Device for improving communication reliability of cross-cloud data center network

Abstract

The invention provides a device for improving the reliability of cross-cloud data center network communication, and belongs to the field of computer network communication, a proxy gateway group realizes cross-cloud data center network interconnection, a flow monitoring device monitors the cross-cloud data center flow load condition and reports the cross-cloud data center flow load condition to a central controller, and the central controller is connected with the proxy gateway group. And the central controller carries out elastic capacity expansion and shrinkage on the proxy gateway group, and load sharing of flow is realized through an ECMP (Equal Coordinated Multi-Point). In addition, the traffic detection device can also monitor the survival state of the proxy gateway and start or stop the proxy gateway in time, so that the reliability of cross-network communication is improved.

Description

Technical Field

[0001] The present invention relates to the field of computer network communications, and in particular to a device for improving the reliability of network communications across cloud data centers. Background Art

[0002] With the advent of the digital and intelligent era, cloud data center networks are becoming more diverse. The traditional method of achieving cross-cloud data center network communication is to connect lines through network operators, but this method is often time-consuming, labor-intensive and costly, and is not suitable for the current era of rapidly changing network interconnection.

[0003] Currently, most of the connection lines based on cross-cloud data center networks adopt single-line mode, single-gateway mode or other complex protocols such as BGP for networking. Line abnormalities or gateway failures during interconnection communications will directly lead to communication interruptions between cross-data center networks.

[0004] Therefore, improving the reliability of cross-cloud data center network communications has become a top priority in Internet solutions at this stage. The present invention will propose a device for improving the reliability of cross-network communications based on the cross-cloud data center network interconnection scenario. Summary of the invention

[0005] In order to solve the above technical problems, the present invention provides a device for improving the reliability of network communication across cloud data centers. Communication between two cross-cloud data center networks is realized through a proxy gateway group, and the traffic load of the proxy gateway is detected by a traffic detection device to elastically expand and shrink the proxy gateway. Load sharing of traffic between proxy gateways is realized through equal-cost routing ECMP. The complexity of network communication across cloud data centers is simplified, network resources are fully utilized, and the reliability of network communication across cloud data centers is improved.

[0006] The technical solution of the present invention is:

[0007] A device for improving the reliability of network communication across cloud data centers, comprising:

[0008] Proxy gateway group to achieve network interconnection across cloud data centers,

[0009] The traffic monitoring device monitors the traffic load across cloud data centers and reports it to the central controller, which elastically scales the proxy gateway group and implements traffic load sharing through equal-cost routing (ECMP).

[0010] Furthermore,

[0011] The flow detection device can also monitor the survival status of the proxy gateway and enable or disable the proxy gateway in time.

[0012] Furthermore,

[0013] The next hop of the route from the cloud data center network to the destination network segment of the peer network passes through the proxy gateway. The proxy gateway and the cloud data center network are configured to be interconnected at Layer 2. Here, the proxy gateway realizes Layer 2 interconnection by carrying a private IP address in the cloud data center network.

[0014] The proxy gateway communicates with the two cross-cloud data centers through the VLAN physical network, or through elastic public IP and GRE tunnel.

[0015] The proxy gateway groups are deployed at the edge of the cloud data center and are managed by their respective central controllers.

[0016] Furthermore,

[0017] The router for communication between cloud data center networks reduces the number of proxy gateways when the traffic load pressure between the two cloud data centers is lower than the set value. If the load pressure is overloaded, the central controller elastically expands the number of proxy gateways and balances the data traffic between each proxy gateway through equal-cost routing.

[0018] The equivalent route is set on the internal router of the cloud data center network and is issued based on the gateway interface of the virtual private network. The next hop is specified to the private IP address assigned by the proxy gateway, namely the hip address. It is occupied by the proxy gateway when the virtual private network is created in the cloud data center. After the proxy gateway group is created, it is mounted on the proxy gateway.

[0019] The traffic load pressure is monitored by the traffic monitoring device. By building a message counter in the proxy gateway, the traffic of each hip interface is counted. The traffic monitoring device periodically extracts the counter information on the proxy gateway and the CPU usage of the proxy gateway to determine the load of the proxy gateway. When a certain threshold is reached, it is reported to the central controller. The central controller controls the elastic expansion and contraction of the proxy gateway and adjusts the configuration of equivalent routes on the internal routers of the cloud data center network.

[0020] The proxy gateway groups of each cloud data center communicate with each other through the VLAN physical network. Each hip interface corresponds to a VLAN interface. The two interfaces are divided into the same network namespace, and network isolation is performed through the network namespace.

[0021] Deploy proxy gateway groups and traffic monitoring devices between cloud data centers to achieve network communication between cloud data centers;

[0022] The proxy gateway group mounts the hip interface in the virtual private network to forward the traffic in the virtual private network to the proxy gateway;

[0023] The qrouter of the virtual private network sends equal-cost routing ECMP to achieve traffic load sharing;

[0024] The proxy gateway isolates traffic between different networks through namespaces;

[0025] The proxy gateway starts a counter to calculate the number of messages under each namespace and reports it to the traffic monitoring device. The traffic monitoring device calculates the load pressure of each namespace on the proxy gateway and reports it to the central controller. The central controller elastically expands and shrinks the proxy gateway group.

[0026] The beneficial effects of the present invention are

[0027] The present invention is based on cross-cloud data center network communication. By deploying proxy gateway groups and traffic monitoring devices between cross-cloud data centers, the traffic load between cross-cloud data centers is monitored in real time, the proxy gateway groups are elastically expanded and shrunk, and the traffic is shared to multiple proxy gateways through equal-cost routing ECMP, thereby improving network resource utilization, realizing traffic load sharing, and improving the reliability of network communication. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 Establish an initial proxy gateway group and local virtual machine private network for cross-cloud data center communication. The next hop of the forwarded traffic on the local qrouter is a single hip interface, the proxy gateway group is in stand-alone mode, and the traffic between cross-cloud data centers is a single path;

[0029] Figure 2 The flowchart of elastic expansion and contraction of the proxy gateway group is shown in the figure. The central controller receives the traffic monitoring information of the proxy gateway group and elastically expands or contracts the proxy gateway according to the remaining resources in the current environment.

[0030] Figure 3 When the proxy gateway group is overloaded, it is elastically expanded to active-active mode. At this time, traffic load sharing is achieved through equal-cost routing ECMP, which reduces the load pressure of a single proxy gateway and improves the reliability of network communication across data centers. DETAILED DESCRIPTION

[0031] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0032] The present invention provides a device for improving the reliability of network communication across cloud data centers. Communication between networks across cloud data centers is achieved through a proxy gateway group. The proxy gateway is monitored by a traffic monitor for its load condition and survival status, and reported to a central controller according to the load condition. The central controller controls the proxy gateway to perform elastic expansion and contraction, and load sharing of traffic is achieved through equal-cost routing ECMP, thereby improving the reliability of network communication across data centers and avoiding communication interruptions caused by line or gateway downtime.

[0033] The present invention realizes the role of elastic expansion and contraction of the proxy gateway and traffic load sharing by deploying a proxy gateway group and a traffic monitoring device between cross-cloud data center networks, thereby reducing message discarding and improving the reliability of cross-cloud data center network communication. The cloud data center network needs to be configured to pass through the proxy gateway as the next hop of the route to the destination network segment of the opposite network. The proxy gateway and the cloud data center network need to be configured as Layer 2 intercommunication. Here, the proxy gateway can achieve Layer 2 intercommunication by carrying a private IP address in the cloud data center network. The proxy gateway and two cross-cloud data centers can be connected through a VLAN physical network, or communicate through an elastic public network IP and a GRE tunnel. The present invention pulls through cross-network communication based on a VLAN physical network.

[0034] The proxy gateway group is deployed at the edge of the cloud data center and is managed by its own central controller. It is a router for communication between cloud data center networks. When the traffic load pressure between the two cloud data centers is low, the number of proxy gateways is reduced. When the load pressure is overloaded, the central controller elastically expands the number of proxy gateways and balances the data traffic between the proxy gateways through equal-cost routing ECMP. This equal-cost routing is set on the internal router of the cloud data center network and is issued based on the gateway interface of the virtual private network. The next hop is specified to the private network IP address assigned by the proxy gateway. This IP address is subsequently called the hip address and is occupied by the proxy gateway when the virtual private network is created in the cloud data center. After the proxy gateway group is created, it is mounted on the proxy gateway.

[0035] The traffic load pressure mentioned above is monitored by the traffic monitoring device. By building a message counter in the proxy gateway, the traffic of each hip interface is counted. The traffic monitoring device periodically extracts the counter information on the proxy gateway and the CPU usage of the proxy gateway to determine the load of the proxy gateway. When a certain threshold is reached, it is reported to the central controller. The central controller controls the elastic expansion and contraction of the proxy gateway and adjusts the configuration of equivalent routes on the internal routers of the cloud data center network.

[0036] The proxy gateway groups of each cloud data center communicate with each other through the VLAN physical network. Each hip interface corresponds to a VLAN interface. The two interfaces are divided into the same network namespace. Network isolation is performed through the network namespace, which can not only realize the intercommunication between two specific network segments across cloud data centers, but also prevent network series connection.

[0037] The present invention will be further described below based on the description of the drawings.

[0038] like Figure 1 As shown in the figure, in the stand-alone mode of the proxy gateway group, cross-cloud data center network communication is required. Cloud data center 1 and cloud data center 2 need to achieve cross-segment mutual access. The following takes the communication between the virtual private network vpc1:192.168.0.0 / 24 in cloud data center 1 and the virtual private network vpc2:172.16.31.0 / 24 in cloud data center 2 as an example to illustrate.

[0039] When creating a virtual private network, a hip address is assigned to the proxy gateway. The number of hips in each virtual private network is assigned according to the default configuration. This configuration will determine the data of the proxy gateway in the proxy gateway group, that is, the maximum number of proxy gateways that can be expanded. Generally, the number of hip addresses is set to 3. In vpc1, the hip addresses are: 192.168.0.5, 192.168.0.6, 192.168.0.7, and in vpc2, the hip addresses are: 172.16.31.5, 172.16.31.6, 172.16.31.7. In the stand-alone mode of the proxy gateway group, one of the hip interfaces is mounted to the proxy gateway.

[0040] Before creating a proxy gateway group, you need to allocate a VLAN physical network in advance for communication between the proxy gateway groups of the two cloud data centers. In this example, the allocated VLAN network segment is: 100.100.1.0 / 24. When the cross-segment network of the two cloud data centers needs to establish communication needs, configure a VLAN network IP address for each of the two proxy gateways to open up the proxy gateway group at Layer 2.

[0041] The process of establishing cross-segment communication request between vpc1 in cloud data center 1 and vpc2 in cloud data center 2 is as follows:

[0042] (1) Create a proxy gateway group. Two cloud data centers establish proxy gateway groups respectively. The proxy gateway is a virtual machine of the Linux system. The hip interface and the port of the VLAN network are mounted to the proxy gateway group. The hip address mounted on the proxy gateway group of cloud data center 1 is: 192.16.0.5, and the VLAN network address is: 100.100.1.11. The hip address mounted on the proxy gateway group of cloud data center 2 is: 172.16.31.5, and the VLAN network address is: 100.100.1.12.

[0043] (2) The proxy gateway group creates a namespace and adds the hip interface and the VLAN network interface to the namespace;

[0044] (3) In the namespace of cloud data center 1, the following network segment routes are issued based on the destination network segment to guide forwarding: destination network segment 172.16.31.0 / 24, next hop: 100.100.1.12; destination network segment: 192.168.0.0 / 24, next hop: 192.168.0.1;

[0045] (4) In the namespace of cloud data center 2, the following network segment routing is issued based on the destination network segment to guide forwarding: destination network segment 192.168.0.0 / 24, next hop: 100.100.1.11; destination network segment: 172.16.31.0 / 24, next hop: 172.16.31.1;

[0046] (5) Add the network segment route for guiding route forwarding in the qrouter of the virtual private network vpc1: destination network segment: 172.16.31.0 / 24, next hop: 192.168.0.5; add the network segment route for guiding route forwarding in the qrouter of the virtual private network vpc2: destination network segment: 192.168.0.0 / 24, next hop: 172.16.31.5

[0047] Through the above configuration, the communication between cloud data center 1 and cloud data center 2 is realized in the stand-alone mode of the proxy gateway group. The virtual machine under the virtual private network vpc1 requests the virtual machine under the virtual private network vpc2, and the traffic reaches the proxy gateway group 1 through the qrouter of vpc1. The ns1 namespace of the proxy gateway group 1 directs the traffic to the ns1 namespace of the proxy gateway group 2 through forwarding routing. The ns1 namespace of the proxy gateway 2 directs the traffic to the qrouter of the vpc2 of this cloud center through the destination network segment 172.16.31.0 / 24 and the next hop: 172.16.31.1. The qrouter then forwards the traffic to the destination virtual machine. The destination virtual machine parses the message and re-encapsulates the reply message to complete this communication.

[0048] like Figure 2As shown in the figure, the central controller controls the elastic expansion and contraction of the proxy gateway group. The specific process is as follows:

[0049] (1) Each proxy gateway has a counter on the hip interface of the namespace, which reports the number of messages in the namespace to the traffic monitoring device every minute;

[0050] (2) The traffic monitoring device counts the number of messages in each namespace, calculates the load pressure, displays the traffic status of the proxy gateway on the user interface in real time, and reports the load pressure status of the proxy gateway to the central controller every 5 minutes;

[0051] (3) After the central controller receives the traffic load pressure reported by the traffic monitoring device, it elastically expands or contracts the proxy gateway according to the remaining resources in the current environment, expands the namespace with heavy load pressure, and makes real-time adjustments to adapt to different traffic pressure conditions.

[0052] like Figure 3 As shown in the figure, after the traffic load pressure between cloud data centers is overloaded, the proxy gateway group is elastically expanded to active-active mode, the ns1 namespace with high load pressure is expanded to the new proxy gateway, and equal-cost routes are added on qrouter to evenly distribute the traffic to the namespaces of the two proxy gateways. The specific process is as follows:

[0053] (1) The central controller issues an instruction to create a proxy gateway. The proxy gateway is a virtual machine of the Linux system, which is the same as the initialization. Proxy gateway 3 and proxy gateway 1 form the proxy gateway group of cloud data center 1, and proxy gateway 4 and proxy gateway 2 form the proxy gateway group of cloud data center 2.

[0054] (2) The hip interface of vpc1 is mounted on proxy gateway 3, and a port of a VLAN network is allocated. Namespace ns1' is created on proxy gateway 3, and the hip interface and the port of the VLAN network are added to ns1', and the network segment routing of cloud data center 2 is sent down. The hip interface of vpc2 is mounted on proxy gateway 4, and a port of a VLAN network is allocated. A namespace is created, and the interface is added to the namespace, and the network segment routing of cloud data center 1 is sent down.

[0055] (3) Modify the network segment route of the qrouter of the virtual private network vpc1 to guide the forwarding of the route: destination network segment: 172.16.31.0 / 24, next hop: 192.168.0.5 and 192.168.0.6; modify the network segment route of the qrouter of the virtual private network vpc2 to guide the forwarding of the route: destination network segment: 192.168.0.0 / 24, next hop: 172.16.31.5 and 172.16.31.6;

[0056] By elastically scaling the proxy gateway group, the traffic from vpc1 in cloud data center 1 to vpc2 in cloud data center 2 is load-balanced to the two proxy gateways, which greatly reduces the traffic load pressure and improves the reliability of traffic communication across cloud data centers.

[0057] The above description is only a preferred embodiment of the present invention, which is only used to illustrate the technical solution of the present invention, and is not used to limit the protection scope of the present invention. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention are included in the protection scope of the present invention.

Claims

1. A device for improving the reliability of network communication across cloud data centers, characterized in that: include Proxy gateway group to achieve network interconnection across cloud data centers, The traffic monitoring device monitors the traffic load across cloud data centers and reports it to the central controller, which elastically scales the proxy gateway group and implements traffic load sharing through equal-cost routing (ECMP).

2. The device according to claim 1, characterized in that The flow detection device can also monitor the survival status of the proxy gateway and enable or disable the proxy gateway in time.

3. The device according to claim 1, characterized in that The next hop of the route from the cloud data center network to the destination network segment of the peer network passes through the proxy gateway. The proxy gateway and the cloud data center network are configured to be interconnected at Layer 2. Here, the proxy gateway realizes Layer 2 interconnection by carrying a private IP address in the cloud data center network.

4. The device according to claim 3, characterized in that The proxy gateway communicates with the two cross-cloud data centers through the VLAN physical network, or through elastic public IP and GRE tunnel.

5. The device according to any one of claims 1 to 4, characterized in that: The proxy gateway groups are deployed at the edge of the cloud data center and are managed by their respective central controllers.

6. The device according to claim 1, characterized in that The router for communication between cloud data center networks reduces the number of proxy gateways when the traffic load pressure between the two cloud data centers is lower than the set value. If the load pressure is overloaded, the central controller elastically expands the number of proxy gateways and balances the data traffic between each proxy gateway through equal-cost routing.

7. The device according to claim 6, characterized in that The equivalent route is set on the internal router of the cloud data center network and is issued based on the gateway interface of the virtual private network. The next hop is specified to the private IP address assigned by the proxy gateway, namely the hip address. It is occupied by the proxy gateway when the virtual private network is created in the cloud data center. After the proxy gateway group is created, it is mounted on the proxy gateway.

8. The device according to claim 6, characterized in that The traffic load pressure is monitored by the traffic monitoring device. By building a message counter in the proxy gateway, the traffic of each hip interface is counted. The traffic monitoring device periodically extracts the counter information on the proxy gateway and the CPU usage of the proxy gateway to determine the load of the proxy gateway. When a certain threshold is reached, it is reported to the central controller. The central controller controls the elastic expansion and contraction of the proxy gateway and adjusts the configuration of equivalent routes on the internal routers of the cloud data center network.

9. The device according to claim 7, characterized in that The proxy gateway groups of each cloud data center communicate with each other through the VLAN physical network. Each hip interface corresponds to a VLAN interface. The two interfaces are divided into the same network namespace, and network isolation is performed through the network namespace.

Images