CN118689788A - Data sharing method, device and equipment - Google Patents

Abstract

A data sharing method, device and equipment, in the present application, a computing system includes a first host, a second host, a storage controller, a storage device of the first host and a storage device of the second host. The first host sends a read request to the storage controller, and the read request is used to request to read the first data from the first address on the storage device of the first host. After receiving the read request, the storage controller determines that the first address is mapped to the second address on the storage device of the second host according to the sharing information of the first host. The storage controller reads the first data from the second address and feeds the first data back to the first host. The storage controller directly reads the first data from the second address and feeds it back to the first host, without moving the first data itself to the storage device of the first host, ensuring that the first host obtains the first data in time, thereby improving the data sharing efficiency between the first host and the second host.

Description

Data sharing method, device and equipment

CROSS-REFERENCE TO RELATED APPLICATIONS

This application claims priority to the Chinese patent application filed with the Intellectual Property Office of the People's Republic of China on March 22, 2023, with application number 202310293968.7 and invention name "A Computing System", the entire contents of which are incorporated by reference in this application.

Technical Field

The present application relates to the field of communication technology, and in particular to a data sharing method, device and equipment.

Background Art

Compute Express Link (CXL) is an evolution of the Peripheral Component Interconnect Express (PCIe) standard. Regarding the CXL protocol, the initial CXL1.0 protocol will gradually transition to the CXL2.0 protocol.

In the CXL 2.0 protocol, storage devices in a computing system built on CXL can be divided into multiple logical devices, which can be bound to different hosts in the computing system. The host can access the host's logical devices through CXL to read and write data.

However, in the CXL 2.0 protocol, data sharing between different hosts is accompanied by the migration of the data itself. Take the data sharing between host 1 and host 2 in the computing system as an example. When host 1 needs to share the data in the logical device of host 2, the data in the logical device of host 2 needs to be migrated to the logical device of host 1, that is, after copying the data from the logical device of host 2, the data is written to the logical device of host 1, and the data itself needs to be migrated. Host 1 reads the data from the logical device of host 1 by accessing its own logical device. The migration of data leads to low efficiency of data sharing between different hosts in the CXL 2.0 protocol.

Summary of the invention

The present application provides a data sharing method, device and equipment to improve the data sharing efficiency between hosts.

In a first aspect, an embodiment of the present application provides a data sharing method, which is applied to a computing system, wherein the computing system includes a first host, a second host, a storage controller, a storage device of the first host, and a storage device of the second host. The first host can access the storage device of the first host through the storage controller, and the second host can access the storage device of the second host through the storage controller. In an embodiment of the present application, the storage space in the storage device of the first host can be mapped to the storage space of the storage device of the second host. In this way, the first host can access the storage space of the storage device of the second host to which the storage space in the storage device of the first host is mapped. The access process is described below:

The first host may send a read request to the storage controller, where the read request is used to request to read first data (the first data may be data 1 or data 2 in the embodiment of the present application) from a first address on the storage device of the first host.

After receiving the read request, the storage controller can determine the mapping of the first address to the second address on the storage device of the second host according to the sharing information of the first host, where the sharing information of the first host indicates the mapping relationship between the first address and the second address.

After determining the second address according to the shared information of the first host, the storage controller reads the first data from the second address and feeds the first data back to the first host.

In the above description, only the first host is used as an example. Similarly, the storage space in the storage device of the second host can also be mapped to the storage space of the storage device of the first host. In this way, the second host can access the storage space of the storage device of the first host to which the storage space in the storage device of the second host is mapped. The access process is similar to the way the first host accesses the storage space of the storage device of the second host. For details, please refer to the above description, which will not be repeated here.

Through the above method, since the mapping relationship between the first address and the second address is indicated in the sharing information of the first host, the storage controller only needs to read the first data directly from the second address after determining the second address and feed it back to the first host. There is no need to move the first data itself to the storage device of the first host, which reduces data movement, ensures that the first host can obtain the first data in time, and improves the data sharing efficiency between the first host and the second host.

In a possible implementation, before the first host initiates a read request, the first data may be migrated from the second address to the first address through the storage controller. The process is as follows:

The first host sends a data migration instruction to the storage controller, where the data migration instruction is used to request to migrate the first data from the second address to the first address.

After receiving the data migration instruction, the storage controller updates the sharing information of the first host, and the updated sharing information of the first host indicates the mapping relationship between the first address and the second address.

Through the above method, when the host instructs the storage controller to migrate the first data from the second address to the first address, the storage controller does not need to migrate the first data itself, but only needs to add a mapping relationship between the first address and the second address in the shared information of the first host, thereby avoiding the high power consumption caused by the migration of the data itself, and the data migration efficiency is higher. It also ensures that when the first host subsequently reads the first data, the storage controller can read the first data from the second host.

In a possible implementation, the first host may also write data into the storage device of the first host. Here, the second data is taken as an example, and the second data corresponds to the data 3 mentioned in the embodiment of the present application.

The first host sends a write request to the storage controller, where the write request is used to request to write second data at a first address.

After receiving the write request, the storage controller deletes the mapping relationship between the first address and the second address in the shared information of the first host. The storage controller accesses the storage device of the first host and writes the second data at the first address.

Through the above method, after receiving the write request, the storage controller promptly deletes the mapping relationship between the first address and the second address, thereby avoiding failure to read the second data when subsequently reading data from the first address.

In a possible implementation, the sharing information of the second host may record whether the storage space in the storage device of the second host is allowed to be shared. When the storage controller updates the sharing information of the first host, the storage controller may query the sharing information of the second host to determine whether the second address is allowed to be shared. If the sharing information of the second host indicates that the second address is allowed to be shared, the storage controller updates the sharing information of the first host.

Through the above method, the storage controller determines whether the second address allows sharing before updating the sharing information of the first host, so as to ensure the security of the data at the second address in the storage device of the second host and avoid data leakage of the storage device of the second host.

In a possible implementation, the storage controller receives a data migration instruction from the first host, the data migration instruction is used to request to migrate the first data from the second address to the first address; the storage controller can query the sharing information of the second host to determine whether the second address allows sharing. If the sharing information of the second host indicates that the second address prohibits sharing, the storage controller notifies the first host that the data migration fails. Since the second address does not allow sharing, the storage controller avoids leakage of data at the second address by notifying the first host of the data migration failure.

In a possible implementation, the computing system may be deployed in multiple computing devices, and the first host, the second host, the storage controller, the storage device of the first host, and the storage device of the second host may be distributed in multiple computing devices. For example, the first host and the storage device of the first host may be located in different computing devices, and the first host and the second host may be located in different computing devices. The computing system may also be deployed in one computing device, that is, the first host, the second host, the storage controller, the storage device of the first host, and the storage device of the second host are deployed in one computing device.

Through the above methods, there are multiple deployment methods for computing systems, suitable for different scenarios.

In a possible implementation, the storage device of the first host and the storage device of the second host may be physical storage devices, or may be logical devices formed by virtualization of physical storage devices.

Through the above method, the storage device of the first host and the storage device of the second host are more flexible in form, which effectively expands the applicable scenarios of the method.

In a possible implementation, the first address and the second address are both HPA. When the storage controller reads the first data from the second address, it can first query the host device address mapping table of the second host, convert the second address into the DPA of the second host, and read the first data from the DPA, wherein the host device address mapping table of the second host describes the mapping relationship between the HPA and DPA of the second host.

Through the above method, the storage controller can first query the sharing information of the first host, and then query the host device address mapping table of the second host to ensure that the first data can be read from the second address.

In a possible implementation, the storage controller saves the sharing information of each host. Taking the sharing information of the first host as an example, the sharing information of the first host includes the HPA of the first host, the HPA of the second host that has a mapping relationship with the HPA of the first host, and the information of the second host. The HPA of the first host can represent the storage space of the storage device of the first host. The HPA of the second host can represent the storage space of the storage device of the second host.

Through the above method, the storage controller can maintain the shared information of each host to ensure that data can be shared between the hosts.

In a possible implementation manner, the sharing information of the first host further includes a sharing mark, and the sharing mark is used to indicate whether the HPA of the first host allows sharing.

Through the above method, by setting an explicit sharing mark, it can be made clear whether the HPA of the first host allows sharing.

In a possible implementation, if the sharing information of the first host does not indicate a mapping relationship between the first address and the second address; that is, there is no mapping relationship between the first address and the second address. After receiving the read request, the storage controller can access the storage device of the first host, read the first data from the first address, and feed the first data back to the first host.

Through the above method, after receiving a read request, if the first shared information does not indicate the address mapped by the first address, the storage controller can access the storage device of the first host according to the normal data reading process, read the first data from the first address, and feed the first data back to the first host.

In a possible implementation, the first address and the second address are both HPA, and the shared information of the first host is represented by DPA to represent the storage space of the storage device of the first host or the storage space of the storage device of the second host. After receiving the read request, the storage controller first queries the host device address mapping table of the first host and converts the first address into the DPA of the first host, wherein the host device address mapping table of the first host describes the mapping relationship between the HPA and DPA of the first host;

The storage controller determines that the DPA of the first host is mapped to the DPA of the storage device of the second host according to the shared information of the first host, wherein the second address and the DPA of the second host have a mapping relationship. Afterwards, the storage controller reads the first data from the DPA of the second host.

Through the above method, the storage controller can first query the host device address mapping table of the first host, and then query the shared information of the first host to ensure that the first data can be read from the second address.

In a second aspect, an embodiment of the present application further provides a computing system, the computing system comprising a first host, a second host, a storage controller, a storage device of the first host, and a storage device of the second host, the computing system may be connected based on a computing interconnect bus. The storage controller and the first host cooperate to implement the behavior in the method example of the first aspect above, and the beneficial effects can be referred to the description of the first aspect and will not be repeated here.

The first host is used to send a read request to the storage controller, where the read request is used to request to read first data from a first address on the storage device of the first host.

A storage controller is used to receive a read request, determine the mapping of a first address to a second address on a storage device of a second host according to shared information of a first host, wherein the shared information of the first host indicates a mapping relationship between the first address and the second address; read first data from the second address, and feed the first data back to the first host.

In a possible implementation, the first host sends a data migration instruction to the storage controller before sending the read request, where the data migration instruction is used to request to migrate the first data from the second address to the first address.

After receiving the data migration instruction, the storage controller updates the sharing information of the first host, and the updated sharing information of the first host indicates the mapping relationship between the first address and the second address.

In a possible implementation, the first host sends a write request to the storage controller, where the write request is used to request to write the second data at the first address.

After receiving the write request, the storage controller deletes the mapping relationship between the first address and the second address in the shared information of the first host; accesses the storage device of the first host, and writes the second data at the first address.

In a possible implementation, before updating the sharing information of the first host, the storage controller may query the sharing information of the second host, and the sharing information of the second host indicates that the second address allows sharing.

In one possible implementation, the storage controller receives a data migration instruction from the first host, where the data migration instruction is used to request migration of the first data from the second address to the first address; when the storage controller determines that the sharing information of the second host indicates that sharing of the second address is prohibited, the storage controller notifies the first host that the data migration has failed.

In a possible implementation, the computing system may be deployed on one computing device or on multiple computing devices.

In a possible implementation, the storage device of the first host and the storage device of the second host may be logical devices virtually formed by physical storage devices, or the storage device of the first host and the storage device of the second host may also be physical storage devices.

In one possible implementation, the first address and the second address are both HPA. When the storage controller reads the first data from the second address, it queries the host device address mapping table of the second host, converts the second address into the DPA of the second host, and reads the first data from the DPA, wherein the host device address mapping table of the second host describes the mapping relationship between the HPA and DPA of the second host.

In a possible implementation, the storage controller stores sharing information of the first host, where the sharing information of the first host includes the HPA of the first host, the HPA of the second host that is mapped to the HPA of the first host, and information of the second host.

In a possible implementation manner, the sharing information of the first host further includes a sharing mark, and the sharing mark is used to indicate whether the HPA of the first host allows sharing.

In a possible implementation, if the sharing information of the first host does not indicate a mapping relationship between the first address and the second address; after receiving a read request, the storage controller accesses the storage device of the first host, reads the first data from the first address, and feeds the first data back to the first host.

In a possible implementation, the first address and the second address are both HPA, and the shared information of the first host uses DPA to represent the storage space of the storage device of the first host or the storage space of the storage device of the second host.

After receiving the read request, the storage controller first queries the host device address mapping table of the first host and converts the first address into the DPA of the storage device of the first host, wherein the host device address mapping table of the first host describes the mapping relationship between the HPA and DPA of the first host.

The storage controller determines, based on the shared information of the first host, that the DPA of the first host is mapped to the DPA of the second host, and that there is a mapping relationship between the second address and the DPA of the second host. Afterwards, the storage controller reads the first data from the DPA of the second host.

In the third aspect, the embodiment of the present application further provides a data sharing device, which has the function of implementing the behavior in the method example of the first aspect above. The beneficial effects can be found in the description of the first aspect and will not be repeated here. The functions can be implemented by hardware, or by hardware executing corresponding software implementations. The hardware or software includes one or more modules corresponding to the above functions. In one possible design, the structure of the device includes a receiving module, a processing module, and a sending module, which can execute the corresponding functions in the method example of the first aspect above. Please refer to the detailed description in the method example for details, which will not be repeated here.

In a fourth aspect, the present application further provides a computing device, the computing device comprising a processor and a power supply circuit, the power supply circuit being used to power the processor, the processor being able to execute the method provided in the first aspect or any possible implementation of the first aspect. The processor may also call computer program instructions stored in a memory to execute the method provided in the first aspect or any possible implementation of the first aspect, the memory being coupled to the processor, and storing the necessary computer program instructions and data in the data sharing process.

In a fifth aspect, the present application provides a computing device system, which includes at least one computing device. Each computing device includes a memory and a processor. The processor of at least one computing device is used to access the code in the memory to execute the method provided by the first aspect or any possible implementation of the first aspect.

In a sixth aspect, the present application provides a non-transitory readable storage medium, and when the non-transitory readable storage medium is executed by a computing device, the computing device executes the method provided in the aforementioned first aspect or any possible implementation of the first aspect. The storage medium stores a program. The storage medium includes but is not limited to a volatile memory, such as a random access memory, a non-volatile memory, such as a flash memory, a hard disk drive (HDD), and a solid state drive (SSD).

In a seventh aspect, the present application provides a computing device program product, the computing device program product comprising computer instructions, when executed by a computing device, the computing device performs the method provided in the aforementioned first aspect or any possible implementation of the first aspect. The computer program product may be a software installation package, and when the method provided in the aforementioned first aspect or any possible implementation of the first aspect needs to be used, the computer program product may be downloaded and executed on a computing device.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG1 is a schematic diagram of the architecture of a computing system provided by the present application;

FIG2 is a schematic diagram of a data sharing method provided by the present application;

3 to 7 are schematic diagrams of shared information of a host provided by the present application;

FIG8 is a schematic diagram of the structure of a data sharing device provided by the present application;

9 and 10 are schematic diagrams of the structure of a computing device provided by the present application.

DETAILED DESCRIPTION

Before introducing a data sharing method, system, and device provided in the embodiments of the present application, some concepts involved in the embodiments of the present application are first explained.

(1) Virtualization technology, computing instance, container, virtual machine (VM).

Virtualization is a resource management technology that abstracts and transforms the host's physical resources, such as processors, memory, and interfaces, and presents them. Virtualization is a resource configuration method from a logical perspective and is a logical abstraction of physical resources.

The computing device can form a software module with an independent operating environment on the computing device by means of virtualization technology. In the embodiment of the present application, the software module with an independent operating environment formed on the computing device is called a computing instance. The computing instance can be a virtual machine or a container.

A virtual machine is a "complete computer" that runs in a completely isolated environment and has complete hardware system functions, which is simulated by virtualization technology. All the work that can be done in a physical computer can be done in a virtual machine. A virtual machine has a processor (the processor is also called a virtual processor), memory, hard disk and other components. The processor, memory, hard disk and other components of the virtual machine are virtualized from the processor, memory, hard disk and other components of the host.

A container is an independent operating environment simulated by virtualization technology. A container is similar to a lightweight sandbox that shields software and hardware outside the container. The container implements virtualization at the operating system level and directly reuses the host's operating system.

For computing devices, computing instances are considered to be a special kind of "process". This "process" will perform computing tasks and occupy the computing device's processor, memory, hard disk and other resources.

(2) Direct access to equipment and direct access to equipment.

Device passthrough refers to passing a device directly to the host for use. In order to facilitate passing a device directly to the host, some virtualization technologies at the hardware level have been developed. Virtualization technologies based on the hardware level can virtualize the hardware into multiple "devices". These "devices" are recognized as independent hardware devices by computing devices (such as operating systems) or computing instances (such as virtual machines and containers). Both computing devices and computing instances will use these virtual devices as real hardware. Single-root I/O virtualization (SR-IOV) technology is a common hardware-level virtualization technology.

In the embodiment of the present application, with the help of hardware-level virtualization technology, a physical storage device can be virtualized into multiple logical devices. Each logical device has the function of a storage device and can realize data storage. The host can store its own data on different logical hard disks.

(3) Compute Express Link (CXL).

CXL is evolved from the Peripheral Component Interconnect Express (PCIe). CXL runs on the PCIE physical layer and has the same electrical characteristics as PCIE. Compared with PCIE, CXL optimizes cache and memory, and can support lower latency and higher bandwidth.

Under the CXL standard, "multi-logic device (MLD)" is proposed. The MLD component is a component that implements the logical division of resources under the CXL standard. The MLD component can be understood as a software module and a hardware module. For storage resources such as storage devices, the MLD component can divide the storage device into 16 or more logical devices, assign a logical device identifier (LD-ID) to each logical device, and each logical device can operate as an independent CXL device (CXL device refers to a device connected through CXL). The host can identify the logical device and access it.

As shown in FIG1 , it is a schematic diagram of the structure of a computing system provided in an embodiment of the present application, in which a host 100, a storage controller 200 and a storage device 300 are included in the computing system. The number of hosts 100 and storage devices 300 included in the computing device is not limited in the embodiment of the present application, and the computing system may include one host 100 or multiple hosts 100. The computing system may include one storage device 300 or multiple storage devices 300. FIG1 only shows three hosts 100 and three storage devices 300 by way of example.

In the computing system, the host 100, the storage controller 200 and the storage device 300 can be connected based on a computing interconnect bus, wherein the computing interconnect bus can be a PCIe-based line, or a bus of CXL, unified bus (Ubus or UB) or cache coherent interconnect for accelerators (CCIX) protocol or other protocols.

The host 100 is the main data processing unit in the computing system and is the main part of the computing system that undertakes the business. The host 100 can be a computing device such as a desktop computer, a server, a laptop, and a mobile device. The host 100 can be a bare metal server. The host 100 can also be a computing instance such as a container or a virtual machine deployed on a computing device, and the host 100 can also be a processor or a processor core.

Among them, the processor can be a central processing unit (CPU), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA), an artificial intelligence (AI) chip, a system on chip (SoC) or a complex programmable logic device (CPLD), a graphics processing unit (GPU), a data processing unit (DPU), etc., and any module with processing functions can be used as a processor. The processor core is the core in the processor, and the embodiment of the present application does not show the type of the processor core. The processor core can be an advanced reduced instruction set machine (ARM) or a microprocessor without interlocked pipeline stages (MIPS).

The present application does not limit the specific form of the host 100. For example, when the computing system is deployed on a single computing device, the multiple hosts 100 can be deployed on the same computing device, and any host 100 can be a container or virtual machine deployed on the computing device, or a processor or processor core in the computing device. For another example, when the computing system is distributed and deployed on multiple computing devices, the multiple hosts 100 can be distributed on the multiple computing devices, and any host 100 can be a container or virtual machine deployed on a computing device, or a processor or processor core in a computing device, or a computing device.

The storage device 300 has a storage function and provides storage resources for the host 100. The storage device 300 can be used to store data required by the host 100. In the embodiment of the present application, the storage device 300 (or the logical device virtually formed by the storage device 300) can be used as the "memory" of the host 100 to temporarily store data required for the operation of the host 100. The storage device 300 (or the logical device virtually formed by the storage device 300) can also be used as the "hard disk" of the host 100, which can help the host 100 temporarily and persistently store data.

The embodiment of the present application does not limit the type of storage device 300. For example, the storage device 300 can be a memory in a computing device, such as a read-only memory (ROM), a dynamic random access memory (DRAM), or a storage class memory (SCM), a static random access memory (SRAM), a dual in-line memory module or a dual in-line memory module (DIMM). The storage device 300 can also be a persistent memory in a computing device, such as a hard disk drive (HDD) or a solid state drive (SSD).

The storage controller 200 is used to manage the storage device 300. The management of the storage device 300 by the storage controller 200 is manifested in the following aspects:

1) Virtualization of the storage device 300.

The storage controller 200 can virtualize the storage device 300 into multiple logical devices (logical devices, LD), each of which can be bound to a host 100. The logical device is similar to the storage device 300, has a data storage function, and can store data for the bound host 100. The host 100 can access the bound logical device through the storage controller 200. In the embodiment of the present application, the logical device bound to the host 100 is referred to as the logical device of the host 100. Any logical device can be located on the same storage device 300 or on different storage devices 300.

The embodiment of the present application does not limit the manner in which the storage controller 200 virtualizes the storage device 300 into multiple logical devices. A multi-logic device (MLD) component may be deployed on the storage controller 200, and the MLD component virtualizes the storage device 300 into multiple logical devices. The storage controller 200 may also use hardware-level virtualization technology to virtualize the storage device 300 into multiple logical devices and directly pass the logical devices to the host 100.

2) Maintenance of shared information of host 100.

In the embodiment of the present application, each host 100 is provided with sharing information. The sharing information of the host 100 records the shared storage space in the storage device 310 of the host 100, and the storage device 310 of the host 100 also records the mapping relationship between the shared storage space in the storage device 310 of the host 100 and the storage space in the storage device 310 of other hosts 100.

It is worth noting that in the embodiment of the present application, the storage device 310 of the host 100 has two possible forms. For example, the storage device 310 of the host 100 can be a physical storage device 300 in the computing system, or a logical device formed by virtualizing the storage device 300 in the computing system. Regardless of whether the storage device 310 of the host 100 is a physical storage device 300 or a logical device, the storage device 310 of the host 100 means that the storage space of the storage device is used by the host 100, and the host 100 can determine whether the storage space in the storage device 310 of the host 100 can be shared. Therefore, the storage space of the storage device 310 of the host 100 includes shared storage space and non-shared storage space.

The shared storage space in the storage device 310 of the host 100 refers to the storage space that is allowed to be shared with the host 100 other than the host 100. The shared storage space in the storage device 310 of the host 100 can be mapped to the storage space in the storage device 310 of the host 100 other than the host 100, and there is a mapping relationship between the shared storage space in the storage device 310 of the host 100 and the storage space in the storage device 310 of the host 100 other than the host 100. On the contrary, the non-shared storage space in the storage device 310 of the host 100 refers to the storage space that is not allowed to be shared with the host 100 other than the host 100. The non-shared storage space in the storage device 310 of the host 100 is not allowed to be mapped to the storage space in the storage device 310 of the host 100 other than the host 100, and there is no mapping relationship between the non-shared storage space in the storage device 310 of the host 100 and the storage space in the storage device 310 of the host 100 other than the host 100.

The shared storage space in the storage device 310 of the host 100 has two states, one state is a shared state, and the other state is a to-be-shared state.

The shared state means that the shared storage space in the storage device 310 of the host 100 has been shared with other hosts 100, and a mapping relationship has been established between the shared storage space in the storage device 310 of the host 100 and the storage space in the storage devices 310 of other hosts 100. The read operation of the host 100 on the shared storage space can be converted by the storage controller 200 into a read operation on the storage space in the storage device 310 of other hosts 100 that has a mapping relationship with the shared storage space. In other words, when the host 100 initiates a read request for the shared storage space (such as the address of the data carried in the read request belongs to the shared storage space), the storage management controller 200 receives the read request, can read data from the storage space in the storage device 310 of other hosts 100 that has a mapping relationship with the shared storage space, and feed the read data back to the host 100.

The state to be shared means that the shared storage space in the storage device 310 of the host 100 has not been shared with other hosts 100, and a mapping relationship has not been established between the shared storage space in the storage device 310 of the host 100 and the storage space in the storage devices 310 of other hosts 100. For the shared storage space in the state to be shared, the host 100 can initiate a data migration instruction for the shared storage resource, and the data migration instruction is used to request that the data in the storage space of the storage device 310 of other hosts 100 be migrated to the shared storage space (for example, the data migration instruction carries the address of the data in the storage device 310 of other hosts 100 and the address of the storage device 310 of the host 100, and the address of the storage device 310 of the host 100 belongs to the shared storage space to be shared). When the storage controller 200 receives the data migration instruction, the storage controller 200 does not need to perform actual data migration, but only needs to update the sharing information of the host 100 and establish a mapping relationship between the shared storage space in the storage device 310 of the host 100 and the storage space in the storage device 310 of other hosts 100, so that the shared storage space in the to-be-shared state is converted into the shared storage space in the shared state.

The storage controller 200 may maintain the shared information of each host 100, and update the shared storage space in the storage device 310 of the host 100 recorded in the shared information of the host 100, wherein the update of the shared information of the host 100 includes but is not limited to:

Increase the shared storage space in the storage device 310 of the host 100, delete the shared storage space in the storage device 310 of the host 100, configure the state of the shared storage space to be shared as a shared state (that is, establish a mapping relationship between the shared storage space in the storage device 310 of the host 100 and the storage space in the storage devices 310 of other hosts 100), configure the state of the shared shared storage space as a to-be-shared state (that is, delete the mapping relationship between the shared storage space in the storage device 310 of the host 100 and the storage space in the storage devices 310 of other hosts 100), and modify the mapping relationship between the shared storage space in the storage device 310 of the host 100 and the storage space in the storage devices 310 of other hosts 100.

①: Increase the shared storage space in the storage device 310 of the host 100.

For any host 100, which storage spaces in the storage device 310 of the host 100 are allowed to be shared and which storage spaces are not allowed to be shared can be determined by the host 100. When the host 100 needs to increase the shared storage space in the storage device 310 of the host 100, it can initiate a sharing instruction to the storage controller 200. The sharing instruction is used to request to increase the shared storage space. The sharing instruction carries information for indicating the shared storage space to be increased, such as the address of the shared storage space. After receiving the sharing instruction, the storage controller 200 can update the sharing information of the host 100 and add the shared storage space to the sharing information of the host 100, that is, add the information of the shared storage space to the sharing information of the host 100.

②: Delete the shared storage space in the storage device 310 of the host 100.

For any host 100, which storage spaces in the storage device 310 of the host 100 are allowed to be shared and which storage spaces are not allowed to be shared can be determined by the host 100. When the host 100 needs to prohibit the shared storage space recorded in the sharing information from being shared, it can initiate a prohibit sharing instruction to the storage controller 200. The prohibit sharing instruction is used to request the deletion of the shared storage space. The prohibit sharing instruction carries information for indicating the shared storage space to be deleted, such as the address of the shared storage space. After receiving the prohibit sharing instruction, the storage controller 200 can update the sharing information of the host 100, delete the shared storage space in the sharing information of the host 100, such as deleting the mapping relationship between the shared storage space in the storage device 310 of the host 100 and the storage space in the storage device 310 of other hosts 100 in the sharing information of the host 100, deleting the information of the shared storage space in the storage device 310 of the host 100, or marking the status of the shared storage space as prohibited from sharing.

In an embodiment of the present application, there is also a scenario of deleting a shared storage space in the storage device 310 of the host 100. When the host 100 initiates a write request to a shared storage space in a shared state, the storage management controller 200 receives the write request and can update the sharing information of the host 100 and delete the shared storage space in the sharing information of the host 100, such as deleting the mapping relationship between the shared storage space and the storage space in the storage device 310 of other hosts 100 in the sharing information of the host 100, or marking the status of the shared storage space as prohibited from sharing, and writing data in the shared storage space according to the write request.

③: Configure the status of the shared storage space to be shared as shared.

See the above content for the relevant description of the host 100 initiating the data migration instruction. In the process of the host 100 initiating the data migration instruction, the storage controller 200 can establish a mapping relationship between the shared storage space in the storage device 310 of the host 100 and the storage space in the storage device 310 of other hosts 100 in the sharing information, and convert the shared storage space in the to-be-shared state into the shared storage space in the shared state.

④: Set the shared storage space to a waiting-to-share state.

When the host 100 initiates a write request to a shared storage space in a shared state, the storage management controller 200 receives the write request and, in addition to deleting the shared storage space in the sharing information of the host 100, can also configure the state of the shared shared storage space to a to-be-shared state and delete the mapping relationship between the shared shared storage space and the storage space in the storage device 310 of the other host 100.

After the shared storage space in the storage device 310 of the host 100 has been shared with other hosts 100, the host 100 may decide not to share the shared storage space with the other hosts 100, and the host 100 may initiate a sharing change instruction to the storage controller 200, the sharing change instruction is used to request to release the sharing relationship between the shared storage space and other hosts 100, and the sharing change instruction carries information indicating the shared storage space, such as the address of the shared storage space. After receiving the sharing change instruction, the storage controller 200 may update the sharing information of the host 100, configure the state of the shared shared storage space to the waiting-to-be-shared state, and delete the mapping relationship between the shared shared storage space and the storage space in the storage device 310 of the other host 100.

⑤: Modify the mapping relationship between the shared storage space in the storage device 310 of the host 100 and the storage space in the storage device 310 of other hosts 100 .

When the host 100 initiates a data migration instruction request to migrate the data in the storage space of the storage device 310 of other hosts 100 to the shared storage space in a shared state (e.g., the data migration instruction carries the address of the data in the storage device 310 of other hosts 100 and address 1 of the storage device 310 of the host 100, and address 1 of the storage device 310 of the host 100 belongs to the shared shared storage space). After receiving the data migration instruction, the storage controller 200 does not need to perform real data migration, but only needs to update the sharing information of the host 100, modify the mapping relationship between the shared storage space in the storage device 310 of the host 100 and the storage space in the storage device 310 of other hosts 100 in the sharing information of the host 100, establish a mapping relationship between the storage space indicated by the address 1 in the shared storage space and the storage space where the data in the storage device 310 of other hosts 100 is located, and keep the original shared state for the storage space in the shared storage space except the storage space indicated by the address 1.

3) Read and write data to the storage device 310 of the host 100.

When the host 100 needs to access the storage device 310 of the host 100, a data access request may be initiated, such as a write request for requesting to write data, and a read request for requesting to read data. After receiving the data access request, the storage controller 200 may access the storage device 310 according to the data access request. For example, when the data access request is a write request, the storage controller 200 may write the data carried by the write request to the location indicated by the address of the data carried by the write request. For example, when the data access request is a read request, if the address of the data carried by the read request belongs to a shared storage space, and the shared storage space is in a shared state, the storage controller 200 may read data from the storage space of the storage device 310 of other hosts 100 that have a mapping relationship with the shared storage space, and feed it back to the host 100. If the address of the storage space carried by the read request is not the address of the shared storage space, the storage controller 200 accesses the storage device 310 of the host, reads data from the location indicated by the address of the storage space carried by the read request, and feeds the read data back to 100.

The embodiment of the present application does not limit the specific form of the storage controller 200. The storage controller 200 may be a software module, such as the storage controller 200 may be software allowed on a processor. The storage controller 200 may also be a hardware module, such as the storage controller 200 may be an ASIC, FPGA, AI chip, SoC or complex programmable logic device (CPLD, programmable logic device, PLD), etc.

The “mapping relationship between the shared storage space in the storage device 310 of the host 100 and the storage space in the storage devices 310 of other hosts 100” mentioned in the above content can be expressed as a mapping relationship between the address of the shared storage space in the storage device 310 of the host 100 and the address of the storage space in the storage devices 310 of other hosts 100.

The addresses mentioned in the embodiments of the present application (such as the address of the shared storage space, the address of the storage space, the address of the data, etc.) are related to two types of addresses in the present application. One type is the address required for the host 100 to interact with the storage controller 200. This type of address is called the host physical address (HPA). The address carried in the data access request (such as a write request or a read request) and various instructions (sharing instructions, data migration instructions, prohibition of sharing instructions, sharing change instructions, etc.) initiated by the host is the host physical address. The HPA is the address used by the host 100 side to represent the storage location of the data. The other type is the address required for the storage controller 200 to interact with the storage device 310300 (or storage device 310) is called the device physical address (DPA). The DPA is only the address used by the storage device 300 or storage device 310 side to represent the storage location of the data. Therefore, the DPA is an address visible to the storage device 300 or storage device 310 side, and is not visible to the host 100.

HPA needs to be converted to DPA. In the embodiment of the present application, the address conversion operation can be performed by the storage controller 200, and the storage controller 200 can obtain the host device address mapping table (host device mapping, HDM) of the host 100. The host device address mapping table of the host 100 records the mapping relationship between the HPA and DPA of the host 100. The storage controller 200 can convert the HPA to the DPA by querying the HDM.

For example, after the storage controller 200 receives a data access request (such as a write request or a read request) initiated by the host, it obtains the HPA carried in the data access request, queries the HDM to convert the HPA into a DPA, and then accesses the storage device 310 of the host 100 according to the converted DPA.

In view of these two different types of addresses, the shared information of the host 100 is represented in the following two forms.

Form 1: HPA is used to represent storage space in the shared information of the host 100 (the storage space refers to the shared storage space or the storage space that has a mapping relationship with the shared storage space). That is, the address of the shared storage space recorded in the shared information of the host 100 is HPA, and the address of the storage space that has a mapping relationship with the shared shared storage space is also HPA.

In this form, when the storage controller 200 receives a data access request (such as a write request or a read request) and various instructions (sharing instructions, data migration instructions, sharing prohibition instructions, sharing change instructions, etc.) initiated by the host, it can first query the sharing information of the host 100. If address conversion is required, the HDM is queried after the sharing information of the host 100 is queried (the scenario in which the HDM needs to be queried mainly refers to the scenario in which a data access request initiated by the host is received).

For a data access request initiated by a host, the storage controller 200 may query the sharing information of the host 100 to determine whether the HPA carried in the data access request is the address of the shared storage space. If the HPA carried in the data access request is not the address of the shared storage space, the storage controller 200 converts the HPA into the DPA of the storage device 310 of the host 100 (which may be referred to as the DPA of the host 100) according to the HDM, and then accesses the storage device 310 of the host 100 to write or read data at the DPA. If the HPA carried in the data access request is the address of the shared storage space, for a read request, the storage controller 200 will query the HPA of other hosts 100 that have a mapping relationship with the HPA, convert the HPA of other hosts 100 into the DPA of other hosts 100, and then access the storage device 310 of the other hosts 100 to read data at the DPA, and feed the read data back to the host 100. For a write request, the storage controller 200 updates the sharing information of the host 100, deletes the HPA recorded in the sharing information, or configures the state of the HPA to be shared, converts the HPA into the DPA of the host 100 according to the HDM, accesses the storage device 310 of the host 100, and writes data at the DPA.

For instructions initiated by the host (sharing instructions, data migration instructions, prohibit sharing instructions, sharing change instructions, etc.), the storage controller 200 can query the sharing information of the host 100 and update the sharing information according to the instructions initiated by the host. For sharing instructions, the storage controller 200 adds the HPA carried by the instruction to the sharing information of the host 100 and marks the HPA as a state to be shared. For data migration instructions, the storage controller 200 adds the mapping relationship between the HPA carried by the instruction and the HPA of other hosts 100 in the sharing information of the host 100. For prohibit sharing instructions, the storage controller 200 deletes the HPA carried by the instruction from the sharing information of the host 100 or marks the HPA as a state to be prohibited from sharing. For sharing change instructions, the storage controller 200 updates the sharing information of the host 100, configures the state of the shared HPA to a state to be shared, and deletes the mapping relationship between the shared HPA and the HPA of other hosts 100.

Form 2: The address of the storage space is represented by DPA in the shared information of the host 100 (the storage space refers to the shared storage space or the storage space that has a mapping relationship with the shared storage space). That is, the address of the shared storage space recorded in the shared information of the host 100 is DPA, and the address of the storage space that has a mapping relationship with the shared shared storage space is also DPA.

In this form, when the storage controller 200 receives a data access request (such as a write request or a read request) and various instructions (sharing instructions, data migration instructions, prohibit sharing instructions, sharing change instructions, etc.) initiated by the host, it needs to first perform address conversion, that is, query the HDM, and convert the HPA of the host 100 carried in the data access request and various instructions into the DPA of the host 100. After that, query the sharing information of the host 100 and perform corresponding operations.

For a data access request initiated by a host, the storage controller 200 converts the HPA of the host 100 carried in the data access request into the DPA of the host 100, and then queries the sharing information of the host 100 to determine whether the DPA carried in the data access request is the address of the shared storage space. If the DPA carried in the data access request is not the address of the shared storage space, the storage controller 200 accesses the storage device 310 of the host 100 and writes or reads data at the DPA. If the DPA carried in the data access request is the address of the shared storage space, for a read request, the storage controller 200 will query the DPA of the storage device 310 of other hosts 100 that have a mapping relationship with the DPA, access the storage device 310 of the other host 100, read data at the DPA, and feed the read data back to the host 100. For a write request, the storage controller 200 updates the sharing information of the host 100, deletes the DPA recorded in the sharing information, or configures the state of the DPA to be shared.

For instructions initiated by the host (sharing instructions, data migration instructions, sharing prohibition instructions, sharing change instructions, etc.), after the storage controller 200 converts the HPA of the host 100 carried in the instruction into the DPA of the host 100, it can query the sharing information of the host 100 and update the sharing information according to the instructions initiated by the host. For sharing instructions, the storage controller 200 adds the DPA carried by the instruction to the sharing information of the host 100 and marks the DPA as a state to be shared. For data migration instructions, the storage controller 200 adds the mapping relationship between the DPA carried by the instruction and the DPA of other hosts 100 in the sharing information of the host 100. For sharing prohibition instructions, the storage controller 200 deletes the DPA carried by the instruction in the sharing information of the host 100 or marks the DPA as a state to be shared. For sharing change instructions, the storage controller 200 updates the sharing information of the host 100, configures the state of the shared DPA to a state to be shared, and deletes the mapping relationship between the shared DPA and the DPA of other hosts 100.

A data sharing method provided in an embodiment of the present application is described below in conjunction with FIG2 . In FIG2 , a data sharing method between two hosts is taken as an example for description. For the convenience of description, the two hosts are respectively labeled as host A and host B, and the shared information of the hosts adopts the aforementioned form 1.

The data sharing method when the host's shared information adopts the aforementioned form 2 is similar to the data sharing method when the host's shared information adopts the aforementioned form 1. The difference lies in the order in which the storage controller queries the HDM and the host's shared information, which will not be repeated here.

The data sharing method shown in FIG2 includes three processes, one process is a data reading process, specifically, see step 201 to step 206, one process is a data migration process, specifically, see step 207 to step 210, and one process is a data writing process, specifically, see step 211 to step 214. These three processes are independent of each other, and in practical applications, these three processes can be executed independently. In the embodiment shown in FIG2, in order to reflect the relationship between the data reading process, data migration process, and data writing process executed for different shared storage spaces, these three processes are integrated together.

Step 201: the host 100A initiates a read request, the read request is used to request to read data 1, and the read request carries the HPA of the data 1. The HPA of the data 1 includes the starting host physical address of the data 1 and the size of the data 1.

Step 202: the storage controller 200 receives the read request, queries the sharing information of the host 100A, and determines whether the HPA of the data 1 is the address of the shared storage space. If so, execute step 203; otherwise, execute step 205.

As shown in FIG3 , a host 100A sharing information provided in an embodiment of the present application records the addresses of multiple shared storage spaces in the sharing information of the host 100A. The address of each shared storage space can be represented by the starting host physical address of the shared storage space and the size of the shared storage space. For the shared shared storage space, the host 100A sharing information also records the host 100 to which the storage space with which the shared storage space has a mapping relationship belongs, and the address of the storage space with which the shared storage space has a mapping relationship. Similarly, the address of the storage space with which the shared storage space has a mapping relationship can be represented by the starting host physical address of the storage space and the size of the storage space. The sharing relationship also includes a sharing mark, which indicates whether the storage space is allowed to be shared. For example, for a shared storage space, the sharing mark indicates that the storage space is allowed to be shared. Exemplarily, the sharing mark can also indicate that the shared storage space is in a state to be shared or in a shared state. For a non-shared storage space, or a storage space that is prohibited from sharing (such as a storage space indicated by a prohibition of sharing instruction), the sharing mark indicates that the storage space indicates that sharing is prohibited.

In FIG3 , the sharing information of the host 100A records three shared storage spaces. Assuming that the minimum granularity of data storage in the storage device 310 of the host 100A is 16 bytes, the starting host physical address of the first shared storage space is 1001, and the size is 64 bytes. The starting host physical address of the second shared storage space is 1005, and the size is 64 bytes. The difference between the starting host physical address of the first shared storage space and the starting host physical address of the second shared storage space is 4, and the size of the first shared storage space is just equal to 4 minimum granularities of data storage, that is, the first shared storage space and the second shared storage space are adjacent. The starting host physical address of the third shared storage space is 1010, and the size is 4 bytes. The sharing information of the host 100A also marks the status of the three shared storage spaces, wherein the status of the first shared storage space and the third shared storage space is shared, and the status of the second shared storage space is to be shared. The sharing information of the host 100A also records the addresses of the two storage spaces that have a mapping relationship with the first shared storage space and the third shared storage space. Among them, the storage space that has a mapping relationship with the first shared storage space belongs to the host 100B, and the starting host physical address of the storage space that has a mapping relationship with the first shared storage space is 2002. The storage space that has a mapping relationship with the third shared storage space belongs to the host 100B, and the starting host physical address of the storage space that has a mapping relationship with the third shared storage space is 2020. Since the size of the storage space that has a mapping relationship with the shared storage space is consistent with the size of the shared storage space, the size of the storage space that has a mapping relationship with the shared storage space is omitted in the sharing information of the host 100A shown in FIG. 3, and only the size of the corresponding shared storage space needs to be referred to.

If the HPA of the data 1 belongs to the first shared storage space or any of the third shared storage spaces, execute step 203; if the HPA of the data 1 does not belong to the first shared storage space or any of the third shared storage spaces, execute step 205.

It should be noted that only the host 100A and the host 100B are used as examples for explanation here. In actual applications, in the sharing information of the host 100A, the multiple storage spaces that have a mapping relationship with the multiple shared storage spaces may all belong to the host 100B, or may be distributed in multiple different hosts 100. In other words, the host 100A shares the storage space of the storage device 310 with multiple different hosts 100, but one shared storage space of the storage device 310 of the host 100A can only be shared with one host 100, that is, one shared storage space of the storage device 310 of the host 100A can only be mapped with the storage space of the storage device 310 of one host 100.

Step 203: The storage controller 200 determines the HPA of the host 100B that has a mapping relationship with the HPA of the data 1 according to the sharing information of the host 100A.

The HPA of the data 1 belongs to the shared storage space in the storage device 310 of the host 100A. Then, the storage space in the storage device 310 of the host 100B that has a mapping relationship with the shared storage space will include the HPA of the host 100B that has a mapping relationship with the HPA of the data 1. The position of the HPA of the data 1 in the shared storage space is consistent with the position of the HPA of the host 100B that has a mapping relationship with the HPA of the data 1 in the storage space.

Taking the shared information of the host 100A shown in FIG3 as an example, if the starting host physical address of the HPA of data 1 is 1001, the size of data 1 is 4 bits. The HPA of data 1 belongs to the first shared storage space, and the starting host physical address of the HPA of the host 100B that has a mapping relationship with the HPA of data 1 is 2002, and the size of data 1 is 4 bits. If the starting host physical address of the HPA of data 1 is 1002, the offset from the starting physical address of the first shared storage space is 16 bytes, and the size of data 1 is 4 bits. The HPA of data 1 still belongs to the first shared storage space, then the starting host physical address of the HPA of the host 100B that has a mapping relationship with the HPA of data 1 is 2003, and the offset from the starting physical address of the storage space that has a mapping relationship with the first shared storage space is also 16 bytes, and the size of data 1 is 4 bits.

Step 204: After obtaining the HPA of the host 100B, the storage controller 200 queries the HDM of the host 100B, converts the HPA of the host 100B into the DPA of the host 100B, accesses the storage device 310 of the host 100B, and reads data 1 from the DPA.

Step 205: The storage controller 200 queries the HDM of the host 100A, converts the HPA of the data 1 into the DPA of the host 100A, accesses the storage device 310 of the host 100A, and reads the data 1 from the DPA.

Step 206: The storage controller 200 feeds back the data 1 to the host 100A.

At this point, the data reading process ends. In the data reading process, it can be found that due to the setting of the shared information of the host 100A, the host 100A can access the storage device 310 of the host 100B and read the data in the storage device 310 of the host 100B. The host 100A does not perceive that the read data comes from the storage device 310 of the host 100B.

Step 207: Host 100A initiates a data migration instruction, which is used to request to migrate data 2 in the storage device 310 of host 100B to the target HPA of host 100A. The data migration instruction carries the HPA of data 2 in host 100B and the target HPA of host 100A.

Step 208: the storage controller 200 receives the data migration instruction, queries the sharing information of the host 100A, and determines whether the target HPA of the host 100A is the address of the shared storage space. If so, execute step 209; otherwise, execute step 210.

Still taking the sharing information of the host 100A shown in Fig. 3 as an example, if the target HPA of the host 100A belongs to one of the three shared storage spaces, then execute step 209. If the target HPA of the host 100A does not belong to any shared storage space, then execute step 210.

Step 209: The storage controller 200 establishes a mapping relationship between the target HPA of the host 100A and the HPA of the data 2 in the host 100B in the shared information of the host 100A.

If the shared storage space to which the target HPA in the storage device 310 of the host 100A belongs is the shared storage space to be shared, the storage controller 200 can directly add the mapping relationship between the target HPA of the host 100A and the HPA of the host 100B for data 2 in the sharing information of the host 100A. In some scenarios, the target HPA of the host 100A is only a part of the storage space of the shared storage space to be shared. In this scenario, the storage controller 200 adds the mapping relationship between the target HPA of the host 100A and the HPA of the host 100B for data 2 in the sharing information of the host 100A, and the remaining storage space of the shared storage space to be shared remains in the state of being shared.

For example, the target HPA of the host 100A belongs to the second shared storage space in the storage device 310 of the host 100A, and the starting host physical address of the target HPA of the host 100A is 1005, and the size is 16 bytes. The starting host physical address of the data 2 in the HPA of the host 100B is 2006, so the storage controller 200 adds the mapping relationship between the target HPA of the host 100A (that is, the starting host physical address is 1005, the size is 16 bytes) and the HPA of the data 2 in the host 100B (that is, the starting host physical address is 2006, the size is 16 bytes) in the sharing information of the host 100A, and the remaining storage space of the shared storage space to be shared (that is, the starting host physical address is 1006, the size is 48 bytes) remains in the state of being shared.

As shown in Figure 4, this is the updated sharing information of the host 100A. The updated sharing information of the host 100A divides the original second shared storage space into two shared storage spaces, one of which has a starting host physical address of 1005 and a size of 16 bytes. It is in a shared state and has a mapping relationship with the storage space in the storage device 310 of the host 100B (i.e., the starting host physical address is 2006 and the size is 16 bytes). The other shared storage space (with a starting host physical address of 1006 and a size of 48 bytes) is in a state to be shared.

If the shared storage space to which the target HPA of the host 100A belongs is a shared shared storage space, the storage controller 200 may modify the mapping relationship between the shared storage space and the storage space in the storage device 310 of the host 100B in the sharing information of the host 100A. The modified mapping relationship between the shared storage space and the storage space in the storage device 310 of the host 100B needs to include the mapping relationship between the target HPA of the host 100A and the data 2 in the HPA of the host 100B. In some scenarios, the target HPA of the host 100A is only a portion of the storage space of the shared shared storage space. In such a scenario, the storage controller 200 modifies the mapping relationship between the portion of the storage space and the storage space in the storage device 310 of the host 100B in the sharing information of the host 100A to the mapping relationship between the target HPA of the host 100A and the data 2 in the HPA of the host 100B, and the remaining storage space of the shared storage space remains in the original shared state.

For example, the target HPA of the host 100A belongs to the first shared storage space in the storage device 310 of the host 100A, and the starting host physical address of the target HPA of the host 100A is 1001, and the size is 16 bytes. The starting host physical address of the data 2 in the HPA of the host 100B is 2006, so the storage controller 200 modifies the mapping relationship between the first shared storage space of the shared information of the host 100A and the storage space of the host 100B, and adds the mapping relationship between the target HPA of the host 100A (that is, the starting host physical address is 1001, the size is 16 bytes) and the data 2 in the HPA of the host 100B (that is, the starting host physical address is 2006, the size is 16 bytes), and the remaining storage space of the shared storage space (that is, the starting host physical address is 1002, the size is 48 bytes) remains in the shared state.

As shown in Figure 5, this is the updated sharing information of the host 100A. The updated sharing information of the host 100A divides the original first shared storage space into two shared storage spaces, one of which has a starting host physical address of 1001, a size of 16 bytes, and is in a shared state. There is a mapping relationship between the shared storage space and the storage space in the storage device 310 of the host 100B (i.e., the starting host physical address is 2006, and the size is 16 bytes). The other shared storage space has a starting host physical address of 1002, a size of 48 bytes, and maintains the original mapping relationship between the shared storage space and the storage space in the storage device 310 of the host 100B (i.e., the starting host physical address is 2003, and the size is 48 bytes).

Step 210: the storage controller 200 accesses the storage device 310 of the host 100B, reads data 2 from the storage device 310 of the host 100B, and writes data 2 to the location indicated by the target HPA of the host 100A.

If the target HPA in the storage device 310 of the host 100A does not belong to the shared storage space, the storage controller 200 obtains the HDM of the host 100B, converts the data 2 in the HPA of the host 100B into the DPA of the data 2 in the host 100B, and reads the data 2 from the DPA. The storage controller 200 obtains the HDM of the host 100A, converts the target HPA of the host 100A into the DPA, and writes the data 2 to the DPA.

In some scenarios, after receiving the data migration instruction, the storage controller 200 may not execute step 208 and step 210, but directly execute step 209. That is, the storage controller 200 does not need to determine whether the target HPA in the storage device 310 of the host 100A is the address of the shared storage space. The storage controller 200 can directly establish a mapping relationship between the target HPA of the host 100A and the HPA of the data 2 in the host 100B in the shared information of the host 100A. This situation can be understood as the storage space of the storage device 310 of the host 100A is a shared storage space or the host 100A can clearly know that the target HPA belongs to the shared storage space (in this case, there is no branch of step 210).

In some scenarios, after receiving the data migration instruction, the storage controller 200 may also query the sharing information of the host 100B to determine whether the data 2 in the HPA of the host 100B belongs to the shared storage space. If so, the storage controller 200 may execute steps 208 to 210, or the storage controller 200 may not execute steps 208 and 210 and directly execute step 209. If not, the storage controller 200 may notify the host 100A that the data migration has failed.

The data migration process is now complete. It can be seen from the data migration process that, due to the setting of the shared information of the host 100A, when the data in the storage device 310 of the host 100B is migrated to the storage device 310 of the host 100A, it is not necessary to copy the data from the storage device 310 of the host 100B and then write the copied data to the storage device 310 of the host 100A. The effect of data migration can be achieved by simply updating the shared information of the host 100A and adding the corresponding mapping information, and the data migration method is more convenient.

After the data migration is performed, the host 100A may initiate a read request for requesting to read data 2, and the read request carries the target HPA of the host 100A. After receiving the read request, the storage controller 200 may query the shared information of the host 100A, and determine the HPA of the host 100B that has a mapping relationship with the target HPA of the host 100A; after determining the HPA of the host 100B, the HPA of the host 100B is converted into the DPA of the host 100B, and data 2 is read from the DPA and fed back to the host 100A. The process of the host 100A reading data 2 is similar to the process of the host 100A reading data 1, and each step will not be described in detail here. For details, please refer to the above content.

Step 211: the host 100A initiates a write request, where the write request is used to request writing of data 3, and the read request carries the HPA of the data 3.

Step 212: the storage controller 200 receives the write request, queries the sharing information of the host 100A, and determines whether the HPA of the data 3 is the address of the shared storage space. If so, execute step 213; otherwise, execute step 214.

Taking the sharing information of the host 100A shown in FIG5 as an example, if the HPA of the data 3 belongs to any shared storage space, step 213 and step 214 are executed. Otherwise, step 214 is executed.

Step 213: The storage controller 200 deletes the HPA of data 3 in the shared information of the host 100A.

If the host 100A's sharing information records the mapping relationship between the HPA of data 3 and the HPA of data 2 in the storage device 310 of the host 100B, that is, the HPA of data 3 is in a shared state, the storage controller 200 can delete the mapping relationship or mark the HPA of data 3 as prohibited from sharing. If the host 100A's sharing information records that the HPA of data 3 is in a state to be shared, the storage controller 200 can delete the HPA of data 3 or mark the HPA of data 3 as prohibited from sharing. The following is an explanation of different situations:

Case 1: When the shared storage space to which the HPA of data 3 belongs is a shared shared storage space.

The storage controller 200 may modify the mapping relationship between the shared storage space and the storage space in the storage device 310 of the host 100B in the sharing information of the host 100A. The modified mapping relationship between the shared storage space and the storage space in the storage device 310 of the host 100B no longer includes the mapping relationship between the HPA of the data 3 and the HPA of the data 2 in the host 100B. In some scenarios, the HPA of the data 3 is only a part of the storage space of the shared shared storage space. In such a scenario, the storage controller 200 deletes the mapping relationship between the part of the storage space and the storage space in the storage device 310 of the host 100B in the sharing information of the host 100A, and the remaining storage space of the shared storage space remains in the original shared state.

Taking the sharing information of the host 100A shown in FIG6 as an example, it is assumed that the HPA of data 3 belongs to a shared storage space in the storage device 310 of the host 100A, and the starting host physical address of the HPA of data 3 is 1002, and the size is 16 bytes. Then, the storage controller 200 modifies the mapping relationship between the shared storage space and the storage space of the storage device 310 of the host 100B in the sharing information of the host 100A, deletes the mapping relationship between the HPA of data 3 (that is, the starting host physical address is 1002, the size is 16 bytes) and the storage space in the storage device 310 of the host 100B (that is, the starting host physical address is 2003, the size is 16 bytes) or marks the HPA of data 3 as prohibited from sharing, and the remaining storage space of the shared storage space (that is, the starting host physical address is 1003, the size is 32 bytes) remains in the original shared state. The modified sharing information of the host 100A is shown in FIG6.

Case 2: the shared storage space to which the HPA of data 3 belongs is the shared storage space to be shared.

The storage controller 200 may delete the HPA of the data 3 in the sharing information of the host 100A or mark the HPA of the data 3 as prohibited from being shared. In some scenarios, the HPA of the data 3 is only a part of the storage space of the shared storage space to be shared. In such a scenario, the storage controller 200 deletes the HPA of the data 3 in the sharing information of the host 100A, and the remaining storage space of the shared storage space to be shared remains in the state of being shared.

Taking the sharing information of the host 100A shown in FIG4 as an example, it is assumed that the HPA of data 3 belongs to a shared storage space to be shared in the storage device 310 of the host 100A, and the starting host physical address of the HPA of data 3 is 1006, and the size is 16 bytes. Then, the storage controller 200 deletes the HPA of data 3 (that is, the starting host physical address is 1006, and the size is 16 bytes) in the sharing information of the host 100A, and the remaining storage space of the shared storage space to be shared (that is, the starting host physical address is 1007, and the size is 32 bytes) remains in the state of being shared. The modified sharing information of the host 100A is shown in FIG7.

Step 214: the storage controller 200 writes the data 3 to the HPA of the data 3 according to the write request.

The storage controller 200 obtains the HDM of the host 100A, converts the HPA of the data 3 into the DPA, and writes the data 2 into the DPA.

Based on the same inventive concept as the method embodiment, the embodiment of the present application also provides a data sharing device, which is used to execute the method executed by the storage controller in the above method embodiment. As shown in Figure 8, the data sharing device 800 includes a receiving module 801, a processing module 802, and a sending module 803. Specifically, in the API testing device, each module is connected through a communication path.

The receiving module 801 is used to receive a read request from a first host, where the read request is used to request to read first data from a first address on a storage device of the first host.

The processing module 802 is used to determine the mapping of the first address to the second address on the storage device of the second host according to the sharing information of the first host, the sharing information of the first host indicates the mapping relationship between the first address and the second address; and read the first data from the second address.

The sending module 803 is used to feed back the first data to the first host.

As a possible implementation, the receiving module 801 receives a data migration instruction from the first host, where the data migration instruction is used to request to migrate the first data from the second address to the first address.

The processing module 802 updates the sharing information of the first host, and the updated sharing information of the first host indicates a mapping relationship between the first address and the second address.

As a possible implementation, the receiving module 801 receives a write request from the first host, where the write request is used to request to write the second data at the first address.

The processing module 802 deletes the mapping relationship between the first address and the second address in the shared information of the first host; accesses the storage device of the first host, and writes the second data at the first address.

As a possible implementation, the processing module 802 queries the sharing information of the second host before updating the sharing information of the first host, and the sharing information of the second host indicates that the second address allows sharing.

As a possible implementation, the receiving module 801 receives a data migration instruction from the first host, the data migration instruction is used to request to migrate the first data from the second address to the first address. The processing module 802 determines that the sharing information of the second host indicates that the second address is prohibited from being shared; and the sending module 803 notifies the first host that the data migration fails.

As a possible implementation manner, the storage device of the first host and the storage device of the first host are logical devices virtually formed by physical storage devices, and may also be physical storage devices.

As a possible implementation, the first address and the second address are both host physical addresses HPA. When the processing module 802 reads the first data from the second address, it queries the host device address mapping table of the second host, converts the second address into the DPA of the second host, and reads the first data from the DPA, wherein the host device address mapping table of the second host describes the mapping relationship between the HPA and DPA of the second host.

As a possible implementation, the processing module 802 stores the shared information of the first host, which includes the HPA of the first host, the HPA of the second host that is mapped to the HPA of the first host, and the information of the second host.

As a possible implementation manner, the sharing information of the first host further includes a sharing mark, and the sharing mark is used to indicate whether the HPA of the first host allows sharing.

As a possible implementation, if the sharing information of the first host does not indicate the mapping relationship between the first address and the second address, the processing module 802 accesses the storage device of the first host, reads the first data from the first address, and feeds the first data back to the first host.

The division of modules in the embodiments of the present application is schematic and is only a logical function division. There may be other division methods in actual implementation. In addition, each functional module in each embodiment of the present application may be integrated into a processor, or may exist physically separately, or two or more modules may be integrated into one module. The above-mentioned integrated modules may be implemented in the form of hardware or in the form of software functional modules.

If the integrated module is implemented in the form of a software function module and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions to enable a terminal device (which can be a personal computer, a mobile phone, or a network device, etc.) or a processor (processor) to execute all or part of the steps of the method of each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk and other media that can store program code.

The present application also provides a computing device 900 as shown in FIG9 . The computing device 900 includes a bus 901, a processor 902, a communication interface 903, and a memory 904. The processor 902, the memory 904, and the communication interface 903 communicate with each other via the bus 901. Although not shown, the computing device 900 may also include a power supply circuit, which is used to supply power to the processor 902.

Among them, the processor 902 can be a CPU, ASIC, FPGA, AI chip, SoC or CPLD, GPU, DPU, etc.

The memory 904 may include a volatile memory, such as a random access memory (RAM). The memory 904 may also include a non-volatile memory, such as a ROM, a flash memory, a HDD or a SSD. The memory 904 may store computer program instructions.

The processor 902 can execute the executable code burned on the processor 902 to execute the method executed by the storage controller 200 in the embodiment shown in FIG. 2 .

The processor 902 may also call the computer program instructions stored in the memory 904 to execute the method executed by the storage controller 200 in the embodiment shown in FIG2 . The memory 904 may also include software modules required for other running processes such as an operating system (such as multiple modules in the data sharing device 800 ). The operating system may be LINUX ^TM , UNIX ^TM , WINDOWS ^TM , etc.

The present application also provides a computing device system, which includes at least one computing device 1000 as shown in FIG10. The computing device 1000 includes a bus 1001, a processor 1002, a communication interface 1003, and a memory 1004. The processor 1002, the memory 1004, and the communication interface 1003 communicate with each other through the bus 1001. At least one computing device 1000 in the computing device system communicates with each other through a communication path.

Among them, the types of processor 1002 and memory 1004 are similar to those of processor 902 and memory 904. For details, please refer to the above description and will not be repeated here.

The processor 1002 may execute executable codes burned on the processor 1002 to execute part or all of the method executed by the storage controller 200 in the embodiment shown in FIG. 2 .

The processor 1002 may also call the computer program instructions stored in the memory 1004 to execute part or all of the methods executed by the storage controller 200 in the embodiment shown in FIG2 . The memory 1004 may also include software modules required for other running processes such as an operating system (such as part or all of the multiple modules in the data sharing device 800). The operating system may be LINUX ^TM , UNIX ^TM , WINDOWS ^TM , etc.

At least one computing device 1000 in the computing device system establishes communication with each other via a communication network, and any one or multiple modules in the data sharing device 800 are run on each computing device 1000.

The descriptions of the processes corresponding to the above-mentioned figures have different emphases. For parts that are not described in detail in a certain process, please refer to the relevant descriptions of other processes.

In the above embodiments, all or part of the embodiments may be implemented by software, hardware, firmware or any combination thereof. When implemented by software, all or part of the embodiments may be implemented in the form of a computer program product. The computer program product includes computer program instructions, and when the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in FIG. 2 of the embodiment of the present invention are generated.

The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions may be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions may be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via a wired (e.g., coaxial cable, optical fiber, digital subscriber line, or wireless (e.g., infrared, wireless, microwave, etc.) method. The computer-readable storage medium may be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more available media integrated therein. The available medium may be a magnetic medium (e.g., a floppy disk, a hard disk, a magnetic tape), an optical medium (e.g., a DVD), or a semiconductor medium (e.g., an SSD).

Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the claims of the present application and their equivalents, the present application is also intended to include these modifications and variations.

Claims (22)

1. A data sharing method, characterized in that the method uses a computing system, the computing system includes a first host, a second host, a storage controller, a storage device of the first host, and a storage device of the second host, and the method includes: The storage controller receives a read request from the first host, wherein the read request is used to request to read the first data from a first address on a storage device of the first host; The storage controller determines, according to the sharing information of the first host, that the first address is mapped to a second address on the storage device of the second host, wherein the sharing information of the first host indicates a mapping relationship between the first address and the second address; The storage controller reads the first data from the second address and feeds the first data back to the first host. 2. The method according to claim 1, wherein before the storage controller receives the read request from the first host, the method further comprises: The storage controller receives a data migration instruction from the first host, where the data migration instruction is used to request to migrate the first data from the second address to the first address; The storage controller updates the sharing information of the first host, and the updated sharing information of the first host indicates a mapping relationship between the first address and the second address. 3. The method according to claim 1 or 2, characterized in that the method further comprises: The storage controller receives a write request from the first host, where the write request is used to request writing the second data at the first address; The storage controller deletes the mapping relationship between the first address and the second address in the sharing information of the first host; The storage controller accesses the storage device of the first host and writes the second data at the first address. 4. The method according to claim 2, wherein before the storage controller updates the sharing information of the first host, the storage controller further comprises: The storage controller queries the sharing information of the second host, and the sharing information of the second host indicates that the second address allows sharing. 5. The method according to claim 1, further comprising: The storage controller receives a data migration instruction from the first host, where the data migration instruction is used to request to migrate the first data from the second address to the first address; When determining that the sharing information of the second host indicates that the second address is prohibited from being shared, the storage controller notifies the first host that the data migration fails. 6. The method according to any one of claims 1 to 5, characterized in that the computing system is deployed in a computing device. 7. The method according to any one of claims 1 to 6, characterized in that the storage device of the first host and the storage device of the second host are logical devices virtually formed by physical storage devices. 8. The method according to any one of claims 1 to 6, wherein the first address and the second address are both host physical addresses HPA, and the storage controller reads the first data from the second address, comprising: The storage controller queries the host device address mapping table of the second host, converts the second address into the DPA of the second host, and reads the first data from the DPA, wherein the host device address mapping table of the second host describes the mapping relationship between the HPA and DPA of the second host. 9. The method according to any one of claims 1 to 8, characterized in that the storage controller stores sharing information of the first host, the sharing information of the first host includes the HPA of the first host, the HPA of the second host that is mapped to the HPA of the first host, and information of the second host. 10. The method according to claim 9, wherein the sharing information of the first host further includes a sharing tag, and the sharing tag is used to indicate whether the HPA of the first host allows sharing. 11. A data sharing device, characterized in that the device comprises: a receiving module, configured to receive a read request from the first host, wherein the read request is used to request to read the first data from a first address on a storage device of the first host; a processing module, configured to determine, according to the sharing information of the first host, that the first address is mapped to a second address on the storage device of the second host, wherein the sharing information of the first host indicates a mapping relationship between the first address and the second address; and read the first data from the second address; A sending module is used to feed back the first data to the first host. 12. The device according to claim 11, characterized in that The receiving module is further used to receive a data migration instruction from the first host, wherein the data migration instruction is used to request to migrate the first data from the second address to the first address; The processing module is further used to update the sharing information of the first host, and the updated sharing information of the first host indicates a mapping relationship between the first address and the second address. 13. The device according to claim 11 or 12, characterized in that The receiving module is further used to receive a write request from the first host, wherein the write request is used to request to write the second data at the first address; The processing module is further used to delete the mapping relationship between the first address and the second address in the shared information of the first host; access the storage device of the first host, and write the second data at the first address. 14. The device according to claim 12, wherein before updating the shared information of the first host, the processing module is further configured to: The sharing information of the second host is queried, and the sharing information of the second host indicates that the second address allows sharing. 15. The device according to claim 11, characterized in that The receiving module is further used to receive a data migration instruction from the first host, wherein the data migration instruction is used to request to migrate the first data from the second address to the first address; The processing module is further used to determine that the sharing information of the second host indicates that the second address is prohibited from being shared; The sending module is used to notify the first host that the data migration fails. 16. The device according to any one of claims 11 to 15, wherein the storage device of the first host and the storage device of the second host are logical devices virtually formed by physical storage devices. 17. The device according to any one of claims 11 to 16, wherein the first address and the second address are both host physical addresses HPA, and the processing module reads the first data from the second address, for: Query the host device address mapping table of the second host, convert the second address into the DPA of the second host, and read the first data from the DPA, wherein the host device address mapping table of the second host describes the mapping relationship between the HPA and the DPA of the second host. 18. The device according to any one of claims 11 to 17, characterized in that the storage controller stores sharing information of the first host, the sharing information of the first host includes the HPA of the first host, the HPA of the second host that is mapped to the HPA of the first host, and information of the second host. 19. The apparatus according to claim 18, wherein the sharing information of the first host further comprises a sharing mark, and the sharing mark is used to indicate whether the HPA of the first host allows sharing. 20. A computing device, characterized in that the computing device comprises a processor and a power supply circuit; The power supply circuit is used to supply power to the processor; The processor executes the method according to any one of claims 1 to 10. 21. A non-transitory computer-readable storage medium, characterized in that when the non-transitory computer-readable storage medium is executed by a computing device, the computing device executes the method according to any one of claims 1 to 10. 22. A computing device program product, characterized in that the computing device program product comprises computer instructions, and when executed by a computing device, the computing device performs the method according to any one of claims 1 to 10.