JP7548609B1 - Information processing device, information processing method, and program - Google Patents

Abstract

[Problem] To reduce degradation of access performance and load concentration to a storage device. [Solution] A host control unit 31 transmits and receives data to and from a host device 10. A disk control unit 32 is connected to a disk unit 40, and reads and writes data to a plurality of HDDs 41, ..., 41. The plurality of HDDs 41, ..., 41 configure a plurality of RAID configurations 42, ..., 42 for each predetermined unit. A deduplication data monitor 33 periodically or when instructed by a user refers to a deduplication management table 34, and determines the optimal logical disk from among a plurality of logical disks 51, 52 as the reference source for the deduplication data based on the number of reads and writes of the deduplication data, moves the deduplication data, and then updates the information in the deduplication management table 34. [Selected Figure] FIG.

Description

The present invention relates to an information processing device, an information processing method, and a program.

In recent years, with the increase in the amount of data handled by storage devices, it has become common to reduce the amount of data stored in a volume or the amount of data to be stored by using compression technology, deduplication technology, etc., in order to improve the utilization efficiency of the volume capacity. In addition, when storing data, storage systems that utilize not only hard disk drives (HDDs) and solid state drives (SSDs) within the storage device, but also external storage and cloud storage have become widespread. In the following, the storage device, external storage, cloud storage, etc. included in the storage system will be collectively referred to as the storage device, and when distinguishing between a storage device directly connected to a host device, it will be referred to as the local storage device.

In such storage systems, the aforementioned data deduplication technology is applied across multiple storage devices, improving the utilization efficiency of the volume capacity of the entire storage system (for example, see Patent Document 1).

JP 2020-106999 A

In data management using such deduplication technology, it is necessary to move the deduplication data depending on the access status of the deduplication data, and therefore processing such as updating the addresses of the reference source and reference destination of the deduplication data occurs. However, the deduplication technology in Patent Document 1 does not take into account the occurrence of data communication processing between storage devices and the effects of line delays in data communication, which causes problems such as reduced access performance and load concentration to the storage devices.

The present invention aims to provide an information processing device, information processing method, and program that reduce degradation of access performance and load concentration to a storage device.

In order to solve the above-mentioned problems, one aspect of the present invention is characterized in that, when storing and managing data that overlaps among multiple original data on multiple logical disks, a deduplication means stores the duplicated data as deduplication data on one logical disk of the multiple logical disks for one of the original data, and stores address information for referencing the deduplication data on other logical disks of the multiple logical disks for original data other than the one original data, a monitoring means for monitoring access status to the multiple logical disks, a destination determination means for determining a logical disk to which the deduplication data is to be moved in accordance with the access status, a moving means for moving the deduplication data to the determined destination logical disk for the deduplication data, and a setting information update means for updating information about the logical disk on which the deduplication data is stored and address information for referencing the deduplication data stored on the logical disk for the other original data in accordance with the determined destination of the deduplication data.

In addition, one aspect of the present invention is characterized in that, when storing and managing data that overlaps among multiple original data on multiple logical disks, the method includes the steps of: storing the overlapping data as deduplication data on one logical disk among the multiple logical disks for one of the original data, and storing address information for referencing the deduplication data on other logical disks among the multiple logical disks for the original data other than the one original data; monitoring access status to the multiple logical disks; determining a logical disk to which the deduplication data is to be moved according to the access status; moving the deduplication data to the determined logical disk to which the deduplication data is to be moved; and updating information about the logical disk on which the deduplication data is stored and address information for referencing the deduplication data stored on the logical disk for the other original data according to the determined destination of the deduplication data.

In addition, one aspect of the present invention is characterized in that, when storing and managing data that overlaps among multiple original data on multiple logical disks, a computer of an information processing device is configured to function as a deduplication function that stores the duplicated data as deduplication data on one logical disk of the multiple logical disks for one of the original data, and stores address information for referencing the deduplication data on logical disks of the other original data other than the one original data, a monitoring function that monitors the access status of the multiple logical disks, a destination determination function that determines a logical disk to which the deduplication data is to be moved in accordance with the access status, a moving function that moves the deduplication data to the determined destination logical disk for the deduplication data, and a setting information update function that updates information about the logical disk on which the deduplication data is stored and address information for referencing the deduplication data stored on the logical disk for the other original data in accordance with the determined destination of the deduplication data.

As explained above, this has the advantage of reducing degradation of access performance and load concentration to storage devices.

1 is a block diagram showing a configuration of an information processing system 1 according to an embodiment of the present invention. 3 is a conceptual diagram showing the data configuration of a deduplication management table 34 according to the present embodiment. FIG. 4 is a flowchart for explaining the operation of the information processing system 1 according to the present embodiment. FIG. 2 is a conceptual diagram for explaining a first operation example of the information processing system 1 according to the present embodiment. FIG. 2 is a conceptual diagram for explaining a first operation example of the information processing system 1 according to the present embodiment. FIG. 11 is a conceptual diagram for explaining a second operation example of the information processing system 1 according to the present embodiment. FIG. 11 is a conceptual diagram for explaining a second operation example of the information processing system 1 according to the present embodiment. FIG. 13 is a conceptual diagram for explaining a third operation example of the information processing system 1 according to the present embodiment. FIG. 13 is a conceptual diagram for explaining a third operation example of the information processing system 1 according to the present embodiment. FIG. 2 is a block diagram showing a minimum configuration of an information processing system according to the present embodiment.

The following describes an embodiment of the present invention with reference to the drawings.

A. Configuration of the embodiment Fig. 1 is a block diagram showing the configuration of an information processing system according to an embodiment of the present invention.
1, an information processing system 1 includes a host device 10, a storage device 20, an external storage 200, a cloud storage 300, and a network 400. The host device 10 transmits and receives various data to and from the storage device 20 to which it is directly connected.

The storage device 20 has a controller 30 and a disk unit 40. The controller 30 has a host control unit 31, a disk control unit 32, a deduplication data monitoring unit (deduplication means, monitoring means, destination determination means, movement means, configuration information update means) 33, and a deduplication management table 34. The disk unit 40 is made up of multiple HDDs (Hard Disk Drives) 41, ..., 41.

The host control unit 31 is connected to the host device 10 and transmits and receives data to and from the host device 10. There may be one or more host devices 10. The disk control unit 32 is connected to the disk unit 40 and reads and writes data to the multiple HDDs 41, ..., 41. The multiple HDDs 41, ..., 41 configure multiple RAID configurations 42, ..., 42 for each specified unit. The deduplication data monitoring unit 33 and the deduplication management table 34 will be described in detail later.

The storage device 20 is also connected to an external storage 200 and a cloud storage 300 via a network 400. The external storage 200 has a volume 210 consisting of a large-capacity storage medium, and the cloud storage 300 has a volume 310 consisting of a large-capacity storage medium.

The storage device 20 configures an LD (logical disk) 50 in a plurality of RAID configurations 42, ..., 42, and the volume 210 of the external storage 200 and the volume 310 of the cloud storage 300. The LD 50 is composed of a plurality of LDs 51, ..., 51, 52 ... 52 identified by an LDN (logical disk number; a number for identifying an LD). In the illustrated example, the LDs 51, ..., 51 correspond to the RAID configurations 42, ..., 42, respectively, and the LDs 52, ..., 52 correspond to the volume 210 of the external storage 200 and the volume 310 of the cloud storage 300, respectively. The actual data is stored in either the RAID configurations 42, ..., 42, the volume 210 of the external storage 200, or the volume 310 of the cloud storage 300. Note that HDD41, ..., 41 may be configured with other storage media, such as an SSD (Solid State Disk).

2 is a conceptual diagram showing the data configuration of the deduplication management table 34 according to this embodiment. In FIG. 2, the deduplication management table 34 is configured with a list for each LD 61 (#0, #1, ..., #N). The list is configured with LBA (logical block address) 62 for each unit (32 KB), deduplication status 63, number of deduplications (number of targets for deduplication processing) 64, reference LDN 65, reference LBA 66, data storage storage 67, number of reads 68, and number of writes 69, and holds each information for each LBA 62. The LBA 62 indicates the logical block address of the LD. The deduplication status 63 stores information indicating whether the deduplication data (actual data) is the reference source, the reference destination, or the deduplication has not been performed. Note that "reference source" indicates that deduplication data is stored, and "reference destination" indicates that deduplication data is being referenced. The number of deduplications 64 indicates the number of data entities indicating the same data that is the target of deduplication processing.

The deduplication management table 34 is composed of a list of LBAs 62 for identifying volumes every 32 KB, and whether data is duplicated (duplication judgment) is judged in units of LBAs 62, that is, every 32 KB of data. The duplication judgment is performed by a known method (for example, the method shown in the specification of U.S. Patent Application Publication No. 2007/0101074). The duplication judgment is a judgment of whether the data contained in the LBAs 62 match. The units of each LDN and LBAs 62 and the duplication judgment method are not limited to the values and methods described above. If it is judged that data is duplicated in multiple LBAs 62 as a result of the duplication judgment, the storage device 1 performs a deduplication process. As a result, the actual data is deleted from one LBA 62 that stores the duplicated data, and the one LBA 62 refers to the other LBA 62, and the actual data is stored only in the other LBA 62. In this case, the other LBA 62 is the reference source. As a result of the deduplication process, a flag indicating the source or destination of the reference is recorded in the deduplication status 63 column in the deduplication management table 34. The source of the reference indicates the LBA 62 that stores the actual data, and the destination of the reference indicates the LBA 62 that is referencing the actual data stored in the source of the reference as the destination of the reference.

The referenced LDN 65 indicates the LD 61 that is the reference of the deduplication data when deduplication is being performed. The referenced LBA 66 indicates the LBA 62 that is the reference of the deduplication data when deduplication is being performed. The referenced LDN 65 and the referenced LBA 66 indicate the address where the deduplication data is stored. Furthermore, the data storage 67 indicates the identifier of the storage device where the deduplication data is stored when deduplication is being performed. The read count 68 and the write count 69 store the number of read and write accesses to the LBA in question from other devices such as the host device 10 within a certain period of time.

Note that LBA (logical block address) is a method of specifying addresses in external storage devices such as HDDs, in which a serial number address is assigned to every unit (sector) that can be accessed on the storage medium, making it possible to access a recording location using that serial number. In this embodiment, too, the address where data is stored is specified using the LDN, which indicates the LD number, and the LBA, which indicates the address within the LD.

The deduplication data monitor 33 periodically or when instructed by the user checks the read count 68 and write count 69 of the deduplication data for each LD 61 and LBA 62 in the deduplication management table 34. The deduplication data monitor 33 determines the optimal LBA (logical storage) 62 as the reference source of the deduplication data based on the read count 68 and/or write count 69, and if the newly determined reference source LBA 62 is different from the current reference source LBA 62 of the deduplication data, moves the deduplication data to the storage area indicated by the newly determined reference source LBA 62. Thereafter, the deduplication status 43, deduplication count 44, reference destination LDN 45, reference destination LBA 46, and data storage storage 47 are updated for the reference source and reference destination LBA 62 of the moved deduplication data.

B. Operation of the embodiment Fig. 3 is a flowchart for explaining the operation of the information processing system 1 according to the present embodiment. Also, Figs. 4, 5, 6, 7, 8, and 9 are conceptual diagrams for explaining specific operation examples of the information processing system 1 according to the present embodiment.

As a specific example for the purpose of explanation, the storage device 20 is storage A, and the external storage 200 or cloud storage 300 is storage B. The storage device 20 is directly connected to the host device 10, and is therefore used to store deduplication data that is accessed relatively frequently. On the other hand, the external storage 200 or cloud storage 300 is accessed via the network 400, and is therefore susceptible to the effects of access performance, and is therefore used to store deduplication data that is accessed relatively infrequently.

The deduplication data monitoring unit 33 executes the following process for each LD 61 and LBA 62 in the deduplication management table 34 periodically or when instructed by the user. The deduplication data monitoring unit 33 first determines whether the deduplication status 63 of the corresponding LBA 62 of the monitored LD 61 is "not performed" (step S10). If the deduplication status 63 is "not performed" (YES in step S10), the deduplication data monitoring unit 33 sets the monitoring target to the next LBA (next 32 KB) (step S28) and returns to step S10.

On the other hand, if the deduplication status 63 is not "not yet performed", i.e., if the deduplication status 63 of the LBA 62 is "source" or "destination" (NO in step S10), the deduplication data monitor 33 obtains the read count 68 and write count 69 (I/O characteristics) for the monitored LBA 62 and all other LBAs 62 for which deduplication is being performed in relation to that LBA 62 from the deduplication management table 34 (step S12). In other words, the deduplication data monitor 33 obtains the read count 68 and write count 69 (I/O characteristics) for each LBA 62 that is in a reference source and reference destination relationship from the deduplication management table 34.

Next, the deduplication data monitoring unit 33 judges whether the read count 68 and the write count 69 are both equal to or less than a threshold (for example, the read count threshold is 10 and the write count threshold is 5) (step S14). If both are equal to or less than the threshold (YES in step S14), the deduplication data reference source is determined to be the external storage 200 or cloud storage 300 accessed via the network 400, i.e., LBA 62 of the data storage storage 67 "B" (step S20). In other words, for deduplication data that is accessed relatively infrequently, even if it is stored in the external storage 200 or cloud storage 300 "B" accessed via the network 400 (a storage with a relatively low capacity as a storage), a decrease in access performance is unlikely to occur. Note that even if a storage connected via the network 400 has a large amount of data transmission per unit time on the network line, the storage may be determined to have a relatively high capacity and another storage with a relatively low capacity may be determined as the new reference source.

Now, assume that the current (after deduplication processing) deduplication management table 34 is as shown in FIG. 4, for example, in LD 61 "#0", at LBA 62 "1000", the deduplication status 63 is "reference source", the deduplication count 64 is "3", the data storage storage 67 is "A", the read count 68 is "5", and the write count 69 is "0". Also assume that in LD 61 "#0", at LBA 62 "2000", the deduplication status 63 is "not performed", the data storage storage 67 is "A", the read count 68 is "10", and the write count 69 is "0". Note that in the figure, "-" means invalid (no data).

Similarly, in LD 61 "#1", at LBA 62 "11000", the deduplication status 63 is "referenced", the deduplication count 64 is "3", the referenced LDN is "0", the referenced LBA is "1000", the data storage storage 67 is "A", the read count 68 is "0", and the write count 69 is "0". Also, in LD 61 "#1", at LBA 62 "12000", the deduplication status 63 is "not performed", the data storage storage 67 is "A", the read count 68 is "5", and the write count 69 is "5". Note that the referenced LDN is "0" and the referenced LBA is "1000", indicating the referenced LD and LBA. In other words, this means that the actual deduplication data is stored in LBA 62 "1000" of LD 61 "#0".

In LD 61 "#200", at LBA 62 "201000", the deduplication status 63 is "referenced", the deduplication count 64 is "3", the referenced LDN is "0", the referenced LBA is "1000", the data storage storage 67 is "B", the read count 68 is "5", and the write count 69 is "0". In addition, in LD 61 "#200", at LBA 62 "202000", the deduplication status 63 is "not performed", the data storage storage 67 is "B", the read count 68 is "10", and the write count 69 is "0". The referenced LDN 65 is "0" and the referenced LBA 66 is "1000" as described above.

In the example shown in FIG. 4, in LD 61 "#0", the read count 68 at LBA 62 "1000" is "5" and the write count 69 is "0" (see the thick-lined rectangle). In this case, since both the read count 68 and the write count 69 are below the threshold, the access frequency is low, and therefore LBA 62 "201000" in LD "#200", which is data storage storage 67 "B (external storage 200 or cloud storage 300)" connected via the network, is determined as the reference source for deduplication data (see the thick-lined rectangle).

Next, the deduplication data monitoring unit 33 determines whether there has been a change from the current reference source LBA 62 to the newly determined reference source LBA 62, i.e., whether the deduplication status 63 of the newly determined reference source LBA 62 is already "reference source" (step S22). If there has been no change from the current reference source LBA 62 to the newly determined reference source LBA 62, i.e., if the deduplication status 63 of the newly determined reference source LBA 62 is already "reference source" (NO in step S22), the deduplication data monitoring unit 33 sets the next LBA (next 32 KB) to be monitored (step S28) and returns to step S10.

On the other hand, if the deduplication status 63 of the LBA 62 determined as the reference source is "reference destination" (YES in step S22), the deduplication data monitor 33 determines that the reference source has changed, and moves the actual deduplication data stored in the current reference source LBA 62 to the LBA 62 of the LD 61 determined as the new reference source (step S24). Next, the deduplication data monitor 33 updates each item in the deduplication management table 34 (step S26).

In the example shown in FIG. 4, the actual deduplication data stored in LBA 62 "1000" in LD 61 "#1", which is the current reference source, is moved to the storage area indicated by LBA 62 "201000" in LD 61 "#200", which has been newly determined as the reference source. Also, as shown by the diagonal hatching in FIG. 5, in LD 61 "#0", at LBA 62 "1000", the deduplication status 63 is updated from "reference source" to "referenced", the referenced LDN 65 is updated from invalid "-" to "200", and the referenced LBA 66 is updated from invalid "-" to "201000". Also, in LD 61 "#1", at LBA 62 "11000", the referenced LDN 65 is updated from "0" to "200", and the referenced LBA 66 is updated from "1000" to "201000". Then, in LD 61 "#200", at LBA 62 "201000", the deduplication status 63 is updated from "referenced" to "referenced", the referenced LDN 65 is updated from "0" to invalid "-", and the referenced LBA 66 is updated from "1000" to invalid "-".

Then, the deduplication data monitoring unit 33 sets the monitoring target to the next LBA (next 32 KB) (step S28) and returns to step S10.

In this way, when the read count 68 and write count 69 are both below the threshold, the reference source of the deduplication data is changed to the external storage 200 or cloud storage 300 accessed via the network 400, which is less affected by access performance, i.e., the LD 61 whose data storage storage 67 is "B" is changed to "#200". This maintains the access performance to other data with large read counts 68 and write counts 69, thereby reducing the degradation of access performance and load concentration to the storage device.

In the above-mentioned step S14, if the read count 68 and the write count 69 are not equal to or less than the threshold (NO in step S14), it is determined whether the read count 68 is equal to or greater than the threshold (for example, the threshold for the read count is 50) (step S16). If the read count 68 is equal to or greater than the threshold (YES in step S16), the reference source of the deduplication data is determined to be the storage device 20 directly connected to the host device 10, which is relatively easy to access (fast, relatively high capacity), that is, the LBA 62 where the data storage storage 67 is "A" (step S20). In other words, for deduplication data that is accessed relatively frequently, the storage device 20 directly connected to the host device 10 is less likely to cause a decrease in access performance than the external storage 200 or cloud storage 300 "B" accessed via the network 400.

Now, assume that the current (after deduplication processing) deduplication management table 34 is as shown in FIG. 6, for example, in LD 61 "#0", at LBA 62 "4000", the deduplication status 63 is "reference", the deduplication count 64 is "3", the reference LDN 65 is "200", the reference LBA 66 is "204000", the data storage storage 67 is "A", the read count 68 is "50", and the write count 69 is "0". Also assume that in LD 61 "#0", at LBA 62 "5000", the deduplication status 63 is "not performed", the data storage storage 67 is "A", the read count 68 is "10", and the write count 69 is "0". Note that in the figure, "-" means invalid (no data). Note that the referenced LDN is "200" and the referenced LBA is "204000" which indicates the referenced LD and LBA. In other words, this means that the deduplication data is stored at LDN "#200" and LBA 62 "204000".

Similarly, in LD 61 "#1", at LBA 62 "14000", the deduplication status 63 is "referenced", the deduplication count 64 is "3", the referenced LDN is "200", the referenced LBA is "204000", the data storage storage 67 is "A", the read count 68 is "20", and the write count 69 is "0". Also, in LD 61 "#1", at LBA 62 "15000", the deduplication status 63 is "not performed", the data storage storage 67 is "A", the read count 68 is "5", and the write count 69 is "5". The referenced LDN 65 is "200" and the referenced LBA 66 is "204000" as described above.

In LD 61 "#200", at LBA 62 "204000", the deduplication status 63 is "reference source", the deduplication count 64 is "3", the data storage storage 67 is "B", the read count 68 is "40", and the write count 69 is "0". In addition, in LD 61 "#200", at LBA 62 "205000", the deduplication status 63 is "not performed", the data storage storage 67 is "B", the read count 68 is "10", and the write count 69 is "0".

In the deduplication management table 34 shown in FIG. 6, the actual deduplication data is stored in the storage area indicated by LBA 62 "204000" in LD 61 "#200" (data storage storage "B" = external storage 200 or cloud storage 300), and this means that the actual deduplication data is referenced from LBA 62 "4000" in LD 61 "#0" and LBA 62 "14000" in LD 61 "#1".

In the example shown in FIG. 4, in LD 61 "#0", the read count 68 at LBA 62 "4000" is "50" (see the thick-lined rectangle). In this case, since the read count 68 is greater than or equal to the threshold, LD "#0", LBA 62 "4000", where the data storage storage 47 is "A (storage device 20)", is determined to be the reference source for the deduplication data (see the thick-lined rectangle).

Next, as described above, the deduplication data monitoring unit 33 determines whether there has been a change from the current reference source LBA 62 to the newly determined reference source LBA 62, i.e., whether the deduplication status 63 of the newly determined reference source LBA 62 is already "reference source" (step S22). If the deduplication status 63 of the newly determined reference source LBA 62 is already "reference source" (NO in step S22), the deduplication data monitoring unit 33 sets the next LBA (next 32 KB) to be monitored (step S28) and returns to step S10. In other words, if the deduplication status 63 of the newly determined reference source LBA 62 is already "reference source", there is no need to move the deduplication data, so the deduplication data monitoring unit 33 proceeds to the next LBA (next 32 KB) without updating the deduplication management table 34.

On the other hand, if the deduplication status 63 of the LBA 62 determined as the reference source is "reference destination" (YES in step S22), the deduplication data monitoring unit 33 moves the actual deduplication data to the LBA 62 of the LD 61 determined as the new reference source (step S24), as described above, and updates each item in the deduplication management table 34 (step S26).

In the example shown in FIG. 6, the deduplication data stored at LBA 62 "204000" in LD 61 "#200", which is the current reference source, is moved to the storage area indicated by LBA 62 "4000" in LD 61 "#0", which has been newly determined as the reference source. Also, as shown by the diagonal hatching in FIG. 7, in LD 61 "#0", at LBA 62 "4000", the deduplication status 63 is updated from "referenced" to "referenced", the referenced LDN 65 is updated from "200" to invalid "-", and the referenced LBA 66 is updated from "204000" to invalid "-". Also, in LD 61 "#1", at LBA 62 "14000", the referenced LDN 65 is updated from "200" to "0", and the referenced LBA 66 is updated from "204000" to "4000". Then, in LD 61 "#200", at LBA 62 "204000", the deduplication status 63 is updated from "source" to "destination", the destination LDN 65 is updated from invalid "-" to "0", and the destination LBA 66 is updated from invalid "-" to "4000".

Then, the deduplication data monitoring unit 33 sets the monitoring target to the next LBA (next 32 KB) (step S28) and returns to step S10.

In this way, when it is determined in step S16 that the read count 68 is equal to or greater than the threshold, the reference source of the deduplication data is changed to the storage device 20 that is directly connected to the host device 10 and whose access performance is unlikely to deteriorate, i.e., the LD 61 whose data storage storage 67 is "A" is changed to "#0", thereby reducing the deterioration of access performance and the concentration of load on the storage device.

In the above-mentioned step S16, if the number of reads 68 is not equal to or greater than the threshold (NO in step S16), it is determined whether the number of writes 69 to the reference source LBA is equal to or greater than the threshold (for example, the threshold for the number of writes is 30) (step S18). Then, if the number of writes 69 is equal to or greater than the threshold (YES in step S18), the reference source of the deduplication data is determined to be the LBA 62 with a relatively low access frequency (the sum of the number of reads and the number of writes is the smallest) (step S20). In other words, if the number of writes 69 to the reference source LBA is relatively high (greater than a predetermined threshold), the deduplication data is likely to be changed, and if it is changed, it is inappropriate to leave the actual deduplication data as the reference source in the first place. In other words, in the LBA 62 that refers to the reference source, it may be inappropriate to change the reference source, and in such a case, it is necessary to determine a new other LBA 62 to store the reference source. Therefore, it becomes necessary to execute the deduplication process to move the deduplication data and update the deduplication management table 34. Therefore, when the number of writes 69 is relatively high (higher than a predetermined threshold), it is more difficult to cause a decrease in access performance by changing the number of deduplication-depleted LBAs that is large and has the smallest total number of reads and writes to the LBA that is the reference source for the deduplication data.

Now, assume that the current (after deduplication processing) deduplication management table 34 is as shown in FIG. 8, for example, in LD 61 "#0", at LBA 62 "8000", the deduplication status 63 is "reference source", the deduplication count 64 is "5", the data storage storage 67 is "A", the read count 68 is "20", and the write count 69 is "30". Also, in LD 61 "#0", at LBA 62 "9000", the deduplication status 63 is "not performed", the data storage storage 67 is "A", the read count 68 is "5", and the write count 69 is "5". Note that in the figure, "-" means invalid (no data).

Similarly, in LD 61 "#1", at LBA 62 "18000", the deduplication status 63 is "referenced", the deduplication count 64 is "5", the referenced LDN is "0", the referenced LBA is "8000", the data storage storage 67 is "A", the read count 68 is "20", and the write count 69 is "35". Also, in LD 61 "#1", at LBA 62 "19000", the deduplication status 63 is "not performed", the data storage storage 67 is "A", the read count 68 is "5", and the write count 69 is "5". Note that the referenced LDN "0" and referenced LBA "8000" indicate the referenced LD and LBA; in other words, this means that the deduplication data is stored in LBA 62 "8000" of LDN "#0".

In LD 61 "#200", at LBA 62 "208000", the deduplication status 63 is "referenced", the deduplication count 64 is "5", the referenced LDN 65 is "0", the referenced LBA 66 is "8000", the data storage storage 67 is "B", the read count 68 is "5", and the write count 69 is "5". In LD 61 "#200", at LBA 62 "209000", the deduplication status 63 is "referenced", the deduplication count 64 is "5", the referenced LDN 65 is "0", the referenced LBA 66 is "8000", the data storage storage 67 is "B", the read count 68 is "5", and the write count 69 is "0". In LD 61 "#200", at LBA 62 "20A000", the deduplication status 63 is "referenced", the deduplication count 64 is "5", the referenced LDN 65 is "0", the referenced LBA 66 is "8000", the data storage 67 is "B", the read count 68 is "5", and the write count 69 is "0". The referenced LDN 65 is "0" and the referenced LBA 66 is "8000" as described above.

In the example shown in FIG. 8, in LD 61 "#0", the write count 69 at LBA 62 "8000" is "30" (see the thick rectangle). In this case, since the write count 69 is equal to or greater than the threshold, the number of LBAs being deduplicated is large, and LBA 42 "20A000" with the smallest total of read count 68 and write count 69 is selected as the reference source for the deduplication data (see the thick rectangle). Note that LBA 42 "209000" may also be used, and here LBA 42 "20A000" was randomly selected as the reference source for the deduplication data.

Next, as described above, the deduplication data monitoring unit 33 determines whether there has been a change from the current reference source LBA 62 to the newly determined reference source LBA 62, i.e., whether the deduplication status 63 of the newly determined reference source LBA 62 is already "reference source" (step S22). If the deduplication status 63 of the newly determined reference source LBA 62 is already "reference source" (NO in step S22), the deduplication data monitoring unit 33 sets the next LBA (next 32 KB) to be monitored (step S28) and returns to step S10. In other words, if the deduplication status 63 of the newly determined reference source LBA 62 is already "reference source", there is no need to move the deduplication data, so the deduplication data monitoring unit 33 proceeds to the next LBA (next 32 KB) without updating the deduplication management table 34.

On the other hand, if the deduplication status 63 of the LBA 62 determined as the reference source is "reference destination" (YES in step S22), the deduplication data monitoring unit 33 moves the deduplication data to the LBA 62 of the LD 61 newly determined as the reference source (step S24), as described above, and updates each item in the deduplication management table 34 (step S26).

In the example shown in FIG. 8, the deduplication status 63 of the newly determined reference source LBA 62 is "reference destination", so the deduplication data stored in LBA 62 "8000" of the current reference source LD 61 "#0" is moved to the storage area indicated by LBA 62 "20A000" of the newly determined reference source LD 61 "#200".

Also, as shown by the diagonal hatching in FIG. 9, in LD 61 "#0", at LBA 62 "8000", the deduplication status 63 is updated from "reference source" to "reference destination", the reference destination LDN 65 is updated from invalid "-" to "200", and the reference destination LBA 66 is updated from invalid "-" to "20A000". Also, in LD 61 "#1", at LBA 62 "18000", the reference destination LDN 65 is updated from "0" to "200", and the reference destination LBA 66 is updated from "8000" to "20A000".

In addition, in LD 61 "#200", at LBA 62 "208000", the referenced LDN 65 is updated from "0" to "200", and the referenced LBA 66 is updated from "8000" to "20A000". In LD 61 "#200", at LBA 62 "209000", the referenced LDN 65 is updated from "0" to "200", and the referenced LBA 66 is updated from "8000" to "20A000". In LD 61 "#200", at LBA 62 "20A000", the deduplication status 63 is updated from "referenced" to "source", the referenced LDN 65 is updated from "0" to invalid "-", and the referenced LBA 66 is updated from "8000" to invalid "-".

Then, the deduplication data monitoring unit 33 sets the monitoring target to the next LBA (next 32 KB) (step S28) and returns to step S10.

In this way, when the write count 69 is equal to or greater than the threshold, the reference source of the deduplication data is changed to the LBA with the largest number of deduplication-enabled LBAs and the smallest total of the read count 68 and write count 69. This reduces the frequency with which deduplication processing is performed when deduplication data is changed, and ultimately reduces degradation of access performance and load concentration on the storage device.

In the above embodiment, the reference source is determined using the read count 48 and/or write count 49, but the present invention is not limited to this and may be determined using all parameters that affect access performance, such as the load state of the storage device 20 and the access frequency of each host device 10.

According to the above-described embodiment, the deduplication data monitoring unit 33 determines the appropriate logical disk 50 to which the deduplication data is to be moved in accordance with a priority order according to the access status to the logical disk 50, thereby reducing the concentration of access to a specific storage device. The priority order can be set as follows according to the access status to the logical disk 50 (number of reads and/or writes):

According to the above-described embodiment, when the number of accesses (read and write counts) to the deduplicated data is equal to or less than the first threshold, the destination of the deduplicated data is set to the external storage 200 or cloud storage 300 connected via the network, thereby reducing the concentration of accesses to the storage device 20 itself.

Furthermore, according to the above-described embodiment, when the access (number of reads) to the deduplicated data is equal to or greater than the second threshold, the destination of the deduplicated data is set to the local storage device 20 that can transmit and receive data without going through a network that may cause line delays in data communication, thereby making it possible to avoid a decrease in access performance.

In addition, when the access (write count) to the deduplicated data is equal to or greater than the third threshold, the destination of the deduplicated data is the LBA of the logical disk 50 that has the largest number of deduplicated LBAs and the smallest total value of the read count 68 and the write count 69. This reduces the frequency with which the deduplicated data is changed and, as a result, reduces the degradation of access performance and the load concentration on the entire storage device.

FIG. 10 is a block diagram showing the minimum configuration of an information processing system according to this embodiment.
The information processing system according to the present embodiment may at least include a plurality of logical disks 90, 90, ..., 90, a deduplication means 100, a monitoring means 101, a migration destination determination means 102, a migration means 103, and a configuration information update means 104. The plurality of logical disks 90, 90, ..., 90 are constructed from a plurality of physical storage devices. When storing and managing data that is duplicated among a plurality of original data in the plurality of logical disks 90, 90, ..., 90, the deduplication means 100 stores the duplicated data as deduplication data in one logical disk 90 for one original data among the plurality of logical disks 90, 90, ..., 90, and stores address information for referencing the deduplication data in one logical disk 90 for other original data among the plurality of logical disks 90, 90, ..., 90. The monitoring means 101 monitors the access status to the plurality of logical disks 90, 90, ..., 90. The destination determination means 102 determines a destination logical disk 90 for the deduplication data according to the access status. The movement means 103 moves the deduplication data to the determined destination logical disk 90 for the deduplication data. The setting information update means 104 updates information about the logical disk 90 on which the deduplication data is stored and address information for referencing the deduplication data stored on the logical disk 90 for the other original data according to the determined destination for the deduplication data.

Although several embodiments of the present invention have been described above, the present invention is not limited to these and includes the inventions described in the claims and their equivalents.
The inventions described in the claims of this application are set forth below.

(Appendix 1)
1. An information processing apparatus comprising: a deduplication unit for storing and managing duplicate data among a plurality of original data on a plurality of logical disks, the deduplication unit storing the duplicate data as deduplication data in one logical disk among the plurality of logical disks for one of the original data, and storing address information for referencing the deduplication data in logical disks among the plurality of logical disks for other original data other than the one original data; a monitoring unit for monitoring access status to the plurality of logical disks; a destination determination unit for determining a logical disk to which the deduplication data is to be moved in accordance with the access status; a moving unit for moving the deduplication data to the determined logical disk to which the deduplication data is to be moved; and a setting information update unit for updating, in accordance with the determined destination of the deduplication data, information regarding the logical disk on which the deduplication data is stored and address information for referencing the deduplication data stored on the logical disk for the other original data.

(Appendix 2)
The information processing device described in Appendix 1, characterized in that the destination determination means determines a logical disk with relatively low access performance as the destination for the deduplicated data when the number of reads and writes to the multiple logical disks is below a predetermined first threshold.

(Appendix 3)
3. The information processing device according to claim 1, wherein the logical disk having a relatively low access performance is a storage device connected via a network.

(Appendix 4)
The information processing device according to any one of claims 2 to 3, characterized in that the destination determination means determines a logical disk with relatively high access performance as the destination for the deduplicated data when the number of reads to the plurality of logical disks is equal to or greater than a predetermined second threshold.

(Appendix 5)
5. The information processing device according to claim 2, wherein the logical disk having a relatively high access performance is a storage device that is directly connected without going through a network.

(Appendix 6)
The information processing device according to any one of appendices 1 to 5, characterized in that, when both the number of reads and the number of writes for the plurality of logical disks are equal to or greater than a predetermined third threshold, the destination determination means determines as the destination for the deduplicated data a logical disk having a large number of deduplicated data and the smallest total number of reads and writes.

(Appendix 7)
7. The information processing device of claim 1, wherein the destination determination means determines a logical disk with relatively low access performance as the destination for the deduplicated data if the number of reads and writes for the plurality of logical disks is below a predetermined first threshold, determines a logical disk with relatively high access performance as the destination for the deduplicated data if the number of reads for the plurality of logical disks is equal to or greater than a predetermined second threshold, and determines a logical disk with relatively high access performance as the destination for the deduplicated data if the number of reads and writes for the plurality of logical disks is equal to or greater than a predetermined third threshold.

In addition, a program for implementing the functions of the processing unit in the present invention may be recorded on a computer-readable recording medium, and the program recorded on the recording medium may be read into a computer system and executed to control the bonus information. The term "computer system" as used herein includes hardware such as an OS and peripheral devices. The term "computer system" may also include a plurality of computer devices connected via a network including the Internet, a WAN, a LAN, a dedicated line, and other communication lines. The term "computer-readable recording medium" refers to portable media such as flexible disks, optical magnetic disks, ROMs, and CD-ROMs, and storage devices such as hard disks built into a computer system. The term "computer-readable recording medium" also refers to storage devices that hold a program for a certain period of time, such as volatile memory (RAM) inside a computer system that becomes a server or client when a program is transmitted over a network. The program may also be one for implementing part of the above-mentioned functions. Furthermore, the program may be a so-called differential file (differential program) that can implement the above-mentioned functions in combination with a program already recorded in the computer system.

In addition, some or all of the above-mentioned functions may be realized as an integrated circuit such as an LSI (Large Scale Integration). Each of the above-mentioned functions may be individually processed, or some or all of them may be integrated into a processor. The integrated circuit method is not limited to LSI, and may be realized by a dedicated circuit or a general-purpose processor. Furthermore, if an integrated circuit technology that can replace LSI appears due to advances in semiconductor technology, an integrated circuit based on that technology may be used.

REFERENCE SIGNS LIST 1 Information processing system 10 Host device 20 Storage device 30 Controller 31 Host control unit 32 Disk control unit 33 Deduplication data monitoring unit 34 Deduplication management table 40 Disk unit 41 HDD
42 RAID configuration 50 LD (logical disk)
51, 52 LD (Logical Disk)
200 External storage 210 Volume 300 Cloud storage 310 Volume 61 LD (logical disk)
62 LBA (Logical Block Address)
63 Deduplication status 64 Number of deduplications 65 Reference LDN (logical disk number)
66 Reference LBA (Logical Block Address)
67 Data storage 68 Read count 69 Write count

Claims (9)

a deduplication means for storing and managing duplicate data among a plurality of original data on a plurality of logical disks, the deduplication means storing the duplicate data as deduplication data in one logical disk among the plurality of logical disks corresponding to one of the original data, and storing address information for referencing the deduplication data in other logical disks among the plurality of logical disks corresponding to the original data other than the one of the original data;
a monitoring means for monitoring an access status to the plurality of logical disks;
a destination determination unit that determines a destination logical disk to which the deduplication data is to be moved in accordance with the access status;
a migration means for migrating the deduplication data to the determined destination logical disk for the deduplication data;
a configuration information update unit that updates information about the logical disk in which the deduplication data is stored and address information for referencing the deduplication data stored on the logical disk for the other original data in accordance with the determined destination of the deduplication data;
An information processing device comprising: 2. The information processing device according to claim 1, wherein the destination determination means determines a logical disk determined to have low access performance as the destination of the deduplication data when the number of reads and writes to the plurality of logical disks is equal to or less than a predetermined first threshold. 3. The information processing apparatus according to claim 2, wherein the logical disk determined to have low access performance is a storage device connected via a network. 2. The information processing device according to claim 1, wherein the destination determination means determines a logical disk that is determined to have high access performance as the destination of the deduplication data when a number of reads from the plurality of logical disks is equal to or greater than a predetermined second threshold. 5. The information processing apparatus according to claim 4, wherein the logical disk determined to have high access performance is a storage device that is directly connected without going through a network. The information processing device according to claim 1, characterized in that, when both the number of reads and the number of writes for the plurality of logical disks are equal to or greater than a predetermined third threshold, the destination determination means determines, as the destination for the deduplicated data, a logical disk having a large number of deduplicated data and the smallest total number of reads and writes. The destination determination means
determining, when a number of reads and writes to the plurality of logical disks is equal to or less than a first predetermined threshold, a logical disk determined to have low access performance as a destination for the deduplication data;
determining, when a number of reads from the plurality of logical disks is equal to or greater than a second threshold, a logical disk determined to have high access performance as a destination for the deduplication data;
2. The information processing device according to claim 1, wherein when both the number of reads and the number of writes for the plurality of logical disks are equal to or greater than a predetermined third threshold, a logical disk having a large number of deduplicated data and the smallest total number of reads and writes is determined as the destination for the deduplicated data. a step of storing and managing duplicate data among a plurality of original data on a plurality of logical disks, storing the duplicate data as deduplication data in one logical disk among the plurality of logical disks for one of the original data, and storing address information for referencing the deduplication data in other logical disks among the plurality of logical disks for the original data other than the one of the original data;
monitoring an access status to the plurality of logical disks;
determining a logical disk to which the deduplication data is to be moved according to the access status;
migrating the deduplication data to the determined destination logical disk for the deduplication data;
updating information about the logical disk on which the deduplication data is stored and address information for referencing the deduplication data stored on the logical disk for the other original data, in accordance with the determined destination of the deduplication data. The computer of the information processing device,
a deduplication function for storing and managing duplicate data among a plurality of original data on a plurality of logical disks, the duplicate data being stored as deduplication data in one logical disk among the plurality of logical disks for one of the original data, and storing address information for referencing the deduplication data in other logical disks among the plurality of logical disks for the original data other than the one of the original data;
a monitoring function for monitoring access status to the plurality of logical disks;
a migration destination determination function that determines a logical disk to which the deduplication data is to be migrated in accordance with the access status;
a migration function for migrating the deduplication data to the determined destination logical disk of the deduplication data;
a setting information update function that updates information about the logical disk in which the deduplication data is stored and address information for referencing the deduplication data stored on the logical disk for the other original data, in accordance with the determined destination of the deduplication data;
A program characterized by causing the program to function as a

Images