Flow distribution system based on redundant switch

Abstract

The invention provides a traffic distribution system based on redundant switches, and relates to the field of switches, the system comprises an access layer, a convergence layer and a core layer, the access layer is used for configuring an IP address and distributing traffic for a device to be distributed, and transmitting a traffic feature vector corresponding to the device to be distributed to a target convergence switch; the convergence layer is used for integrating the traffic of the access layer and forwarding the traffic of the core layer; the core layer is used for distributing and issuing traffic; the flow distribution system of the redundant switch is higher in reliability and usability, and the actual value of the flow distribution system of the redundant switch in the application of various industries is larger.

Description

Technical Field

[0001] This application relates to the technical field of switches, and in particular, to a traffic distribution system based on redundant switches. Background Art

[0002] A switch is a network device used to connect multiple devices in a computer network and implement data frame forwarding. In a system without switch redundancy, its reliability is poor because if the switch fails, all connected servers will lose network connection, resulting in service interruption. Summary of the Invention

[0003] In view of the above technical problems, this application provides a traffic distribution system based on redundant switches, which at least partially solves the problems existing in the prior art.

[0004] In the first aspect of this application, a traffic distribution system based on redundant switches is provided. The system includes: an access layer, an aggregation layer, and a core layer. Among them, both the aggregation layer and the core layer include at least two switches; each access switch in the access layer is connected to several devices to be allocated; each aggregation switch in the aggregation layer is connected to all access switches included in the access layer; each core switch in the core layer is connected to all aggregation switches included in the aggregation layer;

[0005] The access switch is used to configure IP addresses for the devices to be allocated and distribute traffic, and transmit the traffic feature vectors corresponding to the devices to be allocated to the target aggregation switch; among them, the traffic feature vectors of the devices to be allocated are determined according to the traffic corresponding to the devices to be allocated within a preset time window, the CPU usage rate and the disk read and write rate corresponding to each preset collection time point; the target aggregation switch is the aggregation switch currently in the working state; the target aggregation switch is any aggregation switch;

[0006] The aggregation switch is used to integrate the traffic feature vectors of the devices to be allocated uploaded by all access switches included in the access layer, and send the integrated traffic feature vectors to the target core switch, and when receiving the traffic sent by the target core switch, forward the traffic sent by the target core switch to each access switch in the access layer; the target core switch is the core switch currently in the working state; the target core switch is any core switch;

[0007] The core switch is used to obtain the integrated traffic feature vectors sent by the target aggregation switch, generate traffic distribution information according to the traffic feature vectors of each device to be allocated, and send the traffic distribution information to the target aggregation switch.

[0008] Optionally, the multi-spanning tree protocol and the virtual router redundancy protocol are used from the access layer to the core layer of the system, where:

[0009] Multiple Spanning Tree Protocol, used to create multiple spanning tree instances;

[0010] Virtual Router Redundancy Protocol, used to create virtual routers.

[0011] Optionally, the core switch is used to perform the following steps:

[0012] Obtain the standardized traffic feature vector corresponding to the traffic feature vector of each device to be allocated;

[0013] Obtain the derived feature vector corresponding to each standardized traffic feature vector;

[0014] Based on the traffic feature vector and the derived feature vector corresponding to each device to be allocated, obtain the traffic allocation information corresponding to each device to be allocated;

[0015] Send the traffic classification information to the target aggregation switch.

[0016] Optionally, the derived feature vector includes a traffic change rate feature, a traffic moving average difference feature, and a joint change rate feature.

[0017] Optionally, the traffic change rate feature is:

[0018]

[0019] where, is the standardized traffic of the device to be allocated corresponding to the t-th preset collection time point, is the standardized traffic of the device to be allocated corresponding to the (t - 1)-th preset collection time point; avg() is a preset average value determination function; t = 1, 2,..., m; m is the number of preset collection time points included in the preset time window.

[0020] Optionally, the traffic moving average difference feature is:

[0021]

[0022] where, is the moving average value, and the moving average value is:

[0023]

[0024] where, is the standardized traffic of the device to be allocated corresponding to the i-th preset collection time point, and n is the number of preset collection time points used to calculate the moving average value.

[0025] Optionally, the joint change rate feature is:

[0026]

[0027] Among them, is the combined change rate feature, is the weight of the disk read / write rate change; is the weight of the CPU usage rate change, is the change rate of the disk read / write rate, is:

[0028]

[0029] Among them, is the disk read / write rate corresponding to the t-th preset acquisition time point, is the disk read / write rate corresponding to the (t - 1)-th preset acquisition time point;

[0030] is the CPU usage change rate, is:

[0031]

[0032] Among them, is the CPU usage rate corresponding to the t-th preset acquisition time point, is the CPU usage rate corresponding to the (t - 1)-th preset acquisition time point.

[0033] Optionally, if the target aggregation switch fails, any aggregation switch other than the target aggregation switch in the aggregation layer is switched to the working state.

[0034] Optionally, if the target core switch fails, any core switch other than the target core switch in the core layer is switched to the working state.

[0035] Optionally, any two aggregation switches or any two core switches are connected by at least one link.

[0036] This application has at least the following beneficial effects:

[0037] The traffic distribution system based on redundant switches provided by this application, by setting up the aggregation layer, shares the burden of the core layer with heavy tasks. In addition, redundant switches are respectively set for each aggregation switch and each core switch, so that the traffic distribution system based on redundant switches formed by this application can be stable and available. In addition, the core switch is used to obtain the integrated traffic feature vector sent by the target aggregation switch, generate traffic distribution information according to the traffic feature vector of each device to be allocated, and send the traffic distribution information to the target aggregation switch, so that the traffic utilization rate is higher and the distribution is more objective and reasonable. Description of the Drawings

[0038] To more clearly illustrate the technical solutions in the embodiments of the present application, the following briefly introduces the drawings required for the description of the embodiments. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0039] Figure 1 It is a structural block diagram of a traffic distribution system based on redundant switches provided by an embodiment of the present application;

[0040] Figure 2 It is a system architecture diagram of a traffic distribution system based on redundant switches provided by an embodiment of the present application. Specific embodiments

[0041] The following will clearly and completely describe the technical solutions in the embodiments of the present application in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only some embodiments of the present application, rather than all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative efforts fall within the scope of protection of the present application.

[0042] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and do not necessarily need to describe a specific order or sequence. It should be understood that such used data can be interchanged under appropriate circumstances so that the embodiments of the present application described here can be implemented in an order other than those illustrated or described here. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or server including a series of steps or units does not necessarily need to be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products, or devices.

[0043] It should be noted that the following describes various aspects of the embodiments within the scope of the appended claims. It should be obvious that the aspects described herein can be embodied in a wide variety of forms, and any specific structure and / or function described herein is illustrative only. Based on the present application, those skilled in the art should understand that one aspect described herein can be implemented independently of any other aspect, and two or more of these aspects can be combined in various ways. For example, any number of aspects described herein can be used to implement a device and / or practice a method. In addition, this device and / or this method can be implemented using other structures and / or functions in addition to one or more of the aspects described herein.

[0044] Please refer to Figure 1 As shown, an embodiment of the present application provides a traffic allocation system 100 based on redundant switches. The above system includes: an access layer 110, an aggregation layer 120, and a core layer 130; wherein, each access switch in the access layer is connected to several devices to be allocated.

[0045] Specifically, the access layer contains several access switches, each access switch is connected to several devices to be allocated, and IP addresses and traffic are allocated to the corresponding devices to be allocated through each access switch.

[0046] The aggregation layer contains at least two switches; each aggregation switch in the aggregation layer is connected to all the access switches included in the access layer.

[0047] Specifically, the aggregation layer contains at least two aggregation switches, and each aggregation switch is connected to all the access switches included in the access layer; that is, it is used to achieve redundant aggregation switches; under normal circumstances, one aggregation switch is in the working state. When the aggregation switch in the working state fails, any other aggregation switch can be switched to the working state, and all access switches can continue to work normally, ensuring the stable availability of the aggregation layer network.

[0048] In addition, strategies such as VLAN and access control list (ACL) can be configured on the aggregation layer to optimize network performance and enhance security. In this embodiment, the number of switches directly managed by the core switch is reduced, improving the scalability and management efficiency of the network.

[0049] The core layer all contains at least two switches; each core switch in the core layer is connected to all the aggregation switches included in the aggregation layer.

[0050] Specifically, the core layer contains at least two core switches, and each core switch in the core layer is connected to all the aggregation switches included in the aggregation layer, that is, it is used to achieve redundant core switches; under normal circumstances, one core switch is in the working state. When the core switch in the working state fails, any other core switch can be switched to the working state, ensuring the stable availability of the core layer network.

[0051] In this embodiment, the core switch is connected to the aggregation switch through a high-speed link to build a backbone network. The core switch is the key node of the entire network and must have high throughput and low latency to ensure the stable operation of critical services.

[0052] In an exemplary embodiment of the present application, the multi-spanning tree protocol and the virtual router redundancy protocol are used from the access layer to the core layer of the system, wherein: the multi-spanning tree protocol is used to create multiple spanning tree instances; the virtual router redundancy protocol is used to create virtual routers.

[0053] Specifically, from the access layer to the core layer, MSTP (Multiple Spanning Tree Protocol) + VRRP (Virtual Router Redundancy Protocol) is used to improve the network reliability. Through the MSTP protocol, multiple spanning tree instances are created to achieve load balancing among different VLANs, enabling traffic in different VLANs to be forwarded along different paths. At the same time, combined with the VRRP protocol, redundant backup is achieved, ensuring a quick switch to the standby router in case of a failure of the primary router, thus guaranteeing the continuous availability of the network. Multiple backup groups are created in the VRRP protocol, and different masters and backups are specified for each backup group to achieve load balancing of virtual routes. As Figure 2 shown, in this embodiment, for VLAN10, 20, and 30, Core Switch 1 is the master and Core Switch 2 is the backup. For VLAN40 and 50, it is the opposite.

[0054] MSTP ensures that there are no loops in the network and can quickly converge when the network topology changes to guarantee network connectivity and stability. VRRP provides the virtual router redundancy function, enabling a quick switch to the standby router in case of a failure of the primary router, thus ensuring the high availability and connectivity of network devices.

[0055] By optimizing the network topology structure, MSTP can reduce the complexity of spanning tree calculation and improve network performance. In addition, MSTP supports fast convergence and can quickly recalculate the spanning tree when the network topology changes, thereby reducing the impact of network failures on performance. VRRP provides redundancy in the network and can quickly switch to the standby device in case of a failure of the primary device, ensuring network connectivity and stability.

[0056] Configure Instance 1 and Instance 2 under all access switches. Instance 1: Make the primary root of VLAN10, 20, and 30 be on Core Switch 1 and the secondary root be on Core Switch 2. Instance 2: Make the primary root of VLAN40 and 50 be on Core Switch 2 and the secondary root be on Core Switch 1. Configure VRRP on the core switches and configure Eth-Trunk link aggregation to improve the link reliability, so that the normal communication of the network is not affected when a switch or link has problems.

[0057] Configure under all access switches:

[0058] stp region-configuration / / Configure STP region parameters

[0059] region-name ceshi / / Set the region name to "ceshi"

[0060] revision-level 1 / / Set the version number of the configured area to 1

[0061] instance 1 vlan 10 20 30 / / Configure instance 1 to include VLAN 10, 20, and 30

[0062] instance 2 vlan 40 50 / / Configure instance 2 to include VLAN 40 and 50

[0063] active region-configuration / / Activate the STP area configuration

[0064] The configuration of other access switches is the same as this one.

[0065] Aggregation switch configuration:

[0066] interface Vlanif10

[0067] ip address 192.168.10.1 255.255.255.0

[0068] vrrp vrid 1 virtual-ip 192.168.10.254 / / Configure the VRRP group ID as 1 and the virtual IP address as 192.168.10.254

[0069] vrrp vrid 1 priority 120 (The default priority of VRRP is 100. Setting it to 120 makes it the master for VLAN10)

[0070] The configurations of Vlanif20 and 30 are the same by analogy.

[0071] stp region-configuration / / Configure STP area parameters

[0072] region-name ceshi / / Set the area name as "ceshi"

[0073] revision-level 1 / / Set the version number of the area to 1

[0074] instance 1 vlan 10 20 30 / / Configure instance 1 to include VLAN 10, 20, and 30

[0075] instance 2 vlan 40 50 / / Configure instance 2 to include VLAN 40 and 50

[0076] active region-configuration / / Activate STP region configuration

[0077] stp instance 1 root primary / / Set the root of STP instance 1 as the primary root

[0078] stp instance 2 root secondary / / Set the root of STP instance 2 as the secondary root

[0079] interface Eth-Trunk0 (Configure Eth-Trunk link aggregation)

[0080] port link-type trunk / / Configure the port link type as Trunk mode

[0081] port trunk allow-pass vlan 2 to 4094 / / Allow all VLANs passing through Trunk to range from 2 to 4094.

[0082] In an exemplary embodiment of the present application, the link redundancy between switches not only provides fault protection but also is used for load balancing. Traffic is distributed among multiple links, reducing the burden on a single link or switch and improving the efficiency of data transmission.

[0083] In an exemplary embodiment of the present application, an access switch is used to configure an IP address and allocate traffic for a device to be allocated, and transmit the traffic feature vector corresponding to the device to be allocated to a target aggregation switch; wherein, the traffic feature vector of the device to be allocated is determined according to the traffic corresponding to the device to be allocated within a preset time window, the CPU usage rate corresponding to each preset collection time point, and the disk read / write rate; the target aggregation switch is the aggregation switch currently in the working state; the target aggregation switch is any aggregation switch.

[0084] Specifically, the preset time window contains several consecutive preset time periods; the traffic corresponding to the device to be allocated refers to the total traffic used by the device to be allocated within the preset time window, and the traffic refers to the amount of data in network transmission; the CPU usage rate refers to the degree to which the CPU is used per unit time; the disk read / write rate refers to the speed at which the device disk can read or write data per unit time.

[0085] A convergence switch is used to integrate the traffic feature vectors of the devices to be allocated uploaded by all access switches included in the access layer, send the integrated traffic feature vectors to the target core switch, and when receiving the traffic sent down by the target core switch, forward the traffic sent down by the target core switch to each access switch in the access layer; the target core switch is the core switch currently in the working state; the target core switch is any core switch.

[0086] Specifically, the convergence switch integrates the traffic feature vectors uploaded by all access switches in the access layer into a traffic feature vector list, and the traffic feature vector of each device to be allocated is an item in the list, and then uploads the traffic feature vector list to the core switch in the working state in the core layer, so that the core switch will only trigger the data reception and processing process once, greatly reducing the working pressure of the core switch.

[0087] Therefore, in this embodiment, the convergence layer and the core layer have corresponding redundant switch connections, improving the reliability and stability of the networks in the convergence layer and the core layer.

[0088] In the embodiment of the present application, a traffic allocation system based on redundant switches is provided, and the core switch is used to perform the following steps:

[0089] S100, obtain the standardized traffic feature vectors corresponding to the traffic feature vectors of each device to be allocated.

[0090] Specifically, the standardized traffic feature vector standardizes the feature values of each feature in the traffic feature vector, so that the feature values of different magnitudes are converted into the interval of [0, 1].

[0091] S200, obtain the derivative feature vectors corresponding to each standardized traffic feature vector.

[0092] Specifically, the derivative feature vectors include: traffic change rate feature, traffic moving average difference feature, and joint change rate feature.

[0093] S300, according to the traffic feature vectors and derivative feature vectors corresponding to each device to be allocated, obtain the traffic allocation information corresponding to each device to be allocated.

[0094] Specifically, put the traffic feature vectors and derivative feature vectors corresponding to each device to be allocated into the traffic allocation model to obtain the specific traffic volume that each device to be allocated should be allocated.

[0095] S400, send the traffic classification result to the target convergence switch.

[0096] In summary, according to the above steps, traffic is allocated to the devices to be allocated, and the network performance and traffic utilization rate of the constructed redundant switch traffic allocation system are higher.

[0097] In an embodiment of the present application, a traffic allocation system based on redundant switches is provided, and the traffic change rate feature is:

[0098]

[0099] where, is the standardized traffic of the device to be allocated corresponding to the t-th preset collection time point, is the standardized traffic of the device to be allocated corresponding to the (t - 1)-th preset collection time point; avg() is a preset average value determination function; t = 1, 2,..., m; m is the number of preset collection time points included in the preset time window.

[0100] Specifically, based on the standardized data, the traffic change rate feature is obtained. The traffic change rate can more intuitively reflect the relative change trend of traffic on a standard scale, helping the model to more accurately capture the rhythm of traffic changes. For example, when analyzing the fluctuation of network traffic over time, the standardized traffic change rate can be used to judge the severity of traffic changes to plan network resources in advance. By learning this feature, the model can predict the trend of traffic in advance, such as upcoming traffic peaks or troughs. For example, during an e-commerce promotion event, the traffic change rate may suddenly increase. After the model detects this feature change, it can predict that the traffic will grow rapidly, thus providing a lead for network resource scheduling and avoiding network congestion. This accurate grasp of traffic trends greatly improves the accuracy of the model in predicting future traffic states.

[0101] Since it is calculated based on standardized data, this feature eliminates the influence of the original magnitude of traffic data and highlights the relative degree of change. This enables the model to measure the severity of traffic changes with a unified standard in different scenarios. For example, whether in a low-traffic environment of a small network or a high-traffic environment of a large data center, the standardized traffic change rate can accurately reflect the traffic changes. Based on this feature, the model can more accurately evaluate the impact of traffic changes on network performance, thereby improving the accuracy of prediction.

[0102] In an embodiment of the present application, a traffic allocation system based on redundant switches is provided, and the traffic moving average difference feature is:

[0103]

[0104] where, is the moving average value.

[0105] Specifically, the flow movement difference is the flow at the t-th preset collection time point minus the moving average at the t-th preset collection time point, and the flow deviation at the t-th preset collection time point can be obtained.

[0106] The moving average is:

[0107]

[0108] Wherein, is the standardized flow of the device to be allocated corresponding to the i-th preset collection time point, and n is the number of preset collection time points used to calculate the moving average.

[0109] Specifically, the moving average is the average of the flows corresponding to the preset collection time points from i to t, and the specific number is n; as an example: when t is 5 and n = 3, the moving average is the average of the flows corresponding to the 3rd, 4th, and 5th preset collection time points, that is, the average of the flows in the recent several times closer to t.

[0110] This feature can highlight the deviation degree of the current flow relative to the recent average flow, and more clearly show the abnormal fluctuation of the flow under the standard scale. For example, when detecting network traffic anomalies, if exceeds a certain threshold, it may mean that there is an abnormal traffic pattern, which helps to detect network attacks or abnormal business activities in a timely manner.

[0111] The standardized flow moving average difference highlights the deviation degree of the current flow relative to the recent average flow. Under the standard scale, the model can clearly identify the abnormal fluctuation of the flow. For example, when the network is attacked or there are abnormal business activities, the flow will deviate from its normal fluctuation range, and this feature will increase or decrease significantly. By learning this feature, the model can timely detect these abnormal situations and predict possible network problems, such as bandwidth exhaustion and service quality degradation, so as to take countermeasures in advance and improve the accuracy and timeliness of prediction.

[0112] The moving average difference can not only reflect anomalies, but also help the model better understand the stability of the flow trend. Under a stable flow trend, the moving average difference is relatively small and the fluctuation range is narrow; while when the trend changes, the moving average difference will show obvious fluctuations. By analyzing the changes of this feature, the model can more accurately predict whether the flow trend will continue to be stable or is about to turn, thus providing a more reliable prediction basis for the long-term planning of network resources.

[0113] In the embodiment of the present application, a flow allocation system based on redundant switches is provided, and the joint change rate feature is:

[0114]

[0115] Among them, is the combined change rate feature, is the weight of the disk read / write rate change; is the weight of the CPU usage rate change, is the change rate of the disk read / write rate, is the change rate of the CPU usage.

[0116] Specifically, the combined change rate is determined by the change rate of the CPU usage rate and the change rate of the disk read / write rate; the higher the change rate of the CPU usage rate, the higher the combined change rate; the higher the change rate of the disk read / write rate, the higher the combined change rate. represents the importance degree of the disk read / write rate when determining the combined change rate; represents the importance degree of the CPU usage rate when determining the combined change rate.

[0117] is:

[0118]

[0119] Among them, is the disk read / write rate corresponding to the t-th preset acquisition time point, is the disk read / write rate corresponding to the (t - 1)-th preset acquisition time point.

[0120] Specifically, the change rate of the disk read / write rate at the t-th preset acquisition time point is positively correlated with the change of the disk read / write rate at the t-th preset acquisition time point. The greater the change of the disk read / write rate, the greater the change rate of the disk read / write rate; taking an average of the change rates of the disk read / write rates at all preset acquisition time points can obtain the average change rate of the disk read / write rate within the preset time window.

[0121] is:

[0122]

[0123] Among them, is the CPU usage rate corresponding to the t-th preset acquisition time point, is the CPU usage rate corresponding to the (t - 1)-th preset acquisition time point.

[0124] Specifically, the change rate of the CPU usage rate at the t-th preset acquisition time point is positively correlated with the change of the CPU usage rate at the t-th preset acquisition time point. The greater the change of the CPU usage rate, the greater the change rate of the CPU usage rate; taking an average of the change rates of the CPU usage rates at all preset acquisition time points can obtain the average change rate of the CPU usage rate within the preset time window.

[0125] This combined change rate feature comprehensively reflects the dynamic change relationship between disk I / O and CPU usage under the standard scale, which helps the model better understand the resource requirement changes of the server during data reading, writing, and processing, thereby more accurately predicting the server performance changes and providing a basis for resource scheduling.

[0126] In an embodiment of the present application, a traffic distribution system based on redundant switches is provided. If the target aggregation switch fails, any aggregation switch other than the target aggregation switch within the aggregation layer is switched to the working state.

[0127] Specifically, when the target aggregation switch fails, any aggregation switch other than the target aggregation switch can continue to complete the work, ensuring the stable and reliable operation of the aggregation layer network.

[0128] In an embodiment of the present application, a traffic distribution system based on redundant switches is provided. If the target core switch fails, any core switch other than the target core switch within the core layer is switched to the working state.

[0129] Specifically, when the target core switch fails, any core switch other than the target core switch can continue to complete the work, ensuring the stable and reliable operation of the core layer network.

[0130] In an embodiment of the present application, a traffic distribution system based on redundant switches is provided. Any two aggregation switches or any two core switches are connected by at least one link.

[0131] Specifically, a link is added directly between two core switches to complete the communication between the two core switches. By sending messages to each other between the two core switches within a preset time interval, it is confirmed whether the other core switch is faulty, thereby realizing the redundancy of the switches.

[0132] The program code for performing the operations of the present application can be written in any combination of one or more programming languages. The programming languages include object-oriented programming languages such as Java, C++, etc., and also include conventional procedural programming languages such as the "C" language or similar programming languages. The program code can be executed entirely on the user computing device, partially on the user device, executed as an independent software package, partially on the user computing device and partially on a remote computing device, or entirely on a remote computing device or server. In the case of a remote computing device, the remote computing device can be connected to the user computing device through any type of network, including a local area network (LAN) or a wide area network (WAN), or can be connected to an external computing device (for example, by using an Internet service provider to connect through the Internet).

[0133] In addition, the above-mentioned drawings are only schematic illustrations of the processes included in the method according to the exemplary embodiments of the present application, rather than for limiting purposes. It is easy to understand that the processes shown in the above-mentioned drawings do not indicate or limit the chronological order of these processes. Additionally, it is also easy to understand that these processes can be executed synchronously or asynchronously in, for example, multiple modules.

[0134] It should be noted that although several modules or units of the device for action execution are mentioned in the above detailed description, such a division is not mandatory. In fact, according to the embodiments of the present application, the features and functions of two or more of the above-mentioned modules or units can be embodied in one module or unit. Conversely, the features and functions of one module or unit described above can be further divided and embodied by multiple modules or units.

[0135] The above is only the specific embodiment of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions that can be easily thought of by those skilled in the art within the technical scope disclosed in the present application should be covered by the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.

Claims

1. A traffic distribution system based on redundant switches, characterized in that: The system comprises an access layer, an aggregation layer and a core layer, wherein the aggregation layer and the core layer each comprise at least two switches; each access switch in the access layer is connected to a number of devices to be allocated; each aggregation switch in the aggregation layer is connected to all access switches in the access layer; each core switch in the core layer is connected to all aggregation switches in the aggregation layer; The access switch is used to configure IP addresses and allocate traffic for the devices to be allocated, and transmit the traffic feature vectors corresponding to the devices to be allocated to the target aggregation switch; wherein the traffic feature vectors of the devices to be allocated are determined according to the traffic corresponding to the devices to be allocated within a preset time window, the CPU usage rate and the disk read / write rate corresponding to each preset collection time point; the target aggregation switch is the aggregation switch currently in working state; the target aggregation switch is any aggregation switch; The aggregation switch is used to integrate the traffic feature vectors of the devices to be allocated uploaded by all access switches included in the access layer, and send the integrated traffic feature vectors to the target core switch, and when receiving the traffic sent by the target core switch, forward the traffic sent by the target core switch to each access switch of the access layer; the target core switch is the core switch currently in working state; the target core switch is any core switch; The core switch is used to obtain the integrated traffic feature vector sent by the target aggregation switch, generate traffic distribution information according to the traffic feature vector of each device to be distributed, and send the traffic distribution information to the target aggregation switch.

2. The traffic distribution system based on redundant switches according to claim 1 is characterized in that: The access layer to the core layer of the system uses the Multiple Spanning Tree Protocol and the Virtual Routing Redundancy Protocol, where: Multiple Spanning Tree Protocol, used to create multiple spanning tree instances; Virtual Router Redundancy Protocol, used to create virtual routers.

3. The traffic distribution system based on redundant switches according to claim 1 is characterized in that: The core switch is used to perform the following steps: Obtain a standardized flow characteristic vector corresponding to the flow characteristic vector of each device to be allocated; Obtain a derived feature vector corresponding to each standardized flow feature vector; Obtaining flow distribution information corresponding to each device to be distributed according to the flow characteristic vector and the derived characteristic vector corresponding to each device to be distributed; The traffic classification information is sent to the target aggregation switch.

4. The traffic distribution system based on redundant switches according to claim 3 is characterized in that: The derived feature vector includes a flow rate change feature, a flow moving average difference feature and a joint change rate feature.

5. The traffic distribution system based on redundant switches according to claim 4 is characterized in that: The flow rate change characteristic for: in, is the standardized device flow to be allocated corresponding to the t-th preset collection time point, is the standardized device flow to be allocated corresponding to the t-1th preset collection time point; avg() is the preset average value determination function; t=1, 2, ..., m; m is the number of preset collection time points included in the preset time window.

6. The traffic distribution system based on redundant switches according to claim 4 is characterized in that: The flow moving average difference characteristic for: in, is a moving average, and the moving average is: in, is the standardized device flow to be allocated corresponding to the i-th preset collection time point, and n is the number of preset collection time points used to calculate the moving average.

7. The traffic distribution system based on redundant switches according to claim 4 is characterized in that: The combined rate of change characteristic for: in, is the joint rate of change feature, Change weights for disk read and write rates; The weight for CPU usage change, is the disk read and write rate change rate, for: in, is the disk read / write rate corresponding to the tth preset collection time point, The disk read / write rate corresponding to the t-1th preset acquisition time point; is the CPU usage change rate, for: in, is the CPU usage corresponding to the tth preset collection time point, It is the CPU usage corresponding to the t-1th preset collection time point.

8. The traffic distribution system based on redundant switches according to claim 1, characterized in that: If the target aggregation switch fails, any aggregation switch in the aggregation layer except the target aggregation switch will be switched to a working state.

9. The traffic distribution system based on redundant switches according to claim 1, characterized in that: If the target core switch fails, any core switch in the core layer except the target core switch will be switched to a working state.

10. The traffic distribution system based on redundant switches according to claim 1, characterized in that: Any two aggregation switches or any two core switches are connected through at least one link.

Images