JP2008186141A - Data management method, data management program, data management system and configuration management device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To implement load balancing throughout a data management system. <P>SOLUTION: A data management method is provided for the data management system 2 connecting an access request device 21, a pair of data storage devices 23 and data access devices 22, and a configuration management device 24 for storing configuration information 25. In response to the input of a request to read a given file, the access request device 21 issues a read request to one of the original data access device 22 and duplicate data access devices 22 that manage the given file. The configuration management device 24 adds the configuration information 25 for each pair of data storage device 23 storing copies of the files on which loads are concentrated and the duplicate data access devices 22. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to a data management method, a data management program, a data management system, and a configuration management apparatus.

  Data management systems play an important role in social infrastructure. In a data management system that continues to accumulate data, the amount of data handled and the number of accesses increase. This creates a load on the data management system.

  The load on the data management system is due to the fact that access is concentrated on some of the entire data. As a tendency of data access, for example, the data reading rate may be much larger than the data writing rate. Another tendency is that access is concentrated as data having a shorter elapsed time since writing is written and data is not accessed as time passes.

  A data storage device constituting the data management system stores data to be accessed. Several methods for alleviating the load concentration on the data storage device have been proposed.

  For example, RAID (Redundant Arrays of Inexpensive Disks) described in Non-Patent Document 1 or the like allows a plurality of physical data storage devices to be recognized as a single logical data storage device and a plurality of physical data storage devices. By distributing the access to the storage devices in parallel, the load concentration on the data storage device is alleviated.

  A method of preventing deterioration of access performance to the data storage device 13 by creating a copy of the data in the data storage device and distributing the access to the same data to a plurality of data storage devices for data on which access is concentrated (See Patent Document 1). Furthermore, striping reduces the load concentration on the data storage device by fragmenting one file and distributing and storing it in separate data storage devices.

And when multiple accesses to the same data occur frequently, by setting the conditions that trigger the creation of duplicate data in advance, the creation of duplicate data and the switching of the access destination are automatically performed within the system. A technique for performing this is disclosed (see Patent Document 2).
US Pat. No. 6,961,727 JP 2005-338893 A Patterson, David A / Gibson, Garth A / Katz, Randy H, `` A Case for Redundant Arrays of Inexpensive Disks '', CSD-87-391

  Simply distributing the load to the data storage devices that make up the data management system is not sufficient to achieve the load distribution of the entire data management system.

  The data management system is connected to a data access device that instructs access to the data storage device. The data access device mediates access of data stored in the data storage device to the access request source in accordance with the access request.

  When the access request amount of the data access device increases and exceeds the limit of the processing capability, the data access device waits for processing even if the processing capacity of the data storage device is sufficient. That is, the data access device becomes a bottleneck of the entire data management system.

  Therefore, the main object of the present invention is to achieve load distribution of the entire data management system.

  In order to solve the above-described problems, the present invention provides an access request device that receives an input of an access request that is a read request or a write request, a data storage device that stores a file that is a target of the access request, and the data storage device A set of data access devices dedicated to the access right, and the original data access device that manages the original file of the access request target among the data access device, and the data access device, A data management method using a data management system for connecting a duplicate data access device that manages a duplicate file to a configuration management device that stores configuration information associated with a file stored in the data storage device, the access request The device inputs the read request to a predetermined file. Then, out of the original data access device and the duplicate data access device that manages a predetermined file, one of the data access devices is selected from the configuration information, the read request is made, and the write request to the predetermined file is made In response to the input, the write request is made to the original data access device and the duplicate data access device that manage a predetermined file, and the configuration management device stores the duplicate file of the file on which the load is concentrated The data storage device and the duplicate data access device are added to the configuration information in units of sets. Other means will be described later.

  According to the present invention, load distribution of the entire data management system can be achieved.

  Hereinafter, a data management system according to an embodiment of the present invention will be described with reference to the drawings.

  FIG. 1 is a configuration diagram showing a data management system 2 according to an embodiment of the present invention. This data management system 2 is applied to an order information management system, for example.

  The data management system 2 includes an access request device 21, a configuration management device 24, a data access device 22, and a data storage device 23 connected to each other via a network.

  In the data management system 2, the desirable number of devices is as follows. There are as many access request devices 21 as there are users who make access requests. The configuration management device 24 is basically only one, but may be multiplexed for redundancy or load distribution.

  Since the data access device 22 and the data storage device 23 are paired with each other, the same number exists. The number of sets is prepared in advance so that the entire access amount to the data management system 2 can be processed with a margin. This is because the load distribution performed by the data management system 2 of the present embodiment is realized by jointly burdening another data access device 22 with concentration of access to a specific data access device 22. In other words, it is difficult to realize load balancing in a computer environment where there is no data access device 22 to be shared jointly (in other words, there is no room for processing capability of all the data access devices 22).

  The access request device 21 makes an access request to the data access device 22. The access request is data writing to the data storage device 23 and data reading from the data storage device 23. The data writing is a process in which the data to be accessed changes. For example, “INSERT INTO (insert record data)” or “UPDATE to SET (update record data)” of an SQL (Structured Query Language) instruction. , And “DELETE FROM”. On the other hand, the data reading is a process in which the data to be accessed does not change, and corresponds to, for example, the SQL instruction “SELECT to FROM to WHERE (record search)”.

  The access request device 21 analyzes an access request from a general user. When the access request is a data write request, the data access device 22 is instructed to write data to the data storage device 23. When the access request is a data read request, the data access device 22 is instructed to read data into the data storage device 23.

  Further, the access request apparatus 21 performs load distribution of access requests. The load distribution is to select one data access device 22 to be an access request target from a plurality of data access device 22 candidates (acquired from configuration information 25 described later). As a result, the concentration of the load only on the specific data access device 22 is suppressed. Hereinafter, three specific load balancing calculation methods (1) to (3) will be exemplified.

  The load distribution method (1) is round robin. First, each of a plurality of candidates is ordered. Then, among the plurality of candidates, the candidate that is ordered next to the previously selected candidate is selected as the current candidate.

  The load distribution method (2) is a hash method. A hash function that selects one candidate among a plurality of candidates is calculated, and the candidate selected as the calculation result is selected as the current candidate. A hash function that allows all candidates to be selected equally is desirable.

  The load distribution method (3) is a method of selecting at random. For example, a method of using the current time as a parameter for selecting at random will be described. The access request device name is converted into a character code, and the sum of each character code is obtained. The current time is converted into the total seconds from 0:00, and the sum of the character codes and the total seconds is divided by the value obtained by adding 1 to the number of corresponding data access devices, and the remainder is obtained. Then, the data access device 22 that issues a request is specified from the remainder value.

  The data access device 22 instructs the data storage device 23 to access the access request. The data storage device 23 receives an instruction from the data access device 22, accesses the data stored in the device itself, and returns the result to the data access device 22. The data storage device 23 stores an original file 23a or a duplicate file 23b as data. The original file 23a is a file created first as a definitive one for storing data, and the duplicate file 23b is a file having the same contents as the original file 23a. Here, the data access device 22 that manages the data storage device 23 that stores the original file 23 a is defined as the original data access device 22. The data access device 22 that manages the data storage device 23 that stores the duplicate file 23b is defined as a duplicate data access device 22.

  First, the case where the access request is a write request will be described. The access request device 21 transmits a write request to the data access device 22. The data access device 22 requests the subordinate data storage device 23 to write the data specified by the write request to the file specified by the write request. The data storage device 23 that has received the request writes the designated data in the designated file, and notifies the data access device 22 of the completion of the data writing. The data access device 22 that has received the notification notifies the access request device 21 of the completion of data writing.

  Next, a case where the access request is a read request will be described. The access request device 21 transmits a read request to the data access device 22. The data access device 22 requests the subordinate data storage device 23 to read the designated data for the file designated by the read request. Upon receiving the request, the data storage device 23 reads the data of the designated file, and notifies the data access device 22 of the read data and the data read completion. The data access device 22 that has received the notification notifies the access request device 21 of the read data and the completion of data reading.

  The set of the data access device 22 and the data storage device 23 has a shared nothing (dedicated access) system relationship. In other words, the data access device 22 has exclusive access rights to the data stored in the data storage device 23 with respect to the subordinate data storage devices 23. That is, each data access device 22 can access all data stored in the data storage device 23 under its control.

  Instead of the data access device 22, the configuration management device 24 that has received the access request from the data access device 22 may proxy the access request to the data access device 22.

  The configuration management device 24 holds configuration information 25. The configuration management device 24 returns configuration information 25 in response to the configuration information read request. In addition, the configuration management device 24 receives a change request for the configuration information 25 from the administrator user, and changes the configuration information 25.

表１は、構成情報２５のデータ構造の一例を示す図である。構成情報２５の各レコードは、データアクセス装置２２およびその配下のデータ記憶装置２３の組が追加された場合に追加され、組が削除された場合に削除される。

Table 1 is a diagram illustrating an example of the data structure of the configuration information 25. Each record of the configuration information 25 is added when the set of the data access device 22 and the data storage device 23 under the data access device 22 is added, and is deleted when the set is deleted.

  The configuration information 25 manages the data access device name, the file name accessible by the data access device name, and the attribute of the file name in association with each other. The file attribute takes two types of values, “original” or “replica”, and means that the file handled by the data access device name 31 is either an original file or a duplicate file. Note that zero or more duplicate files are created for one original file.

  By searching the configuration information 25 using the file name as a key, it is possible to search for a duplicate file corresponding to a predetermined original file or an original file corresponding to a predetermined replica file.

  The data access device 22 that processes the data read request is one original data access device 22 or one duplicate data access device 22 among the original data access device 22 and the duplicate data access device 22 for a predetermined file. . The data access device 22 that processes the data write request is all of the original data access device 22 and the duplicate data access device for a predetermined file.

  FIG. 2 is an explanatory diagram of hardware resources for configuring each device (access request device 21, configuration management device 24, data access device 22, and data storage device 23) of FIG. The device 1 includes a CPU (Control Processing Unit) 11 and a memory 12 and is connected to a storage unit 13. In the memory 12, an OS (Operating System) 121 and an application program 122 are arranged.

  The CPU 11 executes the OS 121 and the application program 122 in the memory 12. The device 1 accesses the data 131 in the storage unit 13 and executes different application programs for each role in the system. The storage unit 13 corresponds to a device capable of reading and writing data, such as a hard disk. The data reading is, for example, data acquisition. The data write is, for example, data write, change, and deletion.

  Note that a set of the data access device 22 and the data storage device 23 is newly added due to insufficient performance of the data storage device 23, for example. Insufficient performance occurs when the amount of data that can be stored in the data storage device 23 is insufficient, or when the access performance of the data storage device 23 cannot satisfy the requirements of the entire system.

  Here, when there are a plurality of sets of the data access device 22 and the data storage device 23, load imbalance may occur between the sets. For example, when recently accessed data is intensively stored in the newly added data storage device 23, access to the new data storage device 23 is concentrated. This is due to the tendency that newer data is used more frequently.

  Therefore, in the present embodiment, the load concentration is reduced by duplicating the data managed by the data access device 22 and the data storage device 23 in which the load concentration has occurred to a new set. Jointly bear access to the data that occurs. As a result, efficient storage utilization called ILM (Information Lifecycle Management) is realized. In other words, in order to improve the access performance of the entire system, the number of computer resources for data that is hardly accessed is reduced, and the number of computer resources for data with a large number of accesses is increased to change the configuration according to the number of accesses for each data. I do.

  FIG. 3 is an explanatory diagram for explaining the change of the configuration information 25 of the present embodiment. FIG. 3A shows the file “p” before addition, and FIG. 3B shows the file “p” after addition.

  First, in FIG. 3A, each file to be accessed is assigned a set of a data access device 22 and a data storage device 23 for processing the access. For example, a set “1” for managing the original file and a set “2” for managing the duplicate file are allocated to the file “1”.

  The storage pool manages a set of the data access device 22 and the data storage device 23 in which the file to be accessed is not stored in the data storage device 23 (in other words, the processing is vacant).

  It should be noted that a set is returned to the storage pool by deleting a set where the load is not concentrated (for example, set “5”) from the set already used for specific file processing. The returned set is reused for another file processing. Details of deletion of a set will be described in the “deletion of duplicate data access device 22” procedure described later.

  Next, in FIG. 3B, a set to which the file “p” is assigned is secured from the storage pool. The set “6” is allocated to manage the original file of the file “p”. Details will be described in the “addition of original data access device 22” procedure described later. The group “7” and the group “5” are assigned to manage the duplicate file of the file “p”. Details will be described in the “addition of duplicate data access device 22” procedure described later.

  As described above, the set “5” which is the computer resource of the file “n” where access is not concentrated is utilized as the computer resource of the file “p” where access is concentrated. As a result, it is possible to efficiently load balance access to the file “p”. Note that when assigning a pair in a storage pool, as a method to specify which pair to assign, a method in which an administrator user explicitly designates a pair, a pair in a storage pool is ordered, and a previously specified pair is assigned. There is a round robin method that specifies the next set of items.

  FIG. 4 is a flowchart showing an access request procedure by the access request device 21.

  The access request device 21 receives a request from a general user (S101). Here, the request from the general user is either “data writing” or “data reading”. The data write request is a request for writing data designated by the user to a file designated by the user, and includes a data body to be written and a file name of the write destination. A data read request is a request to read data from a file specified by the user, and includes the name of the file to be read.

  The access request device 21 branches the process according to the request type from the general user (S102), and executes the data write procedure (S103) or the data read procedure (S104).

  FIG. 5 is a flowchart showing a data writing procedure.

  The access request device 21 requests the configuration information 25 from the configuration management device 24 (S201). The configuration management device 24 returns the configuration information 25 to the access request device 21 (S202).

  The access requesting device 21 specifies the original data access device 22 that manages the original file of the data write target file from the configuration information 25 (S211). Specifically, among the data access devices 22 that can access a file having the same file name as the write destination file name specified by the user based on the acquired configuration information 25, the data access device 22 having the original file attribute is selected. Identify.

  The access requesting device 21 requests the specified original data access device 22 to write the original file (S212). The request message includes the data body of the original file and the file name to which the original file is written.

  The original data access device 22 controls the subordinate data storage device 23 to write data (S213). The original data access device 22 notifies the configuration management device 24 of the request message in S212 and requests the writing of the duplicate file (S221).

  The configuration management device 24 identifies the duplicate data access device 22 that manages the duplicate file of the data write target file from the configuration information 25 (S222). The configuration management device 24 sequentially selects the specified duplicate data access device 22 (S223), and performs loop control so as to request the selected duplicate data access device 22 to write a duplicate file (S224). The duplicate data access device 22 controls the subordinate data storage device 23 to write data (S225).

  The configuration management device 24 notifies the original data access device 22 of the completion of writing the duplicate file (S226). The original data access device 22 notifies the access request device 21 of the completion of writing data (original file and duplicate file) (S231). The access requesting device 21 notifies the general user of the completion of data writing (S232).

  FIG. 6 is a flowchart showing a data reading procedure.

  The access request device 21 requests the configuration information 25 from the configuration management device 24 (S301). The configuration management device 24 returns the configuration information 25 to the access request device 21 (S302).

  The access request apparatus 21 determines whether or not the duplicate data access apparatus 22 exists (S311). Specifically, the data access device 22 that can access a file having the same file name as the read-destination file name specified by the user based on the acquired configuration information 25 is specified.

  First, if Yes in S311, the access requesting device 21 determines the duplicate data access device 22 to be accessed (S312). When there are a plurality of duplicate data access devices 22, one access target is selected by the above selection algorithm. On the other hand, if No in S311, the access requesting device 21 determines the original data access device 22 to be accessed (S313).

  The access request device 21 requests the data access device 22 to be accessed to read data (S314). The access target data access device 22 controls the subordinate data storage device 23 to read data (S315). The data access device 22 to be accessed notifies the access request device 21 of the completion of data reading (S321). The notification message includes the data content of the file to be read. The access requesting device 21 notifies the general user of the completion of data reading (S322).

  FIG. 7 is a flowchart showing a configuration management request procedure.

  The access request device 21 transmits a configuration management request from the administrator user to the configuration management device 24 (S401). Here, the configuration management request is one of the following (1) to (3).

  The configuration management request (1) is “addition of original data access device 22”. A set of a data storage device 23 that stores an original file and an original data access device 22 that manages the data storage device 23 is newly added to the data management system 2. This is issued as the person in charge of management when the original file is newly managed. The configuration management request includes the name of the data access device 22 to be added and the file name of the original file.

  The configuration management request (2) is “addition of duplicate data access device 22”. A set of a data storage device 23 that stores a duplicate file and a duplicate data access device 22 that manages the data storage device 23 is added to the data management system 2 for the original file that has already been managed. This is issued for the purpose of mitigating the concentrated access to the original file and its duplicate files. The configuration management request includes the name of the data access device 22 to be added and the file name of the original file to which the duplicate file is added.

  The configuration management request (3) is “deletion of duplicate data access device 22”. For a duplicate file whose data is already managed, the set of the data storage device 23 that stores the duplicate file and the duplicate data access device 22 that manages the data storage device 23 is deleted from the data management system 2. This is issued for the purpose of effective use of the apparatus because a duplicate file added because access has been concentrated before has been processed because the access has been relaxed. The configuration management request includes the name of the data access device 22 to be deleted.

  The access request device 21 determines the type of the configuration management request (S402), and “addition of the original data access device 22” (S403), “addition of the duplicate data access device 22” (S404), or “replication data access” “Deleting device 22” (S405) is performed.

  FIG. 8 is a flowchart showing a procedure of “adding original data access device 22”.

  The configuration management device 24 determines whether the configuration management request is valid (S501). The fact that the configuration management request is valid indicates that the same data access device name and file name as those specified by the user in the configuration management request have already been written in the configuration information 25.

  If No in S501, the configuration management apparatus 24 notifies the administrator user of an error indicating that the configuration management request is invalid (S502). Then, the process ends.

  If Yes in S501, the configuration management device 24 instructs the original data access device 22 to manage the original file specified in the configuration management request (S503). Then, an application for managing the original file is installed and deployed in the original data access device 22.

  The original data access device 22 instructs the subordinate data storage device 23 to prepare to store the original file specified by the configuration management request (S504). The preparation refers to a process for securing an area for storing data and making the data accessible.

  The configuration management device 24 registers a set of the newly prepared data storage device 23 and the original data access device 22 that manages the data storage device 23 in the configuration information 25 (S505). For example, when the added data access device name is “data access device p” and the original file name to be added is “file p”, the data access device name item of the configuration information 25 includes “data access device p”, file “File p” is written in the name item, and “original” is written in the file attribute item. The configuration management device 24 notifies the administrator user of the completion of processing of the configuration management request (S506).

  FIG. 9 is a flowchart showing the procedure of “addition of duplicate data access device 22”.

  The configuration management device 24 determines whether or not the configuration management request is valid (S601). The fact that the configuration management request is valid indicates that the configuration information 25 includes the same data access device name and file name as the data access device name and file name specified by the user in the configuration management request.

  If No in S601, the configuration management apparatus 24 notifies the administrator user of an error indicating that the configuration management request is invalid (S602). Then, the process ends.

  If Yes in S601, the configuration management device 24 instructs the duplicate data access device 22 to manage the duplicate file specified in the configuration management request (S603). Then, an application for managing a duplicate file is installed and deployed in the duplicate data access device 22. The duplicate data access device 22 prepares the subordinate data storage device 23 to store the duplicate file designated by the configuration management request.

  The configuration management device 24 identifies the original data access device 22 that manages the original file for the duplicate file specified in the configuration management request from the configuration information 25 (S611). Specifically, the data access device 22 having the original file attribute is identified from among the data access devices 22 that can access the file having the same file name as the duplicate file specified by the user.

  The configuration management device 24 requests the specified original data access device 22 to read the original file (S612). The original data access device 22 controls the subordinate data storage device 23 to read data (S613). The original data access device 22 notifies the configuration management device 24 of the completion of reading of the original file (S614).

  The configuration management device 24 requests the duplicate data access device 22 added in S603 to write the duplicate file based on the data content of the original file (S621). The duplicate data access device 22 controls the subordinate data storage device 23 to write data (S622). The duplicate data access device 22 notifies the configuration management device 24 of the completion of writing the duplicate file (S623).

  The configuration management device 24 registers the added duplicate data access device 22 in the configuration information 25 (S624). For example, if the added data access device name is “data access device p1” and the original file name corresponding to the duplicate file to be added is “file p”, the data access device name item of the configuration information 25 is “data access device”. "Device p1", "File p" is written in the file name item, and "Duplicate" is written in the file attribute item. The configuration management device 24 notifies the administrator user of the completion of processing of the configuration management request (S625).

  FIG. 10 is a flowchart showing a procedure of “deletion of duplicate data access device 22”.

  The “deletion of duplicate data access device 22” procedure is executed, for example, when a configuration management request is explicitly input from an administrator user. Alternatively, the configuration management device 24 may measure the processing capability of the duplicate data access device 22 and issue a configuration management request when detecting that the processing capability is idle. In the state where the processing capacity is idle, for example, the administrator user designates the conditions in advance, for example, a period in which the CPU usage rate is 20% or less continues for 10 hours.

  The configuration management device 24 determines whether or not the configuration management request is valid (S701). “Valid” means that the configuration information 25 includes the same data access device name as the data access device name designated by the administrator user, and the file attribute of the file handled by the data access device 22 designated by the administrator user is duplicated. This is the state.

  If No in S701, the configuration management device 24 notifies an error indicating that the configuration management request is invalid (S702). Then, the process ends.

  If Yes in S701, the configuration management device 24 instructs the duplicate data access device 22 to remove the duplicate file specified in the configuration management request from the management target (S703). Then, an application for managing the duplicate file is uninstalled in the duplicate data access device 22. In the configuration management request, the administrator user may specify the number of duplicate data access devices 22 instead of explicitly specifying the data access device name. When the number is specified, the configuration management device 24 refers to the configuration information 25 and acquires the specified number of data access device names.

  The duplicate data access device 22 instructs the subordinate data storage device 23 to delete the duplicate file specified in the configuration management request (S704).

  According to the present embodiment described above, the access request apparatus 21 is a candidate for allocating requests not only to the original data access apparatus 22 but also to the duplicate data access apparatus 22 in response to the input of the read request. In addition, for a file on which a load is concentrated due to many access requests, a duplicate data access device 22 for processing the file can be added. Therefore, it is possible to efficiently assign the duplicate data access device 22 to read requests that vary from file to file. In addition, the bottleneck to the data access device 22 is eliminated by adding not only the data storage device 23 but also the duplicate data access device 22 in charge of the data storage device 23 in units of sets. Therefore, load distribution of the entire data management system can be achieved.

  Note that execution of the write request to the data access device 22 in charge causes a slight processing overhead compared to the method of writing to one data access device. However, since the frequency of access requests is overwhelmingly higher for read requests than for write requests, the load on the write requests is relatively negligible.

It is a block diagram which shows the data management system regarding one Embodiment of this invention. It is explanatory drawing which shows the hardware resource of each apparatus of the data management system regarding one Embodiment of this invention. It is explanatory drawing which shows the change of the structure information regarding one Embodiment of this invention. It is a flowchart which shows the access request procedure by the access request apparatus regarding one Embodiment of this invention. It is a flowchart which shows the data-writing procedure regarding one Embodiment of this invention. It is a flowchart which shows the data reading procedure regarding one Embodiment of this invention. It is a flowchart which shows the structure management request | requirement procedure regarding one Embodiment of this invention. It is a flowchart which shows the "addition of an original data access apparatus" procedure regarding one Embodiment of this invention. It is a flowchart which shows the "addition of a replication data access apparatus" procedure regarding one Embodiment of this invention. It is a flowchart which shows the "deletion of a replication data access apparatus" procedure regarding one Embodiment of this invention.

Explanation of symbols

1 device 11 CPU
12 Memory 121 OS
122 Application Program 13 Storage Unit 131 Data 2 Data Management System 21 Access Request Device 22 Data Access Device 23 Data Storage Device 23a Original File 23b Replicated File 24 Configuration Management Device 25 Configuration Information 9 Communication Network

Claims (6)

An access request device that accepts an input of an access request that is a read request or a write request;
A set of a data storage device that stores the file that is the target of the access request, and a data access device that has exclusive access rights to the data storage device;
Among the data access devices, an original data access device that manages an original file of a file that is a target of the access request, and a duplicate data access device that manages a duplicate file among the data access devices, the data storage device A configuration management device that stores configuration information associated with a file stored by
A data management method by a data management system for connecting
The access request device includes:
In response to the input of the read request to a predetermined file, one of the original data access device and the duplicate data access device that manages the predetermined file is selected from the configuration information and the read Make a request,
In response to the input of the write request to the predetermined file, the write request is made to the original data access device and the duplicate data access device that manage the predetermined file,
The configuration management device includes:
A data management method comprising: adding to the configuration information in units of a set of the data storage device that stores the duplicate file of the file on which the load is concentrated and the duplicate data access device.   The configuration management apparatus manages the group units that are not processing the access request in a storage pool in order, and the load concentrates the group units that are not processing the access request according to the ordering. 2. The data management method according to claim 1, wherein the data management method is assigned as a file processing target.   The configuration management device deletes from the configuration information and returns to the storage pool in units of the data storage device that stores the duplicate file of a file that is not concentrated in load and the duplicate data access device. The data management method according to claim 2, wherein:   A data management program for causing a computer to execute the data management method according to any one of claims 1 to 3. An access request device that accepts an input of an access request that is a read request or a write request;
A set of a data storage device that stores the file that is the target of the access request, and a data access device that has exclusive access rights to the data storage device;
Among the data access devices, an original data access device that manages an original file of a file that is a target of the access request, and a duplicate data access device that manages a duplicate file among the data access devices, the data storage device A configuration management device that stores configuration information associated with a file stored by
A data management system for connecting
The access request device includes:
In response to an input of the read request to a predetermined file, one of the original data access device and the duplicate data access device that manages the predetermined file is selected from the configuration information and the read request is selected. And
In response to the input of the write request to the predetermined file, the write request is made to the original data access device and the duplicate data access device that manage the predetermined file,
The configuration management device includes:
A data management system, wherein the data storage device that stores the duplicate file of the file on which the load is concentrated and the duplicate data access device are added to the configuration information. An access request device that accepts an input of an access request that is a read request or a write request;
A set of a data storage device that stores the file that is the target of the access request, and a data access device that has exclusive access rights to the data storage device;
Among the data access devices, an original data access device that manages an original file of a file that is a target of the access request, and a duplicate data access device that manages a duplicate file among the data access devices, the data storage device A configuration management device that stores configuration information associated with a file stored by
A configuration management device used in a data management system for connecting
The access request device includes:
In response to an input of the read request to a predetermined file, one of the original data access device and the duplicate data access device that manages the predetermined file is selected from the configuration information and the read request is selected. And
In response to the input of the write request to the predetermined file, the write request is made to the original data access device and the duplicate data access device that manage the predetermined file,
The configuration management device includes:
A configuration management device, wherein the data storage device that stores the duplicate file of the file on which the load is concentrated and the duplicate data access device add the configuration information to the configuration information.

Images