JP6372111B2 - Storage system - Google Patents

Description

  The present invention relates to a storage system, and more particularly to a storage system that stores data in a distributed manner in a plurality of storage devices.

  In recent years, backup operation of business data from the viewpoint of disaster recovery has been accelerated, and the number of large-capacity storage devices used has increased accordingly. When data is backed up, in order to increase the reliability of stored data, redundancy is increased and distributed arrangement is performed in a plurality of storage devices.

  However, the storage device consumes more power than a normal server, and there is a problem of power shortage in the summer. Therefore, the demand for power saving is increasing in each workplace and business place of a company. For this reason, how to reduce the power consumption of the storage apparatus is a problem in backup operation.

JP 2009-271828 A

  Here, PARAID is a technology for realizing power saving of storage. This technology realizes power saving by reducing the number of disks in the active state by biasing the parity arrangement of each disk constituting the RAID.

  Japanese Patent Application Laid-Open No. 2004-228561 discloses a technique for adding updated data and read data in order from a small address. As a result, unaccessed data is consolidated in the smaller address, and the disk device that collects the unaccessed data is set to the power saving mode.

  However, any of the above-described techniques allows data to be read and written only to a specific disk. For this reason, the load with respect to a specific disk increases, and the problem that performance falls or a lifetime becomes short arises.

  For this reason, an object of the present invention is to solve the above-described problems such as the reduction in performance and the shortening of the life of the storage system while reducing the power consumption of the storage system.

A storage system according to an aspect of the present invention
Storage processing content management means for storing and managing job data including the number of distributions representing the number of storage devices to which data is distributed;
Distributed storage control means for distributing and storing data to be stored in the distributed number of storage devices based on the job data,
The distributed storage control means determines at least the storage devices of the distributed number based on the distributed number, and turns on only the determined storage device;
The configuration is as follows.

Moreover, the program which is the other form of this invention is:
In the information processing device,
Storage processing content management means for storing and managing job data including the number of distributions representing the number of storage devices to which data is distributed;
Distributed storage control means for distributing and storing data to be stored in the distributed number of storage devices based on the job data,
The distributed storage control means determines at least the storage devices of the distributed number based on the distributed number, and turns on only the determined storage device;
The configuration is as follows.

In addition, a distributed storage method according to another aspect of the present invention includes:
Represents the contents of the distributed storage processing of data, stores and manages job data including the number of distributions representing the number of storage devices that distribute the data,
A distributed storage method for storing data to be stored in a distributed number of storage devices based on the job data,
Determining at least the number of storage devices based on the number of distributions, turning on only the determined storage devices, and performing distributed storage processing based on the job data;
The configuration is as follows.

  With the configuration as described above, the present invention can suppress a decrease in performance and a shortened life while saving power in the storage system.

It is a functional block diagram which shows the structure of the backup device in Embodiment 1 of this invention. It is a figure which shows an example of the data structure of the job table disclosed in FIG. It is a figure which shows an example of the data structure of the switching time table disclosed in FIG. It is a figure which shows an example of the data structure of the directory information table disclosed in FIG. FIG. 2 is an explanatory diagram for explaining an operation at the time of backup in the backup device disclosed in FIG. 1. It is a figure which shows an example of the data processed with the backup apparatus disclosed in FIG. It is a figure which shows an example of the data processed with the backup apparatus disclosed in FIG. It is a figure which shows an example of the data processed with the backup apparatus disclosed in FIG. It is a figure which shows an example of the data processed with the backup apparatus disclosed in FIG. It is explanatory drawing for demonstrating the mode of the storage node by the backup apparatus disclosed in FIG. It is a figure which shows an example of the data processed with the backup apparatus disclosed in FIG. It is a figure which shows an example of the data processed with the backup apparatus disclosed in FIG. It is explanatory drawing for demonstrating the mode of the storage node by the backup apparatus disclosed in FIG. FIG. 7 is an explanatory diagram for explaining an operation at the time of restoration in the backup device disclosed in FIG. 1. It is a figure which shows an example of the data processed with the backup apparatus disclosed in FIG. It is a block diagram which shows the structure of the storage system in attachment 1 of this invention.

<Embodiment 1>
A first embodiment of the present invention will be described with reference to FIGS. 1 to 4 are diagrams for explaining the configuration of the backup device. FIG. 5 is a diagram for explaining the operation at the time of backup of the backup device. 6 to 9 are diagrams illustrating an example of data processed by the backup device. FIG. 10 is a diagram for explaining the state of the storage node. 11 to 12 are diagrams illustrating an example of data processed by the backup device. FIG. 13 is a diagram for explaining the state of the storage node. FIG. 14 is a diagram for explaining the operation at the time of restoration of the backup device. FIG. 15 is a diagram illustrating an example of data processed by the backup device.

[Constitution]
The backup device 100 according to the present invention is a storage system that receives data to be backed up from the backup target device 40 and distributes and stores the data in storage devices. The backup device 100 also has a function of restoring data stored in the storage device to another storage device serving as the restore destination 50.

  Specifically, as shown in FIG. 1, the backup device 100 includes a backup server 10 and a grid-type storage node group 20 in which data is distributed and redundantly arranged.

  The backup server 10 is an information processing apparatus including an arithmetic device, and stores data in a job management unit 11 that manages a backup job, which is built by incorporating a program in the arithmetic device, and mainly in the storage node group 20. And a storage control unit 12 for writing.

  The storage node group 20 includes a plurality of storage nodes 21. The storage node 21 is an information processing apparatus provided with a storage device, and as will be described later, data to be backed up is distributed and stored from the backup server 10 by distributed storage processing. As shown in FIG. 1, the name (identification information) of the storage node 21 is “No. 0” for the first unit and “No. (N-1)” for the N unit. In the example of FIG. 1, a case where six (No. 0 to No. 5) storage nodes 21 are provided is illustrated, but the number of storage nodes 21 is not limited to this, and at least one storage node 21 is provided. The storage node 21 may be configured.

  The job management unit 11 (storage processing content management means) has a job table 11a for storing backup job data (job data) registered by the user 30. When the user 30 registers a backup job in the backup server 10, the backup job data is added to the job table 11a as a new record. The job table 11a is data representing the content of distributed processing of backup target data by a backup job, and is configured as shown in FIG. 2, for example. In particular, each record of the job table 11a includes the number of distributed nodes (the number of distributions) that represents the number of storage nodes that are the targets of distributed storage of backup target data. Then, the job management unit 11 schedules the backup job along the “next start time” in the job table 11 a and transmits the job data to be executed to the storage control unit 12.

  When the storage control unit 12 receives job data from the job management unit 11, the storage control unit 12 determines a storage node 21 to be a write destination (storage target) based on the number of distributed nodes, as will be described later. Then, after turning on the power of the storage node 21 that is the write destination as necessary, the backup target data is read from the backup target device 40, and distributed to the storage nodes 21 to be written (stored). To do). At this time, the storage node 21 for which the storage control unit 12 decides to turn on the power stores not only the storage node 21 that is the write destination but also the latest generation data that was the write destination in the previous backup. The existing storage node 21 is also targeted.

  In addition, the storage control unit 12 changes and determines the combination of the storage nodes 21 every time backup is executed. A specific example of determining the combination of storage nodes 21 will be described in detail later when the operation is described.

  The storage control unit 12 has a switching time table 12a and a directory information table 12b. The switching time table 12a stores, as the system switching time, a time during which backup is not performed within one day, which is data registered when the user 30 uses the system for the first time. For example, the switching time table 12a is configured as shown in FIG. Is done. The switching time table 12a does not add a new record arbitrarily by the user.

  In the directory information table 12b, after writing the backup target data to the storage node 21, the job name, the storage position of the written data on the storage node 21, and the like are added as a new record. The directory information table 12b has information on items shown in FIG. 4 in addition to the directory information of a general file system.

[Operation]
Next, a backup operation and a restore operation by the backup device 1 in this embodiment will be described. In addition, in this operation description, the function of each component of the backup device 1 described above will also be described in detail.

(Daily backup)
First, an operation when performing a daily backup, that is, a daily backup will be described with reference to FIGS. It is assumed that this daily backup job ends in one day.

  First, the user 30 registers in the backup server 10 a time zone (in minutes) during which backup is not executed in one day (arrow A1 in FIG. 5). This procedure is always performed when the backup device 1 is used for the first time. In this example, it is assumed that this time is the first use, and for example, the user 30 inputs “12:00”. Thereafter, as shown in FIG. 6, the storage control unit 12 of the backup server 10 inputs “12:00” as the backup switching time in the switching time table 12a and sets “0” as the number of system operating days in the switching time table 12a. input.

  Next, the user 30 registers a backup job (hereinafter referred to as a job) in the backup server 10 (arrow A2 in FIG. 5). When registering a job, the user 30 inputs all the data items shown in FIG. 7A for each job. Here, it is assumed that the user 30 inputs the value shown in FIG. 7B as a job. Thereafter, the job management unit 11 of the backup server 10 calculates the “next start time” from the input “schedule” time shown in FIG. 7B and the current system time. Then, based on the input information shown in FIG. 7B and the calculated “next start time”, the job management unit 11 adds the job of the new record to the job table 11a as shown in FIG. 8A. to add.

  The job management unit 11 refers to the “next start time” column of the job registered in the job table 11a every minute. If there is data in the “next start time” column that is equal to the current time of the system, a backup job for the record of that data is started. In the example of the job table 11a in FIG. 8A, the backup job is started at “2013/11/11 21:00”. Then, at the start of the backup job, the job management unit 11 inputs the current time of the system to the “final start time” of the job table 11a as shown in FIG. 8B. Further, the job management unit 11 calculates the “next start time” from the current time of the system and the “job type”, and overwrites the “next start time” item in the job table 11 a. In this example, since the “job type” is a daily backup, the date and time after one day is overwritten with the “next start time”.

  Thereafter, the job management unit 11 transmits a part of the record of the job to be executed in the job table 11 to the storage control unit 12 (arrow A3 in FIG. 5). Here, as shown in FIG. 8C, the job name, job type, data to be backed up, the number of distributed nodes, the number of parity, and the generation are transmitted to the storage control unit 12 from the job record.

  When the storage control unit 12 receives partial data of the job record shown in FIG. 8C, the storage control unit 12 determines a virtual node group. Here, the virtual node group to be determined is a combination of the storage nodes 21 that are turned on in the distributed storage processing of the backup target data, and at least a plurality of storage nodes that are the distribution destinations for distributing the backup target data 21 is included. Note that the storage node 21 determined as the virtual node group may be composed of only the storage node 21 that is the distribution destination, and includes the storage node 21 that was the storage destination at the time of the previous backup, as will be described later. You may go out.

  Specifically, the storage control unit 12 uses the “number of distributed nodes” included in the job record shown in FIG. 8C and the “system operating days” in the switching time table 12a shown in FIG. A virtual node group is determined based on the following calculation formula (1).

The calculation formula (1) for determining the virtual node group is set so as to calculate a numerical value to be the name (No.) of the storage node 21 as follows.
・ Calculation formula (1)
[No. (system working days% number of storage nodes), No. ((system working days + 1)% number of storage nodes), ..., No. ((system working days + (number of distributed nodes-1))% storage Number of nodes)]

  Note that “%” in the above formula (1) is the remainder when the sum of (system operating days) and (number of distributed nodes: 0 to (maximum value−1)) is divided by (number of storage nodes), Represents. Therefore, when the first backup job is executed, the combination of the storage nodes 21 “No. 0” to “No. (number of distributed nodes −1)” is determined as a virtual node group. In this example, since the “number of distributed nodes” is “3”, the three storage nodes 21 such as “No. 0, No. 1, No. 2” are virtual nodes as shown by the hatched lines in FIG. Determined as a group. However, the method of determining the virtual node group is not limited to the above-described calculation formula (1), and may be determined by any method.

  When the virtual node group is determined, the storage control unit 12 confirms whether the storage nodes 21 constituting the virtual node group can be accessed. When there is a storage node 21 that cannot be accessed due to the power being turned off, the storage node 21 is controlled to be turned on (arrow A4 in FIG. 4). In this example, only the “No.0, No.1, No.2” storage node is powered on. Note that the storage control unit 12 may turn off the power of other storage nodes that have not been determined as a virtual node group.

  After confirming that the virtual node group can be accessed, the storage control unit 12 reads data from the backup target device 40 with reference to “backup target data” shown in FIG. 8C (arrow in FIG. 5). A5). Then, the parity of the “number of parity” value is added to the backup target data, and distributed to each storage node 21 of the determined virtual node group “No. 0, No. 1, No. 2”. (Distributed redundant arrangement) (arrow A6 in FIG. 5). At this time, since only the “distributed node” storage nodes 21 are determined as the virtual node group at the time of the first backup, all the storage nodes 21 constituting the virtual node group are targeted for distributed storage.

  The storage control unit 12 adds “1” to the “generation” of the record shown in FIG. 8C after the completion of the above-mentioned write, and writes the information of the written data into the directory as shown in FIG. 9A. A new record is added to the information table 12b. The data consisting of this record is called directory information and is provided in response to an information browsing request from the user. Further, the storage control unit 12 returns the “Write completion time” and “generation” to the job management unit 11, and the job management unit 11 sets the received values to the job table 11a as shown in FIG. 9B. Enter the “Last completion time” and “Generation” of the corresponding record.

  Next, the daily backup operation on the next day after the second time will be described. The storage control unit 12 refers to the “backup switching time” in the switching time table 12a shown in FIG. 6 every minute. When the current time is equal to the “backup switching time”, “1” is added to “system operating days” in the switching time table, and “number of distributed nodes” in the directory information table 12b shown in FIG. 9A. Referring to the column, obtain the maximum value of “number of distributed nodes”.

  The storage control unit 12 determines a new virtual node group based on the following calculation formula (2) using the “system operating days” and the “distributed node number”. In the calculation formula (2), the names of storage nodes that are removed from the previous virtual node group and turned off, and storage nodes that are newly added to the virtual node group and turned on are as follows: No.) is set to calculate the numerical value.

・ Calculation formula (2)
Power ON: No. (system operating days + (number of distributed nodes -1))% number of storage nodes Power OFF: No. (system operating days -2)% number of storage nodes

  In the above formula (2), “%” is obtained by dividing “(the number of system operating days) and (the number of distributed nodes−1)” or “(the system operating days) −2” by (the number of storage nodes). Represents the surplus. Therefore, the “system operating days” is “1” for the three storage nodes “No.0, No.1, No.2” that were previously determined as virtual node groups and turned on. Therefore, the power of the storage node “No. 3” is newly turned on by the above calculation formula (2). On the other hand, none of the storage nodes are turned off. As a result, a combination of four storage nodes 21 of “No. 0, No. 1, No. 2, No. 3” different from the previous combination is determined as a new virtual node group.

  Then, the storage control unit 12 performs a power ON / OFF operation on the storage node 21 determined to be turned ON / OFF as described above. In the above case, only the power-on operation of the storage node “No. 3” is performed. Then, as indicated by the oblique lines in FIG. 10B, the power supplies of the four storage nodes 21 “No. 0, No. 1, No. 2, No. 3” are turned on. In this case, only the newly added “No. 3” and three storage nodes “No. 1” and “No. 2” with larger numbers are targets for distributed storage. On the other hand, the storage node of “No. 0” is not the target of distributed storage, but is the target of the previous distributed storage, so the latest generation data is stored. Therefore, data can be read immediately by turning on the power of the storage node “No. 0”.

  Furthermore, on the next day, since “system working days” is “2”, “No.1, No.2, No.3, The four storage nodes 21 of “No. 4” are determined as virtual storage nodes. That is, the power of the storage node “No. 4” is turned on, and the power of the storage node “No. 0” is turned off. Only the newly added “No. 4” and the three storage nodes with the larger numbers “No. 2” and “No. 3” are subject to distributed storage, and the storage node of “No. 1” The power is on to read the latest generation of data.

  As described above, in the present embodiment, the above-mentioned calculation formula (2) including “system operating days” in which a value is incremented by 1 every day is used to determine whether a power-on storage node and a power-off storage node are I will decide. As a result, the combination of storage nodes constituting the virtual node group is rotated one day at a time, and power saving can be achieved while equalizing the capacity of data stored in each storage node. As a result, it is possible to suppress a load from being concentrated on a specific storage node, and it is possible to suppress a decrease in performance and a shortened life of the storage system.

  In addition, in this embodiment, the storage node storing the latest generation data is also included in the virtual node group and the power is turned on. As a result, the latest generation data can be read immediately while saving power. The method for determining the virtual node group is not limited to using the above-described calculation formula (2), and may be determined by any method.

(Weekly backup)
Next, an operation when a weekly backup, that is, a backup once a week is executed will be described with reference to FIGS. 5 and 11 to 12. Here, a description will be given by taking as an example a process of registering a weekly backup on the day after the daily backup described above and executing the schedule. For this reason, registration is not performed during a time period when the backup indicated by arrow A1 in FIG. 5 is not executed. Note that the backup job in this description is assumed to end in one day.

  The user 30 registers a job in the backup server 10 (arrow A2 in FIG. 5). At the time of registration, the user 30 inputs all items shown in FIG. 7A for one job. An example of a case where a registered job is added as a new record to the job table 11a is shown in the third line of FIG.

  Thereafter, according to the “next start time” of the job, the backup job is started at “2013/11/12 23:30”. Then, as shown in FIG. 11B, the job management unit 11 transmits a part of the record of the job table 11a of the job to be executed to the storage control unit 12 (arrow A3 in FIG. 5).

  The storage control unit 12 receives the data shown in FIG. 11B, refers to the “number of distributed nodes” in this, and the “system operating days” in the switching time table 12a shown in FIG. A virtual node group that is a combination of the previous storage nodes is determined. At this time, by using the calculation formula (1) described above, in this example, the three storage nodes 21 of “No. 1, No. 2, No. 3” are determined as virtual node groups.

  The storage control unit 12 confirms whether these virtual node groups are accessible, and when there is a storage node that cannot be accessed (a node whose power is OFF), turns on the power of the storage node (the arrow in FIG. 5). A4). Thus, only the storage node determined as the virtual node group is turned on. In this example, there is no node that is newly powered on.

  Thereafter, the storage control unit 12 reads data from the backup target device 40 with reference to “backup target data” shown in FIG. 11B (arrow A5 in FIG. 5). Then, the parity of the value of the number of parity is added to the backup target data, and writing (distributed redundant arrangement) to each storage node of “No.1, No2, No3” constituting the virtual node group calculated earlier (FIG. 5). Arrow A6)

  After completion of writing, “1” is added to “generation” shown in FIG. 11B, and the information of the written data is added as a new record to the directory information table 12b as shown in FIG. The storage control unit 12 returns the “Write completion time” and “generation” to the job management unit 11, and the job management unit 11 sets the received values as “#” in the job table 11 a as shown in FIG. Enter in “Final Completion Time” and “Generation” respectively.

  The storage control unit 12 performs parity regeneration processing when parity regeneration of weekly data is necessary. This is because the weekly data write (backup) is executed once a week, so the storage node that wrote the latest generation of weekly data is turned off by rotating the storage node with the power on and off as described above. Depending on the case. That is, if the latest weekly data is not in the power ON node, the data cannot be read immediately. Such processing will be described with reference to FIG.

  First, in the above example, the “system operating days” when the weekly data is written is “1”, and the virtual node group to which the data has been written is indicated by the diagonal lines in FIG. No.1, No2, No, 3 ". Thereafter, when the “system operating days” becomes “3”, the storage node “No. 1” is turned off and the storage node “No. 5” is turned off as shown by the hatched line in FIG. Is turned on. At this time, the storage control unit 12 refers to the “job type” column and the “virtual node group” column of the directory information table 12b shown in FIG. 12A, the “job type” is “Weekly”, and the power is turned off earlier. The storage node that has been set is searched for data included in the “virtual node group”. If the corresponding data exists, the parity of the data on the storage node that is powered off is regenerated on the node that is powered on. In the example of FIG. 12A, since “Job02” is applicable, the data stored in the “No. 1” storage node in which the data of “Job02” is written and the power is newly turned off is stored. Parity is regenerated in the No. 5 storage node that is newly powered on (see arrow in FIG. 13C). At this time, the data stored in the storage node “No. 1” is reproduced using the data stored in the storage node where the other power is turned on. Alternatively, the data stored in the storage node “No. 1” may be moved to the storage “No. 5” as it is.

(restore)
Next, an operation when restoring backup data from the storage node group 20 will be described with reference to FIGS. Here, the processing for restoring the daily backup data described above will be described as an example.

  First, the user 30 inputs all items shown in FIG. 15A to the backup server 10 (arrow B1 in FIG. 14). For example, it is assumed that the user 30 inputs a value shown in FIG. Thereafter, the storage control unit 12 of the backup server 10 searches the directory information table 12b for “job name” and “generation” of the data shown in FIG.

  When the storage control unit 12 finds the data, the storage control unit 12 refers to the “virtual node group” and “data storage location” of the directory information table 12b, identifies the storage location of the backup data to be restored, and stores the data from the storage node group 20 Is read (arrow B2 in FIG. 14). At this time, if the storage node storing the backup data to be read cannot be accessed (when the power is OFF), the power of the storage node is turned on, and then the backup data is read.

  After the data is read, as shown in FIG. 15B, the data is restored to the restore destination 50 input by the user 30 (arrow B3 in FIG. 14). Thereafter, if there is a storage node whose power is turned on in the above-described processing, the power of the storage node is turned off.

  Here, the backup device in which the backup server and the grid type storage are integrated has been described above. However, a general-purpose backup server can be substituted for the backup server portion. In such a case, when a backup execution command is sent from the backup server to the storage node, information such as data to be backed up, the number of distributed nodes, the number of parity, etc. may be passed by a script. In this case, the directory information table 12b and the switching time table 12a of the storage control unit 12 need to be copied and arranged on each storage node.

  In addition, when the immediacy of new data write and the latest generation data read is not required, the power consumption of the storage node can be expected to be further reduced by starting up the storage node only at the time of data write / read. .

  In the above, it is assumed that the backup ends in one day. However, by setting the backup switching time in the switching time table 11a as every two days, etc., it is possible to support backups of two days or more. It is. Accordingly, the calculation formulas (1) and (2) may be changed as appropriate.

<Appendix>
Part or all of the above-described embodiment can be described as in the following supplementary notes. The outline of the configuration of the storage system (see FIG. 16), program, and distributed storage method in the present invention will be described below. However, the present invention is not limited to the following configuration.

(Appendix 1)
Storage processing content management means 101 for storing and managing job data including the number of storages representing the content of the data distributed storage processing and the number of storage devices to which the data is distributed;
Distributed storage control means 102 for storing the data to be stored in the distributed number of storage devices 110 based on the job data,
The distributed storage control means 102 determines at least the storage devices of the distributed number based on the distributed number, and turns on only the determined storage device.
Storage system 100.

(Appendix 2)
The storage system according to attachment 1, wherein
The distributed storage control means determines a storage device of a combination different from the determined storage device when the distributed storage processing was performed in the past;
Storage system.

(Appendix 3)
The storage system according to appendix 2,
The distributed storage control means includes a storage device that is a target of distributed storage among the storage devices determined when the distributed storage process was performed last time, and a combination of storages different from the determined storage device Determine a device, turn on the determined storage device, and perform distributed storage processing on the distributed number of storage devices among the determined storage devices;
Storage system.

(Appendix 4)
The storage system according to attachment 3, wherein
The distributed storage control means turns off the storage device that was not the target of the distributed storage among the storage devices that were turned on when the distributed storage process was performed last time, With the storage device turned on, the storage device that is the target of the distributed storage and a storage device that is different from the storage device are determined and the other storage device is determined. Turn on the power of the storage device, and perform distributed storage processing on the distributed number of storage devices including the other storage devices among the storage devices that are turned on.
Storage system.

(Appendix 5)
The storage system according to appendix 3 or 4,
The distributed storage control means is distributed and stored in the storage device when the power is turned off without determining the storage device that was the target of the distributed storage when the distributed storage processing was performed last time. Regenerate the data from the data stored in the storage device that is turned on and store it in the storage device that is turned on;
Storage system.

(Appendix 6)
The storage system according to any one of appendices 2 to 5,
The distributed storage control means determines a storage device based on the number of operating days of the storage system and the distributed number.
Storage system.

  According to the above invention, only the number of storage devices based on the number of distributions included in the job data are turned on, and the data is distributed and stored. At this time, since the storage device to be turned on is determined according to the number of distributions included in each job data, the storage device to be distributed and stored is not fixed. In particular, by appropriately changing the combination of storage devices for distributed storage to be rotated, it is possible to reduce power consumption and prevent data writing and reading from being concentrated on a specific storage device. The capacity of the stored data can be made uniform. As a result, it is possible to suppress a reduction in performance and a shortened life of the storage system while saving power.

  Furthermore, by including the storage device that was the target of distributed storage in the previous distributed processing in the storage device that is turned on, the latest data can be read at high speed while saving power. . In addition, even if the power of the storage device targeted for distributed storage in the previous distributed processing is off, the data stored in the storage device is regenerated from the storage device that is powered on. Thus, as described above, the latest data can be read at high speed while saving power.

(Appendix 7)
In the information processing device,
Storage processing content management means for storing and managing job data including the number of distributions representing the number of storage devices to which data is distributed;
Distributed storage control means for distributing and storing data to be stored in the distributed number of storage devices based on the job data,
The distributed storage control means determines at least the storage devices of the distributed number based on the distributed number, and turns on only the determined storage device;
program.

(Appendix 8)
The program according to appendix 7,
The distributed storage control means determines a storage device of a combination different from the determined storage device when the distributed storage processing was performed in the past;
program.

(Appendix 9)
Represents the contents of the distributed storage processing of data, stores and manages job data including the number of distributions representing the number of storage devices that distribute the data,
A distributed storage method for storing data to be stored in a distributed number of storage devices based on the job data,
Determining at least the number of storage devices based on the number of distributions, turning on only the determined storage devices, and performing distributed storage processing based on the job data;
Distributed storage method.

(Appendix 10)
The distributed storage method according to attachment 9, wherein
Determining a combination of storage devices different from the storage device determined when the distributed storage processing was performed in the past;
Distributed storage method.

(Appendix 11)
The distributed storage method according to appendix 10, wherein
Among the storage devices determined when the distributed storage processing was performed last time, the storage device that is the target of the distributed storage is determined, and a storage device of a combination different from the determined storage device is determined, and the storage device Turn on the power of the determined storage device, and perform distributed storage processing on the distributed number of storage devices among the determined storage devices.
Distributed storage method.

(Appendix 12)
The distributed storage method according to attachment 11, wherein
Of the storage devices that were turned on when the distributed storage process was performed last time, the storage devices that were not the target of distributed storage were turned off, and the power of the storage device that was the target of distributed storage The storage device that is a combination of the storage device that is the target of the distributed storage and another storage device that is different from the storage device is determined and the power of the other storage device is turned on. The distributed storage process is performed on the distributed number of storage devices including the other storage devices among the storage devices that are turned on.
Distributed storage method.

(Appendix 13)
The distributed storage method according to appendix 11 or 12,
If the storage device that was the target of distributed storage when the distributed storage processing was performed last time and the power was turned off, the data stored in the distributed storage device is turned on. Regenerated from the data distributed and stored in the storage device, and store it in the storage device that is turned on.
Distributed storage method.

  Note that the above-described program is stored in a storage device or recorded on a computer-readable recording medium. For example, the recording medium is a portable medium such as a flexible disk, an optical disk, a magneto-optical disk, and a semiconductor memory.

  Although the present invention has been described with reference to the above-described embodiment and the like, the present invention is not limited to the above-described embodiment. Various changes that can be understood by those skilled in the art can be made to the configuration and details of the present invention within the scope of the present invention.

DESCRIPTION OF SYMBOLS 1 Backup apparatus 10 Backup server 11 Job management part 11a Job table 12 Storage control part 12
12a Switching time table 12b Directory information table 20 Storage node group 21 Storage node 100 Storage system 101 Storage processing content management means 102 Distributed storage control means

Claims (6)

Storage processing content management means for storing and managing job data including the number of distributions representing the number of storage devices to which data is distributed;
Distributed storage control means for distributing and storing data to be stored in the distributed number of storage devices based on the job data,
The distributed storage control means determines at least the storage device of the distributed number based on the distribution number, turns on only the determined storage device , and then performs the determination when the distributed storage processing is performed last time Among the storage devices that were previously included in the storage device that was the target of distributed storage, and determined a storage device of a combination different from the determined storage device, power on the determined storage device, A distributed storage process is performed on the storage devices of the distributed number among the determined storage devices.
Storage system. The storage system according to claim 1,
The distributed storage control means turns off the storage device that was not the target of the distributed storage among the storage devices that were turned on when the distributed storage process was performed last time, With the storage device turned on, the storage device that is the target of the distributed storage and a storage device that is different from the storage device are determined and the other storage device is determined. Turn on the power of the storage device, and perform distributed storage processing on the distributed number of storage devices including the other storage devices among the storage devices that are turned on.
Storage system. The storage system according to claim 1 or 2,
The distributed storage control means is distributed and stored in the storage device when the power is turned off without determining the storage device that was the target of the distributed storage when the distributed storage processing was performed last time. Regenerate the data from the data stored in the storage device that is turned on and store it in the storage device that is turned on;
Storage system. The storage system according to any one of claims 1 to 3,
The distributed storage control means determines a storage device based on the number of operating days of the storage system and the distributed number.
Storage system. In the information processing device,
Storage processing content management means for storing and managing job data including the number of distributions representing the number of storage devices to which data is distributed;
Distributed storage control means for distributing and storing data to be stored in the distributed number of storage devices based on the job data,
The distributed storage control means determines at least the storage device of the distributed number based on the distribution number, turns on only the determined storage device , and then performs the determination when the distributed storage processing is performed last time Among the storage devices that were previously included in the storage device that was the target of distributed storage, and determined a storage device of a combination different from the determined storage device, power on the determined storage device, A distributed storage process is performed on the storage devices of the distributed number among the determined storage devices.
program. An information processing device stores and manages job data including the number of storages representing the number of storage devices to which the data is distributed, representing the content of the data distributed storage processing,
A distributed storage method in which the information processing device stores data to be stored in a distributed number of storage devices based on the job data,
The information processing apparatus, on the basis of the distributed number determines at least the distributed number of storage devices, the power storage device only with the determined turn on, have rows distributed storage processing based on the job data, then among the determined storage device when performing a distributed storage processing in the previous, wherein the storage device as a target of the distributed storage in previous and determines the storage device of a different combination from that of the determined storage device Then, power on the determined storage device is turned on, and distributed storage processing is performed for the distributed number of storage devices among the determined storage devices.
Distributed storage method.

Images