Configuration and optimization of HTTP connection pool in SpringBoot

1. The core value of HTTP connection pool

In microservice architecture and distributed system scenarios, frequent creation/disconnection of HTTP clients can cause significant performance losses. Through connection pooling technology, it can be achieved:

Multiplexing TCP connections: Reduce the overhead of three handshakes and TLS negotiations

Resource control: Prevent system overload from burst traffic

Performance improvement: JMeter test shows that multiplexed connections can increase request throughput by 30%+

Connection management: Unified monitoring and exception handling

2. Spring Boot Integration Solution

Solution 1: Apache HttpClient (recommended)

1. Add dependencies

<dependency>
    <groupId></groupId>
    <artifactId>httpclient</artifactId>
    <version>4.5.13</version>
</dependency>

2. Connection pool configuration class

@Configuration
public class HttpPoolConfig {
    
    @Bean
    public CloseableHttpClient httpClient() {
        PoolingHttpClientConnectionManager cm = 
            new PoolingHttpClientConnectionManager();
        
        // Global maximum number of connections        (200); 
        // Maximum number of connections per route        (50);
        // Idle connection survival time (seconds)        (30_000);

        RequestConfig requestConfig = ()
                .setConnectTimeout(5000)
                .setSocketTimeout(10000)
                .build();
        return ()
                .setConnectionManager(cm)
                .setDefaultRequestConfig(requestConfig)
                .build();
    }

​​​​​​​    @Bean
    public RestTemplate restTemplate() {
        return new RestTemplate(new HttpComponentsClientHttpRequestFactory(httpClient()));
    }
}

Solution 2: OkHttp3

@Bean
public OkHttpClient okHttpClient() {
    return new ()
        .connectionPool(new ConnectionPool(100, 5, ))
        .connectTimeout((5))
        .readTimeout((10))
        .retryOnConnectionFailure(true)
        .build();
}

3. Detailed explanation of key configuration parameters

4. Best practices in production environment

Capacity planning:

• Calculation formula: Maximum number of connections = QPS × Average response time (seconds)
• Reserve 20% of the buffer space to deal with the peak flow

Monitoring alarm:

// Get the connection pool statusPoolingHttpClientConnectionManager mgr = 
    (PoolingHttpClientConnectionManager) ();
("Available Connections:" + ().getAvailable());
("Rent connection:" + ().getLeased());

Exception handling:

try {
    // Business request} catch (ConnectTimeoutException e) {
    // Connection timeout processing} catch (SocketTimeoutException e) {
    // Read and write timeout processing} finally {
    // Make sure to release the connection}

Lifecycle management:

@PreDestroy
public void destroy() {
    ();
}

5. Performance optimization skills

Connection warm-up: Partial connection is established in advance when the system starts

Dynamic parameter adjustment: implement configuration hot update based on Apollo/Nacos

DNS cache: Set a reasonable DNS refresh policy

Connection Evicting: Regularly clean out abnormal connections

6. Troubleshooting of FAQs

Check server load

Confirm the firewall policy

Verify DNS resolution

Increase maxTotal value

Optimize server response time

Add circuit downgrade strategy

3. Connection leak

Use() to ensure release

Add a connection tracking log

Integrated LeakCanary detection tool

7. New generation solutions

For responsive programming scenarios, it is recommended to use WebClient + Reactor Netty:

@Bean
public WebClient webClient() {
    return ()
        .clientConnector(new ReactorClientHttpConnector(
            ()
                .baseUrl("")
                .responseTimeout((5))
                .runOn(("http-loop", 4, true))
        ))
        .build();
}

By rationally configuring the HTTP connection pool, the network communication performance of Spring Boot applications can be significantly improved. It is recommended to conduct stress testing based on actual business scenarios and continuously optimize the connection pool parameter configuration.

This is the end of this article about the configuration and optimization of HTTP connection pools in SpringBoot. For more related contents of SpringBoot HTTP connection pools, please search for my previous articles or continue browsing the related articles below. I hope everyone will support me in the future!