Hierarchical response ranking algorithm for app peak access system and method

Abstract

This invention provides a hierarchical response level algorithm-based system and method for peak-hour app access, including an app client, a P-zone (user access point), and an app backend server. When the app client's access volume reaches the peak capacity of the app backend server, a hierarchical algorithm links the app client's access to the P-zone, which is linked to the app backend server. In the P-zone, a leveling algorithm classifies new access links according to different user factors, determines the priority level for processing by the app backend server, and writes it into the P-zone database. The hierarchical and leveling algorithms in this invention fully consider factors such as user membership level, app level, and network speed to prioritize users. When data access volume is high, higher-level users can access services first, providing a better user experience and increasing user stickiness and app acceptance.

Description

Technical Field

[0001] This invention relates to the field of computer technology, and more specifically, to an APP peak access system and method based on a hierarchical response level algorithm. Background Technology

[0002] With the widespread adoption of mobile apps, people can quickly and effectively solve many business problems through these apps, greatly optimizing business processing efficiency. However, as business processing efficiency improves, the number of access links that app servers need to handle is also increasing, often leading to server overload and a worse service experience for customers. Therefore, how to effectively handle customer connections even when access data reaches server peak levels, without affecting basic business needs, and provide a better user experience for customers, especially high-level users, to improve user stickiness and recognition of the app, is a problem that many market players using apps are looking to solve.

[0003] Patent document CN107888509A discloses a method for exam registration that avoids server congestion. This method reduces server-side congestion and ensures a smooth registration process by forwarding excessive data packets to a virtual storage server and then to the user via a proxy server. However, this method essentially involves setting up a separate storage server and server to address data overload. This significantly increases costs and leads to severe resource idleness and waste when data is not overloaded. Furthermore, because the virtual storage server has no processing restrictions and can directly transmit processed data through a proxy server, data security cannot be guaranteed, contradicting users' requirements for data confidentiality. Summary of the Invention

[0004] To address the shortcomings of existing technologies, the purpose of this invention is to provide a hierarchical response level algorithm for APP peak access system and method. The hierarchical algorithm and the level algorithm fully consider factors such as user membership level, APP level, and network speed to prioritize users. When the data access volume is high, users with higher levels can access services first, which provides a better user experience for high-level users, improves user stickiness and APP recognition.

[0005] A hierarchical response level algorithm-based APP peak access system according to the present invention includes an APP client, a P zone, and an APP backend server.

[0006] When the number of visits from the APP client reaches the peak capacity of the APP backend server, the access from the APP client is linked to the P area which is connected to the APP backend server through a layered algorithm.

[0007] In Zone P, access links from the APP client are classified into different levels based on different factors of the APP client user information using a ranking algorithm. Then, the priority level of the APP backend server is determined according to the ranking, and the priority level data is written into the Zone P database.

[0008] Furthermore, the service functions of the P zone are triggered by the APP backend server when the access volume of the APP client reaches the peak capacity of the APP backend server; the data access between the APP client and the P zone is achieved without difference through a layered algorithm.

[0009] Furthermore, the service functions of the P zone include taking over the access links of the APP client after the P zone is triggered, providing login authentication services, providing basic page display services for its own cache, and providing basic browsing functions, which can avoid occupying the process of the APP background server.

[0010] Furthermore, the different factors of the APP client user information include APP user level, APP membership level, network speed, and response time.

[0011] Furthermore, the ranking algorithm reads the user level x and membership level y from the APP user information in the P zone; according to the network speed requirements of different APPs, it divides the network speed of the user accessing the P zone into network speed levels, and takes the corresponding network speed level value as the network speed value z for the user accessing the P zone based on the read network speed, and sets the network speed coefficient C; the P zone starts timing from the time the user establishes a connection with the P zone, which is the user's waiting time t, and the time coefficient M is determined based on the waiting time t; when the P zone is activated, the influence of each user's real-time user level x and membership level y on the APP is calculated using the Euclidean distance formula, and then the user influence coefficient A and membership influence coefficient B are calculated;

[0012] Through the function:

[0013] f(x,y,z,t)=Ax+By+Cz+Mt

[0014] Calculate the user's response level. Specifically, the response level is divided into five levels: Level 1, Level 2, Level 3, Level 4, and Level 5.

[0015] Furthermore, based on the response level of the grading algorithm, users in area P are divided into priority levels equal to the number of response levels, with lower response levels having higher priority. The user and their priority level are then written into the area P database.

[0016] Furthermore, it also includes a load balancing server; when the highest priority user disconnects from the APP backend server, that is, when the APP backend server has idle capacity to process new responses, the load balancing server notifies the P zone to delete the highest priority user from the P zone database. Then, the P zone selects the connection response of the current highest priority user to respond to and forwards the connection response to the load balancing server for response; the APP backend server receives the connection response from the P zone from the load balancing server and responds to the connection response by transmitting it back to the P zone through the load balancing server, thus realizing the connection between the user and the APP backend server.

[0017] Furthermore, the calculation methods for the user influence coefficient A and the member influence coefficient B are as follows:

[0018] According to the Euclidean distance formula:

[0019]

[0020] The resulting influence formula is:

[0021]

[0022] u i Individual Influence

[0023] Average Influence

[0024] g(u: influence value; N: the total number of u values ​​involved in the influence calculation in the g(u) function; i = 1, 2, 3…N;

[0025] Therefore, we can conclude that:

[0026] User influence g A (u):

[0027]

[0028] Member influence g B (u):

[0029]

[0030] N A N represents the total number of users. B Indicates the total number of members, u Ai This represents the influence of the i-th user; N represents A The average influence of each user; u Bi This represents the influence of the i-th member; N represents B The average influence of each member;

[0031] Based on the two influence formulas, two coefficients are calculated: User Influence Coefficient A and Member Influence Coefficient B.

[0032]

[0033]

[0034] This invention also provides a method for handling peak-hour access to an app using a hierarchical response level algorithm. The method, which employs the hierarchical response level algorithm described above, further includes the following steps:

[0035] Step 1: When the APP client accesses the service and starts connecting to the APP backend server, it normally establishes a connection with the APP backend server and requests the service.

[0036] During peak access periods, the traffic to the APP's backend server reaches the server's processing peak, triggering the P-zone's service function. The P-zone will then take over the response of all newly established APP client links, while links established before the trigger will not be affected.

[0037] When a user of the APP client connects to the P area, the P area reads the user's information, including the APP user level, membership level, network speed, and waiting time.

[0038] Step 2: Calculate the response level and prioritize users who have established links with area P.

[0039] Step 3: After the highest priority user disconnects from the APP backend server, the load balancing server notifies Zone P to delete the highest priority user from the Zone P database;

[0040] In zone P, the user with the highest priority is selected to respond to the connection response, and the connection response is forwarded to the load balancer server to wait for a response.

[0041] Step 4: The load balancer server receives the request response from the user with the established connection and forwards the request response to the APP backend server, which then processes the relevant business.

[0042] Step 5: The APP user completes the request and exits the APP. The APP backend server disconnects from the user, releases the occupied memory, and continues to select the highest-level user's connection response from the P zone to respond.

[0043] Furthermore, after the APP client in step 1 establishes a link with the P area, the APP backend server can provide login authentication services; the P area will perform a preliminary login operation with the APP client that has established the link; the P area provides its own cached basic page display services, providing basic browsing functions without occupying the process of the APP backend server; users who have established a link with the P area maintain a long link and wait for the APP backend server to process.

[0044] Compared with the prior art, the present invention has the following beneficial effects:

[0045] 1. The hierarchical algorithm of this invention can provide login services when traffic peaks and the backend server reaches its maximum processing capacity by using the P zone, allowing users to log in to the APP normally and access certain pages even during peak access periods.

[0046] 2. The hierarchical and graded algorithms of this invention fully consider factors such as user membership level, APP level, and network speed to prioritize users. When the data access volume is large, users with higher levels can access services first, which provides a better user experience for high-level users and improves user stickiness and recognition of the APP.

[0047] 3. The tiered algorithm of this invention solves the problem of server crashes due to a large influx of users during peak traffic periods by employing a method of prioritizing users in a tiered manner during peak traffic periods and processing request connections by the server according to priority. Attached Figure Description

[0048] Other features, objects, and advantages of the present invention will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings:

[0049] Figure 1 This is an overall diagram of the hierarchical response algorithm system for the APP of this invention;

[0050] Figure 2 Here is the response flowchart for area P;

[0051] Figure 3 Response level flowchart. Detailed Implementation

[0052] The present invention will now be described in detail with reference to specific embodiments. These embodiments will help those skilled in the art to further understand the present invention, but do not limit the invention in any way. It should be noted that those skilled in the art can make several changes and improvements without departing from the concept of the present invention. These all fall within the protection scope of the present invention.

[0053] This invention provides a hierarchical response level algorithm-based APP peak access system, including an APP client, a P zone, and an APP backend server;

[0054] When the number of visits from the APP client reaches the peak capacity of the APP backend server, the access from the APP client is linked to the P area which is connected to the APP backend server through a layered algorithm.

[0055] In Zone P, access links from the APP client are categorized into different levels based on various factors of the APP client user information using a ranking algorithm. The priority level for processing by the APP's backend server is then determined based on this ranking, and the priority data is written to the Zone P database. These different factors for the APP client user include APP user level, APP membership level, network speed, and response time.

[0056] The services in the P zone are triggered by the APP's backend server when the number of visits from the APP client reaches the peak capacity of the APP's backend server. A layered algorithm ensures seamless data access between the APP client and the P zone. The P zone's services include taking over access links from the APP client after the P zone is triggered, providing login authentication services, providing basic page display services with its own cache, and providing basic browsing functionality, all without consuming processing time on the APP's backend server.

[0057] It also includes a load balancing server; when the highest priority user disconnects from the APP backend server, that is, when the APP backend server has idle capacity to process new responses, the load balancing server notifies the P zone to delete the highest priority user from the P zone database. Then, the P zone selects the connection response of the current highest priority user to respond to and forwards the connection response to the load balancing server for response. The APP backend server receives the connection response from the P zone from the load balancing server and responds to the connection response by transmitting it back to the P zone through the load balancing server, thus realizing the connection between the user and the APP backend server.

[0058] The ranking algorithm reads the user level x and membership level y from the APP user information in the P zone; based on the network speed requirements of different APPs, it divides the network speed of the user accessing the P zone into network speed levels, and takes the corresponding network speed level value as the network speed value z for the user's access to the P zone, and sets the network speed coefficient C; the P zone starts timing from the time the user establishes a connection with the P zone, which is the user's waiting time t, and the time coefficient M is determined based on the waiting time t; when the P zone is activated, the influence of each user's real-time user level x and membership level y on the APP is calculated using the Euclidean distance formula, and then the user influence coefficient A and membership influence coefficient B are calculated;

[0059] According to the Euclidean distance formula:

[0060]

[0061] The resulting influence formula is:

[0062]

[0063] u i Individual Influence

[0064] Average Influence

[0065] g(u) influence value; N: represents the total number of u values ​​involved in the influence calculation in the g(u) function; i = 1, 2, 3…N;

[0066] Therefore, we can conclude that:

[0067] User influence g A (u):

[0068]

[0069] Member influence g B (u):

[0070]

[0071] N A N represents the total number of users. B Indicates the total number of members, u Ai This represents the influence of the i-th user; N represents A The average influence of each user; u Bi This represents the influence of the i-th member; N represents B The average influence of each member;

[0072] Based on the two influence formulas, two coefficients are calculated: User Influence Coefficient A and Member Influence Coefficient B.

[0073]

[0074]

[0075] Through the function:

[0076] f(x,y,z,t)=Ax+By+Cz+Mt

[0077] Calculate the user's response level. Specifically, the response level is divided into five levels: one, two, three, four, and five.

[0078] The response level algorithm assigns a priority level to users in zone P by setting those with lower response levels to higher priority levels. The number of priority levels is equal to the number of response levels. The user and their priority level are then written into the zone P database.

[0079] This embodiment also specifically provides a method for APP peak access based on a hierarchical response level algorithm. The APP peak access system using the hierarchical response level algorithm described above also includes the following steps:

[0080] Step 1: When the APP client accesses the service and starts connecting to the APP backend server, it normally establishes a connection with the APP backend server and requests the service.

[0081] During peak access periods, the traffic to the APP's backend server reaches the server's processing peak, triggering the P-zone's service function. The P-zone will then take over the response of all newly established APP client links, while links established before the trigger will not be affected.

[0082] When a user of the APP client connects to the P area, the P area reads the user's information, including the APP user level, membership level, network speed, and waiting time.

[0083] Specifically: The APP client establishes a connection with the P-zone, and the APP backend server provides login authentication services; the P-zone will perform an initial login operation with the linked APP client; the P-zone provides its own cached basic page display services, offering basic browsing functions without consuming APP backend server processes. Specific business logic requires processing by the APP backend server; users who have established a connection with the P-zone maintain a persistent connection and wait for processing by the APP backend server.

[0084] Step 2: Calculate the response level and prioritize users who have established links with area P.

[0085] Step 3: After the highest priority user disconnects from the APP backend server, the load balancing server notifies Zone P to delete the highest priority user from the Zone P database;

[0086] In zone P, the user with the highest priority is selected to respond to the connection response, and the connection response is forwarded to the load balancer server to wait for a response.

[0087] Step 4: The load balancer server receives the request response from the user with the established connection and forwards the request response to the APP backend server, which then processes the relevant business.

[0088] Step 5: The APP user completes the request and exits the APP. The APP backend server disconnects from the user, releases the occupied memory, and continues to select the highest-level user's connection response from the P zone to respond.

[0089] The working principle of this invention is as follows:

[0090] The hierarchical response ranking algorithm consists of two parts: a hierarchical algorithm and a ranking algorithm. The hierarchical algorithm involves both the APP client and the APP backend server. It is used when the APP client accesses the server during peak periods, and the traffic surge reaches the server's peak capacity. During peak APP client access periods, this algorithm links client access to a waiting area—the P-zone—connecting to the server. In the P-zone, the ranking algorithm categorizes accessing clients based on factors such as APP user level, APP membership level, and network speed. Lower-priority clients (i.e., higher-priority) can access the server first. After a high-priority client successfully connects to the server, it initiates a request-response mechanism while the lower-priority client initiates a connection response. Simultaneously, the server responds to the client's request and lower-priority clients' connection responses. When a high-priority client completes its access and sends a disconnect response, the highest-priority client sends a connection response, the server releases the high-priority client's resources, and responds to the highest-priority connection response in the P-zone.

[0091] The specific implementation steps are as follows:

[0092] Step 1: When the APP client accesses the service and starts connecting to the APP backend server, it normally establishes a connection with the APP backend server to request service processing. During peak access periods, the server traffic reaches its processing capacity. The server cannot process the newly established connection responses in time. At this time, the APP backend server triggers the P-zone function. The P-zone service will take over all newly established connection responses, and connections established before the trigger are unaffected.

[0093] When a user connects to the P area, the P area reads data such as the user's APP user level, membership level, and network speed.

[0094] like Figure 1 The diagram shows the overall system diagram of the APP distributed response algorithm.

[0095] Specifically:

[0096] 1.1 The APP client establishes a connection with the P zone, and the server can provide login authentication services.

[0097] 1.2. Area P will perform an initial login operation with the linked APP client.

[0098] 1.3. The P-zone calculates user priority based on different factors (user level, membership level, network speed, response time, etc.) and saves the results in the database. Priority is divided into five levels: Level 1, Level 2, Level 3, Level 4, and Level 5.

[0099] 1.4. The P zone provides basic page display services through its own cache, offering basic browsing functionality without consuming processes on the app's backend server. Specific business logic requires processing by the app's backend server.

[0100] 1.5 Users who have established a connection with the P area should maintain a long-term connection and wait for the APP's backend server to process the request.

[0101] Step 2: Calculate and sort the response levels of users who have established links with area P.

[0102] Specifically:

[0103] 2.1. In the P zone, the user level x and membership level y are read directly from the user information.

[0104] 2.2. Based on the network speed requirements of different apps, four key values ​​are selected for network speed. These four key values ​​are used to classify user network speeds into five levels, corresponding to values ​​of 1, 5, 10, 15, and 20 respectively. The P area reads the user's network speed, compares it with these four key values, and selects the corresponding value as the user's network speed value z.

[0105] 2.3. The time taken for P area starts from the moment the user establishes a connection with P area, and is recorded as the user's waiting time t.

[0106] 2.4 When the P zone is activated, the influence of each user's real-time user level and membership level on the APP will be calculated.

[0107] According to the Euclidean distance formula:

[0108]

[0109] The formula for influence is derived from the transformation:

[0110]

[0111] u i Individual Influence

[0112]

[0113] g(u): Influence value

[0114] Therefore, we can conclude that:

[0115] User influence g A (u):

[0116]

[0117] Member influence g B (u):

[0118]

[0119] N A N represents the total number of users. B Indicates the total number of members, u Ai This represents the influence of the i-th user; N represents A The average influence of each user; u Bi This represents the influence of the i-th member; N represents B The average influence of each member;

[0120] Based on the two influence formulas, two coefficients are calculated: User Influence Coefficient A and Member Influence Coefficient B.

[0121]

[0122]

[0123] 2.5. The network speed coefficient is a fixed value: C = 0.3

[0124] 2.6 The time coefficient M, the value of which depends on the user's waiting time t:

[0125] When t < 10s, M = 0

[0126] When t >= 10 s, M = 0.5

[0127] 2.7. Based on the above data, using the response level algorithm f(x,y,z,t):

[0128] f(x,y,z,t)=Ax+By+Cz+Mt (7)

[0129] Calculate the user's response level. Where:

[0130] x: APP user level

[0131] y:APP Membership Level

[0132] z: Internet speed entering zone P

[0133] t: Waiting time to enter area P

[0134] A: User coefficient

[0135] B: Membership coefficient

[0136] C: Internet speed coefficient

[0137] M: Time coefficient

[0138] 2.8. Based on the above response level algorithm, the response levels of users in area P are divided into five levels: one, two, three, four, and five. The user and their level are then written into the database of area P.

[0139] Step 3: After the highest priority user disconnects from the APP backend server, i.e., when the APP backend server has idle capacity to process new responses and deletes the user from the P-zone database, it selects the highest priority user's connection response from the P-zone for response, and forwards the connection response from the P-zone to the load balancer server for response. The process is as follows: Figure 2 and Figure 3 .

[0140] Specifically:

[0141] 3.1 After the highest priority user disconnects from the APP backend server, the load balancer server notifies the P-zone database to delete the user's data.

[0142] 3.2. From step 2, the P zone selects the connection response data of the user with the highest current level and forwards it to the load balancer server.

[0143] 3.3 The APP backend server receives the connection response from area P and transmits the response back to area P, indicating that the user has successfully connected with the APP backend server.

[0144] Step 4: The APP's backend server continues to process business.

[0145] Specifically:

[0146] 4.1 The load balancer server receives the request response from the user with the established connection in step 3.3 and transfers it to the backend server to continue processing the relevant business.

[0147] Step 5: The user completes the request and exits the APP. The APP's backend server disconnects from the user, releases the occupied memory, and continues to select the highest-level user's connection response from the P zone to respond.

[0148] Specifically:

[0149] 5.1 Once the user completes all requests and exits the app, the app's backend server disconnects from the user.

[0150] 5.2 The APP backend server releases the memory occupied by the user, and the load balancer server notifies zone P.

[0151] 5.3. From step 2.2, P zone selects to forward the connection response data of the highest-level user to the load balancer server.

[0152] Taking the response level algorithm as an example, which divides responses into five levels, the communication relationship between Zone P and the APP backend server is as follows:

[0153] P zone → APP backend server: Sends a level 1 connection response according to the level algorithm (level 1 is determined as "yes").

[0154] APP backend server → P zone: Handles connection responses of level 1.

[0155] P zone → APP backend server: Level 1 establishes connection, sends Level 1 request response, and sends Level 2 connection response (Level 2 is determined to be "yes").

[0156] APP backend server → P zone: handles level 1 request responses and level 2 connection responses.

[0157] P zone → APP backend server: Sends a level 1 disconnect response, a level 2 connection establishment response, sends a level 2 request response, and sends a level 3 connection response (level 3 is determined to be "yes").

[0158] APP backend server → P zone: Handles disconnection responses of level 1, releases resources occupied by level 1, handles request responses of level 2, and handles connection responses of level 3.

[0159] P zone → APP backend server: Sends a level 2 disconnect response, a level 3 connection establishment response, sends a level 3 request response, and sends a level 4 connection response (level 4 is determined to be "yes").

[0160] APP backend server → P zone: Handles level 2 disconnect responses, releases level 2 occupied resources, handles level 3 request responses, and handles level 4 connection responses.

[0161] P zone → APP backend server: Sends a level 3 disconnect response, a level 4 connection establishment response, sends a level 4 request response, and sends a level 5 connection response (level 5 is determined to be "yes").

[0162] APP backend server → P zone: Handles level 3 disconnect responses, releases level 3 occupied resources, handles level 4 request responses, and handles level 5 connection responses.

[0163] P zone → APP backend server: Sends a level 4 disconnect response, establishes a level 5 connection, and sends a level 5 request response.

[0164] APP backend server → P zone: Handles level 4 disconnection responses, releases resources occupied by level 4, and handles level 5 request responses.

[0165] P zone → APP backend server: Send a level 5 disconnect response.

[0166] APP backend server → P zone: Handles disconnection responses at level 5 and releases resources occupied by level 5.

[0167] Those skilled in the art will understand that, besides implementing the system and its various devices, modules, and units provided by this invention in the form of purely computer-readable program code, the same functions can be achieved entirely through logical programming of the method steps, making the system and its various devices, modules, and units of this invention function in the form of logic gates, switches, application-specific integrated circuits, programmable logic controllers, and embedded microcontrollers. Therefore, the system and its various devices, modules, and units provided by this invention can be considered as a hardware component, and the devices, modules, and units included therein for implementing various functions can also be considered as structures within the hardware component; alternatively, the devices, modules, and units for implementing various functions can be considered as both software modules implementing the method and structures within the hardware component.

[0168] Specific embodiments of the present invention have been described above. It should be understood that the present invention is not limited to the specific embodiments described above, and those skilled in the art can make various changes or modifications within the scope of the claims, which do not affect the essence of the present invention. Unless otherwise specified, the embodiments and features described in this application can be arbitrarily combined with each other.

Claims

1. A hierarchical response level algorithm for APP peak access system, characterized in that, This includes the APP client, P zone, and APP backend server; When the number of visits from the APP client reaches the peak capacity of the APP backend server, the access from the APP client is linked to the P area which is connected to the APP backend server through a layered algorithm. In Zone P, access links from the APP client are classified into different levels based on different factors of the APP client user information using a ranking algorithm. Then, the priority level of the APP backend server is determined according to the ranking, and the priority level data is written into the Zone P database. The different factors of the APP client user information include APP user level, APP membership level, network speed, and response time. The grading algorithm reads the user grading level from the APP user information based on the P zone. x and membership level y Based on the network speed requirements of different apps, the network speed of users accessing the P-zone is divided into speed levels. The corresponding speed level value is then retrieved based on the user's network speed and assigned as the network speed value for the user's access to the P-zone. z Set network speed coefficient C The P-zone starts timing from the moment a user establishes a connection with the P-zone, representing the user's waiting time. t Based on waiting time t Determine the time coefficient M When zone P is activated, the real-time user level of each user is calculated using the Euclidean distance formula. x and membership level y The influence of the app is used to calculate the user influence coefficient. A Member Influence Coefficient B ; Through the function: Calculate the user's response level; Based on the response level of the grading algorithm, users in area P are divided into priority levels equal to the number of response levels, with lower response levels having higher priority. The user and their priority level are then written into the area P database. The user influence coefficient A The member influence coefficient B The calculation method is as follows: According to the Euclidean distance formula: The resulting influence formula is: Influence value; N: represents The total number of u values ​​involved in the influence calculation within the function; =1,2,3…N; Therefore, we can conclude that: User influence : Member influence : N A N represents the total number of users. B Indicates the total number of members. Indicates the first Individual user influence; N represents A The average influence of each user's influence; Indicates the first Individual member influence; N represents B The average influence of each member; According to the two influence formulas, two coefficients are calculated: a user influence coefficient A , and a member influence coefficient B : 。 2. The tiered response algorithm APP peak visit system of claim 1, wherein, The service functions of the P zone are triggered by the APP backend server when the access volume of the APP client reaches the peak capacity of the APP backend server; the data access between the APP client and the P zone is achieved without difference through a layered algorithm.

3. The tiered response algorithm APP peak visit system of claim 2, wherein, The service functions of the P zone include taking over the access links of the APP client after the P zone is triggered, providing login authentication services, providing basic page display services for its own cache, and providing basic browsing functions, which can be done without occupying the process of the APP background server.

4. The APP peak access system based on the hierarchical response level algorithm according to claim 1, characterized in that, It also includes a load balancing server; when the highest priority user disconnects from the APP backend server, the load balancing server notifies the P zone to delete the highest priority user from the P zone database, and then the P zone selects the connection response of the current highest priority user to respond to and forwards the connection response to the load balancing server for response. The APP backend server receives the connection response from zone P from the load balancer server and transmits the response back to zone P through the load balancer server, thus establishing the connection between the user and the APP backend server.

5. A layered response level algorithmic APP peak hour access method, characterized by, The APP peak access system employing the hierarchical response level algorithm as described in any one of claims 1-4 further includes the following steps: Step 1: When the APP client accesses the service and starts connecting to the APP backend server, it normally establishes a connection with the APP backend server and requests the service. During peak access periods, the traffic to the APP's backend server reaches the server's processing peak, triggering the P-zone's service function. The P-zone will then take over the response of all newly established APP client links. When a user of the APP client connects to the P area, the P area reads the user's information, including the APP user level, membership level, network speed, and waiting time. Step 2: Calculate the response level and prioritize users who have established links with area P; Step 3: After the highest priority user disconnects from the APP backend server, the load balancing server notifies Zone P to delete the highest priority user from the Zone P database; In zone P, the user with the highest priority is selected to respond to the connection response, and the connection response is forwarded to the load balancer server to wait for a response. Step 4: The load balancer server receives the request response from the user with the established connection and forwards the request response to the APP backend server, which then processes the relevant business. Step 5: The APP user completes the request and exits the APP. The APP backend server disconnects from the user, releases the occupied memory, and continues to select the highest-level user's connection response from the P zone to respond.

6. The tiered response algorithm APP peak visit method of claim 5, wherein, After the APP client establishes a connection with the P area in step 1, the APP backend server can provide login authentication services; the P area will perform a preliminary login operation with the APP client that has established the connection; the P area provides basic page display services with its own cache, providing basic browsing functions without occupying the process of the APP backend server; users who have established a connection with the P area maintain a long connection and wait for the APP backend server to process.

Images