A performance optimization method for large-scale service network architecture

Abstract

The present invention discloses a performance optimization method for a large-scale service network architecture, including: constructing a tree network based on a service routing node as a backbone node and expanding service switching nodes outward; constructing one or more minimum spanning tree multicast networks based on the network communication cost between the service routing nodes, the node with the smallest total network communication cost with other service routing nodes in the minimum spanning tree multicast network is used as a controller node, and the x service routing nodes adjacent to each service routing node are used as the dynamic neighbor node set of the service routing node, the controller node updates the dynamic neighbor node according to regular network communication cost detection, and ensures that the total network communication cost of the minimum spanning tree multicast network is minimized; the tree network performs hierarchical adjustment on all minimum spanning tree multicast networks based on a threshold. The method provided by the present invention realizes the de-neutralization of the service network, thereby improving the overall availability and efficiency of the service network.

Description

Technical Field

[0001] The present invention belongs to the technical field of service networks, and in particular relates to a performance optimization method for a large-scale service network architecture. Background Art

[0002] The service network was first proposed by the European Union in the 7th Framework Program (FP7) in 2007. With the gradual maturity of the Internet of Everything and Everything as a Service, the interconnection between massive services has formed a service network. The service network provides infrastructure support for the interconnection between massive services. Each service entity is marked and published in the service network and exists in the form of mutual association or partial combination. Service consumers can discover various services in the service network through the retrieval mechanism, and can combine or aggregate services according to needs to form more complex service solutions. With the increasing complexity of user needs in the Internet era, the services provided by a single organization are far from meeting the needs, and often require different service providers in multiple industries and regions to provide collaborative services to meet the needs. Therefore, the service network is actually the next development direction of the Internet and the future network infrastructure.

[0003] Currently, the world's largest centralized service network ProgrammableWeb.com contains more than 24,000 APIs and more than 8,000 combined services, covering more than 400 different service types. This architecture generally has a single service registration center, where service providers register service information according to certain requirements, and service users query the service registration center for target services that meet the requirements, and then make subsequent service calls through the access methods provided by the service.

[0004] As an early form of service network, the centralized architecture achieved good results at the time. However, as the number of services on the network increases and their scale becomes larger, the architecture has the following two problems in such a scenario:

[0005] 1. Single point of failure problem, that is, if there is a problem with the service registration center, the entire service network will be paralyzed. Especially in large-scale scenarios, intensive client requests will greatly increase the probability of single point of failure, thereby seriously reducing the availability of the entire service network.

[0006] 2. Performance bottleneck problem: The request processing performance of a single service registration center is limited. When the number of services is very large, the request processing capacity of the service registration center will drop significantly, and the service information access performance will also be difficult to support.

[0007] In a distributed service network, service information is generally stored in multiple directory entities, so there is no single point of failure. Its typical underlying implementation is a P2P network. Since unstructured P2P networks often use methods such as limiting the number of hops, flooding, or random walks for routing and forwarding, which will occupy a large amount of network bandwidth, researchers generally use structured P2P networks to implement service networks, such as Chord. There are many service network implementations using Chord, and their underlying principles are basically similar, such as the literature (He Q, Yan J, Yang Y, et al. A decentralized service discovery approach on peer-to-peer networks [J]. IEEE Transactions on Services Computing, 2011, 6 (1): 64-75.). In this network, service providers register service information in a global hash table, and service users can accurately locate the service by hashing the service information they need to find. Although this network improves routing efficiency, it still has some defects in large-scale scenarios:

[0008] 1 When nodes join and leave, the relationship between service information and storage addresses in the network needs to be maintained, so the addresses need to be re-hashed and all service information in the mobile network needs to be re-allocated, which brings a lot of network load and reduces network availability. In large-scale service scenarios, nodes join and leave very frequently, so this is a very serious problem.

[0009] 2 Since service information will be stored in network nodes that are not provided by the service provider, there is a problem of information tampering, and the information security in the network cannot be guaranteed.

[0010] Academic Literature Research on Key Technologies of Cross-border Service Networks [D]. Zhejiang University, 2019. An ExtendNet network is disclosed, which can be imaged as a tree structure. Its leaf nodes are service exchange nodes, which connect to service providers to open services. The internal nodes are service routing nodes, which are responsible for message routing. The top layer of the network is a super node, which is responsible for storing service information. The top layer of ExtendNet forms a structured P2P network through the DHT protocol, and the service information is distributed and stored in the top-level network node cluster. ExtendNet solves the problems of cross-border service integration, combination, retrieval, routing, etc. Summary of the invention

[0011] In order to solve the above problems, the present invention provides a performance optimization method for a large-scale service network architecture, which can solve the single point failure and performance bottleneck problems existing in the centralized architecture. At the same time, decentralization also solves the information security problems existing in the traditional distributed architecture.

[0012] A performance optimization method for a large-scale service network architecture, comprising:

[0013] A tree network is constructed based on a service routing node as a backbone node and an outwardly extending service switching node, wherein the service information of the tree network is stored in the service switching node.

[0014] Based on the network communication cost between service routing nodes, one or more minimum spanning tree multicast networks are constructed. The node with the smallest total network communication cost with other service routing nodes in the minimum spanning tree multicast network is used as a controller node, and the x service routing nodes adjacent to each service routing node are used as a dynamic neighbor node set of the service routing node. The controller node updates the dynamic neighbor nodes based on regular network communication cost detection to ensure that the total network communication cost of the minimum spanning tree multicast network is minimized.

[0015] The tree network performs hierarchical adjustment on all minimum spanning tree multicast networks based on a threshold value, wherein the threshold value includes a maximum number of nodes and a minimum number of nodes:

[0016] When the number of service routing nodes in the minimum spanning tree multicast network is greater than the maximum number of nodes, two service routing nodes with the largest communication cost in the minimum spanning tree multicast network are used as base points, and the network communication costs between other service routing nodes and the two base points are used to re-divide the network to obtain two new minimum spanning tree multicast networks.

[0017] When the number of child nodes managed by the service routing node in the minimum spanning tree multicast network is greater than the maximum number of nodes, the service routing node is replicated, and half of the service switching nodes originally carried are allocated to the replicated new nodes.

[0018] When the sum of the node sizes in all minimum spanning tree multicast networks managed by a service routing node is less than the minimum number of nodes, the service routing node will be merged with the nearest neighbor service routing node.

[0019] The present invention constructs a tree network framework and reduces network bandwidth and network load based on a minimum spanning tree multicast network with minimum dynamic delay, thereby achieving de-neutralization of the service network and improving the overall availability and efficiency of the service network.

[0020] Preferably, the adjacent node selects the first x nodes with the smallest network communication cost based on the network communication cost between the service routing node and other service routing nodes.

[0021] Specifically, the x is 10.

[0022] Specifically, the controller node detects the network communication cost periodically, and the specific process is as follows:

[0023] Regularly detect the network communication cost between the service routing node and the nodes in the dynamic neighbor node set, and randomly detect the network communication cost between the service routing node and a non-dynamic neighbor node;

[0024] If the network communication cost with the non-dynamic neighbor node is lower than any node in the dynamic neighbor node set, the non-dynamic neighbor node is used as a replacement, and the replaced dynamic neighbor node set and the corresponding network communication cost are sent to the controller node;

[0025] If the network communication cost with the non-dynamic neighbor node is higher than any node in the dynamic neighbor node set, the dynamic neighbor node set and the corresponding network communication cost are directly sent to the controller node;

[0026] The controller node uses the collected network communication cost data of all dynamic neighbor nodes to establish an undirected weighted graph representing the network, and calculates the corresponding minimum spanning tree;

[0027] A judgment is made based on the difference between the total weight of the minimum spanning tree and the total weight of the previous detection. When it exceeds a stable value, the multicast routing table of each node in the minimum spanning tree multicast network is updated based on the dynamic neighbor node set replaced in this detection process.

[0028] Specifically, the maximum number of nodes is 6400 network nodes.

[0029] Specifically, the minimum number of nodes is 3200 network nodes.

[0030] Specifically, the re-dividing is performed based on the network communication cost between other service routing nodes and the two base points, and the specific process is as follows:

[0031] According to the network communication cost between the service routing node and the two base points, the service routing node is divided to the base point with the smaller network communication cost, until all the service routing nodes are divided, and two new minimum spanning tree multicast networks are obtained;

[0032] When the minimum spanning tree multicast network before re-division has no father node, a new father node is created for the two newly divided minimum spanning tree multicast networks, thereby increasing the total network level by 1.

[0033] Specifically, the process of merging the service routing node with the nearest neighbor service routing node is as follows:

[0034] Merge the service routing node with other nearest service routing nodes in the minimum spanning tree multicast network, and merge the routing table and other information;

[0035] When the minimum spanning tree multicast network where the service routing node is located is the top layer of the tree network and the scale is merged to 1, the top layer is deleted and the nodes in the top layer are merged with the next layer to obtain a new top layer, thereby reducing the total network level by 1.

[0036] Compared with the prior art, the present invention has the following beneficial effects:

[0037] The decentralized hierarchical service network constructed by the performance optimization method provided by the present invention can provide a high-performance service network architecture for large-scale service scenarios. The decentralized characteristics of the service network well solve the single point failure and performance bottleneck problems existing in the centralized architecture, and decentralization also solves the information security problems existing in the traditional distributed architecture. The high scalability of the service network also solves the problem of large-scale service information migration existing in the traditional distributed service network. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 A schematic diagram of the tree network framework provided in this embodiment;

[0039] Figure 2 A degradation degree diagram of the minimum spanning tree multicast network provided in this embodiment;

[0040] Figure 3 This is a performance comparison chart of the tree network provided in this embodiment and the Chord service-based network. DETAILED DESCRIPTION

[0041] The present invention is further described below in conjunction with specific examples. It should be understood that these examples are only to help those of ordinary skill in the art understand the principles and knowledge of the present invention, and are not intended to limit the scope of the present invention, and cannot be considered to limit the application scenarios of the present invention. In addition, it should be understood that after reading the content taught by the present invention, those skilled in the art may make various changes or modifications to the present invention, but the deformation, changes and conversions made to the embodiments based on the principles and purposes of the present invention also fall within the scope defined by the claims attached to this application. And it is obvious that this specification is only exemplified by preferred embodiments, and there is no need to exhaust all embodiments.

[0042] like Figure 1 As shown, a tree network based on service routing nodes as backbone nodes and service switching nodes expanding outward.

[0043] Service routing nodes form the backbone of the network and have the following functions and features:

[0044] 1. When the service routing node is located at the second-to-last layer in the network, it is responsible for access control of multiple service switching nodes, and the number of switching nodes accessed by a single routing node does not exceed a certain threshold.

[0045] 2. Manage the routing nodes or switching nodes of the lower layer, control the joining and exit of the lower layer nodes, and ensure that the number of lower layer nodes that a single routing node can access does not exceed a certain threshold. Generally speaking, a routing node contains an abbreviated representation or index of all services of the lower layer nodes.

[0046] 3. Responsible for service routing, service discovery, service caching, service traffic monitoring and other functions, forming the core skeleton of the entire service network.

[0047] 4. Dynamically perceive changes in single-domain network resources, provide load balancing functions, and optimize routing paths in real time.

[0048] Service exchange node, which has the following functions and features:

[0049] 1. As a network edge node, it is responsible for the access of service providers, generally enterprises, developers, etc. Service providers directly register services on the service exchange node and store them through the built-in database. Therefore, the service exchange node is responsible for the maintenance and update of service information.

[0050] 2. The service exchange node is also responsible for processing service query requests and searching and matching in the internal database according to the request information. Since the service scale of each service provider is not too large, the service search efficiency at the service exchange node is higher.

[0051] 3. It assumes the function of decentralization in the service network. All service information is stored at the service provider at the edge of the network, thus avoiding the problems caused by centralized storage. The localized storage of services also saves the cost of copying service information in the network. The cost of registering, updating and deleting service information is lower, so the network overhead caused by information maintenance is very small.

[0052] The tree network contains one or more minimum spanning tree multicast networks, which construct an undirected weighted graph representing the network by real-time monitoring of the network communication cost between nodes, where the edge weight is the node communication cost, and the communication cost is represented by the round-trip delay of direct network communication between two network nodes at the application layer.

[0053] By calculating the minimum spanning tree of the undirected weighted graph in real time, the network can update the optimal multicast routing table for each node, which ensures that the total network load caused by multicast is minimized. In an ideal minimum spanning tree multicast network, each network node needs to maintain the communication cost from the node to all other nodes in the network.

[0054] In a small-scale network, such as a network with less than 100 nodes, maintaining this data will not cause too much network load. However, when the network scale becomes larger, since the time complexity of each real-time update of this data is O(n^2), where n represents all nodes in the network, frequent data updates will cause serious bandwidth occupancy and network load.

[0055] When the network scale used in this embodiment is less than 6400 nodes, only 10 neighbor nodes need to be maintained to form a global optimal minimum spanning tree, and at this time, there is no degradation in the total network delay.

[0056] The construction process of the minimum spanning tree multicast network is as follows:

[0057] The node with the smallest total communication cost with other nodes in the network is selected as the controller node in the multicast network.

[0058] Network nodes periodically detect and update the network communication costs with all neighboring nodes.

[0059] The network node periodically randomly selects a node other than the neighbor node to detect the network communication cost. If the communication cost is lower than any neighbor, the neighbor is replaced.

[0060] The network node sends the updated neighbor communication cost data to the controller node.

[0061] The controller node uses the collected network communication cost data of all nodes to establish an undirected weighted graph representing the network and calculates the minimum spanning tree of the graph.

[0062] The controller node determines the difference between the total weight of the minimum spanning tree and the last calculated result. If it exceeds a certain threshold, it sends an instruction to update the routing table of all network nodes.

[0063] In this construction process, the controller node improves the stability of the network by setting the update threshold and minimizes the network bandwidth occupation caused by the routing table update. It is worth noting that since the minimum spanning tree caused by maintaining a constant level of neighbor network communication cost does not guarantee to reach the global optimum, the network node needs to randomly select other nodes in the network for network delay testing at regular intervals. If the communication cost is less than the existing neighbor, it will be replaced. Through this measure, the minimum spanning tree network can continuously approach the global optimum.

[0064] In general, the dynamic delay-based minimum spanning tree multicast network brings the following benefits: Compared with the flooding, random walk, hop limit and other methods used in traditional unstructured networks, the minimum spanning tree structure greatly reduces the network bandwidth usage and the network load, thereby improving the overall availability and efficiency of the service network. Compared with the ordinary minimum spanning tree multicast network, the nodes in this network only need to maintain the constant level of neighbor communication cost data and can gradually approach the global optimum, thus further reducing the network load and supporting medium and large network scales. Compared with IP layer multicast, the application layer multicast network is easy to deploy and can ignore the heterogeneity of the underlying network.

[0065] The de-neutralization function of the tree network achieves a dynamic balance of network level increase and decrease through the single domain splitting algorithm, routing node splitting algorithm and network level reduction algorithm. This is a process based on passive triggering.

[0066] Single domain splitting algorithm:

[0067] 1. When the controller node in the minimum spanning tree multicast network determines that the single domain scale exceeds the maximum number of nodes, it proceeds to the next step, otherwise it continues to monitor the minimum spanning tree multicast network;

[0068] 2. Select the two nodes node1 and node2 that are the farthest apart in the minimum spanning tree multicast network;

[0069] 3. Two new minimum spanning tree multicast networks are established with node1 and node2 as the central nodes. Other nodes closest to the central node of the new minimum spanning tree multicast network in the old minimum spanning tree multicast network are selected to join the new minimum spanning tree multicast network respectively until all nodes in the old minimum spanning tree multicast network are selected.

[0070] 4. The controller node in the two newly formed minimum spanning tree multicast networks selects the service routing node with the smallest sum of distances to all nodes;

[0071] 5. If the minimum spanning tree multicast network before the split has no father node, then a new father node is created, whose child nodes are the two minimum spanning tree multicast networks, and the total network level increases by 1.

[0072] Routing node splitting algorithm:

[0073] 1. The node continuously monitors the number of child nodes. When the number of child nodes exceeds the threshold c2, it proceeds to the next step;

[0074] 2. The node copies its own running container;

[0075] 3. Use the minimum spanning tree multicast network splitting algorithm to evenly distribute the original child nodes to the new nodes. After the distribution is completed, the two nodes each manage half of the original child nodes.

[0076] Network level reduction algorithm:

[0077] 1. The node continuously monitors the size of each minimum spanning tree multicast network it manages, automatically merges the smaller minimum spanning tree multicast networks, and maintains the size of the minimum spanning tree multicast network at a medium level. If the sum of the sizes of all minimum spanning tree multicast networks is less than 3200, proceed to the next step.

[0078] 2. The node merges with other service routing nodes in the minimum spanning tree multicast network that are the nearest neighbors, and merges the routing table and other information. If the minimum spanning tree multicast network where the node is located is the top layer and the scale has been reduced to 1, proceed to the next step.

[0079] 3. The service network deletes the top layer, and the top layer nodes are merged with the second layer in the network into a minimum spanning tree multicast network as the new top layer of the network. The total number of network layers is reduced by 1.

[0080] The service network routing strategy adopted by the tree network is as follows:

[0081] Service routing in a service network refers to the process in which, after a user initiates a service call request, the network uses a certain strategy to forward the request until it reaches the service entity. On the one hand, service routing allows users to not care about the specific address of the service provider, thereby having sufficient flexibility to optimize request forwarding. On the other hand, service routing enables the service network to have the ability to supervise service traffic, thereby providing trust mechanisms and traffic prediction functions on this basis.

[0082] A domain name is assigned to all service routing nodes, service switching nodes and service entities in the tree-like network. The subordinate relationship between nodes in the global domain name tree is consistent with the service network. Only the network nodes are replaced with the domain names corresponding to the nodes. In addition, the domain name tree has an additional layer of the network bottom layer composed of service domain names.

[0083] The global domain name tree needs to meet the following conditions: for any node in the network i , node j and node k For node i The lower-level nodes managed, and i≠j, there is node j .name≠node k .name is valid, where node.name represents the domain name of the node.

[0084] The global domain name tree needs to maintain three data at each node in the service network: prefix domain name prefix, father node address node.father and child domain name hash table node.children. Among them, node.children maps the domain names of all child nodes to the corresponding network addresses, which is implemented using a hash table. The prefix domain name is a data structure used to accelerate service routing. It is defined as starting from the top node, finding the shortest path to the current network node, and connecting the domain names of each node on the path to form a string.

[0085] Based on the domain name tree, the service entity ID can be constructed by the following algorithm:

[0086] 1. Set the starting point to the service exchange node where the service entity is located, and let id = node.name;

[0087] 2. Determine whether node.father is empty. If it is empty, it means that the top layer has been reached. At this time, the service entity ID is constructed and the id is returned. Otherwise, go to step 3.

[0088] 3. Add the domain name of the current network node to the id header and connect the two using the '.^' character;

[0089] 4. Let node = node.father and return to step 2.

[0090] Based on the global domain name tree, the service user gives the destination service ID, and the corresponding service routing algorithm is as follows:

[0091] 1. Set the starting point to the service exchange node where the service user is located, and the destination service id is represented as id;

[0092] 2. Use '.^' as the separator to split the id to form a string list names with a length of n;

[0093] 3. Match names with the current node domain name prefix. If the match is successful, go to step 5, otherwise go to step 4;

[0094] 4. If the current node has a father node, then set node = node.father, move to the father node, and go to step 3; otherwise, it means that the current node is at the top layer of the service network. At this time, the routing request is distributed through the minimum spanning tree multicast network, and the service routing node with a matching domain name prefix is ​​found. Node is set to this node, and then go to step 5;

[0095] 5. Match names with the domain name of the child node. If the match is successful, enter the child node and repeat step 5. If there is no child node, it means that the destination service exchange node has been reached.

[0096] like Figure 2 As shown, the ratio of the growth of the total network delay of the sub-minimum spanning tree based on the constant neighbor table is referred to as the degradation degree in this embodiment, where the neighbor table is the neighbor table with the minimum communication cost, and the neighbor table is random when the service network is initialized.

[0097] As can be seen in Figure (a), as the number of neighbor nodes gradually increases, the degradation of the total network delay gradually decreases. When the number of neighbor nodes is 10, the total network delay has almost no degradation. Therefore, when the node scale is less than 6400, the network nodes in the minimum spanning tree multicast network only need to maintain the network communication cost data of 10 neighbors to ensure the minimum total network delay.

[0098] Figure (b) shows that when the number of neighbor nodes is 10, as time continues to grow, the single domain can randomly select network nodes to detect communication costs, thereby continuously optimizing the random neighbor table and eventually converging to the optimal state.

[0099] like Figure 3 As shown in the figure, the horizontal axis represents the number of nodes contained in the service network, and the vertical axis represents the average length of the routing path. It can be observed from the figure that when the number of network nodes reaches 16,000, the average length of the routing path for unicast operations in the Chord service network is 6.729, and the average length of the routing path for unicast operations in the decentralized hierarchical service network is 3.008. The latter reduces the routing path length by 55.3% compared with the former. Therefore, the decentralized hierarchical service network proposed in this paper effectively optimizes the performance of unicast operations in large-scale scenarios.

Claims

1. A performance optimization method for a large-scale service network architecture, characterized in that: include: Constructing a tree network based on a service routing node as a backbone node and expanding outwardly a service exchange node, wherein service information of the tree network is stored in the service exchange node; Based on the network communication cost between service routing nodes, one or more minimum spanning tree multicast networks are constructed, wherein the node with the smallest total network communication cost with other service routing nodes in the minimum spanning tree multicast network is used as a controller node, and the x service routing nodes adjacent to each service routing node are used as a dynamic neighbor node set of the service routing node, and the controller node updates the dynamic neighbor nodes according to regular network communication cost detection to ensure that the total network communication cost of the minimum spanning tree multicast network is minimized; The tree network performs hierarchical adjustment on all minimum spanning tree multicast networks based on a threshold value, wherein the threshold value includes a maximum number of nodes and a minimum number of nodes: When the number of service routing nodes in the minimum spanning tree multicast network is greater than the maximum number of nodes, two service routing nodes with the largest communication cost in the minimum spanning tree multicast network are used as base points, and the network communication costs between other service routing nodes and the two base points are used to re-divide the network to obtain two new minimum spanning tree multicast networks; When the number of child nodes managed by the service routing node in the minimum spanning tree multicast network is greater than the maximum number of nodes, the service routing node is replicated, and half of the service switching nodes originally carried are allocated to the replicated new nodes; When the sum of the node sizes in all minimum spanning tree multicast networks managed by a service routing node is less than the minimum number of nodes, the service routing node will be merged with the nearest neighbor service routing node.

2. The performance optimization method of a large-scale service network architecture according to claim 1, characterized in that: The neighboring node selects the first x nodes with the smallest network communication cost based on the network communication cost between the service routing node and other service routing nodes.

3. The performance optimization method of a large-scale service network architecture according to claim 1 or 2, characterized in that: The x is 10.

4. The performance optimization method of a large-scale service network architecture according to claim 1, characterized in that: The controller node detects the network communication cost periodically, and the specific process is as follows: Regularly detect the network communication cost between the service routing node and the nodes in the dynamic neighbor node set, and randomly detect the network communication cost between the service routing node and a non-dynamic neighbor node; If the network communication cost with the non-dynamic neighbor node is lower than any node in the dynamic neighbor node set, the non-dynamic neighbor node is used as a replacement, and the replaced dynamic neighbor node set and the corresponding network communication cost are sent to the controller node; If the network communication cost with the non-dynamic neighbor node is higher than any node in the dynamic neighbor node set, the dynamic neighbor node set and the corresponding network communication cost are directly sent to the controller node; The controller node uses the collected network communication cost data of all dynamic neighbor nodes to establish an undirected weighted graph representing the network, and calculates the corresponding minimum spanning tree; A judgment is made based on the difference between the total weight of the minimum spanning tree and the total weight of the previous detection. When it exceeds a stable value, the multicast routing table of each node in the minimum spanning tree multicast network is updated based on the dynamic neighbor node set replaced in this detection process.

5. The performance optimization method of a large-scale service network architecture according to claim 1, characterized in that: The maximum number of nodes is 6400 network nodes.

6. The performance optimization method of a large-scale service network architecture according to claim 1, characterized in that: The minimum number of nodes is 3200 network nodes.

7. The performance optimization method of a large-scale service network architecture according to claim 1, characterized in that: The specific process of re-dividing the network communication cost between other service routing nodes and two base points is as follows: According to the network communication cost between the service routing node and the two base points, the service routing node is divided to the base point with the smaller network communication cost, until all the service routing nodes are divided, and two new minimum spanning tree multicast networks are obtained; When the minimum spanning tree multicast network before re-division has no father node, a new father node is created for the two newly divided minimum spanning tree multicast networks, thereby increasing the total network level by 1.

8. The performance optimization method for a large-scale service network architecture according to claim 1, characterized in that: The specific process of merging the service routing node with the nearest neighbor service routing node is as follows: Merge the service routing node with other nearest service routing nodes in the minimum spanning tree multicast network, and merge the routing table and other information; When the minimum spanning tree multicast network where the service routing node is located is the top layer of the tree network and the scale is merged to 1, the top layer is deleted and the nodes in the top layer are merged with the next layer to obtain a new top layer, thereby reducing the total network level by 1.

Images