Bandwidth data processing system, method and readable storage medium

Abstract

The application provides a bandwidth data processing system, method and readable storage medium, and relates to the technical field of communication. The processing system comprises a routing gateway (RGW), a gateway cluster and a virtual network address translation (VNAT) device cluster; the RGW is used for acquiring target bandwidth data in a cloud-in direction or a cloud-out direction; the target bandwidth data comprises bandwidth traffic, a bandwidth type and a service type; the RGW is further used for performing separation processing on the target bandwidth data through a target classification algorithm, and performing flow limiting processing on the separated data through the gateway cluster or the VNAT device cluster. According to the scheme, flow separation or flow limiting can be performed on different target bandwidth data, the problem of super-large flow bandwidth can be solved, and the bandwidth processing capacity of a single cluster is greatly improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of communication, in particular to a bandwidth data processing system, method and readable storage medium. BACKGROUND

[0002] In the prior art, a cloud tenant provides public network service by independently subscribing to public network Internet Protocol (IP) related services; and corresponding capabilities are provided by underlying network address translation gateway (NAT GW) and quality of service gateway (QoS GW). With the acceleration of digital transformation and computing power network construction, the number of public network IP of the tenant and bandwidth demand grow rapidly.

[0003] The prior art has the following disadvantages:

[0004] 1) The current NAT GW and QoS GW are formed by combining a plurality of minimum forwarding units (the maximum processing capacity of a single forwarding unit is 10G) to form an overall forwarding capacity, and the overall bandwidth capacity of a single device is 200G. The traffic is distributed among the minimum units inside. It depends on the multi-queue receiving (RSS) of the network card and the gateway device internal switching matrix Fabric technology, and uses a Hash algorithm to distribute traffic. However, the Hash algorithm itself cannot distinguish between service traffic and cannot automatically adjust the traffic among the processing units, resulting in the possibility of mixing large bandwidth and small bandwidth on the same minimum processing unit, which may cause the large bandwidth to burst the single forwarding unit, affecting other traffic on the forwarding unit, and causing network jitter, packet loss and other phenomena.

[0005] 2) The tenant purchases public network IP bandwidth or shared bandwidth. In order to ensure accurate speed limiting, a Hash algorithm based on IP is used to ensure that multiple flows of the same IP are distributed to the same minimum processing unit. The current device has a minimum processing capacity of less than or equal to 10G, which cannot meet the needs of single public network IP customers with larger bandwidth. SUMMARY

[0006] The purpose of the present application is to provide a bandwidth data processing system, method and readable storage medium to solve the disadvantages in the prior art. The system separates or limits the flow of different target bandwidth data, can handle super large flow bandwidth problems, and greatly improves the bandwidth processing capacity of a single cluster.

[0007] To achieve the above purpose, an embodiment of the present application provides a bandwidth data processing system, comprising: a routing gateway RGW, a gateway cluster and a virtual network address translation VNAT device cluster.

[0008] The RGW is configured to acquire target bandwidth data in a cloud-in direction or a cloud-out direction, and the target bandwidth data comprises bandwidth traffic, bandwidth type and service type.

[0009] The RGW is further configured to separate the target bandwidth data by a target classification algorithm, and perform traffic limiting processing on the separated data by the gateway cluster or the VNAT device cluster.

[0010] Optionally, the routing gateway RGW comprises:

[0011] a first determination module configured to determine the bandwidth type or the service type of the target bandwidth data according to the target bandwidth data;

[0012] a first separation module configured to separate the target bandwidth data in the cloud-out direction by a first target classification algorithm based on source address information, or separate the target bandwidth data in the cloud-in direction by a second target classification algorithm based on destination address information;

[0013] a first processing module configured to send the separated data to a corresponding processing module for traffic limiting processing according to the bandwidth type or the service type of the separated data.

[0014] Optionally, the routing gateway RGW further comprises:

[0015] a second processing module configured to perform traffic limiting processing by the gateway cluster when the service type is centralized service traffic;

[0016] a third processing module configured to perform proportional network address translation on the target bandwidth data and perform bandwidth limiting processing by the routing gateway RGW when the bandwidth type is first type data;

[0017] a fourth processing module configured to perform limiting processing by a network address translation gateway NAT GW of the gateway cluster when the bandwidth type is second type data;

[0018] a fifth processing module configured to perform limiting processing by a quality of service gateway Qos GW of the gateway cluster of the gateway cluster when the bandwidth type is third type data;

[0019] wherein the bandwidth traffic corresponding to the first type data is greater than the bandwidth traffic corresponding to the second type data, and the bandwidth traffic corresponding to the second type data is greater than the bandwidth traffic corresponding to the third type data.

[0020] Optionally, the processing system is provided with a preset open interface through which a bandwidth initial value of the first type of data, a bandwidth initial value of the second type of data and a bandwidth initial value of the third type of data are configured.

[0021] Optionally, the number of RGWs includes two groups, including:

[0022] The RGW located at the LAN side is used to obtain target bandwidth data in the cloud-out direction; in the target bandwidth data in the cloud-out direction, the source address is a dummy IP address corresponding to a virtual machine; the dummy IP address has a first mapping relationship with a public IP address bound to the virtual machine; and the destination address is an external public IP address.

[0023] The RGW located at the WAN side is used to obtain target bandwidth data in the cloud-in direction; in the target bandwidth data in the cloud-in direction, the source address is an external public IP address; and the destination address is a public IP address bound to the virtual machine.

[0024] The LAN-side RGW and the WAN-side RGW are connected to a VNAT device cluster, respectively.

[0025] To achieve the above object, the embodiments of the present application further provide a bandwidth data processing method, applied to the bandwidth data processing system as described in any one of the above, and the method comprises:

[0026] Target bandwidth data in a cloud-in direction or a cloud-out direction is obtained; the target bandwidth data includes bandwidth traffic, bandwidth type and service type.

[0027] The target bandwidth data is separated by a target classification algorithm, and the separated data is subjected to flow limiting processing by a gateway cluster or a VNAT device cluster.

[0028] Optionally, the target bandwidth data is separated by a target classification algorithm, and the separated data is subjected to flow limiting processing by a gateway cluster or a VNAT device cluster, including:

[0029] According to the target bandwidth data, the bandwidth type or the service type of the target bandwidth data is determined.

[0030] For target bandwidth data in the cloud-out direction, a first target classification algorithm based on source address information is used for flow separation, or for target bandwidth data in the cloud-in direction, a second target classification algorithm based on destination address information is used for flow separation.

[0031] According to the bandwidth type or service type corresponding to the separated data, the data is sent to a corresponding processing module for flow limiting processing.

[0032] Optionally, according to the bandwidth type or service type corresponding to the separated data, the data is sent to a corresponding processing module for flow limiting processing, comprising:

[0033] When the service type is centralized service traffic, the flow limiting processing is performed through a gateway cluster;

[0034] When the bandwidth type is first type data, the target bandwidth data is subjected to equal proportion network address translation, and the bandwidth limiting processing is performed through a routing gateway RGW;

[0035] When the bandwidth type is second type data, the limiting processing is performed through a network address translation gateway NATGW of the gateway cluster;

[0036] When the bandwidth type is third type data, the limiting processing is performed through a quality of service gateway Qos GW of the gateway cluster of the gateway cluster;

[0037] The bandwidth flow corresponding to the first type data is greater than the bandwidth flow corresponding to the second type data, and the bandwidth flow corresponding to the second type data is greater than the bandwidth flow corresponding to the third type data.

[0038] Optionally, the method further comprises:

[0039] The bandwidth initial value of the first type data, the bandwidth initial value of the second type data and the bandwidth initial value of the third type data are configured through a preset open interface.

[0040] Optionally, the target bandwidth data in the cloud-in direction or the cloud-out direction is obtained, comprising:

[0041] The target bandwidth data in the cloud-out direction is obtained through an RGW located at a local area network LAN side; wherein, in the target bandwidth data in the cloud-out direction, a source address is a dummy internet protocol fip address corresponding to a virtual machine; the dummy fip address has a first mapping relationship between the virtual machine and a public internet protocol IP address bound to the virtual machine; and a destination address is an external public IP address;

[0042] The target bandwidth data in the cloud-in direction is obtained through an RGW located at a wide area network WAN side; wherein, in the target bandwidth data in the cloud-in direction, the source address is the external public IP address; and the destination address is the public IP address bound to the virtual machine;

[0043] The RGW at the LAN side and the RGW at the WAN side are connected with a VNAT device cluster respectively.

[0044] Optionally, the method further comprises:

[0045] The target bandwidth data of the second type of data is processed using independent resource processing, and the target bandwidth data of the third type of data is processed using shared resource processing.

[0046] To achieve the above object, the embodiments of the present application further provide a readable storage medium having a program or instructions stored thereon, the program or instructions being executed by a processor to implement the steps in the bandwidth data processing method according to any one of the above.

[0047] The beneficial effects of the above technical solutions of the present application are as follows:

[0048] The processing system of the embodiments of the present application comprises a routing gateway RGW, a gateway cluster, and a virtual network address translation VNAT device cluster; the RGW is configured to obtain target bandwidth data in a cloud-in direction or a cloud-out direction; the target bandwidth data comprises bandwidth traffic, bandwidth type, and service type; the RGW is further configured to perform separation processing on the target bandwidth data by using a target classification algorithm, and perform flow limiting processing on the separated data by using the gateway cluster or the VNAT device cluster; the present application can separate or limit the flow of different target bandwidth data, can process super-large flow bandwidth, and greatly improves the bandwidth processing capacity of a single cluster. BRIEF DESCRIPTION OF DRAWINGS

[0049] Figure 1 A structural schematic diagram of a bandwidth data processing system provided by the embodiments of the present application is provided.

[0050] Figure 2 A structural schematic diagram of a routing gateway provided by the embodiments of the present application is provided.

[0051] Figure 3 A system architecture diagram of flow separation and service offloading provided by the embodiments of the present application is provided.

[0052] Figure 4 A flowchart of a bandwidth data processing method provided by the embodiments of the present application is provided. DETAILED DESCRIPTION

[0053] To make the technical problems, technical solutions, and advantages of the present application clearer, the following will be described in detail with reference to the drawings and specific embodiments.

[0054] It should be understood that every reference to an "embodiment" or "one embodiment" throughout the specification means that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least one embodiment of the application. Therefore, appearances of the phrases "in one embodiment" or "in an embodiment" in various places throughout the specification are not necessarily all referring to the same embodiment. Furthermore, the particular features, structures, or characteristics can be combined in any suitable manner in one or more embodiments.

[0055] In various embodiments of the present application, it should be understood that the size of the serial number of the following processes does not mean the order of execution, and the execution order of the processes should be determined by its function and inherent logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0056] In the embodiments provided in the present application, it should be understood that "B corresponding to A" means that B is associated with A, and B can be determined according to A. However, it should also be understood that the determination of B according to A does not mean that B is determined only according to A, but B can also be determined according to A and / or other information.

[0057] Before the present application is further described, the terms and terminology used in the embodiments of the present application are explained, and the terms and terminology used in the embodiments of the present application are applicable to the following explanations.

[0058] The underlay network is a carrier network composed of various physical devices, which guarantees the IP connectivity between its devices by using routing or switching protocols.

[0059] The overlay network is one or more virtual logical networks built on the underlay network through network virtualization technology. Although different overlay networks share the devices and lines in the underlay network, the services in the overlay network are decoupled from the physical networking and interconnection technology in the underlay network. The multi-instance of the overlay network can serve different services of the same tenant (such as multiple departments), or different tenants, and is the core networking technology used in the SDN solution of the cloud data center.

[0060] Software Defined Network (SDN) is a network solution for network virtualization in a cloud data center, which includes SDN controller, SDN gateway, virtual switch and other key components.

[0061] Virtual Switch (VSW) is a device managed by the SDN controller through the OpenFlow communication protocol.

[0062] Underlay switch: the collective term for the core components that provide the underlay network, including aggregation switches and core switches.

[0063] Top-of-rack switch (TOR): a device that can be managed by an SDN controller and accessed by the overlay network.

[0064] Network address translation gateway (NAT GW): a device that provides 1:1 public IP services and public proxy services N:1 NAT proxy for cloud tenants.

[0065] Quality of service gateway (QoS GW): an underlay device that provides shared bandwidth services for cloud tenants. Multiple public IPs share a bandwidth (limited speed), allowing cloud hosts, physical machines, load balancing, and cloud databases (such as MySQL) bound to public IPs to access the public network, saving network operation costs for enterprises and facilitating operation and maintenance statistics.

[0066] P4 programming: programming for the data plane. The name P4 actually comes from "Programming Protocol-Independent Packet Processors", which translates to "Protocol-Independent Packet Processor Programming Language".

[0067] Cloud platform: an entry for tenants to configure cloud services, and also a management entry for administrators to configure FIP bandwidth types.

[0068] In a complete cloud data center network environment, various switches such as access switches, aggregation switches and core switches provide underlay network services, and controllers, virtual switches or TORs, SDN gateways and the like provide cloud data center SDN overlay networks. When cloud hosts or other cloud services purchased by cloud tenants need to access public networks or provide services to public networks, they need to subscribe to public IP address binding public IP bandwidth or purchase NAT proxy services or shared bandwidth services according to their own business scenarios through the cloud platform. All tenant public network related services will be configured to the NAT GW and Qos GW. For example, ten thousand public IP services will have ten thousand public IP policy configurations on the NAT GW. Related technologies cloud tenants subscribe to public IP related services to provide public network service; the corresponding capabilities are provided by the underlying NAT GW and QoS GW. With the acceleration of digital transformation and computing power network construction, the number of public IP and bandwidth demand of tenants grows rapidly, resulting in large bandwidth and small bandwidth processing on a forwarding unit, network jitter, packet loss and other phenomena, or due to the processing capacity of the minimum processing unit, the demand of single public IP processing bandwidth greater than the processing capacity of the customer cannot be met. Therefore, the present application provides a bandwidth data processing system, method and readable storage medium, which separates or limits the flow of different target bandwidth data, can process super large flow bandwidth problems, and greatly improves the bandwidth processing capacity of a single cluster.

[0069] Referring to Figure 1 The bandwidth data processing system of the present application comprises a routing gateway RGW 11, a gateway cluster 12 and a virtual network address translation VNAT device cluster 13.

[0070] The RGW 11 is used to obtain target bandwidth data in the cloud direction or out of the cloud direction; the target bandwidth data includes bandwidth flow, bandwidth type and service type.

[0071] The RGW 11 is also used to separate and process the target bandwidth data through a target classification algorithm, and limit the flow of the separated data through the gateway cluster 12 or the VNAT device cluster 13.

[0072] In the embodiments of the present application, the RGW 11 can receive target bandwidth data from the cloud-in direction or the cloud-out direction, separate and process the target bandwidth data in the cloud-in direction by using a target classification algorithm based on a target address, or separate and process the target bandwidth data in the cloud-out direction by using a target classification algorithm based on a source address, and further limit the flow of the separated data by using the gateway cluster 12 or the VNAT device cluster 13. The separated data can be matched with a corresponding flow limiting processing mode by using the data type or the service type of the separated data, so that the problem of super large flow bandwidth can be solved, and the bandwidth processing capacity of a single cluster is greatly improved.

[0073] Optionally, the gateway cluster 12 includes a network address translation gateway NAT GW and a quality of service gateway Qos GW of the gateway cluster; and the VNAT device cluster 13 includes a plurality of VNAT devices.

[0074] Optionally, as shown in Figure 2 The RGW 11 of the present application includes three physical interfaces in three directions:

[0075] LAG1 or LAG2: LAG port formed by bundling a plurality of physical ports, facing the Underlay switch, forming a large bandwidth processing capacity, processing the flow in the cloud-out direction and the cloud-in direction.

[0076] LAG3: LAG port formed by bundling a plurality of physical ports, facing another RGW device, such as RGW-2, serving as a device dual-active high-reliable link protection.

[0077] 1 / 1 / 1, … 1 / 1 / 6: ordinary physical port: facing the server (NAT GW, QoS GW cluster).

[0078] For example, two RGWs and two Underlay switches are provided in the present application, each RGW is connected to two Underlay switches, and each Underlay switch can transmit a flow of N*100G. Each RGW can distribute N*100G to the NAT GW cluster or the QoS GW cluster corresponding to each RGW through 1*100G, so as to realize the purpose of flow splitting.

[0079] Optionally, the routing gateway RGW includes:

[0080] The first determining module is configured to determine the bandwidth type or the service type of the target bandwidth data according to the target bandwidth data.

[0081] The first separating module is configured to separate the target bandwidth data in the cloud-out direction by using a first target classification algorithm based on source address information, or separate the target bandwidth data in the cloud-in direction by using a second target classification algorithm based on destination address information.

[0082] The first processing module is configured to send the separated data to a corresponding processing module for flow limiting processing according to a bandwidth type or a service type corresponding to the separated data.

[0083] In the embodiment, the RGW separates the target bandwidth data in the cloud direction by using a first target classification algorithm based on source address information, and the source address information includes but is not limited to a source address segment or a source address. The source address can also be a preset address that is bound to have a mapping relationship. The RGW separates the target bandwidth data in the cloud direction by using a second target classification algorithm based on destination address information, and the source address information includes but is not limited to a destination address segment or a destination address. The destination address can be an external public IP address bound to a virtual machine (VM) corresponding to the RGW. In the application, the RGW can classify the traffic in the cloud direction by using a source IP address or a source IP address prefix, and hand over the classified data to a processing module corresponding to a bandwidth type or a service type corresponding to the separated data for processing. The traffic in the cloud direction is classified by using a destination IP address or a destination IP address prefix, and handed over to a processing module corresponding to a bandwidth type or a service type corresponding to the separated data for processing. The P4 chip large channel is used to avoid the disadvantage of limited capacity table items.

[0084] The first processing module described above can be understood as a first virtual routing forwarding (VRF) module arranged in the RGW, and the first VRF module has the capability of separating data.

[0085] Optionally, the routing gateway RGW further includes:

[0086] The second processing module is configured to perform flow limiting processing through a gateway cluster when the service type is centralized service traffic.

[0087] The third processing module is configured to perform equal proportion network address translation on the target bandwidth data when the bandwidth type is first type data, and perform bandwidth limiting processing through the routing gateway RGW.

[0088] The fourth processing module is configured to perform limiting processing through a network address translation gateway (NAT GW) of the gateway cluster when the bandwidth type is second type data.

[0089] The fifth processing module is configured to perform limiting processing through a quality of service gateway (Qos GW) of the gateway cluster when the bandwidth type is third type data.

[0090] The bandwidth flow corresponding to the first type data is greater than the bandwidth flow corresponding to the second type data, and the bandwidth flow corresponding to the second type data is greater than the bandwidth flow corresponding to the third type data.

[0091] It should be noted that, from the perspective of cloud services, a processing mode is determined when the service type is determined as centralized service flow, and three other processing modes are determined when the bandwidth types have different bandwidths.

[0092] In an implementation of the present application, when the service type is determined as centralized service flow, such as the business product NAT proxy service, shared bandwidth, etc., the second processing module needs to separate such flow through data plane programmable technology and send it to the NAT GW cluster or QOS GW of the gateway cluster for processing, receive the corresponding backhaul flow, and perform packet forwarding through the destination route.

[0093] In another implementation of the present application, public IP service flow: 1:1 NAT (public IP) and its corresponding speed limiting service (bandwidth), considering different public IP service bandwidths (speed limits), such services are divided into three types: ultra large bandwidth (Ultra), large bandwidth (Large), and small bandwidth (Mini), which correspond to the bandwidth flow corresponding to the first type data, the bandwidth flow corresponding to the second type data, and the bandwidth flow corresponding to the third type data, respectively.

[0094] When the bandwidth type is the first type data, i.e., the bandwidth type is Ultra large bandwidth, the RGW itself is directly processed, and the third processing module is responsible for 1:1 NAT and speed limiting (bidirectional) of public IP with bandwidth of Ultra and above, and performs bandwidth speed limiting processing through the routing gateway RGW; when the bandwidth type is the second type data, i.e., the bandwidth type is Large, this level considers processing by the NAT GW, the fourth processing module sends the separated data to the NAT GW for speed limiting processing, and here, special resources are reserved on the NAT GW for processing to ensure that large bandwidth is guaranteed with sufficient resources; when the bandwidth type is the third type data, i.e., the bandwidth type is Mini small bandwidth, the fourth processing module sends the separated data to the QOS GW for speed limiting processing, and such flow uses shared resources for processing on the NAT GW.

[0095] In the present application, the public IP addresses of ultra large bandwidth, large bandwidth, and small bandwidth are implemented to complete NAT address conversion and QoS speed limiting on different systems.

[0096] Optionally, the processing system is provided with a preset open interface, through which a bandwidth initial value of the first type of data, a bandwidth initial value of the second type of data and a bandwidth initial value of the third type of data are configured.

[0097] In the embodiments of the application, through the preset open interface, the cloud platform administrator can pre-configure the bandwidth initial values of Ultra super bandwidth, Large bandwidth and Mini bandwidth. When a tenant opens a public network IP service and other centralized service traffic, the above configurations are automatically implemented.

[0098] Through the application, the separation processing of large bandwidth and small bandwidth can be realized, mutual influence is avoided, and the large channel processing capacity of the P4 chip is utilized, so that the single public network IP super large flow bandwidth problem can be processed, and the bandwidth processing capacity of a single cluster is greatly improved.

[0099] In a specific embodiment of the application, the RGW includes five VRF modules, namely Root VRF, Bypass VRF, Ultra VRF, Large VRF and Mini VRF, which correspond to the first to fifth processing modules respectively. Root VRF: responsible for interfacing with the underlay switch outside the cluster, processing the ingress and egress traffic in the out-cloud direction and the in-cloud direction. Bypass VRF: responsible for processing public network IP proxy services, shared rate limiting and other N:1 centralized service traffic. Such traffic is separated by Root VRF based on source IP classification algorithm, and is sent to NAT GW or QoS GW for processing through Bypass VRF. Ultra VRF: responsible for 1:1 NAT and rate limiting (bidirectional) of public network IP with bandwidth above Ultra bandwidth. Such traffic is separated and processed by the source IP classification algorithm of Root VRF. Large VRF: responsible for sending Large bandwidth public network IP traffic to NAT GW for processing, and the traffic is load balanced on multiple NAT GW gateways. Mini VRF: responsible for sending Mini bandwidth public network IP traffic to NAT GW gateway for processing, and the traffic is load balanced on multiple NAT GW gateways.

[0100] Optionally, the number of RGWs includes two groups, including:

[0101] The RGW located on the local area network LAN side is used to obtain target bandwidth data in the out-cloud direction; wherein in the target bandwidth data in the out-cloud direction, the source address is a dummy Internet Protocol address dummyfip corresponding to the virtual machine; the dummyfip address has a first mapping relationship between the virtual machine and the public Internet Protocol IP address bound to the virtual machine; and the destination address is an external public IP address.

[0102] The RGW located at the WAN side is used to acquire target bandwidth data in the cloud direction; wherein, in the target bandwidth data in the cloud direction, the source address is an external public network IP address, and the destination address is a public network IP address bound by a virtual machine.

[0103] The LAN side RGW and the WAN side RGW are connected with the VNAT device cluster respectively.

[0104] In the embodiment of the application, facing the DCN direction, the LAN side RGW is used to process the LAN side service, and can process target bandwidth data in the cloud direction; facing the public network direction, the WAN side RGW is used to process the WAN side service, and can process target bandwidth data in the cloud direction. Here, the LAN side service and the WAN side service are connected with the NAT GW / QoS GW cluster.

[0105] For target bandwidth data (such as traffic) in the cloud direction, the first processing module (such as Root VRF module) of the LAN side RGW separates the traffic based on the source address or source address prefix algorithm, and the fourth processing module (such as Large VRF module) and the fifth processing module (such as Mini VRF module) of the LAN side RGW use the source IP address Hash algorithm to perform load sharing; for target bandwidth data (such as traffic) in the cloud direction, the first processing module (such as Root VRF module) of the WAN side RGW uses the destination IP address or destination IP address prefix algorithm to separate the traffic, and the fourth processing module (such as Large VRF module) and the fifth processing module (such as Mini VRF module) of the WAN side RGW use the destination IP address Hash algorithm to perform load sharing; for ultra large bandwidth traffic (ultra public network IP traffic), the third processing module (such as Ultra VRF module) of the LAN side RGW or the third processing module (such as Ultra VRF module) of the WAN side RGW is used to perform 1:1 NAT and rate limiting.

[0106] Here, the first to fifth processing modules can have multiple in each set of service logic, for example, Mini VRF1 and Mini VRF2 are used to process different IP address segments, to ensure that the capacity can be expanded smoothly later.

[0107] It should be noted that the dummyfip address, the virtual machine, and the public network IP address bound to the virtual machine have a corresponding relationship. For example, the address corresponding to the virtual machine is overlayip, forming a first mapping relationship of overlayip<->dummyfip<->public network IP. For example, a cloud tenant creates a virtual machine VM on a cloud platform. At this time, the virtual machine VM is assigned an overlay IP address, namely overlayip, which is valid and unique in the private network (VPC) of the cloud tenant, but the IP address can be reused by different tenants or different VPCs. The address cannot be used to uniquely identify the virtual machine VM outside the overlay. Therefore, when the virtual machine VM applies for a public network IP, the SDN upper layer orchestrator assigns a dummyfip bound to the virtual machine to the cloud tenant through a developed interface, forming the first mapping relationship described above. In the present application, the dummyfip is an underlay ip address segment planned in advance in the resource pool, and the dummyfip corresponding to the virtual machine has uniqueness, which represents the virtual machine VM.

[0108] It should also be noted that when the virtual machine VM accesses the external network (out-cloud direction), the source address is converted to dummyfip through the SDN GW, and enters the Lan side Root VRF processing in the system. The source address is dummyfip, and the destination address is the external public network IP. When the external network accesses the virtual machine VM (in-cloud direction), it enters the Wan side Root VRF processing in the system. The source address is the external public network IP address, and the destination address is the public network IP address bound to the virtual machine VM.

[0109] In another specific embodiment, referring to FIG. 8, the present application embodiment also provides a processing flow of traffic separation and service offloading. Figure 3

[0110] For out-cloud direction traffic: 1, the LAN side Root VRF separates the traffic based on the source address segment or the source address (dummyfip address corresponding to the virtual machine VM), and the separated traffic enters different VRF (Bypass VRF / Ultra VRF / Large VRF / Mini VRF) processing according to different types. For example, refer to Table 1 for source address allocation corresponding VRF processing

[0111] Table 1:

[0112] LAN Root VRF Source Address dummy fip address / prefix Destination VRF for policy routing 192.168.1.0 / 24 Mini VRF 1 192.168.2.0 / 24 Mini VRF 2 192.168.1.10 / 32 Large VRF 1 192.168.2.9 / 32 Large VRF 2 192.168.2.102 / 32 Bypass VRF-1 192.168.1.12 / 32 Bypass VRF-1 192.168.1.19 / 32 Ultra VRF-1 192.168.1.21 / 32 Ultra VRF-1

[0113] ​2、Go to public network IP proxy business, sharing speed limit and other N: 1 centralized traffic into Bypass VRF for next step processing; Bypass VRF uses multiple physical interface VLAN sub-interface with NAT GW cluster and QoS GW cluster connection, through the equal cost multipath (ECMP) way to send traffic to NAT GW or Qos GW, ensure traffic load balancing to NAT GW or Qos GW cluster for processing.

[0114] 3、For bandwidth of Ultra bandwidth above public network IP, into Ultra VRF for processing; In Ultra VRF, complete 1:1 NAT conversion and bandwidth limiting, return to Root VRF (WAN); By Root VRF (WAN) sent to underlay switch, then send the message out of the public network.

[0115] 4、For bandwidth of Large bandwidth above public network IP, into Large VRF for processing; In Large VRF, multiple physical interface VLAN sub-interface with NAT GW cluster connection, through ECMP way to send traffic to NAT GW cluster, ensure traffic load balancing to NAT GW cluster for processing.

[0116] 5、For bandwidth of Mini bandwidth above public network IP, into Mini VRF for processing; In Mini VRF, multiple physical interface VLAN sub-interface with NAT GW cluster connection, through ECMP way to send traffic to NAT GW, ensure traffic load balancing to NAT GW cluster for processing.

[0117] 6、LAN side Large VRF and Mini VRF use source IP based Hash algorithm for load sharing, a single IP will be HASH to the same NAT device on the NAT GW cluster for unified bandwidth limiting, to ensure the accuracy of the speed limit.

[0118] 7、For Mini bandwidth and Large bandwidth of public network IP traffic, the processing logic of the system is consistent, the difference is that the NAT gateway cluster for Large bandwidth of public network IP using independent resource processing, while the Mini bandwidth of public network IP using shared resource processing, to ensure not to occupy each other's resources.

[0119] It should be noted that the difference between Mini and Large traffic is that Mini VRF traffic uses source IP address prefix through Root VRF, while Large VRF traffic uses source IP address through Root VRF.

[0120] Traffic to the cloud direction:

[0121] 1. WAN side Root VRF separates traffic based on destination address segment or destination address (public IP address corresponding to virtual machine VM), and the separated traffic enters different VRF (Bypass VRF / Ultra VRF / Large VRF / Mini VRF) processing according to different categories. For example, refer to Table 2 for source address allocation corresponding VRF processing

[0122] Table 2:

[0123]

[0124]

[0125] 2. Traffic to public IP proxy services, shared rate limiting, and other N:1 centralized services enters Bypass VRF for further processing; Bypass VRF uses multiple physical interface VLAN sub-interfaces to connect with NAT GW cluster or QoS GW cluster, and sends traffic to NAT GW or QoS GW through ECMP to ensure load balancing of traffic to NAT GW or QoS GW cluster for processing.

[0126] 3. For public IP with bandwidth of Ultra bandwidth or above, enter Ultra VRF for processing; after 1:1 NAT in Ultra VRF converts the destination public IP to dummyfip address of virtual machine VM and performs bandwidth limiting, it is returned to Root VRF (LAN); sent by Root VRF (LAN) to underlay switch, and the message is converted to virtual machine VM overlay IP address by SDN GW, and sent to virtual machine VM.

[0127] 4. For public IP with bandwidth of Large bandwidth or above, enter Large VRF for processing; Large VRF uses multiple physical interface VLAN sub-interfaces to connect with NAT GW cluster, and sends traffic to NAT GW through ECMP to ensure load balancing of traffic to NAT GW cluster for processing.

[0128] 5. For public IP with bandwidth of Mini bandwidth or above, enter Mini VRF for processing; Mini VRF uses multiple physical interface VLAN sub-interfaces to connect with NAT GW cluster, and sends traffic to NAT GW through ECMP to ensure load balancing of traffic to NAT GW cluster for processing.

[0129] 6. On the WAN side, Large VRF and Mini VRF use a hash algorithm based on the destination IP for load balancing. A single IP will be hashed to the same NAT GW device in the NAT GW cluster for unified bandwidth limiting to ensure the accuracy of the rate limiting.

[0130] 7. The system's processing logic is consistent for public IP traffic with Mini bandwidth and Large bandwidth. The difference lies in that the NAT GW cluster uses independent resources to process public IP traffic with Large bandwidth, while it uses shared resources for Mini bandwidth traffic, ensuring that they do not compete for resources.

[0131] In summary, the processing system of this application circumvents the limitation of limited capacity entries. This system handles ultra-high bandwidth traffic (Ultra type), while the NAT GW cluster and QoS GW cluster handle other bandwidth traffic, achieving the goal of traffic separation. The system can classify outbound traffic by source IP address or source IP address prefix and assign it to different VRF processing logics; similarly, it can classify inbound traffic by destination IP address or destination IP address prefix and assign it to different VRFs. It can handle ultra-high bandwidth traffic from a single public IP address, while significantly improving the bandwidth processing capacity of a single cluster.

[0132] Reference Figure 4 As shown in the embodiments of this application, a method for processing bandwidth data is also provided, applicable to, for example... Figure 1 The bandwidth data processing system includes a routing gateway (RGW), a gateway cluster, and a virtual network address translation (VNAT) device cluster; the method includes:

[0133] Step 41: Obtain target bandwidth data for the cloud inbound or cloud outbound direction; the target bandwidth data includes bandwidth traffic, bandwidth type, and service type.

[0134] Step 42: The target bandwidth data is separated using a target classification algorithm, and the separated data is then rate-limited through a gateway cluster or a Virtual Network Address Translation (VNAT) device cluster.

[0135] In the embodiments of the present application, the RGW in the processing system can receive target bandwidth data from the cloud-in direction or the cloud-out direction, separate and process the target bandwidth data in the cloud-in direction by using a target classification algorithm based on a target address, or separate and process the target bandwidth data in the cloud-out direction by using a target classification algorithm based on a source address, and further limit the flow of the separated data by using a gateway cluster or a VNAT device cluster, and the separated data can be matched with a corresponding flow limiting processing mode by using the data type or the service type of the separated data, so that the problem of super large flow bandwidth can be solved, and the bandwidth processing capacity of a single cluster is greatly improved.

[0136] Optionally, the step 42 comprises:

[0137] According to the target bandwidth data, the bandwidth type or the service type of the target bandwidth data is determined.

[0138] For the target bandwidth data in the cloud-out direction, a first target classification algorithm based on source address information is used for flow separation, or for the target bandwidth data in the cloud-in direction, a second target classification algorithm based on destination address information is used for flow separation.

[0139] According to the bandwidth type or the service type of the separated data, the separated data is sent to a corresponding processing module for flow limiting processing.

[0140] In the embodiments of the present application, the RGW separates the target bandwidth data in the cloud-out direction by using a first target classification algorithm based on source address information, and the source address information includes but is not limited to a source address segment or a source address. The source address can also be a preset address that is bound to have a mapping relationship. The RGW separates the target bandwidth data in the cloud-in direction by using a second target classification algorithm based on destination address information, and the source address information includes but is not limited to a destination address segment or a destination address. The destination address can be an external public IP address bound to a virtual machine VM corresponding to the RGW. In the present application, the RGW can classify the flow in the cloud-out direction by using the source IP address or the source IP address prefix, and hand over the classified data to a processing module corresponding to the bandwidth type or the service type of the separated data for processing. The flow in the cloud-in direction is classified by using the destination IP address or the destination IP address prefix, and handed over to a processing module corresponding to the bandwidth type or the service type of the separated data for processing. The P4 chip large channel is used to avoid the disadvantage of limited capacity table entries.

[0141] The first processing module described above can be understood as a first virtual routing forwarding (VRF) module arranged in the RGW, and the first VRF module has the capability of separating data.

[0142] Optionally, according to the bandwidth type or service type corresponding to the separated data, the data is sent to a corresponding processing module for flow limiting processing, including:

[0143] When the service type is centralized service traffic, the flow limiting processing is performed through a gateway cluster.

[0144] When the bandwidth type is first type data, the target bandwidth data is subjected to equal proportion network address translation, and the bandwidth limiting processing is performed through a routing gateway RGW.

[0145] When the bandwidth type is second type data, the limiting processing is performed through a network address translation gateway NATGW of the gateway cluster.

[0146] When the bandwidth type is third type data, the limiting processing is performed through a quality of service gateway Qos GW of the gateway cluster.

[0147] The bandwidth flow corresponding to the first type data is greater than the bandwidth flow corresponding to the second type data, and the bandwidth flow corresponding to the second type data is greater than the bandwidth flow corresponding to the third type data.

[0148] It should be noted that, from the perspective of cloud service, when the service type is determined to be centralized service traffic, one processing mode is determined, and when the bandwidth types have different bandwidths, three other processing modes are determined.

[0149] In an implementation manner of the application, when the service type is determined to be centralized service traffic, for example, a service product NAT proxy service, a shared bandwidth, and the like, the data face programmable technology is used to separate such traffic, and the traffic is sent to a NAT GW cluster or a QOS GW of a gateway cluster for processing. Corresponding backhaul traffic is received, and packet forwarding is performed through a destination route.

[0150] In another implementation manner of the application, public network IP service traffic, that is, 1:1 NAT (public network IP) and corresponding speed limiting service (bandwidth), considers different bandwidths (speed limits) of different public network IP services. Such services are divided into three types: ultra large bandwidth (Ultra), large bandwidth (Large), and small bandwidth (Mini), which correspond to the bandwidth flow corresponding to the first type data, the bandwidth flow corresponding to the second type data, and the bandwidth flow corresponding to the third type data, respectively.

[0151] When the bandwidth type is the first type of data, that is, the bandwidth type is Ultra, the third processing module of the RGW is used to directly process the RGW itself, 1:1 NAT and rate limiting (bidirectional) are performed on the public IP with a bandwidth of Ultra or more, and the bandwidth rate limiting is performed through the routing gateway RGW; when the bandwidth type is the second type of data, that is, the bandwidth type is Large, the fourth processing module of the RGW is used to separate the data and send the separated data to the NAT GW for rate limiting, and the NAT GW is reserved for special resource processing to ensure that the large bandwidth is guaranteed with sufficient resources; when the bandwidth type is the third type of data, that is, the bandwidth type is Mini, the fourth processing module of the RGW is used to separate the data and send the separated data to the QOS GW for rate limiting, and the QOS GW is used for rate limiting, and the shared resource processing is used for the traffic on the NAT GW.

[0152] In the present application, the public IP addresses of the ultra large bandwidth, the large bandwidth and the small bandwidth are separated and processed on different systems to complete the NAT address conversion and the QoS rate limiting.

[0153] Optionally, the method further includes:

[0154] The bandwidth initial value of the first type of data, the bandwidth initial value of the second type of data and the bandwidth initial value of the third type of data are configured through a preset open interface.

[0155] In the present application, the bandwidth initial values of the Ultra, the Large and the Mini are pre-configured for the cloud platform administrator through the preset open interface of the RGW, and the above configuration is automatically realized when the tenant opens the public IP service and other centralized service traffic.

[0156] The present application can separate and process the large bandwidth and the small bandwidth, avoid mutual influence, use the large channel processing capacity of the P4 chip, process the ultra large flow bandwidth of a single public IP, and greatly improve the bandwidth processing capacity of a single cluster.

[0157] In a specific embodiment of the present application, the RGW described above includes five VRF modules, namely Root VRF, Bypass VRF, Ultra VRF, Large VRF and Mini VRF, corresponding to the first to fifth processing modules described above. Root VRF: responsible for interfacing with underlay switches outside the cluster, processing ingress and egress traffic in the cloud direction. Bypass VRF: responsible for processing public IP proxy services, shared rate limiting and other N: 1 centralized service traffic. Such traffic is separated by Root VRF based on source IP classification algorithm, and is sent to NAT GW or QoS GW for processing through Bypass VRF. Ultra VRF: responsible for 1:1 NAT and rate limiting (bidirectional) of public IP with bandwidth above Ultra bandwidth. Such traffic is separated and processed by the source IP classification algorithm of Root VRF. Large VRF: responsible for sending Large bandwidth public IP traffic to NAT GW for processing, and the traffic is load balanced on multiple NAT GW gateways. Mini VRF: responsible for sending Mini bandwidth public IP traffic to NAT GW gateway for processing, and the traffic is load balanced on multiple NAT GW gateways.

[0158] Optionally, the step 41 described above comprises:

[0159] The target bandwidth data in the cloud direction is obtained through the RGW located on the local area network LAN side; wherein in the target bandwidth data in the cloud direction, the source address is a dummy IP address corresponding to the virtual machine; the dummy IP address has a first mapping relationship with the public IP address bound to the virtual machine; and the destination address is an external public IP address.

[0160] The target bandwidth data in the cloud direction is obtained through the RGW located on the local area network LAN side; wherein in the target bandwidth data in the cloud direction, the source address is a dummy IP address corresponding to the virtual machine; the dummy IP address has a first mapping relationship with the public IP address bound to the virtual machine; and the destination address is an external public IP address.

[0161] The LAN side RGW and the WAN side RGW are connected with the VNAT device cluster respectively.

[0162] It should be noted that the dummyfip address, the virtual machine, and the public IP address bound to the virtual machine have a corresponding relationship. For example, the address corresponding to the virtual machine is overlayip, forming a first mapping relationship of overlayip<->dummyfip<->public IP. For example, a cloud tenant creates a virtual machine VM on a cloud platform. At this time, the virtual machine VM is assigned an overlay IP address, namely overlayip, which is valid and unique in the private network (VPC) of the cloud tenant, but the IP address can be reused by different tenants or different VPCs. The address cannot be used to uniquely identify the virtual machine VM outside the overlay. Therefore, when the virtual machine VM applies for a public IP, the SDN upper layer orchestrator assigns a dummyfip bound to the virtual machine to the cloud tenant through a developed interface, forming the first mapping relationship described above. In this application, the dummyfip is an underlay ip address segment planned in advance in the resource pool, and the dummyfip corresponding to the virtual machine has uniqueness, which represents the virtual machine VM.

[0163] It should also be noted that when the virtual machine VM accesses the external network (out-cloud direction), the source address is converted to dummyfip through the SDN GW, and enters the Lan side Root VRF processing in the system. The source address is dummyfip, and the destination address is the external public IP. When the external network accesses the virtual machine VM (in-cloud direction), it enters the Wan side Root VRF processing in the system. The source address is the external public IP address, and the destination address is the public IP address bound to the virtual machine VM.

[0164] In this embodiment, the structure diagram shown in Figure 3 The target bandwidth data in the out-cloud direction can be obtained through the RGW located on the LAN side. For example, the LAN side Root VRF separates the traffic based on the source address segment or the source address (dummyfip address corresponding to the virtual machine VM), and the separated traffic enters different VRF (Bypass VRF / Ultra VRF / Large VRF / Mini VRF) processing according to different types. The target bandwidth data in the in-cloud direction can be obtained through the RGW located on the WAN side. For example, the WAN side Root VRF separates the traffic based on the destination address segment or the destination address (public IP address corresponding to the virtual machine VM), and the separated traffic enters different VRF (Bypass VRF / Ultra VRF / Large VRF / Mini VRF) processing according to different types.

[0165] Optionally, the method further comprises:

[0166] The target bandwidth data of the second type of data is processed using independent resources, and the target bandwidth data of the third type of data is processed using shared resources.

[0167] In the present application, the processing logic of the system for Mini bandwidth and Large bandwidth public IP traffic is consistent, the difference is that the NAT GW cluster uses independent resource processing for Large public IP, and Mini uses shared resource processing, ensuring that resources are not occupied by each other.

[0168] In a specific implementation of the present application, for traffic in the out-cloud direction:

[0169] 1. The LAN side Root VRF separates traffic based on source address segment or source address (dummy fip address corresponding to virtual machine VM), and the separated traffic enters different VRF (Bypass VRF / Ultra VRF / Large VRF / Mini VRF) processing according to different types.

[0170] 2. Traffic to public IP proxy services, shared rate limiting, and other N:1 centralized services enters Bypass VRF for further processing; Bypass VRF uses multiple physical interface VLAN sub-interfaces to connect with the NAT GW cluster and the QoS GW cluster, and sends traffic to the NAT GW or QoS GW through ECMP to ensure load balancing of traffic to the NAT GW or Qos GW cluster for processing.

[0171] 3. For public IP with bandwidth of Ultra bandwidth or above, enter Ultra VRF for processing; after 1:1 NAT conversion and bandwidth rate limiting are completed in Ultra VRF, return to Root VRF (WAN); send the message to the underlay switch by Root VRF (WAN), and then send the message out of the public network.

[0172] 4. For public IP with bandwidth of Large bandwidth or above, enter Large VRF for processing; Large VRF uses multiple physical interface VLAN sub-interfaces to connect with the NAT GW cluster, and sends traffic to the NAT GW cluster through ECMP to ensure load balancing of traffic to the NAT GW cluster for processing.

[0173] 5. For public IP with bandwidth of Mini bandwidth or above, enter Mini VRF for processing; Mini VRF uses multiple physical interface VLAN sub-interfaces to connect with the NAT GW cluster, and sends traffic to the NAT GW through ECMP to ensure load balancing of traffic to the NAT GW cluster for processing.

[0174] 6. On the LAN side, Large VRF and Mini VRF use a source IP-based hash algorithm for load balancing. A single IP will be hashed to the same NAT device in the NAT GW cluster for unified bandwidth limiting to ensure the accuracy of the rate limiting.

[0175] 7. The system's processing logic is consistent for public IP traffic with Mini bandwidth and Large bandwidth. The difference lies in that the NAT gateway cluster uses independent resources to process public IP traffic with Large bandwidth, while it uses shared resources to process public IP traffic with Mini bandwidth, ensuring that they do not compete for resources.

[0176] It should be noted that the difference between Mini and Large traffic is that Mini VRF traffic is extracted through the Root VRF using a source IP address prefix, while Large VRF traffic is extracted through the Root VRF using the source IP address.

[0177] For traffic flowing into the cloud:

[0178] 1. The WAN-side Root VRF performs traffic separation based on the destination address range or destination address (the public IP address corresponding to the virtual machine VM). The separated traffic is processed by different VRFs (Bypass VRF / Ultra VRF / Large VRF / Mini VRF) depending on its type.

[0179] 2. Traffic destined for N:1 centralized services such as public IP proxy services and shared rate limiting enters the Bypass VRF for further processing. The Bypass VRF uses VLAN sub-interfaces of multiple physical interfaces to connect to the NAT GW cluster or QoS GW cluster, and sends the traffic to the NAT GW or QoS GW via ECMP to ensure that the traffic load is balanced to the NAT GW or QoS GW cluster for processing.

[0180] 3. For public IP addresses with bandwidth exceeding Ultra bandwidth, they are processed by the Ultra VRF. Inside the Ultra VRF, a 1:1 NAT is performed to convert the destination public IP address to the dummyfip address of the virtual machine (VM) and bandwidth limiting is applied before the packet is sent back to the Root VRF (LAN). The Root VRF (LAN) then sends the packet to the underlay switch. The SDN GW converts the destination address to the overlay IP address of the VM and sends it to the VM.

[0181] 4. For public network IP with bandwidth of Large or above, enter Large VRF for processing; in Large VRF, use VLAN sub-interfaces of multiple physical interfaces to connect with NAT GW cluster, and send traffic to NAT GW through ECMP mode to ensure load balancing of traffic to NAT GW cluster for processing.

[0182] 5. For public network IP with bandwidth of Mini or above, enter Mini VRF for processing; in Mini VRF, use VLAN sub-interfaces of multiple physical interfaces to connect with NAT GW cluster, and send traffic to NAT GW through ECMP mode to ensure load balancing of traffic to NAT GW cluster for processing.

[0183] 6. Large VRF and Mini VRF on the WAN side use Hash algorithm based on destination IP for load sharing, and a single IP is hashed to the same NAT GW device in the NAT GW cluster for unified bandwidth limiting to ensure the accuracy of bandwidth limiting.

[0184] 7. For public network IP traffic with Mini bandwidth and Large bandwidth, the processing logic of the system is consistent, and the difference is that the NAT GW cluster uses independent resources for processing of Large public network IP, and shared resources for processing of Mini, to ensure that resources are not occupied by each other.

[0185] In summary, the present application solves the limitation that a single public network IP cannot exceed 10G, and can meet the demand of a single public network IP with super large bandwidth for a tenant; through traffic classification routing algorithm, large bandwidth and small bandwidth are separated for processing to avoid mutual influence, and the problem of network packet loss and delay caused by mixed running of large bandwidth and small bandwidth is solved; the capacity of a single cluster can be upgraded from several hundred G to several T, the number of cluster machines is greatly reduced under the condition of the same bandwidth capacity, and the cost is reduced; at the same time, the expansion frequency and difficulty are reduced.

[0186] The readable storage medium of the embodiment of the present application has programs or instructions stored thereon, the programs or instructions are executed by a processor to implement the steps in the bandwidth data processing method described above, and the same technical effects can be achieved. To avoid repetition, details are not described here.

[0187] The processor is the processor in the bandwidth data processing method described in the above embodiments. The readable storage medium includes a computer readable storage medium, such as a computer readable memory (Read-Only Memory, ROM for short), a random access memory (Random Access Memory, RAM for short), a magnetic disk or an optical disk, etc.

[0188] The foregoing exemplary embodiments are described with reference made to the drawings which are provided for the purpose of explanation and illustration and are not intended to limit the scope of the application. Indeed, various modifications and variations that fall within the spirit and scope of the application can become apparent to those skilled in the art upon reading this specification, and it is to be understood that such modifications and variations are intended to fall within the scope of the application. Further, it is to be understood that the phraseology and terminology employed herein are for the purpose of description and explanation and are not intended to be limiting. The use of "including" and "comprising" and variations thereof is meant to encompass the items listed thereafter and equivalents thereof as well as additional items and equivalents thereof. Unless otherwise specified, a range of values includes the beginning and end points of the range and any sub-ranges therebetween.

[0189] The above description is considered that of the preferred embodiments of the application only. Modifications and alterations will occur to others upon reading the description contained herein. It is intended to include all such modifications and alterations insofar as they come within the scope of the application. Accordingly, the application is not to be restricted except in the spirit of the claims.

Claims

1. A bandwidth data processing system, characterized in that, include: Routing gateway RGW, gateway cluster, and virtual network address translation (VNAT) device cluster; RGW is used to obtain target bandwidth data for either cloud inbound or cloud outbound directions; the target bandwidth data includes bandwidth traffic, bandwidth type, and service type. RGW is also used to separate the target bandwidth data using a target classification algorithm, and to perform rate limiting on the separated data through the gateway cluster or VNAT device cluster.

2. The processing system according to claim 1, characterized in that, The routing gateway RGW includes: The first determining module is used to determine the bandwidth type or service type of the target bandwidth data based on the target bandwidth data. The first separation module is used to perform traffic separation on target bandwidth data in the cloud outbound direction using a first target classification algorithm based on source address information, or to perform traffic separation on target bandwidth data in the cloud inbound direction using a second target classification algorithm based on destination address information. The first processing module is used to send the separated data to the corresponding processing module for rate limiting based on the bandwidth type or service type.

3. The processing system according to claim 1, characterized in that, The routing gateway RGW also includes: The second processing module is used to perform rate limiting through the gateway cluster when the service type is centralized service traffic. The third processing module is used to perform proportional network address translation on the target bandwidth data when the bandwidth type is the first type of data, and to perform bandwidth rate limiting through the routing gateway RGW. The fourth processing module is used to perform rate limiting through the Network Address Translation Gateway (NAT GW) of the gateway cluster when the bandwidth type is the second type of data. The fifth processing module is used to perform rate limiting through the Quality of Service (QoS) Gateway (GW) of the gateway cluster when the bandwidth type is the third type of data. Wherein, the bandwidth traffic corresponding to the first type of data is greater than the bandwidth traffic corresponding to the second type of data; the bandwidth traffic corresponding to the second type of data is greater than the bandwidth traffic corresponding to the third type of data.

4. The processing system according to claim 3, characterized in that, The processing system is equipped with a preset open interface, through which the initial bandwidth values ​​of the first type of data, the second type of data, and the third type of data are configured.

5. The processing system according to claim 1, characterized in that, The number of RGWs includes two groups, including: The RGW located on the LAN side is used to obtain target bandwidth data in the cloud outbound direction; wherein, in the target bandwidth data in the cloud outbound direction, the source address is the target Internet Protocol dummyfip address corresponding to the virtual machine; the dummyfip address has a first mapping relationship between the virtual machine and the public Internet Protocol IP address bound to the virtual machine; the destination address is an external public IP address; The RGW located on the WAN side is used to obtain target bandwidth data in the direction of cloud access; in the target bandwidth data in the direction of cloud access, the source address is the external public IP address; the destination address is the public IP address bound to the virtual machine. The RGW on the LAN side and the RGW on the WAN side are respectively connected to the VNAT device cluster.

6. A method for processing bandwidth data, characterized in that, The method, applied to a bandwidth data processing system as described in any one of claims 1 to 5, comprises: Obtain target bandwidth data for either cloud inbound or cloud outbound directions; the target bandwidth data includes bandwidth traffic, bandwidth type, and service type. The target bandwidth data is separated using a target classification algorithm, and the separated data is then rate-limited through a gateway cluster or a Virtual Network Address Translation (VNAT) device cluster.

7. The method according to claim 6, characterized in that, The target bandwidth data is separated using a target classification algorithm, and the separated data is then rate-limited through a gateway cluster or a Virtual Network Address Translation (VNAT) device cluster, including: Based on the target bandwidth data, determine the bandwidth type or service type of the target bandwidth data; For target bandwidth data in the cloud outbound direction, traffic separation is performed using a first target classification algorithm based on source address information; or for target bandwidth data in the cloud inbound direction, traffic separation is performed using a second target classification algorithm based on destination address information. Based on the bandwidth type or service type of the separated data, it is sent to the corresponding processing module for rate limiting.

8. The method according to claim 7, characterized in that, Based on the bandwidth type or service type corresponding to the separated data, it is sent to the corresponding processing module for rate limiting, including: When the service type is centralized service traffic, rate limiting is performed through a gateway cluster; When the bandwidth type is the first type of data, the target bandwidth data is subjected to proportional network address translation, and bandwidth limiting is performed through the routing gateway RGW. When the bandwidth type is the second type of data, rate limiting is performed through the Network Address Translation Gateway (NAT GW) of the gateway cluster. When the bandwidth type is the third type of data, rate limiting is performed through the Quality of Service (QoS) Gateway (GW) of the gateway cluster. Wherein, the bandwidth traffic corresponding to the first type of data is greater than the bandwidth traffic corresponding to the second type of data; the bandwidth traffic corresponding to the second type of data is greater than the bandwidth traffic corresponding to the third type of data.

9. The method according to claim 8, characterized in that, The method further includes: The initial bandwidth values ​​for the first type of data, the second type of data, and the third type of data are configured through a preset open interface.

10. The method according to claim 6, characterized in that, Obtain target bandwidth data for either the inbound or outbound cloud direction, including: The target bandwidth data in the cloud direction is obtained through the RGW located on the LAN side; wherein, in the target bandwidth data in the cloud direction, the source address is the target Internet Protocol dummyfip address corresponding to the virtual machine; the dummyfip address has a first mapping relationship between the virtual machine and the public Internet Protocol IP address bound to the virtual machine; the destination address is the external public IP address. The target bandwidth data for the cloud inbound direction is obtained through the RGW located on the WAN side; in the target bandwidth data for the cloud inbound direction, the source address is the external public IP address; and the destination address is the public IP address bound to the virtual machine. The RGW on the LAN side and the RGW on the WAN side are respectively connected to the VNAT device cluster.

11. The method according to claim 8, characterized in that, The method further includes: The target bandwidth data for the second type of data is processed using independent resources, while the target bandwidth data for the third type of data is processed using shared resources.

12. A readable storage medium having a program or instructions stored thereon, characterized in that, When the program or instructions are executed by the processor, they implement the steps in the bandwidth data processing method as described in any one of claims 6 to 11.

Images