WO2020259418A1

WO2020259418A1 - Data access method, network card and server

Info

Publication number: WO2020259418A1
Application number: PCT/CN2020/097238
Authority: WO
Inventors: 冯宇波
Original assignee: 华为技术有限公司
Priority date: 2019-06-24
Filing date: 2020-06-19
Publication date: 2020-12-30
Also published as: CN115344197A; CN112130748B; CN112130748A

Abstract

A data access method, a network card and a server, which are used to solve the problems in a server cluster in which the time for establishing a connection is too long when a node that provides a shared disk provides the disk to other nodes for use, and in which the processor burden on an access node is too large when the access node uses EC/RAID to access a shared disk provided by another node. In the present solution, the network card acquires from the server an access request to be processed, the access request being used to access a virtual disk generated by the server, and the virtual disk corresponding to a plurality of local disks in the server; and the network card accesses the plurality of local disks according to configuration information of the plurality of local disks and the access request. Since one virtual disk corresponds to a plurality of local disks, the topology of the local disks and the virtual disk may be effectively simplified. When the virtual disk is accessed, the network card in the server executes the access request, and the plurality of local disks corresponding to the virtual disk may be simultaneously accessed, effectively improving the efficiency of data reading and writing.

Description

Data access method, network card and server

Technical field

This application relates to the field of storage technology, in particular to a data access method, network card and server.

Background technique

The server cluster realizes the local disk sharing of each node (a node in the server cluster is a server) through the network-based non-volatile storage standard (non-volatile memory express over Fabrics, NOF). Each node can communicate through a switch, and any node can read and write the local disk of other nodes. For example, node 1 can access the local disk of node 2 through the switch. In order to be able to access the local disk of node 2, node 1 needs to establish connections with multiple local disks of node 2, and set a one-to-one virtual disk corresponding to the local disk of node 2 in node 1. Node 1 accesses the virtual disk, The local disk corresponding to the virtual disk in node 2 can be accessed.

When the scale of the server cluster is large and there are many nodes, since each node has to establish a connection relationship with the local disks of other nodes, the connection establishment time is relatively long, and the network connection topology is more complicated; in addition, when When an access node accesses the local disks of other nodes through redundant arrays of independent drives (RAID)/erasure code (EC), it will increase the burden on the processor of the access node.

Summary of the invention

This application provides a data access method, network card, and server to solve the problem that when a node that provides a shared disk provides a disk for other nodes in a server cluster, it takes too long to establish a connection, and the access node accesses other nodes through EC/RAID When the shared disk is provided, the processor load of the access node is too large.

In the first aspect, this application provides a data access method applied to a network card of a server, and the method includes:

The network card obtains the pending access request from the server. The access request is used to access the virtual disk generated by the server. The virtual disk corresponds to multiple local disks in the server. After that, the network card accesses multiple local disks according to the configuration information and access requests of multiple local disks. Local disks.

Through the above method, since a virtual disk in the server corresponds to multiple local disks of the server, the topology structure of the local disk and the virtual disk can be effectively simplified. When the virtual disk needs to be accessed, the network card in the server executes the access request. At the same time, access to multiple local disks corresponding to the virtual disk can ensure the efficiency of data reading and writing. In addition, because the access operations to multiple local disks no longer require the participation of the processor in the server, but are implemented by the network card, you can Effectively reduce the power consumption of the server.

In a possible design, multiple local disks form a RAID or EC group. When the network card accesses multiple local disks, it can determine the read and write of multiple local disks based on the configuration information of the multiple local disks forming the RAID or EC group. Strategy. Reading and writing data to multiple local disks can improve data access efficiency.

In a possible design, the network card can also obtain the configuration information of multiple local disks from the server after executing the access request. Illustratively, the network card can read the configuration information of multiple local disks from the memory of the server. Furthermore, subsequent data reading and writing can be realized by multiple local disks, which can improve data access efficiency.

In a possible design, the type of access request is different, and the way the network card executes the access request is different. Exemplarily, when the access request is a data read request, the data read request is used to request to read data from the virtual disk. When the network card accesses multiple local disks to perform an access request, it can read data from multiple local disks according to the configuration information, and send the read data to the sender of the data read request, that is, connect to the server through the network card Other servers.

Through the above method, the network card can simultaneously read data from multiple local disks and execute access requests, which can effectively ensure the efficiency of data reading and writing.

In a possible design, the type of access request is different, and the way the network card executes the access request is different. Exemplarily, the access request is a data write request, and the data write request is used to write data to the virtual disk. When the network card accesses multiple local disks to perform an access request, it writes the data in the data write request to the multiple local disks according to the configuration information.

Through the above method, the network card can write data to multiple local disks at the same time, which can realize efficient data reading and writing.

In the second aspect, this application provides a data access method applied to a first server, the first server and the second server are connected through a network card, the method includes: the first server generates a virtual disk, and the virtual disk corresponds to a plurality of Local disk and set the virtual disk to be accessible by the second server; the first server can receive the connection request from the second server, and transmit the virtual disk information to the second server, so that the second server can access the virtual disk according to the virtual disk information .

Through the above method, the first server can notify the second server by sending virtual disk information, so that the second server can access multiple local disks of the first server by accessing the virtual disk, so that the virtual disk It can be associated with multiple local disks instead of a one-to-one correspondence, which can simplify the topology of virtual disks and local disks.

In a possible design, the first server may configure multiple local disks as a RAID or EC group. Ensure that data can be read and written to multiple local disks at the same time in the follow-up, improving the efficiency of data reading and writing.

In a possible design, the first server and the second server communicate through the NOF protocol to realize local disk sharing. When the first server generates virtual disks, it can generate NVMe subsystems for multiple local disks; it can also generate a virtual NVMe subsystem and generate a virtual disk identifier, write the virtual disk identifier into the virtual NVMe subsystem; and create a virtual The relationship between the NVMe subsystem and multiple NVMe subsystems.

Through the above method, the first server can establish the association relationship between the virtual NVMe subsystem and the NVMe subsystems of multiple local disks, so that the virtual disk can be associated with multiple local disks instead of a one-to-one correspondence. Simplify the topology of virtual disks and local disks.

In a third aspect, this application provides a network card, which is located in a server, and the network card has the functions implemented by the network card in the first aspect and any one of the possible designs of the first aspect. The function can be realized by hardware, or by hardware executing corresponding software. The hardware or software includes one or more modules corresponding to the above-mentioned functions. In a possible design, the structure of the device includes a transmission unit and a processing unit, and these units can perform the corresponding functions in the above-mentioned method example of the first aspect. For details, please refer to the detailed description in the method example, which will not be repeated here.

In a fourth aspect, this application provides a first server, and the first server and the second server are connected through a network card. The first server has the functions implemented by the first server in any possible design of the second aspect and the second aspect. The function can be realized by hardware, or by hardware executing corresponding software. The hardware or software includes one or more modules corresponding to the above-mentioned functions. In a possible design, the structure of the device includes a creation unit and a transmission unit, and these units can perform the corresponding functions in the above-mentioned method example of the second aspect. For details, refer to the detailed description in the method example, which will not be repeated here.

In the fifth aspect, this application also provides a network card. For the beneficial effects, please refer to the description of the first aspect and any possible design of the first aspect, which will not be repeated here. The structure of the network card includes a processor and a memory, and the processor is configured to support the network card to perform corresponding functions in the first aspect and any one of the possible design methods of the first aspect. The memory is coupled with the processor, and it stores the necessary program instructions and data of the network card. The structure of the network card also includes a communication interface for communicating with other devices.

In the sixth aspect, the present application also provides a first server. For beneficial effects, please refer to the description of the first aspect and any possible design of the first aspect and will not be repeated here. The structure of the first server includes a processor and a memory, and the processor is configured to support the first server to perform a corresponding function in the second aspect and any one of the possible design methods of the second aspect. The memory is coupled with the processor, and it stores the necessary program instructions and data of the first server. The structure of the first server also includes a communication interface for communicating with other devices.

In a seventh aspect, the present application also provides a computer-readable storage medium that stores instructions in the computer-readable storage medium, which when run on a computer, causes the computer to execute the methods of the foregoing aspects.

In an eighth aspect, the present application also provides a computer program product containing instructions, which when run on a computer, causes the computer to execute the methods of the above aspects.

In a ninth aspect, the present application also provides a computer chip, which is connected to a memory, and the chip is used to read and execute a software program stored in the memory, and execute the methods of the foregoing aspects.

Description of the drawings

Figure 1 is a schematic diagram of a server cluster architecture;

Figure 2 is a schematic diagram of a system architecture provided by this application;

FIG. 3 is a schematic diagram of an association relationship between a virtual NVMe subsystem and an NVMe subsystem provided by this application;

4 is a schematic diagram of a method for mapping multiple local disks to one virtual disk provided by this application;

FIG. 5 is a schematic diagram of a method for reading and writing data provided by this application;

Fig. 6 is a schematic structural diagram of a server provided by this application;

FIG. 7 is a schematic structural diagram of a network card provided by this application.

Detailed ways

As shown in FIG. 1, it is a schematic structural diagram of a server cluster. The server cluster includes multiple nodes. In FIG. 1, two nodes (node 100 and node 200) are taken as an example, and the nodes are connected through a switch 10. .

The structure of each node in the server cluster is basically the same, and the node 100 is taken as an example for description below. The node 100 includes a processor 110, a network card 120, a local disk 130, a virtual disk 140, and a memory 150. Optionally, it may also include a communication interface for communicating with other devices; each virtual disk 140 is associated with a local disk 230 in the node 200, and the process of associating the virtual disk 140 with the local disk 230 in the node 200 will be described in A detailed description is given below. The local disk in the embodiment of the present application is a solid-state drive (SSD), and communicates with the processor 110 through a non-volatile memory express (NVMe) protocol.

The memory 150 is used to store computer program instructions, and the processor 110 can call the computer program instructions stored in the memory 150 to implement main computing functions of the node 100.

Hereinafter, taking the server cluster shown in FIG. 1 as an example, the process of establishing a connection between the node 100 and the local disk in the node 200 and the node 100 locally establishing the virtual disk 140 associated with the local disk 230 in the node 200 is introduced.

First, according to the NVMe protocol, the node 200 establishes an NVMe subsystem (NVMe subsystem) for each local disk 230 in the node 200, and additionally generates an NVMe subsystem with a discovery service (discovery service). The node 200 records the NVMe subsystem information of the local disk 230 that needs to be provided to the node 100 to access in the NVMe subsystem with discovery service. When the node 100 needs to connect to the local disk 230 of the node 200, it first sends a discovery command to the node 200. After receiving the discovery command, the node 200 sends the NVMe subsystem information recorded in the NVMe subsystem with the discovery service to the node 100. The node 100 can establish connections with the NVMe subsystem indicated by the information of the NVMe subsystem with the discovery service one by one according to the information of the NVMe subsystem.

When establishing a connection with one of the NVMe subsystems, the node 100 sends a connection command (connection cammand) to the node 200. The connection command carries the NVMe qualified name (NVMe qualified name, host NQN) of the host, the NQN of the NVMe subsystem, the host identifier, and related parameters of the submission queue (submission quene) and completion queue (completion quene). The host NQN and host identifier are used to identify the node 100. The NQN of the NVMe subsystem is used to identify the NVMe subsystem to establish a connection this time. After the node 200 receives the connection command, it establishes the submission queue and the completion queue used when the node 100 communicates with the NVMe subsystem. After the submission queue and the completion queue are established, it returns related information about successful connection. Information to node 100. The node 100 establishes the virtual disk 140 and creates the NVMe subsystem of the virtual disk 140, and records the information of the NVMe subsystem of the local disk 230 in the node 200 in the NVMe subsystem of the virtual disk 140; the node 100 accesses the virtual disk 140 can access a local disk 230 of the node 200.

When the node 200 includes multiple local disks 230, the node 100 and each local disk 230 must establish a connection according to the method described above. In this way, it takes a relatively long time to establish a connection, and the topology of the node 100 that needs to be maintained It is also more complicated. In addition, for data reliability and read and write efficiency, the node 100 will perform RAID or EC on a virtual disk group composed of multiple virtual disks, which will increase the load of the node 100. Among them, RAID refers to a disk array composed of multiple independent disks, which allows multiple local disks of the disk array to be read and written at the same time;

RAID configuration configures a data read and write strategy for the disk array to improve data reliability and input/output (I/O) performance. EC configuration is an extension of RAID configuration. To a certain extent, EC configuration can be regarded as the general formula of RAID configuration, which means that any RAID configuration can be converted to a specific EC configuration.

In order to solve the above problem, in the embodiment of the present invention, multiple local disks in a node can be mapped to a virtual disk. As shown in FIG. 2, a server cluster includes two nodes (node 300 and node 400) as an example , The architecture of the server cluster to which the embodiment of the present application is applied is described, and the nodes are connected through the switch 10. The structure of the node 300 and the node 400 is similar to that of the node 100 and the node 200 in FIG. 1, except that one virtual disk 340 in the node 300 can be associated with multiple local disks 430 in the node 400, and one virtual disk in the node 400 The disk 440 may be associated with multiple local disks 330 in the node 300.

The memory 350 is used to store computer program instructions. The processor 310 calls the computer program instructions stored in the memory 350 to perform some operations. The processor 310 can call the computer program instructions stored in the memory 350 to execute the local disks and virtual disks provided in the embodiments of this application. Method of association. For example, the operations performed by the node 300 in the embodiment shown in FIG. 4 in the embodiment of the present application can be performed. Similarly, the memory 450 is used to store computer program instructions, and the processor 410 calls the computer program instructions stored in the memory 450 to perform some operations, for example, it can execute the execution of the node 400 in the embodiment shown in FIG. 4 in the embodiment of the present application. operating.

In the embodiment of the present application, the network card 320 may have an Ethernet card (rdma network interface controller, RNIC) with a remote direct memory access (RDMA) engine. The following takes the network card 420 as an example, one possibility for the network card in the node As shown in FIG. 2, the network card 420 includes a processor 4201 and a memory 4202. The processor 4201 may call the computer execution instructions stored in the memory 4202, so that the network card can execute the data access method provided in the embodiment of the present application (as shown in FIG. 5). Optionally, the network card 420 may also include a communication interface. When the processor 4201 is communicating with other devices, data transmission may be performed through the communication interface.

The processor 4201 is configured to call computer program instructions stored in the memory 4202 to implement access to the local disk 430. For example, to execute the data access method provided in the embodiment of the present application, see the embodiment shown in FIG. 5 for details. The processor 4201 may be a central processing unit (central processing unit, CPU) or the like.

The memory 4202 is used to store computer program instructions. The memory 4202 can be a volatile memory, such as a random access memory; the memory 4202 can also be a non-volatile memory, such as a read-only memory, flash memory, or other storable computer program instructions. Medium.

In another embodiment of the present application, the network card 420 includes a programmable logic chip, and program instructions are burned in the programmable logic chip, and the programmable logic chip can execute the program burned therein. Execute the data access method provided in the embodiment of this application.

Please refer to FIGS. 3 and 4 at the same time. FIG. 3 is a schematic diagram of mapping multiple local disks 430 in the node 400 to a virtual disk 340 in the node 300, and FIG. 4 shows the mapping of multiple local disks 430 in the node 400 as nodes A flowchart of a method of virtual disk 340 in 300.

Step 401: The node 400 establishes an NVMe subsystem for each local disk 430, such as NVMe subsystem 01, VVMe subsystem 02, and NVMe subsystem 03 in FIG. 3.

Step 402: The node 400 establishes the virtual NVMe subsystem 001.

When establishing the virtual NVMe subsystem, the node 400 also generates a virtual disk identifier, such as dev/NVMe001, and writes the virtual disk identifier in the virtual NVMe subsystem.

Step 403: The node 400 selects multiple NVMe subsystems of the local disk 430 that can be provided to the node 300 to access, and associates the NVMe subsystems of the multiple local disks 430 with the virtual network subsystem 001, and records the association relationship.

When the node 400 associates the virtual NVMe subsystem 001 with the NVMe subsystems of the multiple local disks 430, the node 400 adds the information of the NVMe subsystems of the multiple local disks 430 to the device list of the virtual NVMe subsystem 001.

The information of the NVMe subsystem of the local disk 430 may be the NQN of the NVMe, or other information used to indicate the NVMe subsystem, which is not limited in the embodiment of the present application.

The manner in which the node 400 selects the NVMe subsystem of the multiple local disks 430 is not limited in the embodiment of the present application. The node 400 may be selected randomly or according to a preset rule.

Exemplarily, one or more local disk groups may be pre-configured in the node 400, and each local disk group includes multiple local disks 430, and the node 400 may select a local disk group from the multiple local disk groups; after that, Associate the NVMe subsystems of the multiple local disks 430 in the local disk group with the virtual NVMe subsystem 001.

The node 400 may also perform RAID/EC configuration on the multiple local disks 430 associated with the virtual NVMe subsystem 001, and allocate the role of each local disk 430 in RAID/EC, for example, the local disk 430 is used to implement the original data Storage, or storage of verification data.

After the node 400 establishes the association relationship between the virtual NVMe subsystem 001 and the NVMe subsystem of the multiple local disks 430, the association relationship and the RAID/EC configuration of the multiple local disks 430 can be stored locally, such as in the memory 450 The network card 420 can read the association relationship and the RAID/EC configuration of the multiple local disks 430 from the memory 450, and perform processing on the multiple local disks 430 according to the association relationship and the RAID/EC configuration of the multiple local disks 430 Data reading and writing, and the data access method executed by the network card 420 will be described in detail in the embodiment shown in FIG. 5 below.

Step 404: The node 300 sends a discovery command to the node 400.

Step 405: After receiving the discovery command sent by the node 300, the node 400 sends the identifier of the virtual NVMe subsystem 001 to the node 300. In this way, the virtual disk created by the node 400 for the virtual NVMe subsystem is allowed to be accessed by the node 300.

The node 300 may first send a discovery command to the node 400, and the discovery command is used to query the NVMe subsystem of the local disk 430 in the node 400 that the node 300 is allowed to access. In the embodiment of the present invention, since the node 400 constructs the virtual NVMe subsystem 001, when receiving the discovery command, the related information of the virtual NVMe subsystem 001 is sent to the node 300, and the virtual NVMe subsystem 001 The related information of the subsystem 001 includes the identification of the virtual NVMe subsystem 001 and the identification of the corresponding virtual disk.

Step 406: After receiving the identifier of the virtual NVMe subsystem 001 fed back by the node 400, the node 300 may send a connection command to the node 400 to request the establishment of a connection with the virtual NVMe subsystem. The connection command carries the NVMe NQN of the host, the identifier of the virtual NVMe subsystem 001, the host identifier, and related parameters of the submission queue and the completion queue. The host NQN and host identifier are used to identify the node 300. The identifier of the virtual NVMe subsystem 001 is used to identify the virtual NVMe subsystem 001 to which the connection is established this time. For example, the identifier of the virtual NVMe subsystem 001 may be the NVMe NQN of the virtual NVMe subsystem 001.

Step 407: After receiving the connection command, the node 400 establishes the submission queue and the completion queue used when the node 300 communicates with the virtual NVMe subsystem 001, and returns after the submission queue and completion queue are established. The information related to the successful connection is given to the node 300. The node 300 establishes a virtual disk 340, and creates an NVMe subsystem of the virtual disk 340, records the information of the virtual NVMe subsystem in the node 400 in the NVMe subsystem of the virtual disk 340; establishes a connection between the virtual disk 340 and the virtual NVMe subsystem The association relationship, and further, the multiple local disks 430 of the node 400 are mapped to one virtual disk 340 of the node 300. The node 300 can access multiple local disks 430 of the node 400 by accessing the virtual disk 340.

As shown in FIG. 5, it is a flowchart of a method for accessing multiple local disks 430 in a node 400 by accessing a virtual disk 340 according to an embodiment of this application. The method includes:

Step 501: The node 300 generates an access request for accessing the virtual disk 340, and sends the access request to the node 400.

In the node 300, since the association relationship between the virtual disk 340 and the virtual NVMe subsystem is recorded, the node 300 will convert the access request for accessing the virtual disk 140 into the access request for accessing the virtual NVMe subsystem. request.

Step 502: The network card 420 obtains an access request from the node 400.

Specifically, after the virtual disk 340 establishes a connection with the virtual NVMe subsystem, the sending queue and the completion queue are generated in the node 300, and the receiving queue and the submission queue are generated in the node 400. The sending queue and the receiving queue correspond one-to-one, and the submission queue and the completion queue correspond one-to-one. After the node 300 generates the access request, it puts the access request in the sending queue of the node 300 and waits to be sent to the node 400. After the access request is sent to the node 400, the node 400 places the access request in the receiving queue, and the network card 420 obtains the access request from the receiving queue.

Step 503: The network card 420 accesses the multiple local disks 430 according to the association relationship between the virtual NVMe subsystem 001 and the NVMe subsystem of the multiple local disks 430 and the RAID/EC configuration of the multiple local disks 430 to execute the access request .

In addition to obtaining the access request, the network card 420 can also obtain multiple local disks 430 and the RAID/EC configuration of the multiple local disks 430 corresponding to the NVMe subsystem 001, and access multiple local disks according to the RAID/EC configuration of the local disk 430 430.

In the embodiment of the present invention, the network card 420 has a remote direct memory access (RDMA) function. The network card 420 can read the access request from the memory 450 and execute step 503. When the access request is a data write request, in step 503, the network card 400 acquires the data write request, acquires the data in the data write, and writes the data in the write request to the node 400 in the local disk 430. The network card 400 writes the data in the write request to the multiple local disks 430 corresponding to the NVMe subsystem 001 according to the RAID/EC configuration of the multiple local disks 430 and the multiple local disks 430 corresponding to the NVMe subsystem 001.

When the access request is a data read request, in step 503, when the node 400 receives the data read request, the network card 420 reads the data read request from the memory 450, according to the NVMe sub The multiple local disks 430 and the RAID/EC configuration of the multiple local disks 430 corresponding to the system 001 read data from the multiple local disks 430 corresponding to the NVMe subsystem 001.

Since the network card 420 can execute the access request in step 503, the load of the node 400 can be effectively reduced.

Based on the same inventive concept as the method embodiment, the embodiment of the present application also provides a first server, the first server (such as the node 400 in the embodiment of this application) and the second server (such as the node 300 in the embodiment of this application) ) Connected through a network card, the first server can be used to execute the method executed by the node 400 in the method embodiment shown in FIG. 4, and related features can be found in the above method embodiment, which will not be repeated here, as shown in FIG. The server includes a creation unit 601 and a transmission unit 602:

The creation unit 601 is configured to generate virtual disks, the virtual disks correspond to multiple local disks in the first server, and the virtual disks are set to be accessible by the second server; for example, the creation unit 601 can execute the embodiment shown in FIG. 4 For details of the method executed by the node 400 in steps 401 to 403, refer to the foregoing content, which is repeated here.

The transmission unit 602 is configured to receive a connection request from the second server, and transmit the virtual disk information to the second server, so that the second server can access the virtual disk according to the virtual disk information. For example, the transmission unit 602 can execute the embodiment shown in FIG. 4 In steps 405 to 407, the node 400 performs sending and receiving operations, such as receiving a discovery command, a connection command, and sending an identification of the virtual NVMe subsystem 001. For details, please refer to the foregoing content, which will be repeated here.

Optionally, the creation unit 601 may configure multiple local disks as a RAID group or an EC group.

Optionally, the first server and the second server communicate through the NOF protocol. When creating a virtual disk, the creation unit may first generate NVMe subsystems for multiple local disks; generate a virtual NVMe subsystem and generate a virtual disk. Disk identification, write the virtual disk identification into the virtual NVMe subsystem; associate the virtual NVMe subsystem with multiple NVMe subsystems.

Based on the same inventive concept as the method embodiment, the embodiment of the application also provides a network card, which is used to execute the method executed by the network card 420 in the above method embodiment. For related features, please refer to the above method embodiment, and will not be repeated here. As shown in FIG. 7, the network card includes a transmission unit 701 and a processing unit 702:

The transmission unit 701 is used to obtain a pending access request from the server, the access request is used to access a virtual disk, and the virtual disk corresponds to multiple local disks in the server; the transmission unit 701 can perform the steps in the embodiment shown in FIG. 5 For the operation of obtaining the access request in 502, refer to the foregoing content for details, which will be repeated here.

The processing unit 702 is configured to access multiple local disks according to the configuration information and access requests of the multiple local disks. The processing unit 702 may perform the operation of executing the access request in step 503 in the embodiment shown in FIG. 5, for details, please refer to the foregoing content, which will be repeated here.

Optionally, the configuration information of multiple local disks is the configuration information of multiple local disks forming a RAID or EC group.

Optionally, the processing unit 702 may also obtain configuration information of multiple local disks.

Optionally, the type of access request is different, and the access operation performed by the processing unit is different. The following two cases are introduced separately:

First, the access request is a data read request, and the data read request is used to request to read data from the virtual disk;

When the processing unit accesses multiple local disks, it can read data from the multiple local disks according to the configuration information, and send the read data to another server connected to the server through a network card. The server sends to the server.

Second, the access request is a data write request, and the data write request is used to write data to the virtual disk;

When the processing unit accesses multiple local disks, it writes the data in the data write request to the multiple local disks according to the configuration information.

It should be noted that the division of units in the embodiments of the present application is illustrative, and is only a logical function division, and there may be other division methods in actual implementation. The functional units in the embodiments of the present application may be integrated into one processing unit, or each unit may exist alone physically, or two or more units may be integrated into one unit. The above-mentioned integrated unit can be implemented in the form of hardware or software functional unit.

The foregoing embodiments can be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented by software, the above-mentioned embodiments may be implemented in the form of a computer program product in whole or in part. The computer program product includes one or more computer instructions. When the computer program instructions are loaded or executed on the computer, the processes or functions according to the embodiments of the present invention are generated in whole or in part. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions may be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions may be transmitted from a website, computer, server, or data center. Transmission to another website site, computer, server or data center via wired (such as coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (such as infrared, wireless, microwave, etc.). The computer-readable storage medium may be any available medium that can be accessed by a computer or a data storage device such as a server or a data center that includes one or more sets of available media. The usable medium may be a magnetic medium (for example, a floppy disk, a hard disk, a magnetic tape), an optical medium (for example, a DVD), or a semiconductor medium. The semiconductor medium may be a solid state drive (SSD).

Based on the same inventive concept as the method embodiment, the embodiment of the present application also provides a server for executing the method executed by the node 400 in the above method embodiment. For the structure of the server, refer to the structure of the node 400 in FIG. 2.

The functions/implementation processes of the transmission unit and the creation unit in FIG. 6 can be implemented by the processor 410 in FIG. 2 calling instructions stored in the memory 450. Alternatively, the function/implementation process of the creation unit in FIG. 6 may be implemented by the processor 410 in FIG. 2 calling computer program instructions stored in the memory 450, and the function/implementation process of the transmission unit in FIG. 6 may be the communication in the server. Interface to achieve.

Based on the same inventive concept as the method embodiment, the embodiment of the present application also provides a network card for executing the method performed by the network card 420 in the above method embodiment. For the structure of the network card, refer to the structure of the network card 420 in FIG. 2.

The functions/implementation processes of the transmission unit and the processing unit in FIG. 7 can be implemented by the processor 4201 in FIG. 2 calling instructions stored in the memory 4202. Alternatively, the function/implementation process of the processing unit in FIG. 7 may be implemented by the processor 4201 in FIG. 2 calling computer program instructions stored in the memory 4202, and the function/implementation process of the transmission unit in FIG. 6 may be implemented by the network card 420. Communication interface.

In the several embodiments provided in this application, it should be understood that the disclosed system, device, and method may be implemented in other ways. For example, the device embodiments described above are only illustrative. For example, the division of the units is only a logical function division, and there may be other divisions in actual implementation, for example, multiple units or components can be combined or It can be integrated into another system, or some features can be ignored or not implemented. In addition, the displayed or discussed mutual coupling or direct coupling or communication connection may be indirect coupling or communication connection through some interfaces, devices or units, and may be in electrical, mechanical or other forms.

The units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the objectives of the solutions of the embodiments.

The above are only specific embodiments of the present invention. Those skilled in the art can think of changes or substitutions according to the specific implementation manners provided by the present invention, which should all fall within the protection scope of the present invention.

Claims

A data access method, characterized in that it is applied to a network card of a server, and the method includes:

Acquiring a pending access request from the server, the access request being used to access a virtual disk, the virtual disk corresponding to multiple local disks in the server; and

Accessing the multiple local disks according to the configuration information of the multiple local disks and the access request.
The method according to claim 1, wherein the configuration information is configuration information of a redundant independent disk array (RAID) or an error correction code EC group formed by the multiple local disks.
The method according to claim 1 or 2, wherein the method further comprises:

Obtain the configuration information of the multiple local disks from the server.
The method according to any one of claims 1 to 3, wherein the method further comprises:

The access request is a data read request, and the data read request is used to request to read data from the virtual disk;

The accessing the multiple local disks according to the configuration information of the multiple local disks and the access request includes:

Read data from the multiple local disks according to the configuration information, and send the read data to another server connected to the server through a network card, and the data read request is made by the other server Send to the server.
The method according to any one of claims 1 to 3, wherein the method further comprises:

The access request is a data write request, and the data write request is used to write data to the virtual disk;

The accessing the multiple local disks according to the configuration information of the multiple local disks and the access request includes:

Writing the data in the data write request to the multiple local disks according to the configuration information.
A data access method, applied to a first server, the first server and the second server are connected through a network card, characterized in that the method includes:

Generating a virtual disk, the virtual disk corresponding to multiple local disks in the first server, and setting the virtual disk to be accessible by the second server;

Receiving a connection request from the second server, and transmitting the virtual disk information to the second server, so that the second server accesses the virtual disk according to the virtual disk information.
The method of claim 6, wherein the multiple local disks are configured as a RAID group or an EC group.
The method according to claim 6 or 7, wherein the first server and the second server communicate through a network-based non-volatile storage standard NOF protocol, and the generating a virtual disk includes:

Generating a non-volatile storage standard NVMe subsystem for the multiple local disks;

Generate a virtual NVMe subsystem, generate a virtual disk identifier, and write the virtual disk identifier into the virtual NVMe subsystem;

Associating the virtual NVMe subsystem with the multiple NVMe subsystems.
A network card, characterized in that the network card is located in a server, and the network card includes a transmission unit and a processing unit:

The transmission unit is configured to obtain a pending access request from the server, the access request is used to access a virtual disk, and the virtual disk corresponds to multiple local disks in the server;

The processing unit is configured to access the multiple local disks according to the configuration information of the multiple local disks and the access request.
The network card according to claim 9, wherein the configuration information is the configuration information of the redundant array of independent disks (RAID) or error correction code EC group formed by the multiple local disks.
The network card according to claim 9 or 10, wherein the processing unit is further configured to:

Obtain the configuration information of the multiple local disks from the server.
The network card according to any one of claims 9 to 11, wherein the access request is a data read request, and the data read request is used to request to read data from the virtual disk;

The processing unit is specifically configured to access the multiple local disks according to the configuration information of the multiple local disks and the access request:

Read data from the multiple local disks according to the configuration information, and send the read data to another server connected to the server through a network card, and the data read request is made by the other server Send to the server.
The network card according to any one of claims 9 to 11, wherein the access request is a data write request, and the data write request is used to write data to the virtual disk;

The processing unit is specifically configured to access the multiple local disks according to the configuration information of the multiple local disks and the access request:

Writing the data in the data write request to the multiple local disks according to the configuration information.
A first server, wherein the first server and the second server are connected through a network card, wherein the first server includes a creation unit and a transmission unit;

The creating unit is configured to generate a virtual disk, the virtual disk corresponding to multiple local disks in the first server, and setting the virtual disk to be accessible by the second server;

The receiving unit is configured to receive a connection request from the second server, and transmit the virtual disk information to the second server, so that the second server accesses the virtual disk according to the virtual disk information.
The first server of claim 14, wherein the multiple local disks are configured as a RAID group or an EC group.
The first server according to claim 14 or 15, wherein the first server and the second server communicate through a network-based non-volatile storage standard NOF protocol, and the creation unit specifically uses in:

Generating an NVMe subsystem for the multiple local disks;

Generate a virtual NVMe subsystem, generate a virtual disk identifier, and write the virtual disk identifier into the virtual NVMe subsystem;

Associating the virtual NVMe subsystem with the multiple NVMe subsystems.
A network card, characterized by comprising a memory and a processor; the memory stores program instructions, and the processor runs the program instructions to execute the method according to any one of claims 1 to 5.
A server, characterized by comprising a memory and a processor; the memory stores program instructions, and the processor runs the program instructions to execute the method according to any one of claims 6-8.