WO2014201623A1

WO2014201623A1 - Method, apparatus and system for data transmission, and physical network card

Info

Publication number: WO2014201623A1
Application number: PCT/CN2013/077412
Authority: WO
Inventors: 赵晓雷
Original assignee: 华为技术有限公司
Priority date: 2013-06-18
Filing date: 2013-06-18
Publication date: 2014-12-24
Also published as: CN103609077A; CN103609077B

Abstract

The embodiment of the present invention provides a method, apparatus and system for data transmission, and a physical network card. The method comprises: determining a main physical function (PF) module in a physical function (PF) module group of a first physical network card, the PF module group comprising at least two PF modules; determining a target virtual function (VF) module cluster managed by the main PF module on the basis of virtual function (VF) module groups corresponding to each of the PF modules in the PF module group; and allocating VF modules in the target VF module cluster to at least one virtual machine, enabling the at least one virtual machine to realize data transmission. The method, apparatus and system for data transmission in the embodiment of the present invention can avoid system failures and reduce resource waste.

Description

Method, device and system for data transmission and physical network card

The present invention relates to the field of communications and, more particularly, to a method, apparatus and system for data transmission and a physical network card. Background technique

As the foundation of the rapid development of cloud computing, virtualization technology has gradually matured in computing, storage, network and cluster management, and promotes the application of cloud computing in various fields.

With the development of the central processing unit (CPU) chip technology, the capability of a single physical CPU is getting stronger and stronger, the number of virtual machines that can be provided is increasing, and the performance requirements of the physical network card are higher. The number of virtual network cards is increasing, to meet the requirements of more users, and to improve the utilization of cloud resources. In network virtualization, NIC virtualization is the key technology to solve the last mile of virtualization. It has experienced software simulation, Virtual Machine Device Queue (VMDq), and single root I/O virtualization (Single-Root I). /O

Virtualization, SR-IOV) After three phases of development, SR-IOV technology has become the ideal device to solve the last mile of virtualization. SR-IOV is used to interconnect a high-speed peripheral component (Peripheral Component Interconnect Express, PCIe). The NIC device is virtualized into at least two PCIe NIC devices, and each virtual PCIe NIC device provides services to the upper layer software of the virtual machine just like the physical PCIe NIC device.

SR-IOV is a 10 virtualization technology standard that allows one physical device to support at least two virtual NICs. SR-IOV introduces two new types of functions, Physical Function (Physical Function,

PF) module and Virtual Function (VF) module. Among them, the PF module has all the PCIe functions of the SR-IOV, and can perform SR-IOV function configuration and management. The VF module has a lightweight PCIe function, including resources necessary for data transmission and a small amount of configuration resources. As a separate PCIe device, the VF module can be managed via Virtual Machine Management (Virtual Machine Management,

VMM) is assigned to the virtual machine (VM).

However, as a key node of the system, the PF module will cause system failure and waste of resources in the event of a failure. Summary of the invention Embodiments of the present invention provide a method, an apparatus, and a system for data transmission, and a physical network card, which can avoid system failure and reduce resource waste.

The first aspect provides a method for data transmission, including: determining, from a physical function PF module group of a first physical NIC, a primary PF module, where the PF module group includes at least two PF modules, and the PF module group Each of the PF modules corresponds to a different virtual function VF module group; based on the VF module group corresponding to each PF module in the PF module group, the target VF module group managed by the main PF module is determined, wherein the VF in the target VF module group The module is configured to be allocated to at least one virtual machine VM such that the at least one VM implements data transmission.

With reference to the first aspect, in a first possible implementation manner of the first aspect, determining, by the PF module group of the first physical network card, the primary PF module includes: determining, according to a reference of each PF module in the PF module group a parameter, the primary PF module is determined, wherein the reference parameter of the PF module includes at least one of a PF module failure number, a number of VF modules in the VF module group corresponding to the PF module, a static bandwidth of the PF module, and a static bandwidth of the PF module. .

With reference to the first aspect, or the first possible implementation manner of the first aspect, in a second possible implementation manner of the first aspect, the determining, according to the VF module group corresponding to each PF module in the PF module group, determining the primary The target VF module group managed by the PF module includes: selecting at least one VF module from the VF module group corresponding to each PF module in the PF module group to form the target VF module group.

With reference to the first aspect, the first possible implementation manner of the first aspect, or the second possible implementation manner of the first aspect, in a third possible implementation manner of the first aspect, the method further includes: determining The first PF module is re-enabled or fault-recovered, and the first PF module is determined as the PF module in the PF module group. Based on the first PF module, the VF module group corresponding to the first PF module is established.

With reference to the first aspect, or any one of the first to the third possible implementation manners of the first aspect, in a fourth possible implementation manner of the first aspect, the method further includes: determining the second PF module Deleting from the PF module group of the first physical network card; releasing the VF module group corresponding to the second PF module.

With reference to the first aspect, or any one of the first to fourth possible implementations of the first aspect, in a fifth possible implementation of the first aspect, the method further includes: determining the first physics When the third PF module of the network card fails, determining whether the VF module in the VF module group corresponding to the third PF module is available; when at least one VF module in the VF module group corresponding to the third PF module is available, The at least one VF available in the VF module group corresponding to the first PF module The module is determined to continue to be available.

In conjunction with the first aspect, or any one of the first to fifth aspects of the first aspect, in a sixth possible implementation of the first aspect, the method further comprises: determining the target

The first VF module allocated to the first VM of the at least one VM in the VF module group is faulty; from the VF module group corresponding to each PF module in the PF module group, selecting a second different from the first VF module VF module; update the target VF module group.

With reference to the first aspect, or any one of the first to the sixth possible implementation manners of the first aspect, in the seventh possible implementation manner of the first aspect, the first physical network card occupies at least two high speeds The peripheral component interconnects the standard PCIe bus, the method further includes: determining that data of the VF module connected to the first PCIe bus of the at least two PCIe buses in the target VF module group needs to be sent to the target VF module group a VF module connected to the second PCIe bus of the at least two PCIe buses; and the data of the VF module connected to the first PCIe bus is sent to the VF module connected to the second PCIe bus through the switch of the first physical network card.

With reference to the first aspect, or any one of the first to the seventh possible implementation manners of the first aspect, in the eighth possible implementation manner of the first aspect, the method further includes: determining the primary PF module A fault occurs; from the PF module group, a PF module other than the main PF module is selected as a new main PF module; and the new main PF module is determined based on the VF module group corresponding to each PF module in the PF module group. A new target VF module group is managed, wherein the VF module in the new target VF module group is used to allocate to at least one virtual machine VM, so that the at least one VM implements data transmission.

The second aspect provides an apparatus for data transmission, including: a first determining unit, configured to determine, from a physical function PF module group of the first physical network card, a primary PF module, where the PF module group includes at least two PF module, each PF module in the PF module group corresponds to a different virtual function VF module group; and a second determining unit is configured to determine the main PF module management based on the virtual function VF module group corresponding to each PF module in the PF module group The target VF module group, wherein the VF module in the target VF module group is used to allocate to at least one virtual machine VM, so that the at least one VM implements data transmission.

With reference to the second aspect, in a first possible implementation manner of the second aspect, the first determining unit is specifically configured to: determine, according to a reference parameter of each PF module in the PF module group, the primary PF module, where The reference parameters of the PF module include at least the number of PF module failures, the number of VF modules in the VF module group corresponding to the PF module, the static bandwidth of the PF module, and the static bandwidth of the PF module. One.

With reference to the second aspect, or the first possible implementation manner of the second aspect, in the second possible implementation manner of the second aspect, the second determining unit is specifically configured to: use each PF module from the PF module group At least one VF module is selected from the corresponding VF module group to form the target VF module group.

With reference to the second aspect, the first possible implementation manner of the second aspect, or the second possible implementation manner of the second aspect, in a third possible implementation manner of the second aspect, the apparatus further includes: a determining unit, configured to determine that the first PF module is re-enabled or fault-recovered, determining the first PF module as a PF module in the PF module group; and establishing a unit, configured to establish the first PF module based on the first PF module A VF module group corresponding to a PF module.

With reference to the second aspect, or any one of the first to the third possible implementation manners of the second aspect, in a fourth possible implementation manner of the second aspect, the device further includes: a fifth determining unit, The second PF module is configured to be deleted from the PF module group of the first physical NIC; and the release unit is configured to release the VF module group corresponding to the second PF module.

With reference to the second aspect, or any one of the first to the fourth possible implementation manners of the second aspect, in a fifth possible implementation manner of the second aspect, the device further includes: a sixth determining unit, When it is determined that the third PF module of the first physical network card is faulty, determining whether the VF in the VF module group corresponding to the third PF module is available; the holding unit, configured to be in the VF module group corresponding to the third PF module When at least one VF module is available, the at least one VF module in the VF module group corresponding to the third PF module is determined to continue to be available.

With reference to the second aspect, or any one of the first to the fifth possible implementation manners of the second aspect, in a sixth possible implementation manner of the second aspect, the device further includes: a seventh determining unit, Determining that a first VF module allocated to the first VM of the at least one VM in the target VF module group is faulty; and selecting a unit for selecting from a VF module group corresponding to each PF module in the PF module group a second VF module different from the first VF module; an update unit, configured to update the target VF module group, and assign the second VF module to the first VM.

With reference to the second aspect, or any one of the first to the sixth possible implementation manners of the second aspect, in the seventh possible implementation manner of the second aspect, the first physical network card occupies at least two high speeds The peripheral component interconnects the standard PCIe bus, the device further includes: an eighth determining unit, configured to determine that data of the VF module connected to the first PCIe bus of the at least two PCIe buses in the target VF module group needs to be sent to a VF module connected to the second PCIe bus of the at least two PCIe buses in the target VF module group; a sending unit, configured to connect the VF to the first PCIe bus The data of the module is sent to the VF module connected to the second PCIe bus through the switch of the first physical network card.

With reference to the second aspect, or any one of the first to the seventh possible implementation manners of the second aspect, in the eighth possible implementation manner of the second aspect, the device further includes a third determining unit, Determining that the main PF module is faulty; the first determining unit is further configured to: select, from the PF module group, a PF module other than the main PF module as a new main PF module; the second determining unit further uses And: determining, according to the VF module group corresponding to each PF module in the PF module group, a new target VF module group managed by the new main PF module, where the VF module in the new target VF module group is used to allocate to At least one virtual machine VM, such that the at least one VM implements data transmission.

In a third aspect, an apparatus for data transmission is provided, including a processor and a memory; wherein the memory stores program code, the processor is configured to call the program code stored in the memory, and performs the following operations: Physical NIC module physical group PF module group, the main PF module is determined, the PF module group includes at least two PF modules, each PF module in the PF module group corresponds to a different virtual function VF module group; based on each of the PF module groups a virtual function VF module group corresponding to the PF module, determining a target VF module group managed by the main PF module, wherein the target

The VF module in the VF module group is used to allocate to at least one virtual machine VM, so that the at least one

The VM implements data transfer.

With reference to the third aspect, in a first possible implementation manner of the third aspect, the processor is configured to invoke the program code stored in the memory, and specifically perform the following operations: based on a reference of each PF module in the PF module group a parameter, the primary PF module is determined, wherein the reference parameters of the PF module include the number of failures of the PF module, the number of VF modules in the VF module group corresponding to the PF module,

At least one of a static bandwidth of the PF module and a static bandwidth of the PF module.

In conjunction with the third aspect or the first possible implementation manner of the third aspect, in a second possible implementation manner of the third aspect, the processor is configured to invoke the program code stored in the memory, and specifically perform the following operations: At least one VF module is selected from the VF module group corresponding to each PF module in the PF module group to form the target VF module group.

With reference to the third aspect, the first possible implementation manner of the third aspect, or the second possible implementation manner of the third aspect, in a third possible implementation manner of the third aspect, the processor is used to invoke the The program code stored in the memory further performs the following operations: determining that the first PF module is re-enabled or failback, determining the first PF module as the PF module in the PF module group; establishing based on the first PF module The VF module group corresponding to the first PF module. With reference to the third aspect, or any one of the first to the third possible implementation manners of the third aspect, in a fourth possible implementation manner of the third aspect, the processor is configured to invoke the storage in the memory The program code also performs the following operations: determining that the second PF module is deleted from the PF module group of the first physical network card; releasing the VF module group corresponding to the second PF module.

With reference to the third aspect, or any one of the first to fourth possible implementation manners of the third aspect, in a fifth possible implementation manner of the third aspect, the processor is configured to invoke the storage in the memory The program code is further configured to: determine whether the VF module in the VF module group corresponding to the third PF module is available when the third PF module of the first physical network card is faulty; and corresponding to the third PF module When at least one VF module in the VF module group is available, the at least one VF module available in the VF module group corresponding to the first PF module is determined to be continuously available.

With reference to the third aspect, or any one of the first to fifth possible implementation manners of the third aspect, in a sixth possible implementation manner of the third aspect, the processor is configured to invoke the storage in the memory The program code is further configured to: determine that the first VF module allocated to the first VM of the at least one VM in the target VF module group is faulty; from the VF module group corresponding to each PF module in the PF module group Selecting a second VF module different from the first VF module; updating the target VF module group, and assigning the second VF module to the first VM.

With reference to the third aspect, or any one of the first to the sixth possible implementation manners of the third aspect, in the seventh possible implementation manner of the third aspect, the first physical network card occupies at least two high speeds The peripheral component interconnects a standard PCIe bus, the processor is configured to call the program code stored in the memory, and performs the following operations: determining that the target PCI module is required to be connected to the first PCIe bus of the at least two PCIe buses The data of the VF module is sent to the VF module in the target VF module group and connected to the second PCIe bus of the at least two PCIe buses; and the data of the VF module connected to the first PCIe bus passes through the first physical network card. The switch sends to the VF module connected to the second PCIe bus.

With reference to the third aspect, or any one of the first to seventh possible implementation manners of the third aspect, in an eighth possible implementation manner of the third aspect, the processor is configured to invoke the storage in the memory The program code also performs the following operations: determining that the main PF module is faulty; from the PF module group, selecting a PF module other than the main PF module as a new main PF module; based on each PF in the PF module group a VF module group corresponding to the module, determining a new target VF module group managed by the new main PF module, wherein the VF module in the new target VF module group is used to allocate to at least one virtual machine VM, so that the at least one The VM implements data transfer. In a fourth aspect, a physical network card is provided, including: a PF module group having at least two PF modules, and at least two VF module groups; wherein, one of the at least two PF modules corresponds to the at least one One of the two VF module groups, one of the at least two PF modules can manage the VF module in the VF module group to which the other PF modules are connected after being determined as the main PF module.

With reference to the fourth aspect, in a first possible implementation manner of the fourth aspect, the VF module in each VF module group of the at least two VF module groups includes two types of spatial resources; a type of space resource is used to access the PF module corresponding to the VF module group to which the VF module to which the two types of space resources belong before the primary PF module is determined; the other two of the two types of space resources are accessed. A type of spatial resource is used to establish a connection with the primary PF module for access by the primary PF module after the primary PF module is determined.

With reference to the fourth aspect, or the first possible implementation manner of the fourth aspect, in a second possible implementation manner of the fourth aspect, the physical network card occupies at least two high-speed peripheral components interconnecting a standard PCIe bus, The physical network card further includes a built-in switch; wherein the built-in switch is configured to send data of the VF module connected to the first PCIe bus of the at least two PCIe buses to the at least two PCIe buses The VF module connected to the second PCIe bus.

With reference to the fourth aspect, the first possible implementation manner of the fourth aspect, or the second possible implementation manner of the fourth aspect, in a third possible implementation manner of the fourth aspect, the physical network card passes at least two a port is connected to the external switch; the physical network card is further configured to receive a data stream sent by the device for data transmission through the aggregated link, and uniformly send the data stream to the external switch by using each of the at least two ports .

A fifth aspect, a system for data transmission, comprising the apparatus for data transmission in any of the possible implementations of the second aspect or the second aspect, and the fourth aspect or the fourth aspect A physical network card in a possible implementation, or a device for data transmission in any of the possible implementations of the third aspect or the third aspect, and any one of the fourth aspect or the fourth aspect The physical network card in the implementation manner; wherein the first physical network card or the third aspect or the third aspect of the possible implementation in any one of the possible implementation manners in the second aspect or the second aspect The first physical network card referred to in the manner is the physical network card referred to in the fourth aspect or any possible implementation manner of the fourth aspect.

Therefore, in the embodiment of the present invention, one physical network card has a PF module group composed of at least two PF modules, and one PF can be selected from at least two PF modules in the PF module group. The module acts as the main PF module, which avoids system failures and reduces resource waste due to the many opportunities for selection. DRAWINGS

In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings to be used in the embodiments or the description of the prior art will be briefly described below. Obviously, the drawings in the following description are only the present invention. For some embodiments, other drawings may be obtained from those of ordinary skill in the art without departing from the drawings.

1 is a schematic flow chart of a method for data transmission according to an embodiment of the present invention.

2 is a schematic flow chart of a method for data transmission according to another embodiment of the present invention. 3 is a schematic block diagram of an apparatus for data transmission in accordance with another embodiment of the present invention. 4 is a schematic block diagram of an apparatus for data transmission in accordance with another embodiment of the present invention. FIG. 5 is a schematic block diagram of an apparatus for data transmission according to another embodiment of the present invention. 6 is a schematic block diagram of a physical network card in accordance with another embodiment of the present invention.

7 is a schematic block diagram of a system for data transmission in accordance with another embodiment of the present invention. detailed description

The technical solutions in the embodiments of the present invention are clearly and completely described in the following with reference to the accompanying drawings in the embodiments of the present invention. It is obvious that the described embodiments are a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments obtained by a person of ordinary skill in the art based on the embodiments of the present invention without making creative labor are within the scope of the present invention.

1 is a method for data transmission in accordance with an embodiment of the present invention. As shown in FIG. 1, the method 100 includes:

S110: Determine a primary PF module from a first PF module group of the first physical NIC, where the first PF module group includes at least two PF modules, and each PF module in the first PF module group corresponds to a different VF module group.

S120. Determine, according to the VF module group corresponding to each PF module in the PF module group, a target VF module group managed by the main PF module, where the VF module in the target VF module group managed by the main PF module is allocated to at least one virtual Machine, so that the at least one virtual machine implements data transmission.

A device for data transmission, for example, a device having at least one processor for data transmission function, such as a host, or a host group composed of at least two hosts, which can be from the first physics Determining a main PF module in a PF module group including at least two PF modules of the network card; and then determining, from the VF module group corresponding to each PF module in the PF module group, a target VF module group that needs to be managed by the main PF module, where The VF module in the target VF module group can then be assigned to at least one virtual machine, so that each virtual machine can implement data transmission through the VF module.

Therefore, in the embodiment of the present invention, one physical network card has at least two PF modules to form a first PF module group, and one PF module can be selected as a main PF module from at least two PF modules in the first PF module group. Since there are at least two PF modules to choose from, the reliability of the main PF module is increased, thereby reducing the probability of system failure and reducing resource waste caused by system failure.

The physical network card mentioned in the embodiment of the present invention may also be referred to as an SR-IOV network card, which may include two functional modules, a PF module and a VF module, which those skilled in the art should know, also referred to as PF and VF. Among them, the PF module supports the PCIe function of the SR-IOV extended function, which is used to configure and manage the SR-IOV function; the VF module implements the PCIe function. After the primary PF module is determined from the PF module group and the target VF module group managed by the primary PF module is determined, the virtual machine management (VMM) can treat each VF module in the target VF module group as A PCIe device is assigned to a virtual machine (VM). When the VM performs control plane access to the assigned VF module, the VM initiates a request, the primary PF module and the VMM respond to the request, and return the result to the VM. When the VM sends and receives data packets from the VF module, it combines the input/output memory management unit (IMMMU) technology provided by the CPU to realize complete and independent transmission and reception. The main PF module and the VMM do not need to participate.

In the embodiment of the present invention, each PF module in the PF module group has a corresponding relationship with the VF module group. Before the main PF module is selected, each PF module can be used to manage all VF modules in the corresponding VF module group. That is to say, the VF modules in the corresponding VF module group are hung under each PF module. After the main PF module is selected, the main PF module can implement unified management and resource allocation for the VF modules in the VF module group corresponding to all PF modules in the PF module group, so that the VF modules corresponding to all PF modules can be obtained. All or part of the VF modules in the group are selected to form a target VF module group, and the VF module in the target VF module group is allocated to the virtual machine by the VMM.

In the embodiment of the present invention, the VF module may include two types of space resources, and the first type of space resource is controlled by the directly corresponding PF module. Before the primary PF module is determined, the directly corresponding PF module accesses the first type of space resource. Manage the VF module; the second type of space resource is in the system After the primary PF module is selected, it is remapped to the area visible to the primary PF module (that is, the connection between the second spatial resource and the primary PF module is established), and the primary PF module can directly access the second type of spatial resource to implement The VF module is managed and is not controlled by the directly corresponding PF module.

In the embodiment of the present invention, determining the primary PF module from the PF module group of the first physical network card in S110 may include: determining the primary PF module based on a reference parameter of each PF module in the PF module group, where The reference parameter may include at least one of a number of failures of the PF module, a number of VF modules in the VF module group corresponding to the PF module, a static bandwidth of the PF module, and a dynamic bandwidth of the PF module.

Specifically, when it is required to determine the main PF module from the PF module group of the first physical network card, the apparatus for data transmission may determine each of the PF module groups based on the reference parameter of each PF module in the PF module group. The performance value of a PF module, and based on the performance value of each PF module, the priority ordering of the PF modules in the PF module group can be determined, and then the main PF module can be determined from the PF module group based on the priority ordering. In the embodiment of the present invention, the reference parameter of each PF module can be obtained according to different reference performance values with respect to the PF module. For example, if the reference parameter of the PF module is the number of PF module failures and the number of VF modules in the VF module group corresponding to the PF module, the baseline performance value of one failure can be set to -10, and the reference performance value of the VF module is 5, if the number of failures of PF module 0 is 2 and the number of VF modules in the corresponding VF module group is 6, the number of failures of PF module 1 is 3, and the number of VF modules in the corresponding VF module group is 5, then It can be calculated that the performance value of PF module 0 is 10, and the performance value of PF module 1 is -5, then the performance value of PF module 0 is greater than the performance value of PF module 1, and thus, the priority of PF module 0 is higher than that of PF module. 1 , you can select PF module 0 as the main PF module. It should be understood that the foregoing is only an example, and the embodiment of the present invention is not limited thereto.

In the embodiment of the present invention, the following principles can be determined for how to determine the primary PF module: If the reference parameter is only the number of PF module failures, the fewer the number of failures occurs, the higher the priority of the PF module is, and the more likely it is to be determined as the primary PF. Module; If the reference parameter is only the number of VF modules in the VF module group corresponding to the PF module, the more the number of VF modules in the VF module group corresponding to the PF module, the higher the priority of the PF module is, the more likely it is to be If the reference parameter is only the static bandwidth of the PF module, the greater the static bandwidth of the PF module, the higher the priority of the PF module, the more likely it is to be determined as the primary PF module; if the reference parameter is only the PF module Dynamic bandwidth, the larger the dynamic bandwidth of the PF module, the higher the priority, the more likely it is to be determined as the primary PF module; if the reference parameter is the number of PF module failures, the number of VF modules in the VF module group corresponding to the PF module PF mode At least two of the block dynamic bandwidth and the PF module static bandwidth can be prioritized by considering two factors to determine the main PF module.

In the embodiment of the present invention, the target VF module group managed by the main PF module is determined according to the VF module group corresponding to each PF module in the PF module group, and may include: each PF module from the first PF module group. At least one VF module is selected from the corresponding VF module group to form the target VF module group.

For example, when at least two VF modules are required to implement the link aggregation function (otherwise, in other scenarios), after the primary PF module is determined, it may be from the VF module group corresponding to each PF module in the first PF module group. Select at least one VF module to form the target VF module group. In the embodiment of the present invention, if a certain PF module fails, the normal operation of the VF module in the VF module group corresponding to the other PF modules is not affected, so that the VF modules of different VF module groups can be isolated from each other, so that the present invention Embodiments Selecting a VF module from a VF module group corresponding to different PF modules can improve system reliability and stability, thereby ensuring system bandwidth.

In the embodiment of the present invention, the aggregated link data stream needs to be sent (the aggregated link data stream refers to the data stream jointly sent by the at least two VF modules, that is, the channel through which at least two VF modules send and receive data can form an aggregated link. The aggregated link data stream can be sent to the external switch through multiple ports on the physical NIC. In this case, the traffic of each port can be monitored in real time. For example, the physical NIC has two ports connected to the external switch. During a certain period of time, the sending scheduling module in the physical network card detects that the traffic of port 0 is 0.2 G, and the traffic of port 1 is 1 G. In order to achieve link bandwidth balancing, the aggregation link sent to port 1 can be reduced. The traffic is increased to the traffic of the aggregated link of port 0. For example, the traffic of port 0 and port 1 can be adjusted to 0.6G.

It should be understood that for non-aggregated link data streams, the data stream can also be sent to the external switch through multiple ports, and the traffic of the port can be monitored and adjusted in real time.

It should be understood that, in the embodiment of the present invention, after determining the main PF module, the VF module group corresponding to the main PF module may be directly determined as the target VF module group managed by the main PF module, or one or more other ones or more The VF module group corresponding to the PF module is determined as the target VF module group managed by the main PF module. Specifically, the target VF module group is determined from the VF module group corresponding to each PF module in the PF module group, and the present invention is not limited.

In the embodiment of the present invention, the method 100 may further include: determining that the first PF module is re-enabled or failing, and determining the first PF module as the PF module in the first PF module group; The first PF module establishes a VF module group corresponding to the first PF module. Alternatively, the method 100 may further include: determining that the second PF module is deleted from the first PF module group of the first physical network card; releasing the VF module group corresponding to the second PF module.

Specifically, the means for data transmission can monitor all VF modules of the first physical network card and manage the VF modules in groups. For example, the VF module connected to the PF module 0 can be composed of the VF module group 0, that is, the VF module 0, the VF module 1 and the VF module 127 form the VF module group 0, and the VF module connected to the PF module 1 constitutes the VF module group. 1 , that is, the VF module 128, the VF module 129, the VF module 130... The VF module 255 constitutes the VF module group 1. After the first PF module in the first physical network card is re-enabled or the fault is recovered, the device for data transmission may establish a VF corresponding to the first PF module based on the first PF module and the VF module connected to the first PF module. Module group. When the second PF module in the PF module group of the first physical network card is deleted from the system (for example, the second PF module is deleted or the user actively deletes), the established second PF module may be released. Group of VF modules.

Therefore, in the embodiment of the present invention, after the PF module is re-enabled or the fault is re-established, the VF module group corresponding to the re-enabled or fault-recovering PF module may be established, or after the PF module is deleted, the deleted is deleted. The VF module group corresponding to the PF module can dynamically adjust the VF module that can be used in the physical network card, making the system more flexible.

In the embodiment of the present invention, the method 100 may further include: determining whether the VF module in the VF module group corresponding to the third PF module is available when the third PF module of the first physical network card is faulty; When at least one VF module in the VF module group corresponding to the third PF module is available, the at least one VF available in the VF module group corresponding to the first PF module is determined to be continuously available.

Specifically, in the embodiment of the present invention, when a certain PF module of a physical network card fails, the VF module group corresponding to the PF module may not be directly released, and the state of the VF module under the PF module may be further determined. If all the VF modules under the PF module are faulty, the VF module group corresponding to the PF module can be directly released. If a part of the VF module under the PF module fails, only the failed VF module needs to be removed from the VF. Excluded from the module group. Therefore, in the embodiment of the present invention, when the PF module fails, if the VF module that is connected to the PF module is available, the available VF module that is connected to the failed PF module is determined to continue to be available, thereby avoiding resource waste. .

It should be understood that the first PF module, the second PF module and the third PF mentioned in the embodiments of the present invention Modules can be the same PF module or different PF modules.

In the embodiment of the present invention, the method 100 may further include: determining that a first VF module allocated to the first VM of the at least one VM in the target VF module group is faulty; corresponding to each PF module in the PF module group In the VF module group, a second VF module is determined, the second VF module is different from the VF module included in the target VF module group; the target VF module group is updated, and the second VF module is allocated to the first VM.

Specifically, when a VF module that is allocated in a VM fails, a new VF module can be reselected from the VF module group corresponding to each PF module in the PF module group of the physical network card, and the reselected VF is selected. The module is different from the VF module included in the original target VF module group; then, the target VF module group is updated, that is, the reselected VF module is used as a member of the target VF module group, and the reselected VF module is assigned to The VM.

In the embodiment of the present invention, the method 100 may further include: determining that data of the VF module connected to the first PCIe bus in the target VF module group needs to be sent to the VF connected to the second PCIe bus in the target VF module group. And transmitting, by the switch of the first physical network card, the data of the VF module connected to the first PCIe bus to the VF module connected to the second PCIe bus.

That is, the first physical network card can occupy at least two PCIe buses. If one VF module data on one PCIe bus needs to be sent to another VF module on the same PCIe bus, the data can be implemented through an internal switch of the physical network card. Send. If the VF module data connected to one PCIe bus needs to be sent to the VF module connected to another PCIe bus, the data can also be transmitted through the internal switch of the physical network card.

Therefore, in the embodiment of the present invention, data exchange between VF modules connected to different PCIe buses in the same physical network card, or data exchange between VF modules connected to the same PCIe bus in the same physical network card may pass the data exchange. The switch built in the physical NIC can save network bandwidth and reduce data transmission and reception delay.

In the embodiment of the present invention, the method 100 may further include: determining that the primary PF module is faulty; and selecting, from the PF module group, a PF module other than the primary PF module as a new primary PF module; a VF module group corresponding to each PF module in the module group, determining a new target VF module group managed by the new main PF module; assigning the VF module in the new target VF module group to at least one virtual machine, so that At least one virtual machine implements data transfer.

Specifically, during the running of the system, the specific operation information of the main PF module can be tracked in real time, and the state of the main PF module is monitored. If the main PF module fails, it can be from PF Reselecting a PF module as a new main PF module in the module group, updating the recorded operation information to the new main PF module, and re-determining the new target VF module group for the new main PF module, and The VF module in the new target VF module group is allocated to at least one virtual machine, so that the at least one virtual machine implements data transmission.

Therefore, in the embodiment of the present invention, when the original primary PF module fails, a new primary PF module can be reselected from the PF module group having at least two PF modules, which can further avoid system failure and reduce resource waste.

It should be understood that the present invention can also be combined with a software simulation virtual technology, that is, a second PF module group can also exist in the network card, and the PF module in the second PF module group can virtualize at least two by software simulation. vNICs for use by legacy virtual machines.

In order to more clearly understand the present invention, a method for data transmission according to an embodiment of the present invention will be described below with reference to FIG.

First, the scene shown in Fig. 2 will be described. In this scenario, the PF module group 1 includes the PF module 0; the PF module group 2 includes the PF module 1 and the PF module 2, the PF module 1 has the VF module 1 and the VF module 2 attached thereto, and the PF module 2 has the VF module 3 attached thereto. And VF module 4. PF module 0, PF module 1 and PF module 2 are shown as PF0, PF1 and PF2 in the figure; VF module 1, VF module 2, VF module 3 and VF module 4 are respectively shown as VF 1, VF2, VF3 and VF4. The VF DD shown in Figure 2 refers to the drive of the VF module (Device Driver), and the PF DD refers to the drive of the PF module. The VMM can be run by the host for VM creation, deletion, management of hardware resources, allocation of VM resources, and so on. The method according to an embodiment of the present invention will be specifically described below with reference to FIG.

The VMM can select a PF module as the main PF module from the PF module group 2 in the physical network card (NIC). Specifically, the VMM can be based on the number of failures of each PF module, the number of VF modules in the corresponding VF module group, and the PF module. At least one of the static bandwidth and the dynamic bandwidth of the PF module determines the primary PF module. For example, as shown in FIG. 2, the PF module 1 may be determined as the primary PF module. After the main PF module is determined, the VMM can determine the VF module to be managed by the main PF module. If the link aggregation function is required, the VF module can be selected from the VF modules connected to different PF modules. The target VF module group managed by the PF module, for example, as shown in FIG. 2, the VF module 2, the VF module 3, and the VF module 4 can be determined as the target VF module group. Then, the VMM can allocate each VF module in the target VF module group that needs to be managed by the main PF module to each VM for subsequent data transmission of each VM, for example, as shown in FIG. 2, The VF module 2 is assigned to the VM 2, the VF module 3 is assigned to the VM 3, and the VF module 4 is assigned to the VM 4. When the selected main PF module fails, the VMM can reselect a PF module from the PF module group as the main PF module, and can reselect the target VF module group for the main PF module. In the embodiment of the present invention, the VMM may also virtualize one or at least two vNICs through the PF module based on the software simulation manner. For example, as shown in FIG. 2, the vNIC1 and the vNIC 2 are virtualized by the PF module 0, respectively. Legacy) VM1 and legacy (legacy) VM2 services. After the VF module or vNIC is assigned to the VM, each VM can implement data transmission. The following describes the implementation of the four main data transmission paths:

Data Stream 1: Data exchange between Legacy VMs is achieved through VMM's built-in virtual software switch (Vi r tua l SW swi tch ).

Data Stream 2: The data channels sent and received by the virtual NIC in the Legacy VM are: External Switch Internal Switch - MAC - PF - VMM - VM; VM - VMM - PF - MAC - Internal Switch - External Switch.

Data Stream 3: The data exchange between the VMs using the VF module is implemented by the switch built in the NIC. Whether the VF module is connected to the same PCIe or a different PCIe, it can be implemented through the switch built into the NIC.

Data Stream 4: The VM data transceiver channel using the VF module is: External Switch - Internal Switch - Media Access Control (MAC) → VF; VF - MAC - Internal Switch - External Switch.

Data stream 5: The channel through which the primary PF performs resource management and control for the VF is: VM→VMM - primary PF-VF; VF-primary PF→VMM→VM.

Therefore, in the embodiment of the present invention, a physical network card has a PF module group composed of at least two PF modules, and one PF module can be selected as a main PF module from at least two PF modules in the PF module group, thereby Due to the many opportunities for selection, system failures can be avoided and resource waste can be reduced.

3 is a schematic block diagram of an apparatus 200 for data transmission in accordance with an embodiment of the present invention. As shown

As shown in FIG. 3, the apparatus 200 includes a first determining unit 210 and a second determining unit 220. The first determining unit 210 is configured to determine, from the physical function PF module group of the first physical network card, the main PF module, where the PF module group includes at least two PF modules, and each PF module in the PF module group corresponds to a different VF module group. The second determining unit 220 is configured to determine, according to the virtual function VF module group corresponding to each PF module in the PF module group, the target VF module group managed by the main PF module, where the VF module in the target VF module group is used for Assigned to at least one virtual machine VM so that the at least one VMs implement data transfer.

Optionally, the first determining unit 210 is specifically configured to: determine, according to a reference parameter of each PF module in the PF module group, the primary PF module, where the reference parameter of the PF module includes a PF module failure number, a PF module At least one of the number of corresponding VF modules, the static bandwidth of the PF module, and the static bandwidth of the PF module.

Optionally, the second determining unit 220 is specifically configured to: select at least one VF module from the VF module group corresponding to each PF module in the PF module group to form the target VF module group.

As shown in FIG. 4, the apparatus 200 further includes a fourth determining unit 252 and an establishing unit 254. The fourth determining unit 252 is configured to determine that the first PF module is re-enabled or fault-recovered, and the first PF module is determined as a PF module in the PF module group; and the establishing unit 254 is configured to be based on the first PF module. Establishing a VF module group corresponding to the first PF module.

Optionally, as shown in FIG. 4, the apparatus 200 further includes a fifth determining unit 256 and a releasing unit 258. The fifth determining unit 256 is configured to determine that the second PF module is deleted from the PF module group of the first physical network card; and the releasing unit 258 is configured to release the VF module group corresponding to the second PF module.

Optionally, as shown in FIG. 4, the apparatus 200 further includes a sixth determining unit 262, a holding unit 264, and a holding unit 264. The sixth determining unit 262 is configured to determine whether the VF in the VF module group corresponding to the third PF module is available when the third PF module of the first physical network card is faulty; the holding unit 264 is configured to be in the When at least one VF module in the VF module group corresponding to the third PF module is available, the at least one VF module in the VF module group corresponding to the third PF module is determined to continue to be available.

Optionally, as shown in FIG. 4, the apparatus 200 further includes a seventh determining unit 272, a selecting unit 274, and an updating unit 276. The seventh determining unit 272 is configured to determine that the first VF module allocated to the first VM of the at least one VM in the target VF module group is faulty; the selecting unit 274 is configured to use the VF corresponding to each PF module in the PF module group. In the module group, a second VF module different from the first VF module is selected; the update unit 276 is used to update the target VF module group.

Optionally, the first physical network card occupies at least two high-speed peripheral components to interconnect the standard PCIe bus. As shown in FIG. 4, the apparatus 200 further includes: an eighth determining unit 282, configured to determine that the target needs to be Transmitting, by the VF module group, the data of the VF module connected to the first PCIe bus of the at least two PCIe buses to the VF module of the target VF module group and the second PCIe bus of the at least two PCIe buses; a sending unit 284, configured to connect to the first PCIe bus The data of the connected VF module is sent to the VF module connected to the second PCIe bus through the switch of the first physical network card.

Optionally, as shown in FIG. 4, the apparatus 200 further includes a third determining unit 240. The third determining unit 240 is configured to determine that the primary PF module is faulty; the first determining unit 210 is further configured to: select, from the PF module group, a PF module other than the primary PF module as a new primary PF. The second determining unit 220 is further configured to: determine, according to the VF module group corresponding to each PF module in the PF module group, a new target VF module group managed by the new primary PF module, where the new target VF The VF module in the module group is used to allocate to at least one virtual machine VM, so that the at least one VM implements data transmission.

Figure 5 is a schematic block diagram of an apparatus 300 for data transmission in accordance with an embodiment of the present invention. As shown in FIG. 5, the apparatus 300 includes: a processor 310310 and a memory 320320; wherein the memory 320 stores program code, and the processor 310 is configured to invoke the program code stored in the memory 320 to perform the following operations: The physical function of a physical network card in the PF module group, determine the main

a PF module, the PF module group includes at least two PF modules, and each PF module in the PF module group corresponds to a different virtual function VF module group; and based on the virtual function VF module group corresponding to each PF module in the PF module group, a target VF module group managed by the main PF module, wherein, the target

The VM implements data transfer.

Optionally, the processor 310 is configured to invoke the program code stored in the memory 320, and specifically perform the following operations: determining the primary PF module based on a reference parameter of each PF module in the PF module group, where the PF The reference parameters of the module include at least one of the number of PF module failures, the number of VF modules corresponding to the PF module, the static bandwidth of the PF module, and the static bandwidth of the PF module.

Optionally, the processor 310 is configured to invoke the program code stored in the memory 320, and specifically perform the following operations: selecting at least one VF module from the VF module group corresponding to each PF module in the PF module group, The target VF module group.

Optionally, the processor 310 is configured to invoke the program code stored in the memory 320, and further perform the following operations: determining that the first PF module is re-enabled or failback, and the first PF module is The PF module in the PF module group is determined; based on the first PF module, the VF module group corresponding to the first PF module is established.

Optionally, the processor 310 is configured to invoke the program code stored in the memory 320, and further perform the following operations: determining that the second PF module is deleted from the PF module group of the first physical network card; releasing the second PF The VF module group corresponding to the module.

Optionally, the processor 310 is configured to invoke the program code stored in the memory 320, and further perform the following operations: when determining that the third PF module of the first physical network card fails, determining a VF corresponding to the third PF module Whether the VF module in the module group is available; when at least one VF module in the VF module group corresponding to the third PF module is available, determining the at least one VF module available in the VF module group corresponding to the first PF module To continue to be available.

Optionally, the processor 310 is configured to invoke the program code stored in the memory 320, and further perform the following operations: determining that the first VF module allocated to the first VM of the at least one VM in the target VF module group appears The second VF module different from the first VF module is selected from the VF module group corresponding to each PF module in the PF module group; and the target VF module group is updated.

Optionally, the first physical NIC occupies at least two high-speed peripheral components interconnecting a standard PCIe bus, and the processor 310 is configured to invoke the program code stored in the memory 320, and performs the following operations: determining that the target needs to be And transmitting, by the VF module group, the data of the VF module connected to the first PCIe bus of the at least two PCIe buses to the VF module of the target VF module group and the second PCIe bus of the at least two PCIe buses; The data of the VF module connected to the first PCIe bus is sent to the VF module connected to the second PCIe bus through the switch of the first physical network card.

Optionally, the processor 310 is configured to invoke the program code stored in the memory 320, and further perform the following operations: determining that the primary PF module is faulty; and selecting, from the PF module group, other than the primary PF module The PF module is used as a new main PF module; based on the VF module group corresponding to each PF module in the PF module group, a new target VF module group managed by the new main PF module is determined, wherein the new target VF module group The VF module is configured to be allocated to at least one virtual machine VM, so that the at least one VM implements data transmission.

Therefore, in the embodiment of the present invention, a physical network card has a PF module group composed of at least two PF modules, and one PF module can be selected as a main PF module from at least two PF modules in the PF module group, thereby Due to the many opportunities to choose, you can avoid system failures. And reduce waste of resources.

6 is a schematic block diagram of a physical network card in accordance with an embodiment of the present invention. As shown in FIG. 6, the physical network card 400 includes: a PF module group 410 composed of at least two PF modules 411, and at least two VF module groups 420; wherein one of the at least two PF modules 411 is a PF module. 411 corresponds to one of the at least two VF module groups 420, and any one of the at least two PF modules 411 can manage other after being determined to be the main PF module.

The VF module 421 in the VF module group 420 to which the PF module 411 is connected.

Optionally, each PF module 411 in the PF module group 410 is connected to each other, and any PF module 411 in the PF module group 410 can manage the VF module group connected to the other PF module 411 through other PF modules 411. VF module 421 in 420.

Optionally, the VF module 421 in each of the at least two VF module groups 420 includes two types of spatial resources; wherein, one of the two types of spatial resources is used in the Before the primary PF module is determined, the PF module 411 corresponding to the VF module group 420 to which the VF module 421 to which the two types of spatial resources belong belongs is accessed; another type of spatial resource in the two types of spatial resources is used in the After the primary PF module is determined, establishing a connection with the primary PF module is accessed by the primary PF module.

Optionally, the physical network card 400 occupies at least two high-speed peripheral components interconnecting a standard PCIe bus, and the physical network card 400 further includes a built-in switch; wherein the built-in switch is used to connect with the at least two PCIe buses The data of the VFE module connected to the first PCIe bus is transmitted to the VF module connected to the second PCIe bus of the at least two PCIe buses.

Optionally, the physical network card is connected to the external switch by using at least two ports; the physical network card is further configured to receive a data stream sent by the device for data transmission through the aggregation link, and pass each port of the at least two ports. Equalizing the data stream to the external switch.

FIG. 7 is a schematic block diagram of a system 500 for data transmission in accordance with an embodiment of the present invention. As shown in FIG. 7, the system 400 includes a device 410 for data transmission and a physical network card 420; wherein the device 510 for data transmission may correspond to the device 200 or 300 for data transmission in the embodiment of the present invention, physical The network card 520 can correspond to the physical network card 400, and the device 200 for data transmission The first physical network card mentioned in 300 or 300 is the physical network card 400. For the sake of cleanliness, we will not repeat them here. Therefore, in the embodiment of the present invention, a physical network card has a PF module group composed of at least two PF modules, and one PF module can be selected as a main PF module from at least two PF modules in the PF module group, thereby Due to the many opportunities for selection, system failures can be avoided and resource waste can be reduced.

Those of ordinary skill in the art will appreciate that the elements and algorithm steps of the various examples described in connection with the embodiments disclosed herein can be implemented in a combination of electronic hardware or computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the solution. A person skilled in the art can use different methods for implementing the described functions for each particular application, but such implementation should not be considered to be beyond the scope of the present invention.

It will be apparent to those skilled in the art that, for the convenience of the description and the cleaning process, the specific operation of the system, the device and the unit described above may be referred to the corresponding processes in the foregoing method embodiments, and details are not described herein again.

In the several embodiments provided herein, it should be understood that the disclosed systems, devices, and methods may be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the unit is only a logical function division. In actual implementation, there may be another division manner, for example, multiple units or components may be combined or Can be integrated into another system, or some features can be ignored, or not executed. In addition, the coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection through some interface, device or unit, and may be electrical, mechanical or otherwise.

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

In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

The functions may be stored in a computer readable storage medium if implemented in the form of a software functional unit and sold or used as a standalone product. Based on such understanding, the technical solution of the present invention, or the part contributing to the prior art, or the part of the technical solution, may be embodied in the form of a software product stored in a storage medium. A number of instructions are included to cause a computer device (which may be a personal computer, server, or network device, etc.) to perform all or part of the steps of the methods described in various embodiments of the present invention. The foregoing storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and the like, which can store program codes. .

The above is only the specific embodiment of the present invention, but the scope of the present invention is not limited thereto, and any person skilled in the art can easily think of changes or substitutions within the technical scope of the present invention. It should be covered by the scope of the present invention. Therefore, the scope of the invention should be determined by the scope of the claims.

Claims

Rights request

1. A method for data transmission, characterized by including:

Determine the main PF module from the physical function PF module group of the first physical network card. The PF module group includes at least two PF modules, and each PF module in the PF module group corresponds to a different virtual function VF module group;

Based on the VF module group corresponding to each PF module in the PF module group, determine the target VF module group managed by the main PF module, wherein the VF modules in the target VF module group are used to be allocated to at least one virtual machine VM , so that the at least one VM implements data transmission.

2. The method according to claim 1, characterized in that: the slave first physical network card

In the PF module group, determine the main PF module, including:

The main PF module is determined based on the reference parameters of each PF module in the PF module group, where the reference parameters of the PF module include the number of PF module failures and the number of VF modules in the VF module group corresponding to the PF module. , at least one of PF module static bandwidth and PF module static bandwidth.

3. The method according to claim 1 or 2, characterized in that, based on the VF module group corresponding to each PF module in the PF module group, determining the target VF module group managed by the main PF module includes:

Select at least one VF module from the VF module group corresponding to each PF module in the PF module group to form the target VF module group.

4. The method according to any one of claims 1 to 3, characterized in that, the method further includes:

Determine whether the first PF module is re-enabled or fault restored, and determine the first PF module as the PF module in the PF module group;

Based on the first PF module, a VF module group corresponding to the first PF module is established.

5. The method according to any one of claims 1 to 4, characterized in that, the method further includes:

Determine that the second PF module is deleted from the PF module group of the first physical network card; release the VF module group corresponding to the second PF module.

6. The method according to any one of claims 1 to 5, characterized in that, the method further includes: When it is determined that the third PF module of the first physical network card fails, determine whether the VF module in the VF module group corresponding to the third PF module is available;

When at least one VF module in the VF module group corresponding to the third PF module is available, the at least one VF module available in the VF module group corresponding to the first PF module is determined to continue to be available.

7. The method according to any one of claims 1 to 6, characterized in that, the method further includes:

Determining that the first VF module assigned to the first VM in the at least one VM in the target VF module group fails;

Select a second VF module that is different from the first VF module from the VF module group corresponding to each PF module in the PF module group;

Update the target VF module group.

8. The method according to any one of claims 1 to 7, characterized in that the first physical network card occupies at least two high-speed peripheral component interconnection standard PCIe buses, and the method further includes: determining the need to The first of the target VF module group and the at least two PCIe buses

The data of the VF module connected to the PCIe bus is sent to the VF module in the target VF module group connected to the second PCIe bus of the at least two PCIe buses;

The data of the VF module connected to the first PCIe bus is sent to the VF module connected to the second PCIe bus through the switch of the first physical network card.

9. The method according to any one of claims 1 to 8, characterized in that, the method further includes:

It is determined that the main PF module is faulty;

From the PF module group, select a PF module other than the main PF module as the new main PF module;

Based on the VF module group corresponding to each PF module in the PF module group, determine the new primary

A new target VF module group managed by the PF module, wherein the VF modules in the new target VF module group are used to be assigned to at least one virtual machine VM, so that the at least one VM implements data transmission.

10. A device for data transmission, characterized in that it includes:

The first determination unit is used to determine the master module from the physical function PF module group of the first physical network card.

PF module, the PF module group includes at least two PF modules, each PF in the PF module group The modules correspond to different virtual function VF module groups;

The second determination unit is configured to determine the target VF module group managed by the main PF module based on the virtual function VF module group corresponding to each PF module in the PF module group, wherein, the VF module in the target VF module group Used to be allocated to at least one virtual machine VM so that the at least one VM implements data transmission.

11. The device according to claim 10, characterized in that the first determining unit is specifically used to:

The main PF module is determined based on the reference parameters of each PF module in the PF module group, where the reference parameters of the PF module include the number of PF module failures, the number of VF modules in the VF module group corresponding to the PF module. Number, at least one of PF module static bandwidth and PF module static bandwidth.

12. The device according to claim 10 or 11, characterized in that the second determination unit is specifically used to:

13. The method according to any one of claims 10 to 12, characterized in that the device further includes:

The fourth determination unit is used to determine the re-enabling or fault recovery of the first PF module, and determine the first PF module as the PF module in the PF module group;

An establishment unit, configured to establish, based on the first PF module, the corresponding

VF module set.

14. The device according to any one of claims 10 to 13, characterized in that the device further includes:

The fifth determination unit is used to determine that the second PF module is deleted from the PF module group of the first physical network card;

A release unit is used to release the VF module group corresponding to the second PF module.

15. The method according to any one of claims 10 to 14, characterized in that the device further includes:

The sixth determination unit is used to determine whether the VF in the VF module group corresponding to the third PF module is available when the third PF module of the first physical network card fails;

Holding unit, used for at least one VF in the VF module group corresponding to the third PF module When a module is available, the at least one VF module in the VF module group corresponding to the third PF module is determined to continue to be available.

16. The device according to any one of claims 10 to 15, characterized in that the device further includes:

A seventh determination unit, configured to determine that the first VF module assigned to the first VM in the at least one VM in the target VF module group fails;

A selection unit configured to select a second VF module different from the first VF module from the VF module group corresponding to each PF module in the PF module group;

An update unit, used to update the target VF module group.

17. The device according to any one of claims 10 to 16, wherein the first physical network card occupies at least two high-speed peripheral component interconnection standard PCIe buses, and the device further includes:

The eighth determination unit is used to determine that the data of the VF module in the target VF module group connected to the first PCIe bus of the at least two PCIe buses needs to be sent to the target VF module group connected to the at least two PCIe buses. The VF module connected to the second PCIe bus in the first PCIe bus;

A sending unit, configured to send the data of the VF module connected to the first PCIe bus to the VF module connected to the second PCIe bus through the switch of the first physical network card.

18. The device according to any one of claims 10 to 17, characterized in that the device further includes a third determination unit for determining that the main PF module is faulty;

The first determination unit is also used to: select a PF module other than the main PF module from the PF module group as a new main PF module;

The second determination unit is also configured to: determine a new target VF module group managed by the new main PF module based on the VF module group corresponding to each PF module in the PF module group, wherein, the new target The VF module in the VF module group is used to be assigned to at least one virtual machine VM, so that the at least one VM implements data transmission.

19. A device for data transmission, characterized in that it includes a processor and a memory; wherein the memory stores program code, and the processor is used to call the program code stored in the memory to perform the following operations :

Determine the main PF module from the physical function PF module group of the first physical network card. The PF module group includes at least two PF modules, and each PF module in the PF module group corresponds to a different virtual function VF module group; Based on the virtual function VF module group corresponding to each PF module in the PF module group, the target VF module group managed by the main PF module is determined, wherein the VF modules in the target VF module group are used to be allocated to at least A virtual machine VM, so that the at least one VM implements data transmission.

20. The device according to claim 19, wherein the processor is configured to call the program code stored in the memory, and specifically perform the following operations:

21. The device according to claim 19 or 20, characterized in that the processor is configured to call the program code stored in the memory, and specifically perform the following operations:

22. The device according to any one of claims 19 to 21, characterized in that the processor is configured to call the program code stored in the memory, and also perform the following operations:

23. The device according to any one of claims 19 to 22, characterized in that the processor is configured to call the program code stored in the memory, and also perform the following operations:

24. The device according to any one of claims 19 to 23, characterized in that the processor is configured to call the program code stored in the memory, and also perform the following operations:

When it is determined that the third PF module of the first physical network card fails, determine whether the VF module in the VF module group corresponding to the third PF module is available;

25. The device according to any one of claims 19 to 24, characterized in that: The processor is configured to call the program code stored in the memory, and also perform the following operations: determine that the first VF module assigned to the first VM in the at least one VM in the target VF module group fails;

Update the target VF module group.

26. The device according to any one of claims 19 to 25, wherein the first physical network card occupies at least two high-speed peripheral component interconnection standard PCIe buses, and the processor is used to call the memory The program code stored in also performs the following operations:

Determine the need to place the target VF module in the cluster with the at least two PCIe buses in the first

27. The method according to any one of claims 19 to 26, characterized in that the processor is configured to call the program code stored in the memory, and also perform the following operations:

It is determined that the main PF module is faulty;

A new target VF module group managed by the PF module, wherein the VF modules in the target VF module group are used to be allocated to at least one virtual machine VM, so that the at least one VM implements data transmission.

28. A physical network card, characterized in that it includes: a PF module group composed of at least two PF modules, and at least two VF module groups; wherein, one of the at least two PF modules

The PF module corresponds to one VF module group among the at least two VF module groups, and any PF module among the at least two PF modules can manage VF module groups connected to other PF modules after being determined as the main PF module. VF module in .

29. The physical network card according to claim 28, wherein the VF modules in each of the at least two VF module groups include two types of space resources; wherein,

One type of space resource among the two types of space resources is used before the main PF module is determined. Accessed by the PF module corresponding to the VF module group to which the VF module of the two types of space resources belongs;

The other type of space resource among the two types of space resources is used to establish a connection with the main PF module for access by the main PF module after the main PF module is determined.

30. The physical network card according to claim 28 or 29, characterized in that, the physical network card occupies at least two high-speed peripheral components interconnecting standard PCIe buses, and the physical network card also includes a built-in switch; wherein, the built-in The switch is configured to transmit data of the VF module connected to the first PCIe bus of the at least two PCIe buses to the VF module connected to the second PCIe bus of the at least two PCIe buses.

31. The physical network card according to any one of claims 28 to 31, characterized in that, the physical network card is connected to an external switch through at least two ports; the physical network card is also used to receive data transmission through an aggregation chain. The data stream is sent through the port, and the data stream is evenly sent to the external switch through each port of the at least two ports.

32. A system for data transmission, characterized by comprising the device for data transmission according to any one of claims 10 to 18 and the physical network card according to any one of claims 28 to 31 , or include the device for data transmission according to any one of claims 19 to 27 and the physical network card according to any one of claims 28 to 31; wherein, any one of claims 10 to 18 The first physical network card referred to in the claims or the first physical network card referred to in any one of claims 19 to 27 is the physical network card described in any one of claims 28 to 31.