Method for processing I / O request, storage array and host

By installing the proxy module in the host, intercepting the mirror disk identity reported by the multipath module and setting a virtual path, the problem of data migration between heterogeneous storage arrays is solved, and online migration is realized, avoiding business interruptions and data loss.

CN112965659BActive Publication Date: 2024-11-15HUAWEI TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202110171553.3
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Priority Date
2017-10-10
Filing Date
2017-11-13
Publication Date
2024-11-15
Estimated Expiration
2037-11-13

AI Technical Summary

Technical Problem

Existing I/O request processing methods cannot implement online data migration between heterogeneous storage arrays, resulting in the risk of business interruption or data loss.

Method used

Install the proxy module in the host, intercept the mirrored disk identifier reported by the multipath module through the proxy module, set the virtual path, and map the identification of the disk to be migrated to the mirrored disk identifier, realizing online migration of data between heterogeneous storage arrays.

Benefits of technology

Online data migration between heterogeneous storage arrays is realized, which avoids business interruptions and ensures data security and integrity.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN112965659B_ABST
    Figure CN112965659B_ABST
Patent Text Reader

Abstract

A method for processing an I / O request is executed by a proxy module of a host. The proxy module sends the disk identifier of the disk to be migrated to the second storage array corresponding to the array identifier, so that the second storage array builds a mirror disk of the disk to be migrated according to the disk identifier of the disk to be migrated. After the mirror disk is built, the proxy module sets virtual path information for the path reporting the disk identifier of the mirror disk, and reports the virtual path information to the multipath module. The proxy module receives a disk identifier query request sent by the multipath module through a virtual path, and sends the disk identifier of the disk to be migrated to the multipath module, so that the multipath module manages the virtual path as the path of the disk to be migrated for priority access path, and sets the virtual path as the priority access path of the disk to be migrated.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the field of storage, and in particular to a method for processing I / O requests, a storage array, and a host. Background Art

[0002] In IT operation and maintenance, it is often necessary to replace old storage arrays with new ones, which requires migrating data from the old storage arrays to the new ones. However, in many scenarios, such as the financial industry and the telecommunications operator industry, users hope that the I / O (Input / Output) request processing between the old and new storage arrays is performed online to avoid business interruption or data loss.

[0003] The existing I / O request processing method can realize online data migration between homogeneous storage arrays (storage arrays that can be managed by the same multipath module, and the multipath module is software executed by the host), but cannot realize online data migration between heterogeneous storage arrays (storage arrays that cannot be managed by the same multipath module). Summary of the invention

[0004] The embodiments of the present invention provide an I / O request processing method, device, and host for realizing online migration of data between heterogeneous storage arrays.

[0005] A first aspect of an embodiment of the present invention provides an I / O request processing method for online migration of data in a disk to be migrated in a first storage array to a second storage array. In the implementation of this aspect, an agent module is installed in the host, and the agent module executes the I / O request processing method. After detecting a data migration request sent to the host, the agent module obtains the disk identifier of the disk to be migrated and the array identifier of the second storage array from the data migration request, and sends the disk identifier of the disk to be migrated to the second storage array corresponding to the array identifier. In this way, the second storage array can build a mirror disk of the disk to be migrated according to the disk identifier of the disk to be migrated. After the second storage array builds the mirror disk, the agent module notifies the second storage array to report the disk identifier of the disk belonging to the host, sets virtual path information for the path reporting the disk identifier of the mirror disk, and reports the virtual path information to the multipath module of the host. Afterwards, the agent module receives a disk identifier query command sent by the host through the virtual path indicated by the virtual path information. After receiving the disk identification query command, the proxy module sends the disk identification of the disk to be migrated to the multipath module. In this way, the multipath module can use the virtual path as the path of the disk to be migrated and set the virtual path as the priority access path of the disk to be migrated. In other words, the input and output I / O requests to access the disk to be migrated are preferentially sent through the virtual path.

[0006] By installing a proxy module in the host, the proxy module intercepts the disk identifier of the mirror disk reported by the second storage array to the multipath module, and sets a virtual path pointing to the proxy module for the path of the reported mirror disk. When the multipath queries the disk identifier of the mirror disk through the virtual path, the proxy module reports the disk identifier of the disk to be migrated to the multipath module through the virtual path, so that the multipath module uses the virtual path as the path of the disk to be migrated and uses the virtual path as the priority access path. In this way, the host can send the received IO accessing the disk to be migrated to the second storage array where the mirror disk is located through the virtual path, thereby realizing the online migration of data in the disk to be migrated.

[0007] In a possible implementation of the first aspect, after obtaining the disk identifier of the mirror disk sent by the second storage array after building the mirror disk, the proxy module will establish a mapping relationship between the disk identifier of the mirror disk and the disk identifier of the disk to be migrated. In this way, after receiving the disk identifier query request, the proxy module obtains the disk identifier of the mirror disk according to the disk code of the mirror disk in the disk identifier query request, obtains the disk identifier of the disk to be migrated corresponding to the disk identifier of the mirror disk according to the mapping relationship, and then sends the disk identifier of the disk to be migrated to the multipath module.

[0008] By establishing a mapping relationship between the disk identifier of the disk to be migrated and the disk identifier of the mirror disk, the proxy module can quickly find the disk identifier of the disk to be migrated corresponding to the disk identifier of the mirror disk when receiving the disk identifier of the mirror disk.

[0009] In combination with any one of the above possible implementations of the first aspect, in a possible implementation, when the proxy module notifies the second storage array to report the disk identification of the disk belonging to the host, it first triggers the host operating system to send a disk reporting command to the second storage array, and the disk reporting command instructs the second storage array to generate reporting information according to the reporting command and report the reporting information; then, after receiving the reporting information, obtain the path information in the reporting information, and the path information is the path information of the path for reporting the reporting information, which is generated in the process of reporting the reporting information; finally, send a disk identification query command to the second storage array through the path determined by the path information to notify the second storage array to report the disk identification of the disk belonging to the host.

[0010] Through this possible implementation, after triggering the operating system to send a disk reporting command to the second storage array, the proxy module intercepts the reporting information reported by the second storage array to the multipath module, and obtains the disk identifier in the second storage array through the proxy module, so that the proxy module can set a virtual path for accessing the disk to be migrated for the path of the disk identifier of the reported mirror disk.

[0011] In combination with any one of the above possible implementations of the first aspect, in a possible implementation, the proxy module obtains the disk identifier of the mirror disk sent by the second storage array after building the mirror disk; establishes a mapping relationship between the disk identifier of the mirror disk and the disk identifier of the disk to be migrated; after receiving the disk identifier of the disk belonging to the host reported by the second storage array, if the reported disk identifier exists in the mapping relationship, it is determined that the reported disk identifier is the disk identifier of the mirror disk. After determining the disk identifier of the mirror disk, virtual path information can be set for the path of the disk identifier of the mirror disk reported.

[0012] In combination with any one of the above possible implementations of the first aspect, in one possible implementation, the virtual path information includes a path number and a path pointer, the path number is used to represent the path for reporting the mirror disk identifier, and the path pointer points to the path for reporting the mirror disk identifier.

[0013] By setting the path pointer, the request sent by the multipath module through the virtual path can be sent to the proxy module.

[0014] In combination with any one of the above possible implementations of the first aspect, in a possible implementation, the first storage array and the second storage array are manufactured by different manufacturers, and the identifier of the disk to be migrated and the disk identifier of the mirror disk include manufacturer information.

[0015] The present invention is more conducive to solving the online migration of data when the first storage array and the second storage array are from different manufacturers.

[0016] In combination with any one of the above possible implementations of the first aspect, in a possible implementation, the proxy module sends the disk identifier of the disk to be migrated to the second storage array through a private command in a communication protocol between the host and the second storage array.

[0017] In combination with any one of the above possible implementations of the first aspect, in one possible implementation, the private command includes a command description block and a cache, the command description block instructs the second storage array to generate the mirror disk, the cache includes a sending information part and a return information part, the sending information part is used to write the disk identifier of the disk to be migrated; the return information part is used to write the disk identifier of the mirror disk after the second storage array builds the mirror disk.

[0018] In combination with any one of the above possible implementations of the first aspect, in a possible implementation, the migration method also includes, after the multi-path module sets the virtual path as the priority access path of the disk to be migrated, the proxy module sending a migration command to the second storage array, the migration command instructing the second storage array to migrate the data of the disk to be migrated, and the migration command includes the disk identifier of the mirror disk or the disk to be migrated.

[0019] The second aspect of the present invention provides an I / O request processing method for online migrating data in a disk to be migrated in a first storage array to a second storage array. The processing method is executed by a host. When the host obtains the disk identifier of the disk to be migrated and the array identifier of the second storage array according to the detected data migration request, and sends the disk identifier of the disk to be migrated to the second storage array corresponding to the array identifier, so that the second storage array builds a mirror disk of the disk to be migrated according to the disk identifier of the disk to be migrated, wherein the disk to be migrated is built in the first storage array. Afterwards, the host sets the mirror disk as the priority disk of the disk object of the disk to be migrated, and then receives an input / output I / O request to access the disk to be migrated, and sends the I / O request to the mirror disk according to the name of the disk object of the disk to be migrated carried in the I / O request.

[0020] The effect achieved by the I / O request processing method provided in the second aspect of the present invention is the same as that of the I / O request processing method provided in the first aspect, and will not be described in detail here.

[0021] In combination with the second aspect of the present invention, in a first possible implementation, a proxy module and a multipath module are run on the host. When the host sets the mirror disk as the disk with priority access to the disk object of the disk to be migrated, the proxy module notifies the second storage array to report the disk identifier of the disk belonging to the host, and sets virtual path information for the path reporting the disk identifier of the mirror disk, and then reports the virtual path information to the multipath module. The proxy module receives the disk identifier query request sent by the multipath module through the virtual path, and the virtual path is determined according to the virtual path information, and the disk identifier of the disk to be migrated is sent to the multipath module, so that the multipath module uses the virtual path as the path of the disk object of the disk to be migrated. After that, the virtual path is set as the priority access path of the disk object of the disk to be migrated.

[0022] In combination with the first possible implementation method of the second aspect of the present invention, in a possible implementation, when the I / O request is sent to the mirror disk, the priority access path of the disk object is first determined based on the disk object identifier, and then the I / O request is sent to the mirror disk through the priority access path.

[0023] In a possible implementation of the first aspect, after obtaining the disk identifier of the mirror disk sent by the second storage array after building the mirror disk, the proxy module will establish a mapping relationship between the disk identifier of the mirror disk and the disk identifier of the disk to be migrated. In this way, after receiving the disk identifier query request, the proxy module obtains the disk identifier of the mirror disk according to the disk code of the mirror disk in the disk identifier query request, obtains the disk identifier of the disk to be migrated corresponding to the disk identifier of the mirror disk according to the mapping relationship, and then sends the disk identifier of the disk to be migrated to the multipath software.

[0024] By establishing a mapping relationship between the disk identifier of the disk to be migrated and the disk identifier of the mirror disk, the proxy module can quickly find the disk identifier of the disk to be migrated corresponding to the disk identifier of the mirror disk when receiving the disk identifier of the mirror disk.

[0025] In combination with any one of the above possible implementations of the first aspect, in a possible implementation, when the proxy module notifies the second storage array to report the disk identification of the disk belonging to the host, it first triggers the host operating system to send a disk reporting command to the second storage array, and the disk reporting command instructs the second storage array to generate reporting information according to the reporting command and report the reporting information; then, after receiving the reporting information, obtain the path information in the reporting information, and the path information is the path information of the path for reporting the reporting information, which is generated in the process of reporting the reporting information; finally, send a disk identification query command to the second storage array through the path determined by the path information to notify the second storage array to report the disk identification of the disk belonging to the host.

[0026] Through this possible implementation, after triggering the operating system to send a disk reporting command to the second storage array, the proxy module intercepts the reporting information reported by the second storage array to the multipath module, and obtains the disk identifier in the second storage array through the proxy module, so that the proxy module can set a virtual path for accessing the disk to be migrated for the path of the disk identifier of the reported mirror disk.

[0027] In combination with any one of the above possible implementations of the second aspect, in one possible implementation, the proxy module obtains the disk identifier of the mirror disk sent by the second storage array after building the mirror disk; establishes a mapping relationship between the disk identifier of the mirror disk and the disk identifier of the disk to be migrated; after receiving the disk identifier of the disk belonging to the host reported by the second storage array, if the reported disk identifier exists in the mapping relationship, it is determined that the reported disk identifier is the disk identifier of the mirror disk. After determining the disk identifier of the mirror disk, virtual path information can be set for the path of the disk identifier of the mirror disk reported.

[0028] In combination with any possible implementation of the second aspect above, in one possible implementation, the virtual path information includes a path number and a path pointer, the path number is used to represent the path for reporting the mirror disk identifier, and the path pointer points to the path for reporting the mirror disk identifier.

[0029] By setting the path pointer, the request sent by the multipath module through the virtual path can be sent to the proxy module.

[0030] In combination with any one of the above possible implementations of the second aspect, in a possible implementation, the first storage array and the second storage array are manufactured by different manufacturers, and the identifier of the disk to be migrated and the disk identifier of the mirror disk include manufacturer information.

[0031] The present invention is more conducive to solving the online migration of data when the first storage array and the second storage array are from different manufacturers.

[0032] In combination with any one of the above possible implementations of the second aspect, in a possible implementation, the proxy module sends the disk identifier of the disk to be migrated to the second storage array through a private command in a communication protocol between the host and the second storage array.

[0033] In combination with any one of the above possible implementations of the second aspect, in one possible implementation, the private command includes a command description block and a cache, the command description block instructs the second storage array to generate the mirror disk, the cache includes a sending information part and a return information part, the sending information part is used to write the disk identifier of the disk to be migrated; the return information part is used to write the disk identifier of the mirror disk after the second storage array builds the mirror disk.

[0034] In combination with any possible implementation of the second aspect above, in one possible implementation, the migration method also includes sending a migration command to the second storage array after the multipath module sets the virtual path as the priority access path of the disk to be migrated, the migration command instructing the second storage array to migrate the data of the disk to be migrated, and the migration command includes the disk identifier of the mirror disk.

[0035] The third aspect of the present invention provides an I / O request processing method for online migrating data in a disk to be migrated in a first storage array to a second storage array. The processing method is executed by a host. The host obtains the disk identifier of the disk to be migrated and the array identifier of the second storage array according to the detected data migration request, and sends the disk identifier of the disk to be migrated to the second storage array corresponding to the array identifier, so that the second storage array builds a mirror disk of the disk to be migrated according to the disk identifier of the disk to be migrated, wherein the disk to be migrated is built in the first storage array. Then the mirror disk is set as the disk that is preferentially accessed by the disk object of the disk to be migrated. When receiving an input / output I / O request to access the disk to be migrated, the host sends the I / O request to the mirror disk according to the name of the disk object of the disk to be migrated carried in the I / O request.

[0036] The third aspect of the embodiment of the present invention provides an I / O request processing method applied to a host. The migration method provided by the second aspect of the embodiment of the present invention is different from the migration method provided by the first aspect in that, in the embodiment of the second aspect, the proxy module generates a comprehensive command after obtaining the migration request. The comprehensive command includes two commands, namely, a mirror disk generation command and a migration command. The mirror disk generation command and the migration command are implemented in the same way as those provided in the first aspect, and the execution order and triggering conditions of the mirror disk generation command and the migration command are defined in the comprehensive command. For example, it can be defined that the mirror disk generation command is executed first, and then the migration command is executed, and the execution condition of the migration command is defined as after the second storage array has established the virtual path, or after waiting for a preset time.

[0037] Through the implementation method provided in the third aspect, the proxy module only needs to send a command to the second storage array once, which can reduce the interaction between the host and the storage array.

[0038] The fourth aspect of the present invention provides an I / O request processing method applied to a host, which is used to migrate data in a disk to be migrated in a first storage array to a second storage array online. In the implementation of this aspect, a proxy module is installed in the host, and the I / O request processing method is executed by the proxy module. After detecting a data migration request sent to the host, the proxy module obtains the disk identifier of the disk to be migrated from the data migration request, and then returns the disk identifier to the client, so that the client obtains the disk identifier, and sends the disk identifier of the disk to be migrated to the second storage array corresponding to the array identifier, so that the second storage array builds a mirror disk of the disk to be migrated according to the disk identifier of the disk to be migrated. After receiving the feedback command that the second storage array has built the mirror disk, the proxy module receives the command to establish a virtual path sent by the client, notifies the second storage array to report the disk identifier of the disk belonging to the host according to the command, sets virtual path information for the path reporting the disk identifier of the mirror disk, and then reports the virtual path information to the multipath module of the host, and the virtual path reporting information indicates that the host sends a disk identifier query command to the proxy module through the virtual path indicated by the virtual path information. After receiving the disk identifier query command, the proxy module sends the disk identifier of the disk to be migrated to the multipath module, so that the multipath module uses the virtual path as the path of the disk to be migrated, and sets the virtual path as the priority access path of the disk to be migrated, and the input and output I / O requests to access the disk to be migrated are preferentially issued through the virtual path. The proxy module sends a feedback command to the client that the virtual path is established, and the client sends a migration command to the second storage array, and the migration command instructs the second storage array to migrate the data of the disk to be migrated.

[0039] By installing a proxy module in the host, the proxy module intercepts the disk identifier of the mirror disk reported by the second storage array to the multipath module, and sets a virtual path pointing to the proxy module for the path of the reported mirror disk. When the multipath module queries the disk identifier of the mirror disk through the virtual path, the proxy module reports the disk identifier of the disk to be migrated to the multipath module through the virtual path, so that the multipath module uses the virtual path as the path of the disk to be migrated and uses the virtual path as the priority access path. In this way, the host can send the received IO accessing the disk to be migrated to the second storage array where the mirror disk is located through the virtual path, thereby realizing the online migration of data in the disk to be migrated.

[0040] The fifth aspect of the present invention provides a method for processing input / output I / O requests. The method is executed by a target storage array. The target storage array is connected to a host, and the client is connected to the host and the target storage array. When receiving a mirror disk generation command sent by the client, the target storage array constructs a mirror disk of the disk to be migrated according to the disk identifier of the disk to be migrated carried in the mirror disk generation command, wherein the disk to be migrated is constructed in the source storage array. Then, a disk identifier query request sent by the host is received, and the disk identifier of the disk to be migrated is reported to the host through the access path of the mirror disk, so that the host records the access path of the mirror disk as the access path of the disk to be migrated, and the client sets the access path of the mirror disk as the priority access path of the disk to be migrated.

[0041] A mirror disk is established for the disk to be migrated in the target storage array. When the host queries the disk identifier of the mirror disk through the access path of the mirror disk, the disk identifier of the disk to be migrated is sent to the host, so that the host records the access path of the mirror disk as the access path of the disk to be migrated, and the client sets the access path of the mirror disk as the priority access path of the disk to be migrated. In this way, the host can send the received IO for accessing the disk to be migrated to the second storage array through the access path of the mirror disk, thereby realizing the online migration of data in the disk to be migrated.

[0042] According to the fifth aspect of the present invention, in a possible implementation, after the mirror disk of the disk to be migrated is constructed, the target storage array establishes a mapping relationship between the disk identifier of the mirror disk and the disk identifier of the disk to be migrated, and before reporting the disk identifier of the disk to be migrated to the host, after receiving the disk identifier query request, the target storage array obtains the disk identifier of the mirror disk according to the disk code of the mirror disk in the disk identifier query request, and then obtains the disk identifier of the disk to be migrated corresponding to the disk identifier of the mirror disk according to the mapping relationship.

[0043] According to the fifth aspect of the present invention, in a possible implementation, before receiving the disk identification query request, the target storage array sends a feedback command to the client after the mirror disk is built, so that the client triggers the host to send a disk reporting command to the target storage array. After receiving the disk reporting command, the target storage array obtains the disk code corresponding to the host port information according to the host port information in the reporting command, and reports the disk code of the mirror disk to the host, wherein the path for reporting the mirror disk code is the access path of the mirror disk, so that the host sends the disk identification query request through the access path of the mirror disk.

[0044] According to any possible implementation manner of the fifth aspect of the present invention, the target storage array receives a migration command sent by a client, and migrates the data of the disk to be migrated to the mirror disk according to the migration command.

[0045] A sixth aspect of an embodiment of the present invention provides a host. The host includes a mirror indication unit, a path reporting unit, and a disk identification reporting unit. The mirror indication unit is used to obtain the disk identification of the disk to be migrated and the array identification of the second storage array according to the detected data migration request, and send the disk identification of the disk to be migrated to the second storage array corresponding to the array identification, so that the second storage array builds a mirror disk of the disk to be migrated according to the disk identification of the disk to be migrated, and the disk to be migrated is built in the first storage array. The path reporting unit is used to notify the second storage array to report the disk identification of the disk belonging to the host, set virtual path information for the path of reporting the disk identification of the mirror disk, and report the virtual path information to the multi-path module of the host. The disk identification reporting unit is used to receive a disk identification query request sent by the host's multipath module through a virtual path. The virtual path is determined based on the virtual path information, and the disk identification of the disk to be migrated is sent to the multipath module, so that the multipath module manages the virtual path as the path of the disk to be migrated, and sets the virtual path as the priority access path of the disk to be migrated. Input and output I / O requests to access the disk to be migrated are preferentially sent through the virtual path.

[0046] Other possible implementations of the migration device provided in the sixth aspect correspond to the implementations of the migration method provided in the first aspect and are not described in detail here.

[0047] A seventh aspect of an embodiment of the present invention provides a host, which is connected to a first storage array and a second storage array. The host includes a mirror indication unit, a setting unit, and an I / O sending unit. The mirror indication unit obtains the disk identifier of the disk to be migrated and the array identifier of the second storage array according to the detected data migration request, and sends the disk identifier of the disk to be migrated to the second storage array corresponding to the array identifier, so that the second storage array builds a mirror disk of the disk to be migrated according to the disk identifier of the disk to be migrated, wherein the disk to be migrated is built in the first storage array. The setting unit is used to set the mirror disk as the disk with priority access to the disk object of the disk to be migrated. The I / O sending unit is used to receive an input / output I / O request to access the disk to be migrated, and send the I / O request to the mirror disk according to the name of the disk object of the disk to be migrated carried in the I / O request.

[0048] Other possible implementations of the host provided in the seventh aspect correspond to the implementations of the migration method provided in the second aspect and are not described in detail here.

[0049] An eighth aspect of an embodiment of the present invention provides a target storage array, wherein the target storage array is connected to a host, and a client is connected to the host and the target storage array. The target storage array includes a mirror disk generation unit and a reporting unit. The mirror disk generation unit is used to receive a mirror disk generation command sent by the client, and to construct a mirror disk of the disk to be migrated according to the disk identifier of the disk to be migrated carried in the mirror disk generation command, wherein the disk to be migrated is constructed in the source storage array. The reporting unit is used to receive a disk identifier query request sent by the host, and to report the disk identifier of the disk to be migrated to the host through the access path of the mirror disk. In this way, the host can record the access path of the mirror disk as the access path of the disk to be migrated, and the client can set the access path of the mirror disk as the priority access path of the disk to be migrated.

[0050] Other possible implementations of the target storage device provided in the eighth aspect correspond to implementations of the migration method provided in the second aspect and are not described in detail here.

[0051] A ninth aspect of an embodiment of the present invention provides a host. The host includes a processor and a memory. The memory is connected to the processor and is used to store program instructions. The processor runs the program instructions to execute the I / O request processing method in any possible implementation provided in the first to sixth aspects.

[0052] A tenth aspect of an embodiment of the present invention provides a computer-readable medium, comprising computer execution instructions. When a computer processor executes the computer execution instructions, the computer executes the I / O request processing method in any possible implementation of the first to sixth aspects. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art are briefly introduced below.

[0054] Figure 1 The following is an architecture diagram of a storage system used in an embodiment of the present invention.

[0055] Figure 2 for Figure 1 The multipath module in the host accesses Figure 1 A flow chart of a method for managing paths to disks in a storage array.

[0056] Figure 3 The diagram is a schematic diagram of a storage system after a new storage array is connected to the storage system when data in an old storage array needs to be migrated.

[0057] Figure 4 In the first embodiment of the present invention, Figure 3 A flow chart of a method for migrating data from a first storage array to a second storage array.

[0058] Figure 5 This is a schematic diagram of an interface for selecting a disk to be migrated provided by a client for a user in an embodiment of the present invention.

[0059] Figure 6 The invention provides a description table of the mirror disk generation command defined by the private command in the SCSI protocol in the embodiment of the present invention.

[0060] Figure 7 Schematic diagram of a process of establishing a virtual path in an embodiment of the present invention.

[0061] Figure 8 Flow chart of a method for migrating data from a first storage array to a second storage array according to a third embodiment of the present invention.

[0062] Fig. 9 A block diagram of a host provided in accordance with an embodiment of the present invention.

[0063] Fig.10 A structural diagram of a host provided in an embodiment of the present invention.

[0064] Fig.11 is a schematic diagram of a storage system in a fourth embodiment of the present invention.

[0065] Fig.12 Flow chart of a method for migrating data from a first storage array to a second storage array according to a fourth embodiment of the present invention.

[0066] Fig.13 FIG. 4 is a block diagram of a second storage array in a fourth embodiment of the present invention.

[0067] Fig.14 FIG. 4 is a structural diagram of a second storage array in a fourth embodiment of the present invention. DETAILED DESCRIPTION

[0068] like Figure 1 As shown, it is an architecture diagram of a storage system 100 used in an embodiment of the present invention. In the storage system 100, multiple hosts 101, 102 and 103 are connected to a storage array 130 through a switch 120. In this embodiment, the structures of each host are basically the same. In the following description, the host 101 is taken as an example for description. Various application software 111 are running on the host 101 to provide different application services for customers. The storage array 130 stores the business data generated by the application in the host during operation. The storage array 130 can be a flash memory array using a non-volatile memory standard (Non-Volatile Memory express, NVMe) interface, or it can be a storage array using a small computer system interface (Small Computer System Interface, SCSI). The client 300 can be connected to each host through a network 200 to implement operation or control of the host.

[0069] The storage array 130 can be divided into at least one disk, such as disks 1 to n, and the storage array 130 can communicate with the host using different interfaces, such as NVMe interface and SCSI interface. When the storage array 130 is a flash memory array using an NVMe interface, the host 100 communicates with the storage array 130 through the NVMe protocol, and the disk can be represented by the namespace (Namespace) defined in the NVMe protocol. When the storage array 130 is a storage array using a SCSI interface, the host 101 communicates with the storage array 130 through the SCSI protocol, and the disk can be represented by the logical unit number (Logic Unit Number, LUN) defined in the SCSI protocol. After the disks are divided, the storage array 130 will assign a disk identification (ID) and a disk code to each disk. Among them, the disk identifier can uniquely identify the disk, for example, the disk identifier includes a World Wide Name (WWN) applied to the SCSI protocol, vendor information, and product information; or the disk identifier includes a Globally Unique Identifier (GUID), vendor information, and product information applied to the NVMe protocol. The disk code is used to distinguish different disks in the storage array, such as LUN numbers LUN1, LUN2... under the SCSI protocol, and Namespace numbers under the NVMe protocol, such as Namespace1, Namespace2... After the disk identifier and disk code of the disk are allocated, the storage array 130 will configure the mapping relationship between the disk and the host, specifically, the mapping relationship between the disk code and the host port information (for example, HBA card port information), and the host to which the disk belongs can be determined according to the mapping relationship.

[0070] The host 101 further includes a multipath module 112, which may be a software program executed by the host 101. When the host 101 finds that a new device is connected, the multipath module 112 manages the disk in the new device and establishes a path to access the disk. The specific method is as follows: Figure 2 As shown in the flowchart.

[0071] Step S201, when the operating system 114 of the host 101 detects that the storage array 130 is connected to the host 101, the operating system 114 sends a disk reporting command to the storage array 130 through multiple paths between the host 101 and the storage array 130, and the disk reporting command includes the host port information for transmitting the disk reporting command.

[0072] In step S202, the storage array 130 receives a disk reporting command, obtains host port information in the disk reporting command, obtains the disk code corresponding to the host port according to the host port information and a mapping relationship between the disk code and the host port information, generates reporting information for the disk corresponding to the disk code, carries the disk code in the above reporting information, and reports the reporting information to the host 101.

[0073] According to the provisions of the communication protocol between the storage array 130 and the host 101, during the reporting process of the reported information, when the reported information passes through the storage array port (not shown) and the host port (not shown), the storage array port and the host port will respectively add the array port ID of the storage array port and the host port ID of the host port to the reported information, and the reported information includes the path information for reporting the reported information. For example, if the communication protocol between the host 101 and the storage array 130 is the SCSI protocol, the reporting command is the Report LUN command defined by the SCSI protocol.

[0074] Step S203, after receiving the reporting information, the driver software of the host port sets logical path information for the path reporting the reporting information according to the path information in the reporting information, adds the logical path information to the reporting information, and reports the reporting information with the added logical path information to the multipath module 112.

[0075] When setting the logical path information for the path for reporting the reported information, the driver software of the host port will first obtain the disk code in the received reported information, and set the logical path information for the path for reporting the reported information. The logical path information includes a path number and a path pointer. The path number can be determined according to the number of times the reported information containing the disk code is received. For example, if it is received for the first time, the path number is set to path 1, and if it is received for the second time, the path number is set to path 2. The path pointer points to the path information for reporting the reported information. The path information includes the array port ID and the host port ID in the reported information. After the path number is set, the path number can be used to identify the path for reporting the reported information.

[0076] Step S204: the multipath module 112 receives the reported information and obtains the logical path information in the reported information.

[0077] Step S205: the multipath module 112 sends a disk identification query command according to the path determined by the logical path information, wherein the query command includes the disk code in the reported information.

[0078] Step S206: the storage array 130 receives the query command, obtains the disk identifier corresponding to the disk code according to the disk code in the query command, and reports the disk identifier corresponding to the disk code through the path of sending the query command.

[0079] Since the storage array 130 allocates a disk identifier and a disk code to each disk when creating the disk, the storage array 130 can obtain the disk identifier according to the disk code after receiving the disk code.

[0080] Step S207: the multipath module 112 determines a path for accessing the disk corresponding to the disk identifier according to the reported disk identifier, and manages the determined path.

[0081] Specifically, when the multi-path module 112 receives a disk identifier that is different from other disk identifiers for the first time, it will create a disk object based on the disk identifier, including assigning a disk object name to the disk object, establishing a mapping relationship between the disk object name and the disk identifier, and recording the path of the reported disk identifier as the path of the disk object. Furthermore, other relevant information of the disk, such as address space, capacity, etc., can also be recorded in the disk object. If the same disk identifier is received subsequently, the path of the subsequent reported disk identifier can be recorded as another path of the disk object. In this way, the multi-path module 112 can manage multiple paths to access the disk through the disk object, such as discovering new paths, deleting disconnected paths, selecting paths to issue IO requests according to preset policies, etc. The preset policy can be a preset path priority, or load balancing of each path, etc.

[0082] For example, if the multipath module first receives the disk identifier of the disk whose disk code is LUN1 in path 1, a disk object is created for LUN1, and a disk object name sda ​​is assigned to the disk object, and a mapping relationship between sda and the disk identifier is established, and path 1 is recorded as the first path of the disk object. If the disk identifier of LUN1 is received on path 2 and path 3 later, path 2 and path 3 are also recorded as the paths of the disk object.

[0083] After the multipath module 112 manages the paths for accessing the disks in the storage array 130, the multipath module can send the IO sent by the host to the storage array 130 through the paths, and the storage array reads and writes the data in the disks.

[0084] When the original storage array needs to be replaced with a new storage array, the data of the original storage array needs to be migrated to the new storage array. In the prior art, when the new storage array and the original storage array are isomorphic storage arrays, online data migration can be achieved, and the specific implementation method is: a mirror disk of the disk to be migrated in the old storage array is constructed on the new storage array, and the multipath module of the host obtains the disk identification of the mirror disk, and the disk identification includes the manufacturer code. The multipath module determines whether the mirror disk can be managed according to the manufacturer code in the disk identification of the obtained mirror disk. If the manufacturer code in the disk identification is the manufacturer code preset in the multipath module, the multipath module can identify the mirror disk as a disk on a storage array that is isomorphic to the current storage array. The multipath module can manage the mirror disk and set the path of the disk identification reported for the mirror disk as the path to access the mirror disk. In order to transfer the IO requests of the old storage array to the new storage array, the multipath module aggregates the disk to be migrated and the mirror disk into an aggregate disk, and sets the path to access the disk to be migrated and the path to access the mirror disk as the access path of the aggregate disk, so that all IO requests to access the disk to be migrated will be sent to the aggregate disk. In addition, the multipath module sets the path to access the mirror disk as the priority access path of the aggregate disk, so that the IO requests sent to the aggregate disk will be sent to the mirror disk through the path to access the mirror disk.

[0085] Since the multipath module is generally provided by the array manufacturer, the manufacturer code of the array manufacturer is generally preset in the multipath module. Therefore, the multipath module will usually only recognize the storage array of the manufacturer. However, if the new storage array and the old storage array are heterogeneous storage arrays, since the multipath module provided by the manufacturer of the old storage array is installed on the host, the mirror disk built on the new storage array cannot be recognized. Therefore, it is impossible to transfer IO requests to the new storage array, and the client business running on the host can only be stopped, and the data on the old storage array can be migrated to the new storage array after disconnecting the connection between the old storage array and the host. After the data migration is completed, the new storage array is configured, and the client business is started based on the new storage array, that is, online data migration cannot be achieved.

[0086] In the embodiment of the present invention, when data on an old storage array is migrated to a new storage array that is heterogeneous with the old storage array, there is no need to stop the host's services. Instead, the host's services are sent to the new storage array, thereby achieving online migration of data in the heterogeneous storage array. Figure 3The diagram shows a schematic diagram of the storage system after the new storage array is connected to the storage system when the data in the old storage array needs to be migrated. Among them, the first storage array 140 is the old storage array, and the multipath module 112 installed in the host has managed the path of the disk 141 to be migrated that accesses the first storage array 140. The second storage array 150 is the new storage array, and the first storage array 140 and the second storage array 150 are in communication connection, and the communication protocol between the two can be the NVNe protocol or the SCSI protocol. Before the second storage array 150 is connected to the host 101, an agent module 113 is installed on the host 101. The agent module 113 is a software program executed by the host and is located at the lower layer of the multipath module 112. That is, when the host 101 interacts with the second storage array 150 through the multipath module 112, it must first pass through the agent module 113. By running the agent module 113, the host 101 can realize the online data migration of heterogeneous storage arrays without changing the existing multipath module 112. For specific methods, please refer to Figure 4 Description.

[0087] The following will be combined Figure 4 The flowchart shown illustrates a method for performing heterogeneous storage array data migration according to a first embodiment of the present invention.

[0088] First, in step S301, the user selects the disk to be migrated, that is, the disk to be migrated 141 and the target storage array to which the data of the disk to be migrated is migrated through the client 300, and sends a migration request to the host 101 to which the disk to be migrated belongs. In this embodiment, the target storage array is the second storage array 150. The migration request includes the disk name of the disk to be migrated 141 and the array identifier of the second storage array 150, and the array identifier is used to uniquely identify the second storage array 150.

[0089] When the client 300 is connected to the host 101 through the network 200 , the client 300 will obtain and store relevant information of the host 101 and the disks belonging to the host 101 , so the user can select the disk to be migrated through the display interface provided by the client 300 .

[0090] like Figure 5 In the example shown, the display interface 40 of the client 300 displays the disk name 41 of the disk belonging to each host 101. The user selects the corresponding disk name 41 to select the disk 141 to be migrated represented by the disk name 41, and clicks the migration button 42 provided by the system to generate a migration request, and carries the name 41 of the disk to be migrated, such as Sdb, in the migration request and sends it to the host 101 to which the disk to be migrated belongs. Figure 5The interface shown is only an example, and different interfaces may be used in different systems to display the hosts 101 in the storage system and the disks belonging to the hosts 101 .

[0091] The client 300 and the second storage array 150 are also connected via a network, so the client 300 can obtain the array identifier of the second storage array 150 from the second storage array 150 .

[0092] Step S302 , after detecting that the host 101 has received the migration request, the proxy module 113 obtains the disk name of the disk to be migrated 141 and the array identifier of the second storage array 150 from the migration request.

[0093] Step S303 : the proxy module 113 obtains the disk identifier of the disk to be migrated 141 according to the disk name of the disk to be migrated 141 .

[0094] like Figure 2 As described in step S206 of , after the storage array reports the disk ID of the disk, the host 101 generates a disk name for the disk and establishes a mapping relationship between the disk name and the disk ID. In this way, after the host 101 obtains the disk name of the disk 141 to be migrated, the disk ID of the disk 141 to be migrated can be obtained.

[0095] Step S304: the proxy module 113 generates a mirror disk generation command, which carries the disk identifier of the disk to be migrated, and sends the mirror disk generation command to the second storage array 150 determined according to the array identifier of the second storage array.

[0096] Generally, the communication protocols between the host and the storage array include the SCSI protocol and the NVMe protocol. These protocols include private commands, and the user can customize the private commands. In this embodiment, the user can customize the private command as the mirror disk generation command and carry the disk identifier of the disk to be migrated.

[0097] like Figure 6 As shown, it is a description table of the mirror disk generation command defined by the private command Inquiry VPD Page 0xC8 in the SCSI protocol.

[0098] The description table includes the name of the private command, the implemented function, command parameters and buffer. For example, in this embodiment, a private command named Inquiry VPD Page 0xC8 in the SCSI protocol is used. The function of the private command can be described in the function column of the description table as realizing the function of notifying the target array to generate a mirror disk. The parameter part is to define the function implemented by the private command through bytes, and the value here is negotiated in advance by the communicating parties. That is, after receiving the private command, the other end (such as the second storage array) can identify the function indicated by the private command according to the value of the command descriptor block (CDB) in the parameter part. For example, in this embodiment, the value 0x12 of the command descriptor block CDB[0] is used to indicate that the private command is an inquiry command in the SCSI command set; the value 0xc8 of the command descriptor block CDB[2] is used to indicate that the target array is notified to generate a mirror disk. The buffer includes two parts, the first part is the sending information part, which is used to fill in the information carried by the private command. In this embodiment, the private command carries the disk identification of the disk to be migrated, and the carried identification of the disk to be migrated can be filled into the sending information part in the cache. For example, the disk identification of the disk to be migrated includes the ID of the manufacturer of the disk to be migrated, which is AAA and occupies byte 0; the product name of the disk to be migrated, which is BBB and occupies byte 1; the WWN of the disk to be migrated, XXXXXXXXX, occupies bytes 3-66. The return information part is used to write the disk identification of the mirror disk after the second storage array has established the mirror disk of the disk to be migrated. For details, please refer to the description of step S306.

[0099] Step S305, after receiving the mirror disk generation command, the second storage array 150 obtains the disk identifier of the disk to be migrated carried in the mirror disk generation command, and then builds a mirror disk 151 for the disk to be migrated 141 (the process of building a mirror disk for the disk to be migrated 141 according to the disk identifier of the disk to be migrated will be described below), and allocates a disk identifier, a disk code, and a host port to the mirror disk 151.

[0100] The first storage array 140 and the second storage array 150 can also communicate via the SCSI protocol or the NVMe protocol. When the second storage array 150 constructs the mirror disk of the disk to be migrated 141, the second storage array 150 sends a disk reporting command to the first storage array 140. After receiving the disk reporting command, the first storage array 140 generates a reporting information for each disk on the first storage array 140, carries the disk code of the corresponding disk in each reporting information, and reports each reporting information to the second storage array 150. In the process of reporting the reporting information, the first array port and the second array port through which the reporting information passes will add their respective port IDs to the reporting information to generate path information of the path for reporting the reporting information. The second storage array 150 sends a disk identification query command to the first storage array 140 through the path represented by the path information. The query command includes the disk code in the path information. The first storage array 140 obtains the disk identifier corresponding to the disk code according to the disk code in the query command, and sends the disk identifier to the second storage array 150 through the path represented by the path information. If the disk identifier reported by the first storage array 140 is the same as the disk identifier of the disk to be migrated pre-stored in the second storage array 150, the second storage array 150 manages the path of the disk identifier reported for the disk to be migrated as the path to access the disk to be migrated. After the second storage array 150 manages the path to access the disk to be migrated, the second storage array 150 can obtain the relevant information of the disk to be migrated 141, such as the disk size, address space, etc., and then construct the mirror disk 151 of the disk to be migrated in the second storage array 150 according to this information. The size of the constructed mirror disk 151 can be greater than or equal to the disk to be migrated. The way in which the second storage array 150 constructs the mirror disk according to the disk information of the disk to be migrated is the same as the prior art, and will not be repeated here.

[0101] Step S306 , after the mirror disk is constructed, the second storage array 150 carries the disk identifier of the mirror disk 151 in the feedback command of the mirror disk generation command, and sends the feedback command to the host 101 .

[0102] like Figure 6As shown, the cache in the private command also includes the return information part. After the second storage array has established the mirror disk, the disk identifier of the mirror disk can be written into the return information part. For example, the manufacturer information CCC of the mirror disk is written into the 67th byte, the product information DDD of the mirror disk is written into the 68th byte, and the WWN of the mirror disk is written into the 69th to 132nd bytes. After the disk identifier of the mirror disk is written into the private command, a feedback command of the private command is formed, and the feedback command is sent to the host.

[0103] Step S307, after receiving the feedback command, the proxy module 113 obtains the disk identifier of the mirror disk 151, and establishes a mapping relationship between the disk identifier of the disk to be migrated 141 and the disk identifier of the mirror disk 151. For the convenience of description, the mapping relationship here is referred to as a disk identifier mapping relationship hereinafter.

[0104] Step S308, after the disk identification mapping relationship is established, the proxy module 113 triggers the operating system 114 to send a disk reporting command through one or more paths between the host 101 and the second storage array 150, and the disk reporting command includes the host port information for transmitting the disk reporting command.

[0105] Usually, when a storage array is connected to a host, or a hard disk is inserted into or removed from a storage array connected to the host, the host's operating system 114 will send a disk reporting command to the storage array to update the access path of the disk belonging to the host. In this embodiment, after the above disk identification mapping relationship is established, the proxy module 113 will also trigger the operating system 114 to send a disk reporting command to the second storage array through one or more paths between the host 101 and the second storage array 150.

[0106] Step S309, after receiving the disk report command, the second storage array 150 obtains the host port information in the disk report command, queries the mapping relationship between the disk code and the host port information to obtain the disk code corresponding to the host port, generates report information for the disk corresponding to the disk code, carries the disk code in the report information, and reports the generated report information to the host 101. In the process of reporting the report information, path information of the path for reporting the report information will be generated. For details, please refer to Figure 2 Description of steps S203 and 204 in .

[0107] When there are multiple paths between the host 101 and the second storage array 150, the host 101 sends a disk report command to the second storage array 150 through the multiple paths. The second storage array 150 generates a report message according to each disk report command received.

[0108] In step S310, the proxy module 113 receives the reported information, obtains and records the path information in the reported information, and then sends a disk identification query command to the second storage array 150 through the path determined by the path information, wherein the disk identification query command carries the disk code in the reported information.

[0109] The path information here is Figure 2 The logical path information described in step 203.

[0110] In the prior art, since the proxy module is not installed, the reported information will be directly reported to the multipath module, and the multipath module will send a disk identification query command to the second storage array according to the reported information. However, in this embodiment, since the proxy module is installed in the lower layer of the multipath module, the reported information will be intercepted by the proxy module, and after intercepting the reported information, the proxy module will send a disk identification query command to the second storage array 150.

[0111] Step S311, after receiving the disk identification query command, the second storage array 150 obtains the disk identification corresponding to the disk code according to the disk code carried in the disk identification query command, and reports the disk identification to the proxy module S113 through the path of sending the query command.

[0112] In the prior art, the disk identifier is directly reported to the multipath module. However, in this embodiment, due to the existence of the proxy module, the disk identifier is also reported to the proxy module.

[0113] Step S312: If the proxy module 113 determines that the reported disk identifier exists in the disk identifier mapping relationship, the reported disk identifier is considered to be the disk identifier of the mirror disk, and the path of the disk identifier of the mirror disk is reported (i.e., Figure 2 The logical path corresponding to the logical path information described in step S203, hereinafter referred to as the "disk identifier reporting path") sets virtual path information, and reports the set virtual path information to the multipath module 112.

[0114] As described in step S308, after establishing the disk identifier mapping relationship, the proxy module 113 triggers the operating system 114 to send a disk reporting command to the second storage array 150, so the disk identifiers received by the proxy module 113 are all disk identifiers of the disks in the second storage array 150. As long as the reported disk identifier exists in the mapping relationship of the disk identifiers, the reported disk identifier can be considered as the target disk of this migration, that is, the disk identifier of the mirror disk.

[0115] The virtual path information includes a path number, a path pointer, and a disk code. The path number is specified by the proxy module 113, and may be different from or the same as the path number of the path reporting the disk code of the mirror disk. The path pointer points to the path information of the disk identifier reporting path recorded in the proxy module 113, and the disk code is the disk code of the mirror disk.

[0116] Step S313, after receiving the virtual path information, the multipath module 112 sends a disk identification query command of the mirror disk to the proxy module through the virtual path determined by the virtual path information, and the disk identification query command includes the disk code of the mirror disk.

[0117] Since the virtual path information includes a path pointer, and the path pointer points to the path information of the disk identifier reporting path recorded in the proxy module, the multipath module 112 sends the disk identifier query command to the proxy module 113 .

[0118] In this embodiment, the multi-path module receives information reported by the proxy module, and the reported information includes virtual path information, and the path pointer of the virtual path information points to the proxy module. Therefore, the disk identification query command will be sent to the proxy module, and will not be sent to the second storage array as in the prior art.

[0119] Step S314, after receiving the disk identification query command, the proxy module 113 obtains the disk code of the mirror disk carried in the disk identification query command, obtains the disk identification of the mirror disk according to the disk code of the mirror disk, and obtains the disk identification of the disk to be migrated 141 according to the disk identification mapping relationship established in step 307, and then reports the disk identification of the disk to be migrated to the multipath module 112.

[0120] Step S315 , after obtaining the disk identifier of the disk to be migrated, the multipath module 112 sets the path represented by the virtual path information as the access path of the disk to be migrated.

[0121] It should be noted that, when the host 101 subsequently queries the identifier of the mirror disk through the virtual path represented by the virtual path information, the disk identifier of the disk to be migrated is also returned.

[0122] Step S316: The multipath module 112 sets the virtual path as a priority access path.

[0123] Generally, the multipath module or the host system has the function of setting the path priority. Here, the proxy module 113 may also carry an instruction to set the virtual path as the priority access path when reporting the identifier of the disk to be migrated. The multipath module 112 may set the virtual path as the priority access path according to the instruction, and the mirror disk accessed through the priority access path is the priority access disk.

[0124] After the virtual path is set as the priority access path, when the host 101 receives an IO request, the multipath module 112 sends the IO request to the proxy module 113 through the path pointer in the virtual path information. The proxy module 113 finds the disk identification reporting path pointed to by the virtual path according to the path pointer, and then sends the IO to the second storage array 150 through the actual path. The second storage array 150 accesses the data in the mirror disk 151 according to the IO request. In this way, it is ensured that the service of the host 101 is not interrupted when the second storage array 150 subsequently migrates the data of the disk 141 to be migrated.

[0125] It should be noted that, if the path to access the mirror disk is interrupted during the process of sending the IO request through the virtual path, the multipath module 112 will send the IO request through the path to access the disk to be migrated. After the path to access the mirror disk is restored or the virtual path is re-established, the IO request will be sent through the newly established virtual path, and the migration of data in the disk to be migrated will be restarted.

[0126] The process of establishing a virtual path can be further referenced Figure 7 Example shown.

[0127] For example, before migrating the data of the disk to be migrated in the first storage array, the host is only connected to the first storage array, and the agent module is not installed on the host. Figure 2In the method shown, since the multipath module receives the disk identifier (ID1) of the disk to be migrated on path 1 and path 2, the multipath module uses path 1 and path 2 as paths to access the disk to be migrated. When it is necessary to migrate the data of the disk to be migrated to the second storage array, the second storage array is connected to the host, and the proxy module is installed on the host as the lower software of the multipath module. The proxy module can be installed on the host before the second storage array is connected to the host, or it can be installed on the host after the second storage array is connected to the host. Figure 3 After the mirror disk of the disk to be migrated is constructed in steps 302 to 305, the second storage array returns the disk identifier of the mirror disk to the host. The proxy module of the host establishes a mapping relationship between the disk identifier of the disk to be migrated and the disk identifier of the mirror disk. After the mapping relationship is established, the proxy module triggers the operating system of the host to send a disk reporting command to the second storage array.

[0128] The second storage array first sends the reporting information to the port driver software. The port driver software receives the reporting information reported by the second storage array in succession, and generates corresponding logical path information for each reporting information according to the received reporting information. The logical path information includes the path code assigned to the logical path and the path pointer. The path code assigned to the same disk is not repeated. In this embodiment, if the port driver software receives two reports containing the disk code of the mirror disk, the path codes path 3 and path 4 are respectively assigned to the paths reporting the disk code of the mirror disk, and the path information of the two logical paths is reported to the proxy module. After receiving the two logical path information, the proxy module sends a disk identification query command carrying the disk code to the second storage array according to the path 3 and path 4 indicated by the two logical path information. The second storage array obtains the disk identification according to the disk code in the disk query command, and then sends the disk identification (for simplicity, only WWN2 is represented in the figure) to the proxy module. The proxy module determines that the disk identifier exists in the previously set mapping relationship between the disk identifier of the disk to be migrated and the disk identifier of the mirror disk, and then sets virtual paths for the path 3 and the path 4, namely, path 5 and path 6. The path information of the virtual path 5 and the virtual path 6 includes path pointers, which point to the path 3 and the path 4 respectively, and also includes the disk code of the disk to be migrated. The proxy module reports the path information of the path 5 and the path 6 to the multipath module. After receiving the path information of the path 5 and the path 6, the multipath module sends a disk identifier query command carrying the disk code of the mirror disk to the proxy module through the path 5 and the path 6. After receiving the query command, the proxy module obtains the disk identifier corresponding to the disk code of the mirror disk, obtains the disk identifier of the disk to be migrated according to the disk identifier mapping relationship, and then reports the disk identifier of the disk to be migrated to the multipath module through the path 5 and the path 6. After receiving the disk identifier of the disk to be migrated, the multipath module can set the path 5 and the path 6 as the access path of the disk to be migrated, and use the path 5 and the path 6 as the priority access path.

[0129] In step S317 , the proxy module 113 sends a migration command to the second storage array 150 , and carries the disk identifier of the mirror disk 151 in the migration command.

[0130] In other embodiments, the migration command may carry the identifiers of the disk to be migrated and the mirror disk at the same time, or only carry the identifier of the disk to be migrated.

[0131] Step S318, after receiving the migration command from the host 101, the second storage array 150 obtains the disk identifier of the disk to be migrated 140 corresponding to the mirror disk 151 according to the disk identifier of the mirror disk 151. Since the second storage array has established an access path to the disk to be migrated when establishing the mirror disk 151 for the disk to be migrated 141, the second storage array can directly access the disk to be migrated and migrate the data therein to the mirror disk 151.

[0132] It should be noted that the operation of the host sending a migration command to the second storage array to instruct the second storage array to perform data migration can also be performed before the virtual path is established. If the data migration operation is performed before the virtual path is established, a cache needs to be set in the first storage array, and the cache is used to store IO requests sent to the first storage array. After the virtual path is established, the IO requests cached in the cache are forwarded to the second storage array for execution.

[0133] Step S319 , after the data in the disk to be migrated 141 is migrated, the second storage array 150 feeds back a migration completion command to the host 101 .

[0134] In step S320, the host 101 returns a migration completion feedback command to the client 300. When the host returns the migration completion command, the name of the disk to be migrated may be carried.

[0135] Step S321: the client receives the command indicating that the migration is completed, and the data migration process of the disk to be migrated ends.

[0136] exist Figure 4In the described embodiment, by installing an agent module in the lower layer of the multipath module, the reporting information for reporting the disk code of the second storage array reported to the multipath module in the prior art will be intercepted by the agent module, and the agent module will further query the disk identifier of the disk belonging to the host in the second storage array according to the received reporting information, and then identify the disk identifier of the mirror disk from the reported disk identifier, set virtual path information for reporting the path of the mirror disk, the virtual path information points to the agent module, and report the disk code of the mirror disk and the virtual path information to the multipath module. In the prior art, the path information in the reporting information received by the multipath module is the path information from the second storage array to the multipath module, so the multipath module will send a disk identifier query command to the second storage array, but in this embodiment, since the information reported to the multipath module is virtual path information, the multipath module will send the disk identifier query command to the agent module through the virtual path indicated by the virtual path information. The proxy module will obtain the disk identifier of the disk to be migrated based on the disk identifier of the mirror disk queried by the multipath module, and report the disk identifier of the disk to be migrated to the multipath module. In this way, the multipath module can manage the virtual path as the path of the disk to be migrated, and after setting the virtual path as the priority access path, the IO received by the host can be sent to the mirror disk through the virtual path. In this way, when migrating the data of the disk to be migrated to the mirror disk, the host service does not need to be interrupted, thereby realizing online migration of data.

[0137] In the second embodiment of the present invention, the proxy module generates a comprehensive command in step S304, and the comprehensive command includes two commands, namely, a mirror disk generation command and a migration command. The mirror disk generation command and the migration command are the same as those in the first embodiment, and the execution order and trigger conditions of the mirror disk generation command and the migration command are defined in the comprehensive command. For example, it can be defined that the mirror disk generation command is executed first, and then the migration command is executed, and the execution condition of the migration command is defined as after the second storage array 150 has established the virtual path, or after waiting for a preset time.

[0138] After the comprehensive command is sent to the second storage array 150, the second storage array 150 first executes the mirror disk generation command, that is, executes steps S305 and S306. If the execution condition of the migration command defined in the comprehensive command is after the virtual path is established, the multipath module will send a notification command to the second storage array after the virtual path is established in step S316, and the second storage array can execute the migration command according to the notification command, that is, migrate the data in the disk to be migrated. For the specific migration process, please refer to the description of step S318. If the execution condition of the migration command defined in the comprehensive command is after waiting for a preset time, the second storage array starts timing after receiving the comprehensive command, and starts executing the migration command when the timing reaches the preset time, and migrates the data in the disk to be migrated. When executing the migration command, if the virtual path has been established, the IO received during the data migration process is directly sent to the second storage array through the virtual path. If the virtual path has not been established, the IO received during the data migration process is cached in the cache of the first storage array and transferred to the second storage array after the virtual path is established.

[0139] Figure 8 This is a flow chart of a method for performing heterogeneous storage array data migration in a third embodiment of the present invention.

[0140] The main difference between the third embodiment and the first embodiment is that in the third embodiment, the migration of data is mainly controlled by the client 300. The specific scheme is as follows:

[0141] Steps 601 to 603 are the same as steps S301 to S303 of the first embodiment. Since the client controls the data migration process in this embodiment, after the host obtains the disk identifier of the disk to be migrated in step S603, it returns the disk identifier of the disk to be migrated to the client in step S604.

[0142] In step S605, after receiving the disk identifier of the disk to be migrated, the client generates a mirror disk generation command and directly sends the mirror disk generation command to the second storage array.

[0143] In step S606, the second storage array generates a mirror disk according to the mirror disk generation command. The method of generating the mirror disk by the second storage array is the same as that of the first embodiment. For details, please refer to Figure 4 Step S305.

[0144] In step S607, the second storage array sends a feedback command carrying the mirror disk identifier to the client.

[0145] In step S608, the client sends a command for establishing a virtual path to the host, where the command for establishing a virtual path includes a disk identifier of the mirror disk.

[0146] In step S609, the proxy module establishes a mapping relationship between the identifier of the disk to be migrated and the identifier of the mirror disk.

[0147] Steps S610 to S618 and Figure 4 Steps 308 and 316 correspond to each other and perform the same functions. Please refer to Figure 4 The relevant description in will not be repeated here.

[0148] In step S619, the proxy module sends a feedback command to the client.

[0149] In step S620, after receiving the feedback command, the client sends a migration command to the second storage array, and the migration command carries the disk identifier of the disk to be migrated.

[0150] In step S621, the second storage array can migrate the data in the disk to be migrated according to the migration command. Figure 4 Description of step S318 in .

[0151] In step S622, if the second storage array completes the migration of the data in the disk to be migrated, a feedback instruction of the completion of the migration is sent to the client.

[0152] In step S623, the client receives a feedback instruction indicating that the migration is completed, and the data migration process ends.

[0153] It should be noted that, during the process of data migration, when a write request is received, the data to be written in the write request can be written to the mirror disk through the established logical path. When a read request is received, the read request can also be sent to the second storage array through the logical path. When the second storage array receives the read request, it reads the data in the mirror disk. If the read data has not been migrated from the disk to be migrated, the second storage array reads the data from the disk to be migrated in the first storage array.

[0154] In addition, during the data migration process, if it is found that new data has been written into the virtual disk, the migrated data does not need to overwrite the new data that has been written.

[0155] Fig. 9 FIG. 1 is a functional block diagram of the host 110 according to the first embodiment of the present invention.

[0156] The host 110 includes a mirror indication unit 901, a mapping establishment unit 902, a path reporting unit 903, a disk identification reporting unit 904, a migration indication unit 905, and an I / O sending unit 906. The mirror indication unit 901 is used to obtain the disk name of the disk to be migrated and the array identification of the second storage array from the migration request after detecting that the host has received a migration request, obtain the disk identification of the disk to be migrated according to the disk name of the disk to be migrated, and then generate a mirror disk generation command, and send the mirror disk generation command to the second storage array determined according to the array identification of the second storage array. The mirror disk generation command carries the disk identification of the disk to be migrated, and is used to instruct the second storage array to create a mirror disk for the disk to be migrated. For details, please refer to Figure 4 Description of steps S301-S304 in .

[0157] The mapping establishment unit 902 is used to receive the second storage array after building the mirror disk (the construction method of the mirror disk is shown in FIG. Figure 4 When the disk identifier of the mirror disk reported by the system (described in step S305 and S306 in the description of step S307) is obtained, a mapping relationship between the disk identifier of the disk to be migrated and the disk identifier of the mirror disk is established. Please refer to the description of step S307 for details.

[0158] The path reporting unit 903 is used to trigger the operating system to send a disk reporting command through one or more paths between the host and the second storage array. After receiving the reporting information generated and reported by the second storage array according to the reporting command (the process of the second storage array generating the reporting information according to the reporting command, please refer to Figure 4 The path reporting unit 903 acquires and records the path information in the reported information, and then sends a disk identification query command to the second storage array 150 through the path determined by the path information. The disk identification query command carries the disk code in the reported information. After receiving the disk identification reported by the second storage array (the process of the second storage array reporting the disk identification is described in detail in the description of step S309 ... Figure 4 ), if the path reporting unit 903 determines that the reported disk identifier exists in the disk identifier mapping relationship, it considers that the reported disk identifier is the disk identifier of the mirror disk, and sets virtual path information for the path of reporting the disk identifier of the mirror disk (hereinafter referred to as the "disk identifier reporting path"), and reports the set virtual path information to the multi-path module. For details, please refer to the relevant descriptions of steps S308, S310, and S312. After receiving the virtual path information reported by the path reporting unit 903, the multi-path module will send a disk identifier query command. For details, please refer to Figure 4Description of step S313 in .

[0159] The disk identification reporting unit 904 is used to obtain the disk code of the mirror disk carried in the disk identification query command after receiving the disk identification query command sent by the multipath module, obtain the disk identification of the mirror disk according to the disk code of the mirror disk, and replace the disk identification of the mirror disk with the disk identification of the disk to be migrated according to the disk identification mapping relationship, and then report the disk identification of the disk to be migrated to the multipath module. For details, please refer to Figure 4 Description of step S314 in .

[0160] After receiving the disk identifier of the disk to be migrated, the multipath module sets the path represented by the virtual path information as the access path of the disk to be migrated, and sets the virtual path as the priority access path. In this way, the I / O sending unit 906 can send the IO sent by the host to the second storage array through the virtual path, and the second storage array executes the IO to access the data in the mirror disk. For details, please refer to Figure 4 Description of steps S314 and S315 in .

[0161] The migration instruction unit 905 is used to send a migration command to the second storage array, and the migration command carries the disk identifier of the mirror disk, which is used to instruct the second storage array to migrate the data of the disk to be migrated to the mirror disk. Figure 4 Description of step S317 in .

[0162] After receiving the migration command, the second storage array can start to migrate the data in the disk to be migrated. After the data migration is completed, the migration completion command is fed back to the host, and the host continues to feed back the migration completion command to the client, thus completing the migration of the data in the disk to be migrated. For details, please refer to Figure 4 Description of steps S318 to S321 in .

[0163] It should be noted that the path reporting unit 903 and the disk identification reporting unit 904 may also be included in one setting module.

[0164] Fig.10 1 is a hardware structure diagram of a host 110 in an embodiment of the present invention. The host includes a processor 20, a memory 30, a system bus 40 and a host port 50. The processor 20, the memory 30 and the host port 50 are connected through the system bus 40 and communicate with each other. Figure 3 The multipath module 112 , the proxy module 113 and the operating system 114 in the host are all stored in the memory 30 in the form of program instructions 60 .

[0165] When the host 110 is running, the processor 20 executes the program instructions in the memory 30 to execute Figure 4 The method flow is shown.

[0166] The embodiment of the present invention further provides a computer-readable medium, including program instructions. When the processor of the host executes the program instructions, the host executes Figure 4 The method flow is described.

[0167] In a fourth embodiment of the present invention, a storage system 400 is provided. Fig.11 As shown, the architecture of the storage system 400 is different from the storage system 100 in the first embodiment only in that the host 410 in the fourth embodiment does not include an agent module, but includes an agent module 452 in the second storage array 450. The other parts are the same as the first embodiment and will not be repeated here.

[0168] The following will be combined Fig.12 The flowchart shown illustrates the method for data migration in the fourth embodiment.

[0169] In step S1201, the client 300 responds to the user's data migration operation by first sending a mirror disk generation command to the second storage array 450. The mirror disk generation command carries the disk identifier of the disk to be migrated.

[0170] Similar to the first embodiment, the first storage array 140 is an old storage array, the second storage array 450 is a new storage array, and the host 410 has been Figure 2 The method shown manages the path for accessing the disk mounted to the host in the first storage array 150. The host 410 stores the relevant information of the disk mounted to the host 410 in the first storage array 150, such as the disk name, disk code, disk identifier, etc. of the disk mounted to the host 410. When the user needs to migrate the data in the disk to be migrated 141 in the first storage array 140, it is necessary to first obtain the disk identifier of the disk to be migrated 141 from the host 410 through the client 300, and then send the disk identifier of the disk to be migrated to the second disk array 450 by carrying it in the mirror disk generation command.

[0171] Step S1202, after receiving the mirror disk generation command, the second storage array 450 obtains the disk identifier of the disk to be migrated carried in the mirror disk generation command, then builds a mirror disk 151 for the disk to be migrated 141, and allocates a disk identifier, disk code, and host port to the mirror disk 151.

[0172] An agent module 452 and a mirror module (not shown) are installed in the second storage array 450. The agent module 452 first receives the mirror disk generation command, and after parsing the mirror disk generation command, sends the parsed mirror disk generation command to the mirror module. The mirror module can generate the mirror disk 151 of the disk to be migrated according to the mirror disk generation command.

[0173] The process of building the mirror disk is the same as that in the first embodiment. For details, please refer to Figure 4 The description of step S305 is not repeated here.

[0174] Step S1203: the second storage array 450 establishes a mapping relationship between the disk identifier of the disk to be migrated and the disk identifier of the mirror disk.

[0175] After the mirror module builds the mirror disk, the mirror module 452 transmits the disk identifier of the mirror disk to the proxy module, and the proxy module establishes a mapping relationship between the disk identifier of the mirror disk and the disk identifier of the disk to be migrated.

[0176] Step S1204: the second storage array 450 sends a feedback command to the client 300 indicating that the mirror disk is established. The feedback command carries the disk identifier of the disk to be migrated and the disk code of the mirror disk.

[0177] After the proxy module 452 establishes the mapping relationship between the disk identifier of the mirror disk and the disk identifier of the disk to be migrated, it carries the disk identifier of the disk to be migrated and the disk code of the mirror disk in a feedback command and sends it to the client 300 .

[0178] Step S1205, after receiving the feedback command, the client 300 confirms that the mirror disk of the disk to be migrated has been established according to the disk identifier of the disk to be migrated, stores the disk code of the mirror disk, then logs in to the host 410 and performs a disk scan operation on the host.

[0179] It should be noted that after receiving the feedback command, the client 300 can automatically log in to the host to trigger the host to send a disk report command, or generate a notification to inform the user that the mirror disk of the disk to be migrated has been established, and the user can log in to the host 410 through the client 300 and perform a disk sweep operation on the host 410. The notification can be a prompt box, that is, the client 300 pops up a prompt box, and the content of the prompt box is that the mirror disk of the disk to be migrated has been established.

[0180] In step S1206 , the host 410 sends a disk report command to the second storage array 450 in response to the disk scan operation.

[0181] Step S1207: the second storage array 450 receives the disk report command and generates report information according to the disk report command. The report information includes the disk code of the reported disk. Figure 2 The relevant description of step S202 is not repeated here.

[0182] Step S1208: the second storage array 450 reports the reporting information of the mirror disk, and in the process of reporting the reporting information, generates path information of the path for reporting the reporting information.

[0183] When the generated reporting information passes through the proxy module 452, the proxy module 452 determines the reporting information of the mirror disk according to the disk code in the reporting information, and reports the reporting information of the mirror disk to the host. Since the second storage array will allocate a disk code and a disk identifier to the mirror disk after the mirror disk is constructed, the proxy module determines the disk identifier corresponding to the disk code according to the disk code in the reporting information. If the disk identifier corresponding to the disk code exists in the mapping relationship between the disk identifier of the mirror disk and the disk identifier of the disk to be migrated, the reporting information including the disk code is the reporting information of the mirror disk.

[0184] S1209, after receiving the report information of the mirror disk, the multipath module of the host sends a disk identification query command to the second storage array through the path indicated by the path information in the report information. The path indicated by the path information in the report information is the access path of the mirror disk.

[0185] S1210: After receiving the disk identification query command, the second storage array obtains the disk identification of the disk to be migrated according to the disk code of the mirror disk and the mapping relationship between the disk identification of the disk to be migrated and the disk identification of the mirror disk.

[0186] The disk identification query command carries the disk code of the mirror disk. The proxy module 452 determines the disk identification of the mirror disk according to the disk code of the mirror disk in the disk identification query command, and obtains the disk identification of the disk to be migrated according to the mapping relationship.

[0187] S1211: Report the disk identifier of the disk to be migrated to the host through the access path of the mirror disk.

[0188] S1212: The host uses the access path of the mirror disk as a path for accessing the disk to be migrated.

[0189] When the multipath module of the host determines that the received disk identifier is the disk identifier of the disk to be migrated, it will record the path information of the access path of the mirror disk of the disk to be migrated in the path information corresponding to the disk object (for example, sda) corresponding to the disk to be migrated, thereby using the access path of the mirror disk as the path to access the disk to be migrated.

[0190] S1213, after step S1205, the client starts to detect whether the path information of the disk to be migrated in the host contains the path information of the access path of the mirror disk.

[0191] The client queries the path information corresponding to the disk to be migrated in the host at a predetermined time after sending the disk identification reporting command, for example, 10 seconds later, or periodically. If the disk code included in the path information of the disk to be migrated is consistent with the disk code of the mirror disk reported by the second storage array in step S1204, the path represented by the path information is considered to be the access path of the mirror disk.

[0192] S1211: If the path information of the disk to be migrated contains path information of the access path of the mirror disk, the client sets the access path of the mirror disk as a priority access path of the disk to be migrated.

[0193] S1211: The client sends a migration command to the second storage array. The migration command may carry a disk identifier of the mirror disk or a disk identifier of the disk to be migrated.

[0194] S1212: The second storage array migrates the data in the disk to be migrated.

[0195] The process of migrating the data in the disk to be migrated by the second storage array is the same as that in the first embodiment. Figure 4 The description of step S318 in will not be repeated here.

[0196] S1213: After the data migration in the disk to be migrated is completed, the second storage array sends a migration completion command to the client.

[0197] S1214: After receiving the migration completion command, the client can complete the data migration process of the disk to be migrated.

[0198] During the data migration process, the host can send the I / O to the second storage array through the access path of the mirror disk, and the second storage array accesses the data in the mirror disk.

[0199] like Fig.13 As shown, it is a functional block diagram of the second storage array corresponding to the data migration method in the fourth embodiment. The second storage array includes a mirror disk generation unit 1301, a reporting unit 1302, and a migration unit 1303. The mirror disk generation unit 1301 is used to receive the mirror disk generation command sent by the client, and build a mirror disk of the disk to be migrated according to the disk identifier of the disk to be migrated carried in the mirror disk generation command, wherein the disk to be migrated is built in the first storage array, and after the mirror disk of the disk to be migrated is built, a mapping relationship between the disk identifier of the mirror disk and the disk identifier of the disk to be migrated is established, and a feedback command of the mirror disk establishment completion is sent to the client 300. The functions performed by the mirror disk generation module 1301 are the same as those of the first storage array. Fig.12 Corresponding to steps S1202 and S1204 in.

[0200] The reporting unit 1302 is used to receive a disk reporting command sent by the host, and generate reporting information according to the disk reporting command, wherein the reporting information includes the disk code of the reported disk. After receiving the reporting information of the mirror disk, the host sends a disk identification query command to the second storage array through the path indicated by the path information in the reporting information. Then, after receiving the disk identification query command, the reporting unit 1302 obtains the disk identification of the disk to be migrated according to the disk code of the mirror disk and the mapping relationship between the disk identification of the disk to be migrated and the disk identification of the mirror disk, and reports the disk identification of the disk to be migrated to the host through the access path of the mirror disk. The functions performed by the unit are the same as those of the host. Fig.12 Corresponding to 1208, 1210 and 1211.

[0201] In this way, the host uses the access path of the mirror disk as the path for accessing the disk to be migrated.

[0202] When the multipath module of the host determines that the received disk identifier is the disk identifier of the disk to be migrated, the path information of the access path of the mirror disk of the disk to be migrated will be recorded in the path information corresponding to the disk object (e.g., sda) corresponding to the disk to be migrated, thereby using the access path of the mirror disk as the path to access the disk to be migrated. Further, the client starts to detect whether there is path information of the access path of the mirror disk in the path information of the disk to be migrated in the host. If there is path information of the access path of the mirror disk in the path information of the disk to be migrated, the client sets the access path of the mirror disk as the priority access path of the disk to be migrated, and sends a migration command to the second storage array.

[0203] The migration unit 1303 is configured to migrate the data in the disk to be migrated after receiving the migration command from the client.

[0204] like Fig.13 14 is a block diagram of a second storage array 450 in a fourth embodiment of the present invention. The second storage array includes a processor 1401, a memory 1402, a system bus 1405, and an array port 1404. The processor 1401, the memory 1402, and the array port 1404 are connected and communicate with each other via the system bus 1405. The proxy module 452 is stored in the memory 1402 in the form of program instructions.

[0205] When the second storage array 450 is running, the processor 1401 executes the program instructions in the memory 1402 to execute Fig.12 The method flow is shown.

[0206] The fourth embodiment of the present invention further provides a computer-readable medium, including program instructions. When the processor 1401 of the second storage array 450 executes the program instructions, the second storage array executes Fig.12 The method flow executed by the second storage array.

[0207] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art who is familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A method for processing an input / output (I / O) request, executed by a proxy module of a host, wherein the host is connected to a first storage array and a second storage array, wherein the first storage array and the second storage array are heterogeneous storage arrays, and the method comprises: Acquire a disk identifier of a first disk from the second storage array, where the first disk is a mirror disk of a second disk in the first storage array in the second storage array; Setting a virtual path according to the disk identifier of the first disk, reporting the virtual path information to the multipath module, and disguising the virtual path as a path for accessing the second disk so as to transmit an I / O request for accessing the second disk to the first disk, wherein the virtual path includes a path between the first disk and the proxy module; The disguising the virtual path as an access path of the second disk comprises: receiving a disk identification query request sent by the multipath module through the virtual path, wherein the disk identification query request carries a disk code of the first disk; Sending the disk identifier of the second disk to the multipath module; receiving an input / output I / O request through the virtual path; The I / O request is sent to the first disk through the virtual path.

2. The method according to claim 1, characterized in that When it is determined that the data of the first disk needs to be migrated to the second storage array, the second storage array is notified to construct the first disk.

3. The method according to claim 2, characterized in that The method further comprises: receiving a disk identifier of the first disk sent by the second storage array after the first disk is constructed; Establishing a mapping relationship between the disk identifier of the first disk and the disk identifier of the second disk; The obtaining of the disk identifier of the first disk from the second storage array includes: receiving the disk identifier of the disk belonging to the host reported by the second storage array, and if the disk identifier exists in the mapping relationship, determining that the disk identifier is the disk identifier of the first disk.

4. The method according to claim 1, characterized in that The virtual path information includes a path number and a path pointer, the path number is a path label set for the virtual path, and the path pointer points to a path for accessing the first disk.

5. An input / output (I / O) request processing device, applied to a host, the host is connected to a first storage array and a second storage array, the first storage array and the second storage array are heterogeneous storage arrays, the processing device comprising: an acquiring unit, configured to acquire a disk identifier of a first disk from the second storage array, where the first disk is a mirror disk of a second disk in the first storage array in the second storage array; a camouflage unit, configured to set a virtual path according to the disk identifier of the first disk, report the virtual path information to the multipath module, and camouflage the virtual path as a path for accessing the second disk so as to transmit an I / O request for accessing the second disk to the first disk, wherein the virtual path includes a path between the first disk and the proxy module; The disguising the virtual path as an access path of the second disk comprises: receiving a disk identification query request sent by the multipath module through the virtual path, wherein the disk identification query request carries a disk code of the first disk; Sending the disk identifier of the second disk to the multipath module; The migration unit is configured to receive an input / output (I / O) request through the virtual path; and send the I / O request to the first disk through the virtual path.

6. The processing device according to claim 5, characterized in that The system further comprises a mirror instructing unit for instructing the second storage array to construct the first disk when it is determined that the data of the first disk needs to be migrated to the second storage array.

7. The processing device according to claim 6, characterized in that Also includes: A mapping establishment unit, configured to receive a disk identifier of the first disk sent by the second storage array after the first disk is constructed; Establishing a mapping relationship between the disk identifier of the first disk and the disk identifier of the second disk; The obtaining of the disk identifier of the first disk from the second storage array includes: receiving the disk identifier of the disk belonging to the host reported by the second storage array, and if the disk identifier exists in the mapping relationship, determining that the disk identifier is the disk identifier of the first disk.

8. The processing device according to claim 5, characterized in that The virtual path information includes a path number and a path pointer, the path number is a path label set for the virtual path, and the path pointer points to a path for accessing the first disk.

9. A host, comprising: processor, and A memory, connected to the processor, for storing program instructions, wherein the processor runs the program instructions to execute the I / O request processing method according to any one of claims 1 to 4.

10. A computer-readable medium comprising computer-executable instructions. When a processor of a computer executes the computer-executable instructions, the computer executes the I / O request processing method according to any one of claims 1 to 4.