WO2014087465A1 - Storage device and storage device migration method - Google Patents

Abstract

A migration destination primary storage device (MD-PDKC) is connected to a migration source primary storage device (MS-PDKC) that provides a first primary logical volume (a first PVOL) that forms a pair with a secondary logical volume (SVOL) of a secondary storage device (SDKC). The migration destination primary storage device has a second PVOL and, by means of first identification information, manages a virtual primary logical volume (PVVOL) that virtualizes the first PVOL, and copies data from the PVVOL to the second PVOL. When a first write request specifying the first identification information is received prior to the completion of the aforementioned copying, the MD-PDKC transmits the write data subject to the first write request to the MS-PDKC. When copying is completed the MD-PDKC changes the identification information for the second PVOL to the first identification information, and when a second write request specifying the first identification information is received, the MD-PDKC writes the write data subject to the second write request to the second PVOL, and transfers this data to the SDKC.

Description

Storage device and storage device migration method

The present invention relates to a technique for migrating a volume of a primary or secondary storage device to another storage device in a remote copy system having a primary storage device and a secondary storage device.

A remote copy system having a primary storage device (PDKC) and a secondary storage device (SDKC) for copying data from a primary logical volume (PVOL) of PDKC to a secondary logical volume (SVOL) of SDKC is known. Remote copy includes synchronous remote copy that writes data to the SVOL in synchronization with the writing of data to the PVOL, and asynchronous remote copy that writes data to the SVOL asynchronously with writing of data to the PVOL.

For example, Patent Document 1 discloses a technique related to synchronous remote copy, which is a technique for migrating a PVOL to another storage apparatus.

JP 2005-182222 A

In the technology according to Patent Document 1, it is necessary to install a virtual management network in the remote copy system and to have a function for the migration source PDKC to migrate the PVOL.

Here, the PVOL migration situation is that the old model storage device that manages the synchronous remote copy PVOL currently in operation is the migration source, and the newly installed storage device is the migration destination. It is conceivable to migrate PVOL. In such a situation, in the technique according to Patent Document 1, it is necessary for the migration source PDKC to have a function for migrating the PVOL, but the migration source PDKC does not always have such a function in advance. First, it is necessary to provide the migration source PDKC with this function, which is troublesome. In some cases, it may be impossible to provide the necessary functions in the source PDKC.

The above problem may also exist in the case where the SVOL is migrated from the migration source SDKC to the migration destination SDKC.

The migration destination PDKC connected to the migration source PDKC that provides the first PVOL that forms a pair with the SVOL of the SDKC manages the second PVOL and the virtual primary logical volume (PVVOL) in which the first PVOL is virtualized using the first identification information, And a control device that performs copying of data from the PVVOL to the second PVOL. When the control device receives the first write request designating the first identification information before the copying is completed, the control device transmits the write data according to the first write request to the migration source PDKC. After the copy is completed, the control device changes the identification information of the second PVOL to the first identification information, and when receiving the second write request designating the first identification information, the control device changes the write data according to the second write request, Write to the second PVOL and transfer to SDKC.

According to the present invention, the migration source PDKC can be migrated to the migration destination PDKC while continuing the synchronous remote copy without adding a special function to the migration source PDKC.

FIG. 1 is a diagram illustrating an example of an overview of the first embodiment. FIG. 2 is a diagram illustrating a configuration example of the computer system according to the first embodiment. FIG. 3 is a diagram illustrating an example of a state of the computer system before the start of migration. FIG. 4 is a diagram for explaining the introduction of MD-DKC into the computer system. FIG. 5 is a diagram for explaining port virtualization in MD-DKC. FIG. 6 is a diagram for explaining generation of VVOL in MD-DKC. FIG. 7 is a diagram for explaining VVOL virtualization in MD-DKC. FIG. 8 is a diagram for explaining the addition of a path to the MD-DKC in the host. FIG. 9 is a diagram for explaining the deletion of the path to the MS-DKC in the host. FIG. 10 is a diagram for explaining reading of pair information from the MS-DKC. FIG. 11 is a diagram for explaining addition of an RC path. FIG. 12 is a diagram for explaining the RC path viewed from the MS-DKC. FIG. 13 is a diagram for explaining deletion of an RC path between MS and DKC. FIG. 14 is a first diagram illustrating data migration of the MS-VOL according to the first embodiment. FIG. 15 is a second diagram illustrating data migration of the MS-VOL according to the first embodiment. FIG. 16 is a third diagram for explaining data migration of the MS-VOL according to the first embodiment. FIG. 17 is a diagram for explaining stop of remote copy between MS and DKC. FIG. 18 is a diagram illustrating an example of the state of the computer system after completion of migration. FIG. 19 is a first flowchart of write processing according to the first embodiment. FIG. 20 is a second flowchart of the write process according to the first embodiment. FIG. 21 is a third flowchart of the write process according to the first embodiment. FIG. 22 is a diagram illustrating addition of an RC path according to the second embodiment. FIG. 23 is a diagram for explaining data migration of the MS-VOL according to the second embodiment. FIG. 24 is a diagram illustrating addition of an RC path according to the third embodiment. FIG. 25 is a diagram for explaining data migration of the MS-VOL according to the third embodiment.

Hereinafter, some examples will be described.

Hereinafter, in this specification, terms will be used as follows.
(*) “PDEV” is an abbreviation for a physical storage device (typically a nonvolatile storage device). The PDEV includes, for example, a storage medium (typically a non-volatile storage medium) and a controller that controls data input / output (I / O) for the storage medium. The nonvolatile storage medium may be a hard disk group (one or more hard disks) or an NVM group (one or more nonvolatile semiconductor memories (NVM)). That is, the PDEV may be, for example, an HDD (Hard Disk Drive) or an NVM device such as a flash memory (FM) device (for example, SSD (Solid State Device)). The NVM may be, for example, FM or phase change memory. For example, the FM may include a plurality of blocks (physical blocks), and each block may include a plurality of pages (physical pages). In NVM (typically FM), data is written in units of pages. If data in a page is not erased, data cannot be written in the page, and data is erased in units of blocks (for example, NAND) Type FM).
(*) “VOL” is an abbreviation for logical volume and is a logical storage device. The VOL may be a substantial VOL (RVOL) or a virtual VOL (VVOL). The VOL is an online VOL provided to an external device (for example, a host computer) connected to the DKC having the VOL, and an offline VOL not provided to the external device (not recognized by the external device). There is good. “RVOL” is a physical storage resource (for example, RAID (Redundant Array of Independent (or Inexpensive) Disks) configured by a plurality of non-volatile physical storage devices (PDEVs)) included in the storage apparatus having the RVOL. VOL based on (Group). “VVOL” is, for example, an externally connected VOL (EVOL) that is a VOL based on storage resources (for example, a logical volume) of an external storage device connected to the storage device having the VVOL and is in accordance with the storage virtualization technology. There may be a VOL (TPVOL) configured with a plurality of virtual pages (virtual storage areas) and following a capacity virtualization technology (typically Thin Provisioning). The TPVOL is typically an online VOL. A real page is allocated from the pool to the TPVOL. A “pool” is a storage resource composed of a plurality of real pages (substantial storage areas). The pool is composed of a plurality of pool VOLs, for example, and each pool VOL is divided into two or more real pages. The “pool VOL” is a VOL that is a component of the pool. The pool VOL may be an RVOL or an EVOL. The pool VOL is typically an offline VOL.
(*) “PVOL” is an abbreviation for the primary logical volume, and is a word representing the copy-source VOL. “SVOL” is an abbreviation for a secondary logical volume, and is a word representing a copy-destination VOL. A pair of PVOL and SVOL is typically a VOL pair. Data is copied from PVOL to SVOL.
(*) “DKC” is a word representing a storage device. The DKC has a VOL and a controller that controls data I / O to the VOL. The DKC may include one or a plurality of PDEVs (for example, one or RAID group), and the VOL included in the DKC may be an RVOL or a VVOL.
(*) “PDKC” is an abbreviation for primary DKC, and is a DKC having a PVOL. “SDKC” is an abbreviation for secondary DKC, and is an abbreviation for DKC having SVOL. The remote copy system typically has a PDKC and an SDKC, and the PVOL and SVOL constituting the VOL pair are present in different DKCs.
(*) “MS” is an abbreviation for migration source. For example, “MS-PDKC” means the source PDKC.
(*) “MD” is an abbreviation for migration destination. For example, “MD-PDKC” means a migration destination PDKC.

In the following description, various types of information may be described using the expression “aaa table”, but the various types of information may be expressed using a data structure other than a table. The “aaa table” can be referred to as “aaa information” to indicate that the various types of information do not depend on the data structure. Further, when a program or software is used as a subject, it is understood that a processor or the like actually executes the program or software.

FIG. 1 is a diagram illustrating an example of an outline of the first embodiment.

This computer system is a computer system for transferring PVOL 105 and SVOL 108 to MD-PVOL 107 and MD-SVOL 110 while maintaining a remote copy.

The computer system has one or more hosts (host computers) 200 and a remote copy system. The remote copy system has a plurality of DKCs (storage devices) 100. In this embodiment, the remote copy system includes an MS-PDKC 100A that manages the PVOL 105 that is the MS (migration source), an MS-SDKC 100B that manages the SVOL 108 that is the copy destination VOL of the PVOL 105, and an MD (migration destination) of the PVOL 105. The MD-PDKC 100C that manages the MD-PVOL 107 that is the MD and the MD-SDKC 100D that manages the MD-SVOL 110 that is the MD of the SVOL 108 are included.

MS-PDKC 100A and MS-SDKC 100B are connected via a communication network 60. The MD-PDKC 100C and the MD-SDKC 100D are connected via a communication network 60. In this embodiment, the PVOL 105 and SVOL 108 of the MS-DKC group 30 including the MS-PDKC 100A and the MS-SDKC 100B are transferred to the MD-DKC group 40 including the MD-PDKC 100C and the MD-SDKC 100D.

Furthermore, the MS-PDKC 100A and the MD-PDKC 100C are installed, for example, in the same local site 10 and are connected so as to communicate with each other. In this embodiment, the MD-PDKC 100C virtualizes the communication interface device (port) so that the MS-PDKC 100A can be seen as the MS-SDKC 100B. Further, the MS-SDKC 100B and the MD-SDKC 100D are installed at a remote site 20 which is a place distant from the local site 10, and are connected so as to be communicable with each other. In this embodiment, the MD-SDKC 100D virtualizes the communication interface device (port) with respect to the MS-SDKC 100B so that it can be seen as the MS-PDKC 100A.

MD-PDKC 100C manages PVVOL 106 obtained by virtualizing externally connected PVOL 105 based on storage virtualization technology. The host 200 and the MD-PDKC 100C are connected via the communication network 50. The host 200 recognizes a path that can access the PVVOL 106 of the MD-PDKC 100C. The MD-SDKC 100D manages the SVVOL 109 obtained by virtualizing the externally connected SVOL 108 based on the storage virtualization technology.

In this computer system, the MD-PDKC 100C executes a PVOL migration process for migrating (for example, copying) the data of the PVOL 105 of the MS-PDKC 100A to the MD-PVOL 107 (FIG. 1 (7)). An SVOL migration process (FIG. 1 (8)) for migrating the data of the SVOL 108 of the MS-SDKC 100B to the MD-SVOL 110 is executed.

In the computer system, the host 200 can write (write) to the PVOL 105 in parallel with the PVOL migration processing and the SVOL migration processing. Specifically, when writing to the PVOL 105 of the MS-PDKC 100A (during host I / O), the host 200 transmits a write request for the PVVOL 106 to the MD-PDKC 100C ((1) in FIG. 1).

When the MD-PDKC 100C receives a write request for the PVVOL 106 from the host 200, the MD-PDKC 100C transfers the write request to the MS-PDKC 100A ((2) in FIG. 1).

MS-PDKC 100A executes a write process to PVOL 105 in accordance with the received write request and also executes a remote copy corresponding to the write request. In this embodiment, the MS-PDKC 100A is connected to the MS-SDKC 100B via the communication network 60 (FIG. 1 (3)) and the MD-PDKC 100C as a path for remotely copying the PVOL 105 to the SVOL 108. It is recognized that there is a path (FIG. 1 (4)).

Therefore, at the time of remote copy, a write request related to remote copy is transmitted using any of these paths. The path to be used is selected according to, for example, a round robin method.

For example, when a write request is transmitted through the path shown in FIG. 1 (3), the MS-SDKC 100B writes the write target data (may be described as “WR data” in the figure) to the SVOL 108 according to the write request. .

On the other hand, when a write request is transmitted through the path shown in FIG. 1 (4), the MD-PDKC 100C transmits the write request transmitted from the MS-PDKC 100A to the MD-SDKC 100D (FIG. 1 (5)). Upon receiving this write request, the MD-SDKC 100D recognizes that it is a write request to the SVVOL 109, and transmits the write request to the MS-SDKC 100B (FIG. 1 (6)).

When the MS-SDKC 100B receives this write request, it writes the write target data to the SVOL 108 in accordance with the write request. Accordingly, remote copy can be executed from the MS-PDKC 100A to the SVOL 108 of the MS-SDKC 100B via the MD-PDKC 100C.

In this embodiment, when all the data of the PVOL 105 of the MS-PDKC 100A is transferred to the MD-PVOL 107 of the MD-PDKC 100C, the MD-PDKC 100C is set to handle the access to the PVVOL 106 as the access to the MD-PVOL 107. To do. Thereafter, when the host 200 accesses the PVVOL 106, the MD-PDKC 100C executes the process as an access to the MD-PVOL 107. Specifically, when a write request is received from the host 200 for the PVVOL 106 before setting, the MD-PDKC 100C executes remote copy to the MD-SDKC 100D without transferring the write request to the MS-PDKC 100A. After that, the write target data is stored in the MD-PVOL 107.

As described above, according to the present embodiment, it is possible to transfer the data of the PVOL 105 to the MD-PVOL 107 of the MD-PDKC 100C while continuing the remote copy, and thereafter appropriately execute the remote copy of the MD-PVOL 107. Can do.

Hereinafter, the computer system according to the present embodiment will be described in detail with reference to the drawings.

FIG. 2 is a diagram illustrating a configuration example of the computer system according to the first embodiment.

The computer system has one or more hosts 200 and a plurality of DKCs 100.

The host 200 is a general computer and includes a processor, a cache, a storage device, and the like. The host 200 issues an access request for the VOL managed by the DKC 100 to the DKC 100. The host 200 has one or more ports 201. The port 201 is an example of a communication interface device, and is an interface for connecting to a network. In the port 201, identification information (for example, WWN (World Wide Name) etc.) that can uniquely identify the port is set. The host 200 associates the identification information (for example, device number) of the VOL to be accessed with the identification information (port information) of the port 101 of the DKC 100 that is accessed when accessing the VOL. Is managing.

As the DKC 100, in this embodiment, the MS-PDKC 100A that manages the PVOL 105 that is the MS (migration source), the MS-SDKC 100B that manages the SVOL 108 that is the copy destination VOL of the PVOL 105, and the MD (migration destination) of the PVOL 105 An MD-PDKC 100C that manages the MD-PVOL 107 and an MD-SDKC 100D that manages the MD-SVOL 110 that is the MD of the SVOL 108 are included.

MS-PDKC 100A and MS-SDKC 100B are connected via a communication network 60. The MD-PDKC 100C and the MD-SDKC 100D are connected via a communication network 60. In this embodiment, the PVOL 105 and SVOL 108 of the MS-DKC group 30 including the MS-PDKC 100A and the MS-SDKC 100B are transferred to the MD-DKC group 40 including the MD-PDKC 100C and the MD-SDKC 100D.

Also, the MS-PDKC 100A and the MD-PDKC 100C are installed, for example, in the same local site 10 and are connected so as to communicate with each other. Further, the MS-SDKC 100B and the MD-SDKC 100D are installed at a remote site 20 which is a place distant from the local site 10 and are communicably connected to each other.

The DKC 100 includes, for example, a plurality of ports 101, one or more MPPKs (microprocessor packages) 102, one or more CMPKs (cache memory packages) 103, an internal network 104, and a PDEV that constitutes one or more VOLs. Prepare. A VOL is composed of one or more PDEVs. The port 101, the MPPK102, the CMPK103, and the PDEV configuring the VOL are connected via the internal network 104 so as to be able to communicate with each other.

For example, as a VOL managed by the DKC 100, the MS-PDKC 100A manages the PVOL 105. The MS-SDKC 100B manages the SVOL 108. The MD-PDKC 100C manages the PVVOL 106 and the MD-PVOL 107. The MD-SDKC 100D manages the SVVOL 109 and the MD-SVOL 110.

The port 101 is an example of a communication interface device, and is an interface for connecting to a network. Identification information (for example, WWN (World Wide Name), etc.) that can uniquely identify the port is set in the port 101. Virtually arbitrary identification information can be assigned to the port 101.

In this embodiment, for example, the MS-PDKC 100A has a port 101 for connecting to the host 200, a port 101 for connecting to the MD-PDKC 100C, and a port 101 for connecting to the MS-SDKC 100B.

The MD-PDKC 100C has a port 101 for connecting to the host 200, a port 101 for connecting to the MS-PDKC 100A, and a port 101 for connecting to the MD-SDKC 100D.

The MS-SDKC 100B has a port 101 for connecting to the host 200, a port 101 for connecting to the MS-PDKC 100A, and a port 101 for connecting to the MD-SDKC 100D.

The MD-SDKC 100D has a port 101 for connecting to the host 200, a port 101 for connecting to the MD-PDKC 100C, and a port 101 for connecting to the MS-SDKC 100B.

The MPPK102 is a device having one or a plurality of processors. The processor executes various processes according to the program. The MPPK 102 may have a memory that stores programs to be executed by the processor and data to be used. In this embodiment, the MPPK 102 of the MD-PDKC 100C and MD-SDKC 100D has a program for realizing a new function in this embodiment described later. The MPPK 102 of the MS-PDKC 100A and the MS-SDKC 100B only needs to have an existing program.

The CMPK 103 includes one or more cache memories (CMs). One or more CMs are volatile memory and / or nonvolatile memory. One or more CMs have a storage area (hereinafter referred to as a shared area) for storing information used by the processor of the MPPK 102 in addition to a storage area (hereinafter referred to as a cache area) for temporarily storing data input to and output from the PDEV. You may have.

FIG. 3 is a diagram showing an example of the state of the computer system before the start of migration. In the figure, the copy source is described as “CS” and the copy destination is described as “CD”.

The computer system before the start of migration has a host 200, an MS-PDKC 100A, and an MS-SDKC 100B.

The MS-PDKC 100A has ports 101 with preset port information “1A”, “1B”, and “1C” (hereinafter referred to as ports 1A, 1B, and 1C, respectively. Shown by method.) The MS-SDKC 100B has ports 2A, 2B, and 2C. The MS-PDKC 100A is connected to the host 200 via the port 1B. Further, the port 1A of the MS-PDKC 100A and the port 2A of the MS-SDKC 100B are connected.

The MS-PDKC 100A manages the PVOL 105 whose device information is “0101”. The MS-SDKC 100B manages the SVOL 108 whose device information that is the copy destination of the PVOL 105 is “0201”. In the computer system, remote copy for copying the PVOL 105 to the SVOL 108 is performed. The remote copy is a synchronous remote copy.

The MPPKs 102 of the MS-PDKC 100A and the MS-SDKC 100B each store a remote copy table (RC table) 120.

The RC table 120 includes an entry having fields of a number (#) 121, copy source port information 122, copy source device information 123, copy destination port information 124, and copy destination device information 125 for each remote copy. Store.

The number 121 stores a remote copy number for identifying a remote copy. The copy source port information 122 stores identification information (port information) for identifying the port of the copy source DKC 100. The copy source device information 123 stores identification information (device information) for identifying a device that is a copy source (here, PVOL 105). The copy destination port information 124 stores identification information (port information) for identifying the copy destination DKC 100 port. The copy destination device information 125 stores identification information (device information) for identifying a device that is a copy destination (here, the SVOL 108).

3 stores a remote copy in which the device 0101 of the port 1A is the copy source and the device 0201 of the port 2A is the copy destination.

Accordingly, in the computer system before the start of migration, when the MS-PDKC 100A receives a write request for the PVOL 105 from the host 200, a write request for writing the write target data corresponding to the write request is sent from the port 1A to the MS-SDKC 100B. By transmitting to the port 2A, the write target data is copied to the SVOL of the MS-SDKC 100B.

FIG. 4 is a diagram for explaining the introduction of MD-DKC into the computer system.

For example, the MD-PDKC 100C (DKC 100C) to which the administrator migrates the PVOL 105 is placed in the local site 10 to which the MS-PDKC 100A belongs and the SVOL 108 is migrated to the computer system shown in FIG. ) Is arranged at, for example, the remote site 20 to which the MS-SDKC 100B belongs. The MD-PDKC 100C has a port 3A, a port 3B, and a port 3C. The MD-SDKC 100D has a port 4A and a port 4B.

Then, the administrator connects the port 3B of the MD-PDKC 100C and the host 200 port 201 via the network 50. Further, the port 3C of the MD-PDKC 100C and the port 1C of the MS-PDKC 100A are connected via a cable or a network. Further, the port 3A of the MD-PDKC 100C and the port 4A of the MD-SDKC 100D are connected via the network 60. Further, the port 4C of the MD-SDKC 100D is connected to the port 2C of the MS-SDKC 100B via a cable or a network.

FIG. 5 is a diagram for explaining port virtualization in MD-DKC.

For each port 101 of the MD-PDKC 100C and the MD-SDKC 100D, identification information (port information such as WWN) that can uniquely identify the port 101 is set in advance. Here, the preset port information is referred to as actual port information. In this embodiment, in MD-PDKC 100C and MD-SDKC 100D, virtual port information (virtual port information) (VP information) different from actual port information (RP information) is assigned to each port 101. It is possible to communicate using the virtual port information. In order to manage such virtual port information, the MPPK 102 of the MD-PDKC 100C and MD-SDKC 100D stores a port virtual information management table (PV table) 130 for managing the virtual port information of the port 101. ing.

The port virtual information management table 130 stores an entry having fields of item number 131, actual port information 132, and virtual port information 133 for each port.

Entry number 131 stores the entry number. The actual port information 132 stores actual port information of the port corresponding to the entry. The virtual port information 133 stores virtual port information assigned to the port corresponding to the entry.

In this embodiment, after the computer system is brought into the state shown in FIG. 4, the actual port information of the port 2C of the MS-SDKC 100B is allocated as the virtual port information of the port 3C of the MD-PDKC 100C, and the virtual port of the port 3A is allocated. As information, actual port information of the port 1C of the MS-PDKC 100A is assigned. Further, the actual port information “1C” of the port 1C of the MS-PDKC 100A is assigned as the virtual port information of the port 4C of the MD-SDKC 100D, and the actual port information of the port 2C of the MS-SDKC 100B is allocated as the virtual port information of the port 4A. Port information “2C” is assigned. Specifically, the actual port information “2C” of the port 2C of the MS-SDKC 100B is stored in the virtual port information 133 of the entry corresponding to the port 3C of the MD-PDKC 100C in the port virtual information management table 130 of the MD-PDKC 100C. Then, the actual port information “1C” of the port 1C of the MS-PDKC 100A is stored in the virtual port information 133 of the entry corresponding to the port 3A. Further, the actual port information “2C” of the MS-SDKC 100B is stored in the virtual port information 133 of the entry corresponding to the port 4A of the MD-SDKC 100D in the port virtual information management table 130 of the MD-SDKC 100D, and corresponds to the port 4C. The actual port information “1C” of the MS-PDKC 100A is stored in the virtual port information 133 of the entry to be entered.

In order to set the port virtual information management table 130 of the MD-PDKC 100C and MD-SDKC 100D in this way, the administrator can connect via the SVP (Service Processor) connected to the MD-PDKC 100C or the MD-SDKC 100D, or via a network. It may be set using a terminal such as a management device, or may be set by the MPPK 102 processor of the MD-PDKC 100C and MD-SDKC 100D collecting and determining various information. Note that an administrator may use a terminal such as SVP to operate, set, etc., with “administrator” as the subject.

When the MD-PDKC 100C uses the virtual port information assigned as described above for communication with the MS-PDKC 100A, the MS-PDKC 100A recognizes the MD-PDKC 100C as the MS-SDKC 100B. Further, when the MD-SDKC 100D uses the virtual port information assigned as described above for communication with the MS-SDKC 100B, the MS-SDKC 100B recognizes the MD-SDKC 100D as the MS-PDKC 100A. Therefore, in MS-PDKC 100A and MS-SDKC 100B, it is not necessary to grasp MD-PDKC 100C and MD-SDKC 100D, and it is not necessary to add a new function to MS-PDKC 100A and MS-SDKC 100B.

FIG. 6 is a diagram for explaining generation of VVOL in MD-DKC.

After the computer system is brought into the state shown in FIG. 5, the administrator generates a VVOL 106 based on the MS-PVOL 105 in the MD-PDKC 100C, and generates a VVOL 109 based on the MS-SVOL 108 in the MD-SDKC 100D. Note that the function of generating the VVOL based on the RVOL of the MS-PDKC 100A (MS-SDKC 100B) in the MD-PDKC 100C (MD-SDKC 100D) can be realized by using a storage virtualization technology. Here, unique device information (actual device information) is set in each of the VVOL 106 and the VVOL 109. In this embodiment, “0301” is set as device information in the VVOL 106, and “0401” is set as device information in the VVOL 109.

FIG. 7 is a diagram illustrating VVOL virtualization in MD-DKC.

The VVOL 106 and the VVOL 109 can set virtual device information as their device information, and can be used as devices corresponding to the virtual device information. The MD-PDKC 100C and the MD-SDKC 100D store a device virtual information management table (DV table) 140 in the memory of each MPPK 102 in order to manage virtual device information about the VVOL.

The device virtual information management table 140 stores an entry including a number (#) 141, actual device information (RD information) 142, virtual device information (VD information) 143, and a type 144 field for each device.

The number 141 stores a number corresponding to the entry. The actual device information 142 stores actual device information of the device corresponding to the entry. The virtual device information 143 stores virtual device information set for the device corresponding to the entry. The type 144 stores the type of device corresponding to the entry. In this embodiment, when the device corresponding to the entry is a VVOL, “virtual” is stored in the type 144.

In the present embodiment, after the computer system is set to the state shown in FIG. 6, the actual device information “0101” of the PVOL 105 of the MS-PDKC 100A is set as the virtual device information of the VVOL 106 generated in the MD-PDKC 100C. The actual device information “0201” of the SVOL 108 of the MS-SDKC 100B is set as the virtual device information of the VVOL 109 generated in the SDKC 100D. Specifically, in accordance with an instruction from the management apparatus, the MD-PDKC 100C adds an entry to the device virtual information management table 140, stores the actual device information “0301” of the virtual device 106 in the actual device information 142, and the PVOL 105 The actual device information “0101” is stored in the virtual device information 143, and “virtual” is stored in the type 144. Further, in response to an instruction from the management apparatus, the MD-SDKC 100D adds an entry to the device virtual information management table 140, stores the actual device information “0401” of the virtual device 109 in the actual device information 142, and the actual device information of the SVOL 108. The device information “0201” is stored in the virtual device information 143, and “virtual” is stored in the type 144.

Thus, since the device information of the VVOL 106 can be handled as the same as the actual device information “0101” of the PVOL 105, an access request to the PVOL 105 can be handled as an access request to the VVOL 106. Further, since the device information of the VVOL 109 can be handled as the same as the actual device information “0201” of the SVOL 108, an access request for the SVOL 108 can be handled as an access request for the VVOL 109.

FIG. 8 is a diagram illustrating the addition of a path to the MD-DKC in the host.

After the computer system is brought into the state shown in FIG. 7, the administrator adds a path (access path) for accessing the device (VOL) of the device information “0101” in the host 200. Specifically, a path from the port 201 of the host 200 to the port 3B of the MD-PDKC 100C is added as a path for accessing the device having the device information “0101”. As a result, the host 200 newly adds a path (first path) via the port 1B of the MS-PDKC 100A as an access path for accessing the device (originally PVOL 105) of the device information “0101”. It is recognized that there is a path (second path) via the port 3B of the added MD-PDKC 100C. Here, in accessing the device having the device information “0101”, the host 200 selects and uses either the first path or the second path according to, for example, the round robin method. It will be.

In the computer system in the state shown in FIG. 8, when the host 200 selects and transmits the write request for the device information “0101”, the MS-PDKC 100A accepts the write request via the port 1B. . The MS-PDKC 100A saves the write target data corresponding to the write request in the CM, and sends a write request for remote copy of the write target data to the device information “0201” of the port 2A according to the contents of the RC table 120. Send to SDKC100B. As a result, the MS-SDKC 100B receives the write request via the port 2A, writes the write target data corresponding to the write request to the device having the device information “0201”, that is, the SVOL 108, and returns a response to the MS-PDKC 100A. When receiving a normal response from the MS-SDKC 100B, the MS-PDKC 100A stores the write target data saved in the CM in the device having the device information “0101”, that is, the PVOL 105, and returns the response to the host 200.

On the other hand, when the host 200 selects and transmits the write request for the device having the device information “0101” by selecting the second path, the MD-PDKC 100C accepts the write request via the port 3B. The MD-PDKC 100C transfers a write request for the device information “0101” to the MS-PDKC 100A by a switching operation. The MS-PDKC 100A saves the write target data corresponding to the write request in the CM, and sends a write request for remote copy of the write target data to the device information “0201” of the port 2A according to the contents of the RC table 120. Send to SDKC100B. As a result, the MS-SDKC 100B receives the write request via the port 2A, writes the write target data corresponding to the write request to the device having the device information “0201”, that is, the SVOL 108, and returns a response to the MS-PDKC 100A. Upon receiving a normal response from the MS-SDKC 100B, the MS-PDKC 100A stores the write target data saved in the CM in the device having the device information “0101”, that is, the PVOL, and returns the response to the MD-PDKC 100C. The MD-PDKC 100C returns a response to the write request to the host 200.

In this way, the write target data corresponding to the write request can be appropriately stored in the PVOL 105 and the SVOL 108 regardless of which path the write request from the host 200 takes.

FIG. 9 is a diagram for explaining the deletion of the path to the MS-DKC in the host.

After the computer system is brought into the state shown in FIG. 8, the administrator deletes the path through the port 1B of the MS-PDKC 100A that accesses the device of the device information “0101” in the host 200. As a result, access requests to the PVOL 105 can be aggregated and transmitted on the path via the port 3B of the MD-PDKC 100C. That is, the MD-PDKC 100C can acquire all access requests to the PVOL 105.

FIG. 10 is a diagram for explaining reading of pair information from the MS-DKC.

After the computer system is brought into the state shown in FIG. 9, the MD-PDKC 100C and the MD-SDKC 100D extract the entry information (copy pair information) of the RC table 120 from the MS-PDKC 100A and the MS-SDKC 100B, respectively. Register in the RC table 120. Here, the MD-PDKC 100C and the MD-SDKC 100D store the device information “0101” of the PVOL 105 in the copy source device information 123 and the device information “0201 of the SVOL 106 in the copy destination device information 125 in the respective RC tables 120. Is stored.

FIG. 11 is a diagram for explaining the addition of the RC path.

After the computer system is brought into the state shown in FIG. 10, the administrator adds a new remote copy path (RC path) via the MD-PDKC 100C and the MD-SDKC 100D.

Here, in order to manage a new RC path, the MD-PDKC 100C stores the primary remote copy management table (PRC management table) 150 in the MPPK 102, and the MD-SDKC 100D stores the secondary remote copy management table (SRC management table). ) 160 is stored in the MPPK102.

The PRC management table 150 includes fields of number (#) 151, migration source primary port (MS-PP) information 152, reception source virtual port (RS-VP) information 153, and PVOL device information 154 for each remote copy. Store the entry. The number 151 stores the number for the remote copy. The MS-PP information 152 stores port information of the port of the MS-PDKC 100A to which a remote copy write request is transmitted. The transfer data receiving source virtualization port information 153 stores the virtualization port information of the port of the MD-PDKC 100C from which a remote copy write request is received. The PVOL device information 154 stores PVOL device information related to remote copy. If there is no RC path via MD-PDKC 100C and MD-SDKC 100D for remote copy, MS-PP information 152, transfer data receiver virtualization port information 153, and PVOL device information 154 include: For example, “-” is stored.

The SRC management table 160 includes fields of number (#) 161, transmission source virtual port (TS-VP) information 162, migration source secondary port (MS-SP) information 163, and SVOL device information 164 for each remote copy. Store the entry. The number 161 stores the number for the remote copy. The source virtualization port information 162 stores the virtualization port information of the MD-SDKC 100D port that transmits a remote copy write request. The MS-SP information 163 stores port information of the port of the MS-SDKC 100B to which a remote copy write request is transmitted. The SVOL device information 164 stores SVOL device information related to remote copy.

Next, a process for adding a new RC path via the MD-PDKC 100C and the MD-SDKC 100D will be described in detail.

The MD-SDKC 100D transmits a path addition request for adding a remote copy path from the MD-SDKC 100D to the MS-SDKC 100B to the MS-SDKC 100B (FIG. 11 (1)). The path addition request includes contents in which the virtual port information “1C” of the port 4C of the MD-SDKC 100D is the copy source and the port 2C of the MS-SDKC 100B is the copy destination. When the MS-SDKC 100B receives the path addition request, the virtual port of the port 4C of the MD-SDKC 100D is added to the copy source port information 122 of the vacant entry (here, the second entry) in the RC table 120 of the MS-SDKC 100B. Information (“1C”), the device information “0101” of the PVOL 105 is stored in the copy source device information 123, the port information “2C” of the port 2C of the MS-SDKC 100B is stored in the copy destination port information 124, and the copy SVOL device information “0201” is stored in the destination device information 125. Further, the MD-SDKC 100D sets the virtual port information “1C” of the port 4C of the MD-SDKC 100D to the transfer data transmission source virtualization port information 162 for the entry corresponding to the added RC copy number in the SRC management table 160. The port information “2C” of the port 2C of the MS-SDKC 100B is stored in the MS-SP information 163, and the device information “0201” of the SVOL 108 is stored in the SVOL device information 164.

Next, the MD-PDKC 100C forms a remote path between the MD-PDKC 100C and the MD-SDKC 100 (FIG. 11 (2)). Specifically, the MD-PDKC 100C stores the port information “1C” of the port 1C of the MS-PDKC 100A in the MS-PP information 152 of the entry corresponding to the RC copy number formed in the PRC management table 150, and the transfer data The virtual port information “2C” of the port 3C of the MD-PDKC 100C is stored in the reception source virtual port information 153, and the device information “0101” of the PVOL 105 is stored in the PVOL device information 154. Upon receiving an instruction from the MD-PDKC 100C, the MD-SDKC 100D virtualizes the port 3A of the MD-PDKC 100C in the copy source port information 122 for the entry corresponding to the copy path formed in the RC table 120 of the MD-SDKC 100D. The port information “1C” is stored, and the virtual port information “2C” of the port 4A of the MD-SDKC 100D is stored in the copy destination port information 124.

Next, according to an instruction from the administrator, the MS-PDKC 100A adds the port information “1C” of the port 1C of the MS-PDKC 100A to the copy source port information 122 of the entry corresponding to the copy path to be added in the RC table 120 of the MS-PDKC 100A. And the virtual port information “2C” of the port 3C of the MD-PDKC 100C is stored in the copy destination port information 124. Further, the MS-PDKC 100A transmits a path addition request for adding a remote copy path from the MS-PDKC 100A to the MD-PDKC 100C to the MD-PDKC 100C (FIG. 11 (3)). Upon receiving the path addition request, the MD-PDKC 100C stores the virtual port information “1A” of the port 3A of the MD-PDKC 100C in the copy source port information 122 of the entry corresponding to the copy path to be added in the RC table 120. In the copy destination port information 124, the virtual port information “2C” of the port 4A of the MD-SDKC 100D is stored.

As a result of this processing, as a path for remote copying from the PVOL 105 to the SVOL 108, the path from the MS-PDKC 100A to the MS-SDKC 100B without passing through the MD-DKC, and the MS from the MS-PDKC 100A to the MD-PDKC 100C and the MD-SDKC 100D via MS -The path to the SDKC 100B is managed. In the above description, the processing has been described in the order of (1), (2), and (3) in FIG. 11, but is not limited to this, and (1), (2), and (3) are not limited thereto. The processes may be performed in parallel. In short, it is only necessary that the process (3) is not performed before the processes (1) and (2). This is because the remote copy write request is issued from the MS-PDKC 100A to the MD-PDKC 100C when the process (3) is executed first and the remote path to the MS-SDKC 100B is not completed. Is to prevent.

FIG. 12 is a diagram for explaining the RC path as seen from the MS-DKC.

In the computer system, two RC paths are managed as shown in FIG. 11. In the MS-PDKC 100A and the MS-SDKC 100B, each path is port information of the MS-PDKC 100A and the MS-SDKC 100B. Therefore, as shown in FIG. 12, it is recognized that there are simply two remote paths between each other. Therefore, in MS-PDKC100A and MS-SDKC100B, the presence of MD-PDKC100C and MD-SDKC100D cannot be recognized. In other words, it is not necessary to recognize the presence of MD-PDKC100C and MD-SDKC100D.

FIG. 13 is a diagram for explaining the deletion of the RC path between MS and DKC.

After the computer system is brought into the state shown in FIG. 11, the administrator deletes the access path from the MS-PDKC 100A to the MS-SDKC 100B without going through the MD-PDKC 100C and the MD-SDKC 100D. That is, the administrator copies the copy source port information 122 of the entry corresponding to the access path from the MS-PDKC 100A to the MS-SDKC 100B in the RC table 120 of the MS-PDKC 100A and the MS-SDKC 100B without going through the MD-PDKC 100C and the MD-SDKC 100D. "-" Is set in the copy source device information 123, the copy destination port information 124, and the copy destination device information 125. Accordingly, remote copy write requests from the PVOL 105 to the SVOL 108 can be aggregated and transmitted from the MS-PDKC 100A to the MS-SDKC 100B via the MD-PDKC 100C and MD-SDKC 100D. That is, the MD-PDKC 100C can obtain all remote copy write requests from the PVOL 105 to the SVOL 108.

FIG. 14 is a first diagram illustrating MS-VOL data migration according to the first embodiment.

For example, after the computer system is changed to the state shown in FIG. 13, as shown in FIG. 14, the PVOL 105 of the MS-PDKC 100A is transferred to the MD-PVOL 107 of the MD-PDKC 100C, and the SVOL 108 of the MS-SDKC 100B is changed to the MD-SDKC 100D. Data migration processing for migrating to the MD-SVOL 110 is started.

In the data migration processing, the MD-PDKC 100C sequentially migrates from the data in the top area of the VVOL 106 based on the MS-PVOL 105 to the PVOL 107, and the MD-SDKC 100D sequentially performs the SVOL 110 from the data in the top area of the VVOL 109 based on the MS-SVOL 108. Make the transition to

In the middle of this data migration processing, if there is a write request from the host 200 to the MD-PDKC 100C, the MD-PDKC 100C transfers the write request to the MS-PDKC 100A. The MS-PDKC 100A transmits a remote copy write request through the RC path shown in FIG. As a result, the write to the PVOL 105 can be appropriately reflected in the SVOL 108. Here, when the remote copy write request received from the MS-PDKC 100A is a write for an area that has already been copied from the PVOL 105 to the MD-PVOL 107, the MD-PDKC 100C reads the MD-PDKC 100C in that area of the MD-PVOL 107. On the other hand, write target data. In addition, when the remote copy write request received from the MD-PDKC 100C is a write for an area that has already been copied from the SVOL 105 to the MD-SVOL 110, the MD-SDKC 100D applies the write request to the MD-SVOL 110 to that area. Write the data to be written. As a result, it is possible to appropriately reflect the write for the copied data in the MD-VOL.

FIG. 15 is a second diagram illustrating data migration of the MS-VOL according to the first embodiment.

FIG. 15 shows data migration processing when data migration from the SVOL 108 to the MD-SVOL 110 is completed in the data migration processing shown in FIG.

When the data migration from the SVOL 108 to the MD-SVOL 110 is completed, the MD-SDKC 100D replaces the device information of the MD-SVOL 110 and the VVOL 109 to make the SVOL using the MD-SVOL 110.

Thereafter, upon receiving a remote copy write request from the MD-PDKC 100C, the MD-SDKC 100D writes the write target data corresponding to the write request to the MD-SVOL 110, and also remotely transmits to the MS-SVOL 108. Send a copy write request. When receiving the write request, the MS-SDKC 100B writes the write target data to the SVOL. Here, the write target data is also written to the SVOL 108. If the migration of the PVOL 105 is not completed, the state of the SVOL 108 is made the same as that of the PVOL 105, and the PVOL 105 is appropriately restored based on the SVOL 108 when a failure occurs. This is to make it possible.

FIG. 16 is a third diagram for explaining data migration of the MS-VOL according to the first embodiment.

FIG. 16 shows the data migration process immediately after the data migration from the PVOL 105 to the MD-PVOL 107 is completed in the data migration process shown in FIG.

When the data migration from the PVOL 105 to the MD-PVOL 107 is completed, the MD-PDKC 100C replaces the device information of the MD-PVOL 107 and the VVOL 106 to make the PVOL that uses the MD-PVOL 107. As a result, the host 200 can appropriately access the MD-PVOL 107 without any particular attention thereafter.

Thereafter, when there is a write request from the host 200 to the MD-PDKC 100C, the MD-PDKC 100C saves the write target data in the CM without transferring the write request to the MS-PDKC 100A. Next, the MD-PDKC 100C transmits a remote copy write request to the MD-SDKC 100D.

When the MD-SDKC 100D receives a remote copy write request from the MD-PDKC 100C, it writes the write target data corresponding to the write request to the MD-SVOL 110, and also sends a remote copy write request to the MS-SVOL 108. Send. When receiving the write request, the MS-SDKC 100B writes the write target data to the SVOL. Here, the write target data is also written to the SVOL 108. If the migration of the PVOL 105 is not completed, the state of the SVOL 108 is made the same as that of the PVOL 105, and the PVOL 105 is appropriately restored based on the SVOL 108 when a failure occurs. This is to make it possible.

When the MD-SDKC 100D receives a write request response from the MS-SDKC 100B, it returns a write request response to the MD-PDKC 100C. When the MD-PDKC 100C receives a response indicating normal termination as a response to the write request from the MD-SDKC 100D, the MD-PDKC 100C writes the write target data saved in the CM to the corresponding area of the MD-PVOL 107 and sends it to the host 200. And return a write request response.

FIG. 17 is a diagram for explaining the stop of remote copy between MS and DKC.

After the state shown in FIG. 16 is reached, the MD-PDKC 100C and the MD-SDKC 100D stop the remote copy between the MS and DKC. Specifically, the MD-PDKC 100C sets “−” in the MS-PP information 152, the transfer data receiving source virtualization port information 153, and the PVOL device information 154 of the entry regarding the migrated PVOL 105 in the PRC management table 150. This deletes the setting for transferring the remote copy. Further, the MD-SDKC 100D sets “−” in the transfer data transmission source virtualization port information 162, the MS-SP information 163, and the SVOL device information 164 of the entry related to the migrated SVOL 108 in the SRC management table 160, Delete the settings for transferring a remote copy.

Thus, when the MD-SDKC 100D receives a remote copy write request from the MD-PDKC 100C, the MD-SDKC 100D does not transmit a remote copy write request to the MS-SVOL 108.

FIG. 18 is a diagram showing an example of the state of the computer system after completion of the migration.

After the computer system reaches the state shown in FIG. 17, the administrator can remove the MS-PDKC 100A and the MS-SDKC 100B from the computer system as shown in FIG.

Hereinafter, processing performed in this embodiment will be described.

In this embodiment, the MD-PDKC 100C and the MD-SDKC 100D execute processes according to the common flowcharts shown in FIGS. 19 to 21, so that the MD-PDKC 100C is necessary for the MD-PDKC in FIGS. The processing is executed, and the MD-SDKC 100D executes processing necessary for the MD-SDKC in FIGS. In this case, the same program may be executed by the MD-PDKC 100C and the MD-SDKC 100D. Here, in FIG. 19 to FIG. 21, the MD-PDKC 100C and the MD-SDKC 100D may execute the processing, and therefore, the main body that executes the processing will be described as the DKC 100.

FIG. 19 is a first flowchart of the write process according to the first embodiment. FIG. 20 is a second flowchart of the write process according to the first embodiment. FIG. 21 is a third flowchart of the write process according to the first embodiment. Here, the flowchart of FIGS. 19 and 20 is connected to the portion of the terminal 1 indicated by the pentagon in the drawing, and the flowchart of FIGS. 20 and 21 is connected to the portion of the terminal 2 indicated by the pentagon in the drawing. Has been.

When the DKC 100 (specifically, the processor of the MPPK 102 of the DKC 100) receives a write request including the write target data (write target data) (step S11), the write request corresponds to the host I / O from the host 200. It is determined whether the request is a write request to be made or a write request corresponding to remote copy (step S12).

As a result, when the write request is a write request corresponding to remote copy (step S12: remote copy), the DKC 100 advances the process to step S26 in FIG.

On the other hand, if the write request is a write request corresponding to the host I / O (step S12: host I / O), this means that the DKC 100 is the MD-PDKC 100C. It is determined whether or not the target of RVOL is RVOL (step S13).

As a result, if the target of the write request is RVOL (step S13: Y), for example, it indicates that the write request to the MD-PVOL 107 has been completed after the migration of the PVOL 105, so the DKC 100 performs the processing. Proceed to step S14. On the other hand, when the target of the write request is not RVOL (step S13: N), the DKC 100 indicates that the migration of the PVOL 105 has not been completed, so the process proceeds to step S21.

In step S14, the DKC 100 saves and stores the write target data in the CM of the CMPK 103. Next, the DKC 100 transmits write target data to the SVOL (step S15). In this embodiment, the SVOL is the MD-SVOL 110 that has been migrated from the MS-SDKC 100B to the MD-SDKC 100D.

Next, the DKC 100 waits for a response from the SVOL, that is, a response from the MD-SDKC 100D (step S16), and determines whether or not there is a Good response from the SVOL indicating that storage of the write target data has been completed (step S17). ). As a result, when there is a Good response (step S17: Y), the DKC 100 reflects the write target data saved in the CM of the CMPK 103 in the MD-PVOL 107 (step S18), and the host that made the write request In step 200, the normal end is returned as a response to the write request, and the write process ends. On the other hand, when there is no Good response (step S17: N), the DKC 100 returns an abnormal end as a response to the write request to the host 200 that has made the write request (step S20), and ends the write process.

In step S21, the DKC 100 transfers a write request including the write target data to the MS-PDKC 100A. Next, the DKC 100 waits for a response from the MS-PDKC 100A (step S22), and determines whether there is a Good response from the MS-PDKC 100A indicating that the storage of the write target data has been completed (step S23). As a result, if there is a Good response (step S23: Y), normal termination is returned as a response to the write request to the host 200 that has made the write request, and the write processing is terminated. As a result, even when the host 200 makes a write request to the PVOL 105 to the MD-PDKC 100C, the MS 200 properly responds to the host 200 that the MS-PDKC 100A has successfully stored the write target data in the PVOL 105. can do. On the other hand, when there is no Good response (step S23: N), the DKC 100 returns an abnormal end as a response to the write request to the host 200 that has made the write request (step S25), and ends the write process.

In step S26 of FIG. 20, the DKC 100 determines whether or not the write request target is RVOL.

As a result, when the target of the write request is RVOL (step S26: Y), this indicates that the DKC 100 is the MD-SDKC 100D and the data transfer from the SVOL 108 to the MD-SVOL 110 has been completed. The DKC 100 advances the process to step S27. Thus, when there is an RVOL, it can be determined that the MD-SDKC 100D is used because the write request corresponding to the remote copy is not sent to the MD-PDKC 100C having the RVOL. On the other hand, when the target of the write request is not RVOL (step S26: N), the DKC 100 advances the process to step S35 in FIG.

In step S27, the DKC 100 saves and stores the write target data in the CM of the CMPK 103. Next, the DKC 100 determines whether or not there is an entry indicating a link with the MS-SDKC 100B in the SRC management table 160 (step S28). Here, in this embodiment, an entry indicating a link with the MS-SDKC 100B is secured in the SRC management table 160 until the migration of the PVOL 105 is completed. This is to maintain the SVOL 108 of the MS-SDKC 100B in a usable state for restoration until the migration of the PVOL 105 is completed.

As a result, when there is no entry indicating the link with the MS-SDKC 100B in the SRC management table 160 (step S28: N), the DKC 100 advances the process to step S32.

On the other hand, if there is an entry indicating the link with the MS-SDKC 100B in the SRC management table 160 (step S28: Y), it is necessary to reflect the write target data in the SDKC 108, so the DKC 100 The write target data is transferred (step S29), a response from the SVOL 108, that is, a response from the MS-SDKC 100B is waited (step S30), and whether or not there is a Good response from the SVOL 108 indicating that the storage of the write target data has been completed. Is determined (step S31). As a result, when there is a Good response (step S31: Y), it means that the write target data is reflected in the SVOL 108, and thus the DKC 100 advances the process to step S32 while there is no Good response. In (Step S31: N), the process proceeds to Step S34.

In step S32, the DKC 100 reflects the write target data saved in the CM of the CMPK 103 on the MD-SVOL 110. Next, the DKC 100 returns a normal end as a response to the write request to the DKC 100 that has made the write request (step S33), and ends the write process.

In step S34, the DKC 100 returns an abnormal end as a response to the write request to the DKC 100 that has made the write request, and ends the write process.

21. In step S35 of FIG. 21, the DKC 100 determines whether the write target data corresponding to the write request is data for the PVOL 105.

As a result, when the write target data of the write request is data for the PVOL 105 (step S35: Y), it indicates that it is necessary to write to the SVOL. Therefore, the DKC 100 advances the process to step S36. If the write target data of the write request is not data for the PVOL 105 (step S35: N), the process proceeds to step S41.

In step S36, the DKC 100 transmits the write target data to the SVOL. In this embodiment, the SVOL is the MD-SVOL 110 or SVVOL 109 that has been migrated from the MS-SDKC 100B to the MD-SDKC 100D.

Next, the DKC 100 waits for a response from the SVOL, that is, a response from the MD-SDKC 100D (step S37), and determines whether or not there is a Good response from the SVOL indicating that storage of the write target data has been completed (step S38). ). As a result, if there is a Good response (step S38: Y), the DKC 100 returns a normal end as a response to the write request to the MS-PDKC 100A that has made the write request (step S39), and the process proceeds to step S46. Proceed to

On the other hand, if there is no Good response (step S38: N), the DKC 100 returns an abnormal end as a response to the write request to the MS-PDKC 100A00A that has made the write request (step S40), and the write processing ends. .

On the other hand, in S41, the DKC 100 transfers a write request including write target data to the MS-SDKC 100B. Next, the DKC 100 waits for a response from the MS-SDKC 100B (step S42), and determines whether there is a Good response from the MS-SDKC 100B indicating that the storage of the write target data has been completed (step S43). As a result, if there is a Good response (step S43: Y), the DKC 100 returns a normal end as a response to the write request to the MD-PDKC 100C that has made the write request (step S41), and the process proceeds to step S46. Proceed to On the other hand, if there is no Good response (step S43: N), the DKC 100 returns an abnormal end as a response to the write request to the MD-PDKC 100C that has made the write request (step S45), and ends the write process. .

In step S46, the DKC 100 is in the process of migrating the data of the PVOL 105 (or SVOL 108) to the MD-PVOL 107 (or MD-SVOL 110) (during data migration), and the write destination of the write target data is the PVOL 105 (or SVOL 108). ) MD-PVOL 107 (or MD-SVOL 110) is determined whether or not the area has been copied.

As a result, when data is being migrated and the write destination of the write target data is an area that has been copied to the MD-PVOL 107 (MD-SVOL 110) of the PVOL 105 (or SVOL 108) (step S46: Y), the DKC 100 The write target data is reflected in the corresponding area of the MD-PVOL 107 (MD-SVOL 110) (step S47), and the process ends. As a result, when a write occurs in the copied area of the MD-PVOL 107 or MD-SVOL 110, it can be appropriately reflected at this point. On the other hand, if data is not being migrated, or the write destination of the write target data is not an area that has already been copied to the MD-PVOL 107 (MD-SVOL 110) of the PVOL 105 (or SVOL 108) (step S46: N), the DKC 100 performs processing. finish.

Next, a computer system according to the second embodiment will be described. Here, the difference from the first embodiment will be mainly described, and the description of the common points with the first embodiment will be omitted or simplified (the same applies to the following third embodiment).

In the second embodiment, the PVOL 105 of the MS-PDKC 100A is migrated and the SVOL 108 of the MS-SDKC 100B is not migrated.

FIG. 22 is a diagram illustrating addition of an RC path according to the second embodiment. FIG. 22 is a diagram corresponding to the addition point of the RC path illustrated in FIG. 11 according to the first embodiment.

In the computer system according to the second embodiment, as shown in FIG. 22, the MD-SDKC 100D does not exist, and the port 3A of the MD-PDKC 100C is connected to the port 2C of the MS-SDKC 100B.

After the computer system is in the state shown in FIG. 10 (however, MD-SDKC 100D does not exist), the administrator adds a new remote copy path (RC path, RC path) via MD-PDKC 100C. .

A process for adding a new RC path via the MD-PDKC 100C and the MD-SDKC 100D will be described in detail.

MD-PDKC 100C transmits to MS-SDKC 100B a path addition request for forming a remote copy path from MD-PDKC 100C to MS-SDKC 100B (FIG. 22 (1)). The path addition request includes contents in which the virtual port information “1C” of the port 3A of the MD-PDKC 100C is the copy source and the port 2C of the MS-SDKC 100B is the copy destination. When the MS-SDKC 100B receives the path addition request, the virtualization of the port 3A of the MD-PDKC 100C is added to the copy source port information 122 of the vacant entry (here, the second entry) in the RC table 120 of the MS-SDKC 100B. The port information (“1C”) is stored, the device information “0101” of the PVOL 105 is stored in the copy source device information 123, the port information “2C” of the port 2C of the MS-SDKC 100B is stored in the copy destination port information 124, The SVOL device information “0201” is stored in the copy destination device information 124. The MD-PDKC 100C stores the port information “1C” of the port 1C of the MS-SDKC 100B in the MS-PP information 151 for the entry corresponding to the RC copy number formed in the PRC management table 150, and receives the transfer data receiving source. The virtual port information “2C” of the port 3C of the MD-PDKC 100C is stored in the virtual port information 153, and the device information “0101” of the PVOL 105 is stored in the PVOL device information 154.

Next, according to an instruction from the administrator, the MS-PDKC 100A adds the port information “1C” of the port 1C of the MS-PDKC 100A to the copy source port information 122 of the entry corresponding to the copy path to be added in the RC table 120 of the MS-PDKC 100A. And the virtual port information “2C” of the port 3C of the MD-PDKC 100C is stored in the copy destination port information 124. Further, the MS-PDKC 100A transmits a path addition request for adding a remote copy path from the MS-PDKC 100A to the MD-PDKC 100C (FIG. 22 (2)). Upon receiving the path addition request, the MD-PDKC 100C stores the virtual port information “1C” of the port 3A of the MD-PDKC 100C in the copy source port information 122 of the entry corresponding to the RC path to be added in the RC table 120. In the copy destination port information 124, the virtual port information “2C” of the port 2C of the MS-SDKC 100B is stored.

As a result of this processing, the path for remote copying from the PVOL 105 to the SVOL 108 is the path from the MS-PDKC 100A to the MS-SDKC 100B without going through the MD-PDKC 100C, and the path from the MS-PDKC 100A to the MS-SDKC 100B through the MD-PDKC 100C. Paths will be managed.

FIG. 23 is a diagram for explaining data migration of MS-VOL according to the second embodiment.

For example, after the computer system is brought into the state shown in FIG. 22, data migration processing for migrating the PVOL 105 of the MS-PDKC 100A to the MD-PVOL 107 of the MD-PDKC 100C is started as shown in FIG.

In the data migration process, the MD-PDKC 100C sequentially migrates data from the top area of the VVOL 106 based on the MS-PVOL 105 to the PVOL 107.

In the middle of this data migration processing, if there is a write request from the host 200 to the MD-PDKC 100C, the MD-PDKC 100C transfers the write request to the MS-PDKC 100A. The MS-PDKC 100A transmits a remote copy write request to the MD-PDKC 100C. The MD-PDKC 100C transmits a remote copy write request to the MS-SDKC 100B. The MS-SDKC 100B stores the write target data in the SVOL 108 in accordance with the write request. As a result, the write to the PVOL 105 can be appropriately reflected in the SVOL 108. Here, if the remote copy write request received from the MS-PDKC 100A is a write request for an area that has already been copied from the PVOL 105 to the MD-PVOL 107, the MD-PDKC 100C indicates that area of the MD-PVOL 107. Write target data to. As a result, the write to the migrated area can be appropriately reflected in the MD-PVOL 107.

Next, a computer system according to the third embodiment will be described.

In Example 3, the SVOL 108 of the MS-SDKC 100B is migrated and the PVOL 105 of the MS-PDKC 100A is not migrated.

FIG. 24 is a diagram illustrating addition of an RC path according to the third embodiment. FIG. 24 is a diagram corresponding to the addition point of the RC path illustrated in FIG. 11 according to the first embodiment.

After the computer system is in the state shown in FIG. 10 (however, MD-PDKC 100C does not exist), the administrator adds a new remote copy path (RC path, RC path) via MD-SDKC 100D. .

Next, a process for adding a new RC path via the MD-SDKC 100D will be described in detail.

The MD-SDKC 100D transmits a path addition request for adding a remote copy path from the MD-SDKC 100D to the MS-SDKC 100B to the MS-SDKC 100B (FIG. 24 (1)). The path addition request includes contents in which the virtual port information “1C” of the port 4C of the MD-SDKC 100D is the copy source and the port 2C of the MS-SDKC 100B is the copy destination. When the MS-SDKC 100B receives the path addition request, the virtual port of the port 4C of the MD-SDKC 100D is added to the copy source port information 122 of the vacant entry (here, the second entry) in the RC table 120 of the MS-SDKC 100B. Information (“1C”), the device information “0101” of the PVOL 105 is stored in the copy source device information 123, the port information “2C” of the port 2C of the MS-SDKC 100B is stored in the copy destination port information 124, and the copy SVOL device information “0201” is stored in the destination device information 124. Further, the MD-SDKC 100D sets the virtual port information “1C” of the port 4C of the MD-SDKC 100D to the transfer data transmission source virtualization port information 162 for the entry corresponding to the added RC copy number in the SRC management table 160. The port information “2C” of the port 2C of the MS-SDKC 100B is stored in the MS-SP information 163, and the device information “0201” of the SVOL 108 is stored in the SVOL device information 164.

Next, according to an instruction from the administrator, the MS-PDKC 100A adds the port information “1C” of the port 1C of the MS-PDKC 100A to the copy source port information 122 of the entry corresponding to the copy path to be added in the RC table 120 of the MS-PDKC 100A. And the virtual port information “2C” of the port 4A of the MD-SDKC 100D is stored in the copy destination port information 124. Further, the MS-PDKC 100A transmits a path addition request for adding a remote copy path from the MS-PDKC 100A to the MD-SDKC 100D (FIG. 24 (2)). Upon receiving the path addition request, the MD-SDKC 100D stores the virtual port information “1C” of the port 4C of the MD-SDKC 100D in the copy source port information 122 of the entry corresponding to the RC path to be added in the RC table 120. The port information “2C” of the port 2C of the MS-SDKC 100B is stored in the copy destination port information 124.

As a result of this process, the remote copy path from the PVOL 105 to the SVOL 108 is the path from the MS-PDKC 100A to the MS-SDKC 100B without going through the MD-SDKC 100D, and the path from the MS-PDKC 100A to the MS-SDKC 100B through the MD-SDKC 100D. Paths will be managed.

FIG. 25 is a diagram for explaining data migration of MS-VOL according to the third embodiment.

For example, after the computer system is brought into the state shown in FIG. 24, data migration processing for migrating the SVOL 108 of the MS-SDKC 100B to the MD-SVOL 110 of the MD-SDKC 100D is started as shown in FIG.

In the data migration process, the MD-SDKC 100D sequentially migrates data from the top area of the VVOL 109 based on the MS-SVOL 108 to the SVOL 110.

If there is a write request from the host 200 to the MS-PDKC 100A during the data migration process, the MS-PDKC 100A transmits a remote copy write request to the MD-SDKC 100D. The MD-SDKC 100D transmits a remote copy write request to the MS-SDKC 100B. The MS-SDKC 100B stores the write target data in the SVOL 108 in accordance with the write request. As a result, the write to the PVOL 105 can be appropriately reflected in the SVOL 108. Here, if the remote copy write request received from the MS-PDKC 100A is a write request for an area that has already been copied from the SVOL 108 to the MD-SVOL 110, the MD-SDKC 100D determines that area of the MD-SVOL 110. Write target data to. As a result, the write to the migrated area can be appropriately reflected in the MD-SVOL 110.

As mentioned above, although several examples were described, it cannot be overemphasized that this invention can be variously changed in the range which is not limited to these Examples and does not deviate from the summary.

100A ... MS-PDKC, 100B ... MS-SDKC, 100C ... MD-PDKC, 100D ... MD-SDKC, 101 ... Port, 102 ... MPPK, 105 ... PVOL, 108 ... SVOL, 200 ... Host

Claims (14)

1. A migration destination primary storage device connected to a migration source primary storage device that provides a first primary logical volume that forms a pair with a secondary logical volume of a secondary storage device;
   A second primary logical volume;
   A control device that manages a virtual primary logical volume obtained by virtualizing the first primary logical volume based on first identification information, and executes copying of data from the virtual primary volume to the second primary logical volume;
   Have
   The control device is
   When the first write request designating the first identification information is received before the copying is completed, the write data according to the first write request is transmitted to the migration source primary storage device,
   After the copying is completed, the identification information of the second primary logical volume is changed to the first identification information, and when the second write request designating the first identification information is received, the second write request is followed. Write data to the second primary logical volume and transfer it to the secondary storage device;
   A migration destination primary storage device.
2. The control device manages the second primary logical volume with second identification information before the copying is completed,
   The control device changes the identification information of the virtual primary logical volume to the second identification information when changing the identification information of the second primary logical volume to the first identification information.
   The migration destination primary storage apparatus according to claim 1.
3. When the control device receives a third write request designating the third identification information of the secondary logical volume from the migration source primary storage apparatus before the copy is completed, the control device receives the write data according to the third write request. Transfer to the secondary storage device,
   The migration destination primary storage apparatus according to claim 1.
4. Furthermore, it has a plurality of communication ports,
   The control device virtualizes identification information of a communication port connected to the migration source primary storage apparatus among the plurality of communication ports into identification information of a communication port of the secondary storage apparatus;
   The migration destination primary storage apparatus according to claim 1.
5. Furthermore, it has a plurality of communication ports,
   The control device virtualizes identification information of a communication port connected to the secondary storage apparatus among the plurality of communication ports into identification information of a communication port of the migration source primary storage apparatus;
   The migration destination primary storage apparatus according to claim 1.
6. When the write destination of the fourth write request received from the migration source primary storage apparatus is the copied area in the second primary logical volume during the copying, the control device writes the data to be written according to the fourth write request Write data to the copied area;
   The migration destination primary storage apparatus according to claim 1.
7. The control device transmits a completion notification to the host in response to the second write request after receiving the completion notification from the secondary storage device.
   The migration destination primary storage apparatus according to claim 1.
8. A method of migrating a migration source primary storage device that provides a first primary logical volume that forms a pair with a secondary logical volume of a secondary storage device to a migration destination primary storage device,
   The migration destination primary storage device manages a virtual primary logical volume obtained by virtualizing the first primary logical volume based on first identification information, and changes from the virtual primary volume to the second primary logical volume of the migration destination primary storage device. Copy data for
   When the migration destination primary storage device receives the first write request designating the first identification information before the copying is completed, the write data according to the first write request is transmitted to the migration source primary storage device. And
   When the migration destination primary storage apparatus changes the identification information of the second primary logical volume to the first identification information after receiving the second write request designating the first identification information after the copy is completed In addition, write data according to the second write request is written to the second primary logical volume and transferred to the secondary storage device.
   A method characterized by that.
9. The migration destination primary storage device manages the second primary logical volume with second identification information before the copying is completed,
   When the migration destination primary storage device changes the identification information of the second primary logical volume to the first identification information, the identification information of the virtual primary logical volume is changed to the second identification information;
   The method according to claim 8, wherein:
10. When the migration destination primary storage apparatus receives a third write request designating the third identification information of the secondary logical volume from the migration source primary storage apparatus before the copying is completed, the migration destination primary storage apparatus follows the third write request. Transferring write data to the secondary storage device;
    The method according to claim 8, wherein:
11. The migration destination primary storage apparatus uses the identification information of the communication port connected to the migration source primary storage apparatus among the plurality of communication ports of the migration destination primary storage apparatus as the identification information of the communication port of the secondary storage apparatus. Virtualize,
    The method according to claim 8, wherein:
12. The migration destination primary storage device uses the identification information of the communication port connected to the secondary storage device among the plurality of communication ports of the migration destination primary storage device as the communication port identification information of the migration source primary storage device. Virtualize,
    The method according to claim 8, wherein:
13. If the write destination of the fourth write request received from the migration source primary storage apparatus is the copied area in the second primary logical volume during the copying, the fourth write request Write the target data to be written to the copied area,
    The method according to claim 8, wherein:
14. The migration destination primary storage device receives a completion notification from the secondary storage device, and then sends a completion notification to the host in response to the second write request.
    The method according to claim 8, wherein:

Images