An address management method, device and related equipment

Abstract

The application relates to the technical field of network communication, in particular to an address management method and device and related equipment. The method is applied to an access device, a DHCP service is integrated on the access device, and an IP address pool corresponding to each network card of a GPU server is preconfigured on the access device accessed by the network card; the method comprises the following steps: receiving an IP address request message sent by a network card of a GPU server, wherein the IP address request message carries MAC address information and network card type information of the network card; judging the network card type of the network card based on the network card type information of the network card; if it is judged that the network card type of the network card is a network card needing to be configured with a static IP address, allocating an IP address for the network card from the IP address pool corresponding to the network card; and maintaining the association relationship between the MAC address of the network card and the IP address into a network card association table needing to be statically configured with an IP address.

Description

Technical Field

[0001] This application relates to the field of network communication technology, and in particular to an address management method, apparatus and related equipment. Background Technology

[0002] In intelligent computing centers, the IP addresses of GPU servers' network interface cards (NICs) have certain fixed requirements. Changes in the IP addresses of storage NICs can cause anomalies in storage access, and changes in the IP addresses of parameter NICs can affect computation calls. In the ever-growing intelligent computing data centers, as computing resources increase, network scale also increases. Server parameter NICs and storage NICs with multiple network interfaces have become the norm. Simultaneously, with the development of virtualization technology in computing platforms, the network scale within containerized devices (PODs) has also grown, increasing the overall number of IP addresses in the network. Static IP configuration strategies are no longer sufficient to manage the network addresses of intelligent computing centers. Summary of the Invention

[0003] This application provides an address management method, apparatus, and related equipment.

[0004] In a first aspect, this application provides an address management method applied to an access device, wherein the access device integrates a DHCP service, and each network card of the GPU server is connected to the access device, which is pre-configured with an IP address pool corresponding to that network card; the method includes:

[0005] Receive an IP address request message sent by the network card of the GPU server, wherein the IP address request message carries the MAC address information and network card type information of the network card;

[0006] Based on the network card type information of the network card, determine the network card type. If the network card type is determined to be a network card that requires a static IP address, then allocate an IP address to the network card from the IP address pool corresponding to the network card.

[0007] The association between the MAC address and the IP address of the network card is maintained in the network card association table that requires static IP address configuration.

[0008] Optionally, after receiving the IP address request packet sent by the network card of the GPU server, and before determining the network card type based on the network card type information of the network card, the method further includes:

[0009] Based on the MAC address of the network card, determine whether there is an entry corresponding to the MAC address in the locally maintained network card association table that requires static IP address configuration;

[0010] If it is determined that there is an entry corresponding to the MAC address, then the IP address included in that entry is assigned to the network card;

[0011] If it is determined that there is no entry in the table corresponding to the MAC address, then the step of determining the network card type based on the network card type information of the network card is executed.

[0012] Optionally, the method further includes:

[0013] If it is determined that the network card type is not a network card that requires a static IP address, then an IP address is allocated to the network card from the target IP address pool corresponding to the network card.

[0014] Optionally, if an IP address in the IP address pool has been assigned to a network interface card (NIC) that requires a static IP address, the IP address is marked as occupied and will not be assigned to other NICs in the future.

[0015] Optionally, the network interface card (NIC) type includes a storage NIC, a parameter NIC, or a container NIC, wherein the storage NIC is defined as a NIC that requires a static IP address to be configured.

[0016] The network interface card (NIC) type information is carried in the sub-option type field of the custom optiontype field in the address request message. If the value of the sub-option type field in the custom optiontype field is 0x01, it indicates that the NIC is a parameter NIC. If the value of the sub-option type field in the custom optiontype field is 0x02, it indicates that the NIC is a storage NIC. If the value of the sub-option type field in the custom optiontype field is 0x04, it indicates that the NIC is a container NIC.

[0017] Secondly, this application provides an address management device applied to an access device, wherein the access device integrates a DHCP service, and each network card of the GPU server is connected to the access device, which is pre-configured with an IP address pool corresponding to that network card; the device includes:

[0018] The receiving unit is used to receive an IP address request message sent by the network card of the GPU server, wherein the IP address request message carries the MAC address information and network card type information of the network card;

[0019] The determination unit is used to determine the network card type of the network card based on the network card type information of the network card;

[0020] If the determination unit determines that the network card type is a network card that requires a static IP address, the allocation unit is used to allocate an IP address to the network card from the IP address pool corresponding to the network card.

[0021] The maintenance unit is used to maintain the association between the MAC address and the IP address of the network card in the network card association table that requires static IP address configuration.

[0022] Optionally, the device determining unit is further configured to:

[0023] Based on the MAC address of the network card, determine whether there is an entry corresponding to the MAC address in the locally maintained network card association table that requires static IP address configuration;

[0024] If the determination unit determines that there is an entry corresponding to the MAC address, the allocation unit is used to allocate the IP address included in the entry to the network card;

[0025] If the determination unit determines that there is no entry corresponding to the MAC address, it executes the step of determining the network card type based on the network card type information of the network card.

[0026] Optionally, if the determining unit determines that the network card type is not a network card that requires a static IP address, the allocation unit is used to allocate an IP address to the network card from the target IP address pool corresponding to the network card.

[0027] Optionally, if an IP address in the IP address pool has been assigned to a network interface card (NIC) that requires a static IP address, the IP address is marked as occupied and will not be assigned to other NICs in the future.

[0028] Optionally, the network interface card (NIC) type includes a storage NIC, a parameter NIC, or a container NIC, wherein the storage NIC is defined as a NIC that requires a static IP address to be configured.

[0029] The network interface card (NIC) type information is carried in the sub-option type field of the custom optiontype field in the address request message. If the value of the sub-option type field in the custom optiontype field is 0x01, it indicates that the NIC is a parameter NIC. If the value of the sub-option type field in the custom optiontype field is 0x02, it indicates that the NIC is a storage NIC. If the value of the sub-option type field in the custom optiontype field is 0x04, it indicates that the NIC is a container NIC.

[0030] Thirdly, embodiments of this application provide an address management device, which includes:

[0031] Memory, used to store program instructions;

[0032] A processor is configured to invoke program instructions stored in the memory and execute the steps of the method as described in any one of the first aspects above, according to the obtained program instructions.

[0033] Fourthly, embodiments of this application also provide a computer-readable storage medium storing computer-executable instructions for causing a computer to perform the steps of the method as described in any of the first aspects above.

[0034] In summary, the address management method provided in this application is applied to an access device, which integrates a DHCP service. Each network card of the GPU server is connected to an access device that is pre-configured with an IP address pool corresponding to that network card. The method includes: receiving an IP address request message sent by the network card of the GPU server, wherein the IP address request message carries the MAC address information and network card type information of the network card; determining the network card type based on the network card type information; if the network card type is determined to be a network card that requires a static IP address, then allocating an IP address to the network card from the IP address pool corresponding to the network card; and maintaining the association between the MAC address and the IP address of the network card in a network card association table that requires a static IP address.

[0035] The address management method provided in this application ensures the static requirement that the storage network address cannot be changed arbitrarily through the custom options of DHCP. Furthermore, the DHCP address allocation avoids the tediousness of manually configuring IP addresses and the problem of address conflicts. The DHCP address allocation also ensures that the host IP and the IP in the container are on the same network segment, and avoids conflicts between the host IP and the IP in the container. Attached Figure Description

[0036] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments of this application or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this application. For those skilled in the art, other drawings can be obtained based on these drawings of the embodiments of this application.

[0037] Figure 1 A detailed flowchart of an address management method provided in an embodiment of this application;

[0038] Figure 2 A schematic diagram of a smart computing center network is provided for an embodiment of this application;

[0039] Figure 3 A schematic diagram of a custom optiontype format provided for an embodiment of this application;

[0040] Figure 4 Another intelligent computing center networking diagram provided for embodiments of this application;

[0041] Figure 5 This is a schematic diagram of the structure of an address management device provided in an embodiment of this application;

[0042] Figure 6 This is a schematic diagram of the hardware architecture of an address management device provided in an embodiment of this application. Detailed Implementation

[0043] The terminology used in the embodiments of this application is for the purpose of describing particular embodiments only and is not intended to limit the application. The singular forms “a,” “the,” and “the” as used in this application and claims are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used herein refers to any and all possible combinations comprising one or more of the associated listed items.

[0044] It should be understood that although the terms first, second, third, etc., may be used to describe various information in embodiments of this application, such information should not be limited to these terms. These terms are only used to distinguish information of the same type from one another. For example, without departing from the scope of this application, first information may also be referred to as second information, and similarly, second information may also be referred to as first information. Depending on the context, the word "if" may also be interpreted as "when," "when," or "in response to a determination."

[0045] Currently, servers in intelligent computing typically support multiple parameter network cards and storage network cards. In order to connect the parameter network and storage network to the network, after completing the PXE installation, users need to manually plan and configure the IP addresses of the server's multiple network cards. The configuration work is cumbersome and the address conflict problem is serious.

[0046] Taking a 4K GPU-based intelligent computing center as an example, deploying servers with 8 network interface cards (NICs), the entire network has 512 servers, corresponding to 4K NICs and 1K storage NICs. If a virtualization computing platform is used, with one GPU server running 8 compute devices (PODs), each POD requires 8 IP addresses. This means that a single server's container network would require 8 * 8 = 64 IP addresses. In summary, a 4K GPU-based intelligent computing center would need to manage a total of:

[0047] 4K (parameter network card address) + 1K (storage network card address) + 32K (address inside container) = 37K addresses;

[0048] Configuring such a large number of addresses manually is time-consuming, labor-intensive, and takes an extremely long time to deploy, and it is also prone to address conflicts.

[0049] In intelligent computing centers, the IP addresses of the network interface cards (NICs) of each server have certain fixed requirements. If the IP address of the storage NIC changes, it will cause anomalies in storage access; if the IP address of the parameter NIC changes, it will affect computation calls. This application provides a solution for automatically allocating IP addresses in intelligent computing centers and fixing NIC IP addresses according to user needs.

[0050] For example, see Figure 1 The diagram shown is a detailed flowchart of an address management method provided in an embodiment of this application. This method is applied to an access device, which integrates a DHCP service. Each network card of the GPU server is connected to an access device pre-configured with an IP address pool corresponding to that network card. The method includes the following steps:

[0051] Step 100: Receive an IP address request message sent by the network card of the GPU server, wherein the IP address request message carries the MAC address information and network card type information of the network card.

[0052] For example, see Figure 2 The diagram shown is a network topology diagram of a smart computing center provided in an embodiment of this application. The parameter network cards (e.g., ens1, ens2, ens3 and ens4) of the GPU server are connected to the access devices of the computing fabric (e.g., Leaf-1, Leaf-2, Leaf-3 and Leaf-4), and the storage network cards (e.g., ens5 and ens6) of the GPU server are connected to the access devices of the storage fabric (Leaf-5 and Leaf-6). When the GPU server is connected to the network, IP addresses need to be assigned to each parameter network card and the storage network card.

[0053] Under the above networking conditions, the ADDC controller (not shown in the figure) enables DHCP server and plans network segments for the access devices. That is, DHCP service is integrated on the access devices, and each access device is a DHCP server, which automatically assigns IP addresses to the network cards of the GPU server connected to the network card.

[0054] Specifically, in practical applications, a Leaf device can connect to one network card of a GPU server. Of course, it can also connect to multiple network cards of the GPU server (through different physical ports). When the DHCP server performs network segment planning, it can configure corresponding network segments (address pools) for each connected network card based on user needs, or it can configure one network segment for multiple network card devices with the same function. In this embodiment of the application, no specific limitation is made here.

[0055] For example, such as Figure 2As shown, network segment 1 is configured for ens1 on Leaf-1, network segment 2 is configured for ens2 on Leaf-2, network segment 3 is configured for ens3 on Leaf-3, network segment 4 is configured for ens4 on Leaf-4, network segment 5 is configured for ens5 on Leaf-5, and network segment 6 is configured for ens6 on Leaf-6.

[0056] When the GPU server comes online, each network interface card (NIC) sends an address request message to its corresponding Leaf device. This address request message can carry the MAC address of the NIC and the NIC type (e.g., parameter NIC, storage NIC, etc.).

[0057] In this embodiment of the application, the network card type information of a network card is carried in the sub-option type field of the custom optiontype field of the address request message. If the value of the sub-option type field in the custom optiontype field is 0x01, it indicates that the network card is a parameter network card. If the value of the sub-option type field in the custom optiontype field is 0x02, it indicates that the network card is a storage network card. If the value of the sub-option type field in the custom optiontype field is 0x04, it indicates that the network card is a container network card.

[0058] For example, see Figure 3 The diagram shown is a custom option type format provided in the embodiment of this application. Specifically, the sub-option type uses 0x01 to identify the network card requesting an IP address, 0x02 to indicate the storage network card requesting an IP address, and 0x04 to indicate the container network card requesting an IP address.

[0059] When the sub-option is 0x01, the sub-option variable carries the network card's MAC address;

[0060] When the sub-option is 0x02, the sub-option variable carries the MAC address of the storage network card; the DHCP server needs to maintain the relationship between the MAC address of the storage network card and the assigned IP address (the storage network card is defined as a network card that needs to be assigned a static IP address).

[0061] When the sub-option is 0x04, the sub-option variable carries the container POD name and the MAC address of the container's network card.

[0062] In this embodiment, for network cards requiring static IP addresses, after allocating an IP address to the network card, the access device locally maintains the association between the network card's MAC address and the allocated IP address, so that the network card can still be allocated an IP address again in the future. Therefore, in this embodiment, after receiving the IP address request message sent by the GPU server's network card, the above method may further include the following steps:

[0063] Based on the MAC address of the network card, determine whether there is an entry corresponding to the MAC address in the locally maintained network card association table that requires static IP address configuration;

[0064] If it is determined that there is an entry corresponding to the MAC address, then the IP address included in that entry is assigned to the network card;

[0065] If it is determined that there is no entry in the table corresponding to the MAC address, then the step of determining the network card type based on the network card type information of the network card is executed.

[0066] In other words, first check the IP address-MAC address association table to see if an IP address has been assigned to the network card. If it has, then the network card is one that needs to be configured with a static IP address. In this case, simply reassign the IP address corresponding to the MAC address of the network card in the association table to the network card.

[0067] If it does not exist, then determine that the network card is applying for an IP address for the first time, or is not a network card that needs to be configured with a static IP address, and proceed to the next step 110.

[0068] Step 110: Determine the network card type based on the network card type information. If the network card type is determined to be a network card that requires a static IP address, then allocate an IP address to the network card from the IP address pool corresponding to the network card.

[0069] In this embodiment, the network interface card (NIC) type includes a storage NIC, a parameter NIC, or a container NIC, wherein the storage NIC is defined as a NIC that requires a static IP address. Of course, parameter NICs or container NICs can also be defined as NICs that require a static IP address, depending on user needs; however, this embodiment does not impose specific limitations.

[0070] The access device parses the address request message, obtains the value of the field used to characterize the network interface card (NIC) type information, and determines the NIC type based on this value.

[0071] If the network card is determined to be a storage network card, then it is determined to be a network card that requires a static IP address, and an available IP address is assigned to the network card from the network segment corresponding to the network card.

[0072] Step 120: Maintain the association between the MAC address and the IP address of the network card in the network card association table that requires static IP address configuration.

[0073] Furthermore, since this network interface card (NIC) requires a static IP address, after assigning an IP address to it, the association between the NIC's MAC address and IP address is added to the locally maintained NIC association table for NICs requiring IP address configuration. If the NIC goes offline and then comes back online, the IP address can be reassigned to it based on this association.

[0074] In this embodiment of the application, if an IP address in the IP address pool has been assigned to a network card that needs to be configured with a static IP address, the IP address is marked as occupied and will not be assigned to other network cards in the future.

[0075] In other words, network card 1 is a network card that needs to be configured with a static IP address. After IP1 is assigned to network card 1, IP1 will not be assigned to other network cards. Even if the network card goes offline and the IP1 address is reclaimed, it will not be assigned to other network cards to avoid the situation where IP1 is occupied by other network cards when the network card comes online again.

[0076] Specifically, in the address pool, if an IP address has already been assigned to a network interface card (NIC) that requires a static IP address, then that IP address will be marked as occupied. Even if the IP address is reclaimed, it will still be marked as allocated to a NIC. Other NICs will not be able to obtain that IP address when requesting an IP address from the address pool. Of course, if the user issues a command to restore the IP address, then the IP address will be allocated to another NIC when it is not in use.

[0077] If it is determined that the network card type is not a network card that requires a static IP address, then an IP address is allocated to the network card from the target IP address pool corresponding to the network card.

[0078] In other words, for network cards that do not require a static IP address, it is sufficient to allocate an available IP address from the address pool corresponding to the network card, without needing to maintain the association between the network card's MAC address and IP address.

[0079] In this embodiment, the container is a container for calculating runtime parameters. Therefore, in the container network, the container network interface card (NIC) and the parameter NIC can be in one-to-one correspondence. For example, assuming the container NIC includes 8 NICs, then the GPU server includes 8 parameter NICs that correspond one-to-one with the 8 NICs of the container NIC. Binding the container NIC to its corresponding parameter NIC means that when planning network segments on the access device, assuming parameter NIC 1 corresponds to container NIC 1, then the corresponding network segment can be planned based on parameter NIC 1 and container NIC 1, and parameter NIC 2 corresponds to container NIC 2. Similarly, the corresponding network segment can be planned based on parameter NIC 1 and container NIC 1. For example, see [link to relevant documentation]. Figure 4 As shown, this is a schematic diagram of the intelligent computing center network provided in the embodiment of this application. ens1 is a parameter network card, which is connected to Leaf-1. The IP address obtained by the container network card corresponding to ens1 and the IP address obtained by ens1 are in the same network segment.

[0080] In other words, the parameter network card and its corresponding container network card obtain IP addresses from the same network segment, ensuring that the IP addresses obtained by the parameter network card and its corresponding container network card are from the same network segment.

[0081] For example, see Figure 5 The diagram shown is a structural schematic of an address management device provided in an embodiment of this application. This device is applied to an access device, which integrates a DHCP service. Each network card of the GPU server is connected to an access device pre-configured with an IP address pool corresponding to that network card. The device includes:

[0082] The receiving unit 50 is used to receive an IP address request message sent by the network card of the GPU server, wherein the IP address request message carries the MAC address information and network card type information of the network card;

[0083] The judgment unit 51 is used to determine the network card type of the network card based on the network card type information of the network card;

[0084] If the determination unit 51 determines that the network card type is a network card that requires a static IP address, the allocation unit 52 is used to allocate an IP address to the network card from the IP address pool corresponding to the network card.

[0085] The maintenance unit 53 is used to maintain the association between the MAC address and the IP address of the network card in the network card association table that requires static IP address configuration.

[0086] Optionally, the device determining unit 51 is further configured to:

[0087] Based on the MAC address of the network card, determine whether there is an entry corresponding to the MAC address in the locally maintained network card association table that requires static IP address configuration;

[0088] If the determination unit 51 determines that there is an entry corresponding to the MAC address, the allocation unit 52 is used to allocate the IP address included in the entry to the network card.

[0089] If the determination unit 51 determines that there is no entry corresponding to the MAC address, then it performs the step of determining the network card type based on the network card type information of the network card.

[0090] Optionally, if the determination unit 51 determines that the network card type is not a network card that requires a static IP address, then the allocation unit 52 is used to allocate an IP address to the network card from the target IP address pool corresponding to the network card.

[0091] Optionally, if an IP address in the IP address pool has been assigned to a network interface card (NIC) that requires a static IP address, the IP address is marked as occupied and will not be assigned to other NICs in the future.

[0092] Optionally, the network interface card (NIC) type includes a storage NIC, a parameter NIC, or a container NIC, wherein the storage NIC is defined as a NIC that requires a static IP address to be configured.

[0093] The network interface card (NIC) type information is carried in the sub-option type field of the custom optiontype field in the address request message. If the value of the sub-option type field in the custom optiontype field is 0x01, it indicates that the NIC is a parameter NIC. If the value of the sub-option type field in the custom optiontype field is 0x02, it indicates that the NIC is a storage NIC. If the value of the sub-option type field in the custom optiontype field is 0x04, it indicates that the NIC is a container NIC.

[0094] These units can be one or more integrated circuits configured to implement the above methods, such as one or more Application Specific Integrated Circuits (ASICs), one or more digital signal processors (DSPs), or one or more Field Programmable Gate Arrays (FPGAs). Alternatively, when one of these units is implemented using processing element scheduler code, the processing element can be a general-purpose processor, such as a Central Processing Unit (CPU) or other processor capable of calling program code. Furthermore, these units can be integrated together to form a system-on-a-chip (SOC).

[0095] Furthermore, regarding the address management device provided in this application embodiment, from a hardware perspective, the hardware architecture diagram of the address management device can be found in [reference needed]. Figure 6 As shown, the address management device may include: a memory 60 and a processor 61.

[0096] The memory 60 is used to store program instructions; the processor 61 calls the program instructions stored in the memory 60 and executes the above method embodiment according to the obtained program instructions. The specific implementation method and technical effect are similar, and will not be described again here.

[0097] Optionally, this application also provides an access device, including at least one processing element (or chip) for performing the above method embodiments.

[0098] Optionally, this application also provides a program product, such as a computer-readable storage medium storing computer-executable instructions for causing the computer to perform the above-described method embodiments.

[0099] Here, a machine-readable storage medium can be any electronic, magnetic, optical, or other physical storage device that can contain or store information, such as executable instructions, data, etc. For example, a machine-readable storage medium can be: RAM (Random Access Memory), volatile memory, non-volatile memory, flash memory, storage drives (such as hard disk drives), solid-state drives, any type of storage disk (such as optical discs, DVDs, etc.), or similar storage media, or combinations thereof.

[0100] The systems, devices, modules, or units described in the above embodiments can be implemented by computer chips or entities, or by products with certain functions. A typical implementation device is a computer, which can take the form of a personal computer, laptop computer, cellular phone, camera phone, smartphone, personal digital assistant, media player, navigation device, email sending and receiving device, game console, tablet computer, wearable device, or any combination of these devices.

[0101] For ease of description, the above devices are described separately by function as various units. Of course, in implementing this application, the functions of each unit can be implemented in one or more software and / or hardware.

[0102] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, embodiments of this application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0103] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0104] Furthermore, these computer program instructions can also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to operate in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in the process. Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0105] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0106] The above description is merely a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. An address management method characterized by comprising: The method is applied to access devices, which integrate DHCP services, and each network card of the GPU server is connected to an access device pre-configured with an IP address pool corresponding to that network card; the method includes: Receive an IP address request message sent by the network card of the GPU server, wherein the IP address request message carries the MAC address information and network card type information of the network card; Based on the network card type information of the network card, the network card type is determined. If the network card type is determined to be a network card that requires a static IP address, then an IP address is allocated to the network card from the IP address pool corresponding to the network card. The network card type includes storage network card, parameter network card, or container network card, wherein the storage network card is defined as a network card that requires a static IP address. The association between the MAC address and the IP address of the network card is maintained in the network card association table that requires static IP address configuration.

2. The method of claim 1, wherein, After receiving the IP address request packet sent by the network card of the GPU server, and before determining the network card type based on the network card type information, the method further includes: Based on the MAC address of the network card, determine whether there is an entry corresponding to the MAC address in the locally maintained network card association table that requires static IP address configuration; If it is determined that there is an entry corresponding to the MAC address, then the IP address included in that entry is assigned to the network card; If it is determined that there is no entry in the table corresponding to the MAC address, then the step of determining the network card type based on the network card type information of the network card is executed.

3. The method as described in claim 1 or 2, characterized in that, The method further includes: If it is determined that the network card type is not a network card that requires a static IP address, then an IP address is allocated to the network card from the target IP address pool corresponding to the network card.

4. The method as described in claim 1 or 2, characterized in that, If an IP address in the IP address pool has already been assigned to a network interface card (NIC) that requires a static IP address, then that IP address will be marked as occupied and will not be assigned to other NICs in the future.

5. The method as described in claim 1 or 2, characterized in that, The network interface card (NIC) type information is carried in the sub-option type field of the custom optiontype field in the address request message. If the value of the sub-option type field in the custom optiontype field is 0x01, it indicates that the NIC is a parameter NIC. If the value of the sub-option type field in the custom optiontype field is 0x02, it indicates that the NIC is a storage NIC. If the value of the sub-option type field in the custom optiontype field is 0x04, it indicates that the NIC is a container NIC.

6. An address management device, characterized in that, It is applied to access devices, which integrate DHCP service. Each network card of the GPU server is connected to an access device that is pre-configured with an IP address pool corresponding to that network card. The device includes: The receiving unit is used to receive an IP address request message sent by the network card of the GPU server, wherein the IP address request message carries the MAC address information and network card type information of the network card; The determination unit is used to determine the network card type of the network card based on the network card type information of the network card. The network card type includes storage network card, parameter network card, or container network card, wherein the storage network card is defined as a network card that needs to be configured with a static IP address. If the determination unit determines that the network card type is a network card that requires a static IP address, the allocation unit is used to allocate an IP address to the network card from the IP address pool corresponding to the network card. The maintenance unit is used to maintain the association between the MAC address and the IP address of the network card in the network card association table that requires static IP address configuration.

7. The apparatus as claimed in claim 6, characterized in that, The device determination unit is also used for: Based on the MAC address of the network card, determine whether there is an entry corresponding to the MAC address in the locally maintained network card association table that requires static IP address configuration; If the determination unit determines that there is an entry corresponding to the MAC address, the allocation unit is used to allocate the IP address included in the entry to the network card; If the determination unit determines that there is no entry corresponding to the MAC address, it executes the step of determining the network card type based on the network card type information of the network card.

8. The apparatus as claimed in claim 6 or 7, characterized in that, If the determination unit determines that the network card type is not a network card that requires a static IP address, then the allocation unit is used to allocate an IP address to the network card from the target IP address pool corresponding to the network card.

9. The apparatus as claimed in claim 6 or 7, characterized in that, If an IP address in the IP address pool has already been assigned to a network interface card (NIC) that requires a static IP address, then that IP address will be marked as occupied and will not be assigned to other NICs in the future.

10. The apparatus as claimed in claim 6 or 7, characterized in that, The network interface card (NIC) type information is carried in the sub-option type field of the custom optiontype field in the address request message. If the value of the sub-option type field in the custom optiontype field is 0x01, it indicates that the NIC is a parameter NIC. If the value of the sub-option type field in the custom optiontype field is 0x02, it indicates that the NIC is a storage NIC. If the value of the sub-option type field in the custom optiontype field is 0x04, it indicates that the NIC is a container NIC.

11. An address management device, characterized in that, The address management device includes: Memory, used to store program instructions; A processor is configured to invoke program instructions stored in the memory and execute the steps of the method as described in any one of claims 1-5 according to the obtained program instructions.

12. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions for causing the computer to perform the steps of the method as described in any one of claims 1-5.

Images