CN117971135A - Storage device access method and device, storage medium and electronic device - Google Patents

Abstract

The embodiment of the application provides a storage device access method, a storage device access device, a storage medium and an electronic device, wherein the method is applied to a storage expansion device, and the storage expansion device is connected between a first storage device and a server and comprises the following steps: receiving a data access request sent by a reference virtual machine; converting the first data access to a second data access by referring to the storage channel in response to the data access request; and controlling the reference storage space of the first storage device to execute the second data access. According to the application, the problem of low storage device expansion efficiency when the storage device connected with the server is expanded in the related technology is solved, and the effect of improving the storage device expansion efficiency when the storage device connected with the server is expanded is achieved.

Description

Storage device access method, device, storage medium and electronic device

Technical Field

The embodiments of the present application relate to the computer field, and in particular, to a storage device access method, apparatus, storage medium and electronic device.

Background technique

In the cloud service scenario of data centers, some applications with strong real-time performance and high IO access density in virtual machines or containers have high requirements for the bandwidth latency and other performance of storage devices. The performance of the centralized storage system deployed remotely cannot meet the requirements, and the local storage device of the server is required to provide support. Since the number of PCIe ports that the server CPU can provide is limited, the number of supported storage devices is also limited. Therefore, how to use limited local storage device resources to support the local access requirements of a large number of virtual machines is one of the problems faced by the current cloud scenario in improving storage performance. In order to solve the above problems, the PCIe Switch expansion solution is adopted in the relevant technology. In this solution, a dedicated Switch chip is used to expand the PCIe port of a CPU into multiple downstream ports to increase the number of storage devices carried by the server, and each downstream port is connected to a storage device. The advantage of this solution is that the local storage capacity of the server is increased in hardware, no additional software cooperation is required, the latency is low, and the versatility is strong. However, the expansion flexibility of this solution is insufficient, and the PCIe Switch model needs to be selected at the beginning of the server motherboard design, and the expansion capacity is fixed accordingly. If more channels are needed in the future, the only option is to replace the PCIe Switch chip with more ports, which requires redesigning the motherboard hardware. If the demand decreases, the additional PCIe channels will be idle, resulting in waste. In addition, expansion is achieved by connecting a storage device to the downstream port of each PCIe Switch, which has a high overall cost.

Summary of the invention

The embodiments of the present application provide a storage device access method, apparatus, storage medium and electronic device to at least solve the problem of low storage device expansion efficiency when expanding a storage device connected to a server in the related art.

According to an embodiment of the present application, a method for accessing a storage device is provided, which is applied to a storage expansion device, wherein the storage expansion device is connected between a first storage device and a server, one or more virtual machines are deployed on the server, one or more storage channels are established in the storage expansion device that are allowed to be identified as a second storage device by the server, the second storage device is assigned to the virtual machine deployed on the server, and the storage channel is assigned a corresponding storage space on the first storage device, the method comprising: receiving a data access request sent by a reference virtual machine, wherein the one or more virtual machines include the reference virtual machine, and the data access request is used to request a first data access to a reference storage device in the second storage device that is assigned to the reference virtual machine; in response to the data access request, converting the first data access into a second data access through a reference storage channel, wherein the reference storage channel is a storage channel identified as the reference storage device in the storage channels established in the storage expansion device, and the second data access is a data access to a reference storage space in the first storage device that corresponds to the reference storage channel; and controlling the reference storage space of the first storage device to perform the second data access.

The converting the first data access into the second data access by referring to the storage channel comprises:

determining a reference storage channel identified as the reference storage device from the storage channels established in the storage expansion device;

forwarding the first data access to the reference storage channel;

The first data access is converted into the second data access through the reference memory channel.

Optionally, a first virtual bridge device is further deployed in the storage expansion device, the first virtual bridge device is connected to the storage channel established in the storage expansion device, the first virtual bridge device is also used to connect to the server and is recognized by the server as being connected to the second storage device, and the first virtual bridge device records the second storage device and the storage channel having a corresponding relationship, wherein,

The step of determining the reference storage channel identified as the reference storage device from the storage channels established in the storage expansion device comprises: searching, through the first virtual bridge device, for the reference storage channel corresponding to the reference storage device from the second storage devices and storage channels having a corresponding relationship;

The forwarding the first data access to the reference storage channel includes: forwarding the first data access to the reference storage channel through the first virtual bridge device.

Optionally, searching, by the first virtual bridge device, for the reference storage channel corresponding to the reference storage device from second storage devices and storage channels having a corresponding relationship, includes:

extracting a reference device identifier of the reference storage device from the data access request through the first virtual bridge device;

The reference storage channel corresponding to the reference device identifier is searched for from storage device identifiers and storage channels having a corresponding relationship through the first virtual bridge device, wherein the storage device identifier is a device identifier assigned to a corresponding storage channel and allowed to be identified as a storage device.

Optionally, the method further includes:

Creating an initial virtual bridge device in the storage expansion device;

Performing a first attribute configuration of a transmission protocol on the initial virtual bridge device in the protocol configuration space of the initial virtual bridge device to obtain the first virtual bridge device, wherein the server and the storage expansion device are connected via a protocol bus that complies with the transmission protocol, and the first attribute is a protocol attribute that allows the server to identify the bridge device as complying with the transmission protocol;

A connection is established between the first virtual bridge device and the storage channel established in the storage expansion apparatus, and a second storage device and a storage channel having a corresponding relationship are recorded.

Optionally, the storage channel includes a second virtual bridge device and a virtual storage device, the first virtual bridge device is connected to the second virtual bridge device, the second virtual bridge device is connected to the virtual storage device, and the virtual storage device is used to connect to the first storage device, wherein,

The converting the first data access into the second data access through the reference storage channel includes:

receiving the first data access through the second virtual bridge device in the reference storage channel;

transmitting the first data access to the virtual storage device in the reference storage channel through the second virtual bridge device in the reference storage channel;

The first data access is converted into the second data access through the virtual storage device in the reference storage channel.

Optionally, converting the first data access into the second data access through the virtual storage device in the reference storage channel includes:

Acquire, through the virtual storage device in the reference storage channel, a first access address indicated by the first data access, wherein the first access address is a virtual address of the reference storage device indicated by the reference virtual machine to access;

querying, through the virtual storage device in the reference storage channel, a first storage address corresponding to the first access address from a virtual address range and a storage address range having a corresponding relationship, wherein the virtual address range is an address range of the reference storage device identified by the server, and the storage address range is an address range of a storage space allocated to the virtual storage device in the reference storage channel on the first storage device;

The first access address in the first data access is replaced with the first storage address to obtain the second data access.

Optionally, querying the first storage address corresponding to the first access address from a virtual address range and a storage address range having a corresponding relationship through the virtual storage device in the reference storage channel includes:

The first storage address mapped by the first access address is queried from the address mapping between the addressing space of the transmission protocol of the reference storage device and the storage space of the storage protocol of the first storage device through the virtual storage device in the reference storage channel, wherein the reference storage device is identified by the server as a device that complies with the transmission protocol, the virtual address range includes the addressing space of the transmission protocol, the first storage device is a storage device of the storage protocol, and the storage address range includes the storage space of the storage protocol.

Optionally, the method further includes:

Creating multiple groups of initial virtual bridging devices and initial virtual storage devices in the storage expansion device;

Performing a second attribute configuration of a transmission protocol on the initial virtual bridge device in the protocol configuration space of the initial virtual bridge device to obtain the second virtual bridge device, wherein the server and the storage expansion device are connected via a protocol bus that complies with the transmission protocol, and the second attribute is a protocol attribute that allows the server to identify the bridge device as complying with the transmission protocol;

The third attribute of the transmission protocol of the initial virtual storage device is configured in the transmission protocol configuration space of the initial virtual storage device, and the fourth attribute of the storage protocol of the initial virtual storage device is configured in the storage protocol configuration space of the initial virtual storage device, to obtain the virtual storage device, wherein the first storage device is a storage device that complies with the storage protocol, the third attribute is a protocol attribute that allows it to be recognized by the server as a terminal device that complies with the transmission protocol, and the fourth attribute is a protocol attribute that allows it to be recognized by the server as a storage device that complies with the storage protocol.

Optionally, a protocol parsing unit is deployed in the virtual storage device, and the configuring the third attribute of the transmission protocol of the initial virtual storage device in the transmission protocol configuration space of the initial virtual storage device includes:

The data link layer attributes and transaction layer attributes of the transmission protocol are configured for the initial virtual storage device in the transmission protocol configuration space of the initial virtual storage device, wherein the third attribute configuration includes: the data link layer attributes and the transaction layer attributes, and the protocol parsing unit is used to perform data link layer protocol parsing on the received data access using the data link layer attributes, and/or, perform transaction layer protocol parsing on the received data access using the transaction layer attributes.

Optionally, configuring a fourth attribute of the storage protocol of the initial virtual storage device in the storage protocol configuration space of the initial virtual storage device includes:

The storage capacity attribute and queue quantity attribute of the storage protocol are configured for the initial virtual storage device in the storage protocol configuration space of the initial virtual storage device, wherein the fourth attribute configuration includes: the storage capacity attribute and the queue quantity attribute, and the storage capacity attribute and the queue quantity attribute are used to be identified by the server as a storage device that complies with the storage protocol.

Optionally, a target bus and a bus interface are further deployed in the storage expansion device, and the bus interface is used to connect the plurality of storage channels connected to the target bus to the first storage device, and the bus interface has a mapping relationship between the bus address space of the target bus and the storage address space of the first storage device, wherein:

The controlling the reference storage space of the first storage device to perform the second data access comprises:

receiving the second data access via the target bus;

The second data access is forwarded to the reference storage space through the bus interface according to the mapping relationship.

Optionally, the bus interface includes: a bus arbitration unit and a storage device driver unit, wherein the storage device driver unit has a mapping relationship between a bus address space of the target bus and a storage address space of the first storage device, wherein:

The step of forwarding the second data access to the reference storage space according to the mapping relationship through the bus interface includes:

determining, by the bus arbitration unit, whether to process bus access of the reference storage channel according to the priority of the reference storage channel;

In the case of determining to process the bus access of the reference storage channel, the second data access is forwarded to the reference storage space according to the mapping relationship by the storage device driver unit.

According to another embodiment of the present application, a storage expansion device is provided, comprising: a server port, a processing chip and a storage device port, wherein the server port is connected to the processing chip, the processing chip is connected to the storage device port, one or more storage channels are established in the processing chip, the server port is used to connect to the server, wherein one or more virtual machines are deployed on the server, the one or more storage channels allow the server to identify them as one or more second storage devices, and the one or more second storage devices are allocated to the virtual machines deployed on the server; the storage device port is used to connect to a first storage device, wherein the storage channel is allocated corresponding storage space on the first storage device; the processing chip is used to implement the aforementioned storage device access step.

Optionally, a first virtual bridge device is deployed in the processing chip, the first virtual bridge device is connected to the one or more storage channels, the first virtual bridge device is also used to connect to the server port, and is recognized by the server as being connected to the second storage device, and the first virtual bridge device records the second storage device and the storage channel having a corresponding relationship, wherein,

The first virtual bridging device is used to search for a reference storage channel corresponding to a reference storage device from the second storage devices and storage channels having a corresponding relationship when receiving a first data access, wherein the reference storage device is a storage device among the one or more second storage devices that is assigned to a reference virtual machine that initiates a data access request, and the reference storage channel is a storage channel among the one or more storage channels that is identified as the reference storage device; and forward the first data access to the reference storage channel.

Optionally, the storage channel includes a second virtual bridge device and a virtual storage device, the second virtual bridge device is connected to the first virtual bridge device, the second virtual bridge device is connected to the virtual storage device, and the virtual storage device is connected to the first storage device, wherein,

The second virtual bridge device is configured to forward the first data access to the virtual storage device connected to the second virtual bridge device when receiving the first data access;

The virtual storage device is used to convert the first data access into a second data access for accessing the storage space corresponding to the reference storage channel in the first storage device.

Optionally, an external storage device is further deployed in the storage expansion device, and the external storage device is connected to the virtual storage device, wherein:

The external storage device is used to cache the data that the virtual storage device is to access in the first storage device.

Optionally, a target processor is also deployed in the processing chip, and the target processor is connected to the one or more storage channels through an internal high-speed bus, and the target processor is also connected to the storage device port, wherein:

The target processor is used to respond to the data access request transmitted by the storage channel and perform storage control and data access on the first storage device.

According to another embodiment of the present application, a storage device access device is provided, which is applied to a storage expansion device, wherein the storage expansion device is connected between a first storage device and a server, one or more virtual machines are deployed on the server, one or more storage channels are established in the storage expansion device that are allowed to be identified as a second storage device by the server, the second storage device is assigned to the virtual machine deployed on the server, and the storage channel is assigned a corresponding storage space on the first storage device, the device comprising: a receiving module, used to receive a data access request sent by a reference virtual machine, wherein the one or more virtual machines include the reference virtual machine, and the data access request is used to request a first data access to a reference storage device in the second storage device that is assigned to the reference virtual machine; a conversion module, used to respond to the data access request and convert the first data access into a second data access through a reference storage channel, wherein the reference storage channel is a storage channel identified as the reference storage device in the storage channels established in the storage expansion device, and the second data access is a data access to a reference storage space in the first storage device that corresponds to the reference storage channel; and a control module, used to control the reference storage space of the first storage device to perform the second data access.

According to another embodiment of the present application, a computer-readable storage medium is provided, in which a computer program is stored, wherein the computer program is configured to execute the steps of any of the above method embodiments when running.

According to another embodiment of the present application, an electronic device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor is configured to run the computer program to execute the steps in any one of the above method embodiments.

According to another embodiment of the present application, a computer program product is provided, including a computer program, characterized in that when the computer program is executed by a processor, the steps in any of the above method embodiments are implemented.

Through the present application, a storage expansion device is deployed between a server and a first storage device deployed on the server, and one or more storage channels established by the storage expansion device are allowed to be identified by the server as one or more second storage devices and allocated to the virtual machines deployed on the server for use, so that each virtual machine can be considered to be connected to a second storage device on the virtual machine side, and then, after the storage expansion device receives a data access request sent by the reference virtual machine on the server, the data access to the reference storage device is converted into data access to the storage space corresponding to the reference storage channel in the first storage device through the one or more storage channels identified as the reference storage device, so that only one storage device needs to be connected to provide services for multiple virtual machines on the server, and the expansion of the storage device connected to the service is realized without changing the server's storage device usage logic and without adding a new storage device, and for the server, each virtual machine is still provided with an independent storage device, so that the expansion performance of the storage device is higher, the versatility is stronger, and it is more flexible and reliable. It can solve the problem of low storage device expansion efficiency when the storage device connected to the server is expanded in the related art, and achieve the effect of improving the storage device expansion efficiency when the storage device connected to the server is expanded.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG1 is a schematic diagram of a method for accessing a storage device according to an embodiment of the present application;

FIG2 is a schematic diagram of a storage expansion device according to an embodiment of the present application;

FIG3 is a schematic diagram of an optional storage device space allocation according to an embodiment of the present application;

FIG4 is a schematic diagram of an optional bus interface according to an embodiment of the present application;

FIG5 is an optional storage device connection system architecture diagram according to an embodiment of the present application;

FIG6 is a flowchart of an optional storage expansion device operation according to an embodiment of the present application;

FIG7 is an optional storage device access flow chart according to an embodiment of the present application;

FIG8 is a structural diagram of an optional storage expansion device according to an embodiment of the present application;

FIG9 is a structural block diagram of an optional storage expansion device according to an embodiment of the present application;

FIG. 10 is a structural block diagram of an access device of a storage device according to an embodiment of the present application.

Detailed ways

The embodiments of the present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

It should be noted that the terms "first", "second", etc. in the specification and claims of this application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence.

In this embodiment, a method for accessing a storage device is provided. FIG. 1 is a schematic diagram of a method for accessing a storage device according to an embodiment of the present application. As shown in FIG. 1 , the method is applied to a storage expansion device, the storage expansion device is connected between a first storage device and a server, one or more virtual machines are deployed on the server, one or more storage channels that are allowed to be identified as a second storage device by the server are established in the storage expansion device, the second storage device is allocated to the virtual machine deployed on the server, and the storage channel is allocated corresponding storage space on the first storage device. The method comprises the following steps:

Step S102: receiving a data access request sent by a reference virtual machine, wherein the one or more virtual machines include the reference virtual machine, and the data access request is used to request a first data access to a reference storage device in the second storage device that is allocated to the reference virtual machine;

Step S104, in response to the data access request, converting the first data access into a second data access through a reference storage channel, wherein the reference storage channel is a storage channel identified as the reference storage device among the storage channels established in the storage expansion device, and the second data access is a data access to a reference storage space in the first storage device corresponding to the reference storage channel;

Step S106: controlling the reference storage space of the first storage device to perform the second data access.

Through the above steps, a storage expansion device is deployed between the server and the first storage device deployed on the server, and one or more storage channels established by the storage expansion device are allowed to be identified by the server as one or more second storage devices and allocated to the virtual machines deployed on the server for use, so that each virtual machine can be considered to be connected to a second storage device on the virtual machine side, and then, after the storage expansion device receives the data access request sent by the reference virtual machine on the server, the data access to the reference storage device is converted into the data access to the storage space corresponding to the reference storage channel in the first storage device through the one or more storage channels identified as the reference storage device, so that only one storage device needs to be connected to provide services for multiple virtual machines on the server, and the expansion of the storage device connected to the service is realized without changing the server's storage device usage logic and without adding new storage devices, and for the server, each virtual machine is still provided with an independent storage device, so that the expansion performance of the storage device is higher, the versatility is stronger, and it is more flexible and reliable. The problem of low storage device expansion efficiency when expanding the storage device connected to the server in the related art can be solved, and the effect of improving the storage device expansion efficiency when expanding the storage device connected to the server can be achieved.

In the technical solution provided in the above step S102, the storage device is used to provide data storage services for the virtual machines on the server. In the embodiment of the present application, the storage device may include but is not limited to hard disk drives, solid-state hard disks, optical disks, flash drives, tapes, etc., such as NVMe (Non-Volatile Memory express, non-volatile memory host controller interface specification) storage devices (such as NVMe SSD (Solid State Drives, solid-state hard disks)), which is not limited in this solution.

Optionally, in an embodiment of the present application, the storage channel established in the storage expansion device may be pre-established, or may be established based on the demand for the storage channel by a virtual machine deployed on the server, and this solution does not limit this.

Optionally, in an embodiment of the present application, one or more storage channels are established in the storage expansion device for accessing the corresponding storage space allocated on the first storage device, and then each storage channel is identified as a storage device on the server side. In the present application, the storage space on the first storage device can be allocated according to, but not limited to, the number of virtual machines deployed on the server and the storage space requirements of each virtual machine, and corresponding storage space is allocated to each virtual machine, and a storage space and a corresponding connection relationship are established. At the same time, the storage expansion device sends the index information of the storage channel (such as the address information of the storage channel, the communication protocol of the storage channel, the data transmission format, etc.) to the server, so that the server can use the index information to call the storage space corresponding to the storage channel.

Optionally, in an embodiment of the present application, a reference virtual machine may be allocated one or more storage channels in the storage expansion device. For example, N storage channels are allocated in the storage expansion device. The conventional allocation method may be to allocate a corresponding storage channel to each reference virtual machine. Alternatively, in order to achieve efficient access of the reference virtual machine to the storage device, M storage channels (where N is greater than or equal to M) may be selected for the reference virtual machine from the N storage channels based on the access requirements of the reference virtual machine to the storage device. The reference virtual machine may then simultaneously perform data access operations on the reference storage devices corresponding to the M reference storage channels by sending data access requests. This solution does not limit this.

Optionally, in an embodiment of the present application, a data access request may include, but is not limited to, an operation request to perform a data read operation on a reference storage device, or may also be a data write request to write data to a reference storage device, which is not limited in this solution.

In the embodiment provided in the above step S104, the reference storage channel is used to implement a one-to-one mapping between the reference storage device and the storage space corresponding to the reference storage channel in the first storage device. Since one or more storage channels in the storage expansion device in this embodiment are recognized by the server as being connected to one or more storage devices, the virtual machine on the server will send a data access request for performing a first data access to the reference storage device to the reference storage device. The storage expansion device, as a transit device connected between the first storage device and the server, converts the first data access into a second data access operation for accessing the storage space corresponding to the reference storage channel on the first storage device.

Optionally, in an embodiment of the present application, the reference storage channel converts the first data access into the second data access by converting the data access address, that is, the virtual machine is connected to the first storage device through the reference storage channel, and then the reference storage channel is regarded as a reference storage device on the virtual machine. The virtual machine needs to access data according to the virtual address of the reference storage device when accessing the storage device. Therefore, the first data access needs to carry the first access address for performing the data access operation. Since the reference storage channel side is connected to the first storage device, the reference storage channel needs to perform a data access operation on the storage area on the first storage device corresponding to the reference storage channel. Therefore, the reference storage channel needs to perform an address conversion operation on the first data access, and convert the access address from the first access address to the second access address of the storage area on the first storage device corresponding to the reference storage channel.

Optionally, in an embodiment of the present application, the reference storage channel converts the first data access into the second data access by accessing a data transmission protocol format, that is, the virtual machine is connected to the first storage device through the reference storage channel, and then the reference storage channel is regarded as a reference storage device on the virtual machine. The virtual machine needs to access data according to the virtual address of the reference storage device when accessing the storage device. Therefore, the first data access needs to be accessed according to the data transmission protocol format of the reference storage device. Since the data transmission protocol format of the reference storage device and the data transmission protocol format of the first storage device are stored differently, the reference storage channel needs to perform data transmission protocol format conversion on the first data access, thereby converting the first data access into the second data access.

Optionally, in the implementation of the present application, the reference storage space may be a continuous storage space corresponding to the reference storage channel in the first storage device, or may also be a plurality of non-continuous storage spaces. For example, the storage space of the first storage device may be divided into five consecutive storage spaces, namely storage space A, storage space B, storage space C, and storage space D, wherein the storage area of storage space A ends and is adjacent to the storage space start position of storage space B, the storage space end position of storage space B is adjacent to the storage space start position of storage space C, and the storage space end position of storage space C is adjacent to the storage space start position of storage space D. Thus, storage space A or storage space A and B may be used as storage space corresponding to a storage channel. When some of the storage spaces in the middle of the four storage spaces are allocated to a storage channel, such as storage space C is allocated to a storage device, then in order to ensure effective use of the storage space, when another storage channel needs two storage spaces, storage space A and storage space C may be allocated together to another storage channel.

In the embodiment provided in the above step S106, a mapping relationship between a reference storage channel and a reference storage space allocated to the reference storage channel in the first storage device is maintained in the storage expansion device, with the purpose of mapping the reference storage device to the reference storage space one-to-one, and then when it is necessary to perform a second data access operation, the reference storage space on the first storage device that has a mapping relationship with the reference channel is found through the mapping relationship.

The embodiment of the present application can be applied to but not limited to the expansion method of NVMe storage devices. The core idea is to expand the number of PCIe channels by adding an NVMe expansion device between the CPU and the NVMe storage device, and at the same time, virtualize an NVMe storage device under each channel, and complete the many-to-one mapping between the physical device, so that when only one NVMe storage device is connected downstream, multiple storage devices are presented to the CPU to achieve the purpose of device expansion. The present invention uses a hardware device to implement the NVMe storage device expansion function without excessive software participation, that is, it has the high performance and versatility of the PCIe Switch expansion solution, and has the flexibility and low cost of the software virtualization solution. The storage device expansion system includes a server, a storage expansion device, and a first storage device. The storage expansion device is connected between the first storage device and the server. One or more virtual machines are deployed on the server, which can present a topology form of multiple virtual NVMe devices (vDevice) mounted under a PCIe root node on the server CPU side, and can dynamically adjust the number of virtual devices expanded according to user needs. In this application, all virtual devices share one NVMe storage device, so the physical NVMe storage space addresses corresponding to multiple virtual devices cannot overlap, and a dedicated address mapping unit is responsible for space management.

As an optional embodiment, converting the first data access into the second data access by referring to the storage channel includes:

determining a reference storage channel identified as the reference storage device from the storage channels established in the storage expansion device;

forwarding the first data access to the reference storage channel;

The first data access is converted into the second data access through the reference memory channel.

Optionally, in an embodiment of the present application, a mapping relationship between virtual machines and storage channels may be maintained in the storage expansion device, but is not limited to, for mapping the virtual machines deployed on the server and the storage channels allocated to the virtual machines. Then, after receiving a data access request sent by a reference virtual machine, the storage expansion device can find out the reference storage channel allocated to the reference virtual machine according to the mapping relationship.

Through the above content, since one or more storage channels are established in the storage expansion device, the storage channels are used to connect the virtual machine and the storage space allocated for the storage channels on the first storage device, so that the server can identify one or more storage channels as one or more second storage devices, and then after receiving the data access request sent by the reference virtual machine, by determining the reference storage channel identified as the reference storage device by the reference virtual machine from one or more storage channels, the storage space on the first storage device that needs to perform the data access operation is found, and then by forwarding the first data access to the reference storage channel, the reference storage channel can convert the first data access that performs the data access operation on the reference storage device into the second data access that performs the data access operation on the reference storage space allocated for the reference storage device on the first storage device, thereby ensuring the accuracy of converting the first data access into the second data access.

As an optional embodiment, a first virtual bridge device is further deployed in the storage expansion device, the first virtual bridge device is connected to the storage channel established in the storage expansion device, the first virtual bridge device is also used to connect to the server and is recognized by the server as being connected to the one or more second storage devices, and the first virtual bridge device records the second storage devices and storage channels having a corresponding relationship, wherein,

The step of determining the reference storage channel identified as the reference storage device from the storage channels established in the storage expansion device comprises: searching, through the first virtual bridge device, for the reference storage channel corresponding to the reference storage device from the second storage devices and storage channels having a corresponding relationship;

The forwarding the first data access to the reference storage channel includes: forwarding the first data access to the reference storage channel through the first virtual bridge device.

Optionally, in an embodiment of the present application, the first virtual bridging device is used to realize the connection between the server and one or more storage channels, and perform protocol parsing on the data sent by the virtual machine or the data that needs to be returned to the virtual machine. In this embodiment, the first virtual bridging device can also perform routing forwarding according to the first pair of data packets in the data packet. For example, the data access request sent by the reference virtual machine carries the address information of the reference storage channel allocated to the reference virtual machine, and then the first virtual bridging device finds the reference storage channel that matches the address information from one or more storage channels according to the address information, and sends the data access request to the reference storage channel. Alternatively, after performing a data access operation on the storage space allocated to the reference storage channel in the first storage device, it is necessary to perform feedback on the data access result (for example, if the data access request is a request to read data, the read data needs to be fed back to the reference virtual machine). At this time, the reference storage channel sends the data access result to the first virtual bridging device, and the first virtual bridging device forwards the data access result sent by the reference storage channel to the reference virtual machine corresponding to the reference storage channel.

Optionally, in an embodiment of the present application, the first virtual bridging device may also maintain a mapping relationship between a virtual machine and a storage channel assigned to the virtual machine, and then implement data forwarding between the virtual machine and the storage channel through the first virtual bridging device, that is, the virtual machine sends a data access request to the first virtual bridging device, and the first virtual bridging device finds a reference storage channel assigned to the virtual machine in one or more storage channels according to the stored mapping relationship, and then forwards the data access request to the reference storage channel, or the storage channel sends a data packet read from the first storage device to the first virtual machine bridging device, and then finds a reference virtual machine corresponding to the storage channel in the virtual machine deployed on the first virtual machine bridging device server, and then forwards the data packet sent by the storage channel to the reference virtual machine.

Through the above content, by deploying a first virtual machine bridging device in the storage expansion device for linking a virtual machine and one or more storage channels established in the storage expansion device, since the first virtual machine bridging device is used to connect to the server and is recognized by the server as connected to one or more second storage devices, the second storage devices and storage channels with corresponding relationships are recorded on the first virtual machine bridging device, so that the reference storage channel corresponding to the reference storage device is searched through the corresponding relationship between the second storage device and the storage channel on the first virtual bridging device, thereby achieving accurate search of the storage channel and avoiding forwarding errors in the first data access.

As an optional embodiment, searching, by the first virtual bridge device, the reference storage channel corresponding to the reference storage device from the second storage devices and storage channels having a corresponding relationship, includes:

extracting a reference device identifier of the reference storage device from the data access request through the first virtual bridge device;

The reference storage channel corresponding to the reference device identifier is searched for from storage device identifiers and storage channels having a corresponding relationship through the first virtual bridge device, wherein the storage device identifier is a device identifier assigned to a corresponding storage channel and allowed to be identified as a storage device.

Optionally, in an embodiment of the present application, the first virtual bridging device is responsible for executing the transit connection between the server and the storage channel. On the server side, the server believes that it is connected to multiple second storage devices, and then generates a data access request to call the second storage device. The data access request carries a reference device identifier of the second storage device to be requested for access. The first virtual bridging device maintains the correspondence between the storage device identifier and the storage channel, and then determines the reference storage channel corresponding to the reference device identifier through the correspondence, and then sends the corresponding data access request to the reference storage channel, thereby calling the first storage device.

Through the above steps, by setting up a first virtual bridge device and using the virtual bridge device as a transit connection device between the server and multiple storage channels, since the multiple storage channels are identified as multiple second storage devices by the server, and then by maintaining the correspondence between the device identifier and the storage channel on the first virtual bridge device, the device identifier of the device requested to be called by the device and the storage channel are mapped, thereby achieving accurate search of the storage channel.

As an optional embodiment, the method further includes:

Creating an initial virtual bridge device in the storage expansion device;

Performing a first attribute configuration of a transmission protocol on the initial virtual bridge device in the protocol configuration space of the initial virtual bridge device to obtain the first virtual bridge device, wherein the server and the storage expansion device are connected via a protocol bus that complies with the transmission protocol, and the first attribute is a protocol attribute that allows the server to identify the bridge device as complying with the transmission protocol;

A connection is established between the first virtual bridge device and the one or more storage channels, and a second storage device and a storage channel having a corresponding relationship are recorded.

Optionally, in an embodiment of the present application, the first attribute configuration of the transmission protocol for the initial virtual bridge device may include but is not limited to register definitions such as PCIe configuration space, Memory base, Memory limit, etc., and through the configuration of the transmission protocol, it can be identified as a normal bridge device during the server CPU enumeration process.

Optionally, in an embodiment of the present application, the virtual bridging device may be, but is not limited to, a virtual PCIe bridging device, which is used to implement bridging between the PCIe buses of the server and the first storage device, or to implement connection between the server and multiple storage channels identified by the server as second storage devices, thereby performing protocol parsing on PCIe data packets from upstream and downstream links, and routing and forwarding upstream and downstream data packets according to the address information in the device. This solution does not limit this.

Through the above content, by configuring the first attribute of the transmission protocol for the initial virtual bridge device in the protocol configuration space of the initial virtual bridge device, the configured first virtual bridge device can be recognized by the server as a bridge device that complies with the transmission protocol, thereby realizing the normal identification of the first virtual bridge device on the CPU, and further providing protocol protection for the connection between the server and multiple storage channels.

As an optional embodiment, the storage channel includes a second virtual bridge device and a virtual storage device, the first virtual bridge device is connected to the second virtual bridge device, the second virtual bridge device is connected to the virtual storage device, and the virtual storage device is used to connect to the first storage device, wherein,

The converting the first data access into the second data access through the reference storage channel comprises:

receiving the first data access via the second virtual bridge device in the reference storage channel;

transmitting the first data access to the virtual storage device in the reference storage channel through the second virtual bridge device in the reference storage channel;

The first data access is converted into the second data access through the virtual storage device in the reference storage channel.

Optionally, in the embodiment of the present application, the virtual storage device has the logical functions of the device link layer and the transaction layer, and can perform functions such as unpacking, encapsulating, and verifying DLP and TLP data packets according to the protocol, and complete bus transactions such as memory read and write, configuration read and write, etc. Before the CPU enumerates the device tree, the channel expansion logic can prepare the necessary information such as the configuration space, address range, and number of devices of the virtual device as needed to cooperate with the upstream system call.

Optionally, in the embodiment of the present application, the core functional unit in the virtual device may include, but is not limited to, a protocol parsing unit, a data cache unit, and an address mapping unit, which are interconnected in the virtual device through a high-bandwidth bus; Protocol parsing unit: This unit completes the protocol parsing, packetization, verification, and data processing of the PCIe data link layer and the transaction layer. At the data link layer, an ACK/NAK response mechanism is used to perform error retransmission and flow control between the sender and the receiver, and DLLP message processing is performed according to the data link layer protocol to achieve reliable data transmission in the channel, and the sending order of different types of messages is defined to ensure that important information is responded to first. At the transaction layer, the unit uses the bus number, device number and function number ID in the TLP data packet to complete routing addressing, realizes the reading and writing of the configuration space content, and performs addressing according to the Address field in the TLP packet to realize the reading and writing of the memory space and IO space; the data cache unit is used to temporarily transmit the data transmitted between the server and the first storage device. When the virtual storage device accesses the PCIe upstream and downstream ports at a high frequency, the problem of untimely data packet processing will occur due to the mismatch of the upstream and downstream port response rates. The present invention sets a data cache unit inside the virtual storage device to deal with the problem of inconsistent data processing speed. The size of the data cache unit of each virtual storage device can be adjusted to adapt to different application scenarios; address mapping unit: in this application, the final data storage medium is the first storage device connected to the downstream port, not the virtual storage device "seen" by the server CPU. Each virtual storage device has a corresponding space in the first storage device, and the data reading and writing of the virtual storage device will eventually act on the space. Therefore, it is necessary to perform bidirectional data forwarding between the virtual storage device and the first storage device, and map the CPU's data reading and writing access to the virtual storage device to the first storage device.

Optionally, in an embodiment of the present application, the virtual storage device is used as a data transfer device between the server and the first storage device, that is, to temporarily store the data access request sent by the server, and perform data access operations on the storage area corresponding to the virtual storage device on the first storage device according to the data access request, and temporarily store the accessed data and pass it to the server. Therefore, in order to achieve this purpose, cache control can be allocated to the virtual device in the storage expansion device, that is, the storage space of the storage expansion device is allocated to multiple virtual storage devices, wherein the storage space of the storage expansion device can be evenly allocated to multiple virtual storage devices or can also be allocated according to the storage space requirements of the business to be handled by each virtual storage device, and this scheme does not limit this.

Through the above content, by setting up a second virtual bridge device and a virtual storage device in the storage channel, the first data access request is received through the second virtual bridge device, and the virtual storage device is used as a transit device for the first data access and the second data access, thereby achieving efficient and accurate access to the storage space of the connected first storage device through the storage channel.

In the embodiment of the present application, the virtual bridge device can be but not limited to a virtual PCIe bridge device, which is used to realize the bridge between two PCIe buses. FIG2 is a schematic diagram of a storage expansion device according to the embodiment of the present application, as shown in FIG2: the storage expansion device includes a channel expansion logic module and an internal bus interface module, the channel expansion logic module is used to connect to the server port (server or virtual machine deployed on the server), the channel expansion logic is also connected to the internal bus interface, the internal bus interface module is connected to the storage device port (first storage device), and the internal bus interface is used to realize the data access of the storage channel created in the storage expansion device to the first storage device. vBridge in the present invention is a virtual PCIe bridge device (including a first virtual bridge device and a second virtual bridge device) for realizing the bridge between two PCIe buses. In the present invention, the vBridge structure includes the PCIe configuration space, Memory base, Memory limit and other register definitions, which can be identified as a normal PCIe Bridge during the CPU enumeration process. In the present invention, vBridge is responsible for establishing a connection relationship between the CPU and the virtual PCIe Device (vDevice (virtual storage device)), performing protocol parsing on PCIe data packets from upstream and downstream links, and routing and forwarding upstream and downstream data packets according to the address information therein.

In this embodiment, the vDevice device (virtual storage device) is first a PCIe Endpoint device and further an NVMe storage device based on it. The vDevice device in the present invention has the logical functions of the link layer and transaction layer of the PCIe Endpoint device, and can perform functions such as unpacking, encapsulating, and verifying DLP and TLP data packets according to the protocol, and complete bus transactions such as memory reading and writing, configuration reading and writing, etc. Before the CPU enumerates the PCIe device tree, the channel extension logic can prepare the necessary information such as the configuration space, address range, number of devices, etc. of the vDevice as needed to cooperate with the upstream system call. The core functional units in the vDevice include a protocol parsing unit, a data cache unit, and an address mapping unit. These three parts are interconnected in the vDevice through a high-bandwidth bus. Details are as follows:

Protocol parsing unit: This unit completes the protocol parsing, packetization, verification and data processing of the PCIe data link layer and transaction layer. At the data link layer, the ACK/NAK response mechanism is used to perform error retransmission and flow control between the sender and the receiver. DLLP message processing is performed according to the data link layer protocol to achieve reliable data transmission in the channel, and the sending order of different types of messages is defined to ensure that important information receives priority response. At the transaction layer, the unit uses the bus number, device number and function number ID in the TLP data packet to complete routing addressing, realize the reading and writing of the configuration space content, and address according to the Address field in the TLP packet to realize the reading and writing of the memory space and IO space.

Based on PCIe, this unit also supports vDevice to have the logical functions of NVMe devices. By implementing the standard NVMe control register group, vDevice has the ability to obtain CQ and SQ queue information from the Host, respond to the Host's DoorBell command, and send MSI interrupts. For data writing and reading of NVMe storage space, this unit will forward data and commands between the upstream server CPU and the downstream NVMe device through the vDevice's data cache unit and address mapping unit, thereby completing the actual read and write operations.

Data cache unit: When vDevice accesses PCIe upstream and downstream ports at a high frequency, data packets may not be processed in time due to the mismatch in the response rates of the upstream and downstream ports. The present invention sets a data cache unit inside the vDevice to address the problem of inconsistent data processing speed. The size of the data cache unit of each vDevice can be adjusted to adapt to different application scenarios. The medium SRAM or DRAM of the data cache unit may be located inside the NVMe expansion device or outside the NVMe expansion device.

Address mapping unit: In the present invention, the final data storage medium is the NVMe storage device connected to the downstream port, rather than the vDevice "seen" by the server CPU. Each vDevice has a corresponding space in the NVMe storage device, and the data read and write of the vDevice will eventually act on this space. Therefore, it is necessary to perform two-way data forwarding between the vDevice and the NVMe storage device to map the CPU's data read and write access to the vDevice to the NVMe storage device. Figure 3 is a schematic diagram of an optional storage device space allocation according to an embodiment of the present application. As shown in Figure 3, the storage space of the first storage device is allocated according to the target granularity (which can be based on the number of virtual storage devices, or the demand for storage space of this virtual storage device), thereby obtaining multiple allocated storage spaces, and then the storage space can be allocated to the corresponding virtual storage device according to the data storage requirements of each virtual machine.

The above work in this application is completed by the address mapping unit in vDevice. The core idea is to map the PCIe BAR space of each vDevice to the part of the BAR space of the NVMe storage device allocated to the vDevice one-to-one. Whenever the vDevice receives a PCIe data packet, it forwards the data packet to the corresponding address of the NVMe storage device according to this mapping relationship. The reverse transmission is also the same. Every time the vDevice receives a data packet from the NVMe storage device, it forwards the data packet to the PCIe BAR space corresponding to the Host. The one-to-one mapping is implemented through a dedicated register group, which uses registers to store the first and last addresses of the PCIe BAR space on the vDevice side and the NVMe storage device side, respectively, and uses hardware logic to implement address matching and offset address.

As an optional embodiment, converting the first data access into the second data access through the virtual storage device in the reference storage channel includes:

Acquire, through the virtual storage device in the reference storage channel, a first access address indicated by the first data access, wherein the first access address is a virtual address of the reference storage device indicated by the reference virtual machine to access;

querying, through the virtual storage device in the reference storage channel, a first storage address corresponding to the first access address from a virtual address range and a storage address range having a corresponding relationship, wherein the virtual address range is an address range of the reference storage device identified by the server, and the storage address range is an address range of a storage space allocated to the virtual storage device in the reference storage channel on the first storage device;

The first access address in the first data access is replaced with the first storage address to obtain the second data access.

Optionally, in an embodiment of the present application, the virtual address range may be, but is not limited to, an address range of a storage space allocated on a network connection device for a corresponding storage channel for caching data.

Optionally, in an embodiment of the present application, in order to improve the utilization rate of the storage space in the first storage device, the storage space of the first storage device can be divided according to the target granularity, thereby dividing the storage space into multiple parts, and then allocating a corresponding storage address range to the virtual storage device according to the data storage capacity and the memory occupancy in the storage space, wherein the storage addresses in the storage address range can be continuous storage addresses or discontinuous storage addresses, that is, in order to improve the utilization rate of the storage space, the storage space corresponding to multiple unoccupied discontinuous storage address ranges in the storage space can be used for data storage, and this scheme does not limit this.

Through the above steps, a virtual address range is allocated to the reference virtual machine corresponding to the reference storage device in the storage space of the storage expansion device, so that the virtual storage device can temporarily store the data transmitted between the server and the first storage device to avoid the problem of inconsistent data delays, and a corresponding storage address range is allocated to the reference virtual machine on the first storage device, and then a virtual address range and a storage address range with a corresponding relationship are configured on the virtual storage device, and then the virtual storage device can replace the first address of the first data access with the first storage address when a data access conversion operation is required according to the mapping relationship, thereby improving the accuracy of data access, and this scheme does not limit this.

As an optional embodiment, querying the first storage address corresponding to the first access address from a virtual address range and a storage address range having a corresponding relationship through the virtual storage device in the reference storage channel includes:

The first storage address mapped by the first access address is queried from the address mapping between the addressing space of the transmission protocol of the reference storage device and the storage space of the storage protocol of the first storage device through the virtual storage device in the reference storage channel, wherein the reference storage device is identified by the server as a device that complies with the transmission protocol, the virtual address range includes the addressing space of the transmission protocol, the first storage device is a storage device of the storage protocol, and the storage address range includes the storage space of the storage protocol.

As an optional embodiment, the method further includes:

Creating multiple groups of initial virtual bridging devices and initial virtual storage devices in the storage expansion device;

Performing a second attribute configuration of a transmission protocol on the initial virtual bridge device in the protocol configuration space of the initial virtual bridge device to obtain the second virtual bridge device, wherein the server and the storage expansion device are connected via a protocol bus that complies with the transmission protocol, and the second attribute is a protocol attribute that allows the server to identify the bridge device as complying with the transmission protocol;

The third attribute of the transmission protocol of the initial virtual storage device is configured in the transmission protocol configuration space of the initial virtual storage device, and the fourth attribute of the storage protocol of the initial virtual storage device is configured in the storage protocol configuration space of the initial virtual storage device, to obtain the virtual storage device, wherein the first storage device is a storage device that complies with the storage protocol, the third attribute is a protocol attribute that allows it to be recognized by the server as a terminal device that complies with the transmission protocol, and the fourth attribute is a protocol attribute that allows it to be recognized by the server as a storage device that complies with the storage protocol.

Optionally, in an embodiment of the present application, data can be transmitted between the server and the storage expansion device through, but is not limited to, the PCIe protocol, that is, the server is a PCIe server, and the virtual bridge device is used to forward data packets between the PCIe server and the virtual storage device, and each second virtual bridge device is connected to a virtual storage device, so it is necessary to set standard PCIe properties in the PCIe configuration space, and information such as the bus number of the downstream device will be automatically determined when the CPU enumerates the PCIe device tree. This work is implemented by the system management unit of the internal bus interface.

Optionally, in the embodiment of the present application, the third attribute configuration may include but is not limited to PCIe attribute configurations such as DID/VID and BAR space.

Optionally, in an embodiment of the present application, the fourth attribute configuration may include but is not limited to NVMe attribute configurations such as storage capacity and number of queues, and private attribute configurations such as cache size and priority.

As an optional embodiment, a protocol parsing unit is deployed in the virtual storage device, and the third attribute configuration of the transmission protocol of the initial virtual storage device in the transmission protocol configuration space of the initial virtual storage device includes:

The data link layer attributes and transaction layer attributes of the transmission protocol are configured for the initial virtual storage device in the transmission protocol configuration space of the initial virtual storage device, wherein the third attribute configuration includes: the data link layer attributes and the transaction layer attributes, and the protocol parsing unit is used to perform data link layer protocol parsing on the received data access using the data link layer attributes, and/or, perform transaction layer protocol parsing on the received data access using the transaction layer attributes.

Optionally, in the embodiment of the present application, the data link layer attributes may include, but are not limited to, data transmission rate, encoding method, frame format, etc., while the transaction layer attributes include data transmission order, confirmation mechanism, error handling, etc. For example, at the data link layer, an ACK/NAK response mechanism is used to perform error retransmission and flow control between the sender and the receiver, DLLP message processing is performed according to the data link layer protocol to achieve reliable data transmission in the channel, and the sending order of different types of messages is defined to ensure that important information is responded to first.

Optionally, in an embodiment of the present application, at the data link layer, the bus number, device number and function number ID in the TLP data packet are used to complete routing addressing to realize reading and writing of the configuration space content, and addressing is performed according to the Address field in the TLP packet to realize reading and writing of the memory space and IO space.

Through the above content, by configuring the data link layer attributes and the transaction layer attributes, the transmission protocol of the initial virtual storage device can be effectively managed and controlled, thereby achieving reliable transmission and processing of data.

As an optional embodiment, configuring the fourth attribute of the storage protocol of the initial virtual storage device in the storage protocol configuration space of the initial virtual storage device includes:

The storage capacity attribute and queue quantity attribute of the storage protocol are configured for the initial virtual storage device in the storage protocol configuration space of the initial virtual storage device, wherein the fourth attribute configuration includes: the storage capacity attribute and the queue quantity attribute, and the storage capacity attribute and the queue quantity attribute are used to be identified by the server as a storage device that complies with the storage protocol.

Optionally, in an embodiment of the present application, the configuration of the storage capacity attribute includes but is not limited to setting the total capacity size of the virtual storage device and the capacity ratio allocated to different storage protocols.

Optionally, in the application embodiment, the configuration of the queue quantity attribute includes but is not limited to setting the read and write queue quantity, queue depth and other parameters of the storage protocol to optimize storage performance and improve concurrent access capabilities.

Through the above content, by configuring the storage capacity attributes and queue quantity attributes of the storage protocol, the virtual storage device can be customized with storage protocol configuration according to actual needs and application scenarios, thereby improving storage performance and flexibility and meeting the storage needs of different applications.

As an optional embodiment, a target bus and a bus interface are further deployed in the storage expansion device, and the bus interface is used to connect the multiple storage channels connected to the target bus to the first storage device, and the bus interface has a mapping relationship between the bus address space of the target bus and the storage address space of the first storage device, wherein:

The controlling the reference storage space of the first storage device to perform the second data access comprises:

receiving the second data access via the target bus;

The second data access is forwarded to the reference storage space through the bus interface according to the mapping relationship.

Optionally, in an embodiment of the present application, the bus interface is a high-bandwidth parallel bus interface connecting each virtual storage device and the storage device port, the bandwidth is much larger than the transmission bandwidth of PCIe, and can realize data transmission between each virtual storage device and the downstream port with very low latency.

As an optional embodiment, the bus interface includes: a bus arbitration unit and a storage device driver unit, wherein the storage device driver unit has a mapping relationship between a bus address space of the target bus and a storage address space of the first storage device, wherein:

The step of forwarding the second data access to the reference storage space according to the mapping relationship through the bus interface includes:

determining, by the bus arbitration unit, whether to process bus access of the reference storage channel according to the priority of the reference storage channel;

In the case of determining to process the bus access of the reference storage channel, the second data access is forwarded to the reference storage space according to the mapping relationship by the storage device driver unit.

Optionally, in an embodiment of the present application, for a bus arbitration unit, when multiple virtual storage devices access a storage device port at the same time, there may be a bus conflict problem. The bus arbitration unit in the present invention performs arbitration according to the different priority settings of each virtual storage device, and decides who will use the internal bus first when there is a bus access conflict to ensure the quality of service. A special priority field is set in the configuration space of each virtual storage device, such as an 8-bit register that can correspond to 256 priority levels, and the smaller the value, the higher the priority.

Optionally, in the embodiment of the present application, the internal bus interface has an NVMe device driver module (storage device driver unit) to interact with the controller of the NVMe storage device to complete the initialization of the NVMe storage device, including BAR space content reading and storage capacity acquisition, interrupt and DMA configuration, AER and other hardware attribute settings, configuration of CQ and SQ queues, etc. At the same time, in the process of presenting the NVMe storage device to multiple vDevices (virtual storage devices) through the internal bus interface, the NVMe driver can realize the mapping of the PCIe address space of the NVMe storage device to the internal bus address space for vDevice access.

In this embodiment, the bus interface is a high-bandwidth parallel bus interface that connects each virtual storage device to the storage device port, and the bandwidth is much greater than the transmission bandwidth of PCIe, and can realize data transmission between each virtual storage device and the downstream port with very low latency. Figure 4 is a schematic diagram of an optional bus interface according to an embodiment of the present application. As shown in Figure 4, the storage expansion device includes an expansion logic module and an internal bus interface module. The channel expansion logic module is used to connect to the server port (server or virtual machine deployed on the server), and the channel expansion logic is also connected to the internal bus interface. The internal bus interface module is connected to the storage device port (first storage device). The channel expansion logic module is used to create one or more storage channels on the storage expansion device that are allowed to be recognized as the second storage device by the server. The internal bus interface includes a bus arbitration unit, an NVMe driver unit (storage device driver unit) and a system management unit, and its functions are as follows:

Bus arbitration unit: When multiple vDevices access the storage device port at the same time, there will be a bus conflict problem. The bus arbitration unit in the present invention arbitrates according to the different priority settings of each vDevices, and decides who will use the internal bus first when there is a bus access conflict to ensure the quality of service. A special priority field is set in the configuration space of each vDevice, such as an 8-bit register that can correspond to 256 priority levels, and the smaller the value, the higher the priority.

NVMe driver unit: In the present invention, the internal bus interface is equipped with an NVMe device driver module, which is used to interact with the controller of the NVMe storage device to complete the initialization of the NVMe storage device, including BAR space content reading and storage capacity acquisition, interrupt and DMA configuration, AER and other hardware attribute settings, configuration of CQ and SQ queues, etc. At the same time, in the process of presenting the NVMe storage device to multiple vDevices through the internal bus interface, the NVMe driver can realize the mapping of the PCIe address space of the NVMe storage device to the internal bus address space for vDevice access.

System management unit: The system management unit is responsible for providing the device management interface of the present invention, realizing user management functions such as vDevice addition and removal, vDevice storage capacity configuration, vDevice priority management, and device and operation information reporting. Since vDevice is entirely implemented by internal logic units and does not involve the external interface of the device, it can be flexibly modified according to the application scenario. In addition to managing the internal bus interface, the system management unit also has the functions of receiving terminal user settings, modifying the parameter settings of the entire system including vBridge and vDevice, and handling system abnormal events to ensure the reliable operation of the system.

As an optional embodiment, after controlling the reference storage space of the first storage device to perform the second data access, the method further includes:

receiving access result information returned by the reference storage space in response to the second data access;

forwarding the access result information to the reference storage channel;

The access result information is sent to the server through the reference storage channel.

Optionally, in an embodiment of the present application, the access result information returned by the reference storage space in response to the second data access may include but is not limited to data read from the reference storage space, or may also be the write result returned after the data protocol to be written to the reference storage space is referenced by the storage space. This solution does not limit this.

Optionally, in an embodiment of the present application, the reference storage channel may include but is not limited to a mapping relationship between the second data access and the virtual machine that initiates the data access on the server, and then the reference storage channel finds the target virtual machine corresponding to the access result information according to the mapping relationship, and then feeds back the access result to the target virtual machine on the server.

Optionally, in an embodiment of the application, the storage expansion device may, but is not limited to, configure a corresponding data cache space for each reference storage channel for caching the data in the reference storage channel. Furthermore, when the data transmission link between the reference storage channel and the server is in an occupied state when receiving access result information, the access result information can be temporarily stored in the data cache space.

As an optional embodiment, forwarding the access result information to the reference storage channel includes:

Extracting the source address from the access result information to obtain the reference storage space;

Determining the reference storage channel corresponding to the reference storage space from the storage channels established in the storage expansion device;

The access result information is sent to the reference storage channel.

Optionally, in an embodiment of the present application, the reference storage channel may be, but is not limited to, determined by the storage expansion device from one or more storage channels according to a mapping relationship between the stored reference storage space and the storage channel, and this solution does not impose any limitation on this.

As an optional embodiment, sending the access result information to the server through the reference storage channel includes:

replacing the source address in the access result information from the reference storage space to the reference storage device to obtain reference access result information;

The reference access result information is sent to the server.

As an optional embodiment,

The method further comprises one of the following:

Determining the number of storage channels established in the storage expansion device according to resource information on the storage expansion device;

Determining the number of storage channels established in the storage expansion device according to the resource information on the storage expansion device and the device information of the first storage device;

The number of channels of the storage channels established in the storage expansion device is determined according to the resource information on the storage expansion device, the device information of the first storage device, and the virtual machine information of the virtual machine deployed on the server.

Optionally, in an embodiment of the present application, the number of storage channels in the storage expansion device can be flexibly configured according to demand.

Optionally, in an embodiment of the present application, the resource information on the storage expansion device may be, but is not limited to, the available space capacity of the storage resource, or the number of multiple candidate storage spaces obtained by dividing the storage resource according to the target granularity, each candidate storage space corresponding to a storage channel.

Optionally, in an embodiment of the present application, the virtual machine information may include, but is not limited to, the total number of virtual machines currently deployed on the server and the number of virtual machines currently storing data access requirements on the server, and this solution does not impose any limitation on this.

As an optional embodiment, determining the number of channels of the storage channels established in the storage expansion device according to the resource information on the storage expansion device and the device information of the first storage device includes:

detecting a connection with the first storage device;

When it is detected that the storage device is connected to the first storage device, loading a storage device driver;

Acquire the device information from the first storage device through the storage device driver;

The channel quantity of the one or more storage channels is determined according to the resource information on the storage expansion device and the device information of the first storage device.

The storage expansion device of the present application is composed of four parts: a server port, a channel expansion logic, an internal bus interface, and a storage device port. FIG5 is an optional storage device connection system architecture diagram according to an embodiment of the present application. As shown in FIG5, the storage expansion device includes four parts: a server port, a channel expansion logic, an internal bus interface, and a PCIe downstream port. Its functions are as follows:

1. Server Port:

The server port is the hardware interface between the NVMe expansion device and the server CPU described in the present invention, and is responsible for establishing a physical connection with the PCIe Root Complex of the CPU. The server port contains the electrical and logical properties of the PCIe physical layer, can complete the transmission and power supply of transceiver, clock, reset, and other signals, implement serial-to-parallel conversion, link training and other functions, and can be in the form of connectors, gold fingers, cables, etc.

2. Storage device port:

The storage device port is the hardware interface between the NVMe expansion device and the NVMe storage device described in the present invention, and is responsible for establishing a connection with the NVMe storage device. Like the server port, the storage device port contains the electrical and logical properties of the physical layer in the PCIe protocol, can transmit and supply power to the transceiver, clock, reset, and other signals, and complete serial-to-parallel conversion, link training and other functions, and can be in the form of a connector, gold finger, cable, etc.

3. Channel expansion logic:

The channel extension logic is an important part of the present invention, which is responsible for dynamically creating multiple virtual PCIe devices (vEP) inside the NVMe extension device, which are mainly composed of virtual PCIe Bridge (vBridge) and virtual PCIe devices (vDevice). The channel extension logic is connected to the storage device port through an internal bus. For the server, these vDevices have the same properties as other physical PCIe devices, and the virtual machine management software (VMM) can easily allocate them to each virtual machine (VM) for use. For the downstream NVMe storage device, the NVMe extension device is embodied as an NVMe master device, which can perform operations such as reading, writing and erasing on the NVMe storage device. The channel extension logic is internally divided into upstream and downstream channels. The data direction of the downstream channel is from the CPU to the NVMe device, and the data direction of the upstream channel is from the NVMe device to the CPU.

FIG. 6 is a flowchart of an optional storage expansion device operation according to an embodiment of the present application. As shown in FIG. 6 , the operation flow chart at least includes the following steps:

S601, internal bus interface configuration. Enumerate and manage NVMe storage devices (first storage device) through downstream PCIe ports, load drivers, and obtain device information, such as PCIe bit width and speed level, storage capacity, etc., to provide information for subsequent vDevice quantity and capacity parameter configuration.

S602, vDevice (virtual storage device) configuration. The content includes PCIe attribute configuration such as DID/VID and BAR space, NVMe attribute configuration such as storage capacity and queue number, and private attribute configuration such as cache size and priority. This work is implemented by the system management unit of the internal bus interface.

S603, vBridge (virtual bridge device, including the first virtual bridge device and the second virtual bridge device) configuration. In the present invention, vBridge is used to forward data packets between PCIe RC and vDevice, and each vBridge is only connected to one vDevice downstream device, so standard PCIe attribute settings need to be performed in the PCIe configuration space, and the bus number and other information of the downstream device will be automatically determined when the CPU enumerates the PCIe device tree. This work is implemented by the system management unit of the internal bus interface.

S604, start the server OS, enumerate the PCIe bus devices, and successfully enumerate N Endpoint (vDevice) and N+1 Bridge (vBridge) devices.

S605, the server loads the NVMe standard driver, further identifies N NVMe devices, and allocates the N NVMe devices to multiple virtual machines under the action of a virtual machine manager (VMM).

S606, the application on each virtual machine accesses the storage device, and finally accesses the physical NVMe storage device through the PCIe upstream port <-> VBridge <-> VDevice <-> internal bus interface <-> PCIe downstream port <-> NVMe storage device data channel.

FIG. 7 is an optional storage device access flow chart according to an embodiment of the present application. As shown in FIG. 7 , the flow chart at least includes the following steps:

S701, the channel extension logic initializes the configuration of vDevice (virtual storage device) and vBridge (virtual bridge device, including the first virtual bridge device and the second virtual bridge device). The CPU enumerates the PCIe device tree, identifies N virtual vDevice devices, and allocates them to each virtual machine by the VMM.

S702, a virtual machine initiates a read/write request to the vDevice to which it belongs. The PCIe RC of the CPU sends the request to the upstream port in the present invention according to the accessed address range.

S703, after being imported by the upstream port, the data packet is parsed by the vBridge protocol and forwarded to the corresponding vDevice according to the PCIe ID number or address space range.

S704, PICe and NVMe protocols are parsed in the vDevice in turn, and then the address access to the vDevice is converted into the access to the internal bus address through the address mapping link.

S705, after the vDevice obtains the right to use the bus through bus arbitration, it converts the internal bus address access into a read/write access to the downstream physical NVMe storage device with the assistance of the NVMe driver, and sends it to the PCIe downstream port.

S706, the access request is sent to the NVMe storage device via the PCIe downstream port and processed. The returned data is transmitted back to the CPU in reverse order through the above links.

In this embodiment, by introducing the NVMe storage device expansion device of the present invention, an ordinary NVMe SSD is equipped with a hardware-supported class virtualization function, and the key links such as protocol parsing, device management and address mapping are implemented by the channel expansion logic in the FPGA, thereby solving the compatibility problems caused by the complex configuration process of traditional virtualization technologies such as SR-IOV.

The embodiments of the present application can be applied to, but not limited to, the expansion of NVMe storage devices. By adding an NVMe channel expansion device in the link, a single NVMe storage device can be presented to the CPU as multiple virtual NVMe devices. Different from the existing software virtualization solutions and hardware PCIe Switch expansion solutions, the present invention has both the high efficiency of hardware expansion solutions and the high flexibility of software virtualization solutions.

The NVMe channel expansion device includes a PCIe upstream port, a PCIe downstream port, a channel expansion logic and an internal bus interface unit.

The PCIe upstream port is the hardware interface between the device and the server, responsible for establishing a physical connection with the PCIe RC of the server CPU;

The PCIe downstream port is the hardware interface between this device and the NVMe storage device, and is responsible for establishing a physical connection with the NVMe storage device.

The channel extension logic creates multiple PCIe bridge devices (vBridge) and Endpoint devices (vDevice) to complete the protocol parsing and forwarding of the respective PCIe link layer and transaction layer, thereby realizing the expansion function of the PCIe channel.

The internal bus interface unit completes the address mapping between the PCIe addressing space and the NVMe storage space, and implements multi-channel access between multiple extended PCIe devices (vDevice) and NVMe storage devices.

The PCIe upstream port and PCIe downstream port have the electrical and logical functions of the PCIe physical layer, can transmit and supply power to transceivers, clocks, resets, and other signals, complete serial-to-parallel conversion, link training and other functions, and can be in the form of connectors, gold fingers, cables, etc.

The vBridge and vDevice described can implement protocol parsing and data forwarding of the PCIe data link layer and transaction layer, and do not implement non-core links such as the physical layer, consistency testing, and power management, thereby saving hardware and logic resources and reducing overall link latency.

The number of the vBridges and vDevice can be flexibly adjusted by the system management unit in the present invention.

The internal bus interface unit is composed of three units: bus president, NVMe driver, and system management.

The bus arbitration unit implements a priority-based bus arbitration mechanism when multiple vDevices communicate with one NVMe storage device.

The NVMe driver unit implements address conversion and access control between the internal bus space and the NVMe storage space.

The system management unit implements management functions such as channel expansion capability configuration, operation information reporting, and exception handling.

In this embodiment, a storage expansion device is also provided. FIG. 8 is a structural diagram of an optional storage expansion device according to an embodiment of the present application. As shown in FIG. 8 , the storage expansion device includes:

A server port, a processing chip and a storage device port, wherein the server port is connected to the processing chip, the processing chip is connected to the storage device port, and one or more storage channels are established in the processing chip.

The server port is used to connect to a server, wherein one or more virtual machines are deployed on the server, the one or more storage channels are allowed to be identified by the server as one or more second storage devices, and the one or more second storage devices are allocated on the server to the virtual machines deployed on the server;

The storage device port is used to connect to a first storage device, wherein the storage channel is allocated a corresponding storage space on the first storage device;

The processing chip is used to implement the steps of the aforementioned storage device access method.

Through the above content, a storage expansion device is deployed between the server and the first storage device deployed on the server, and one or more storage channels established by the storage expansion device are allowed to be identified by the server as one or more second storage devices and allocated to the virtual machines deployed on the server for use, so that each virtual machine can be considered to be connected to a second storage device on the virtual machine side, and then, after the storage expansion device receives the data access request sent by the reference virtual machine on the server, the data access to the reference storage device is converted into the data access to the storage space corresponding to the reference storage channel in the first storage device through the one or more storage channels identified as the reference storage device, so that only one storage device needs to be connected to provide services for multiple virtual machines on the server, and the expansion of the storage device connected to the service is realized without changing the server's storage device usage logic and without adding new storage devices, and for the server, each virtual machine is still provided with an independent storage device, so that the expansion performance of the storage device is higher, the versatility is stronger, and it is more flexible and reliable. It can solve the problem of low storage device expansion efficiency when expanding the storage device connected to the server in the related art, and achieve the effect of improving the storage device expansion efficiency when expanding the storage device connected to the server.

As an optional embodiment, a first virtual bridge device is deployed in the processing chip, the first virtual bridge device is connected to the one or more storage channels, the first virtual bridge device is also used to connect to the server port, and is recognized by the server as being connected to the second storage device, and the first virtual bridge device records the second storage device and the storage channel having a corresponding relationship, wherein,

The first virtual bridging device is used to search for a reference storage channel corresponding to a reference storage device from the second storage devices and storage channels having a corresponding relationship when receiving a first data access, wherein the reference storage device is a storage device among the one or more second storage devices that is assigned to a reference virtual machine that initiates a data access request, and the reference storage channel is a storage channel among the one or more storage channels that is identified as the reference storage device; and forward the first data access to the reference storage channel.

As an optional embodiment, the storage channel includes a second virtual bridge device and a virtual storage device, the second virtual bridge device is connected to the first virtual bridge device, the second virtual bridge device is connected to the virtual storage device, and the virtual storage device is connected to the first storage device, wherein,

The second virtual bridge device is configured to forward the first data access to the virtual storage device connected to the second virtual bridge device when receiving the first data access;

The virtual storage device is used to convert the first data access into a second data access for accessing the storage space corresponding to the reference storage channel in the first storage device.

Optionally, in an embodiment of the present application, data storage resources are deployed in the processing chip for temporarily storing data that the virtual storage device needs to access in the first storage device. For one or more storage channels established in the storage expansion device, the processing chip needs to allocate data storage resources based on the number of storage channels created. When allocating, the data storage resources can be allocated according to the resource requirements of each storage channel, or the data storage resources can be evenly divided according to the number of storage channels created. This solution does not limit this.

As an optional embodiment, an external storage device is further deployed in the storage expansion device, and the external storage device is connected to the virtual storage device, wherein:

The external storage device is used to cache the data that the virtual storage device is to access in the first storage device.

Optionally, in an embodiment of the present application, an external storage device is used to provide data storage services for storage channels when the data storage resources of the processing chip are insufficient. An external storage device can be configured for each storage channel, or multiple storage channels can be connected to a common external storage device and the external storage device is used to provide data storage services for all storage channels.

As an optional embodiment, a target processor is also deployed in the processing chip, and the target processor is connected to the one or more storage channels through an internal high-speed bus, and the target processor is also connected to the storage device port, wherein:

The target processor is used to respond to the data access request transmitted by the storage channel and perform storage control and data access on the first storage device.

Optionally, in an embodiment of the present application, the storage expansion device may be, but is not limited to, configured with a storage device connected to the target processor, such as a DDR memory, running a Linux operating system with the support of DDR, and implementing management and access to the NVMe storage device under the driver of a standard NVMe.

The processing chip of the present application can, but is not limited to, realize the virtualization of NVMe storage devices through an FPGA accelerator in the form of a PCIe Add-in card. It includes two PCIe ports. In this embodiment, one is configured as a PCIe EndPoint interface (server port) to connect the server CPU with a gold finger; the other PCIe port is configured as a PCIe RC interface (storage device port) to connect the downstream NVMe storage device through an MCIO connector. The PCIe channel expansion logic in the present invention is implemented in the IP of the FPGA (processing chip), and the management and access to the NVMe storage device are realized through the NVMe driver IP. Figure 9 is a structural block diagram of an optional storage expansion device according to an embodiment of the present application. As shown in Figure 9, the PCIe upstream port (server port) is in the form of a PCIe gold finger, which is used to insert the PCIe slot of the server and the PCIeRC interface of the connector CPU. The PCIe downstream port (storage device port) is a high-speed connector in the form of an MCIO, which can be connected to an NVMe SSD via a cable. In the channel expansion logic, according to the FPGA resources, a total of 16 vBridges (second virtual bridge devices) and corresponding vDevices are set up. They share an 8TB NVMe storage device (first storage device) connected to the downstream port, and each vDevice (virtual storage device) is allocated an average of 512GB. With the support of the virtual machine manager (VMM), 16 vDevices (virtual storage devices) are allocated to 16 different virtual machines to store user data that changes frequently and has relatively low security requirements. In addition to the logic unit (PL side), the FPGA also contains a processor unit (PS side) with a quad-core ARMA53 processor as the core. The PS side runs the Linux operating system with the support of external DDR and manages and accesses the NVMe storage device under the standard NVMe driver. The PL side is also configured with 8GB of external DDR, which can be used as a data cache for vDevice when the SRAM resources inside the FPGA are insufficient. The system management function is implemented through the SMBUS bus. Through the SMBUS bus, storage virtualization devices can receive out-of-band instructions from the server to implement system performance settings such as the number of virtual machines, vDevice priority settings, data cache adjustments, virtual capacity allocation, etc. It can also read device information such as power consumption and temperature of the entire device, as well as perform operations such as system reset and power on and off.

Through the description of the above implementation methods, those skilled in the art can clearly understand that the method according to the above embodiment can be implemented by means of software plus a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium (such as ROM/RAM, disk, CD), and includes a number of instructions for a terminal device (which can be a mobile phone, computer, server, or network device, etc.) to execute the methods described in each embodiment of the present application.

In this embodiment, a storage device access device is also provided, which is used to implement the above embodiments and preferred implementation modes, and the descriptions that have been made will not be repeated. As used below, the term "module" can implement a combination of software and/or hardware of a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, the implementation of hardware, or a combination of software and hardware, is also possible and conceivable.

FIG. 10 is a structural block diagram of an access device for a storage device according to an embodiment of the present application. As shown in FIG. 10 ,

Applied to a storage expansion device, the storage expansion device is connected between a first storage device and a server, one or more virtual machines are deployed on the server, one or more storage channels that are allowed to be identified as second storage devices by the server are established in the storage expansion device, the second storage device is allocated to the virtual machine deployed on the server, and the storage channel is allocated corresponding storage space on the first storage device, the device comprises:

A receiving module 1002 is configured to receive a data access request sent by a reference virtual machine, wherein the one or more virtual machines include the reference virtual machine, and the data access request is used to request a first data access to a reference storage device in the second storage device that is allocated to the reference virtual machine;

a conversion module 1004, configured to respond to the data access request and convert the first data access into a second data access through a reference storage channel, wherein the reference storage channel is a storage channel identified as the reference storage device in the storage channels established in the storage expansion device, and the second data access is a data access to a reference storage space in the first storage device corresponding to the reference storage channel;

The control module 1006 is used to control the reference storage space of the first storage device to perform the second data access.

Through the above content, a storage expansion device is deployed between the server and the first storage device deployed on the server, and one or more storage channels established by the storage expansion device are allowed to be identified by the server as one or more second storage devices and allocated to the virtual machines deployed on the server for use, so that each virtual machine can be considered to be connected to a second storage device on the virtual machine side, and then, after the storage expansion device receives the data access request sent by the reference virtual machine on the server, the data access to the reference storage device is converted into the data access to the storage space corresponding to the reference storage channel in the first storage device through the one or more storage channels identified as the reference storage device, so that only one storage device needs to be connected to provide services for multiple virtual machines on the server, and the expansion of the storage device connected to the service is realized without changing the server's storage device usage logic and without adding new storage devices, and for the server, each virtual machine is still provided with an independent storage device, so that the expansion performance of the storage device is higher, the versatility is stronger, and it is more flexible and reliable. It can solve the problem of low storage device expansion efficiency when expanding the storage device connected to the server in the related art, and achieve the effect of improving the storage device expansion efficiency when expanding the storage device connected to the server.

Optionally, the conversion module includes:

A first determining unit, configured to determine a reference storage channel identified as the reference storage device from the storage channels established in the storage expansion device;

a forwarding unit, configured to forward the first data access to the reference storage channel;

A first conversion unit is configured to convert the first data access into the second data access through the reference storage channel.

Optionally, a first virtual bridge device is further deployed in the storage expansion device, the first virtual bridge device is connected to the storage channel established in the storage expansion device, the first virtual bridge device is also used to connect to the server and is recognized by the server as being connected to a second storage device established in the storage expansion device, and the first virtual bridge device records the second storage device and the storage channel having a corresponding relationship, wherein,

The first determining unit is configured to: search, through the first virtual bridge device, for the reference storage channel corresponding to the reference storage device from second storage devices and storage channels having a corresponding relationship;

The forwarding the first data access to the reference storage channel includes: forwarding the first data access to the reference storage channel through the first virtual bridge device.

Optionally, the first determining unit is configured to:

extracting a reference device identifier of the reference storage device from the data access request through the first virtual bridge device;

The reference storage channel corresponding to the reference device identifier is searched for from storage device identifiers and storage channels having a corresponding relationship through the first virtual bridge device, wherein the storage device identifier is a device identifier assigned to a corresponding storage channel and allowed to be identified as a storage device.

Optionally, the device further comprises:

A first creation module, used for creating an initial virtual bridge device in the storage expansion device;

A first configuration module is used to configure a first attribute of a transmission protocol for the initial virtual bridge device in a protocol configuration space of the initial virtual bridge device to obtain the first virtual bridge device, wherein the server and the storage expansion device are connected via a protocol bus that complies with the transmission protocol, and the first attribute is a protocol attribute that allows the server to identify the bridge device as complying with the transmission protocol;

The processing module is used to establish a connection between the first virtual bridge device and the storage channel established in the storage expansion device, and record the second storage device and the storage channel having a corresponding relationship.

Optionally, the storage channel includes a second virtual bridge device and a virtual storage device, the first virtual bridge device is connected to the second virtual bridge device, the second virtual bridge device is connected to the virtual storage device, and the virtual storage device is used to connect to the first storage device, wherein,

The conversion module comprises:

A first receiving unit, configured to receive the first data access through the second virtual bridge device in the reference storage channel;

a transmission unit, configured to transmit the first data access to the virtual storage device in the reference storage channel through the second virtual bridge device in the reference storage channel;

A second conversion unit is configured to convert the first data access into the second data access through the virtual storage device in the reference storage channel.

Optionally, the second conversion unit is used to:

Acquire, through the virtual storage device in the reference storage channel, a first access address indicated by the first data access, wherein the first access address is a virtual address of the reference storage device indicated by the reference virtual machine to access;

querying, through the virtual storage device in the reference storage channel, a first storage address corresponding to the first access address from a virtual address range and a storage address range having a corresponding relationship, wherein the virtual address range is an address range of the reference storage device identified by the server, and the storage address range is an address range of a storage space allocated to the virtual storage device in the reference storage channel on the first storage device;

The first access address in the first data access is replaced with the first storage address to obtain the second data access.

Optionally, the second conversion unit is used to:

The first storage address mapped by the first access address is queried from the address mapping between the addressing space of the transmission protocol of the reference storage device and the storage space of the storage protocol of the first storage device through the virtual storage device in the reference storage channel, wherein the reference storage device is identified by the server as a device that complies with the transmission protocol, the virtual address range includes the addressing space of the transmission protocol, the first storage device is a storage device of the storage protocol, and the storage address range includes the storage space of the storage protocol.

Optionally, the device further comprises:

A second creation module, used for creating multiple groups of initial virtual bridging devices and initial virtual storage devices in the storage expansion device;

a second configuration module, configured to configure a second attribute of a transmission protocol for the initial virtual bridge device in a protocol configuration space of the initial virtual bridge device to obtain the second virtual bridge device, wherein the server and the storage expansion device are connected via a protocol bus conforming to the transmission protocol, and the second attribute is a protocol attribute that allows the server to identify the bridge device as conforming to the transmission protocol;

A third configuration module is used to configure the third attribute of the transmission protocol of the initial virtual storage device in the transmission protocol configuration space of the initial virtual storage device, and to configure the fourth attribute of the storage protocol of the initial virtual storage device in the storage protocol configuration space of the initial virtual storage device, so as to obtain the virtual storage device, wherein the first storage device is a storage device that complies with the storage protocol, the third attribute is a protocol attribute that allows the server to identify it as a terminal device that complies with the transmission protocol, and the fourth attribute is a protocol attribute that allows the server to identify it as a storage device that complies with the storage protocol.

Optionally, a protocol parsing unit is deployed in the virtual storage device, and the third configuration module includes:

A configuration unit is used to configure the data link layer attributes and transaction layer attributes of the transmission protocol of the initial virtual storage device in the transmission protocol configuration space of the initial virtual storage device, wherein the third attribute configuration includes: the data link layer attributes and the transaction layer attributes, and the protocol parsing unit is used to use the data link layer attributes to perform data link layer protocol parsing on received data access, and/or use the transaction layer attributes to perform transaction layer protocol parsing on received data access.

Optionally, the configuration unit is used to:

The storage capacity attribute and queue quantity attribute of the storage protocol are configured for the initial virtual storage device in the storage protocol configuration space of the initial virtual storage device, wherein the fourth attribute configuration includes: the storage capacity attribute and the queue quantity attribute, and the storage capacity attribute and the queue quantity attribute are used to be identified by the server as a storage device that complies with the storage protocol.

Optionally, a target bus and a bus interface are further deployed in the storage expansion device, and the bus interface is used to connect the plurality of storage channels connected to the target bus to the first storage device, and the bus interface has a mapping relationship between the bus address space of the target bus and the storage address space of the first storage device, wherein:

The execution module comprises:

a second receiving unit, configured to receive the second data access via the target bus;

A forwarding unit is used to forward the second data access to the reference storage space through the bus interface according to the mapping relationship.

Optionally, the bus interface includes: a bus arbitration unit and a storage device driver unit, wherein the storage device driver unit has a mapping relationship between a bus address space of the target bus and a storage address space of the first storage device, wherein:

The forwarding unit is used for:

determining, by the bus arbitration unit, whether to process bus access of the reference storage channel according to the priority of the reference storage channel;

In the case of determining to process the bus access of the reference storage channel, the second data access is forwarded to the reference storage space according to the mapping relationship by the storage device driver unit.

Optionally, the device further comprises:

A receiving module, configured to receive access result information returned by the reference storage space in response to the second data access after the reference storage space of the first storage device is controlled to perform the second data access;

A forwarding module, used for forwarding the access result information to the reference storage channel;

A sending module is used to send the access result information to the server through the reference storage channel.

Optionally, the forwarding module includes:

An extraction unit, used to extract a source address from the access result information to obtain the reference storage space;

a second determining unit, configured to determine the reference storage channel corresponding to the reference storage space from the storage channels established in the storage expansion device;

A sending unit is used to send the access result information to the reference storage channel.

Optionally, the sending unit is used to:

replacing the source address in the access result information from the reference storage space to the reference storage device to obtain reference access result information;

The reference access result information is sent to the server.

Optionally, the device further comprises one of the following:

A first determining module, configured to determine the number of storage channels established in the storage expansion device according to resource information on the storage expansion device;

A second determining module, configured to determine the number of channels of the storage channels established in the storage expansion device according to the resource information on the storage expansion device and the device information of the first storage device;

The third determination module is used to determine the channel quantity of the storage channels established in the storage expansion device according to the resource information on the storage expansion device, the device information of the first storage device, and the virtual machine information of the virtual machine deployed on the server.

Optionally, the first determining module includes:

a detection unit, configured to detect a connection with the first storage device;

A loading unit, configured to load a storage device driver when detecting that the storage device is connected to the first storage device;

an acquiring unit, configured to acquire the device information from the first storage device through the storage device driver;

A determination unit is used to determine the channel quantity of the storage channels established in the storage expansion device according to the resource information on the storage expansion device and the device information of the first storage device.

It should be noted that the above modules can be implemented by software or hardware. For the latter, it can be implemented in the following ways, but not limited to: the above modules are all located in the same processor; or the above modules are located in different processors in any combination.

An embodiment of the present application further provides a computer-readable storage medium, in which a computer program is stored, wherein the computer program is configured to execute the steps of any of the above method embodiments when run.

In an exemplary embodiment, the computer-readable storage medium may include, but is not limited to, various media that can store computer programs, such as a USB flash drive, a read-only memory (ROM), a random access memory (RAM), a mobile hard disk, a magnetic disk or an optical disk.

An embodiment of the present application further provides an electronic device, comprising a memory and a processor, wherein the memory stores a computer program, and the processor is configured to run the computer program to execute the steps in any one of the above method embodiments.

In an exemplary embodiment, the electronic device may further include a transmission device and an input/output device, wherein the transmission device is connected to the processor, and the input/output device is connected to the processor.

For specific examples in this embodiment, reference may be made to the examples described in the above embodiments and exemplary implementation modes, and this embodiment will not be described in detail herein.

Obviously, those skilled in the art should understand that the above modules or steps of the present application can be implemented by a general computing device, they can be concentrated on a single computing device, or distributed on a network composed of multiple computing devices, they can be implemented by a program code executable by a computing device, so that they can be stored in a storage device and executed by the computing device, and in some cases, the steps shown or described can be executed in a different order from that herein, or they can be made into individual integrated circuit modules, or multiple modules or steps therein can be made into a single integrated circuit module for implementation. Thus, the present application is not limited to any specific combination of hardware and software.

The above description is only the preferred embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the present application may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the principles of the present application shall be included in the protection scope of the present application.

Claims (22)

1. A method for accessing a memory device, characterized in that,

The method is applied to a storage expansion device, the storage expansion device is connected between a first storage device and a server, one or more virtual machines are deployed on the server, one or more storage channels which are allowed to be identified as second storage devices by the server are established in the storage expansion device, the second storage devices are allocated to the virtual machines deployed on the server, and corresponding storage spaces are allocated on the first storage devices by the storage channels, and the method comprises the following steps:

Receiving a data access request sent by a reference virtual machine, wherein the one or more virtual machines comprise the reference virtual machine, and the data access request is used for requesting first data access to a reference storage device allocated to the reference virtual machine in the second storage device;

Responding to the data access request, and converting the first data access into second data access through a reference storage channel, wherein the reference storage channel is a storage channel which is identified as the reference storage device in a storage channel established in the storage expansion device, and the second data access is data access to a reference storage space corresponding to the reference storage channel in the first storage device;

controlling the reference storage space of the first storage device to execute the second data access.

2. The method of claim 1, wherein the step of determining the position of the substrate comprises,

The converting the first data access to a second data access by referring to a storage channel includes:

determining a reference storage channel identified as the reference storage device from storage channels established in the storage expansion means;

forwarding the first data access to the reference storage channel;

And converting the first data access into the second data access through the reference storage channel.

3. The method of claim 2, wherein the step of determining the position of the substrate comprises,

The storage expansion device is further provided with a first virtual bridge device, the first virtual bridge device is connected with a storage channel established in the storage expansion device, the first virtual bridge device is further used for connecting the server and is recognized by the server as being connected with the second storage device, the first virtual bridge device records a second storage device and a storage channel with corresponding relations,

The determining a reference storage channel identified as the reference storage device from the storage channels established in the storage expansion apparatus includes: searching the reference storage channel corresponding to the reference storage device from the second storage device and the storage channel with corresponding relation through the first virtual bridge device;

the forwarding the first data access to the reference storage channel includes: forwarding, by the first virtual bridge device, the first data access to the reference storage channel.

4. The method of claim 3, wherein the step of,

The searching, by the first virtual bridge device, the reference storage channel corresponding to the reference storage device from the second storage device and the storage channel having a corresponding relationship, includes:

Extracting, by the first virtual bridge device, a reference device identifier of the reference storage device from the data access request;

And searching the reference storage channel corresponding to the reference device identifier from the storage device identifiers and the storage channels which have a corresponding relationship through the first virtual bridge device, wherein the storage device identifier is a device identifier which is allocated to the corresponding storage channel and is allowed to be identified as a storage device.

5. The method of claim 3, wherein the step of,

The method further comprises the steps of:

Creating an initial virtual bridge device in the storage expansion device;

Performing first attribute configuration of a transmission protocol on the initial virtual bridging device in a protocol configuration space of the initial virtual bridging device to obtain the first virtual bridging device, wherein the server and the storage expansion device are connected through a protocol bus conforming to the transmission protocol, and the first attribute is a protocol attribute of the bridging device which is allowed to be identified by the server as conforming to the transmission protocol;

and establishing connection between the first virtual bridge equipment and the storage channel established in the storage expansion device, and recording a second storage equipment and the storage channel with corresponding relations.

6. The method of claim 1, wherein the step of determining the position of the substrate comprises,

The storage channel comprises a second virtual bridge device and a virtual storage device, a first virtual bridge device is connected with the second virtual bridge device, the second virtual bridge device is connected with the virtual storage device, the virtual storage device is used for connecting the first storage device,

The converting the first data access to the second data access through the reference storage channel includes:

receiving, by the second virtual bridge device in the reference storage channel, the first data access;

Transmitting, by the second virtual bridge device in the reference storage channel, the first data access to the virtual storage device in the reference storage channel;

the first data access is converted to the second data access by the virtual storage device in the reference storage channel.

7. The method of claim 6, wherein the step of providing the first layer comprises,

Said converting said first data access to said second data access by said virtual storage device in said reference storage channel comprising:

Acquiring a first access address of the access indicated by the first data access through the virtual storage device in the reference storage channel, wherein the first access address is a virtual address of the reference storage device indicated by the reference virtual machine;

querying, by the virtual storage device in the reference storage channel, a first storage address corresponding to the first access address from a virtual address range and a storage address range having a corresponding relationship, where the virtual address range is an address range of the reference storage device identified by the server, and the storage address range is an address range of a storage space allocated on the first storage device by the virtual storage device in the reference storage channel;

And replacing the first access address in the first data access with the first storage address to obtain the second data access.

8. The method of claim 7, wherein the step of determining the position of the probe is performed,

The querying, by the virtual storage device in the reference storage channel, a first storage address corresponding to the first access address from a virtual address range and a storage address range having a corresponding relationship, includes:

Querying, by the virtual storage device in the reference storage channel, the first storage address mapped by the first access address from an address mapping between an addressing space of a transport protocol of the reference storage device and a storage space of a storage protocol of the first storage device, wherein the reference storage device is identified by the server as a device conforming to the transport protocol, the virtual address range includes the addressing space of the transport protocol, the first storage device is a storage device of the storage protocol, and the storage address range includes the storage space of the storage protocol.

9. The method of claim 6, wherein the step of providing the first layer comprises,

The method further comprises the steps of:

Creating a plurality of groups of initial virtual bridging devices and initial virtual storage devices in the storage expansion device;

Configuring a second attribute of a transmission protocol of the initial virtual bridge device in a protocol configuration space of the initial virtual bridge device to obtain the second virtual bridge device, wherein the server and the storage expansion device are connected through a protocol bus conforming to the transmission protocol, and the second attribute is a protocol attribute of the bridge device which is allowed to be identified by the server as conforming to the transmission protocol;

And performing third attribute configuration of a transmission protocol on the initial virtual storage device in a transmission protocol configuration space of the initial virtual storage device, and performing fourth attribute configuration of a storage protocol on the initial virtual storage device in a storage protocol configuration space of the initial virtual storage device to obtain the virtual storage device, wherein the first storage device is a storage device conforming to the storage protocol, the third attribute is a protocol attribute of a terminal device which is allowed to be identified by the server as conforming to the transmission protocol, and the fourth attribute is a protocol attribute of a storage device which is allowed to be identified by the server as conforming to the storage protocol.

10. The method of claim 9, wherein the step of determining the position of the substrate comprises,

The protocol analysis unit is deployed in the virtual storage device, and the performing, in the transport protocol configuration space of the initial virtual storage device, third attribute configuration of a transport protocol on the initial virtual storage device includes:

and performing configuration of data link layer attribute and transaction layer attribute of the transmission protocol on the initial virtual storage device in a transmission protocol configuration space of the initial virtual storage device, wherein the third attribute configuration comprises: the protocol analysis unit is used for carrying out protocol analysis of a data link layer on the received data access by using the data link layer attribute and/or carrying out protocol analysis of a transaction layer on the received data access by using the transaction layer attribute.

11. The method of claim 10, wherein the step of determining the position of the first electrode is performed,

The performing, in the storage protocol configuration space of the initial virtual storage device, a fourth attribute configuration of a storage protocol on the initial virtual storage device includes:

And performing configuration of storage capacity attribute and queue number attribute of the storage protocol on the initial virtual storage device in a storage protocol configuration space of the initial virtual storage device, wherein the fourth attribute configuration comprises: the storage capacity attribute and the queue number attribute are for use by the server to identify storage devices that conform to the storage protocol.

12. The method of claim 1, wherein the step of determining the position of the substrate comprises,

The memory expansion device is further provided with a target bus and a bus interface, wherein the bus interface is used for connecting a plurality of memory channels connected to the target bus to the first memory device, and has a mapping relation between a bus address space of the target bus and a memory address space of the first memory device,

The controlling the reference storage space of the first storage device to perform the second data access includes:

Receiving the second data access via the target bus;

And forwarding the second data access to the reference storage space according to the mapping relation through the bus interface.

13. The method of claim 12, wherein the step of determining the position of the probe is performed,

The bus interface includes: a bus arbitration unit and a memory device driving unit having therein a mapping relationship of a bus address space of the target bus and a memory address space of the first memory device,

The forwarding the second data access to the reference storage space through the bus interface according to the mapping relationship includes:

determining whether to process bus access of the reference memory channel according to the priority of the reference memory channel through the bus arbitration unit;

And forwarding the second data access to the reference memory space by the memory device driving unit according to the mapping relation under the condition of determining to process the bus access of the reference memory channel.

14. A storage expansion device is characterized in that,

Comprising the following steps:

A server port, a processing chip and a storage device port, wherein the server port is connected with the processing chip, the processing chip is connected with the storage device port, one or more storage channels are established in the processing chip,

The server port is used for connecting a server, wherein one or more virtual machines are deployed on the server, the one or more storage channels allow the server to identify the one or more storage channels as one or more second storage devices, and the one or more second storage devices are distributed to the virtual machines deployed on the server;

The storage device port is used for connecting a first storage device, wherein the storage channel is allocated with a corresponding storage space on the first storage device;

the processing chip having disposed thereon the method of any of claims 1 to 13.

15. The storage expansion device of claim 14, wherein the memory is configured to store the data,

The processing chip is provided with a first virtual bridge device, the first virtual bridge device is connected with the one or more storage channels, the first virtual bridge device is also used for connecting the server port and is identified by the server as being connected with the second storage device, the first virtual bridge device records the second storage device and the storage channels with corresponding relations,

The first virtual bridge device is configured to search, when a first data access is received, a reference storage channel corresponding to a reference storage device from the second storage devices and storage channels having a corresponding relationship, where the reference storage device is a storage device allocated to a reference virtual machine that initiates a data access request in the one or more second storage devices, and the reference storage channel is a storage channel identified as the reference storage device in the one or more storage channels; forwarding the first data access to the reference storage channel.

16. The storage expansion device of claim 15, wherein the memory is configured to store the data,

The storage channel comprises a second virtual bridge device and a virtual storage device, the second virtual bridge device is connected with the first virtual bridge device, the second virtual bridge device is connected with the virtual storage device, the virtual storage device is connected with the first storage device, wherein,

The second virtual bridge device is configured to forward the first data access to the virtual storage device connected to the second virtual bridge device when the first data access is received;

The virtual storage device is used for converting the first data access into a second data access for accessing the storage space corresponding to the reference storage channel in the first storage device.

17. The storage expansion device of claim 16, wherein the memory is configured to store the data,

An external storage device is also arranged in the storage expansion device, the external storage device is connected with the virtual storage device, wherein,

The external storage device is used for caching the data to be accessed by the virtual storage device in the first storage device.

18. The storage expansion device of claim 15, wherein the memory is configured to store the data,

A target processor is also disposed in the processing chip, the target processor being coupled to the one or more memory channels via an internal high-speed bus, the target processor also being coupled to the memory device port, wherein,

And the target processor is used for responding to the data access request transmitted by the storage channel and performing storage control and data access on the first storage device.

19. An access device for a storage device, characterized in that,

The storage expansion device is applied to a storage expansion device, the storage expansion device is connected between a first storage device and a server, one or more virtual machines are deployed on the server, one or more storage channels which are allowed to be identified as second storage devices by the server are established in the storage expansion device, the second storage devices are allocated to the virtual machines deployed on the server, the storage channels are allocated with corresponding storage spaces on the first storage devices, and the storage expansion device comprises:

The receiving module is used for receiving a data access request sent by a reference virtual machine, wherein the one or more virtual machines comprise the reference virtual machine, and the data access request is used for requesting to perform first data access on a reference storage device allocated to the reference virtual machine in the second storage device;

The conversion module is used for responding to the data access request and converting the first data access into second data access through a reference storage channel, wherein the reference storage channel is a storage channel which is identified as the reference storage device in a storage channel established in the storage expansion device, and the second data access is data access to a reference storage space corresponding to the reference storage channel in the first storage device;

and the control module is used for controlling the reference storage space of the first storage device to execute the second data access.

20. A computer-readable storage medium comprising,

The computer readable storage medium has stored therein a computer program, wherein the computer program when executed by a processor realizes the steps of the method as claimed in any of claims 1 to 13.

21. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, characterized in that,

The processor, when executing the computer program, implements the steps of the method as claimed in any one of claims 1 to 13.

22. A computer program product comprising a computer program, characterized in that,

Which computer program, when being executed by a processor, carries out the steps of the method as claimed in any one of claims 1 to 13.