Logging is essential for monitoring and troubleshooting running applications. This guide explains how to utilize logback to collect full request/response payloads in a Spring Boot application.
To begin, you create a Maven Project Object Model to enable logback. A Project Object Model or POM is an XML file that contains information about the project and configuration details. Below is a sample specifying the required dependencies.
<properties>
<project.build.sourceEncoding>UTF-8</project.build.sourceEncoding>
<project.reporting.outputEncoding>UTF-8</project.reporting.outputEncoding>
<java.version>1.8</java.version>
<logback.version>1.2.3</logback.version>
</properties>
<dependencies>
<dependency>
<groupId>net.rakugakibox.spring.boot</groupId>
<artifactId>logback-access-spring-boot-starter</artifactId>
<version>2.7.1</version>
</dependency>
<dependency>
<groupId>ch.qos.logback</groupId>
<artifactId>logback-access</artifactId>
<version>${logback.version}</version>
</dependency>
<dependency>
<groupId>ch.qos.logback</groupId>
<artifactId>logback-classic</artifactId>
<version>${logback.version}</version>
</dependency>
</dependencies>
Next create a logback-access.xml under src/main/resources.You can change the fields displayed in an access log. For a full list of available fields refer to the logback documentation.
<configuration>
<appender name="STDOUT" class="ch.qos.logback.core.ConsoleAppender">
<encoder>
<pattern>logging uri: %requestURL | status code: %statusCode | bytes: %bytesSent | elapsed time: %elapsedTime | request-log: %magenta(%requestContent) | response-log: %cyan(%responseContent)</pattern>
</encoder>
</appender>
<appender-ref ref="STDOUT"/>
</configuration>
It is often useful to capture both the client’s request and the server’s response when diagnosing bugs. The TeeFilter servlet filter accomplishes this.
import ch.qos.logback.access.servlet.TeeFilter;
@Configuration
public class FilterConfiguration {
@Autowired
@Bean
public FilterRegistrationBean requestResponseFilter() {
final FilterRegistrationBean filterRegBean = new FilterRegistrationBean();
TeeFilter filter = new TeeFilter();
filterRegBean.setFilter(filter);
filterRegBean.setUrlPatterns("/rest/path");
filterRegBean.setName("Request Response Filter");
filterRegBean.setAsyncSupported(Boolean.TRUE);
return filterRegBean;
}
}
Once this is configured, every request/response payload is logged to your default appender.
There are potential impacts to application performance when this filter is activated. Every request/response payload is copied to an in-memory buffer, creating additional garbage collection and CPU overhead. To reduce overhead or to avoid logging sensitive data, add the following to your application.properties to deactivate access logging by default:
logback.access.enabled=false
You can find out more by reading the Spring Boot documentation on logging or the full logback manual.
Logging has three elements: collection, indexing, and visualization. This guide explains the first element, collection. For indexing and visualization, there’s a wide ecosystem of open-source technologies that can be used. For example, the “EFK stack” (Elasticsearch, Fluentd, and Kibana) is popular for solving this problem.
Two open-source tools that help with logging and visualization are Prometheus and Grafana. Prometheus excels at gathering metrics from a wide array of sources, while Grafana is the go-to tool for visualizing complex time-series data. The following guides explain how to use these tools in Kubernetes environments:
Spring Boot also provides health checks for application monitoring in addition to logging. Learn how to enable health checks using Spring Boot Actuator.