JP2012168840A - Storage system, storage method, and program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To make it possible to execute service in a transfer destination even when data copy is not completed and to prevent the execution speed of the program at the transfer destination from decreasing regardless of a bandwidth and distance of communication.SOLUTION: Access frequency statistic acquisition means 103 acquires a statistic value of a data read frequency in advance. Next, when data replication from a device A 100A, which is a replication source, to a device B, which is a replication destination, is instructed, based on the statistic value of the read frequency acquired by the access frequency statistic acquisition means 103, transfer means 104 transfers data having a high read use frequency from the device A 100A to the device B 100B. Then, data to be short when executing a program at the replication destination is transferred to the replication destination. In addition, as for data that volume management means 105 has written to the replication destination, a message that transfer is not necessary is sent to the replication source and transfer of unnecessary data is suppressed.

Description

  The present invention relates to a storage system for storing data while copying data from a host device to a storage medium, and more particularly to storing data in a disk array system that recognizes and uses a plurality of hard disks as a single hard disk. The present invention relates to a storage system for storing data while copying it to a medium, a method for copying a volume between disk arrays (storage method), and a program.

  In general, in a storage system or a disk array system, data from a host device is written to a storage medium such as a hard disk (HDD: Hard Disk Drive) such as a magnetic disk device or an optical disk device, a silicon disk, or a flash memory. The load on the system is balanced and improved by copying data within each device and between each device. In this case, data is copied between disk arrays for each volume which is a storage area management unit.

  Also, conventional storage systems and disk array systems usually have a RAID (Redundant Arrays of Independent Disks) configuration that handles a plurality of hard disks as a single storage device as a whole. By adopting such a RAID configuration, it is possible to improve the performance of access processing and the reliability of data storage. However, even if the RAID configuration is adopted, depending on the failure state in the device, access to data can always be guaranteed in the case of being connected by the same access interface or the like or in the case of failure of the device. Is not limited. Therefore, in order to prepare for the failure of such a device, a higher level of fault tolerance can be achieved by copying to a storage medium connected to a different interface in the device or copying to another external device. A technique for imparting sex is used. Copying between storage media in a device is called intra-cabinet replication (intra-cabinet replication), and copying between multiple devices is called inter-chassis replication (inter-cabinet replication). In this way, by replicating a database to another database and performing data synchronization, the reliability of data loss is improved.

  For inter-box replication, a data center that is separated by a certain communication distance to prevent the replication service from becoming impossible due to a power failure of the data center in which the device is stored or a disaster of the data center itself. Perform inter-box replication between devices. When the replication service is stopped in one data center, the replication service is continued using the data copied to the other data center.

  Furthermore, in recent years, it has been required to balance the load of the entire storage system by changing the processing load inside and outside the data center or by rebalancing the storage (data storage) capacity. Along with this, not only data movement during storage, but also development of federation (migration) technology that uses virtualization technology to move programs executed on the host device side between servers and data centers. It has been broken.

  The following technologies are disclosed as related technologies. For example, when a RAID group is divided into logical units, a swap area is automatically generated in a recording medium belonging to the RAID group. Further, the number of accesses of the host device is measured in units of blocks constituting the logical unit. When the access frequency of the block exceeds a predetermined number of times, the block is moved to the swap area of another RAID group of the RAID group to which the logical unit including the block belongs, and the block of the block of the host device is moved. Access is performed by the block moved into the swap area (see, for example, Patent Document 1). According to this technology, a configuration for load distribution can be facilitated in a disk array device in which a logical unit is provided in a RAID group.

  In addition, the following technologies are disclosed as other related technologies. For example, in a disk array device that manages addresses for a plurality of pools each composed of storage areas of a plurality of hard disk drives and accesses in response to a request from a host device, the pool is periodically The load status is monitored, and if there is a pool in which frequently accessed data is concentrated, the load is distributed by rearranging the frequently accessed data from the pool to another pool. As a result, a disk array device that can maintain an optimal load state can be realized (see, for example, Patent Document 2).

JP 2003-150324 A JP 2009-252106 A

  However, when performing data replication (duplication) in accordance with program federation (migration) due to load fluctuations, etc., the method of copying data in advance by determining the data replication destination in advance as described above, Whether to execute the program after the copy of the program is completed, or to move the data required by the program, that is, the data required by the federation destination program from the copy source to the copy destination, Need to do either.

  However, in the former case (the program is executed after the data copy is completed), the replication service cannot be executed until the data copy is completed, so the replication service can be executed quickly. Have difficulty. In the latter case (the program is executed after the data is moved from the copy source to the copy destination), the federation (migration) destination program may be changed due to a delay in access due to the network bandwidth or communication distance between devices. Execution speed may be significantly impaired. As a result, the service level of replication is greatly reduced.

  In addition, since the techniques disclosed in Patent Document 1 and Patent Document 2 move or relocate data with high access frequency to another area, the same data does not exist in the original area. Data redundancy is lost.

  The present invention has been made in view of the circumstances as described above, and can execute a replication service at a federation (migration) destination even if data copying is incomplete, An object of the present invention is to provide a storage system, a storage method, and a program that do not decrease the execution speed of the federation destination program regardless of the communication distance.

  The present invention has been made in order to solve the above-described problem. The present invention has one or a plurality of storage media, migrates a program between different higher-level devices, and transfers data from the migration source storage device. A storage system in which a migration destination storage apparatus executes a program based on copy data copied from an access frequency statistics acquisition for collecting read access frequency statistics for a storage area of a management unit accessed by the host apparatus And means for transferring data of only a storage area having a high access frequency to the different storage medium based on a statistical value of the access frequency collected by the access frequency statistics acquisition means when transferring data to a different storage medium And a copy for copying the data transferred by the migration means to the different storage medium in a management unit. To provide a storage system, characterized in that it comprises a chromatography-management unit.

  In addition, the present invention includes one or a plurality of storage media, performs program migration between different higher-level devices, and uses a copy data copied from data in the migration-source storage device as a migration-destination storage device Is a storage method for executing a program, wherein the first step of collecting statistical values of the read access frequency for the storage area of the management unit accessed by the host device, and when transferring data to a different storage medium, Based on the statistics of the access frequency collected in the first step, the second step of transferring the data of only the storage area with the high access frequency to the different storage medium, and the data transferred in the second step And a third step of copying data to the different storage medium in a management unit. To provide.

  In addition, the present invention includes one or a plurality of storage media, performs program migration between different higher-level devices, and uses a copy data copied from data in the migration-source storage device as a migration-destination storage device Is a program executed by an access frequency statistic acquisition unit that collects a statistical value of read access frequency for a storage area of a management unit accessed by the host device, and when transferring data to a different storage medium, Based on the access frequency statistics collected by the access frequency statistics acquisition means, a migration means for transferring only data in a storage area with a high access frequency to the different storage medium, and the data transferred by the migration means is different for each management unit. Providing a program to function as a volume management means for copying to a storage medium To.

  According to the present invention, statistics of read frequency are collected in advance for data to be replicated within a device or between devices, and when replication is instructed, data having a high frequency of read use is transferred, and replication is performed. The data that is missing when the program is executed first is transferred from the replication source to the replication destination. For data written to the replication destination, a message indicating that transfer is not required is sent to the replication source to suppress transfer of unnecessary data. As a result, it is possible to shorten the time until the execution of the program executed at the replication destination, and it is possible to minimize the decrease in performance when the program is executed.

1 is a block diagram showing a configuration of a storage system according to an embodiment of the present invention. It is a block diagram which shows the operation | movement form during the data copy in the storage system which concerns on one Embodiment of this invention. It is a block diagram which shows the operation | movement form after completion | finish of data copy in the storage system which concerns on one Embodiment of this invention.

  Hereinafter, a preferred embodiment of a volume copy method (storage method) between a storage system and a disk array according to the present invention will be described in detail with reference to the drawings.

<Embodiment>
FIG. 1 is a block diagram showing a configuration of a storage system according to an embodiment of the present invention. First, the configuration of the storage system shown in FIG. 1 will be described. This storage system includes a replication-source A device ((migration source storage device) 100A that transfers data, and a replication-destination B device that receives the data transferred from the A device 100A and stores the replicated data ( The storage device 100B is a migration destination storage device 100B, and both the A device 100A and the B device 100B are for reading (reading) / writing (writing) data from a host computer (not shown). Communication connection is established with a host device that issues commands.

  The replication source A apparatus 100A includes host request processing means 101A that transmits and receives commands and data from the host apparatus, communication means 102A that transfers data of the A apparatus 100A to the B apparatus 100B, and access information from the host apparatus. If the access information is read, the access frequency statistics acquisition means 103 for acquiring the access frequency indicating how many times the data on the storage medium 106A has been accessed, and the replication destination B apparatus 100B The configuration includes a transition unit 104 for instructing data transfer and a plurality of storage media 106A for storing data.

  The replication destination B device 100B receives the host request processing unit 101B that transmits and receives commands and data from the host device, the communication unit 102B that receives data transferred from the A device 100A, and the A device 100A. The configuration includes volume management means 105 that performs read / write processing of data in management units, and a plurality of storage media 106B that store data.

  Next, the operation of the storage system shown in FIG. 1 will be described. The access frequency statistics acquisition means 103 of the A apparatus 100A that is the replication source performs read or write processing on the storage medium 106A according to access from the host apparatus. At this time, when a replication instruction to the B apparatus 100B is given to the A apparatus 100A, the migration unit 104 transmits the communication apparatus 102A of the A apparatus 100A to the B apparatus 100B via the communication unit 102B of the B apparatus 100B. Data is copied (replicated) to the storage medium 106B.

  At that time, the migration unit 104 of the A apparatus 100A refers to the information of the access frequency statistics acquisition unit 103, and obtains a list of data in the area of the storage medium 106B having a high read access frequency. At this time, in determining whether the access frequency is high or low, a threshold value for a predetermined number of accesses is set in advance, and a value higher than the threshold value is assumed to be a high access frequency.

  Next, when the migration unit 104 obtains a list of data in the area of the storage medium 106B having a high read access frequency, the migration unit 104 reads the data existing in the list from the storage medium 106A, and reads the data from the storage medium 106A through the communication unit 102A. The transmitted data is transmitted to the B device 100B. As a result, on the B device 100B side, the data is received through the communication means 102B and written to the storage medium 106B through the volume management means 105.

  When a command is issued from the host apparatus to the B apparatus 100B, the host request processing unit 101B of the B apparatus 100B receives the command instruction. Then, the host request processing unit 101B sends a read or write request to the volume management unit 105. Thus, if the data indicated by the read command exists on the storage medium 106B in accordance with the contents of the read command or write command sent from the host device to the host request processing unit 101B, the volume management unit 105 converts the data to the host device. Send to device. If the data indicated by the read command is not on the storage medium 106B, the data is received from the A device 100A through the communication unit 102B, and then transmitted from the host request processing unit 101B of the B device 100B to the host device.

  If the command received by the host request processing unit 101B of the B apparatus 100B from the upper apparatus is a write command, the data indicated by the write command is received from the upper apparatus and this data is written to the storage medium 106B. At the same time, when the volume management means 106 writes the data of the write command received from the host device to the storage medium 106B, if it is not yet received from the A device 100A, the volume management means 106 stores the data in the storage medium 106B. At the same time as writing, a message indicating that it is not necessary to transfer the data is sent to the communication means 102A of the A apparatus 100a through the communication means 102B to suppress the transfer of extra data.

  Next, the overall operation of the storage system according to this embodiment will be described in detail with reference to FIGS. 1, 2, and 3. FIG. 2 is a configuration diagram showing an operation mode during data copying in the storage system according to the embodiment of the present invention, and FIG. 3 is a diagram after data copying is completed in the storage system according to the embodiment of the present invention. It is a block diagram which shows an operation | movement form.

  In the A apparatus 100A, when a volume, which is a storage area management unit, is initially allocated, the access frequency statistics acquisition unit 103 obtains the update frequency of only the read of access to the volume from the host request processing unit 101A. In addition, the number of updates is counted for each management unit (volume). As the management unit, if the access is block access such as a disk array, the minimum management unit is executed, or the management is executed as a chunk unit (a unit of a lump) in which a plurality of the minimum management units are combined.

  At this time, the capacity of the chunk is managed in units of several kilobytes to several gigabytes, and the access frequency is counted if there is read access to a part of the chunk, and if there is access that extends across read and write Counts in each chunk. Further, if the management unit of the storage for storing data is a file unit, the read access is counted for each file. Here, each update frequency count is processed in each chunk or in a table in units of files.

  Here, it is assumed that a data replication (duplication) instruction to the B apparatus 100B is given to the A apparatus 100A after some access to the A apparatus 100A. The state of program movement at this time will be described with reference to FIG. In FIG. 2, a read frequency high area 201 indicating an area with a high read access frequency for one volume (management unit), and a read frequency low indicating an area with a low read access frequency for one volume. The state where the area 202 exists is shown. At this time, whether the access frequency is high or low is determined depending on whether the access frequency is higher or lower than a predetermined threshold.

  Further, in FIG. 2, an area with high update frequency (high read frequency area 201) and an area with low update frequency (low read frequency area 202) are clearly divided, but these sections are virtually divided by a table or the like. Has been done. When a data replication (duplication) instruction is issued to the A apparatus 100A, the access frequency statistics acquisition unit 103 passes the list of the high read frequency area 201, which is an area with high access frequency in the volume on the storage medium 106A, to the migration unit 104. . Although not particularly displayed in the drawings of the present embodiment, the replication instruction as described above is always performed when data replication is performed at the time of program federation (migration) in conjunction with the host device. .

  The migration unit 104 of the A device 100A issues a data replication (duplication) instruction to the B device 100B using the communication unit 102A according to the list of the high read frequency region 201 indicating the high read frequency region. Accordingly, the B device 100B prepares an area for storing the replication data on the storage medium 106B for the replication data from the A device 100A. This area is divided and managed by the volume management unit 105 of the B apparatus 100B, and is divided into a read frequency high area 203 with a high update frequency and other areas 204, and the read update of the A apparatus 100A is high. The area (the high read frequency area 201) is copied to the area 203 of the B device 100B.

  When the B device 100B receives the replication data from the A device 100A through the communication unit 102B, the communication unit 102B of the B device 100B passes the replication data to the volume management unit 105. As a result, the volume management means 105 creates a volume for replication (duplication) on the storage medium 106B (management unit area) and manages which replication data is stored in the storage medium 106B. Do. This management method is performed in units of chunks or files.

  The volume management unit 105 of the B device 100B also manages the write data received from the host device through the host request management unit 101B, and also manages whether the data on the storage medium 106B has been updated by the migrated program. Do. This management method will be described later.

  Here, when all of the read update high areas 201, which are areas with a high volume update frequency in the A apparatus 100A, are copied to the B apparatus 100B, the migration unit 104 of the A apparatus 100A copies all the read update high areas 201. Is sent to the B device 100B.

  Then, the B device 100B receives a message indicating that the copying has been completed, notifies the volume management unit 105 of the message, and at the same time, determines whether the upper device connected to the host request processing unit 101B or the upper device is connected. The execution permission of the program is transmitted to a higher-level upper device to be managed.

  In the host device connected to the host request processing unit 101B of the B device 100B, the program is transferred from the host device connected to the host request processing unit 101A of the A device 100A in advance. Since such program migration is not directly related to the present invention, description thereof will be omitted in the present embodiment, but is generally performed as a virtual machine federation (migration) function. It is what. It should be noted that a conventional method such as distributed program execution may be used instead of a virtual machine.

  Next, the program of the host device on the B device 100B side that has obtained the execution permission performs read or write access processing in order to execute the program through the host request processing means 101B of the B device 100B. At that time, the volume management means 105 of the B apparatus 100B has already been copied from the A apparatus 100A to the storage medium 106B of the B apparatus 100B or written to the storage medium 106B through the host request processing means 101B of the B apparatus 100B. It is determined whether data exists in the area 301 of the B device 100B shown in FIG. That is, the volume management unit 105 determines whether or not data corresponding to data for which access has already been requested exists in the area 301 of FIG.

  If there is data in the area 203 shown in FIG. 3, the data on the area 203 stored in the storage medium 106B is transmitted to the host device through the host request processing means 101B. If the data is uncopied data, the volume management unit 105 requests the A device 100A to read the corresponding uncopied data through the communication unit 102B.

  On the other hand, when receiving a read request through the communication unit 102A, the A device 100A obtains volume (management unit) data from the storage medium 106A through the migration unit 104, and transmits this data to the device 100B through the communication unit 102A. . When the B device 100B receives this data through the communication unit 102B, the volume management unit 105 transmits this data to the host device through the host request processing unit 101B and also sends it to the volume area 204 on the storage medium 106B. Write data. At this time, it is assumed that the management information of the volume management unit 105 has been copied.

  Further, when a write occurs from the host apparatus to the B apparatus 100B, the area 301 for storing the write update data in FIG. 3 is created through the host request processing means 101B of the B apparatus 100B, and the host request processing means 101B. Is written to the area 301, and if the data corresponding to the write data exists in the area 203 and area 204 of the B device 100B shown in FIG. 2, the data is discarded.

  Further, when data is not transferred from the A device 100A to the B device 100B, that is, when there is no data in the region 203, the region 204, or the region 301 of the B device 100B, the data corresponding to the corresponding write data is stored. A message indicating that the transfer is not performed is sent to the volume management unit 105 of the A apparatus 100A through the communication unit 102B of the B apparatus 100B. This prevents extra data from being transferred.

  In this way, when the A device 100A takes statistics of the read frequency in advance and when replication is instructed from the A device 100A to the B device 100B, the read usage frequency from the A device 100A to the B device 100B Transfer of high data (data with high read frequency), and transfer of insufficient data when executing a program at the replication destination (B apparatus 100B) from the replication source (A apparatus 100A) to the replication destination (B apparatus 100B) ) For data written to the replication destination (B apparatus 100B), a message indicating that transfer is not required is sent to the replication source (A apparatus 100A), and unnecessary data transfer is suppressed, thereby replicating the replication destination (B apparatus 100B). It is possible to reduce the time until the execution of the program executed in (1), and to suppress a decrease in performance during execution of the program to a minimum.

  Further, in FIG. 3, after the copy of the area 201 with the high read update frequency of the A apparatus 100A is completed and the program is executed on the B apparatus 100B side, the data of the remaining area 202 with the low read update frequency is stored. By performing the transfer, it is possible to prevent further deterioration in performance. At this time, after the program is executed on the B device 100B side, similarly to the region 201 with a high read update frequency, the region 202 is transferred using the moving unit 104 of the A device 100A and the volume management unit 105 of the B device 100B. To the area 204.

  At this time, which data has been copied is managed by the volume management means 105. For example, if the command received by the host request processing unit 101B from the host device of the B device 100B is an access to the area 204, a data transfer request is sent to the A device 100A through the communication unit 102B as in the case described above. . Then, the data transferred from the A device 100A is written into the area 204 of the B device 100B, and at the same time, the volume management means 105 of the B device 100B holds information that the data has already been transferred.

  It is also conceivable that the program is interrupted from being executed by the host device on the B device 100B side and returned to the A device 100A side. This is the case where, for example, the load on the A device 100A side has decreased, and therefore it is not always necessary to continuously execute replication (duplication) on the B device 100B side. In this way, if it is possible to execute the program with an appropriate load on the A device 100A side, it is not necessary to maintain the environment on the B device 100B side. It is better to return the program. This can be realized by returning the area 301 managed on the B apparatus 100B side, that is, the area where the update by the write has occurred, to the A apparatus 100A side.

  On the A apparatus 100A side, data can be switched back by updating the area 201 or the area 202 in which data corresponding to the write data on the B apparatus 100B side is stored. In this processing, it is assumed that the A device 100A side is in a standby state, that is, the A device 100A side does not execute a program, but the A device 100A side also manages the write update, so that the A device 100A side also It is possible to execute the program and switch back the update of the write data from the B device 100B.

  At that time, a plurality of programs are executed on the A apparatus 100A side, and by having an image area for each write update, that is, by having an area corresponding to the area 301 shown in FIG. Execution and update switchback can be realized. At this time, the volume management means 105 of the B device 100B is also installed on the A device 100A side, and a function for switching the write update area is added by each program.

  In FIG. 1 of the present embodiment, the A device 100A and the B device 100B are written in a configuration having different means. In practice, however, all of these means are the same as the A device 100A and the B device 100B. You may be equipped with. That is, the A device 100A and the B device 100B may be implemented so as to include all of the access frequency statistics acquisition unit 103, the migration unit 104, and the volume management unit 105, respectively. Further, in the description of the storage system according to the present embodiment, the A device 100A and the B device 100B are composed of a plurality of (two) devices, but the means are distributed so that each processing means is provided in the same device. As a result, even in the same apparatus, the same operational effects as those of the above-described embodiment can be realized.

<Summary>
As described above, the storage system according to the embodiment of the present invention has a storage area of a single storage medium, a RAID storage area in which a plurality of storage media are collectively added to increase the apparent capacity, and the like. Statistics of read access frequency are collected for the volume (management unit) in. When data federation (migration) occurs due to virtualization, or when there is a need for data replication (duplication), the data in the storage area with high read access frequency is given priority to the replication destination device. Forwarding.

  At this time, after replication is completed in the replication destination apparatus in all or a part of the storage area where the read access frequency is high, the program is executed in the replication destination apparatus. At this time, when the program is executed at the replication destination, corresponding data is read from the replication source for the deficient data, and the write at the replication destination is held at the replication destination.

  For write, a means for transferring a message indicating that the transfer of the corresponding data to the replication source is unnecessary, or for discarding the corresponding data when the corresponding data is sent from the replication source to the replication destination. Thus, the replication destination apparatus does not need to wait for the transfer of all data on the storage medium. Furthermore, by transferring in advance data that is highly likely to be used from the replication source to the replication destination, it is possible to achieve data replication that minimizes performance degradation, and to execute programs at the replication destination. It is possible to reduce the time required for replication and minimize the performance degradation during replication.

  That is, according to the present invention, regarding a storage system or a disk array system using a storage medium such as a magnetic disk, an optical disk, a silicon disk, or a flash memory, a storage medium in a single device or between two or more devices In replication performed on storage media, even if a replication pair is not set up in advance, replication is not performed in a single device or between devices without waiting for the completion of copying all data at the replication destination. Regardless of the network bandwidth or communication distance used for performing the process, it is possible to prevent a decrease in the execution time and performance of the program at the replication destination.

  In other words, as in the conventional technology, when performing data replication (duplication) in accordance with program migration (federation) due to load fluctuations, the data replication destination is determined in advance and data is copied in advance. In such a method, execute the program after the data copy is completed at the replication destination, or execute the program after moving the data required by the federation destination program from the copy source, Either way must be performed.

  At this time, since the former cannot execute the replication service until the copying is completed, it is difficult to realize a rapid replication service. In the latter case, the execution speed of the replication destination program that federates data may be significantly impaired due to delays in access due to the network bandwidth or communication distance between devices, resulting in a high replication service level. It causes a decrease.

  However, with the storage system means of the present invention, read frequency statistics are collected in advance for data that is replicated within a single device or between multiple devices, and when replication is instructed, the frequency of read usage High priority data transfer. At this time, transfer processing of data that is insufficient when the program is executed at the replication destination is performed from the replication source to the replication destination. For data written to the replication destination, a message indicating that transfer is not required is sent to the replication source to suppress unnecessary data transfer. As a result, it is possible to shorten the time until the execution of the program executed at the replication destination, and to minimize the performance degradation during the execution of the program.

  As described above, the embodiments of the storage system according to the present invention have been described in detail with reference to the drawings. However, the specific configuration of the present invention is not limited to the contents of the above-described embodiments, and the gist of the present invention. Even if there is a design change or the like within a range that does not deviate from the above, they are all included in the present invention.

  The operation of the storage system described above is stored in a computer-readable storage medium in the form of a program, and the above-described processing is realized by the computer reading and executing this program. Here, the computer-readable storage medium refers to a magnetic disk, a magneto-optical disk, a CD-ROM, a DVD-ROM, a semiconductor memory, and the like. Alternatively, the computer program may be distributed to the computer via a communication line, and the computer that has received the distribution may execute the program.

  The program may be for realizing a part of the functions described above. Furthermore, what can implement | achieve the function mentioned above in combination with the program already recorded on the computer system, what is called a difference file (difference program) may be sufficient.

  According to the present invention, a plurality of non-volatile memories such as a disk array device and a flash memory having redundant hard disks such as a magnetic disk device and an optical disk device are provided, and programs, I / O (input / output), etc. are provided between these memories. The present invention can be effectively used in a computer system or the like in which a program is moved at any time due to a change in load or the like.

100A A device (migration source storage device)
100BB device (migration storage device 101A, 101B host request processing means 102A, 102B communication means 103 access frequency statistics acquisition means 104 migration means 105 volume management means 106A, 106B storage medium 201, 202, 203, 204, 301 storage area

Claims (10)

The migration destination storage apparatus has one or a plurality of storage media and performs migration of programs between different host apparatuses, and the migration destination storage apparatus executes the program by copy data copied from the migration source storage apparatus data. A storage system,
Access frequency statistics acquisition means for collecting statistical values of read access frequencies for the storage area of the management unit accessed by the host device;
When migrating data to a different storage medium, based on the access frequency statistics collected by the access frequency statistics acquisition means, a migration means for transferring only data in a storage area with a high access frequency to the different storage medium;
A storage system comprising: volume management means for copying data transferred by the migration means to the different storage medium in a management unit. The volume management means separately manages only the updated data when there is data update by a program in the migration destination storage device,
When the program is executed again by the migration source storage apparatus after execution of the program in the migration destination storage apparatus is completed, the migration means has a write update in the migration destination storage apparatus The storage system according to claim 1, wherein only data is transferred.   When the migration destination storage device returns the updated data to the migration source storage device, the volume management means manages the data updated by writing in the migration source storage device for each program, The storage system according to claim 2, wherein each of the programs is executed individually.   4. The storage system according to claim 3, wherein the volume management means transmits a message to the migration means that data transferred to the migration destination storage apparatus is not required to be transferred.   4. The storage system according to claim 3, wherein the volume management unit discards the data when already written data is transferred from the migration source storage apparatus. The migration destination storage apparatus has one or a plurality of storage media and performs migration of programs between different host apparatuses, and the migration destination storage apparatus executes the program by copy data copied from the migration source storage apparatus data. A storage method,
A first step of collecting a statistical value of a read access frequency for a storage area of a management unit accessed by the host device;
A second step of transferring data of only a storage area having a high access frequency to the different storage medium based on the statistical value of the access frequency collected in the first step when transferring data to a different storage medium; ,
And a third step of copying the data transferred in the second step to the different storage medium in a management unit. A fourth step of separately managing only the updated data when there is a data update by a program in the migration destination storage apparatus;
After the execution of the program in the migration destination storage device is completed, when the program is executed again by the migration source storage device, only the data whose write has been updated in the migration destination storage device is transferred. The storage method according to claim 6, further comprising a fifth step.   When the migration destination storage apparatus returns the updated data to the migration source storage apparatus, the data updated by writing in the migration source storage apparatus is managed for each program, and each of the programs is individually managed. The storage method according to claim 7, further comprising a sixth step of executing.   8. The storage method according to claim 7, further comprising a seventh step of transmitting to the migration source storage apparatus a message indicating that transfer of data written to the migration destination storage apparatus is unnecessary. . A program that has one or a plurality of storage media, performs program migration between different host devices, and is executed by the migration destination storage device using copy data copied from the data of the migration source storage device. And
Computer
Access frequency statistics acquisition means for collecting statistical values of read access frequency for the storage area of the management unit accessed by the host device;
A transfer means for transferring data of only a storage area having a high access frequency to the different storage medium based on a statistical value of the access frequency collected by the access frequency statistics acquisition means when transferring data to a different storage medium;
A program for functioning as volume management means for copying data transferred by the migration means to the different storage medium in a management unit.

Images