JP2012058825A - Storage device and method for improvement in use efficiency of storage resource in storage device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To improve use efficiency of storage resource.SOLUTION: A storage device comprises: a storage area for storing data; a first management part for managing an area of data on a virtual disk and a specific value calculated based on the data in a manner that creates an association between them; and a second management part for managing an area of the data stored in the storage area and the specific value in a manner that creates an association between them. When receiving write data that specify an area on the virtual disk from a host device, the storage device calculates the specific value managed by the first management part based on the write data, and determines whether or not a specific value consistent with the calculated specific value is managed by the second management part. When determining that the consistent specific value is managed by the second management part, the storage device uses the consistent specific value to create an association between the area on the virtual disk managed by the first management part and the area of the data managed by the second management part, and avoids storage processing of the write data into the storage area.

Description

  The present invention relates to a storage apparatus and a method for improving the use efficiency of storage resources of the storage apparatus.

  A technology for effectively utilizing storage resources using a thin provisioning function in a storage apparatus is known. The thin provisioning function is a technology that makes it possible to use more storage capacity allocated to applications and servers than the disk capacity of physical storage that is actually allocated.

  In addition, a method of virtually unifying a plurality of network storages determines the network storage for storing objects based on the hash value generated from the logical identifier of the object to be accessed, and the object logical There is known a method for determining a network storage to process an access request by calculating a hash value of a typical identifier. According to this method, information on the storage position of the object can be acquired without inquiring of the management server (see, for example, Patent Document 1).

JP 2003-216474 A

  By using the thin provisioning function, the storage apparatus can appear to be able to use more capacity than the actual storage capacity for applications and servers. However, the thin provisioning function is configured such that when data is written to an unallocated block, the data is always allocated to the block. Therefore, the storage apparatus always performs a process of assigning data to blocks even if the same data as the written data exists in any area. Therefore, even a storage apparatus having a thin provisioning function still has room for improvement in terms of usage efficiency.

  Further, the technique described in Patent Document 1 does not improve the use efficiency of storage resources of the storage apparatus.

  The present invention has been made in view of the above circumstances, and an object of the present invention is to provide a storage apparatus that can improve the use efficiency of storage resources and a method for improving the use efficiency of storage resources of the storage apparatus.

  The present invention is a storage device that manages data using a virtual disk, and manages a storage area for storing data, an area on the virtual disk of data, and a specific value calculated based on the data in association with each other When the first management unit, the second management unit that manages the data area stored in the storage area in association with the specific value, and the write data designating the area on the virtual disk from the higher-level device are received, A specific value managed by the first management unit is calculated based on the write data, and it is determined whether or not a specific value that matches the calculated specific value is managed by the second management unit. If it is determined that the virtual disk area managed by the first management unit is associated with the data area managed by the second management unit using the matching specific value, the write data to the storage area Case A processing unit to avoid processing, characterized in that it comprises a.

  According to the present invention, it is possible to provide a storage apparatus that can improve the use efficiency of storage resources, and a method for improving the use efficiency of storage resources of the storage apparatus.

It is a figure which shows the whole structure of the storage apparatus which concerns on the 1st Embodiment of this invention. It is a figure which shows the structure of the controller part of the storage apparatus which concerns on the same 1st Embodiment. It is a figure which shows the actual data storage area of the storage apparatus which concerns on the same 1st Embodiment. It is a flowchart which shows the process of the storage apparatus which concerns on the same 1st Embodiment. It is a figure for demonstrating the replication process which concerns on the 2nd Embodiment of this invention. It is a figure for demonstrating the snapshot process which concerns on the 3rd Embodiment of this invention.

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

(First embodiment)
FIG. 1 is a diagram showing a configuration of the storage apparatus 10. As shown in the figure, the storage apparatus 10 has a controller unit 100 and an actual data storage area 200. The controller unit 100 controls storage of data in the actual data storage area 200. The actual data storage area 200 is an area for storing actual data. Since the other configuration of the storage apparatus 10 is the same as that of the conventional one, detailed description thereof is omitted.

  Further, the storage device 10 is connected to be communicable with the host device 20. The host device 20 transmits a write command to the storage device 10 or transmits a read command to the storage device.

  Next, the controller unit 100 will be described with reference to FIG. FIG. 2 is a diagram for explaining the function of the controller unit 100.

  The controller unit 100 includes a processing unit 110, logical disks 120 and 140 that are virtual disks, index tables 130 and 150 that are first management units, and a hash calculation unit 160. In FIG. 2, only two logical disks and index tables are shown for simplicity of explanation.

  When the processing unit 110 receives a write command from the higher-level device 20, more specifically, when receiving write data designating an area on the logical disk 120 (or 140), the processing unit 110 stores the write data in the actual data storage area 200. Process to store in.

  The logical disk 120 is a virtually configured disk. The host device 20 writes data to a predetermined area of the logical disk. When data is written to the logical disk in this way, the write data is processed by the processing unit 110, and the write data is stored in a predetermined area of the actual data storage area 200 corresponding to the predetermined area. The Since the logical disk 140 has the same description, detailed description thereof is omitted.

  The index table 130 is a table provided corresponding to the logical disk 120, and includes a field 131 that holds a logical block address (LBA) and a field 132 that holds a hash value calculated by the hash calculation unit 160. ing. The processing unit 110 manages the logical block address held in the field 131 and the hash value held in the field 132 in association with each other. Note that the description of the index table 150 is substantially the same as the description of the index table 130, and thus detailed description thereof is omitted.

  When data is written from the higher-level device 20 to the logical disk 120 (or 140), the hash calculation unit 160 has a hash value that is a specific value based on the write data held in a cache (not shown). Is calculated. Since the technique for calculating the hash value based on the predetermined data is the same as the conventional technique, the description thereof is omitted.

  Next, the actual data storage area 200 will be described with reference to FIG. FIG. 3 is a diagram for explaining the function of the actual data storage area 200.

  The actual data storage area 200 is an area for storing data and various types of information as described above. The actual data storage area 200 includes a data block information node area 210, which is a second management unit, an uncompressed data area 230, a compressed data area 240, and a data compression unit 250. The uncompressed data area 230 and the compressed data area 240 are storage resources for storing data.

  The data block information node area 210 is an area for holding an information node for each data block stored in the actual data storage area 200. The information node 220 includes various types of information for each hash value. In the present embodiment, the information node 220 is configured to include the number of links 221, access information 222, a location 223, and a location 224. The

  The number of links 221 indicates the number of index tables to be linked. Therefore, if the number of links is “1”, this indicates that the hash value is linked to one index table. If the number of links is “2”, the hash value is linked to two index tables. It is shown that.

  The access information 222 indicates the number of accesses to the data block stored in the actual data storage area 200 (for example, the number of accesses within a predetermined period, that is, the access frequency).

  The location 223 is a pointer indicating a predetermined area of the uncompressed data area 230 that stores actual data. The location 224 is a pointer indicating a predetermined area of the compressed data area 240 that stores the compressed data.

  The uncompressed data area 230 is an area for storing actual data. The compressed data area 240 is an area for storing actual data obtained by compressing data stored in the uncompressed data area 230.

  The data compression unit 250 refers to the access information 222 of each information node 220, compresses the data stored in the location 223 for the information node 220 with low access frequency, generates compressed data, and generates the generated compression Data is stored in the compressed data area 240. A pointer indicating the area where the compressed data is stored is stored in the location 224. Note that when the compressed data is stored in the compressed data area 240, the data compression unit 250 performs a process of releasing the area of the location where the uncompressed data is stored.

  Next, processing of the storage apparatus 10 when data is written by designating an area on the logical disk 120 from the host apparatus 20 will be described with reference to FIG.

  First, the processing unit 110 determines whether the logical block address of the write data is an address already linked to the information node 220 (S11).

  If it is determined that the logical block address has not been linked (S11: NO), the hash calculator 160 calculates a hash value based on the write data held in the cache (not shown) (S12).

  Next, the processing unit 110 stores the hash value calculated by the hash calculation unit 160 in the index table 130 in association with the logical block address (S13).

  Next, the processing unit 110 determines whether or not the information node 220 having the same hash value as the hash value is in the data block information node area 210 (S14).

  When it is determined that there is the same hash value (S14: YES), the processing unit 110 links the information node 220 and the logical block address of the index table 130, and increments the number of links of the information node 220 ( S15).

  When it is determined that there is no hash value with the same value (S14: NO), the processing unit 110 adds a new information node 220 to the data block information node area 210 and writes the contents of the data block to the uncompressed data area 230. Thereafter, a pointer indicating the written location is stored in the location 223 (S16).

  The processing unit 110 links the information node 220 and the logical block address of the index table 130 at the end of writing of the data block in step S16, and increments the number of links 221 to “1” (S17).

  On the other hand, if it is determined in step S11 that the logical block address has been linked, that is, if data is overwritten (S11: YES), the hash calculator 160 calculates a hash value based on the data to be overwritten. (S18).

  Next, the processing unit 110 associates the hash value calculated by the hash calculation unit 160 with the logical block address and holds it in the index table 130 (S19).

  Next, the processing unit 110 determines whether or not the information node 220 having the same hash value as the hash value is in the data block information node area 210 (S20).

  When it is determined that there is a hash value with the same value (S20: YES), the processing unit 110 determines whether the linked information node, that is, the information node 220 corresponding to the data to be overwritten and the logical block address of the index table 130 is the same. The link is released, and the number of links 221 of the information node 220 is decremented (S21).

  Next, the processing unit 110 links the information node 220 having the hash value calculated in step S18 and the logical block address of the index table 130, and increments the link number 221 of the information node 220 (S22).

  On the other hand, when the processing unit 110 determines that there is no hash value with the same value (S20: NO), the processing unit 110 adds a new information node 220 to the data block information node region 210 and adds it to the uncompressed data region 230. After writing the contents of the data block, a pointer indicating the written location is stored in the location 223 (S23).

  The processing unit 110 links the information node 220 and the logical block address of the index table 130 at the end of writing of the data block in step S16, and increments the number of links 221 to “1” (S24).

  Next, a case where the storage apparatus 10 receives an instruction to delete data will be described. For example, when the processing unit 110 receives an instruction to erase, for example, predetermined data on the logical disk 120 from the higher-level device 20, the processing unit 110 is linked to the logical block address of the index table 130 to be erased and the logical block address. The link with the information node 220 is released, and the link number 221 of the information node 220 is decremented.

  In the actual data storage area 200, if the number of links becomes “0” as a result of decrementing the number of links 221 of the information node 220, the data area (location 223 or location 224) of the information node 220 is , Released.

  As described above, the processing unit 110 writes data to the actual data storage area 200 or erases data via the logical disk 120 based on an instruction from the host device 20.

  Next, a process for compressing data stored in the actual data storage area 200 will be described.

  The data compression unit 250 monitors the access information 222 of each information node 220 in the data block information node area 210, and performs a process of compressing data with low access frequency in the background when the load on the storage device 10 is lightened. .

  More specifically, in the compression process, the data compression unit 250 compresses a data block with low access frequency stored in the location 223 to generate compressed data, and stores the generated data in the compressed data area 240. The location where the compressed data is stored is stored in the location 224 and the uncompressed data area 230 storing the data block is released.

  As described above, since the data block with low access frequency is moved from the uncompressed data area 230 to the compressed data area 240, the uncompressed data area 230 is recycled and effectively used according to the access frequency of the data block.

  In addition, when the compressed data is accessed after storing the compressed data in the compressed data area 240, the processing unit 110 holds the developed data in a cache (not shown), and the access frequency thereafter is high. When the predetermined frequency is exceeded, a process of storing the data held in the cache in the uncompressed data area 230 is performed. As a result, data that is temporarily stored in a compressed state with a low access frequency is stored in the uncompressed data area 230 in an uncompressed state when the access frequency exceeds a predetermined frequency. Become so.

  According to the first embodiment, when data is written by designating an area on the logical disk 120, the storage apparatus 10 has the same hash value as the data block information node calculated from the data. If stored in the information node 220 of the area 210, the location of the data managed by the information node 220 is associated with the logical block address of the index table 130, and the process of storing the data in the actual data storage area 200 is avoided. can do. For this reason, the storage apparatus 10 can improve the use efficiency of the actual data storage area 200.

  Further, the data compression unit 250 moves data having an access frequency lower than a predetermined frequency from the uncompressed data area 230 to the compressed data based on the access information 222. For this reason, the capacity of data stored in the actual data storage area 200 is reduced, and the storage apparatus 10 can further improve the use efficiency of the actual data storage area 200.

  Furthermore, since the compression processing of the data compression unit 250 is performed when the load on the storage apparatus 10 is light, the processing capacity of the storage apparatus 10 can be used efficiently, and the processing capacity does not decrease.

(Second Embodiment)
Next, a second embodiment will be described. In the second embodiment, a pair is constituted by a logical disk and another logical disk, and replication is performed from the logical disk to another logical disk. It is different from the form. Therefore, the different points will be described in detail below. In addition, the same code | symbol is attached | subjected to the structure same as 1st Embodiment, and detailed description is abbreviate | omitted.

  FIG. 5 is a diagram for explaining the replication processing. As shown in the figure, a logical disk (hereinafter referred to as a master volume in this second embodiment) 120 that replicates the contents of a logical disk from a logical disk (hereinafter referred to as a master volume in this second embodiment) A storage system having 320 is shown. That is, the master volume 120 and the remote volume 320 form a pair, and the master volume 120 and the remote volume 320 are replicated. Therefore, the storage system is configured to replicate data from the index table 130 to the index table 330 and from the actual data storage area 200 to the actual data storage area 400.

  The master volume 120 and the remote volume 320 may be provided in the same storage device. However, in the second embodiment, the case is provided in different storage devices that are geographically separated. explain.

  Further, in the storage system, when data is written to the master volume 120, synchronous copying is performed at high speed from the index table 130 of the master volume 120 to the index table 330 of the remote volume 320. Asynchronous copying is performed from the data storage area 200 to the actual data storage area 400 of the remote volume 320.

  Next, the replication process will be described.

  First, for example, when the storage apparatus 10 having the master volume 120 receives an instruction to form a pair with the remote volume 320 from the higher-level apparatus 20, the master volume 120 and the remote volume 320 are paired, and the index table 130 Data contents are copied to the table 330.

  In this way, when data is written from the higher level apparatus 20 to the master volume 120 after being paired, the updated contents of the index table 130 are copied to the index table 330 synchronously.

  Next, when the storage apparatus 10 receives a pair release instruction from the host apparatus 20 and the master volume 120 and the remote volume 320 are in a separate state, the index table 330 of the remote volume 320 stores actual data at the time of separation. The contents corresponding to the contents of the area 200 are retained. On the other hand, the contents of the index table 330 are different from the contents of the actual data storage area 400 at the time of separation. However, when the load on the storage apparatus 10 including the master volume 120 is light, data is copied from the actual data storage area 200 to the actual data storage area 400, and the contents of the index table 330 and the contents of the actual data storage area 400 are changed. Match.

  Therefore, the storage apparatus including the remote volume 320 is stored in the actual data storage area 200 via the index table 330 before the copy of the actual data storage area 400 from the actual data storage area 200 is completed after entering the separate state. Refer to the stored data. On the other hand, the storage device including the remote volume 320 is stored in the actual data storage area 400 via the index table 330 after the copy of the actual data storage area 400 from the actual data storage area 200 is completed after entering the separate state. Refer to the stored data.

  According to the second embodiment, when data is written to the master volume 120 in the pair state, only the contents of the index table 130 are synchronously copied, thereby shortening the response delay at the time of replication response, and the storage apparatus 10. The processing performance can be improved.

  Further, when the master volume 120 and the remote volume 320 are in a separate state, copying from the actual data storage area 200 to the actual data storage area 400 is executed in the background when the load on the storage device 10 including the master volume 120 is light. Therefore, data replication can be performed at a remote location while maintaining the performance of the storage apparatus 10 having the master volume 120.

(Third embodiment)
Next, a third embodiment will be described. The third embodiment is different from the first embodiment described above in that it is configured to create a snapshot of the logical disk 120. Therefore, the different points will be described in detail below. In addition, the same code | symbol is attached | subjected to the structure same as 1st Embodiment, and detailed description is abbreviate | omitted.

  FIG. 6 is a diagram for explaining snapshot processing of a logical disk. As shown in the figure, the storage system includes a logical disk (hereinafter referred to as a base volume in this third embodiment, 120), a previous generation logical disk (hereinafter referred to as a third volume in this third embodiment, 121, a logical disk of two generations before (hereinafter referred to as a snapshot volume in this third embodiment) 122, an index table 170, and a generation management as a third management unit A table 180 is provided.

  The index table 170 has a field 171 that holds a logical block address, a field 172 that holds a snapshot generation, and a field 173 that holds a hash value. In addition, the index table 170 associates the logical block address held in the field 171 with the snapshot generation held in the field 172 and the hash value held in the call field 173 so that the logical block address and snapshot generation are related. And hash values are managed.

  The generation management table 180 has a field 181 for holding a snapshot generation, a field 182 for holding a snapshot creation date, and a field 183 for holding management information (pointer) for associating with a snapshot generation of the index table 170. have. Further, the generation management table 180 associates the generation of the snapshot held in the field 181 with the date and time held in the field 182 and the management information held in the field 183, thereby generating the snapshot generation, date and time, and management. It is configured to manage information.

  According to the third embodiment, when an instruction to output a predetermined snapshot is received from the host apparatus 10, the storage apparatus 10 uses the management information corresponding to the predetermined generation using the generation management table 180. Then, the field 172 of the index table 170 is accessed, and the snapshot data of the logical disk 120 stored in the actual data storage area 200 is output to the host device based on the hash value corresponding to the predetermined generation. Can do. Accordingly, the storage apparatus 10 can instantaneously output the snapshot volume of the generation designated by the user to the upper level apparatus 20 using the generation management table 180 and the index table 170.

  Further, as described above, the storage apparatus 10 can switch the output to the snapshot volume of each generation using the generation management table 180 and the index table 170. Therefore, it is not necessary to provide a management table for managing differences between generations used in a conventional nap shot technique.

  In addition, this invention is not limited to the above-mentioned embodiment, A various deformation | transformation is possible in the case of the implementation.

  A part or all of the above embodiment can be described as in the following supplementary notes, but is not limited thereto.

(Appendix 1)
A storage device that manages data using a virtual disk,
A storage area for storing data; and
A first management unit that manages an area of the data on the virtual disk in association with a specific value calculated based on the data;
A second management unit that manages the data area stored in the storage area in association with the specific value;
When write data designating an area on the virtual disk is received from a host device, a specific value managed by the first management unit is calculated based on the write data, and a specific value that matches the calculated specific value It is determined whether or not the value is managed by the second management unit, and if it is determined that the value is managed, the virtual disk area managed by the first management unit using the matching specific value And a processing unit that associates an area of data managed by the second management unit and avoids the storage process of the write data in the storage area,
A storage apparatus comprising:

(Appendix 2)
When the processing unit receives an instruction to delete the write data from the higher-level device, the processing unit cancels the association between the first management unit and the second management unit using the matching specific value. The storage device according to supplementary note 1, which is characterized.

(Appendix 3)
The virtual disk is configured to be paired with another virtual disk that replicates the data of the virtual disk,
When write data is received from the host device via the virtual disk during the pair configuration of the virtual disk and the other virtual disk, the processing unit receives data managed by the first management unit. When data is copied to the first management unit associated with the other virtual disk and the pair is released, the data stored in the storage area is stored in the storage area associated with the second virtual disk. The storage apparatus according to appendix 1 or 2, wherein the storage apparatus is duplicated.

(Appendix 4)
A third management unit for managing one or more snapshots of the virtual disk;
The first management unit manages the specific value for each of the one or more snapshots,
The processing unit, when instructed to output a snapshot of a predetermined generation from the host device, a specific value managed by the first management unit, a specific value managed by the second management unit, and The storage apparatus according to claim 1 or 2, wherein the snapshot of the predetermined generation is output based on a specific value managed by the third management unit.

(Appendix 5)
A method for improving the use efficiency of storage resources of a storage device that manages data using virtual disks,
The storage device
A first management unit that manages an area of the data on the virtual disk in association with a specific value calculated based on the data;
A second management unit that manages the data area stored in the storage area in association with the specific value;
A step of calculating a specific value managed by the first management unit based on the write data when the write data is received from the host device via the virtual disk;
Determining whether a specific value that matches the calculated specific value is managed by the second management unit;
If it is determined that the management is performed in the step, the first management unit and the second management unit are associated with each other using the matching specific value, and the storage process of the write data in the storage area is avoided. And steps to
A method for improving the use efficiency of storage resources of a storage apparatus, comprising:

  The present invention can be widely applied to storage apparatuses that execute data storage processing using logical disks.

10... Host device 20... Storage device 100... Controller unit 110... Processing units 120, 121, 122, 140, 320 ... Logical disks 130, 150, 170, 330. Index table 180 ... generation management table 160 ... hash calculation unit 200, 400 ... actual data storage area 210 ... data block information node area 220 ... information node 230 ... uncompressed data area 240 ... Compressed data area 250 ... Data compression unit

Claims (5)

A storage device that manages data using a virtual disk,
A storage area for storing data; and
A first management unit that manages an area of the data on the virtual disk in association with a specific value calculated based on the data;
A second management unit that manages the data area stored in the storage area in association with the specific value;
When write data designating an area on the virtual disk is received from a host device, a specific value managed by the first management unit is calculated based on the write data, and a specific value that matches the calculated specific value It is determined whether or not the value is managed by the second management unit, and if it is determined that the value is managed, the virtual disk area managed by the first management unit using the matching specific value And a processing unit that associates an area of data managed by the second management unit and avoids the storage process of the write data in the storage area,
A storage apparatus comprising:   When the processing unit receives an instruction to delete the write data from the higher-level device, the processing unit cancels the association between the first management unit and the second management unit using the matching specific value. The storage apparatus according to claim 1. The virtual disk is configured to be paired with another virtual disk that replicates the data of the virtual disk,
When write data is received from the host device via the virtual disk during the pair configuration of the virtual disk and the other virtual disk, the processing unit receives data managed by the first management unit. Replicated to the first management unit associated with the other virtual disk, and when the pair is released, the data stored in the storage area is replicated to the storage area associated with the other virtual disk The storage apparatus according to claim 1 or 2, wherein: A third management unit for managing one or more snapshots of the virtual disk;
The first management unit manages the specific value for each of the one or more snapshots,
The processing unit, when instructed to output a snapshot of a predetermined generation from the host device, a specific value managed by the first management unit, a specific value managed by the second management unit, and The storage apparatus according to claim 1 or 2, wherein the snapshot of the predetermined generation is output based on a specific value managed by the third management unit. A method for improving the use efficiency of storage resources of a storage device that manages data using virtual disks,
The storage device
A first management unit that manages an area of the data on the virtual disk in association with a specific value calculated based on the data;
A second management unit that manages the data area stored in the storage area in association with the specific value;
A step of calculating a specific value managed by the first management unit based on the write data when receiving write data designating an area on the virtual disk from a host device;
Determining whether a specific value that matches the calculated specific value is managed by the second management unit;
If it is determined that the data is managed in the step, the virtual disk area managed by the first management unit and the data area managed by the second management unit using the matching specific value Associating and avoiding storage processing of the write data in the storage area;
A method for improving the use efficiency of storage resources of a storage apparatus, comprising:

Images