Distributed cloud node scheduling management method and device, medium and equipment

Abstract

The application provides a distributed cloud node scheduling management method and device, medium and equipment. The method comprises the following steps: creating a distributed cloud computing task according to actual requirements in an application scenario; splitting the distributed cloud computing task into a plurality of cloud computing subtasks; creating corresponding computing power on a cloud node of a cloud branch of a distributed cloud for each cloud computing subtask, and allocating corresponding IPv6 resources to the computing power; determining the interconnection mode between each cloud computing subtask, creating a corresponding virtual interconnection channel according to the interconnection mode; each computing power executes the corresponding cloud computing subtask, and data interaction is realized based on the IPv6 resources and the virtual interconnection channel in the computing process; and after each cloud computing subtask is executed, the calculation results are summarized and result feedback is performed. The distributed cloud computing power provided based on the IPv6 protocol can meet the needs of cloud development.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of cloud computing, and in particular to a distributed cloud node scheduling management method and device, medium and equipment. BACKGROUND

[0002] With the rapid development and wide application of cloud computing, the form of cloud is constantly evolving, and distributed cloud has become a development trend of future computing form, and is one of the key directions of the future decisive battle of the entire cloud computing industry. However, the distributed cloud computing power provided based on the IPv4 protocol cannot meet the needs of cloud development, for example, the length of IP address in the IPv4 protocol is 32, and the maximum number of addresses is 2^32, which seriously limits the number of IP resources of the distributed cloud. SUMMARY

[0003] In view of at least one of the above technical problems, the present application embodiment provides a distributed cloud node scheduling management method and device, medium and equipment.

[0004] According to a first aspect, the present application embodiment provides a distributed cloud node scheduling management method, which comprises:

[0005] According to the actual needs in the application scenario, a distributed cloud computing task is created;

[0006] The distributed cloud computing task is split into multiple cloud computing sub-tasks;

[0007] For each cloud computing sub-task, a corresponding computing power is created on a cloud node of a cloud branch of the distributed cloud, and the computing power is allocated a corresponding IPv6 resource; wherein the computing power is a container or a virtual machine, and the IPv6 resource includes an IPv6 address, an IPv6 route and an IPv6 switch;

[0008] The interconnection mode between each cloud computing sub-task is determined, and a corresponding virtual interconnection channel is created according to the interconnection mode; wherein the interconnection mode includes the interconnection mode between different computing powers in the same cloud node, the interconnection mode between the cloud nodes of the computing powers, and the interconnection mode between the cloud branches to which the cloud nodes belong; the virtual interconnection channel refers to a virtual connection channel established on the basis of physical connection;

[0009] Each computing power executes the corresponding cloud computing sub-task, and realizes data interaction based on the IPv6 resource and the virtual interconnection channel in the computing process;

[0010] After each cloud computing sub-task is executed, the calculation results are summarized and result feedback is performed.

[0011] In one embodiment, the method provided by the present application embodiment can further comprise:

[0012] The IPv6 network management module is configured to implement at least one of IPv6 resource pool management, IPv6 address management, IPv6 route management, IPv6 switch management, and dual stack management.

[0013] The IPv6 resource pool management refers to management of an IPv6 address segment in the IPv6 resource pool, the IPv6 address management refers to management of an IPv6 address allocated for computing power, the IPv6 route refers to management of IPv6 route information of a cloud node, the IPv6 switch management refers to management of a first IPv6 switch between different cloud nodes in a same cloud branch and a second IPv6 switch between different cloud branches, and the dual stack management refers to management of a correspondence between an IPv6 address and an IPv4 address.

[0014] The different cloud nodes in the same cloud branch are physically connected through the first IPv6 switch, and the first IPv6 switches of different cloud branches are physically connected through the second IPv6 switch.

[0015] In one embodiment, each cloud node has its own node network card, the node network card of each cloud node is connected with the first IPv6 switch in the cloud branch where the cloud node is located, and the node network card of each cloud node is connected with each computing power in the cloud node through a container network interface.

[0016] In one embodiment, the method provided by the embodiment of the application further includes:

[0017] The IPv6 network interconnection module is configured to implement management of various interconnection modes, the interconnection modes include a tunnel protocol-based interconnection mode, a border gateway protocol-based interconnection mode, and an eBPF module-based interconnection mode, and the eBPF module is an extended Berkeley Packet Filter module.

[0018] In one embodiment, each cloud node has its own node network card, virtual network card, and tunnel network card, and the node network card, the virtual network card, and the tunnel network card of each cloud node are connected with each other.

[0019] Each computing power in each cloud node is interconnected with the virtual network card in the cloud node, the virtual network card in each cloud node is interconnected with the tunnel network card in the cloud node through a tunnel protocol, and the tunnel network cards of different cloud nodes are interconnected through the first IPv6 switch.

[0020] The virtual network card in each cloud node is interconnected with the node network card in the cloud node through a border gateway protocol or through the eBPF module, and the node network cards of different cloud nodes are interconnected through the first IPv6 switch.

[0021] In one embodiment, the method provided by the embodiments of the present application further comprises:

[0022] determining the computing power resources required by the plurality of cloud computing sub-tasks;

[0023] determining the scheduling mode of the computing power resources;

[0024] Correspondingly, the creating, for each cloud computing sub-task, of corresponding computing power on a cloud node of a cloud branch of the distributed cloud comprises:

[0025] scheduling the corresponding computing power resources according to the scheduling mode;

[0026] According to the scheduled computing power resources, the corresponding computing power is created for each cloud computing sub-task on a cloud node in a cloud branch of the distributed cloud.

[0027] In one embodiment, the scheduling mode comprises a real-time scheduling mode and a batch scheduling mode.

[0028] According to a second aspect, the distributed cloud node scheduling management apparatus provided by the embodiments of the present application comprises:

[0029] a task creating module configured to create a distributed cloud computing task according to actual requirements in an application scenario;

[0030] a task splitting module configured to split the distributed cloud computing task into a plurality of cloud computing sub-tasks;

[0031] a computing power creating module configured to create, for each cloud computing sub-task, corresponding computing power on a cloud node of a cloud branch of the distributed cloud, and to allocate corresponding IPv6 resources to the computing power; wherein the computing power is a container or a virtual machine, and the IPv6 resources comprise an IPv6 address, an IPv6 route and an IPv6 switch;

[0032] an interconnection opening module configured to determine an interconnection mode between the cloud computing sub-tasks, and to create corresponding virtual interconnection channels according to the interconnection mode; wherein the interconnection mode comprises an interconnection mode between different computing powers in the same cloud node, an interconnection mode between the cloud nodes where the computing powers are located, and an interconnection mode between the cloud branches to which the cloud nodes belong; the virtual interconnection channel is a virtual connection channel established on the basis of physical connection;

[0033] a task computing module configured to execute the respective cloud computing sub-tasks by the computing powers, and to realize data interaction based on the IPv6 resources and the virtual interconnection channels during the computation;

[0034] a result returning module configured to aggregate and feed back the computation results after the execution of the cloud computing sub-tasks.

[0035] According to a third aspect, the embodiments of the present application provide a computer readable storage medium, having stored thereon a computer program which, when executed in a computer, causes the computer to perform the method according to the first aspect.

[0036] According to a fourth aspect, the embodiments of the present application provide a computing device, comprising a memory and a processor, wherein the memory stores executable code, and the processor executes the executable code to implement the method according to the first aspect.

[0037] The distributed cloud node scheduling management method and device, medium and equipment provided by the embodiments of the present application first create a distributed cloud computing task according to actual needs, split the distributed cloud computing task into multiple cloud computing subtasks, then create corresponding computing power on a cloud node of a cloud branch of the distributed cloud for each cloud computing subtask, and allocate corresponding IPv6 resources to the computing power, then determine the interconnection mode between the cloud computing subtasks, and create corresponding virtual interconnection channels according to the interconnection mode; then each computing power executes the corresponding cloud computing subtask, and realizes data interaction based on the IPv6 resources and the virtual interconnection channels in the computing process; and after the execution of each cloud computing subtask, the computing results are summarized and result feedback is performed. It can be seen that the resources allocated to the computing power in the embodiments of the present application are all IPv6 resources, so the entire method is implemented based on the IPv6 protocol. Compared with IPv4, IPv6 has a larger address space, uses a smaller routing table, increases enhanced multicast support and control of flow, adds support for automatic configuration, has higher security, and the like, so the distributed cloud computing power provided based on the IPv6 protocol can meet the needs of cloud development. BRIEF DESCRIPTION OF DRAWINGS

[0038] Figure 1 FIG. 1 is a flowchart of a distributed cloud node scheduling management method according to an embodiment of the present application;

[0039] Figure 2 FIG. 3 is a schematic diagram of management functions involved in an IPv6 network management module according to an embodiment of the present application;

[0040] Figure 3 FIG. 4 is a schematic diagram of an interconnection mode according to an embodiment of the present application;

[0041] Figure 4 FIG. 5 is a structural block diagram of a distributed cloud node scheduling management device according to an embodiment of the present application. DETAILED DESCRIPTION

[0042] In a first aspect, the embodiments of the present application provide a distributed cloud node scheduling management method, as shown in Figure 1The method comprises the following steps S110-S160:

[0043] S110, creating a distributed cloud computing task according to actual requirements in an application scenario;

[0044] S120, splitting the distributed cloud computing task into a plurality of cloud computing subtasks;

[0045] Generally, the computing process of the distributed cloud computing task is relatively complex, and the distributed cloud computing task can be split into a plurality of cloud computing subtasks, each of which can be executed by one computing power. However, in the execution process of one cloud computing subtask, the computing result or intermediate result of another cloud computing subtask can be required, and therefore the interconnection between the cloud computing subtasks needs to be established so as to transmit the required data.

[0046] S130, for each cloud computing subtask, creating a corresponding computing power on a cloud node of one cloud branch of the distributed cloud, and allocating a corresponding IPv6 resource to the computing power; wherein the computing power is a container or a virtual machine, and the IPv6 resource comprises an IPv6 address, an IPv6 route and an IPv6 switch;

[0047] Wherein, the computing power refers to a virtual machine or a container.

[0048] That is to say, for each cloud computing subtask, a corresponding virtual machine or container is created on a cloud node in one cloud branch of the distributed cloud, and then a corresponding IPv6 resource is allocated to the virtual machine or the container.

[0049] It can be understood that, with the continuous expansion of the global IP network scale and the rapid growth of the number of users, the IP network will be unable to continue to develop due to the depletion of address space, and the next generation of Internet Protocol IPv6 emerges as the times require. IPv6 is the focus of the industry today, and in the future development and competition of the world information industry, IPv6 will also become a winning magic weapon.

[0050] Compared with IPv4, IPv6 has the following advantages: IPv6 has a larger address space; IPv6 uses a smaller routing table; IPv6 increases enhanced multicast support and control over flow; IPv6 adds support for automatic configuration; IPv6 has higher security; allows expansion; has a better header format; supports new options to implement additional functions.

[0051] It can be understood that 5G, industrial internet, rail transit and the like are included in new infrastructure, and the integration of the real economy field and the internet is gradually deepened. The device scale and service scale carried by the internet are increased, and higher requirements are put forward for the delay, security and reliability, service level division and the like of network communication. As the information infrastructure of new infrastructure, cloud computing will enter a new era of development. With the development of cloud computing, the scale of distributed cloud is getting larger and larger, and based on the advantages of IPv6 compared with IPv6, supporting IPv6 for ten-level distributed cloud nodes becomes a feasible technical direction.

[0052] Therefore, in the embodiment of the application, the computing power allocation is an IPv6 resource, which includes an IPv6 address, an IPv6 route and an IPv6 switch. The IPv6 address refers to an IP address based on the IPv6 protocol, the IPv6 route refers to a route based on the IPv6 protocol, and the IPv6 switch refers to a switch based on the IPv6 protocol.

[0053] In S140, the interconnection mode between each cloud computing subtask is determined, and a corresponding virtual interconnection channel is created according to the interconnection mode. The interconnection mode includes the interconnection mode between different computing powers in the same cloud node, the interconnection mode between the cloud nodes where the computing powers are located, and the interconnection mode between the cloud branches to which the cloud nodes belong. The virtual interconnection channel refers to a virtual connection channel established on the basis of physical connection.

[0054] Since the computing power corresponding to each cloud computing subtask can be located in the same cloud node, can be located in different cloud nodes of the same cloud branch, and can be located in different cloud branches, the interconnection mode between each cloud computing subtask, that is, the interconnection mode between each computing power, can be multiple. Based on all the interconnection modes, virtual interconnection channels between each computing power are created.

[0055] For example, the computing power a12 corresponding to the cloud computing subtask a11 and the computing power a21 corresponding to the cloud computing subtask a12 are in the same node a, the computing power b12 corresponding to the cloud computing subtask b11 is located in the node b, the computing power c12 corresponding to the cloud computing subtask c11 is located in the node c, the node a and the node b are in the cloud branch 1, and the node c is in the cloud branch 2. The virtual interconnection channel between the computing power a12 and the computing power a21, the virtual interconnection channel between the node a and the node b, and the virtual interconnection channel between the cloud branch 1 and the cloud branch 2 need to be established, so that the computing power a12, the computing power a21, the computing power b12 and the computing power c12 can interact with each other.

[0056] Each computing power has its own IPv6 address, and based on the allocated IPv6 route and IPv6 switch, physical connection between the cloud nodes where each computing power is located can be realized.

[0057] The virtual interconnection channel is established on the basis of physical connection, that is, each computing power needs to have a physical connection, and then a virtual connection is established on the basis of the physical connection, so that each computing power can interact through the virtual interconnection channel when performing the respective cloud computing subtask.

[0058] S150, each computing power performs a respective corresponding cloud computing subtask, and data interaction is realized based on the IPv6 resource and the virtual interconnection channel during the calculation process;

[0059] It can be seen that after the virtual interconnection channel between each computing power is established, each computing power can perform a respective cloud computing subtask, and then during the execution of the cloud computing subtask, data interaction based on the IPv6 resource and the virtual interconnection channel is required to deliver necessary data through interaction.

[0060] S160, after each cloud computing subtask is executed, the calculation results are summarized and result feedback is performed.

[0061] That is, after each cloud computing subtask is executed, a plurality of calculation results can be obtained, and then the calculation results are summarized to obtain the calculation result of the entire distributed cloud computing task, and then the calculation result of the entire distributed cloud computing task is returned to the user, thereby completing the calculation of the distributed cloud computing task.

[0062] In one embodiment, the method provided by the embodiment of the application can further include:

[0063] The IPv6 network management module is used to implement at least one of IPv6 resource pool management, IPv6 address management, IPv6 routing management, IPv6 switch management, and dual stack management.

[0064] The IPv6 resource pool management refers to management of an IPv6 address segment in the IPv6 resource pool, the IPv6 address management refers to management of an IPv6 address allocated to the computing power, the IPv6 routing refers to management of Ipv6 routing information of the cloud node, the Ipv6 switch management refers to management of a first Ipv6 switch between different cloud nodes in the same cloud branch and a second Ipv6 switch between different cloud branches, and the dual stack management refers to management of a corresponding relationship between an IPv6 address and an Ipv4 address.

[0065] The different cloud nodes in the same cloud branch are physically connected through the first Ipv6 switch, and the first Ipv6 switches of different cloud branches are physically connected through the second Ipv6 switch.

[0066] It can be seen that the IPv6 network management module can realize IPv6 network management, including IPv6 resource pool management, IPv6 address management, IPv6 routing management, IPv6 switch management, IPv6-IPv4 dual stack management and other functions, providing greater IP address space for distributed cloud, improving distributed cloud computing network fusion capability, improving computing power communication security, and improving distributed cloud network certainty.

[0067] Among them, the IPv6 resource pool management is the management of IPv6 addresses that can be allocated to computing power, the IPv6 address management refers to the management of IPv6 addresses that have been allocated to computing power, the IPv6 routing information management refers to the management of IPv6 routing information involved in the distributed cloud, and the IPv6 switch management refers to the management of IPv6 switches involved in the distributed cloud.

[0068] It can be seen that the IPv6 network management module uniformly allocates IPv6 resources to each cloud branch in the distributed cloud, supports IPv6 address allocation and routing configuration of each node in the cloud branch, supports network configuration of all switches in the distributed cloud, and ensures network interconnection between each cloud branch based on IPv6 protocol.

[0069] Since the IPv6 network management module can realize unified management of IPv6 resources of virtual machines and containers in the cloud branch, support network interconnection of virtual machines and containers in the cloud branch based on IPv6, fully utilize the characteristics of IPv6 address to support distributed cloud with ten thousand node scale, and support IP address management of million computing power, it can improve the computing power network communication performance in the distributed cloud and strengthen the security of distributed cloud computing data.

[0070] In one embodiment, each cloud node has its own node network card, and the node network card of each cloud node is connected with the first IPv6 switch in the cloud branch where the cloud node is located. The node network card of each cloud node is connected with each computing power in the cloud node through a container network interface.

[0071] Referring to Figure 2Each cloud node has a node network card, and the node network card of a cloud node is connected with a first Ipv6 switch in the cloud branch where the cloud node is located, so that the first Ipv6 switches of the cloud branches can be connected through a second Ipv6 switch. In a cloud node, each computing power is connected with the node network card of the cloud node through a container network interface, which can also be called a CNI interface, which is represented in English as Container Network Interface. The container network interface can also be considered as a virtual machine network card. Through the above connection relationship, each computing power can be connected with the first Ipv6 switch through the container network interface and the node network card. The network card and the switch, and the switch and the switch are physically connected.

[0072] It can be understood that the IPv6 network resources are created for the computing power to support functions such as computing power IP management, routing management, switch management, and cloud branch management, to uniformly allocate IPv6 addresses for the computing power, and to realize cloud branch network data exchange and network data routing functions.

[0073] In one embodiment, the method provided by the embodiment of the application can further include:

[0074] The IPv6 network interconnection module is used to manage various interconnection modes, including a tunnel protocol-based interconnection mode, a border gateway protocol-based interconnection mode, and an eBPF module-based interconnection mode, wherein the eBPF module is an extended Berkeley Packet Filter module.

[0075] The tunnel protocol is a protocol used by a tunnel technology, and the essence of the tunnel technology is how to use a network layer protocol to transmit another network layer protocol. The basic function of the tunnel technology is encapsulation and encryption, which is mainly achieved by using a tunnel protocol. Encapsulation is a basic means of building a tunnel. From the two ends of the tunnel, encapsulation is used to create, maintain and cancel a tunnel to achieve information concealment and abstraction.

[0076] The border gateway protocol, i.e., BGP, is an autonomous system routing protocol running on TCP. BGP is the only protocol used to handle networks as large as the Internet, and is also the only protocol that can properly handle multiple connections between unrelated routing domains.

[0077] The extended Berkeley packet filter module is referred to as an eBPF module, which is a lightweight virtual machine based on the kernel. It allows users to run custom programs in the kernel to monitor and analyze system events in real time. eBPF programs are usually written in C language and compiled into eBPF bytecode by a special compiler. Then, these bytecodes will be loaded into the kernel and run on the eBPF virtual machine.

[0078] It can be seen that the IPv6 network interconnection module can realize the management of various interconnection modes, support tunnel protocol-based interconnection, BGP protocol-based interconnection, eBPF module-based interconnection, and other functions. The distributed cloud computing power realizes the interconnection of tunnel networks, BGP networks, or eBPF networks based on IPv6 technology.

[0079] It can be understood that the interconnection mentioned here is virtual interconnection.

[0080] It can be understood that the IPv6 network interconnection function is realized for computing power, supporting tunnel protocol interconnection, BGP protocol interconnection, eBPF interconnection, and other functions, realizing cloud branch interconnection, cloud node interconnection, computing power interconnection, and other functions, supporting distributed cloud computing power network communication, improving the quality of computing power network service, and realizing the integration of computing power and network.

[0081] Further, referring to Figure 3 Each cloud node has its own node network card, virtual network card, and tunnel network card; wherein:

[0082] Each computing power in each cloud node is interconnected with the virtual network card in the cloud node, and the virtual network card in each cloud node is interconnected with the tunnel network card in the cloud node through a tunnel protocol; the tunnel network cards of different cloud nodes are interconnected through a first IPv6 switch;

[0083] The virtual network card in each cloud node is interconnected with the node network card in the cloud node through a border gateway protocol, or through the eBPF module; the node network cards in different cloud nodes are interconnected through a first IPv6 switch.

[0084] That is, the computing powers in a cloud node can be virtually interconnected through a virtual network card. In a cloud node, the virtual network card can be interconnected with the tunnel network card through a tunnel protocol, and the virtual network card can be interconnected with the node network card through a border gateway protocol or an eBPF module. These are all virtual interconnections.

[0085] Of course, the node network cards on each cloud node in the same cloud branch are connected to the first IPv6 switch, and each cloud branch is connected through the second IPv6 switch, which is a physical connection.

[0086] Based on the above virtual connection and physical connection, the message sent by the virtual network card to the tunnel network card is encapsulated based on the tunnel protocol, sent to the tunnel network card of another cloud node through the first IPv6 switch between cloud nodes, and unpacked based on the tunnel protocol, so as to achieve the IPv6 network interconnection between the computing power of two cloud nodes.

[0087] Moreover, the message sent by the virtual network card to the node network card is sent to the node network card of another node through the first IPv6 switch between cloud nodes. Compared with the tunnel protocol, the BGP protocol reduces the packet unpacking process based on the tunnel protocol of the computing power communication, thereby improving the performance.

[0088] In addition, based on the eBPF module, the network communication between the virtual network card and the node network card can be accelerated, and the message sent by the virtual network card to the node network card can be sent to the node network card of another node through the first IPv6 switch between cloud nodes.

[0089] As can be seen, the distributed network interconnection technology based on IPv6 can balance the performance and reliability of message transmission, and at the same time, through the encryption feature enabled by both parties of communication, the transmission data is encrypted to ensure the security of communication, and is suitable for complex network environment.

[0090] In one embodiment, the method provided by the embodiment of the application can further include:

[0091] Determining the computing power resources required by the plurality of cloud computing sub-tasks;

[0092] Determining the scheduling mode of the computing power resources;

[0093] Correspondingly, the method further includes:

[0094] According to the scheduling mode, the corresponding computing power resources are scheduled;

[0095] According to the scheduled computing power resources, for each cloud computing sub-task, the corresponding computing power is created on a cloud node in a cloud branch of the distributed cloud.

[0096] That is, before creating the computing power, the computing power resources required by each cloud computing subtask need to be determined, and then the scheduling mode of the computing power resources is determined, which can be divided into real-time scheduling mode and batch scheduling mode. The real-time scheduling mode is a way of scheduling IPv6 resources in real time, and the batch scheduling mode is a way of scheduling IPv6 resources for multiple similar cloud computing subtasks. After determining the scheduling mode, the corresponding computing power resources are scheduled, and then the required computing power is created according to the scheduled computing power resources.

[0097] It can be understood that based on the real-time scheduling algorithm or the batch scheduling algorithm, the unified arrangement and scheduling of distributed cloud branch computing power resources are supported.

[0098] In summary, in order to meet the distributed cloud supporting larger scale, providing lower latency network, realizing the real algorithm network fusion capability, providing more smooth computing power for users, the application embodiment proposes a method for realizing management of ten-thousand-level distributed cloud node scheduling based on IPv6 technology. In the method, the network management technology is realized based on IPv6, thereby realizing the unified management of IPv6 resources, the distributed interconnection technology is realized based on IPv6, which is convenient for expanding the distributed cloud, and therefore helps to support the distributed cloud expansion capability of ten-thousand-level node scale. According to the actual demand of distributed cloud computing task, the corresponding computing power and IPv6 resources are allocated for each cloud computing subtask in the distributed cloud computing task, the on-demand scheduling and resource allocation capability of distributed cloud computing power are supported, the distributed cloud node interconnection and scheduling performance are improved, the service quality capability of the stable distributed network is provided, the core technology for building a large-scale cluster scheduling capability, safe and reliable cloud edge-end integrated distributed cloud computing platform is provided, and the automatic, intelligent, high-availability, high-performance distributed cloud computing power supply is realized.

[0099] In a second aspect, the application embodiment provides a distributed cloud node scheduling management device, which is referred to as Figure 4 The device comprises:

[0100] The task creation module is configured to create a distributed cloud computing task according to the actual demand in the application scenario.

[0101] The task splitting module is configured to split the distributed cloud computing task into a plurality of cloud computing subtasks.

[0102] The computing power creation module is configured to create corresponding computing power on a cloud node of a cloud branch of the distributed cloud for each cloud computing subtask, and allocate corresponding IPv6 resources to the computing power. The computing power is a container or a virtual machine, and the IPv6 resources include an IPv6 address, an IPv6 route and an IPv6 switch.

[0103] An interconnection opening module is configured to determine an interconnection mode between each cloud computing subtask, and create a corresponding virtual interconnection channel according to the interconnection mode; wherein the interconnection mode includes an interconnection mode between different computing powers in a same cloud node, an interconnection mode between cloud nodes where the computing powers are located, and an interconnection mode between cloud branches to which the cloud nodes are located; the virtual interconnection channel is a virtual connection channel established on the basis of physical connection;

[0104] A task computing module is configured to execute each corresponding cloud computing subtask by each computing power, and realize data interaction based on the IPv6 resource and the virtual interconnection channel during the computation;

[0105] A result returning module is configured to aggregate and feed back the computation result after each cloud computing subtask is executed.

[0106] In an embodiment, the device provided by the embodiment of the application further includes:

[0107] A first management module is configured to implement at least one of IPv6 resource pool management, IPv6 address management, IPv6 routing management, IPv6 switch management, and dual stack management by the IPv6 network management module; wherein the IPv6 resource pool management refers to management of an IPv6 address segment in the IPv6 resource pool, the IPv6 address management refers to management of an IPv6 address allocated for a computing power, the IPv6 routing refers to management of Ipv6 routing information of a cloud node, the Ipv6 switch management refers to management of a first Ipv6 switch between cloud nodes and a second Ipv6 switch between cloud branches, and the dual stack management refers to management of a corresponding relationship between an IPv6 address and an Ipv4 address.

[0108] Wherein, different cloud nodes in a same cloud branch are physically connected through a first Ipv6 switch, and first Ipv6 switches of different cloud branches are physically connected through a second Ipv6 switch.

[0109] In an embodiment, each cloud node has its own node network card, the node network card of each cloud node is connected with a first Ipv6 switch in a cloud branch where the cloud node is located, and the node network card of each cloud node is connected with each computing power in the cloud node through a container network interface.

[0110] In an embodiment, the device provided by the embodiment of the application further includes:

[0111] A second management module is configured to manage various interconnection modes through the IPv6 network interconnection module; the interconnection modes include a tunnel protocol-based interconnection mode, a border gateway protocol-based interconnection mode, and an eBPF module-based interconnection mode, and the eBPF module is an extended Berkeley Packet Filter module.

[0112] In one embodiment, each cloud node has its own node network card, virtual network card and tunnel network card; wherein:

[0113] Each computing power in each cloud node is interconnected with the virtual network card in the cloud node, the virtual network card in each cloud node is interconnected with the tunnel network card in the cloud node through a tunnel protocol, and the tunnel network cards of different cloud nodes are interconnected through a first IPv6 switch.

[0114] The virtual network card in each cloud node is interconnected with the node network card in the cloud node through a border gateway protocol or through the eBPF module, and the node network cards in different cloud nodes are interconnected through the first IPv6 switch.

[0115] In one embodiment, the device provided by the embodiment of the application can further include:

[0116] A first determination module is configured to determine the computing power resources required by the plurality of cloud computing sub-tasks.

[0117] A second determination module is configured to determine the scheduling mode of the computing power resources.

[0118] Correspondingly, the computing power creation module is specifically configured to: according to the scheduling mode, schedule the corresponding computing power resources; and according to the scheduled computing power resources, create corresponding computing power for each cloud computing sub-task on a cloud node in a cloud branch of the distributed cloud.

[0119] In one embodiment, the scheduling mode includes a real-time scheduling mode and a batch scheduling mode.

[0120] It can be understood that the explanations, specific embodiments, advantages, examples and the like of the system provided by the embodiments of the application can refer to the corresponding parts in the method provided by the first aspect, and will not be repeated here.

[0121] In a third aspect, the embodiments of the application provide a computer readable medium, and the computer readable medium stores computer instructions, and the computer instructions make the processor execute the method provided by the first aspect when executed by the processor.

[0122] Specifically, a system or an apparatus equipped with a storage medium on which a software program code for realizing the functions of any of the above-described embodiments is stored can be provided, and a computer (or CPU or MPU) of the system or the apparatus reads out and executes the program code stored in the storage medium.

[0123] In this case, the program code read out from the storage medium itself realizes the functions of any of the above-described embodiments, and therefore the program code and the storage medium storing the program code constitute a part of the present application.

[0124] Embodiments of the storage medium for providing the program code include a floppy disk, a hard disk, a magneto-optical disk, an optical disk (such as CD-ROM, CD-R, CD-RW, DVD-ROM, DVD-RAM, DVD-RW, DVD+RW), a magnetic tape, a non-volatile memory card, and a ROM. Alternatively, the program code can be downloaded from a server computer via a communication network.

[0125] Further, it should be understood that not only the program code read out by the computer, but also the operating system or the like operating on the computer based on the instructions of the program code can perform part or all of the actual operations to realize the functions of any of the above-described embodiments.

[0126] Further, it should be understood that the program code read out from the storage medium is written into a memory provided in an extension board inserted into the computer or a memory provided in an extension module connected to the computer, and then part or all of the actual operations are performed by a CPU or the like mounted on the extension board or the extension module based on the instructions of the program code to realize the functions of any of the above-described embodiments.

[0127] It should be understood that the explanation of the related contents, the detailed description, the beneficial effects, the examples, and the like of the computer readable medium provided by the embodiments of the present application can refer to the corresponding parts of the method provided by the first aspect, and will not be described here.

[0128] In a fourth aspect, an embodiment of the present specification provides a computing device, comprising a memory and a processor, wherein the memory stores executable code, and the processor executes the executable code to realize the method in any of the embodiments of the specification.

[0129] It should be understood that the explanation of the related contents, the detailed description, the beneficial effects, the examples, and the like of the computing device provided by the embodiments of the present application can refer to the corresponding parts of the method provided by the first aspect, and will not be described here.

[0130] The various embodiments in this specification are described in a progressive manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the apparatus embodiments are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the descriptions of the method embodiments.

[0131] Those skilled in the art will recognize that, in one or more of the examples above, the functions described in this invention can be implemented using hardware, software, widgets, or any combination thereof. When implemented in software, these functions can be stored in a computer-readable medium or transmitted as one or more instructions or code on a computer-readable medium.

[0132] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made on the basis of the technical solution of the present invention should be included within the scope of protection of the present invention.

Claims

1. A distributed cloud node scheduling and management method, characterized in that, include: Create distributed cloud computing tasks based on the actual needs of the application scenario; The distributed cloud computing task is broken down into multiple cloud computing sub-tasks; For each cloud computing subtask, corresponding computing power is created on a cloud node of a cloud branch of the distributed cloud, and corresponding IPv6 resources are allocated to the computing power; wherein, the computing power is a container or virtual machine, and the IPv6 resources include IPv6 addresses, IPv6 routes and IPv6 switches. Determine the interconnection method between each cloud computing subtask, and create corresponding virtual interconnection channels according to the interconnection method; wherein, the interconnection method includes the interconnection method between different computing powers in the same cloud node, the interconnection method between cloud nodes where the computing power is located, and the interconnection method between cloud branches to which the cloud node belongs; the virtual interconnection channel refers to a virtual connection channel established on the basis of physical connection; Each computing power executes its corresponding cloud computing subtask, and data interaction is achieved based on the IPv6 resources and the virtual interconnection channel during the computing process; After each cloud computing subtask is completed, the calculation results are summarized and fed back; this also includes: The IPv6 network management module enables the management of at least one of the following: IPv6 resource pool, IPv6 address, IPv6 route, IPv6 switch, and dual-stack management. The management of the IPv6 resource pool refers to the management of the IPv6 address range in the IPv6 resource pool; the management of the IPv6 address refers to the management of the IPv6 address allocated to computing power; the IPv6 routing refers to the management of the IPv6 routing information of cloud nodes; the management of the IPv6 switch refers to the management of the first IPv6 switch between cloud nodes and the second IPv6 switch between cloud branches; and the dual-stack management refers to the management of the correspondence between IPv6 addresses and IPv4 addresses. Within the same cloud branch, different cloud nodes are physically connected via a first IPv6 switch, while the first IPv6 switches of different cloud branches are physically connected via a second IPv6 switch. Each cloud node has its own node network interface card (NIC), which is connected to the first IPv6 switch in the cloud branch where the cloud node is located. Each cloud node's NIC is also connected to the various computing resources within that cloud node via container network interfaces. Also includes: The IPv6 network interconnection module enables the management of various interconnection methods; these interconnection methods include tunnel protocol-based interconnection, border gateway protocol-based interconnection, and eBPF module-based interconnection, where the eBPF module is an extended Berkeley packet filter module. Each cloud node has its own node network interface card (NIC), virtual NIC, and tunnel NIC; among which: Each computing power in each cloud node is interconnected with the virtual network interface card (NIC) in that cloud node, and the virtual NIC in each cloud node is interconnected with the tunnel NIC in that cloud node through a tunneling protocol; the tunnel NICs of different cloud nodes are interconnected through a first IPv6 switch. The virtual network interface card (NIC) in each cloud node is interconnected with the node NIC in that cloud node through the Border Gateway Protocol (BGP) or through the eBPF module; the node NICs in different cloud nodes are interconnected through the first IPv6 switch.

2. The method according to claim 1, characterized in that, Also includes: Determine the computing resources required for the multiple cloud computing subtasks; Determine the scheduling method for the computing resources; Correspondingly, for each cloud computing subtask, creating corresponding computing power on a cloud node of a cloud branch of the distributed cloud includes: According to the scheduling method, schedule the corresponding computing resources; Based on the scheduled computing resources, for each cloud computing subtask, the corresponding computing power is created on a cloud node in a cloud branch of the distributed cloud.

3. The method according to claim 2, characterized in that, The scheduling methods include real-time scheduling and batch scheduling.

Images