A logical node configuration method based on a cloud management platform and a cloud management platform

Abstract

The application discloses a cloud management platform-based logical node configuration method and a cloud management platform, and can improve the utilization rate of resources occupied by a logical node. The method comprises the following steps: a tenant can send a creation request for a logical node to a cloud management platform, the request being used for indicating the computing resource requirement, the storage resource requirement and the network resource requirement of the logical node, so that the cloud management platform can select physical servers satisfying the computing resource requirement and the storage resource requirement, select intelligent network cards satisfying the network resource requirement, and select SPU for managing the physical servers and the intelligent network cards. Then, the cloud management platform can inform the SPU to create the logical node on the physical servers and the intelligent network cards. Since the intelligent network cards contain port resources one-to-one bound with the physical servers, the port resources can be used for forwarding messages from or to the logical node.

Description

Technical Field

[0001] This application relates to the field of cloud technology, and in particular to a logical node configuration method based on a cloud management platform and a cloud management platform. Background Technology

[0002] With the rapid development of cloud technology, more and more tenants are choosing to use cloud resources provided by cloud providers to complete their business. These cloud resources may include computing resources, storage resources, and network resources provided by physical servers in the cloud. Therefore, tenants can create cloud instances that meet their requirements on these cloud resources to satisfy their business needs.

[0003] In related technologies, cloud resources provided by cloud providers to tenants are typically located on different physical servers. Generally, physical servers can contain various resources such as processors, memory, and network interface cards (NICs). To improve the utilization of these resources, cloud providers can insert additional offload cards into the physical servers to offload some tasks of the processor and NICs to the offload cards for execution. In other words, the offload card can replace the processor to perform resource management functions, and can also replace the NIC to perform data forwarding functions, etc.

[0004] In the above process, since cloud providers can configure corresponding offload cards for several physical servers used to create cloud instances, not every offload card on each physical server needs to perform its corresponding function during the operation of the cloud instance, which leads to low resource utilization of the offload cards. Summary of the Invention

[0005] This application provides a logical node configuration method based on a cloud management platform and a cloud management platform, which can improve the utilization rate of resources occupied by logical nodes.

[0006] The first aspect of this application provides a logical node configuration method based on a cloud management platform. The cloud management platform implementing this method can manage the infrastructure that provides cloud services to tenants. This infrastructure includes a physical server pool, a smart NIC pool, and a service process unit (SPU) pool. Multiple physical servers in the physical server pool, multiple smart NICs in the smart NIC pool, and multiple SPUs in the SPU pool are interconnected within and between pools via a high-speed interconnection network. Each SPU can be used to manage one or more physical servers and one or more smart NICs among the multiple physical servers.

[0007] The method includes:

[0008] When a tenant needs to create its logical node, the tenant can submit a logical node creation request to the creation interface provided by the cloud management platform. In this way, the cloud management platform can receive the logical node creation request sent by the tenant through the creation interface. The logical node creation request includes the resource requirements set by the tenant for the logical node, including the computing resource requirements, storage resource requirements, and network resource requirements for the logical node.

[0009] Upon receiving a logical node creation request, the cloud management platform can determine the tenant's computing resource requirements, storage resource requirements, and network resource requirements for the logical node based on the request. Therefore, the cloud management platform can select at least one physical server from multiple physical servers that meets the computing resource requirements and the storage resource requirements, select at least one smart network interface card from multiple smart network interface cards that meets the network resource requirements, and select at least one SPU from multiple SPUs to manage the at least one physical server and the at least one smart network interface card.

[0010] After identifying at least one SPU, at least one physical server, and at least one smart NIC, the cloud management platform can instruct the at least one SPU to create a tenant's logical node within the at least one physical server and the at least one smart NIC. For any one of the at least one smart NICs occupied by the logical node, the smart NIC contains at least one port resource bound one-to-one with the at least one physical server occupied by the logical node. These port resources can forward packets from or directed to the logical node to meet the tenant's business needs.

[0011] As can be seen from the above method, since each SPU in the SPU pool can be used to manage at least one physical server in the physical server pool and at least one smart NIC in the smart NIC pool, and each smart NIC in the smart NIC pool can provide data forwarding services for at least one physical server in the physical server pool, regardless of the number of physical servers occupied by a tenant's logical node, a smaller number of SPUs and smart NICs can be used to manage and support these physical servers. This makes the logical node more lightweight overall. During the operation of the logical node, there are fewer idle SPUs and smart NICs and more working SPUs and smart NICs, thereby improving the resource utilization rate of SPUs and smart NICs.

[0012] In one possible implementation, at least one port resource of the smart network interface card (NIC) is allocated from the same port of the smart NIC, or at least one port resource of the smart NIC is allocated from different ports of the smart NIC. In the aforementioned implementation, for any one of the at least one smart NICs occupied by the logical node, the smart NIC may have multiple ports. Therefore, one or more of these ports can be allocated as at least one port resource, and this at least one port resource can be bound one-to-one with the at least one physical server occupied by the logical node. In this way, the different port resources of the smart NIC can be allocated to the different physical servers occupied by the logical node. Therefore, even if there is a sudden increase in traffic on a certain physical server, the forwarding is limited only to the port resources of the smart NIC bound to that physical server, and will not affect the traffic forwarding of other physical servers, thus ensuring the stability of the tenant's business.

[0013] In one possible implementation, at least one SPU includes a first SPU, and at least one smart network interface card (NIC) includes the first smart NIC. The method further includes: a cloud management platform determining that the first smart NIC is faulty, notifying the first SPU to select a second smart NIC from the smart NICs managed by the first SPU, wherein the second smart NIC does not belong to at least one smart NIC; the cloud management platform notifying the first SPU to release the first binding relationship between at least one port resource of the first smart NIC and at least one physical server, establishing a second binding relationship between at least one port resource of the second smart NIC and at least one physical server, and controlling at least one port resource of the second smart NIC to re-forward a first packet based on the second binding relationship, wherein the first packet is the packet that at least one port resource of the first smart NIC was forwarding when the first smart NIC was faulty. In the aforementioned implementation, the at least one smart NIC occupied by the logical node includes the first smart NIC, and the first SPU in the at least one SPU is used to manage the first smart NIC. If the first smart network interface card (NIC) is determined to be faulty, the cloud management platform can instruct the first SPU to select a second smart NIC from the smart NICs managed by the first SPU to replace the first smart NIC. This second smart NIC is not part of the at least one smart NIC occupied by the logical node. Since the first SPU records a first binding relationship between at least one port resource of the first smart NIC and the at least one physical server occupied by the logical node, and the first smart NIC is faulty, the cloud management platform can instruct the first SPU to release the first binding relationship and then re-establish a second binding relationship between at least one port resource of the second smart NIC and the at least one physical server occupied by the logical node. Because at least one port resource of the first smart NIC is usually forwarding the first packet when the first smart NIC is faulty, the cloud management platform can instruct the first SPU to control at least one port resource of the second smart NIC to re-forward the first packet based on the second binding relationship. Therefore, for any one of the at least one smart network interface cards (NICs) occupied by a tenant's logical node, if the smart NIC fails, the cloud management platform can notify the SPU managing the smart NIC to promptly select a new smart NIC to replace it, so as to complete the packet forwarding task that the smart NIC should have originally performed, thereby ensuring the continuity and stability of the tenant's business.

[0014] In one possible implementation, at least one SPU includes a first SPU and a second SPU, and at least one smart network interface card (NIC) includes a third smart NIC and a fourth smart NIC. The first SPU records a third binding relationship between at least one port resource of the third smart NIC and at least one physical server, and the second SPU records a fourth binding relationship between at least one port resource of the fourth smart NIC and at least one physical server. The method further includes: a cloud management platform determining that the first SPU is faulty, migrating the third binding relationship between at least one port resource of the third smart NIC and at least one physical server from the first SPU to the second SPU; and the cloud management platform notifying the second SPU to control at least one port resource of the third smart NIC to re-forward a second message based on the third binding relationship, wherein the second message is the message that at least one port resource of the third smart NIC was forwarding when the first SPU was faulty. In the aforementioned implementation, the at least one smart network interface card (NIC) occupied by the logical node includes a third smart NIC and a fourth smart NIC. The at least one system unit (SPU) includes a first SPU for managing the third smart NIC and a second SPU for managing the fourth smart NIC. Therefore, the first SPU records a third binding relationship between at least one port resource of the third smart NIC and the at least one physical server, and the second SPU records a fourth binding relationship between at least one port resource of the fourth smart NIC and the at least one physical server. If the cloud management platform detects a fault in the first SPU, it can migrate the third binding relationship from the first SPU to the second SPU. When the first SPU fails, at least one port resource of the third smart NIC is typically forwarding a second packet; therefore, the cloud management platform can instruct the second SPU to control at least one port resource of the third smart NIC to re-forward the second packet based on the third binding relationship. Therefore, if the SPU used to manage the smart network interface card occupied by the tenant's logical node fails, the cloud management platform can promptly select a new SPU to replace the SPU and let the new SPU take over the smart network interface card managed by the SPU to complete the packet forwarding task, thereby ensuring the continuity and stability of the tenant's business.

[0015] In one possible implementation, at least one SPU includes a first SPU and a second SPU, and at least one smart NIC includes a third smart NIC and a fourth smart NIC. The first SPU records a third binding relationship between at least one port resource of the third smart NIC and at least one physical server, and the second SPU records a fourth binding relationship between at least one port resource of the fourth smart NIC and at least one physical server. The method further includes: a cloud management platform determining that the first SPU is faulty, notifying the second SPU to control at least one port resource of the fourth smart NIC to re-forward a second message based on the fourth binding relationship, wherein the second message is the message that at least one port resource of the third smart NIC is forwarding when the first SPU is faulty. In the aforementioned implementation, the at least one smart NIC occupied by the logical node includes a third smart NIC and a fourth smart NIC, and the at least one SPU includes a first SPU for managing the third smart NIC and a second SPU for managing the fourth smart NIC. Therefore, the first SPU records a third binding relationship between at least one port resource of the third smart NIC and the at least one physical server, and the second SPU records a fourth binding relationship between at least one port resource of the fourth smart NIC and the at least one physical server. If the cloud management platform detects a failure in the first SPU, it can instruct at least one physical server to discard the third smart NIC and use the fourth smart NIC instead. Since at least one port on the third smart NIC is typically forwarding second packets when the first SPU fails, the cloud management platform can instruct the second SPU to control at least one port on the fourth smart NIC to re-forward the second packets based on the fourth binding relationship. Therefore, for at least one smart NIC occupied by a tenant's logical node, if the SPU managing these smart NICs fails, the cloud management platform allows the new SPU to directly utilize the smart NIC it manages to complete the packet forwarding tasks that the original smart NIC should have performed, thus ensuring the tenant's business continuity and stability.

[0016] In one possible implementation, the high-speed interconnect network is implemented via a high-speed serial computer expansion bus standard (peripheral component interconnect express, PCIE) network, an infinite bandwidth (infiniBand, IB) network, or a compute express link (CXL) network.

[0017] In one possible implementation, a logical node contains at least one virtual machine, at least one container, or at least one microvirtual machine.

[0018] A second aspect of this application provides a cloud management platform for managing infrastructure providing cloud services. The infrastructure includes a physical server pool, a smart NIC pool, and an SPU pool. Multiple physical servers in the physical server pool, multiple smart NICs in the smart NIC pool, and multiple SPUs in the SPU pool are interconnected via a high-speed interconnect network. Multiple SPUs manage multiple physical servers and multiple smart NICs. The cloud management platform includes: a receiving module for receiving logical node creation requests input by tenants, wherein the logical node creation request indicates the resource requirements of the logical node, including computing resource requirements, storage resource requirements, and network resource requirements; and a determining module for determining based on logical... The logical node creation request identifies at least one physical server from multiple physical servers that meets the computing and storage resource requirements, at least one smart network interface card (NIC) from multiple smart NICs that meets the network resource requirements, and at least one SPU from multiple SPUs that is used to manage at least one physical server and at least one smart NIC. The creation module is used to notify at least one SPU to create a logical node in at least one physical server and at least one smart NIC, wherein any one of the at least one smart NICs contains at least one port resource bound to at least one physical server, and the at least one port resource of the smart NIC is used to forward packets from the logical node or packets directed to the logical node.

[0019] In one possible implementation, at least one port resource of the smart network interface card (NIC) is allocated from the same port of the smart NIC, or at least one port resource of the smart NIC is allocated from different ports of the smart NIC.

[0020] In one possible implementation, at least one SPU includes a first SPU, at least one smart network interface card (NIC) includes a first smart NIC, and the cloud management platform further includes: a first notification module, used to determine that the first smart NIC is faulty, and to notify the first SPU to select a second smart NIC from the smart NICs managed by the first SPU, wherein the second smart NIC does not belong to at least one smart NIC; and a second notification module, used to notify the first SPU to release the first binding relationship between at least one port resource of the first smart NIC and at least one physical server, to establish a second binding relationship between at least one port resource of the second smart NIC and at least one physical server, and to control at least one port resource of the second smart NIC to re-forward a first message based on the second binding relationship, wherein the first message is the message that at least one port resource of the first smart NIC is forwarding when the first smart NIC is faulty.

[0021] In one possible implementation, at least one SPU includes a first SPU and a second SPU, and at least one smart network interface card (NIC) includes a third smart NIC and a fourth smart NIC. The first SPU records a third binding relationship between at least one port resource of the third smart NIC and at least one physical server, and the second SPU records a fourth binding relationship between at least one port resource of the fourth smart NIC and at least one physical server. The cloud management platform further includes: a migration module, used to determine that the first SPU is faulty and migrate the third binding relationship between at least one port resource of the third smart NIC and at least one physical server from the first SPU to the second SPU; and a third notification module, used to notify the second SPU to control at least one port resource of the third smart NIC to re-forward a second message based on the third binding relationship, wherein the second message is the message that at least one port resource of the third smart NIC is forwarding when the first SPU is faulty.

[0022] In one possible implementation, at least one SPU includes a first SPU and a second SPU, and at least one smart network interface card (NIC) includes a third smart NIC and a fourth smart NIC. The first SPU records a third binding relationship between at least one port resource of the third smart NIC and at least one physical server, and the second SPU records a fourth binding relationship between at least one port resource of the fourth smart NIC and at least one physical server. The cloud management platform further includes a fourth notification module, used to determine that the first SPU is faulty and notify the second SPU to control at least one port resource of the fourth smart NIC to re-forward a second message based on the fourth binding relationship. The second message is the message that at least one port resource of the third smart NIC is forwarding when the first SPU is faulty.

[0023] In one possible implementation, the high-speed interconnect network is implemented via a PCIe network, an IB network, or a CXL network.

[0024] In one possible implementation, the logical node comprises at least one virtual machine, at least one container, or at least one microvirtual machine.

[0025] A third aspect of this application provides a computing device cluster, which includes at least one computing device, each computing device including a processor and a memory: the memory is used to store instructions; the processor is used to cause the computing device cluster to perform the method described in the first aspect or any possible implementation of the first aspect according to the instructions.

[0026] A fourth aspect of this application provides a computer storage medium storing one or more instructions that, when executed by one or more computers, cause the one or more computers to perform the method described in the first aspect or any possible implementation of the first aspect.

[0027] A fifth aspect of this application provides a computer program product storing instructions that, when executed by a computer, cause the computer to perform the method described in the first aspect or any possible implementation of the first aspect.

[0028] In this embodiment, when a tenant needs to create its logical node, the tenant can send a logical node creation request to the creation interface provided by the cloud management platform. Since the logical node creation request indicates the resource requirements set by the tenant for the logical node, including computing resource requirements, storage resource requirements, and network resource requirements, the cloud management platform can select at least one physical server from the physical server pool that meets the computing and storage resource requirements, at least one smart network interface card (NIC) from the smart NIC pool that meets the network resource requirements, and at least one SPU from the SPU pool to manage these at least one physical server and at least one smart NIC. In this way, the cloud management platform can notify these at least one SPU to create the logical node required by the tenant on these at least one physical server and at least one smart NIC. Since any one of these at least one smart NICs contains port resources bound one-to-one with these at least one physical server, these at least one port resources of the smart NIC can be used to forward packets from or to the logical node, so that the logical node meets the tenant's business needs. In the aforementioned process, since each SPU in the SPU pool can be used to manage at least one physical server in the physical server pool and at least one smart NIC in the smart NIC pool, and each smart NIC in the smart NIC pool can provide data forwarding services for at least one physical server in the physical server pool, regardless of the number of physical servers occupied by a tenant's logical node, a smaller number of SPUs and smart NICs can be used to manage and support these physical servers. This makes the logical node more lightweight overall. During the operation of the logical node, there are fewer idle SPUs and smart NICs and more working SPUs and smart NICs, thereby improving the resource utilization of SPUs and smart NICs. Attached Figure Description

[0029] Figure 1 A schematic diagram of the structure of the cloud service system provided in the embodiments of this application;

[0030] Figure 2a A schematic diagram of the infrastructure provided in the embodiments of this application;

[0031] Figure 2b Another structural diagram of the infrastructure provided in the embodiments of this application;

[0032] Figure 3 A flowchart illustrating a logical node configuration method based on a cloud management platform provided in an embodiment of this application;

[0033] Figure 4 This is another schematic diagram of the cloud service system provided in the embodiments of this application;

[0034] Figure 5 This is another schematic diagram of the cloud service system provided in the embodiments of this application;

[0035] Figure 6 This is another schematic diagram of the cloud service system provided in the embodiments of this application;

[0036] Figure 7 This is another schematic diagram of the cloud service system provided in the embodiments of this application;

[0037] Figure 8 This is another schematic diagram of the cloud service system provided in the embodiments of this application;

[0038] Figure 9 A schematic diagram of the structure of the cloud management platform provided in the embodiments of this application;

[0039] Figure 10 A schematic diagram of the structure of a computing device provided in an embodiment of this application;

[0040] Figure 11 A schematic diagram of the structure of a computing device cluster provided in an embodiment of this application;

[0041] Figure 12 This is a schematic diagram illustrating the network connection of computer devices in a computer cluster provided in an embodiment of this application. Detailed Implementation

[0042] This application provides a logical node configuration method based on a cloud management platform and a cloud management platform, which can improve the utilization rate of resources occupied by logical nodes.

[0043] The terms "first," "second," etc., used in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such terms are interchangeable where appropriate; this is merely a way of distinguishing objects with the same attributes in the embodiments of this application. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion, so that a process, method, system, product, or apparatus that comprises a series of elements is not necessarily limited to those elements, but may include other elements not explicitly listed or inherent to those processes, methods, products, or apparatuses.

[0044] With the rapid development of cloud technology, more and more tenants are choosing to use cloud resources provided by cloud providers to complete their business. These cloud resources may include computing resources, storage resources, and network resources provided by physical servers in the cloud. Therefore, tenants can create cloud instances that meet their requirements on these cloud resources to satisfy their business needs.

[0045] In related technologies, cloud resources provided by cloud providers to tenants are typically located on different physical servers. Generally, physical servers can contain various resources such as processors, memory, and network interface cards (NICs). To improve the utilization of these resources, cloud providers can insert additional offload cards into the physical servers to offload some tasks of the processor and NICs to the offload cards for execution. For example, the offload card can contain a service process unit (SPU) and a smart NIC (also known as a smart acceleration NIC, etc.). The SPU can replace the processor to perform resource management functions, and the smart NIC can replace the NIC to perform data forwarding functions, etc.

[0046] In the above process, since the creation of a tenant's cloud instance requires several physical servers, the cloud provider can configure an SPU and a smart NIC for each of these physical servers. During the operation of the cloud instance, not all the SPUs and smart NICs on all the physical servers occupied by the cloud instance need to perform their functions. This results in most of the SPUs and smart NICs on the physical servers being idle, leading to low resource utilization of SPUs and smart NICs.

[0047] To address the aforementioned issues, this application provides a logical node configuration method based on a cloud management platform. This method can be implemented through a cloud service system. Figure 1 A schematic diagram of the cloud service system provided in the embodiments of this application is shown below. Figure 1 As shown, a cloud service system includes the infrastructure that provides cloud services and a cloud management platform that manages this infrastructure. The cloud management platform and the infrastructure are described separately below:

[0048] A cloud management platform can centrally manage the infrastructure of the entire cloud service system (e.g., within the infrastructure, creating one or more logical nodes to serve the tenant according to the tenant's instructions; these logical nodes can be used to complete the tenant's business, etc.). The cloud management platform can also be open to tenants outside the cloud service system and respond to their requests. For example, the cloud management platform can provide various interfaces such as login and creation interfaces for tenant clients (e.g., the terminal device used by the tenant or the browser on the terminal device, etc.) to access. Specifically, the cloud management platform can authenticate the tenant's client through the login interface; after successful authentication, the tenant's client can be allowed to log in to the cloud management platform. For example, the cloud management platform can also create an interface that allows tenant clients to send logical node creation requests to the cloud management platform for the tenant's logical nodes. Since the logical node creation request is used to indicate the (custom) resource requirements set by the tenant for the logical node, and the resource specification requirements are used to describe the specifications of various resources required to create the logical node, such as computing resource requirements, storage resource requirements, and network resource requirements, the cloud management platform can select idle physical devices (such as at least one physical server, at least one SPU, and at least one smart network card, etc.) from various physical device pools that meet the specifications of the logical node, and create the logical node on these physical devices. Therefore, the cloud management platform can subsequently install the operating system image specified by the tenant in the logical node in a certain way, so that the logical node with the operating system image installed can be remotely accessed by the tenant. Thus, the logical node with the operating system image installed can complete the tenant's business based on the tenant's remote access operation.

[0049] The infrastructure comprises multiple physical device pools, including physical server pools, SPU pools, and smart NIC pools. Multiple physical servers in the physical server pool, multiple SPUs in the SPU pool, and multiple smart NICs in the smart NIC pool can all be connected to a high-speed interconnection network. The high-speed interconnection network is used to enable interconnection of these multiple physical servers, multiple SPUs, and multiple smart NICs within and between pools.

[0050] It should be noted that since the multiple physical servers in the physical server pool each contain a central processing unit (CPU), memory, and disk, and these multiple physical servers are interconnected, the CPU, memory, and disk of different physical servers can be shared (that is, the CPU, memory, and disk of one physical server can be provided to the CPU of another physical server, etc.). In other words, the multiple physical servers in the physical server pool constitute a CPU pool, a memory pool, and a disk pool. The CPU pool contains the CPUs of these multiple physical servers (these CPUs can be the same or different in specifications), the memory pool contains the memory of these multiple physical servers (these memory can be the same or different in specifications), and the disk pool contains the disks of these multiple physical servers (these disks can be the same or different in specifications).

[0051] It should also be noted that multiple smart network interface cards (NICs) in the smart NIC pool (these smart NICs may have the same or different specifications; smart NICs can also be called smart acceleration NICs or network process units (NPUs), etc.) can be used to implement data forwarding functions, etc. It is worth noting that the data forwarding functions provided by these multiple smart NICs can be divided into compute offloading, storage offloading, and network offloading functions. Compute offloading refers to the data forwarded by the smart NICs being the data that the physical server needs to transmit during computation (e.g., creating virtual machines or containers on the physical server). Storage offloading refers to the data that the smart NICs forward to be stored externally (e.g., object storage buckets, etc.) by the physical server. Network offloading refers to the data that the smart NICs forward to be transmitted for communication between the physical server and external sources (e.g., other physical servers, etc.), etc. Therefore, according to their different functions, these multiple smart NICs can be divided into general-purpose cards (which can be used to implement compute offloading, storage offloading, and network offloading functions), storage offloading cards (which can be used to implement storage offloading functions), and network offloading cards (which can be used to implement network offloading functions), etc. It is also worth noting that these multiple smart network interface cards (NICs) can be presented in various product forms. For example, these multiple smart NICs can also be service-driven infrastructure (SDI) cards, which include traditional network adapters and microprocessors. Alternatively, these multiple smart NICs can also be data processing units (DPUs), and so on, without specific limitations here.

[0052] It should also be noted that multiple SPUs in the SPU pool can be used to implement resource management functions, etc. In other words, any one of these SPUs can be used to manage at least one physical server and at least one smart network interface card (NIC). These multiple SPUs can be presented in various forms, such as a graphics processing unit (GPU) device pool, a neural network processing unit (NIC) device pool, a tensor processing unit (TPU) device pool, and a DPU device pool, etc., without specific limitations here.

[0053] As can be seen, the cloud management platform can select several physical devices from these physical device pools that meet the tenant's needs and have certain specifications to build the tenant's logical nodes.

[0054] It is also worth noting that these multiple physical device pools are presented and communicate with each other in various ways:

[0055] (1) As Figure 2a As shown ( Figure 2a (A schematic diagram of the infrastructure provided in this application embodiment) In these multiple physical device pools, all physical devices in each physical device pool are randomly and dispersedly arranged on different racks, that is, different types of physical devices may exist in the same rack (for example, physical servers, smart network cards, and SPUs may be stored in the same rack). Therefore, in these multiple physical device pools, any two physical devices in any physical device pool can communicate with each other through a high-speed interconnection device, and physical devices in any two physical device pools can also communicate with each other through a high-speed interconnection device. The high-speed interconnection device may include a high-speed interconnection bus and a high-speed interconnection bus switch, etc. Thus, the high-speed interconnection device connects all physical devices in multiple physical device pools, forming a high-speed interconnection network among all physical devices (the "high-speed" in high-speed interconnection network refers to a bandwidth of not less than 40G / S). In other words, multiple physical servers in the physical server pool, multiple SPUs in the SPU pool, and multiple smart network cards in the smart network card pool are all connected to the high-speed interconnection network, so these physical devices can communicate with each other through the high-speed interconnection network.

[0056] It is also worth noting that after the cloud management platform determines the resource requirements of a tenant's logical nodes—namely, the computing resource requirements (e.g., CPU requirements), storage resource requirements (e.g., memory and disk requirements), and network resource requirements (e.g., network interface card requirements)—the cloud management platform can select several physical devices from a pool of physical devices that meet these resource requirements. Specifically, it can select target physical servers from a pool of physical servers that meet both the computing and storage resource requirements (e.g., ...). Figure 2a The physical server in the dashed box), and the target smart network interface card (NIC) selected from the smart NIC pool that meets the network resource requirements (e.g., Figure 2a The smart network interface card (NIC) within the dashed box), and select the target SPU (e.g., from the SPU pool) for managing the target physical server and the target smart network interface card. Figure 2a (The SPU is shown in the dashed box). In this way, the cloud management platform can instruct the target SPU to create the logical nodes required by the tenant on the target physical server and the target smart NIC. This logical node is used to implement the network formed between the target physical server, the target smart NIC, and the target SPU, which is part of the high-speed interconnect network, or a sub-network of the high-speed interconnect network (e.g., Figure 2a (dashed box).

[0057] (2) Figure 2b As shown ( Figure 2b (Another structural diagram of the infrastructure provided in this application embodiment) In these multiple physical device pools, all physical devices in each physical device pool are arranged in the same rack or adjacent racks according to a certain pattern. The same rack contains physical devices of the same type (for example, one rack contains physical servers, another rack contains smart network cards, and yet another rack contains SPUs). Therefore, in these multiple physical device pools, any two physical devices in any physical device pool can communicate with each other through a high-speed interconnection device, and physical devices in any two physical device pools can also communicate with each other through a high-speed interconnection device. This high-speed interconnection device may include a high-speed interconnection bus and a high-speed interconnection bus switch, etc. Thus, the high-speed interconnection device connects all physical devices in multiple physical device pools, forming a high-speed interconnection network (the "high-speed" in high-speed interconnection network refers to a bandwidth of not less than 40G / S). That is, multiple physical servers in the physical server pool, multiple SPUs in the SPU pool, and multiple smart network cards in the smart network card pool are all connected to the high-speed interconnection network, so these physical devices can communicate with each other through the high-speed interconnection network.

[0058] It is also worth noting that after the cloud management platform determines the resource requirements of a tenant's logical nodes—namely, the computing, storage, and network resource requirements of the logical nodes—the cloud management platform can select several physical devices from a pool of physical devices to meet those resource requirements. Specifically, it can select target physical servers from a pool of physical servers that meet those computing and storage resource requirements (e.g., ...). Figure 2b The physical server in the dashed box), and the target smart network interface card (NIC) selected from the smart NIC pool that meets the network resource requirements (e.g., Figure 2b The smart network interface card (NIC) within the dashed box), and select the target SPU (e.g., from the SPU pool) for managing the target physical server and the target smart network interface card. Figure 2b (The SPU is shown in the dashed box). In this way, the cloud management platform can instruct the target SPU to create the logical nodes required by the tenant on the target physical server and the target smart NIC. This is used to implement the network formed between the target physical server, the target smart NIC, and the target SPU, which is part of the high-speed interconnect network, or a sub-network of the high-speed interconnect network (e.g., the SPU within the dashed box).

[0059] Furthermore, such as Figure 2a As shown, in these multiple physical device pools, each physical device has a certain type of interface. For any given physical device, it can connect to a high-speed interconnect bus switch through its interface, and then communicate with other physical devices through the high-speed interconnect bus.

[0060] Furthermore, for the tenant's logical node, the logical node can typically be regarded as a cloud instance in the infrastructure. These cloud instances can be presented in various ways. For example, these cloud instances can be (one or more) physical servers selected by the cloud management platform in the infrastructure; they can also be bare metal servers selected by the cloud management platform in the infrastructure; they can also be virtual machines (VMs) created by the cloud management platform on physical servers using virtualization technology; they can also be containers (Docker) created by the cloud management platform on physical servers using virtualization technology; they can also be micro VMs created by the cloud management platform on physical servers using virtualization technology, and so on.

[0061] Furthermore, for the aforementioned multiple physical device pools, these multiple physical device pools can be deployed in the same site or different sites. The site can be presented in various forms, such as a region in the infrastructure, an availability zone in the infrastructure, a data center (DC) in the infrastructure, a room in the infrastructure, and so on.

[0062] Furthermore, the high-speed interconnection network built between these multiple physical device pools can be implemented based on multiple high-speed interconnect bus switches. Suppose these multiple physical device pools are deployed across a data center, which often contains multiple racks. These racks are used to house several physical devices from each physical device pool. In order to enable communication between physical devices in different racks, each rack often needs to deploy at least one high-speed interconnect bus switch. In this way, the racks can communicate with each other through the high-speed interconnect bus switches, thereby forming a high-speed interconnection network within the data center.

[0063] Furthermore, the aforementioned high-speed interconnect network can take many forms. For example, it can be a Peripheral Component Interconnect Express (PCIE) network, an InfiniBand (IB) network, a Compute Express Link (CXL) network, or even an interconnect network between devices developed by a cloud vendor (the bandwidth of which also needs to be no less than 40 Gbps), etc. Correspondingly, the high-speed interconnect devices connected to each physical device can be communication devices based on the PCIE, CXL, IB, or cloud vendor-developed communication protocols. Similarly, the interfaces used by the physical devices in each physical device pool to access the high-speed interconnect network can be PCIE interfaces, CXL interfaces, IB interfaces, or cloud vendor-developed interfaces, etc.

[0064] Based on the aforementioned cloud service system, when a tenant needs to create its logical node, the tenant can send a logical node creation request to the creation interface provided by the cloud management platform. Since the logical node creation request indicates the resource requirements set by the tenant for that logical node, including computing, storage, and network resource requirements, the cloud management platform can select at least one physical server from the physical server pool that meets the computing and storage resource requirements, at least one smart network interface card (NIC) from the smart NIC pool that meets the network resource requirements, and at least one SPU from the SPU pool to manage these physical servers and smart NICs. In this way, the cloud management platform can notify these at least one SPU to create the tenant's required logical node on these physical servers and smart NICs. Since any one of these smart NICs contains port resources bound one-to-one with these physical servers, these port resources can be used to forward packets from or to the logical node, ensuring that the logical node meets the tenant's business needs. In the aforementioned process, since each SPU in the SPU pool can be used to manage at least one physical server in the physical server pool and at least one smart NIC in the smart NIC pool, and each smart NIC in the smart NIC pool can provide data forwarding services to at least one physical server in the physical server pool, regardless of the number of physical servers occupied by a tenant's logical node, a smaller number of SPUs and smart NICs can be used to manage and support these physical servers. This makes the logical node more lightweight overall, and during the operation of the logical node, there are fewer idle SPUs and smart NICs and more active SPUs and smart NICs, thereby improving the resource utilization of SPUs and smart NICs. To further understand the workflow of the cloud management platform, the following section... Figure 3 This workflow will be described in further detail. Figure 3 A flowchart illustrating a logical node configuration method based on a cloud management platform provided in this application embodiment is shown below. Figure 3 As shown, this method can be achieved through Figure 1 The cloud service system implementation shown includes infrastructure providing cloud services to tenants and a cloud management platform for managing this infrastructure. This infrastructure may include a pool of physical servers, a pool of smart NICs, and a pool of SPUs. Multiple physical servers in the physical server pool, multiple smart NICs in the smart NIC pool, and multiple SPUs in the SPU pool are interconnected via a high-speed interconnect network. Each SPU manages at least one physical server and at least one smart NIC among the multiple physical servers. The method includes:

[0065] 301. The cloud management platform receives a logical node creation request from a tenant. The logical node creation request indicates the resource requirements of the logical node, including computing resource requirements, storage resource requirements, and network resource requirements.

[0066] In this embodiment, when a tenant needs to create its logical node, the cloud management platform can provide a creation interface to the tenant's client (e.g., a logical node creation section on the tenant's interface). The tenant can then input a logical node creation request for that logical node into the creation interface through their client. In this way, the cloud management platform can receive the logical node creation request sent by the tenant through their client via the creation interface. It is worth noting that the logical node creation request includes the resource requirements set by the tenant for that logical node. These resource requirements include the logical node's computing resource requirements, storage resource requirements, and network resource requirements. Specifically, the computing resource requirements describe the specifications of computing resources such as CPU required to create the logical node; the storage resource requirements describe the specifications of storage resources such as memory and disk required to create the logical node; and the network resource requirements describe the specifications of network resources such as network interface cards required to create the logical node.

[0067] For example, when a tenant needs to create its VM, the tenant can log in to the cloud management platform. The cloud management platform can provide the tenant with a tenant interface, which includes a logical node creation field. Therefore, the tenant can enter the VM creation request specified by the tenant for VM1 to VM8 (i.e., the aforementioned logical nodes) into the logical node creation field. This request is used to indicate the resource requirements of VM1 to VM8. For any VM among VM1 to VM8, the resource requirements include the VM's CPU requirements (e.g., the VM needs a 32-core CPU), memory requirements (e.g., the VM needs 16GB of memory), disk requirements (e.g., the VM needs a 100GB disk), and network card requirements (e.g., the VM needs an 8GB network card), etc.

[0068] 302. Based on the logical node creation request, the cloud management platform determines at least one physical server from multiple physical servers that meets the computing resource requirements and storage resource requirements, determines at least one smart network interface card from multiple smart network interface cards that meets the network resource requirements, and determines at least one SPU from multiple SPUs to manage at least one physical server and at least one smart network interface card.

[0069] Upon receiving a logical node creation request, the cloud management platform can parse the request to determine the resource requirements of the tenant's logical node. Since the resource requirements of the logical node include its computing resource requirements, storage resource requirements, and network resource requirements, the cloud management platform can select at least one idle physical server from multiple physical servers in the physical server pool that meets the computing and storage resource requirements, select at least one idle smart network interface card (NIC) from multiple smart NICs in the smart NIC pool that meets the network resource requirements, and select at least one SPU from multiple SPUs in the SPU pool to manage the at least one physical server and the at least one smart NIC.

[0070] As in the example above, such as Figure 4 As shown ( Figure 4 (This is another structural diagram of the cloud service system provided in the embodiments of this application). Assume that the physical server pool includes physical servers 1 to 16, the smart network interface card pool includes smart network interface cards 1 to 8, and the SPU pool includes SPU1 to SPU2. SPU1 is used to manage physical servers 1, 2, 3, 4, 5, 6, 7, 8, smart network interface cards 1, 2, 3, and 4, and SPU2 is used to manage physical servers 9, 10, 11, 12, 13, 14, 15, 16, smart network interface cards 5, 6, 7, and 8.

[0071] like Figure 5 As shown ( Figure 5 (This is another schematic diagram of the cloud service system provided in this application embodiment). After receiving a VM creation request sent by a tenant, the cloud management platform can determine the resource requirements of VM1 to VM8 to be created by the tenant based on the request. The resource requirements may include the CPU requirements, memory requirements, disk requirements, and network card requirements of VM1 to VM8, etc. Therefore, the cloud management platform can select from physical servers 1 to physical servers 16 that meet these CPU requirements, these memory requirements, and these disk requirements and are idle, assuming they are physical servers 1 to physical servers 8. And select from smart network cards 1 to smart network cards 8 that meet these network card requirements and are idle, assuming they are smart network cards 1, smart network cards 2, smart network cards 5 and smart network cards 6. Therefore, the cloud management platform can further determine the SPUs used to manage physical servers 1 to physical servers 8, smart network cards 1, smart network cards 2, smart network cards 5 and smart network cards 6, namely SPU1 and SPU2.

[0072] 303. The cloud management platform notifies at least one SPU to create a logical node in at least one physical server and at least one smart network interface card (NIC), wherein any one of the at least one smart NICs contains at least one port resource bound to at least one physical server, and the at least one port resource of the smart NIC is used to forward packets from the logical node or packets directed to the logical node.

[0073] After identifying at least one SPU, at least one physical server, and at least one smart NIC, the cloud management platform can instruct the at least one SPU to create a tenant's logical node on the at least one physical server and at least one smart NIC, and obtain the operating system image specified by the tenant, install the operating system image on the logical node, so that the logical node with the operating system image installed can be remotely logged in by the tenant to meet the tenant's business needs.

[0074] Specifically, for any one of the at least one smart network interface cards (NICs) occupied by the logical node, each NIC contains at least one port resource that is bound one-to-one with the at least one physical server occupied by the logical node. Based on this, when the logical node with the operating system image installed needs to send a message to the outside, the at least one port resource of the smart NIC can forward the message from the logical node with the operating system image installed to the outside. When the outside needs to send a message to the logical node with the operating system image installed, the at least one port resource of the smart NIC can forward the message from the outside to the logical node with the operating system image installed, thereby enabling communication between the logical node and the outside, and thus meeting the tenant's business needs.

[0075] More specifically, for any one of the at least one smart network interface cards (NICs) occupied by the logical node, the NIC may contain one or more ports, which can be used for communication with the outside world. Therefore, the at least one SPU can logically allocate one or more of the ports of the smart NIC as at least one port resource bound to the at least one physical server occupied by the logical node. These port resources may also come from the class of service (CoS) resources of these ports. Based on this, for any one of the at least one physical server occupied by the logical node, the physical server can communicate with the outside world (i.e., forward packets) through a port resource bound to it in the smart NIC, thereby enabling communication between the logical node and the outside world.

[0076] Continuing with the example above, after determining SPU1 and SPU2, the cloud management platform can instruct SPU1 and SPU2 to create VM1 to VM8 on physical servers 1 to 8, and smart NICs 1, 2, 5, and 6. Specifically, on physical server 1, VM1 and Virtual Machine Manager (VMM) 1 are deployed. VM1's virtual CPU, virtual memory, and virtual disk are implemented based on the CPU, memory, and disk of physical server 1. VM1's virtual NIC is implemented based on smart NICs 1, 2, 5, and 6. On physical server 2, VM2 and VMM 2 are deployed. VM2's virtual CPU, virtual memory, and virtual disk are implemented based on the CPU, memory, and disk of physical server 2. VM2's virtual NIC is implemented based on smart NICs 1, 2, 5, and 6, and so on. VM7 and VMM7 are deployed on physical server 7. The virtual CPU, virtual memory, and virtual disk of VM7 are implemented based on the CPU, memory, and disk of physical server 7. The virtual network card of VM7 is implemented based on smart network card 1, smart network card 2, smart network card 5, and smart network card 6. VM8 and VMM8 are deployed on physical server 8. The virtual CPU, virtual memory, and virtual disk of VM8 are implemented based on the CPU, memory, and disk of physical server 8. The virtual network card of VM8 is implemented based on smart network card 1, smart network card 2, smart network card 5, and smart network card 6.

[0077] For Smart NIC 1, Smart NIC 1 may contain port 1 and port 2. Suppose that port 1 of Smart NIC 1 is allocated to physical servers 1 through 8. Since these 8 physical servers share port 1, if there is a sudden increase in packets pointing to physical server 1, port 1 will be occupied by packets pointing to physical server 1, causing port 1 congestion, which in turn affects the forwarding of packets pointing to physical servers 2 through 8. Since port 1 contains multiple physical CoS resources such as Cos0, Cos1, ..., Cosn, SPU 1 can logically divide Cos0 into sub-CoS1 through sub-CoS8, and bind sub-CoS1 to physical server 1, sub-CoS2 to physical server 2, ..., sub-CoS7 to physical server 7, and sub-CoS8 to physical server 8. Similarly, SPU1 can perform similar operations on the port of smart NIC 2, and SPU2 can perform similar operations on the ports of smart NIC 5 and smart NIC 6. This allows physical server 1 to be assigned sub-CoS (i.e., sub-CoS1, sub-CoS9, sub-CoS17, and sub-CoS25) of smart NICs 1, 2, 5, and 6. Similarly, physical servers 2 to 8 are also assigned sub-CoS of smart NICs 1, 2, 5, and 6, which will not be elaborated further here.

[0078] Therefore, it can be seen that once VM1 on physical server 1 needs to send or receive a message, it can do so through any one of the sub-CoS1, sub-CoS9, sub-CoS17, or sub-CoS25. Similarly, the same applies to VM2 on physical server 2 through VM8 on physical server 8, which will not be elaborated further here. It should be noted that sub-CoS usually refers to the physical resources of ports in a smart network interface card (NIC). When these sub-CoS are directed to VMs on a physical server, they can be presented as virtual functions (VFs) and (virtual) queues provided by the smart NIC. For example, sub-CoS1 directed to VM1 can be presented as the VF provided by smart NIC1 for VM1 and the queues mounted on the VF. Sub-CoS9 directed to VM1 can be presented as the VF provided by smart NIC2 for VM1 and the queues mounted on the VF. Sub-CoS17 directed to VM1 can be presented as the VF provided by smart NIC5 for VM1 and the queues mounted on the VF. Sub-CoS25 directed to VM1 can be presented as the VF provided by smart NIC6 for VM1 and the queues mounted on the VF. These will not be elaborated further.

[0079] More specifically, the cloud management platform can also perform the following operations:

[0080] For any one of the at least one smart network interface cards (NICs) occupied by this logical node, hereinafter referred to as the first smart NIC, the SPU used to manage the first smart NIC is referred to as the first SPU. The cloud management platform can perform fault monitoring on the first smart NIC in some way. If a fault is determined in the first smart NIC, the cloud management platform can instruct the first SPU to select a second smart NIC from the remaining smart NICs managed by the first SPU to replace the first smart NIC. It should be noted that the second smart NIC is not among the at least one smart NIC occupied by this logical node.

[0081] Since the first SPU records the first binding relationship between at least one port resource of the first smart network card and the at least one physical server occupied by the logical node, and the first smart network card is faulty (i.e., the first smart network card cannot be used), the cloud management platform can notify the first SPU to release the first binding relationship between at least one port resource of the first smart network card and the at least one physical server occupied by the logical node, and then reconstruct the second binding relationship between at least one port resource of the second smart network card (which can be obtained by the first SPU through real-time allocation of one or more ports of the second smart network card) and the at least one physical server occupied by the logical node.

[0082] When the first smart network interface card (NIC) fails, at least one port resource of the first smart NIC is usually forwarding a certain packet, namely the first packet. Therefore, the cloud management platform can, in some way, notify the first SPU to control at least one port resource of the second smart NIC to re-forward the first packet based on its recorded second binding relationship, thereby ensuring the continuity of services.

[0083] As in the example above, such as Figure 6 As shown ( Figure 6 (This is another schematic diagram of the cloud service system provided in the embodiments of this application). For ease of explanation, the following description uses smart network interface card 1 (the aforementioned first smart network interface card) as an example. Port 1 of smart network interface card 1 is divided into sub-CoS1 to sub-CoS8 by SPU1, and sub-CoS1 is bound to physical server 1, sub-CoS2 is bound to physical server 2, ..., sub-CoS7 is bound to physical server 7, and sub-CoS8 is bound to physical server 8. This binding relationship (the aforementioned first binding relationship) is recorded by SPU1 (the aforementioned first SPU).

[0084] Since physical servers 1 to 8 and smart network interface card 1 can be interconnected via high-speed interconnection devices (e.g., PCIe switches), the cloud management platform can enable these high-speed interconnection devices to monitor smart network interface card 1 in real time. If the high-speed interconnection device detects a fault in smart network interface card 1, it can notify the cloud management platform. Based on this information, the cloud management platform can determine that smart network interface card 1 is faulty. The cloud management platform can then notify the SPU1 that manages smart network interface card 1, causing SPU1 to first unbind the aforementioned binding relationship and select smart network interface card 3 (the aforementioned second smart network interface card) from among the smart network interface cards 3 and 4 it manages to replace smart network interface card 1. Then, SPU1 can divide a port of smart network interface card 3 into sub-CoS 33 to sub-CoS 40, bind sub-CoS 33 to physical server 1, sub-CoS 34 to physical server 2, ..., sub-CoS 39 to physical server 7, and sub-CoS 40 to physical server 8, and record this new binding relationship (the aforementioned second binding relationship).

[0085] Assuming that when smart NIC 1 fails, its child CoS1 is forwarding the current packet (i.e., the aforementioned first packet) destined for VM1 from the outside. SPU1 can also notify VMM1 to perform packet verification. That is, when VM1 communicates with the outside, the outside can continuously send multiple consecutive packets to VM1. These packets usually have consecutive IDs, and each packet is sent from child CoS1 to VMM1, and then from VMM1 to VM1. VMM1 can record the ID of each packet. Let the ID of the current packet be 8. Due to the failure of smart NIC 1, child CoS1 failed to successfully send the current packet to VMM1, so VMM1 only records the ID of the previous packet as 7. After SPU1 replaces smart NIC 1 with smart NIC 3, smart NIC 3's CoS33 can continue to receive the next packet from the outside. The ID of the next packet is 9, and it is successfully sent to VMM1. After receiving the next message, VMM1, whose ID is 9, compares it with its recorded message IDs. Finding that it has only received messages with IDs 7 and 9, VMM1 refuses to forward the next message to VM1 and notifies the external system to resend the message with ID 8 (the current message) to the child CoS33. This allows the child CoS33 to forward the current message to VMM1, which in turn forwards it to VM1. In this way, after VMM1 performs message verification and retransmission, VM1 can accurately retrieve the current message, ensuring message continuity and, consequently, service continuity.

[0086] More specifically, the cloud management platform can also perform the following operations:

[0087] (1) Since the at least one SPU is used to manage the at least one physical server and the at least one smart network card occupied by the logical node, it is assumed that the at least one SPU contains two different SPUs, which are referred to as the first SPU and the second SPU below. The at least one smart network card contains two different smart network cards, which are referred to as the third smart network card and the fourth smart network card below. The third smart network card is managed by the first SPU, so the first SPU records the third binding relationship between at least one port resource of the third smart network card and the at least one physical server. The fourth smart network card is managed by the second SPU, so the second SPU records the fourth binding relationship between at least one port resource of the fourth smart network card and the at least one physical server.

[0088] It should be noted that for any one of these at least one physical server, the port resources bound to that physical server in both the third smart network interface card (NIC) and the fourth smart network interface card (NIC) are active (or available). In this case, if the cloud management platform detects a fault in the first SPU (i.e., the first SPU is unusable), the cloud management platform can transfer the third binding relationship between at least one port resource of the third smart network interface card and these at least one physical server from the first SPU to the second SPU. This is equivalent to handing over the management of the third smart network interface card to the second SPU (changing the ownership relationship between the smart network interface card and the SPU), meaning that at least one port resource of the third smart network interface card is now managed by the second SPU. Based on this, for any one of these at least one physical server, it can still use the port resources bound to it in both the third and fourth smart network interface cards to send or receive packets. The same applies to the other physical servers, and will not be elaborated further here.

[0089] When the first SPU fails, at least one port resource of the third smart NIC is usually forwarding a packet, namely the second packet. Therefore, the cloud management platform can, in some way, notify the second SPU to control at least one port resource of the third smart NIC to re-forward the second packet based on its recorded third binding relationship, thereby ensuring the continuity of services.

[0090] As in the example above, such as Figure 7 As shown ( Figure 7(This is another structural diagram of the cloud service system provided in the embodiments of this application). Since SPU1 (the aforementioned first SPU) manages smart network interface card 1 and smart network interface card 2 (these two smart network interface cards are the aforementioned third smart network interface card), SPU1 records the binding relationship between the sub-CoS of smart network interface card 1 and physical server 1, ..., physical server 8 (sub-CoS1 is bound to physical server 1, sub-CoS2 is bound to physical server 2, ..., sub-CoS7 is bound to physical server 7, sub-CoS8 is bound to physical server 8), and the binding relationship between the sub-CoS of smart network interface card 2 and physical server 1, ..., physical server 8 (sub-CoS9 is bound to physical server 1, sub-CoS10 is bound to physical server 2, ..., sub-CoS15 is bound to physical server 7, sub-CoS16 is bound to physical server 8). These two binding relationships are the aforementioned third binding relationship.

[0091] Since SPU2 (the aforementioned second SPU) manages smart NIC 5 and smart NIC 6 (these two smart NICs are the aforementioned fourth smart NIC), SPU2 records the binding relationships between the sub-CoS of smart NIC 5 and physical servers 1, ..., and 8 (sub-CoS17 is bound to physical server 1, sub-CoS18 is bound to physical server 2, ..., sub-CoS23 is bound to physical server 7, and sub-CoS24 is bound to physical server 8), and the binding relationships between the sub-CoS of smart NIC 6 and physical servers 1, ..., and 8 (sub-CoS25 is bound to physical server 1, sub-CoS26 is bound to physical server 2, ..., sub-CoS31 is bound to physical server 7, and sub-CoS32 is bound to physical server 8). These two binding relationships constitute the aforementioned fourth binding relationship.

[0092] For VM1 and VMM1 on physical server 1, since sub-CoS1, sub-CoS9, sub-CoS17, and sub-CoS25 are bound to physical server 1, VMM1 can determine that sub-CoS1, sub-CoS9, sub-CoS17, and sub-CoS25 are port resources that can be provided to VM1. Moreover, all four sub-CoS are active for VMM1, meaning that sub-CoS1, sub-CoS9, sub-CoS17, and sub-CoS25 can enable VM1 to communicate with the outside world.

[0093] If the cloud management platform detects a failure in SPU1, it will migrate the binding relationships between the sub-CoS of smart NIC 1 and physical servers 1, ..., 8, and the binding relationships between the sub-CoS of smart NIC 2 and physical servers 1, ..., 8, from SPU1 to SPU2. In this way, the sub-CoS of smart NIC 1 and smart NIC 2 will be managed by SPU2.

[0094] Assuming that SPU1 fails, the sub-CoS9 of smart NIC 2 is forwarding the current packet (i.e., the aforementioned second packet) destined for VM1 from the outside. The cloud management platform can send a request to SPU2 to notify VMM1 to perform packet verification. In other words, when VM1 communicates with the outside, the outside can continuously send multiple consecutive packets to VM1. These packets usually have consecutive IDs, and each packet is sent from sub-CoS1 to VMM1, and then from VMM1 to VM1. VMM1 can record the ID of each packet. Let the ID of the current packet be 18. Due to the failure of SPU1, sub-CoS9 fails to successfully send the current packet to VMM1, so VMM1 only records the ID of the previous packet as 17. After SPU2 replaces SPU1, the sub-CoS9 of smart NIC 2 can resume normal operation, meaning it can continue to receive the next packet from the outside, with the ID 19, and successfully send it to VMM1. Upon receiving the next message, VMM1, whose ID is 19, compares it with its recorded message IDs. Finding that it has only received messages with IDs 17 and 19, VMM1 refuses to forward the next message to VM1 and notifies the external system to resend the message with ID 18 (the current message) to child CoS9. This allows child CoS9 to forward the current message to VMM1, and then to VM1. In this way, after VMM1 performs message verification and retransmission, VM1 can accurately retrieve the current message, ensuring message continuity and, consequently, service continuity.

[0095] (2) Since the at least one SPU is used to manage the at least one physical server and the at least one smart network interface card occupied by the logical node, it is assumed that the at least one SPU contains two different SPUs, which are referred to as the first SPU and the second SPU below. The at least one smart network interface card contains two different smart network interface cards, which are referred to as the third smart network interface card and the fourth smart network interface card below. The third smart network interface card is managed by the first SPU, so the first SPU records the third binding relationship between at least one port resource of the third smart network interface card and the at least one physical server. The fourth smart network interface card is managed by the second SPU, so the second SPU records the fourth binding relationship between at least one port resource of the fourth smart network interface card and the at least one physical server.

[0096] It should be noted that for any one of these at least one physical server, the port resource bound to that physical server in the third smart network interface card (NIC) is in an active (or available) state, while the port resource bound to that physical server in the fourth smart NIC is in a standby (or unavailable) state. In this scenario, if the cloud management platform detects a fault in the first SPU (i.e., the first SPU is unusable), the cloud management platform can directly notify that physical server to discard the port resource bound to it in the third smart NIC and set the port resource bound to it in the fourth smart NIC to an active state, enabling the physical server to subsequently use the port resource bound to it in the fourth smart NIC to send or receive packets. The same applies to the other physical servers, and will not be elaborated further here.

[0097] When the first SPU fails, at least one port resource of the third smart NIC is usually forwarding a packet, namely the second packet. Therefore, the cloud management platform can notify the second SPU to control at least one port resource of the fourth smart NIC to re-forward the second packet based on its recorded fourth binding relationship, thereby ensuring the continuity of services.

[0098] As in the example above, such as Figure 8 As shown ( Figure 8 (This is another structural diagram of the cloud service system provided in the embodiments of this application). Since SPU1 (the aforementioned first SPU) manages smart network interface card 1 and smart network interface card 2 (these two smart network interface cards are the aforementioned third smart network interface card), SPU1 records the binding relationship between the sub-CoS of smart network interface card 1 and physical server 1, ..., physical server 8 (sub-CoS1 is bound to physical server 1, sub-CoS2 is bound to physical server 2, ..., sub-CoS7 is bound to physical server 7, sub-CoS8 is bound to physical server 8), and the binding relationship between the sub-CoS of smart network interface card 2 and physical server 1, ..., physical server 8 (sub-CoS9 is bound to physical server 1, sub-CoS10 is bound to physical server 2, ..., sub-CoS15 is bound to physical server 7, sub-CoS16 is bound to physical server 8). These two binding relationships are the aforementioned third binding relationship.

[0099] Since SPU2 (the aforementioned second SPU) manages smart NIC 5 and smart NIC 6 (these two smart NICs are the aforementioned fourth smart NIC), SPU2 records the binding relationships between the sub-CoS of smart NIC 5 and physical servers 1, ..., and 8 (sub-CoS17 is bound to physical server 1, sub-CoS18 is bound to physical server 2, ..., sub-CoS23 is bound to physical server 7, and sub-CoS24 is bound to physical server 8), and the binding relationships between the sub-CoS of smart NIC 6 and physical servers 1, ..., and 8 (sub-CoS25 is bound to physical server 1, sub-CoS26 is bound to physical server 2, ..., sub-CoS31 is bound to physical server 7, and sub-CoS32 is bound to physical server 8). These two binding relationships constitute the aforementioned fourth binding relationship.

[0100] For VM1 and VMM1 on physical server 1, since sub-CoS1, sub-CoS9, sub-CoS17, and sub-CoS25 are bound to physical server 1, VMM1 can determine that sub-CoS1, sub-CoS9, sub-CoS17, and sub-CoS25 are port resources available for VM1 to use. Furthermore, sub-CoS1 and sub-CoS9 are both active for VMM1, meaning that sub-CoS1 and sub-CoS9 can currently enable VM1 to communicate with the outside world. Sub-CoS17 and sub-CoS25 are both in standby state for VMM1, meaning that sub-CoS17 and sub-CoS25 cannot currently enable VM1 to communicate with the outside world.

[0101] If the cloud management platform detects a fault in SPU1, it can notify VMM1 to stop using sub-CoS1 and sub-CoS9, and set the status of sub-CoS17 and sub-CoS25 to active at VMM1. Therefore, VMM1 can determine that sub-CoS17 and sub-CoS25 can currently enable VM1 to communicate with the outside world.

[0102] Assuming that SPU1 fails, and the sub-CoS9 of smart NIC 2 is forwarding a current packet (i.e., the aforementioned second packet) destined for VM1 from an external source, the cloud management platform can send a request to SPU2 to instruct VMM1 to perform packet verification. In other words, when VM1 communicates with the external source, the external source can continuously send multiple consecutive packets to VM1. These packets typically have consecutive IDs, and each packet is sent from sub-CoS1 to VMM1, and then from VMM1 to VM1. VMM1 can record the ID of each packet. Let the ID of the current packet be 18. Due to the failure of SPU1, sub-CoS9 fails to successfully send the current packet to VMM1; therefore, VMM1 only records the ID of the previous packet as 17. Since sub-CoS1 and sub-CoS9 are deprecated, and sub-CoS17 and sub-CoS25 are activated, sub-CoS17 of smart NIC 5 can continue to receive the next packet from the outside. The next packet has an ID of 19 and is successfully sent to VMM1. After receiving the next packet, VMM1 compares it with its recorded packet IDs, finding that it has only received packets with IDs of 17 and 19. Therefore, it refuses to forward the next packet to VM1 and notifies the outside to resend the packet with ID 18 (the current packet) to sub-CoS17, so that sub-CoS17 can forward the current packet to VMM1 and then to VM1. In this way, after VMM1 performs packet verification and retransmission, VM1 can accurately retrieve the current packet again, ensuring packet continuity, i.e., service continuity.

[0103] In this embodiment, when a tenant needs to create its logical node, the tenant can send a logical node creation request to the creation interface provided by the cloud management platform. Since the logical node creation request indicates the resource requirements set by the tenant for the logical node, including computing resource requirements, storage resource requirements, and network resource requirements, the cloud management platform can select at least one physical server from the physical server pool that meets the computing and storage resource requirements, at least one smart network interface card (NIC) from the smart NIC pool that meets the network resource requirements, and at least one SPU from the SPU pool to manage these at least one physical server and at least one smart NIC. In this way, the cloud management platform can notify these at least one SPU to create the logical node required by the tenant on these at least one physical server and at least one smart NIC. Since any one of these at least one smart NICs contains port resources bound one-to-one with these at least one physical server, these at least one port resources of the smart NIC can be used to forward packets from or to the logical node, so that the logical node meets the tenant's business needs. In the aforementioned process, since each SPU in the SPU pool can be used to manage at least one physical server in the physical server pool and at least one smart NIC in the smart NIC pool, and each smart NIC in the smart NIC pool can provide data forwarding services for at least one physical server in the physical server pool, regardless of the number of physical servers occupied by a tenant's logical node, a smaller number of SPUs and smart NICs can be used to manage and support these physical servers. This makes the logical node more lightweight overall. During the operation of the logical node, there are fewer idle SPUs and smart NICs and more working SPUs and smart NICs, thereby improving the resource utilization of SPUs and smart NICs.

[0104] Furthermore, in this embodiment, for any one of the at least one smart network interface cards (NICs) occupied by a tenant's logical node, the different port resources of the smart NIC can be allocated to different physical servers occupied by the logical node. Therefore, even if there is a sudden increase in traffic on a certain physical server (packets from or directed to that physical server, etc.), it will only be limited to forwarding on the port resources bound to that physical server in the smart NIC, and will not affect the traffic forwarding of other physical servers (this part of the traffic is forwarded on the port resources bound to the other physical servers in the smart NIC). This can ensure the stability of the tenant's business.

[0105] Furthermore, in this embodiment, for any one of the at least one smart network interface cards (NICs) occupied by a tenant's logical node, if the NIC fails, the cloud management platform can notify the SPU managing the NIC to promptly select a new NIC to replace it, so as to complete the packet forwarding task that the NIC should have originally performed, thereby ensuring the continuity and stability of the tenant's business.

[0106] Furthermore, in this embodiment, if the SPU used to manage these smart network cards fails, the cloud management platform can promptly select a new SPU to replace the SPU and allow the new SPU to take over the smart network cards managed by the SPU to complete the packet forwarding task. Alternatively, the new SPU can directly use the smart network cards it manages to complete the packet forwarding task that the smart network cards managed by the SPU should have originally completed, thereby ensuring the continuity and stability of the tenant's business.

[0107] The above is a detailed description of the logical node configuration method based on the cloud management platform provided in the embodiments of this application. The cloud management platform provided in the embodiments of this application will be introduced below. Figure 9 This is a schematic diagram of a cloud management platform provided in an embodiment of this application. The cloud management platform is used to manage the infrastructure providing cloud services. The infrastructure includes a physical server pool, a smart NIC pool, and an SPU pool. Multiple physical servers in the physical server pool, multiple smart NICs in the smart NIC pool, and multiple SPUs in the SPU pool are interconnected through a high-speed interconnection network. Multiple SPUs are used to manage multiple physical servers and multiple smart NICs. The cloud management platform includes:

[0108] The receiving module 901 is used to receive a logical node creation request input by a tenant. The logical node creation request indicates the resource requirements of the logical node, including computing resource requirements, storage resource requirements, and network resource requirements. For example, the receiving module 901 is used to implement... Figure 3 Step 301 in the illustrated embodiment.

[0109] The determination module 902 is used to determine, based on a logical node creation request, at least one physical server from multiple physical servers that meets the computing and storage resource requirements, at least one smart network interface card (NIC) from multiple smart NICs that meets the network resource requirements, and at least one SPU from multiple SPUs for managing the at least one physical server and the at least one smart NIC; for example, the determination module 902 is used to implement Figure 3 Step 302 in the illustrated embodiment.

[0110] Creation module 903 is used to notify at least one SPU to create a logical node in at least one physical server and at least one smart network interface card (NIC). Each smart NIC contains at least one port resource bound to at least one physical server, and this port resource is used to forward packets from or directed to the logical node. For example, creation module 903 is used to implement... Figure 3 Step 303 in the illustrated embodiment.

[0111] In one possible implementation, at least one port resource of the smart network interface card (NIC) is allocated from the same port of the smart NIC, or at least one port resource of the smart NIC is allocated from different ports of the smart NIC.

[0112] In one possible implementation, at least one SPU includes a first SPU, at least one smart network interface card (NIC) includes a first smart NIC, and the cloud management platform further includes: a first notification module, used to determine that the first smart NIC is faulty, and to notify the first SPU to select a second smart NIC from the smart NICs managed by the first SPU, wherein the second smart NIC does not belong to at least one smart NIC; and a second notification module, used to notify the first SPU to release the first binding relationship between at least one port resource of the first smart NIC and at least one physical server, to establish a second binding relationship between at least one port resource of the second smart NIC and at least one physical server, and to control at least one port resource of the second smart NIC to re-forward a first message based on the second binding relationship, wherein the first message is the message that at least one port resource of the first smart NIC is forwarding when the first smart NIC is faulty.

[0113] In one possible implementation, at least one SPU includes a first SPU and a second SPU, and at least one smart network interface card (NIC) includes a third smart NIC and a fourth smart NIC. The first SPU records a third binding relationship between at least one port resource of the third smart NIC and at least one physical server, and the second SPU records a fourth binding relationship between at least one port resource of the fourth smart NIC and at least one physical server. The cloud management platform further includes: a migration module, used to determine that the first SPU is faulty and migrate the third binding relationship between at least one port resource of the third smart NIC and at least one physical server from the first SPU to the second SPU; and a third notification module, used to notify the second SPU to control at least one port resource of the third smart NIC to re-forward a second message based on the third binding relationship, wherein the second message is the message that at least one port resource of the third smart NIC is forwarding when the first SPU is faulty.

[0114] In one possible implementation, at least one SPU includes a first SPU and a second SPU, and at least one smart network interface card (NIC) includes a third smart NIC and a fourth smart NIC. The first SPU records a third binding relationship between at least one port resource of the third smart NIC and at least one physical server, and the second SPU records a fourth binding relationship between at least one port resource of the fourth smart NIC and at least one physical server. The cloud management platform further includes a fourth notification module, used to determine that the first SPU is faulty and notify the second SPU to control at least one port resource of the fourth smart NIC to re-forward a second message based on the fourth binding relationship. The second message is the message that at least one port resource of the third smart NIC is forwarding when the first SPU is faulty.

[0115] In one possible implementation, the high-speed interconnect network is implemented via a PCIe network, an IB network, or a CXL network.

[0116] In one possible implementation, a logical node contains at least one virtual machine, at least one container, or at least one microvirtual machine.

[0117] It should be noted that the information interaction and implementation process between the modules / units of the above-mentioned device are based on the same concept as the method embodiments of this application, and the resulting technical effects are the same as those of the method embodiments of this application. For details, please refer to the description in the method embodiments shown above in the embodiments of this application, and will not be repeated here.

[0118] Please see Figure 10 , Figure 10 This is a schematic diagram of the structure of a computing device provided in an embodiment of this application. Figure 10 As shown, the computing device 1000 (which can be used to present the aforementioned cloud management platform) includes: a processor 1001, a memory 1002, a communication interface 1003, and a bus 1004. The processor 1001, memory 1002, and communication interface 1003 are coupled via the bus (not shown in the figure). The memory 1002 stores instructions. When the execution instructions in the memory 1002 are executed, the computing device 1000 executes the method performed by the cloud management platform in the above method embodiment.

[0119] The computing device 1000 may be one or more integrated circuits configured to implement the methods described above, such as: one or more application-specific integrated circuits (ASICs), or one or more digital signal processors (DSPs), or one or more field-programmable gate arrays (FPGAs), or a combination of at least two of these forms of integrated circuits. Furthermore, when the units in the device can be implemented in the form of a processing element scheduler, the processing element may be a general-purpose processor, such as a central processing unit (CPU) or other processor capable of calling programs. Alternatively, these units may be integrated together and implemented as a system-on-a-chip (SOC).

[0120] The processor 1001 can be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. A general-purpose processor can be a microprocessor or any conventional processor.

[0121] The memory 1002 can be volatile memory or non-volatile memory, or it can include both. The non-volatile memory can be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. The volatile memory can be random access memory (RAM), which is used as an external cache. By way of example, but not limitation, many forms of RAM are available, such as static random access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDR SDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous linked dynamic random access memory (SLDRAM), and direct rambus RAM (DR RAM).

[0122] The memory 1002 stores executable program code, and the processor 1001 executes the executable program code to implement the functions of the aforementioned receiving module, determining module, and creating module, thereby realizing the above-mentioned logical node configuration method based on the cloud management platform. That is, the memory 1002 stores instructions for executing the above-mentioned logical node configuration method based on the cloud management platform.

[0123] The communication interface 1003 uses transceiver modules such as, but not limited to, network interface cards and transceivers to enable communication between the computing device 1000 and other devices or communication networks.

[0124] In addition to the data bus, the 1004 bus can also include a power bus, a control bus, and a status signal bus. The bus can be a Peripheral Component Interconnect Express (PCIe) bus, an Extended Industry Standard Architecture (EISA) bus, a Unified Bus (Ubus or UB), a Compute Express Link (CXL) bus, a Cache Coherent Interconnect for Accelerators (CCIX) bus, etc. The bus can be divided into address bus, data bus, and control bus.

[0125] Please see Figure 11 , Figure 11 This is a schematic diagram of a computing device cluster provided in an embodiment of this application. Figure 11 As shown, the computing device cluster 1100 includes at least one computing device 1000.

[0126] like Figure 11 As shown, the computing device cluster 1100 includes at least one computing device 1000. The memory 1002 of one or more computing devices 1000 in the computing device cluster 1100 may store the same instructions for executing the above-described logical node configuration method based on the cloud management platform.

[0127] In some possible implementations, the memory 1002 of one or more computing devices 1000 in the computing device cluster 1100 may also store partial instructions for executing the aforementioned logical node configuration method based on the cloud management platform. In other words, a combination of one or more computing devices 1000 can jointly execute the aforementioned logical node configuration method based on the cloud management platform.

[0128] It should be noted that the memory 1002 in different computing devices 1000 within the computing device cluster 1100 can store different instructions, which are used to execute certain functions of the aforementioned cloud management platform. That is, the instructions stored in the memory 1002 of different computing devices 1000 can implement the functions of one or more modules, such as the receiving module, the determining module, and the creating module.

[0129] In some possible implementations, one or more computing devices 1000 in the computing device cluster 1100 can be connected via a network. This network can be a wide area network (WAN) or a local area network (LAN), etc.

[0130] Please see Figure 12 , Figure 12 This is a schematic diagram illustrating the network connection of computer devices in a computer cluster provided in an embodiment of this application. Figure 12 As shown, the two computing devices 1000A and 1000B are connected via a network. Specifically, they are connected to the network through the communication interfaces in each computing device.

[0131] In one possible implementation, the memory in computing device 1000A stores instructions for performing the functions of modules such as the receiving module. Meanwhile, the memory in computing device 1000B stores instructions for performing the functions of modules such as the determining module and the creating module.

[0132] It should be understood that Figure 12 The functions of computing device 1000A shown can also be performed by multiple computing devices. Similarly, the functions of computing device 1000B can also be performed by multiple computing devices.

[0133] This application also relates to a computer storage medium storing a program for signal processing, which, when run on a computer, causes the computer to perform actions such as... Figure 3 The steps performed by the cloud management platform in the illustrated embodiment.

[0134] This application also relates to a computer program product that stores instructions that, when executed by a computer, cause the computer to perform actions such as... Figure 3 The steps performed by the cloud management platform in the illustrated embodiment.

[0135] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0136] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection between apparatuses or units through some interfaces, and may be electrical, mechanical, or other forms.

[0137] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0138] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0139] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

Claims

1. A logical node configuration method based on a cloud management platform, characterized in that, The cloud management platform is used to manage the infrastructure providing cloud services. The infrastructure includes a physical server pool, a smart NIC pool, and a Service Processing Unit (SPU) pool. Multiple physical servers in the physical server pool, multiple smart NICs in the smart NIC pool, and multiple SPUs in the SPU pool are interconnected via a high-speed interconnect network. The multiple SPUs are used to manage the multiple physical servers and the multiple smart NICs. The method includes: The cloud management platform receives a logical node creation request input by a tenant. The logical node creation request is used to indicate the resource requirements of the logical node, including computing resource requirements, storage resource requirements, and network resource requirements. Based on the logical node creation request, the cloud management platform determines at least one physical server from the plurality of physical servers that meets the computing resource requirements and the storage resource requirements, determines at least one smart network interface card from the plurality of smart network interface cards that meets the network resource requirements, and determines at least one SPU from the plurality of SPUs for managing the at least one physical server and the at least one smart network interface card. The cloud management platform notifies the at least one SPU to create the logical node in the at least one physical server and the at least one smart network interface card (NIC), wherein any one of the at least one smart NICs contains at least one port resource bound to the at least one physical server, and the at least one port resource of the smart NIC is used to forward packets from the logical node or packets directed to the logical node.

2. The method according to claim 1, characterized in that, At least one port resource of the smart network card is obtained by allocating the same port of the smart network card, or at least one port resource of the smart network card is obtained by allocating different ports of the smart network card.

3. The method according to claim 1 or 2, characterized in that, The at least one SPU includes a first SPU, the at least one smart network interface card (NIC) includes a first smart NIC, and the method further includes: The cloud management platform determines that the first smart network interface card (NIC) is faulty and notifies the first SPU to select a second smart NIC from the smart NICs managed by the first SPU, wherein the second smart NIC does not belong to the at least one smart NIC. The cloud management platform notifies the first SPU to release the first binding relationship between at least one port resource of the first smart network card and the at least one physical server, establish a second binding relationship between at least one port resource of the second smart network card and the at least one physical server, and control at least one port resource of the second smart network card to re-forward the first message based on the second binding relationship. The first message is the message that at least one port resource of the first smart network card is forwarding when the first smart network card fails.

4. The method according to any one of claims 1 to 3, characterized in that, The at least one SPU includes a first SPU and a second SPU, and the at least one smart network interface card (NIC) includes a third smart NIC and a fourth smart NIC. The first SPU records a third binding relationship between at least one port resource of the third smart NIC and the at least one physical server. The second SPU records a fourth binding relationship between at least one port resource of the fourth smart NIC and the at least one physical server. The method further includes: The cloud management platform determines that the first SPU is faulty and migrates the third binding relationship between at least one port resource of the third smart network card and the at least one physical server from the first SPU to the second SPU. The cloud management platform notifies the second SPU to control at least one port resource of the third smart network card to re-forward the second message based on the third binding relationship, wherein the second message is the message that the third smart network card is forwarding when the first SPU fails.

5. The method according to any one of claims 1 to 3, characterized in that, The at least one SPU includes a first SPU and a second SPU, and the at least one smart network interface card (NIC) includes a third smart NIC and a fourth smart NIC. The first SPU records a third binding relationship between at least one port resource of the third smart NIC and the at least one physical server. The second SPU records a fourth binding relationship between at least one port resource of the fourth smart NIC and the at least one physical server. The method further includes: The cloud management platform determines that the first SPU is faulty and notifies the second SPU to control at least one port resource of the fourth smart network card to re-forward the second message based on the fourth binding relationship. The second message is the message that the third smart network card is forwarding when the first SPU is faulty.

6. The method according to any one of claims 1 to 5, characterized in that, The high-speed interconnect network is implemented through a high-speed serial computer expansion bus standard PCIE network, an unlimited bandwidth IB network, or a compute expansion connection CXL network.

7. The method according to any one of claims 1 to 6, characterized in that, The logical node contains at least one virtual machine, at least one container, or at least one micro virtual machine.

8. A cloud management platform, characterized in that, The cloud management platform is used to manage the infrastructure providing cloud services. The infrastructure includes a physical server pool, a smart NIC pool, and an SPU pool. Multiple physical servers in the physical server pool, multiple smart NICs in the smart NIC pool, and multiple SPUs in the SPU pool are interconnected via a high-speed interconnect network. The multiple SPUs are used to manage the multiple physical servers and the multiple smart NICs. The cloud management platform includes: A receiving module is used to receive a logical node creation request input by a tenant, wherein the logical node creation request is used to indicate the resource requirements of the logical node, and the resource requirements include computing resource requirements, storage resource requirements and network resource requirements; The determination module is configured to, based on the logical node creation request, determine at least one physical server from the plurality of physical servers that meets the computing resource requirements and the storage resource requirements, determine at least one smart network interface card from the plurality of smart network interface cards that meets the network resource requirements, and determine at least one SPU from the plurality of SPUs for managing the at least one physical server and the at least one smart network interface card. A creation module is used to notify the at least one SPU to create the logical node in the at least one physical server and the at least one smart network interface card (NIC), wherein any one of the at least one smart NICs contains at least one port resource bound to the at least one physical server, and the at least one port resource of the smart NIC is used to forward packets from the logical node or packets directed to the logical node.

9. The cloud management platform according to claim 8, characterized in that, At least one port resource of the smart network card is obtained by allocating the same port of the smart network card, or at least one port resource of the smart network card is obtained by allocating different ports of the smart network card.

10. The cloud management platform according to claim 8 or 9, characterized in that, The at least one SPU includes a first SPU, the at least one smart network interface card (NIC) includes a first smart NIC, and the cloud management platform further includes: The first notification module is used to determine that the first smart network interface card (NIC) is faulty and to notify the first SPU to select a second smart network interface card from the smart NICs managed by the first SPU, wherein the second smart network interface card does not belong to the at least one smart network interface card. The second notification module is used to notify the first SPU to release the first binding relationship between at least one port resource of the first smart network card and the at least one physical server, establish a second binding relationship between at least one port resource of the second smart network card and the at least one physical server, and control at least one port resource of the second smart network card to re-forward the first message based on the second binding relationship, wherein the first message is the message that at least one port resource of the first smart network card is forwarding when the first smart network card fails.

11. The cloud management platform according to any one of claims 8 to 10, characterized in that, The at least one SPU includes a first SPU and a second SPU, the at least one smart network interface card (NIC) includes a third smart NIC and a fourth smart NIC, the first SPU records a third binding relationship between at least one port resource of the third smart NIC and the at least one physical server, the second SPU records a fourth binding relationship between at least one port resource of the fourth smart NIC and the at least one physical server, and the cloud management platform further includes: A migration module is used to determine that the first SPU is faulty and to migrate the third binding relationship between at least one port resource of the third smart network card and the at least one physical server from the first SPU to the second SPU. The third notification module is used to notify the second SPU to control at least one port resource of the third smart network card to re-forward the second message based on the third binding relationship, wherein the second message is the message that the third smart network card is forwarding when the first SPU fails.

12. The cloud management platform according to any one of claims 8 to 10, characterized in that, The at least one SPU includes a first SPU and a second SPU, the at least one smart network interface card (NIC) includes a third smart NIC and a fourth smart NIC, the first SPU records a third binding relationship between at least one port resource of the third smart NIC and the at least one physical server, the second SPU records a fourth binding relationship between at least one port resource of the fourth smart NIC and the at least one physical server, and the cloud management platform further includes: The fourth notification module is used to determine that the first SPU is faulty and to notify the second SPU to control at least one port resource of the fourth smart network card to re-forward the second message based on the fourth binding relationship, wherein the second message is the message that the third smart network card is forwarding at least one port resource when the first SPU is faulty.

13. The cloud management platform according to any one of claims 8 to 12, characterized in that, The high-speed interconnect network is implemented through a PCIE network, an IB network, or a CXL network.

14. The cloud management platform according to any one of claims 8 to 13, characterized in that, The logical node contains at least one virtual machine, at least one container, or at least one micro virtual machine.

15. A computing device cluster, characterized in that, The computing device cluster includes at least one computing device, each computing device including a processor and memory: The memory is used to store instructions; The processor is configured to, according to the instructions, cause the computing device cluster to perform the method of any one of claims 1 to 7.

16. A computer storage medium, characterized in that, The computer storage medium stores one or more instructions that, when executed by one or more computers, cause the one or more computers to perform the method of any one of claims 1 to 7.

17. A computer program product, characterized in that, The computer program product stores instructions that, when executed by a computer, cause the computer to perform the method described in any one of claims 1 to 7.

Images