Large-scale concurrent network request management method based on computing power host management platform

Abstract

The invention provides a large-scale concurrent network request management method based on a computing power host management platform, belongs to the technical field of computer networks, and aims to reduce the request pressure on a rear-end host through technologies such as merging same or similar requests, allocating priorities for the requests, maintaining request queues and controlling the request sending frequency. The utilization efficiency of resources is improved, so that the system performance and the user experience are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of computer networks, and particularly to a method for managing large-scale concurrent network requests based on a computing power host management platform. Background Art

[0002] With the rapid development of cloud computing and big data technologies, the computing power host management platform faces severe challenges in processing large-scale concurrent network requests. Especially in high-concurrency business scenarios such as scientific computing, the diverse business needs of user groups put forward higher requirements for the operation efficiency of the computing power network. In the prior art, artificial intelligence tools are used to analyze and parse diverse business needs. Although the processing efficiency is improved to a certain extent, in the case of a large number of business types and a large number of business volumes, it will consume some computing power of the computing power network, resulting in a decrease in operation efficiency. Therefore, how to effectively manage large-scale concurrent network requests and improve the operation efficiency of the computing power host management platform has become an urgent problem to be solved. Aiming at the deficiencies of the prior art, the present invention proposes a management method for large-scale concurrent network requests based on a computing power host management platform to improve the system's ability to process concurrent requests, optimize resource utilization, and reduce response time. Summary of the Invention

[0003] To solve the above technical problems, the present invention provides a method for managing large-scale concurrent network requests based on a computing power host management platform. By techniques such as merging identical or similar requests, assigning priorities to requests, maintaining a request queue, controlling the request sending frequency, and error handling, the request pressure on the backend host is reduced to solve the problems of unreasonable resource allocation and low scheduling efficiency in processing large-scale concurrent network requests in the prior art, improve the efficiency and stability of network request processing, and enhance the user experience.

[0004] The technical solution of the present invention is as follows:

[0005] A method for managing large-scale concurrent network requests based on a computing power host management platform, comprising the following steps:

[0006] (1) Network request classification and priority division;

[0007] (2) Dynamically allocate resources;

[0008] (3) Maintain a network request queue;

[0009] (4) Load balancing and failover mechanism;

[0010] (5) Use a cache to reduce duplicate requests.

[0011] Furthermore,

[0012] Use a hash table to store the requests that have been processed;

[0013] When a new request arrives, first calculate its hash value. If the hash value already exists in the table, discard the request. For requests with similar parameters, merge these parameters into a common request.

[0014] Furthermore,

[0015] Classify the network requests entering the computing power host management platform, and assign corresponding priorities to each request according to the type of request, business scenario, and user attributes.

[0016] Furthermore,

[0017] The computing power host management platform monitors the resource usage of each host in real time, and dynamically allocates computing power host resources to requests according to the priority of the requests and the resource status of each host;

[0018] When a high-priority request arrives, preferentially allocate a host with sufficient resources to this request;

[0019] For low-priority requests, on the premise of ensuring the processing of high-priority requests, utilize the idle resources of the host for processing.

[0020] Furthermore,

[0021] Establish several request queues, and place requests into different queues according to the priority; each queue is processed according to the first-in, first-out (FIFO) principle, but for the high-priority queue, a higher scheduling frequency is given during processing;

[0022] When the host resources are idle, preferentially take out requests from the high-priority queue for processing to ensure that high-priority requests can be responded to in a timely manner. At the same time, set a time threshold. When the waiting time of requests in the low-priority queue exceeds the threshold, increase their processing priority.

[0023] Furthermore,

[0024] Adopt a load balancing algorithm to evenly distribute concurrent requests to each computing power host;

[0025] When a certain host fails, transfer the requests it undertakes to other normal hosts for processing;

[0026] The load balancing algorithm dynamically adjusts the request allocation strategy according to the real-time load and processing capacity of the host.

[0027] Furthermore,

[0028] Set up a cache module in the computing power host management platform to cache frequently accessed data and response results; when receiving the same network request, preferentially obtain data from the cache and return it to the user.

[0029] The cache module uses the LRU algorithm to manage the cache content, and timely eliminates the cache data that has not been accessed for a long time to make room for new data.

[0030] The beneficial effects of the present invention are

[0031] 1. By classifying and prioritizing network requests, the present invention can ensure that important requests are processed first, improving the real-time performance and stability of the service. For example, in the order transaction scenario, it guarantees the rapid processing of transaction requests and reduces transaction delays.

[0032] 2. The resource dynamic allocation strategy enables the rational utilization of computing power host resources, avoiding resource waste and overload phenomena, and improving the overall performance of the system. Allocating requests according to the real-time resource status of the host can give full play to the processing capabilities of each host.

[0033] 3. The request queue management mechanism ensures the orderliness and fairness of request processing, while taking into account the timeliness of high-priority requests and the processing opportunities of low-priority requests, enhancing the user experience. High-priority requests are processed first, and low-priority requests will not be shelved indefinitely.

[0034] 4. The load balancing and failover mechanism enhances the reliability and availability of the system, effectively coping with emergencies such as host failures, and ensuring the continuous processing of large-scale concurrent network requests. When a certain host fails, the requests are automatically transferred to ensure that the service is uninterrupted.

[0035] The cache mechanism greatly improves the response speed, reduces the load on the server, and reduces the occupancy of network bandwidth. Frequently accessed data is obtained from the cache, reducing repeated processing and transmission, and improving the system efficiency. Specific implementation manners

[0036] To make the objectives, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below. Apparently, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0037] The present invention provides a method for managing large-scale concurrent network requests based on a computing power host management platform, including the following steps:

[0038] 1. Merging of identical or similar requests: Use a hash table to store requests that have already been processed. When a new request arrives, first calculate its hash value. If the hash value already exists in the table, discard the request. For requests with similar parameters but slight differences, these parameters can be merged into a common request.

[0039] 2. Request classification and priority assignment: First, classify the network requests entering the computing power host management platform. Based on factors such as the type of request (e.g., HTTP, TCP, etc.), business scenario (e.g., order transactions, file downloads, visual analysis, etc.), and user attributes (e.g., administrator users, ordinary users), assign corresponding priorities to each request. For example, for order transaction requests, due to their high requirements for real-time and accuracy, a higher priority can be assigned; while for ordinary file download requests, a lower priority can be assigned.

[0040] 3. Dynamic resource allocation: The computing power host management platform monitors the resource usage of each host in real time, including CPU usage rate, memory occupancy rate, network bandwidth, etc. Based on the priority of the request and the resource status of each host, dynamically allocate computing power host resources to the request. When a high-priority request arrives, preferentially allocate a host with sufficient resources to this request; for low-priority requests, on the premise of ensuring the processing of high-priority requests, use the idle resources of the host for processing. For example, if the CPU usage rate of a certain host is lower than 50%, the memory occupancy rate is lower than 60%, and there is a large surplus of network bandwidth, it can be allocated to high-priority real-time interaction requests.

[0041] 4. Request queue management: Establish multiple request queues and place requests into different queues according to their priorities. Each queue is processed according to the first-in, first-out (FIFO) principle, but for high-priority queues, a higher scheduling frequency is given during processing. When the host resources are idle, preferentially take out requests from the high-priority queue for processing to ensure that high-priority requests can be responded to in a timely manner. At the same time, to prevent requests in the low-priority queue from not being processed for a long time, set a certain time threshold. When the waiting time of requests in the low-priority queue exceeds the threshold, appropriately increase their processing priority.

[0042] 5. Load balancing and failover: Adopt a load balancing algorithm to evenly distribute concurrent requests to each computing power host, avoiding excessive load on a single host. When a certain host fails, it can quickly transfer the requests it undertakes to other normal hosts for processing to ensure the continuity of request processing. The load balancing algorithm can dynamically adjust the request allocation strategy according to factors such as the real-time load and processing capacity of the host. For example, the round-robin algorithm based on weights assigns different weights to hosts according to their performance. Hosts with higher performance have larger weights and thus receive more requests.

[0043] 6. Caching mechanism: A cache module is set up in the computing power host management platform to cache frequently accessed data and response results. When the same network request is received, data is preferentially retrieved from the cache and returned to the user, reducing repeated calculations and network transmissions and improving the response speed. The cache module uses the LRU (Least Recently Used) algorithm to manage the cache content, and timely eliminates cache data that has not been accessed for a long time to make room for new data.

[0044] The above are only the preferred embodiments of the present invention, which are only used to illustrate the technical solutions of the present invention and are not used to limit the protection scope of the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention are all included in the protection scope of the present invention.

Claims

1. A method for managing large-scale concurrent network requests based on a computing power host management platform, characterized in that: The following steps are involved: (1) Classification and prioritization of network requests; (2) Dynamically allocate resources; (3) Maintaining the network request queue; (4) Load balancing and failover; (5) Use cache to reduce duplicate requests.

2. The method according to claim 1, characterized in that Use a hash table to store processed requests; When a new request arrives, its hash value is calculated first. If the hash value already exists in the table, the request is discarded.

3. The method according to claim 2, characterized in that For requests with similar parameters, merge those parameters into a common request.

4. The method according to claim 1, characterized in that Classify the network requests entering the computing host management platform, and assign corresponding priorities to each request based on the request type, business scenario, and user attributes.

5. The method according to claim 1, characterized in that The computing host management platform monitors the resource usage of each host in real time, and dynamically allocates computing host resources to requests based on the priority of the request and the resource status of each host; When a high-priority request arrives, hosts with sufficient resources are assigned to the request first; For low-priority requests, the host's idle resources are used to process them while ensuring that high-priority requests are processed.

6. The method according to claim 1, characterized in that Establish several request queues and put requests into different queues according to their priorities. Each queue is processed according to the FIFO principle, but for high-priority queues, a higher scheduling frequency is given during processing. When the host resources are idle, the requests are taken out from the high priority queue first for processing to ensure that the high priority requests can be responded to in time. At the same time, a time threshold is set. When the waiting time of the request in the low priority queue exceeds the threshold, its processing priority is increased.

7. The method according to claim 1, characterized in that Use load balancing algorithm to evenly distribute concurrent requests to each computing host; When a host fails, the requests it is responsible for are transferred to other normal hosts for processing; The load balancing algorithm dynamically adjusts the request allocation strategy based on the real-time load and processing capacity of the host.

8. The method according to claim 1, characterized in that A cache module is set up in the computing host management platform to cache frequently accessed data and response results; when the same network request is received, data is obtained from the cache first and returned to the user.

9. The method according to claim 8, characterized in that The cache module uses the LRU algorithm to manage cache content, promptly eliminating cache data that has not been accessed for a long time to make room for new data.