 described above.

The disk information receiving unit 132 receives a list of disk names and detection symbols from the server computer 102 (step S704). This is transmitted in step S603 of the management server reporting unit 173 described above.
The disk information receiving unit 132 determines whether an entry is included in the list received in step S704 (step S705). If the determination is false, the process returns to step S702. If the determination is true, the process proceeds to step S706.

The disc information receiving unit 132 extracts one disc name / detection symbol pair from the list received in step S704 and deletes it from the list (step S706).
The disk information receiving unit 132 determines whether or not the detected symbol acquired in step S706 exists in the disk management master table 145 (step S707).

As shown in FIG. 9, the disk management master table 145 of the disk detection management unit 120 of the management computer 101 is a set of nodes that can be accessed with the disk name corresponding to the detection symbol and the disk name using the detection symbol as a key. It is a table that summarizes.
If the determination in step S707 is true, the disk information receiving unit 132 records the set of the node name received in step S703 and the disk name selected in step S706 for the row found in step S707 (step S708).

If the determination in step S707 is false, the disk information receiving unit 132 adds the detection symbol line acquired in S706 to the master disk management information (step S709).
The disk information receiving unit 132 records the combination of the node name received in S703 and the disk name acquired in S706 in the row added in S709 (step S710).

  The procedure from step S706 to step S710 serves to organize the disk name using the read detection symbol reported from each server computer 102 as a key. As described above, for a shared disk, all server computers 102 that share the disk read the detection symbol written last. In other words, the disk that has read the same detection symbol is the same shared disk. Therefore, the disk management master table 145 writes the detection symbol received from each server computer 102 to the row header column, adds a new row when a new detection symbol is received, and receives a detection symbol that already exists. Processing to add a pair of node and disk name to the table.

  In the system of the first embodiment described above, when the processing of the disk information receiving unit 132 of the disk detection management unit 120 of the management computer 101 is completed, the disk management master table 145 is in a state of disk sharing in the server computer group. Represents. For example, in the example of FIG. 9, the disk sda of the server computer 102 with the node number 1 and the disk sdb of the server computer with the node number 2 are the same shared disk, and the disk sdc of the server computer 102 with the node number 1 is shared. It is shown that no other server computer is present.

  The storage system and shared disk automatic detection method of this embodiment eliminates the need to manually set the existence of a shared disk in an environment having a large number of server computers and shared disks, and allows setting of a program for controlling the shared disk. It is possible to reduce the complexity and the possibility of setting mistakes, and it is possible to realize a computer system in which components in the system using the disk can automatically find the shared disk.

  FIG. 10 is a block diagram showing a configuration of the computer system of the second embodiment. In this embodiment, after the automatic detection of the shared disk explained in the computer system of the first embodiment is completed and the disk management master table is completed in the management computer, a shared file system is constructed using the disk management master table. This is an example.

  The management computer 101 and the server computers 102A and 102B in the computer system of FIG. 10 have all the same components as the computer system of the first embodiment shown in FIG. 10, the detection instruction unit 131, the disk information receiving unit 132, the node list information 141, and the disk attribute information 142, which are the components of the disk detection management unit 120 described in FIG. 1, are not drawn. These are the same operations as the computer of the first embodiment, and are not shown in the figure, and the description is omitted in this embodiment. Similarly, in the server computer 102, the target disk detection unit 171, detection symbol operation unit 172, management server report unit 173, disk attribute condition 180, node number 181, all disk information 182, which are components of the shared disk detection unit 170, Although the disk management table 183 is not depicted in FIG. 2, these actually exist, and the movement thereof is exactly the same as that of the computer of the first embodiment, so that the description thereof is omitted.

Hereinafter, the computer system of the second embodiment shown in FIG. 10 will be described focusing on the differences from the computer system of the first embodiment.
In the figure, a shared file system management unit 1001 in the memory 161 of the server computer 102 is software for constructing a file system using the server computer 102 and the disk device 103.

  The shared file system management unit 1001 in the memory 110 of the management computer 101 has a metadata management unit 1002. The metadata management unit 1002 has a role of receiving a request from the file system client unit 1011 of the server computer 102 to be described later and returning data indicating the location of the file. The processing content of the metadata management unit 1002 will be described later with reference to FIG.

  The metadata information 1003 stored in the storage 113 of the management computer 101 stores information such as the file location. In the configuration of the computer system of FIG. 10, the metadata information 1003 is arranged in the storage 113. However, when the management computer 101 has an accessible external storage connected by a storage area network or the like, the external storage is stored in the external storage. Metadata information may be arranged.

  The shared file system unit 1010 existing in the memory 161 of the server computer 102 receives a request from the user, performs processing, and then returns the result to the user. There is a disk access unit 1012 that performs the following access. The operation of the file system client unit 1011 will be described later with reference to FIG. The disk access unit 1012 is the same as the operation performed by a computer program generally called a file system, and its processing is well known and will not be described here.

FIG. 11 is a flowchart showing processing of the file system client unit 1011 in the computer system of the second embodiment.
The file system client unit 1011 starts processing in response to a file access request from the user (step S1101).

  The file system client unit 1011 inquires of the metadata management unit 1002 of the shared file system management unit 1001 of the management computer 101 about the location where the file has been requested by the user (step S1102).

  The file system client unit 1011 receives a list of the names of the server computers in which the files exist and the names of the disks storing the files in the computers (step S1103). Here, the disk may be shared by multiple server computers, so the same data may be accessible from multiple server computers. In this case, multiple server computer name / disk name pairs are received. Is done.

  The file system client unit 1011 confirms whether or not its server computer name is included in the list of computer name and disk name pairs received in step S1103 (step S1104). The determination in step S1104 checks whether or not an access target file exists on a disk that is directly accessible from the server computer.

If the determination in step S1104 is true, the server computer can reach the file to be accessed by directly referring to the disk. Therefore, the file system client unit 1011 transmits an access request to the disk access unit 1012 to read or write the file (step S1105).
The file system client unit 1011 receives the result of step S1105 (read data and an integer value indicating success or failure).

If the determination in step S1104 is false, the file system client unit 1011 transmits a file access request via the network 104 to one of the server computers included in the list (step S1107).
The file system client unit 1011 receives the result from the network (step S1108).

  In any case of the determination in S1104, the file system client unit 1011 receives the file access result in S1106 or S1108. In step S1109, the file system client unit 1011 returns this result to the user and ends the process.

FIG. 12 is a flowchart showing the processing of the metadata management unit in the management computer in the computer system of the second embodiment.
The metadata management unit 1002 of the management computer 101 starts processing in response to a request from the file system client (step S1201).

The metadata management unit 1002 determines whether or not the request from the file system client is to create a new file (step S1202).
If the determination in step S1202 is true, the metadata management unit 1002 selects an arbitrary row in the disk management master table 145 (step S1203). This means selecting a storage location for a newly created file. In addition to simply selecting a random location, there may be various modifications such as selecting the storage location on a disk close to the client as judged from the node name.

The metadata management unit 1002 creates metadata management information 1003 corresponding to a newly created file in the storage 113 (step S1204).
The metadata management unit 1002 writes the detection symbol written in the row of the disk management master table selected in step S1203 to the metadata management information created in step S1204 (step S1206).

If the determination in step S1202 is false, the request is for reading or writing an existing file. In this case, the metadata management unit 1002 reads the metadata information of the request target file (step S1207).
Further, the metadata management unit 1002 reads the detection symbol of the disk holding the data from the metadata (step S1208).

  Regardless of the determination result of step S1202, the detection symbol to be accessed is obtained at this point. The metadata management unit 1002 refers to the disk management master table 145 to obtain a set of a server computer and a disk name that can access the disk (step S1208).

  The metadata management unit 1002 returns a list of combinations of server computers and disk names to the file system client unit 1011. This reply message corresponds to step S1103 of the file system client unit 1011 described in FIG.

  The computer system of the second embodiment described above is an example of a computer system that constructs a storage system and a shared file system using a disk management master table created using the configuration of the computer system of the first embodiment. It is. By writing the detection symbol written in the disk management master table to the metadata information, it is not necessary to modify the metadata even if there is a change in the server computer or disk connection information. The file system can be managed automatically with the correct configuration automatically by only updating.

  In addition, this invention is not limited to the Example 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 one embodiment can be replaced with the configuration of another embodiment, and the configuration of another embodiment can be added to the configuration of one embodiment. Further, it is possible to add, delete, and replace other configurations for a part of the configuration of each embodiment.

  In addition, each of the above-described configurations, functions, processing units, processing means, etc. exemplifies a case in which some or all of them are realized by software by the processor interpreting and executing a program that realizes each function. However, some or all of them may be realized using hardware by designing with an integrated circuit. In addition, information such as programs, tables, and files for realizing each function can be stored in another recording medium instead of being stored in the memory or the storage. Moreover, although a disk is illustrated as an example of shared storage, other types of storage may be used.

101 management computer 102 server computer 103 disk device 104, 105 network 110 memory 111, 162 network interface 112, 163 processor 113 storage 120 disk detection management unit 121, 189 basic software 131 detection instruction unit 132 disk information reception unit 141 node list information 142 Disk attribute information 145 Disk management master table 164 Storage interface 170 Shared disk detection unit 171 Target disk detection unit 172 Detection symbol operation unit 173 Management server reporting unit 180 Disk attribute condition 181 Node number 182 Total disk information 183 Disk management table 190 Detector number 191 Attribute information 1001 Shared file system management unit 1002 Metadata management unit 1010 Shared file Le system unit 1011 file system client section 1012 disk access unit

Claims (15)

A computer system,
With multiple calculators,
A storage that can be shared by the plurality of computers,
Each of the plurality of computers includes a detection symbol operation unit that records a unique symbol in the storage connected to the computer and outputs the recorded unique symbol from the storage.
A computer system characterized by that. The computer system according to claim 1,
A management computer connected to the plurality of computers via a network;
The plurality of computers are a plurality of server computers;
The storage is one or more shared disks shared by the plurality of server computers.
A computer system characterized by that. The computer system according to claim 2,
Each of the management computer and the plurality of server computers has a processing unit, a storage unit, and an interface unit connected to the network,
The processing unit of each of the plurality of server computers includes the detection symbol operation unit.
A computer system characterized by that. The computer system according to claim 3,
The storage unit of the management computer is
Storing node list information holding a list of the plurality of server computers, disk attribute information of a disk to be detected, and a disk management master table holding a detection result of the shared disk;
The processing unit of the management computer is
Instructing each of the plurality of server computers to detect the shared disk, receiving a detection result from each of the plurality of server computers, and storing the result in the disk management master table.
A computer system characterized by that. A computer system according to claim 4, wherein
The processing unit of each of the plurality of server computers includes a target disk detection unit that creates a list of the detection target disks,
The storage unit of each of the plurality of server computers is
The disk attribute information, a node number uniquely assigned to each of the plurality of server computers, a list of disks accessible from the server computer, and a disk management table for recording a detection result of the target disk detection unit. Remember,
A computer system characterized by that. The computer system according to claim 5,
The detection symbol operation unit of each of the plurality of server computers is
In accordance with an instruction from the processing unit of the management computer, the unique symbol is written to each shared disk, the written unique symbol is read from each shared disk, held in the disk management table, and the held disk Sending the contents of the management table to the management computer;
A computer system characterized by that. A computer system according to claim 6, wherein
The processing unit of the management computer is
The content held in the disk management table received from each of the plurality of server computers is held in the disk management master table for each unique symbol.
A computer system characterized by that. The computer system according to claim 7,
The processing unit of the management computer is
Holding a metadata information for managing the storage location of the file, and having a shared file system management section that returns a list of the storage location of the file in response to a request;
The processing unit of each of the plurality of server computers is
Receiving a file access request from a user, and having a shared file system unit for inputting and outputting information to and from the shared disk connected to each of the plurality of server computers.
A computer system characterized by that. A computer system according to claim 8, wherein
The shared file system management unit
When the file access request is received from the shared file system unit, the metadata information is referred to, the unique symbol of the disk on which the requested file is stored is obtained, the disk management master table is referred to, and the request is made. Create a list of server computers that can access the disk where the file is stored, and send it to the shared file system unit,
The shared file system unit
Receiving a list of the accessible server computers from the shared file system management unit, and performing file access through the server computers described in the list;
Computer system. A method for automatically detecting shared storage in a computer system,
The computer system includes a plurality of server computers, a shared storage that can be shared by the plurality of server computers, and a management computer connected to the plurality of server computers via a network,
Each of the plurality of server computers records a symbol unique to the shared storage connected to the server computer, and outputs the recorded unique symbol from the shared storage;
An automatic detection method characterized by that. The automatic detection method according to claim 10,
The shared storage is one or more shared disks shared by the plurality of server computers,
The management computer is
Node list information that holds a list of the plurality of server computers, disk attribute information of a disk to be detected, and a disk management master table that holds a detection result of the shared disk;
Instructing the shared disk detection operation to each of the plurality of server computers, receiving detection results from the plurality of server computers, and holding them in the disk management master table.
An automatic detection method characterized by that. The automatic detection method according to claim 11, comprising:
Each of the plurality of server computers is
Holding the disk attribute information, a node number uniquely assigned to each of the plurality of server computers, a list of disks accessible from the server computer, and a disk management table for recording the detection results;
According to the command of the management computer, the unique symbol is written to each shared disk, the written unique symbol is read from each shared disk, held in the disk management table, and the contents of the held disk management table To the management computer,
An automatic detection method characterized by that. An automatic detection method according to claim 12, comprising:
The management computer is
Holding the contents of the disk management table received from each of the plurality of server computers in the disk management master table for each unique symbol;
An automatic detection method characterized by that. The automatic detection method according to claim 13,
The management computer is
When metadata information for managing a file storage location is held and a file access request is received from the plurality of server computers, the metadata information is referred to and the unique symbol of the disk on which the requested file is stored is changed. Obtaining, referring to the disk management master table, creating a list of server computers that can access the disk in which the requested file is stored, and transmitting to the plurality of server computers;
An automatic detection method characterized by that. The automatic detection method according to claim 14,
Each of the server computers
When receiving the file access request from the user and inputting / outputting information to / from the shared disk connected to each of the plurality of server computers,
File access is performed through the server computer described in the list of accessible server computers received from the management computer.
An automatic detection method characterized by that.

Images