A high-concurrency access optimization system for an engineering machine rental platform
By implementing modular access request routing, data caching optimization, and database read/write separation, combined with dynamic resource scheduling, the system addresses the issues of response latency and resource waste under high concurrency access in the construction machinery rental platform, thereby achieving high system efficiency, stability, and improved resource utilization.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- CHINA RAILWAY FIRST GROUP CO LTD
- Filing Date
- 2026-04-07
- Publication Date
- 2026-06-30
AI Technical Summary
Construction machinery rental platforms suffer from response latency, resource waste, and instability under high concurrency access, especially issues such as resource contention between core and non-core requests, low cache hit rate, and database connection pool exhaustion.
The system employs an access request splitting module, a data caching optimization module, a database read/write separation module, and a resource dynamic scheduling module. It achieves precise traffic splitting through request priority classification and load balancing, improves data access efficiency through a multi-level caching architecture, reduces database pressure based on read/write separation, and dynamically adjusts resource allocation in conjunction with real-time load, all working together to optimize system performance.
Under 100,000 instantaneous concurrent requests, it shortens the response time of core business, improves cache hit rate, reduces database connection pool exhaustion rate, significantly improves system concurrency processing capacity and resource utilization, and ensures system stability.
Smart Images
Figure CN122309554A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of construction machinery rental service technology, and more specifically to a high-concurrency access optimization system for construction machinery rental platforms. Background Technology
[0002] Currently, with the digital transformation of the construction machinery rental industry, rental platforms have become the core carrier for connecting equipment supply and demand, covering the entire process of services such as equipment information display, online booking, order management, and performance monitoring. In recent years, with the expansion of the user base (including individual tenants, construction companies, and equipment manufacturers) and the expansion of business scenarios (such as emergency engineering equipment dispatch and cross-regional rental needs), the platform has experienced an exponential increase in concurrent access volume, especially during peak construction seasons and equipment promotions, when instantaneous concurrent requests can exceed 100,000.
[0003] The existing technology has the following key drawbacks: The access request processing architecture is simplistic and does not differentiate between request types and priorities. This results in core requests such as order submissions competing for resources with non-core requests such as data statistics, leading to delays or even timeouts in the response of core business processes. The caching strategy is rigid, using only a single level of caching, and cache updates do not take into account the frequency and timeliness of data access. The cache hit rate of frequently accessed data such as device basic information and popular orders is less than 60%, and a large number of requests directly penetrate to the database. The database uses a single-database architecture, and read and write operations are not separated. In scenarios where read requests account for more than 80%, the database connection pool is frequently exhausted, and the query response time exceeds 2 seconds. The static allocation of resources makes it impossible to dynamically adjust resources such as CPU, memory, and number of connections according to real-time load, resulting in some nodes becoming overloaded and crashing, while some resources remain idle, leading to poor system stability.
[0004] Therefore, how to achieve coordinated optimization of request diversion, caching optimization, database decompression and dynamic resource scheduling through modular design, and solve the problems of response latency, resource waste and insufficient stability under high concurrency access of construction machinery rental platforms, is a technical problem that urgently needs to be solved by those skilled in the art. Summary of the Invention
[0005] In view of this, the present invention provides a high-concurrency access optimization system for an engineering machinery rental platform to solve the problems existing in the background art.
[0006] To achieve the above objectives, the present invention adopts the following technical solution: A high-concurrency access optimization system for an engineering machinery rental platform includes: an access request diversion module, a data caching optimization module, a database read / write separation module, and a resource dynamic scheduling module. These modules work together to optimize high-concurrency access. The access request routing module is used to receive client access requests and parse request characteristics, and classify and route requests to the corresponding business service nodes through priority calculation and load balancing strategies. The data caching optimization module establishes a communication connection with the access request splitting module to receive the split access requests, query the target data in the multi-level cache, return the data directly when the cache is hit, and trigger a database access request when the cache is not hit. The database read / write separation module establishes a communication connection with the data cache optimization module. It is used to receive database access requests after a cache miss, route write requests to the master database and read requests to the slave database, and after completing the data read / write operation, it feeds back to the data cache optimization module to update the cache. The resource dynamic scheduling module establishes communication connections with the access request diversion module, the data caching optimization module, and the database read / write separation module, respectively. It is used to collect the operating load indicators of each module in real time, and dynamically adjust the resource allocation of each module based on the load feedback, so as to achieve system performance optimization and stability assurance under high concurrency access.
[0007] Optionally, the request priority calculation function of the access request routing module is implemented using a formula. Implementation, in which P For request priority, T For request type weight, U User level coefficient E To determine the urgency level of the request, The weighting coefficients and .
[0008] Optionally, the data caching optimization module adopts a multi-level architecture of local caching and distributed caching clusters, and the cache hit rate is calculated using the formula... The cache update threshold is calculated using the formula. Calculate, where H is the cache hit rate. To cache the hit count, To cache the number of misses, F For data update frequency, The weighting coefficients and Optionally, the database read / write separation module includes a master and slave cluster. The master and slave databases synchronize data via binlog logs, and the slave read request allocation ratio is determined by a formula. Implementation, where R is the proportion of read requests allocated to the slave database. To the total processing capacity of the slave database cluster, The primary database write processing capacity is given by k, where k is the ratio of historical read / write requests on the platform. Optionally, the resource allocation adjustment function of the dynamic resource scheduling module is implemented using the formula... Implementation, where S is the adjusted resource allocation. Where L is the basic resource quantity and L is the current load. For rated load, For adjustment coefficients and 0 < <1. Optionally, the load balancing strategy is specifically as follows:
[0009] In the formula, For the first i The request allocation ratio for each service node; For the first i The remaining processing capacity of each service node; n is the total number of service nodes of the same type. Optionally, the load warning threshold of the resource dynamic scheduling module is determined by the formula... The configuration is set such that when the resource dynamic scheduling module is currently under load... When this occurs, an early warning is triggered and a resource expansion process is initiated. Optionally, the local cache of the data caching optimization module is deployed on each business service node using the Caffeine framework, with the cache capacity set to 20% of the corresponding node server memory and an expiration time of 5 minutes, used to store frequently accessed hot data within the past 5 minutes; the distributed cache cluster is deployed using Redis master-slave replication and sentinel mode, storing the full amount of high-frequency data in shards according to device ID hash, and dynamically adjusting the data expiration time according to the cache update threshold.
[0010] As can be seen from the above technical solution, compared with the prior art, the present invention discloses a high-concurrency access optimization system for an engineering machinery rental platform, including an access request diversion module, a data caching optimization module, a database read-write separation module, and a resource dynamic scheduling module. It achieves precise diversion through request priority classification and load balancing, improves data access efficiency by adopting a multi-level caching architecture, reduces database pressure based on read-write separation, and dynamically adjusts resource allocation in conjunction with real-time load. The four modules work together to shorten the core business response time, improve the cache hit rate, and reduce the database connection pool exhaustion rate under 100,000 instantaneous concurrent requests. This significantly improves the system's concurrent processing capacity, resource utilization, and operational stability, and is suitable for digital engineering machinery rental platforms of various sizes. Attached Figure Description
[0011] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.
[0012] Figure 1 This is a schematic diagram of the system structure provided by the present invention. Detailed Implementation
[0013] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0014] This invention discloses a high-concurrency access optimization system for an engineering machinery rental platform, such as... Figure 1 As shown, it includes: an access request splitting module, a data caching optimization module, a database read / write separation module, and a resource dynamic scheduling module. These modules work together to optimize high-concurrency access. The access request routing module is used to receive client access requests and parse request characteristics, and classify and route requests to the corresponding business service nodes through priority calculation and load balancing strategies. The data caching optimization module establishes a communication connection with the access request splitting module to receive the split access requests, query the target data in the multi-level cache, return the data directly when the cache is hit, and trigger a database access request when the cache is not hit. The database read / write separation module establishes a communication connection with the data cache optimization module. It is used to receive database access requests after a cache miss, route write requests to the master database and read requests to the slave database, and after completing the data read / write operation, it feeds back to the data cache optimization module to update the cache. The resource dynamic scheduling module establishes communication connections with the access request diversion module, the data caching optimization module, and the database read / write separation module, respectively. It is used to collect the operating load indicators of each module in real time, and dynamically adjust the resource allocation of each module based on the load feedback, so as to achieve system performance optimization and stability assurance under high concurrency access.
[0015] Specifically, the access request routing module, as the core module of the platform's entry layer, is responsible for receiving all access requests from clients (Web, APP, and mini-program), parsing request characteristics (including request type, initiating user level, and urgency), calculating request priority, and routing requests to the corresponding business service nodes (device query node, order processing node, and data statistics node) based on a weighted round-robin load balancing algorithm.
[0016] Working principle: Prioritize requests based on their differentiated characteristics. Core business requests (order submission, fault reporting) are assigned higher processing weights, while non-core business requests (data statistics, historical order queries) are assigned lower weights. This avoids overloading a single node due to mixed requests and achieves "classification routing and priority scheduling" of requests.
[0017] The request priority calculation function uses a formula Implementation, in which P For request priority, T For request type weight, U User level coefficient E To determine the urgency level of the request, The weighting coefficients and .
[0018] The specific load balancing strategy is as follows:
[0019] In the formula, For the first i The request allocation ratio for each service node; For the first i The remaining processing capacity of each service node; n is the total number of service nodes of the same type. The data caching optimization module adopts a multi-level caching architecture of "local caching + distributed caching cluster" and communicates and interfaces with the access request diversion module. The local cache is deployed on each business service node, storing frequently accessed hot data (such as popular device information and current online user data) within the past 5 minutes; the distributed caching cluster is deployed using a Redis cluster, storing all high-frequency data (basic device information, valid order data, and commonly used user configurations), and dynamically adjusting the cache update strategy and expiration time.
[0020] Working principle: Multi-level caching reduces database access penetration. Local caching is responsible for quickly responding to extremely high-frequency short-term requests, while the distributed caching cluster is responsible for global data sharing and medium- to long-term caching. The caching strategy is dynamically adjusted based on cache hit rate and data update frequency to avoid data inconsistency caused by cache expiration, while improving cache resource utilization.
[0021] The cache hit rate is expressed by the formula The cache update threshold is calculated using the formula. Calculate, where H is the cache hit rate. To cache the hit count, To cache the number of misses, F For data update frequency, The weighting coefficients and .
[0022] When data access frequency > When the data access frequency is <, upgrade the data to a distributed cache cluster and extend the expiration time; when the data access frequency is < And if the basic expiration time (default 1 hour) is exceeded, it will be removed from the cache.
[0023] The database read / write separation module adopts a "master database + slave database cluster" architecture and communicates with the data caching optimization module. The master database is responsible for handling all write requests (order creation, device status updates, user information modification), while the slave database cluster (≥3 nodes) is responsible for handling all read requests (device queries, order queries, data statistics). The master and slave databases achieve real-time data synchronization (synchronization latency ≤100ms) through MySQL binlog logs, and the number of slave database nodes is dynamically expanded based on read request pressure.
[0024] Working principle: Taking advantage of the "read-heavy, write-light" business characteristics of the construction machinery rental platform (read requests account for 80%-90%), read and write operations are separated to different database nodes. The master database focuses on ensuring transaction consistency, while the slave database cluster shares the read pressure. At the same time, master-slave synchronization ensures data consistency and improves the overall concurrent processing capability of the database.
[0025] The allocation ratio of read requests from the database is determined by the formula. Implementation, where R is the proportion of read requests allocated to the slave database. To the total processing capacity of the slave database cluster, The primary database write processing capacity is represented by k, which is the ratio of historical read and write requests on the platform.
[0026] Master-slave synchronization delay control:
[0027] In the formula, D represents the master-slave synchronization delay; The commit timestamp for the primary database transaction; The timestamp is used to complete the synchronization from the slave database.
[0028] The resource dynamic scheduling module is deployed at the system management layer. It collects the operating metrics (CPU utilization, memory usage, average response time, number of connections, and QPS) of the access request diversion module, data caching optimization module, and database read / write separation module in real time through the monitoring agent. Based on the load feedback model, it dynamically adjusts the resource allocation of each module (including service node scaling up / down, cache cluster memory expansion, and database connection pool size adjustment).
[0029] Working principle: A closed-loop mechanism of monitoring, analysis, scheduling and feedback is established. When the module load exceeds the rated threshold, resource expansion is automatically triggered; when the load is below the threshold for more than 3 minutes, resource reduction is triggered to avoid resource idleness and ensure the stability and resource utilization of the system under high concurrency fluctuations.
[0030] Among them, the adjustment of resource allocation is achieved through the formula Implementation, where S is the adjusted resource allocation. Where L is the basic resource quantity and L is the current load. For rated load, For adjustment coefficients and 0 < <1. Load warning threshold is determined by the formula. The configuration is set such that when the resource dynamic scheduling module is currently under load... When this happens, an alert is triggered and a resource expansion process is initiated. Specifically, the local cache of the data caching optimization module is deployed on each business service node using the Caffeine framework. The cache capacity is set to 20% of the corresponding node server memory, and the expiration time is 5 minutes. This cache is used to store frequently accessed hot data within the past 5 minutes. The distributed cache cluster is deployed using Redis master-slave replication and sentinel mode. It stores all high-frequency data in shards by device ID hash and dynamically adjusts the data expiration time according to the cache update threshold.
[0031] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the apparatus disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the description is relatively simple; relevant parts can be referred to the method section.
[0032] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims
1. A high-concurrency access optimization system for an engineering machinery rental platform, characterized in that, include: The system includes an access request routing module, a data caching optimization module, a database read / write separation module, and a resource dynamic scheduling module. These modules work together to optimize high-concurrency access. The access request routing module is used to receive client access requests and parse request characteristics, and classify and route requests to the corresponding business service nodes through priority calculation and load balancing strategies. The data caching optimization module establishes a communication connection with the access request splitting module to receive the split access requests, query the target data in the multi-level cache, return the data directly when the cache is hit, and trigger a database access request when the cache is not hit. The database read / write separation module establishes a communication connection with the data cache optimization module. It is used to receive database access requests after a cache miss, route write requests to the master database and read requests to the slave database, and after completing the data read / write operation, it feeds back to the data cache optimization module to update the cache. The resource dynamic scheduling module establishes communication connections with the access request diversion module, the data caching optimization module, and the database read / write separation module, respectively. It is used to collect the operating load indicators of each module in real time, and dynamically adjust the resource allocation of each module based on the load feedback, so as to achieve system performance optimization and stability assurance under high concurrency access.
2. The high-concurrency access optimization system for an engineering machinery rental platform according to claim 1, characterized in that, The request priority calculation function of the access request routing module uses the formula Implementation, in which P For request priority, T For request type weight, U User level coefficient E To determine the urgency level of the request, The weighting coefficients and .
3. The high-concurrency access optimization system for an engineering machinery rental platform according to claim 1, characterized in that, The data caching optimization module adopts a multi-level architecture of local caching and distributed caching clusters. The cache hit rate is calculated using the formula... The cache update threshold is calculated using the formula. Calculate, where H is the cache hit rate. To cache the hit count, To cache the number of misses, F For data update frequency, The weighting coefficients and .
4. The high-concurrency access optimization system for an engineering machinery rental platform according to claim 1, characterized in that, The database read / write separation module includes a master and slave database cluster. Data synchronization between the master and slave databases is achieved through binlog logs. The slave database read request allocation ratio is determined by a formula. Implementation, where R is the proportion of read requests allocated to the slave database. To the total processing capacity of the slave database cluster, The primary database write processing capacity is represented by k, which is the ratio of historical read and write requests on the platform.
5. The high-concurrency access optimization system for an engineering machinery rental platform according to claim 1, characterized in that, The resource allocation adjustment function of the dynamic resource scheduling module is achieved through a formula. Implementation, where S is the adjusted resource allocation. Where L is the basic resource quantity and L is the current load. For rated load, For adjustment coefficients and 0 < <1.
6. The high-concurrency access optimization system for an engineering machinery rental platform according to claim 2, characterized in that, The load balancing strategy is as follows: In the formula, For the first i The request allocation ratio for each service node; For the first i The remaining processing capacity of each service node; n is the total number of service nodes of the same type.
7. A high-concurrency access optimization system for an engineering machinery rental platform according to claim 5, characterized in that, The load warning threshold of the resource dynamic scheduling module is determined by the formula. The configuration is set such that when the resource dynamic scheduling module is currently under load... When this happens, an alert is triggered and the resource expansion process is initiated.
8. The high-concurrency access optimization system for an engineering machinery rental platform according to claim 1, characterized in that, The local cache of the data caching optimization module is deployed on each business service node using the Caffeine framework. The cache capacity is set to 20% of the corresponding node server memory and the expiration time is 5 minutes. It is used to store hot data that is frequently accessed within the past 5 minutes. The distributed cache cluster is deployed using Redis master-slave replication and sentinel mode. It stores the full amount of high-frequency data by hashing the device ID and dynamically adjusts the data expiration time according to the cache update threshold.