Network intercommunication method and apparatus, computer device, and storage medium

Abstract

The application relates to a network interworking method and device, computer equipment and a storage medium. The application can realize the connection of a virtual switch and a hardware switch through a switch control module, establish a neighbor relationship between a computing node and the hardware switch through a kernel routing node to form a dynamic routing neighbor, and realize the conversion between a bare metal network and a cloud host network through the kernel routing node, thereby realizing the two-layer interconnection and interworking of the cloud host and the bare metal server, and providing a two-layer interworking network with high network performance, high cost performance and strong expansibility.

Description

Technical Field

[0001] This application relates to the field of communication technology, and in particular to a network interconnection method, apparatus, computer equipment, and storage medium. Background Technology

[0002] Current cloud computing platforms offer cloud servers and bare metal services to meet users' varying computing power needs. Bare metal services allow users to independently install operating systems and configure networks. To provide better manageability and ease of use, cloud platforms need to level off and manage these two types of resources uniformly. Specifically, at the network level, cloud servers and bare metal servers require unified virtual machines (VPCs) and support Layer 2 and Layer 3 interconnectivity.

[0003] Currently, mainstream cloud platform virtual machine (VPC) networks all use overlay network models such as VxLAN (network virtualization). The core technology of overlay networks is based on MAC IN UDP / TCP (User Datagram Protocol / Transmission Control Protocol), which encapsulates Layer 2 network packets into Layer 4 UDP / TCP packets. Network request packets are encapsulated into overlay packets via the VTEP tunnel endpoint, sent to the peer VTEP tunnel endpoint for decapsulation, and then sent to the business system. Due to different implementation methods, the industry mainly divides overlay networks into two types: host overlay and network overlay. Host overlay uses compute node software (such as Open vSwitch) to implement the overlay tunnel endpoint (VTEP) and OpenFlow flow tables to implement the VxLAN protocol. Network overlay uses hardware switches to implement the overlay tunnel endpoint and BGP-EVPN to implement control plane routing. These different implementation methods lead to difficulties in interconnecting and managing the two types of networks.

[0004] To achieve a unified VPC network for bare metal cloud servers, current mainstream smart NIC solutions suffer from cost-effectiveness issues, software gateways have performance bottlenecks, and overlay networks are dependent on hardware devices. Summary of the Invention

[0005] Based on this, a network interconnection method, apparatus, computer equipment, and storage medium are provided to solve the current mainstream smart NIC solutions in the industry that have cost-effectiveness issues, soft gateways that have performance bottlenecks, and overlay networks that are dependent on hardware devices. This enables cloud hosts and bare metal Layer 2 interconnection and provides a high-performance, cost-effective, and highly scalable Layer 2 interconnection network.

[0006] On the one hand, a network interconnection method is provided, the method comprising:

[0007] Configure a virtual switch to connect to a virtual switch control module, and a hardware switch to connect to a physical switch control module. Connect the virtual switch control module and the physical switch control module through the switch control module. The virtual switch is used to connect to a cloud host, and the hardware switch is used to connect to a bare metal server.

[0008] Configure a kernel routing node, which establishes a dynamic tunnel neighbor relationship between the compute node and the hardware switch;

[0009] The computing nodes are connected to cloud hosts to form a cloud host network, and the hardware switches are connected to bare metal servers to form a bare metal network. The switch control module listens for routing events of the cloud host network through the virtual switch control module, and the switch control module listens for routing events of the bare metal network through the physical switch control module.

[0010] The cloud host network and the bare metal network are connected through the kernel routing node. When a cloud host goes online / offline, the hardware switch is notified through the kernel routing node. When a bare metal server goes online / offline, the virtual switch control module is notified through the kernel routing node.

[0011] In one embodiment, the network interconnection method further includes:

[0012] The physical switch control module is expanded to support VxLAN networks. When a user creates a VxLAN network, it automatically receives user requests, parses user request messages, and sends hardware switch configuration information to configure the hardware switch.

[0013] In one embodiment, the step of distributing hardware switch configuration information to configure the hardware switch includes:

[0014] Create a bridge domain and configure the route distinguisher (RD) and route target (RT);

[0015] Configure the logical interface that implements network virtualization functionality and enable the copying of the Network Identifier (VNI) header;

[0016] Configure the service access point.

[0017] In one embodiment, setting up a kernel routing node, wherein the kernel routing node establishes a dynamic tunnel neighbor relationship between the compute node and the hardware switch, includes:

[0018] Two servers were selected to run dynamic routing software (FRRouting) as kernel routing nodes;

[0019] The compute node runs dynamic routing software (FRRouting), runs dynamic tunneling (BGP EVPN), and establishes neighbor relationships with the kernel routing node;

[0020] The hardware switch establishes a dynamic tunnel (BGP EVPN) neighbor relationship with the kernel routing node.

[0021] In one embodiment, the switch control module listens to routing events of the cloud host network through the virtual switch control module, and the switch control module listens to routing events of the bare metal network through the physical switch control module, including:

[0022] Configure the switch control module to support interface with the virtual switch control module and the dynamic routing software (FRRouting);

[0023] When a bare metal server comes online or goes offline, the switch control module listens for dynamic tunnel (BGP EVPN) address advertisement routing events, generates a logical flow table (OVN flow table) corresponding to the virtual switch control module, and writes it into the database of the virtual switch control module.

[0024] When a cloud host comes online, the switch control module listens for the online event of the computing node, generates a corresponding dynamic tunnel (BGP EVPN) address advertisement route, and writes the dynamic tunnel (BGP EVPN) address advertisement route into the running dynamic routing software (FRRouting) to complete the interconnection between the cloud host network and the bare metal network.

[0025] In one embodiment, the cloud host network and the bare metal network are connected through the kernel routing node. When a cloud host goes online / offline, the hardware switch is notified through the kernel routing node. When a bare metal server goes online / offline, the virtual switch control module is notified through the kernel routing node, including:

[0026] Configure a translation proxy module in the kernel routing node;

[0027] When a cloud host goes online / offline, the conversion proxy module notifies the hardware switch; when a bare metal server goes online / offline, the conversion proxy module notifies the virtual switch control module.

[0028] In one embodiment, the step of notifying the hardware switch through the conversion proxy module when a cloud host goes online / offline, and notifying the virtual switch control module through the conversion proxy module when a bare metal server goes online / offline, includes:

[0029] The conversion proxy module is connected to the virtual switch control module and the running dynamic routing software (FRRouting). The switch control module is used to capture cloud host online and offline action events and update the local flow table.

[0030] The virtual switch control module is configured to listen for events, specifically to monitor the online and offline actions of cloud hosts on the computing node ports.

[0031] The dynamic routing software (FRRouting) is configured to use a callback function that listens for changes in the address advertisement route of the dynamic tunnel (BGP EVPN) for event callbacks. This callback function is used to actively notify the conversion agent module when an address advertisement route event is received.

[0032] When the conversion agent module captures the online / offline operation of the cloud host, it obtains the address table (MAC) and the tunnel endpoint (VTEP) information of the corresponding port of the cloud host and writes it into the system kernel. The running dynamic routing software (FRRouting) listens for kernel routing changes and converts them into dynamic tunnel (BGP EVPN) address advertisement routes, which are then sent to the hardware switch. After receiving the route, the hardware switch updates its local routes.

[0033] When the bare metal server goes online or offline, the hardware switch issues a dynamic tunnel address advertisement route. When the running dynamic routing software (FRRouting) receives the address advertisement route event, the callback function actively notifies the conversion proxy module. The conversion proxy module converts the address advertisement route into routing information that the virtual switch control module can recognize and writes it into the virtual switch control module, updates the virtual switch control module and the logical flow table (OVN flow table), and completes the conversion of the cloud host network to the bare metal network route.

[0034] On the other hand, a network interconnection device is provided, the device comprising:

[0035] The switch connection management module is used to configure virtual switches to connect to the virtual switch control module and hardware switches to connect to the physical switch control module, and to connect the virtual switch control module and the physical switch control module through the switch control module; the virtual switch is used to connect to the cloud host, and the hardware switch is used to connect to the bare metal server.

[0036] The neighbor relationship management module is used to set up kernel routing nodes, which establish dynamic tunnel neighbor relationships between compute nodes and hardware switches.

[0037] The routing event monitoring module is used to connect the computing node to the cloud host to form a cloud host network, and to connect the hardware switch to the bare metal server to form a bare metal network; the switch control module monitors the routing events of the cloud host network through the virtual switch control module, and the switch control module monitors the routing events of the bare metal network through the physical switch control module.

[0038] The interconnection module is used to connect the cloud host network and the bare metal network through the kernel routing node. When a cloud host goes online / offline, the module notifies the hardware switch through the kernel routing node. When a bare metal server goes online / offline, the module notifies the virtual switch control module through the kernel routing node.

[0039] In another aspect, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to perform the following steps:

[0040] Configure a virtual switch to connect to a virtual switch control module, and a hardware switch to connect to a physical switch control module. Connect the virtual switch control module and the physical switch control module through the switch control module. The virtual switch is used to connect to a cloud host, and the hardware switch is used to connect to a bare metal server.

[0041] Configure a kernel routing node, which establishes a dynamic tunnel neighbor relationship between the compute node and the hardware switch;

[0042] The computing nodes are connected to cloud hosts to form a cloud host network, and the hardware switches are connected to bare metal servers to form a bare metal network. The switch control module listens for routing events of the cloud host network through the virtual switch control module, and the switch control module listens for routing events of the bare metal network through the physical switch control module.

[0043] The cloud host network and the bare metal network are connected through the kernel routing node. When a cloud host goes online / offline, the hardware switch is notified through the kernel routing node. When a bare metal server goes online / offline, the virtual switch control module is notified through the kernel routing node.

[0044] In another aspect, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, performs the following steps:

[0045] Configure a virtual switch to connect to a virtual switch control module, and a hardware switch to connect to a physical switch control module. Connect the virtual switch control module and the physical switch control module through the switch control module. The virtual switch is used to connect to a cloud host, and the hardware switch is used to connect to a bare metal server.

[0046] Configure a kernel routing node, which establishes a dynamic tunnel neighbor relationship between the compute node and the hardware switch;

[0047] The computing nodes are connected to cloud hosts to form a cloud host network, and the hardware switches are connected to bare metal servers to form a bare metal network. The switch control module listens for routing events of the cloud host network through the virtual switch control module, and the switch control module listens for routing events of the bare metal network through the physical switch control module.

[0048] The cloud host network and the bare metal network are connected through the kernel routing node. When a cloud host goes online / offline, the hardware switch is notified through the kernel routing node. When a bare metal server goes online / offline, the virtual switch control module is notified through the kernel routing node.

[0049] The aforementioned network interconnection method, apparatus, computer equipment, and storage medium achieve the connection between virtual switches and hardware switches by setting up a switch control module, establish a kernel routing node to form a dynamic routing neighbor relationship between computing nodes and the hardware switches, and realize the conversion between bare metal networks and cloud host networks, thereby achieving Layer 2 interconnection between cloud hosts and bare metal servers, providing a high-performance, cost-effective, and highly scalable Layer 2 interconnection network. Attached Figure Description

[0050] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0051] Figure 1 This is a diagram illustrating the application environment of a network interconnection method in one embodiment of this application;

[0052] Figure 2 This is a flowchart illustrating a network interconnection method in one embodiment of this application;

[0053] Figure 3This is a schematic diagram of a neighbor relationship connection in one embodiment of the present application, in which a kernel routing node is set up to establish a dynamic tunnel neighbor relationship between the compute node and the hardware switch.

[0054] Figure 4 This is a schematic diagram of a network interconnection structure implemented through a kernel routing node in one embodiment of this application;

[0055] Figure 5 This is a schematic diagram illustrating the process by which the conversion proxy module notifies the hardware switch when a cloud host goes online / offline in one embodiment of this application.

[0056] Figure 6 This is a schematic diagram illustrating the process by which the conversion proxy module notifies the hardware switch when a cloud host goes online / offline in one embodiment of this application.

[0057] Figure 7 This is a structural block diagram of a network interconnection device in one embodiment of this application;

[0058] Figure 8 This is an internal structural diagram of a computer device in one embodiment of this application. Detailed Implementation

[0059] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0060] As described in the background section, to achieve a unified VPC network for bare metal and cloud servers, the industry has evolved various technical approaches:

[0061]

[0062]

[0063] As shown in the table above, to achieve a unified VPC network for bare metal cloud hosts, the mainstream smart NIC solutions in the industry have cost-effectiveness issues, software gateways have performance bottlenecks, and overlay networks have hardware device dependencies.

[0064] To address the aforementioned issues, and considering that general-purpose network interface cards (NICs) offer better cost-effectiveness than smart NICs, and that hardware switches have greater network processing capabilities than physical servers while offering higher cost-effectiveness in this scenario, and that Layer 2 networks are the foundation of Layer 3 networks, this invention creatively proposes a network interconnection method. The cloud host network is implemented through host-side software (host overlay), and the bare metal network is implemented through a network overlay. By installing dynamic routing software on the compute node server and establishing dynamic routing neighbors with network devices, and by using agent software to convert between the host overlay and the network overlay, the interconnection between the cloud host and the bare metal Layer 2 network is achieved. This provides a high-performance, cost-effective, and highly scalable method for bare metal cloud host Layer 2 network interconnection.

[0065] The network interconnection method provided in this application can be applied to, for example... Figure 1 The application environment shown is as follows. The cloud host network is implemented based on a host overlay and managed through a virtual switch control module. The bare metal network is implemented based on a network overlay and managed through a hardware switch control module. Interface automation with the cloud platform control panel allows for automated control of both virtual and physical switches. The hardware switch control module is based on NetMiko and can control network devices from multiple vendors, avoiding vendor lock-in.

[0066] The cloud platform adopts the OpenStack Neutron architecture. The virtual switch control module is implemented using OVN and interfaces with Neutron through networking-ovn. Neutron is OpenStack's virtual network service. The physical switch control module uses the OpenStack genericswitch module and implements hardware switch management based on netmiko. Configuration reference: Mechanism_driver = OVN, genericswitch.

[0067] OVN is an SDN controller developed by the OpenvSwitch project team for OpenvSwitch. Compared with other SDN products, OVN has better compatibility and performance with OpenvSwitch and OpenStack. OVN is essentially the Kubernetes equivalent of OVS, providing a centralized OVS controller.

[0068] In one embodiment, such as Figure 2 As shown, a network interconnection method is provided, including the following steps:

[0069] Step S1: Configure the virtual switch to connect to the virtual switch control module and the hardware switch to connect to the physical switch control module. Connect the virtual switch control module and the physical switch control module through the switch control module. The virtual switch is used to connect to the cloud host, and the hardware switch is used to connect to the bare metal server.

[0070] Step S2: Configure a kernel routing node, which establishes a dynamic tunnel neighbor relationship between the compute node and the hardware switch;

[0071] Step S3: Connect the computing node to the cloud host to form a cloud host network, and connect the hardware switch to the bare metal server to form a bare metal network; the switch control module listens for routing events of the cloud host network through the virtual switch control module, and the switch control module listens for routing events of the bare metal network through the physical switch control module.

[0072] Step S4: The cloud host network and the bare metal network are connected through the kernel routing node. When a cloud host goes online / offline, the hardware switch is notified through the kernel routing node. When a bare metal server goes online / offline, the virtual switch control module is notified through the kernel routing node.

[0073] In this embodiment, the network interconnection method further includes:

[0074] The physical switch control module is expanded to support VxLAN networks. When a user creates a VxLAN network, it automatically receives user requests, parses user request messages, and sends hardware switch configuration information to configure the hardware switch.

[0075] The genericswitch provided by the OpenStack community can distribute configurations to hardware switches, but it can only distribute VLANs and does not support VXLANs. Therefore, it is necessary to extend it to support VXLANs so that the physical switch control module can be implemented using genericswitch.

[0076] Extend the OpenStack genericswitch to implement VXLAN networking, enabling automatic distribution of VXLAN configurations to the switch. This not only completes the community's capabilities but also supports hardware switches from multiple vendors, avoiding vendor lock-in and improving method compatibility.

[0077] In this embodiment, the step of distributing hardware switch configuration information to configure the hardware switch includes:

[0078] Create a bridge domain and configure the route distinguisher (RD) and route target (RT);

[0079] Configure the logical interface that implements network virtualization functionality and enable the copying of the Network Identifier (VNI) header;

[0080] Configure the service access point.

[0081] The logical interface for implementing network virtualization is preferably an NVE interface. NVE stands for Network Virtualization Edge (interface, used to connect to the network side and forward VxLAN encapsulation). The NVE interface is used to specify the source and destination addresses of the VxLAN tunnel and to encapsulate data.

[0082] VNI stands for VxLAN Network Identifier, which consists of 24 bits, occupies 24 bits of the VxLAN header, and can be up to 16M of VxLAN segment.

[0083] This involves creating a bridge domain and configuring the route distinguisher (RD) and route target (RT). See the following configuration example:

[0084] bridge-domain 10 (issued by the business and can be the same as VNI)

[0085] vxlanvni10 (issued by business unit)

[0086] evpn

[0087] route-distinguisher 10:2 (generated according to rules)

[0088] vpn-target 100:10export-extcommunity

[0089] vpn-target 100:5010export-extcommunity

[0090] vpn-target 100:10import-extcommunity.

[0091] The configuration includes the logical interface for implementing network virtualization, enabling the replication of the Network Identifier (VNI) header. See the following configuration example:

[0092] interface Nve1

[0093] source 2.2.2.2

[0094] vni 10head-end peer-list protocol bgp.

[0095] Configure the service access point, refer to the following configuration:

[0096] interface GE1 / 0 / 2.10mode I2

[0097] encapsulation dotlq vid 10

[0098] bridge-domain 10.

[0099] like Figure 3 As shown, in this embodiment, setting up a kernel routing node, whereby the kernel routing node establishes a dynamic tunnel neighbor relationship between the compute node and the hardware switch, includes:

[0100] Two servers were selected to run dynamic routing software (FRRouting) as kernel routing nodes;

[0101] The compute node runs dynamic routing software (FRRouting), runs dynamic tunneling (BGP EVPN), and establishes neighbor relationships with the kernel routing node;

[0102] The hardware switch establishes a dynamic tunnel (BGP EVPN) neighbor relationship with the kernel routing node.

[0103] The kernel routing node is abbreviated as RR node, and the reference configuration for the kernel routing node is as follows:

[0104] router bgp BGPID

[0105] bgp router-id RR node IP

[0106] bgp bestpath as-path multipath-relax

[0107] no bgp ebgp-requires-policy

[0108] neighbor fabric peer-group

[0109] neighbor fabric remote-as external

[0110] neighborit compute node 1_JP peer-group fabric

[0111] Neighbor hardware switch_IP peer-group fabric.

[0112] The reference configuration for the compute nodes is as follows:

[0113] router BGP number

[0114] bgp router-id Local node IP address

[0115] no bgp ebgp-requires-policy

[0116] bgp bestpath as-path multipath-relax

[0117] neighbor fabric peer-group

[0118] neighbor fabric remote-as external

[0119] The neighbor RR node IP address is the peer-group fabric.

[0120] The reference configuration for the hardware switch connected to the bare metal server is as follows:

[0121] evpn-overlay enable

[0122] bgp BGPID instance evpn1

[0123] peer RR node IP as-number BGP number

[0124] peer RR node IP connect-interface LoopBack()

[0125] I2vpn-family evpn

[0126] peer RR node IP enable

[0127] commit.

[0128] In this embodiment, the switch control module monitors routing events of the cloud host network through the virtual switch control module, and the switch control module monitors routing events of the bare metal network through the physical switch control module, including:

[0129] Configure the switch control module to support interface with the virtual switch control module and the dynamic routing software (FRRouting);

[0130] When a bare metal server comes online or goes offline, the switch control module listens for dynamic tunnel (BGP EVPN) address advertisement routing events, generates a logical flow table (OVN flow table) corresponding to the virtual switch control module, and writes it into the database of the virtual switch control module.

[0131] When a cloud host comes online, the switch control module listens for the online event of the computing node, generates a corresponding dynamic tunnel (BGP EVPN) address advertisement route, and writes the dynamic tunnel (BGP EVPN) address advertisement route into the running dynamic routing software (FRRouting) to complete the interconnection between the cloud host network and the bare metal network.

[0132] The address advertisement route mentioned above is a MAC / IP address advertisement route, with Type 2 routes being preferred. Type 2 routes, also known as MAC advertisement routes, are used to transmit the host's MAC address or MAC & IP routing information. Devices can use this route to establish relevant forwarding tables.

[0133] In this embodiment, the cloud host network and the bare metal network are connected through the kernel routing node. When a cloud host goes online / offline, the hardware switch is notified through the kernel routing node. When a bare metal server goes online / offline, the virtual switch control module is notified through the kernel routing node, including:

[0134] Configure a translation proxy module in the kernel routing node;

[0135] When a cloud host goes online / offline, the conversion proxy module notifies the hardware switch; when a bare metal server goes online / offline, the conversion proxy module notifies the virtual switch control module.

[0136] In this embodiment, the step of notifying the hardware switch through the conversion proxy module when a cloud host goes online / offline, and notifying the virtual switch control module through the conversion proxy module when a bare metal server goes online / offline, includes:

[0137] The conversion proxy module is connected to the virtual switch control module and the running dynamic routing software (FRRouting). The switch control module is used to capture cloud host online and offline action events and update the local flow table.

[0138] The virtual switch control module is configured to listen for events, specifically to monitor the online and offline actions of cloud hosts on the computing node ports.

[0139] The dynamic routing software (FRRouting) is configured to use a callback function that listens for changes in the address advertisement route of the dynamic tunnel (BGP EVPN) for event callbacks. This callback function is used to actively notify the conversion agent module when an address advertisement route event is received.

[0140] When the conversion agent module captures the online / offline operation of the cloud host, it obtains the address table (MAC) and the tunnel endpoint (VTEP) information of the corresponding port of the cloud host and writes it into the system kernel. The running dynamic routing software (FRRouting) listens for kernel routing changes and converts them into dynamic tunnel (BGP EVPN) address advertisement routes, which are then sent to the hardware switch. After receiving the route, the hardware switch updates its local routes.

[0141] When the bare metal server goes online or offline, the hardware switch issues a dynamic tunnel address advertisement route. When the running dynamic routing software (FRRouting) receives the address advertisement route event, the callback function actively notifies the conversion proxy module. The conversion proxy module converts the address advertisement route into routing information that the virtual switch control module can recognize and writes it into the virtual switch control module, updates the virtual switch control module and the logical flow table (OVN flow table), and completes the conversion of the cloud host network to the bare metal network route.

[0142] Specifically, the running dynamic routing software (FRRouting) uses the zebra component to monitor kernel route changes and converts them into dynamic tunnel (BGP EVPN) address advertisement routes; the running dynamic routing software (FRRouting) uses the zebra component to provide an event callback mechanism, and registers a function in the zebra component to monitor dynamic tunnel (BGP EVPN) address advertisement route changes.

[0143] The network interconnection method further includes:

[0144] The zebra component is used to monitor kernel route changes in the running dynamic routing software (FRRouting), and a function to monitor dynamic tunnel (BGP EVPN) address advertisement route change is registered in the zebra component;

[0145] When the conversion agent module receives a cloud host online / offline action event, the conversion agent module controls the Zebra component to send a dynamic tunnel (BGP EVPN) address announcement route;

[0146] In response to an online / offline event on a bare metal server, the Zebra component sends an Ethernet Virtual Private Network (EVPN) address advertisement route. When the conversion agent module receives the EVPN address advertisement route, it actively notifies the conversion agent module. The conversion agent module converts the EVPN address advertisement route into routing information and notifies the switch control module.

[0147] The conversion agent module uses agent software to convert between host overlay and network overlay, thereby enabling Layer 2 interconnection between cloud hosts and bare metal servers. Network communication between cloud hosts is carried by the host overlay network (virtual switch flow table), communication between bare metal servers is carried by the network overlay (BGP EVPN), and communication between cloud hosts and bare metal servers is achieved by a hybrid overlay.

[0148] like Figure 5 As shown, in this embodiment, the step of notifying the hardware switch through the conversion proxy module when a cloud host goes online / offline includes:

[0149] OVN provides an event listening mechanism that can monitor port online and offline actions. When the Agent captures the online and offline operations of a cloud host, it obtains information such as the MAC address and the VTEP to which the port belongs, and writes it into the Linux kernel. The zebra in FRRouting listens for kernel route changes and converts them into BGP EVPN Type 2 routes, which are then sent to the hardware switch. After receiving the data, the hardware switch updates its local routes, completing the process.

[0150] like Figure 6 As shown, in this embodiment, the step of notifying the hardware switch through the conversion proxy module when a cloud host goes online / offline includes:

[0151] The Zebra component used in FRRouting provides an event callback mechanism. This method registers a function in Zebra to listen for BGP EVPN Type 2 route changes. When bare metal switches go online or offline, they will send out BGP EVPN Type 2 routes. When FRRouting receives the address advertisement route event, the callback function actively notifies the agent. The agent then converts the address advertisement route into OVN-recognizable route information and writes it, completing the conversion.

[0152] This application generates flow tables based on online / offline events, monitors online / offline events based on address announcements and routing changes, and combines a conversion proxy module to implement change notifications for cloud host networks and bare metal networks.

[0153] In the above network interconnection method, the virtual switch and the hardware switch are connected by setting up a switch control module, and the kernel routing node is set up to establish a neighbor relationship between the computing node and the hardware switch to form a dynamic routing neighbor. The kernel routing node realizes the conversion between the bare metal network and the cloud host network, thereby realizing the Layer 2 interconnection between the cloud host and the bare metal server, providing a Layer 2 interconnection network with high network performance, high cost performance and strong scalability.

[0154] High cost-performance ratio: Compared with smart network cards and software gateway solutions, this solution uses a switch as the gateway, which has a price advantage over servers and smart network cards.

[0155] Network performance: This solution uses a switch as a gateway, which has no bottleneck in network performance compared to a software gateway, and offers high performance.

[0156] High scalability: The cloud host network still uses host overlay, which is highly scalable; the hardware switch control module can also be decoupled from the vendor switch, avoiding vendor lock-in.

[0157] In one embodiment, such as Figure 7 As shown, a network interconnection device 10 is provided, including: a switch connection management module 1, a neighbor relationship management module 2, a routing event monitoring module 3, and an interconnection interaction module 4.

[0158] The switch connection management module 1 is used to set up a virtual switch to connect to a virtual switch control module and a hardware switch to connect to a physical switch control module, and to connect the virtual switch control module and the physical switch control module through the switch control module; the virtual switch is used to connect to a cloud host and the hardware switch is used to connect to a bare metal server.

[0159] The neighbor relationship management module 2 is used to set up kernel routing nodes, which establish dynamic tunnel neighbor relationships between compute nodes and hardware switches.

[0160] The routing event monitoring module 3 is used to connect the computing node to the cloud host to form a cloud host network, and to connect the hardware switch to the bare metal server to form a bare metal network; the switch control module monitors the routing events of the cloud host network through the virtual switch control module, and the switch control module monitors the routing events of the bare metal network through the physical switch control module.

[0161] The interconnection module 4 is used to connect the cloud host network and the bare metal network through the kernel routing node. When a cloud host goes online / offline, the kernel routing node is used to notify the hardware switch. When a bare metal server goes online / offline, the kernel routing node is used to notify the virtual switch control module.

[0162] In this embodiment, the network interconnection method further includes:

[0163] The physical switch control module is expanded to support VxLAN networks. When a user creates a VxLAN network, it automatically receives user requests, parses user request messages, and sends hardware switch configuration information to configure the hardware switch.

[0164] In this embodiment, the step of distributing hardware switch configuration information to configure the hardware switch includes:

[0165] Create a bridge domain and configure the route distinguisher (RD) and route target (RT);

[0166] Configure the logical interface that implements network virtualization functionality and enable the copying of the Network Identifier (VNI) header;

[0167] Configure the service access point.

[0168] In this embodiment, setting up a kernel routing node, wherein the kernel routing node establishes a dynamic tunnel neighbor relationship between the compute node and the hardware switch, includes:

[0169] Two servers were selected to run dynamic routing software (FRRouting) as kernel routing nodes;

[0170] The compute node runs dynamic routing software (FRRouting), runs dynamic tunneling (BGP EVPN), and establishes neighbor relationships with the kernel routing node;

[0171] The hardware switch establishes a dynamic tunnel (BGP EVPN) neighbor relationship with the kernel routing node.

[0172] In this embodiment, the switch control module monitors routing events of the cloud host network through the virtual switch control module, and the switch control module monitors routing events of the bare metal network through the physical switch control module, including:

[0173] Configure the switch control module to support interface with the virtual switch control module and the dynamic routing software (FRRouting);

[0174] When a bare metal server comes online or goes offline, the switch control module listens for dynamic tunnel (BGP EVPN) address advertisement routing events, generates a logical flow table (OVN flow table) corresponding to the virtual switch control module, and writes it into the database of the virtual switch control module.

[0175] When a cloud host comes online, the switch control module listens for the online event of the computing node, generates a corresponding dynamic tunnel (BGP EVPN) address advertisement route, and writes the dynamic tunnel (BGP EVPN) address advertisement route into the running dynamic routing software (FRRouting) to complete the interconnection between the cloud host network and the bare metal network.

[0176] In this embodiment, the cloud host network and the bare metal network are connected through the kernel routing node. When a cloud host goes online / offline, the hardware switch is notified through the kernel routing node. When a bare metal server goes online / offline, the virtual switch control module is notified through the kernel routing node, including:

[0177] Configure a translation proxy module in the kernel routing node;

[0178] When a cloud host goes online / offline, the conversion proxy module notifies the hardware switch; when a bare metal server goes online / offline, the conversion proxy module notifies the virtual switch control module.

[0179] In this embodiment, the step of notifying the hardware switch through the conversion proxy module when a cloud host goes online / offline, and notifying the virtual switch control module through the conversion proxy module when a bare metal server goes online / offline, includes:

[0180] The conversion proxy module is connected to the virtual switch control module and the running dynamic routing software (FRRouting). The switch control module is used to capture cloud host online and offline action events and update the local flow table.

[0181] The virtual switch control module is configured to listen for events, specifically to monitor the online and offline actions of cloud hosts on the computing node ports.

[0182] The dynamic routing software (FRRouting) is configured to use a callback function that listens for changes in the address advertisement route of the dynamic tunnel (BGP EVPN) for event callbacks. This callback function is used to actively notify the conversion agent module when an address advertisement route event is received.

[0183] When the conversion agent module captures the online / offline operation of the cloud host, it obtains the address table (MAC) and the tunnel endpoint (VTEP) information of the corresponding port of the cloud host and writes it into the system kernel. The running dynamic routing software (FRRouting) listens for kernel routing changes and converts them into dynamic tunnel (BGP EVPN) address advertisement routes, which are then sent to the hardware switch. After receiving the route, the hardware switch updates its local routes.

[0184] When the bare metal server goes online or offline, the hardware switch issues a dynamic tunnel address advertisement route. When the running dynamic routing software (FRRouting) receives the address advertisement route event, the callback function actively notifies the conversion proxy module. The conversion proxy module converts the address advertisement route into routing information that the virtual switch control module can recognize and writes it into the virtual switch control module, updates the virtual switch control module and the logical flow table (OVN flow table), and completes the conversion of the cloud host network to the bare metal network route.

[0185] In the aforementioned network interconnection device, a switch control module is set up to connect the virtual switch and the hardware switch. A kernel routing node is set up to establish a neighbor relationship between the computing node and the hardware switch to form a dynamic routing neighbor. The kernel routing node realizes the conversion between the bare metal network and the cloud host network, thereby realizing Layer 2 interconnection between the cloud host and the bare metal server, providing a Layer 2 interconnection network with high network performance, high cost performance and strong scalability.

[0186] Specific limitations regarding the network interconnection device can be found in the limitations of the network interconnection method described above, and will not be repeated here. Each module in the aforementioned network interconnection device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in the computer device in hardware form, or stored in the memory of the computer device in software form, so that the processor can call and execute the operations corresponding to each module.

[0187] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, performs the following steps:

[0188] Configure a virtual switch to connect to a virtual switch control module, and a hardware switch to connect to a physical switch control module. Connect the virtual switch control module and the physical switch control module through the switch control module. The virtual switch is used to connect to a cloud host, and the hardware switch is used to connect to a bare metal server.

[0189] Configure a kernel routing node, which establishes a dynamic tunnel neighbor relationship between the compute node and the hardware switch;

[0190] The computing nodes are connected to cloud hosts to form a cloud host network, and the hardware switches are connected to bare metal servers to form a bare metal network. The switch control module listens for routing events of the cloud host network through the virtual switch control module, and the switch control module listens for routing events of the bare metal network through the physical switch control module.

[0191] The cloud host network and the bare metal network are connected through the kernel routing node. When a cloud host goes online / offline, the hardware switch is notified through the kernel routing node. When a bare metal server goes online / offline, the virtual switch control module is notified through the kernel routing node.

[0192] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0193] The network interconnection method also includes:

[0194] The physical switch control module is expanded to support VxLAN networks. When a user creates a VxLAN network, it automatically receives user requests, parses user request messages, and sends hardware switch configuration information to configure the hardware switch.

[0195] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0196] The process of distributing hardware switch configuration information to configure the hardware switch includes:

[0197] Create a bridge domain and configure the route distinguisher (RD) and route target (RT);

[0198] Configure the logical interface that implements network virtualization functionality and enable the copying of the Network Identifier (VNI) header;

[0199] Configure the service access point.

[0200] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0201] The step of setting up a kernel routing node, wherein the kernel routing node establishes a dynamic tunnel neighbor relationship between the compute node and the hardware switch, includes:

[0202] Two servers were selected to run dynamic routing software (FRRouting) as kernel routing nodes;

[0203] The compute node runs dynamic routing software (FRRouting), runs dynamic tunneling (BGP EVPN), and establishes neighbor relationships with the kernel routing node;

[0204] The hardware switch establishes a dynamic tunnel (BGP EVPN) neighbor relationship with the kernel routing node.

[0205] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0206] The switch control module listens to routing events of the cloud host network through the virtual switch control module, and the switch control module listens to routing events of the bare metal network through the physical switch control module, including:

[0207] Configure the switch control module to support interface with the virtual switch control module and the dynamic routing software (FRRouting);

[0208] When a bare metal server comes online or goes offline, the switch control module listens for dynamic tunnel (BGP EVPN) address advertisement routing events, generates a logical flow table (OVN flow table) corresponding to the virtual switch control module, and writes it into the database of the virtual switch control module.

[0209] When a cloud host comes online, the switch control module listens for the online event of the computing node, generates a corresponding dynamic tunnel (BGP EVPN) address advertisement route, and writes the dynamic tunnel (BGP EVPN) address advertisement route into the running dynamic routing software (FRRouting) to complete the interconnection between the cloud host network and the bare metal network.

[0210] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0211] The cloud host network and the bare metal network are connected through the kernel routing node. When a cloud host goes online / offline, the hardware switch is notified through the kernel routing node. When a bare metal server goes online / offline, the virtual switch control module is notified through the kernel routing node.

[0212] Configure a translation proxy module in the kernel routing node;

[0213] When a cloud host goes online / offline, the conversion proxy module notifies the hardware switch; when a bare metal server goes online / offline, the conversion proxy module notifies the virtual switch control module.

[0214] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0215] The step of notifying the hardware switch through the conversion proxy module when a cloud host goes online / offline, and notifying the virtual switch control module through the conversion proxy module when a bare metal server goes online / offline, includes:

[0216] The conversion proxy module is connected to the virtual switch control module and the running dynamic routing software (FRRouting). The switch control module is used to capture cloud host online and offline action events and update the local flow table.

[0217] The virtual switch control module is configured to listen for events, specifically to monitor the online and offline actions of cloud hosts on the computing node ports.

[0218] The dynamic routing software (FRRouting) is configured to use a callback function that listens for changes in the address advertisement route of the dynamic tunnel (BGP EVPN) for event callbacks. This callback function is used to actively notify the conversion agent module when an address advertisement route event is received.

[0219] When the conversion agent module captures the online / offline operation of the cloud host, it obtains the address table (MAC) and the tunnel endpoint (VTEP) information of the corresponding port of the cloud host and writes it into the system kernel. The running dynamic routing software (FRRouting) listens for kernel routing changes and converts them into dynamic tunnel (BGP EVPN) address advertisement routes, which are then sent to the hardware switch. After receiving the route, the hardware switch updates its local routes.

[0220] When the bare metal server goes online or offline, the hardware switch issues a dynamic tunnel address advertisement route. When the running dynamic routing software (FRRouting) receives the address advertisement route event, the callback function actively notifies the conversion proxy module. The conversion proxy module converts the address advertisement route into routing information that the virtual switch control module can recognize and writes it into the virtual switch control module, updates the virtual switch control module and the logical flow table (OVN flow table), and completes the conversion of the cloud host network to the bare metal network route.

[0221] For specific limitations on the steps implemented when a computer program is executed by a processor, please refer to the limitations on network communication methods mentioned above, which will not be repeated here.

[0222] In one embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 8As shown, the computer device includes a processor, memory, network interface, and database connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and database. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The database stores network communication data. The network interface communicates with external terminals via a network connection. When the computer program is executed by the processor, it implements a network communication method.

[0223] Those skilled in the art will understand that Figure 8 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0224] In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to perform the following steps:

[0225] Configure a virtual switch to connect to a virtual switch control module, and a hardware switch to connect to a physical switch control module. Connect the virtual switch control module and the physical switch control module through the switch control module. The virtual switch is used to connect to a cloud host, and the hardware switch is used to connect to a bare metal server.

[0226] Configure a kernel routing node, which establishes a dynamic tunnel neighbor relationship between the compute node and the hardware switch;

[0227] The computing nodes are connected to cloud hosts to form a cloud host network, and the hardware switches are connected to bare metal servers to form a bare metal network. The switch control module listens for routing events of the cloud host network through the virtual switch control module, and the switch control module listens for routing events of the bare metal network through the physical switch control module.

[0228] The cloud host network and the bare metal network are connected through the kernel routing node. When a cloud host goes online / offline, the hardware switch is notified through the kernel routing node. When a bare metal server goes online / offline, the virtual switch control module is notified through the kernel routing node.

[0229] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0230] The network interconnection method also includes:

[0231] The physical switch control module is expanded to support VxLAN networks. When a user creates a VxLAN network, it automatically receives user requests, parses user request messages, and sends hardware switch configuration information to configure the hardware switch.

[0232] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0233] The process of distributing hardware switch configuration information to configure the hardware switch includes:

[0234] Create a bridge domain and configure the route distinguisher (RD) and route target (RT);

[0235] Configure the logical interface that implements network virtualization functionality and enable the copying of the Network Identifier (VNI) header;

[0236] Configure the service access point.

[0237] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0238] The step of setting up a kernel routing node, wherein the kernel routing node establishes a dynamic tunnel neighbor relationship between the compute node and the hardware switch, includes:

[0239] Two servers were selected to run dynamic routing software (FRRouting) as kernel routing nodes;

[0240] The compute node runs dynamic routing software (FRRouting), runs dynamic tunneling (BGP EVPN), and establishes neighbor relationships with the kernel routing node;

[0241] The hardware switch establishes a dynamic tunnel (BGP EVPN) neighbor relationship with the kernel routing node.

[0242] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0243] The switch control module listens to routing events of the cloud host network through the virtual switch control module, and the switch control module listens to routing events of the bare metal network through the physical switch control module, including:

[0244] Configure the switch control module to support interface with the virtual switch control module and the dynamic routing software (FRRouting);

[0245] When a bare metal server comes online or goes offline, the switch control module listens for dynamic tunnel (BGP EVPN) address advertisement routing events, generates a logical flow table (OVN flow table) corresponding to the virtual switch control module, and writes it into the database of the virtual switch control module.

[0246] When a cloud host comes online, the switch control module listens for the online event of the computing node, generates a corresponding dynamic tunnel (BGP EVPN) address advertisement route, and writes the dynamic tunnel (BGP EVPN) address advertisement route into the running dynamic routing software (FRRouting) to complete the interconnection between the cloud host network and the bare metal network.

[0247] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0248] The cloud host network and the bare metal network are connected through the kernel routing node. When a cloud host goes online / offline, the hardware switch is notified through the kernel routing node. When a bare metal server goes online / offline, the virtual switch control module is notified through the kernel routing node.

[0249] Configure a translation proxy module in the kernel routing node;

[0250] When a cloud host goes online / offline, the conversion proxy module notifies the hardware switch; when a bare metal server goes online / offline, the conversion proxy module notifies the virtual switch control module.

[0251] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0252] The step of notifying the hardware switch through the conversion proxy module when a cloud host goes online / offline, and notifying the virtual switch control module through the conversion proxy module when a bare metal server goes online / offline, includes:

[0253] The conversion proxy module is connected to the virtual switch control module and the running dynamic routing software (FRRouting). The switch control module is used to capture cloud host online and offline action events and update the local flow table.

[0254] The virtual switch control module is configured to listen for events, specifically to monitor the online and offline actions of cloud hosts on the computing node ports.

[0255] The dynamic routing software (FRRouting) is configured to use a callback function that listens for changes in the address advertisement route of the dynamic tunnel (BGP EVPN) for event callbacks. This callback function is used to actively notify the conversion agent module when an address advertisement route event is received.

[0256] When the conversion agent module captures the online / offline operation of the cloud host, it obtains the address table (MAC) and the tunnel endpoint (VTEP) information of the corresponding port of the cloud host and writes it into the system kernel. The running dynamic routing software (FRRouting) listens for kernel routing changes and converts them into dynamic tunnel (BGP EVPN) address advertisement routes, which are then sent to the hardware switch. After receiving the route, the hardware switch updates its local routes.

[0257] When the bare metal server goes online or offline, the hardware switch issues a dynamic tunnel address advertisement route. When the running dynamic routing software (FRRouting) receives the address advertisement route event, the callback function actively notifies the conversion proxy module. The conversion proxy module converts the address advertisement route into routing information that the virtual switch control module can recognize and writes it into the virtual switch control module, updates the virtual switch control module and the logical flow table (OVN flow table), and completes the conversion of the cloud host network to the bare metal network route.

[0258] For specific limitations on the steps a processor takes when executing a computer program, please refer to the limitations on network communication methods mentioned above, which will not be repeated here.

[0259] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, the computer program performing the following steps when executed by a processor:

[0260] Configure a virtual switch to connect to a virtual switch control module, and a hardware switch to connect to a physical switch control module. Connect the virtual switch control module and the physical switch control module through the switch control module. The virtual switch is used to connect to a cloud host, and the hardware switch is used to connect to a bare metal server.

[0261] Configure a kernel routing node, which establishes a dynamic tunnel neighbor relationship between the compute node and the hardware switch;

[0262] The computing nodes are connected to cloud hosts to form a cloud host network, and the hardware switches are connected to bare metal servers to form a bare metal network. The switch control module listens for routing events of the cloud host network through the virtual switch control module, and the switch control module listens for routing events of the bare metal network through the physical switch control module.

[0263] The cloud host network and the bare metal network are connected through the kernel routing node. When a cloud host goes online / offline, the hardware switch is notified through the kernel routing node. When a bare metal server goes online / offline, the virtual switch control module is notified through the kernel routing node.

[0264] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0265] The network interconnection method also includes:

[0266] The physical switch control module is expanded to support VxLAN networks. When a user creates a VxLAN network, it automatically receives user requests, parses user request messages, and sends hardware switch configuration information to configure the hardware switch.

[0267] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0268] The process of distributing hardware switch configuration information to configure the hardware switch includes:

[0269] Create a bridge domain and configure the route distinguisher (RD) and route target (RT);

[0270] Configure the logical interface that implements network virtualization functionality and enable the copying of the Network Identifier (VNI) header;

[0271] Configure the service access point.

[0272] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0273] The step of setting up a kernel routing node, wherein the kernel routing node establishes a dynamic tunnel neighbor relationship between the compute node and the hardware switch, includes:

[0274] Two servers were selected to run dynamic routing software (FRRouting) as kernel routing nodes;

[0275] The compute node runs dynamic routing software (FRRouting), runs dynamic tunneling (BGP EVPN), and establishes neighbor relationships with the kernel routing node;

[0276] The hardware switch establishes a dynamic tunnel (BGP EVPN) neighbor relationship with the kernel routing node.

[0277] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0278] The switch control module listens to routing events of the cloud host network through the virtual switch control module, and the switch control module listens to routing events of the bare metal network through the physical switch control module, including:

[0279] Configure the switch control module to support interface with the virtual switch control module and the dynamic routing software (FRRouting);

[0280] When a bare metal server comes online or goes offline, the switch control module listens for dynamic tunnel (BGP EVPN) address advertisement routing events, generates a logical flow table (OVN flow table) corresponding to the virtual switch control module, and writes it into the database of the virtual switch control module.

[0281] When a cloud host comes online, the switch control module listens for the online event of the computing node, generates a corresponding dynamic tunnel (BGP EVPN) address advertisement route, and writes the dynamic tunnel (BGP EVPN) address advertisement route into the running dynamic routing software (FRRouting) to complete the interconnection between the cloud host network and the bare metal network.

[0282] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0283] The cloud host network and the bare metal network are connected through the kernel routing node. When a cloud host goes online / offline, the hardware switch is notified through the kernel routing node. When a bare metal server goes online / offline, the virtual switch control module is notified through the kernel routing node.

[0284] Configure a translation proxy module in the kernel routing node;

[0285] When a cloud host goes online / offline, the conversion proxy module notifies the hardware switch; when a bare metal server goes online / offline, the conversion proxy module notifies the virtual switch control module.

[0286] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0287] The step of notifying the hardware switch through the conversion proxy module when a cloud host goes online / offline, and notifying the virtual switch control module through the conversion proxy module when a bare metal server goes online / offline, includes:

[0288] The conversion proxy module is connected to the virtual switch control module and the running dynamic routing software (FRRouting). The switch control module is used to capture cloud host online and offline action events and update the local flow table.

[0289] The virtual switch control module is configured to listen for events, specifically to monitor the online and offline actions of cloud hosts on the computing node ports.

[0290] The dynamic routing software (FRRouting) is configured to use a callback function that listens for changes in the address advertisement route of the dynamic tunnel (BGP EVPN) for event callbacks. This callback function is used to actively notify the conversion agent module when an address advertisement route event is received.

[0291] When the conversion agent module captures the online / offline operation of the cloud host, it obtains the address table (MAC) and the tunnel endpoint (VTEP) information of the corresponding port of the cloud host and writes it into the system kernel. The running dynamic routing software (FRRouting) listens for kernel routing changes and converts them into dynamic tunnel (BGP EVPN) address advertisement routes, which are then sent to the hardware switch. After receiving the route, the hardware switch updates its local routes.

[0292] When the bare metal server goes online or offline, the hardware switch issues a dynamic tunnel address advertisement route. When the running dynamic routing software (FRRouting) receives the address advertisement route event, the callback function actively notifies the conversion proxy module. The conversion proxy module converts the address advertisement route into routing information that the virtual switch control module can recognize and writes it into the virtual switch control module, updates the virtual switch control module and the logical flow table (OVN flow table), and completes the conversion of the cloud host network to the bare metal network route.

[0293] For specific limitations on the steps implemented when a computer program is executed by a processor, please refer to the limitations on network communication methods mentioned above, which will not be repeated here.

[0294] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0295] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0296] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this patent application should be determined by the appended claims.

Claims

1. A network interconnection method, characterized in that, include: Configure a virtual switch to connect to a virtual switch control module, and a hardware switch to connect to a physical switch control module. Connect the virtual switch control module and the physical switch control module through the switch control module. The virtual switch is used to connect to a cloud host, and the hardware switch is used to connect to a bare metal server. Configure a kernel routing node, which establishes a dynamic tunnel neighbor relationship between the compute node and the hardware switch; The computing nodes are connected to cloud hosts to form a cloud host network, and the hardware switches are connected to bare metal servers to form a bare metal network. The switch control module listens for routing events of the cloud host network through the virtual switch control module, and the switch control module listens for routing events of the bare metal network through the physical switch control module. The cloud host network and the bare metal network are connected through the kernel routing node. When a cloud host goes online / offline, the hardware switch is notified through the kernel routing node. When a bare metal server goes online / offline, the virtual switch control module is notified through the kernel routing node.

2. The network interconnection method according to claim 1, characterized in that, Also includes: The physical switch control module is expanded to support VxLAN networks. When a user creates a VxLAN network, it automatically receives user requests, parses user request messages, and sends hardware switch configuration information to configure the hardware switch.

3. The network interconnection method according to claim 2, characterized in that, The process of distributing hardware switch configuration information to configure the hardware switch includes: Create a bridge domain and configure the route delimiter and route destination; Configure the logical interface to implement network virtualization functionality and enable network identifier header replication; Configure the service access point.

4. The network interconnection method according to claim 1, characterized in that, The step of setting up a kernel routing node, wherein the kernel routing node establishes a dynamic tunnel neighbor relationship between the compute node and the hardware switch, includes: Select two servers to run dynamic routing software, which will serve as kernel routing nodes; The computing node runs dynamic routing software, runs dynamic tunnels, and establishes neighbor relationships with the kernel routing node; the hardware switch establishes dynamic tunnel neighbor relationships with the kernel routing node.

5. The network interconnection method according to claim 4, characterized in that, The switch control module listens to routing events of the cloud host network through the virtual switch control module, and the switch control module listens to routing events of the bare metal network through the physical switch control module, including: The switch control module is configured to support interface with the virtual switch control module and the dynamic routing software. When a bare metal server goes online or offline, the switch control module listens for dynamic tunnel address announcement routing events, generates a logical flow table corresponding to the virtual switch control module, and writes it into the database of the virtual switch control module. When a cloud host comes online, the switch control module listens for the online event of the computing node, generates a corresponding dynamic tunnel address announcement route, and writes the dynamic tunnel address announcement route into the running dynamic routing software to complete the interconnection between the cloud host network and the bare metal network.

6. The network interconnection method according to claim 5, characterized in that, The cloud host network and the bare metal network are connected through the kernel routing node. When a cloud host goes online / offline, the hardware switch is notified through the kernel routing node. When a bare metal server goes online / offline, the virtual switch control module is notified through the kernel routing node. Configure a translation proxy module in the kernel routing node; When a cloud host goes online / offline, the conversion proxy module notifies the hardware switch; when a bare metal server goes online / offline, the conversion proxy module notifies the virtual switch control module.

7. The network interconnection method according to claim 6, characterized in that, The step of notifying the hardware switch through the conversion proxy module when a cloud host goes online / offline, and notifying the virtual switch control module through the conversion proxy module when a bare metal server goes online / offline, includes: The conversion proxy module is connected to the virtual switch control module and the running dynamic routing software. The switch control module is used to capture cloud host online and offline action events and update the local flow table. The virtual switch control module is configured to listen for events, specifically to monitor the online and offline actions of cloud hosts on the computing node ports. The dynamic routing software is configured to use a callback function that listens for changes in the dynamic tunnel address announcement route to perform event callbacks. This is used to actively notify the conversion agent module when an address announcement route event is received. When the conversion agent module captures the online / offline operation of the cloud host, it obtains the address table of the corresponding port of the cloud host and the tunnel endpoint information, and writes it into the system kernel. The dynamic routing software listens for kernel routing changes and converts them into dynamic tunnel address announcement routes, which are then sent to the hardware switch. The hardware switch receives the announcements and updates its local routes. When the bare metal server goes online or offline, the hardware switch issues a dynamic tunnel address advertisement route. When the dynamic routing software receives the address advertisement route event, the callback function actively notifies the conversion proxy module. The conversion proxy module converts the address advertisement route into routing information that the virtual switch control module can recognize and writes it into the virtual switch control module, updates the virtual switch control module and the logical flow table, and completes the conversion of the cloud host network to the bare metal network route.

8. A network interconnection device, characterized in that, The device includes: The switch connection management module is used to configure virtual switches to connect to the virtual switch control module and hardware switches to connect to the physical switch control module, and to connect the virtual switch control module and the physical switch control module through the switch control module; the virtual switch is used to connect to the cloud host, and the hardware switch is used to connect to the bare metal server. The neighbor relationship management module is used to set up kernel routing nodes, which establish dynamic tunnel neighbor relationships between compute nodes and hardware switches. The routing event monitoring module is used to connect the computing node to the cloud host to form a cloud host network, and to connect the hardware switch to the bare metal server to form a bare metal network; the switch control module monitors the routing events of the cloud host network through the virtual switch control module, and the switch control module monitors the routing events of the bare metal network through the physical switch control module. The interconnection module is used to connect the cloud host network and the bare metal network through the kernel routing node. When a cloud host goes online / offline, the module notifies the hardware switch through the kernel routing node. When a bare metal server goes online / offline, the module notifies the virtual switch control module through the kernel routing node.

9. A computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 7.

Images