Back to Guides

Logging configuration

Read about the use of logging API in Quarkus, configuring logging output, and using logging adapters to unify the output from other logging APIs.

This document is part of the Observability in Quarkus reference guide which features this and other observability related components.

Quarkus uses the JBoss Log Manager logging backend for publishing application and framework logs. Quarkus supports the JBoss Logging API and multiple other logging APIs, seamlessly integrated with JBoss Log Manager.

This means you can use your favorite logging API with Quarkus, provided you add the appropriate adapter to your application:

Logging API Adapter

JBoss Logging

Built-in

JDK java.util.logging (JUL)

Built-in

SLF4J

SLF4J adapter

Apache Commons Logging

Commons Logging adapter

Apache Log4j 2

Log4j 2 adapter

Use JBoss Logging for application logging

When using the JBoss Logging API, your application requires no additional dependencies, as Quarkus provides it automatically.

An example of using the JBoss Logging API to log a message:
import org.jboss.logging.Logger;

import jakarta.ws.rs.GET;
import jakarta.ws.rs.Path;
import jakarta.ws.rs.Produces;
import jakarta.ws.rs.core.MediaType;

@Path("/hello")
public class ExampleResource {

    private static final Logger LOG = Logger.getLogger(ExampleResource.class);

    @GET
    @Produces(MediaType.TEXT_PLAIN)
    public String hello() {
        LOG.info("Hello");
        return "hello";
    }
}
While JBoss Logging routes log messages directly to the JBoss Log Manager, one of your libraries might rely on a different logging API. In such cases, you need to use a logging adapter to ensure that its log messages are routed to JBoss Log Manager as well.

Get an application logger

To get an application logger in Quarkus, select one of the following approaches.

Declaring a logger field

With this classic approach, you use a specific API to obtain a logger instance, store it in a static field of a class, and call logging operations upon this instance.

The same flow can be applied with any of the supported logging APIs.

An example of storing a logger instance into a static field by using the JBoss Logging API:
package com.example;

import org.jboss.logging.Logger;

public class MyService {
    private static final Logger log = Logger.getLogger(MyService.class); (1)

    public void doSomething() {
        log.info("It works!"); (2)
    }
}
1 Define the logger field.
2 Invoke the desired logging methods on the log object.

When you create a Logger instance, Quarkus keeps metadata associated with the Logger name in memory for the full lifecycle of the application.

Do not use an unbounded dynamic name for loggers, for example, for (int i …​) { Logger.getLogger("logger-name." + i).info("my log") }.

Simplified logging

Quarkus simplifies logging by automatically adding logger fields to classes that use io.quarkus.logging.Log. This eliminates the need for repetitive boilerplate code and enhances the convenience of logging setup.

An example of simplified logging with static method calls:
package com.example;

import io.quarkus.logging.Log; (1)

class MyService { (2)
    public void doSomething() {
        Log.info("Simple!"); (3)
    }
}
1 The io.quarkus.logging.Log class provides the same methods as JBoss Logging, but as static methods.
2 The class does not declare a logger field. During the application build, Quarkus creates a private static final org.jboss.logging.Logger field in each class that uses the Log API. Quarkus uses the fully qualified name of the calling class as the logger name. In this example, the logger name is com.example.MyService.
3 During the application build, Quarkus rewrites calls to Log methods into JBoss Logging calls on the generated logger field.

Use the Log API only in application classes, not in external dependencies. Log method calls not processed by Quarkus at build time will result in an exception.

Using io.quarkus.logging.Log in extensions:

While the Log API simplifies logging in application classes, do not use it in extension modules or external dependencies.

The following considerations apply:

  • io.quarkus.logging.Log depends on Quarkus bytecode transformations that occur at build time.

  • In extension modules, Log works only if the module includes a Jandex index. However, this behavior is unsupported and can lead to unreliable logging.

    For extension development, use org.jboss.logging.Logger.getLogger(String) instead of io.quarkus.logging.Log.

Injecting a configured logger

Instead of declaring a logger field, you can use an alternative approach and inject a configured org.jboss.logging.Logger instance by using @Inject. This approach applies only to CDI beans.

To control the logger name, choose one of the following injection patterns:

  • Use @Inject Logger log when you want Quarkus to name the logger after the class it is injected into.

  • Use @LoggerName("…​") Logger log when you want to set the logger name explicitly.

@Inject is not required if you already annotated your Logger with @LoggerName("…​").

Once injected, you can use the log object to invoke logging methods.

An example of two different types of logger injection:
package com.example;

import org.jboss.logging.Logger;

@ApplicationScoped
class SimpleBean {

   @Inject
   Logger log; (1)

   @LoggerName("foo")
   Logger fooLog; (2)

   public void ping() {
     log.info("Simple!");
     fooLog.info("Goes to _foo_ logger!");
   }
}
1 The logger is created with org.jboss.logging.Logger.getLogger(SimpleBean.class).
2 In this case, the name "foo" is used as a logger name. The logger is created with org.jboss.logging.Logger.getLogger("foo").

Quarkus caches logger instances internally. When you inject a logger into a @RequestScoped bean, Quarkus shares that logger across bean instances to avoid performance penalties from repeated logger instantiation.

Use log levels

Quarkus provides multiple log levels to control the amount of information logged based on event severity.

Available log levels:
OFF

A special level used in configuration to turn off logging.
FATAL

A critical service failure or total inability to handle any requests.
ERROR

A major issue in processing or an inability to complete a request.
WARN

A non-critical service error or problem that might not require immediate correction.
INFO

Service lifecycle events or other important infrequent information.
DEBUG

Additional information about lifecycle events or events not tied to specific requests, useful for debugging.
TRACE

Detailed per-request debugging information, potentially at a very high frequency.
ALL

A special level to turn on logging for all messages, including custom levels.

You can also configure the following levels for applications and libraries that use java.util.logging:

SEVERE

Same as ERROR.
WARNING

Same as WARN.
CONFIG

Service configuration information.
FINE

Same as DEBUG.
FINER

Same as TRACE.
FINEST

Increased debug output compared to TRACE, which might have a higher frequency.
Table 1. The mapping between the levels
Numerical level value Standard level name Equivalent java.util.logging (JUL) level name

1100

FATAL

Not applicable

1000

ERROR

SEVERE

900

WARN

WARNING

800

INFO

INFO

700

Not applicable

CONFIG

500

DEBUG

FINE

400

TRACE

FINER

300

Not applicable

FINEST

Configure the log level, category, and format

JBoss Logging, integrated into Quarkus, provides a unified configuration for all supported logging APIs through a single configuration file that sets up all available extensions.

To adjust runtime logging, modify the application.properties file.

An example of how you can set the default log level to INFO logging and include Hibernate DEBUG logs:
quarkus.log.level=INFO
quarkus.log.category."org.hibernate".level=DEBUG

Assuming the Quarkus YAML extension is included, the equivalent YAML configuration is:

quarkus:
  log:
    level: INFO
    category:
      org.hibernate: # Using org.hibernate as a single key is required
        level: DEBUG

When you set the log level to below DEBUG, for example TRACE, you must also adjust the minimum log level.

Set the minimum log level globally with the quarkus.log.min-level configuration property, or set it per category:

quarkus.log.category."org.hibernate".min-level=TRACE

The minimum log level is a build-time configuration option that sets the floor level for which Quarkus generates supporting code. This enables further build-time optimizations, such as dead code elimination when building a native image.

An example from a native executable:

Setting INFO as the minimum log level forces checks for lower levels, such as isTraceEnabled(), always to return false.

In turn, code such as if (logger.isDebugEnabled()) callMethod(); is identified as never executed, also known as dead code.

If you add these properties on the command line, ensure the " character is escaped with a backslash:

-Dquarkus.log.category.\"org.hibernate\".level=TRACE

All potential properties are listed in the logging configuration reference section.

Logging categories

Logging is configured per category. Configuration for a category applies recursively to all subcategories unless a more specific subcategory configuration exists.

The parent of all logging categories is called the "root category." As the ultimate parent, this category might contain a configuration that applies globally to all other categories. This includes the globally configured handlers and formatters.

Example 1. An example of a global configuration that applies to all categories:
quarkus.log.handlers=con,mylog

quarkus.log.handler.console.con.enable=true
quarkus.log.handler.file.mylog.enable=true

In this example, the root category is configured to use two named handlers: con and mylog.

Example 2. An example of a per-category configuration:
quarkus.log.category."org.apache.kafka.clients".level=INFO
quarkus.log.category."org.apache.kafka.common.utils".level=INFO

This example shows how to configure the minimal log level for the categories org.apache.kafka.clients and org.apache.kafka.common.utils.

For more information, see Logging configuration reference.

If you want to configure something extra for a specific category, create a named handler like quarkus.log.handler.[console|file|syslog].<your-handler-name>.* and set it up for that category by using quarkus.log.category.<my-category>.handlers.

An example use case can be a desire to use a different timestamp format for log messages that are saved to a file than the format used for other handlers.

For further demonstration, see the outputs of the Attaching named handlers to a category example.

Property Name Default Description

quarkus.log.category."<category-name>".level

INFO [1]

The level to use to configure the category named <category-name>. The quotes are necessary.

quarkus.log.category."<category-name>".min-level

DEBUG

The minimum logging level to use to configure the category named <category-name>. The quotes are necessary.

quarkus.log.category."<category-name>".use-parent-handlers

true

Specify whether this logger should send its output to its parent logger.

quarkus.log.category."<category-name>".handlers=[<handler>]

empty [2]

The names of the handlers that you want to attach to a specific category.

The . symbol separates the specific parts in the configuration property. The quotes in the property name are used as a required escape to keep category specifications, such as quarkus.log.category."io.quarkus.smallrye.jwt".level=TRACE, intact.

Root logger configuration

The root logger category is handled separately, and is configured by using the following properties:

Property Name Default Description

quarkus.log.level

INFO

The default log level for every log category.

quarkus.log.min-level

DEBUG

The default minimum log level for every log category.

  • The parent category is examined if no level configuration exists for a given logger category.

  • The root logger configuration is used if no specific configurations are provided for the category and any of its parent categories.

Although the root logger’s handlers are usually configured directly via quarkus.log.console, quarkus.log.file and quarkus.log.syslog, it can nonetheless have additional named handlers attached to it using the quarkus.log.handlers property.

Logging format

Human-readable text

Quarkus uses a pattern-based logging formatter that generates human-readable text logs by default, but you can also configure the format for each log handler by using a dedicated property.

For the console handler, the property is quarkus.log.console.format.

The logging format string supports the following symbols:

Symbol Summary Description

%%

%

Renders a simple % character.

%c

Category

Renders the category name.

%C

Source class

Renders the source class name.[3]

%d{xxx}

Date

Renders a date with the given date format string, which uses the syntax defined by java.text.SimpleDateFormat.

%e

Exception

Renders the thrown exception, if any.

%F

Source file

Renders the source file name.[3]

%h

Host name

Renders the system simple host name.

%H

Qualified host name

Renders the system’s fully qualified host name, which might be the same as the simple host name, depending on operating system configuration.

%i

Process ID

Render the current process PID.

%l

Source location

Renders the source location information, which includes source file name, line number, class name, and method name.[3]

%L

Source line

Renders the source line number.[3]

%m

Full Message

Renders the log message plus exception (if any).

%M

Source method

Renders the source method name.[3]

%n

Newline

Renders the platform-specific line separator string.

%N

Process name

Render the name of the current process.

%p

Level

Render the log level of the message.

%r

Relative time

Render the time in milliseconds since the start of the application log.

%s

Simple message

Renders just the log message, with no exception trace.

%t

Thread name

Render the thread name.

%t{id}

Thread ID

Render the thread ID.

%z{<zone name>}

Time zone

Set the time zone of the output to <zone name>.

%X{<MDC property name>}

Mapped Diagnostic Context Value

Renders the value from Mapped Diagnostic Context.

%X

Mapped Diagnostic Context Values

Renders all the values from Mapped Diagnostic Context in format {property.key=property.value}.

%x

Nested Diagnostics context values

Renders all the values from Nested Diagnostics Context in format {value1.value2}.

JSON logging format

Add the quarkus-logging-json extension to enable the JSON logging format and its related configuration.

  1. Add this extension to your build file as the following snippet illustrates:

    pom.xml
    <dependency>
    <groupId>io.quarkus</groupId>
    <artifactId>quarkus-logging-json</artifactId>
</dependency>
    build.gradle
    implementation("io.quarkus:quarkus-logging-json")

    By default, this extension overrides the console output format configuration, and any format string or color settings are ignored. Other console configuration items continue to apply, including items that control asynchronous logging and the log level.

    If you prefer human-readable, unstructured logging in dev mode and JSON-structured logging in production mode, configure this by using profiles, as shown in the following configuration.

  2. Disable JSON logging in application.properties for dev and test mode:

    %dev.quarkus.log.console.json.enabled=false
%test.quarkus.log.console.json.enabled=false

  3. Choose the JSON console log format by setting the configuration property:

    quarkus.log.console.json.log-format

    This property sets the JSON log format for the console.

    The possible values are:

    • default: Generates structured logs based on the key-value pairs present in the log record. Mapped Diagnostic Context (MDC) and Nested Diagnostic Context (NDC) data are included under the mdc and ndc fields.

      Set quarkus.log.console.json.mdc.flat-fields=true to write MDC entries as root-level JSON fields instead of nesting them under mdc. This is useful when a log aggregator (such as OpenSearch or Elasticsearch) needs MDC fields to be queryable as top-level document fields.

      Example: nested MDC (default, flat-fields=false)
      {"timestamp":"...","message":"Handling request","mdc":{"requestId":"abc123","userId":"42"}}
      Example: flat MDC (flat-fields=true)
      {"timestamp":"...","message":"Handling request","requestId":"abc123","userId":"42"}

      The option is available for all handler types:

      • console: quarkus.log.console.json.mdc.flat-fields

      • file: quarkus.log.file.json.mdc.flat-fields

      • syslog: quarkus.log.syslog.json.mdc.flat-fields

      • socket: quarkus.log.socket.json.mdc.flat-fields

      • Named handlers, such as quarkus.log.handler.console.<name>.json.mdc.flat-fields

    • ecs: Uses the Elastic Common Schema (ECS) format. This format modifies some default field names and adds other fields to align with ECS.

      Fields include @timestamp, log.logger, log.level, process.pid, process.name, process.thread.name, process.thread.id, host.hostname, event.sequence, error.message, error.stack_trace, ecs.version, data_stream.type, service.name, service.version, and service.environment.

      MDC data is kept in the mdc field, which is not part of the ECS schema. Set mdc.flat-fields to write those entries as root-level fields instead, or add mdc to excluded-keys to leave them out entirely.

    • gcp: Uses the Google Cloud format. This format follows the default format.

      When you use OpenTelemetry, Quarkus flattens tracing data present in the mdc field and copies it to spanId, traceSampled, and trace, with the trace value including a prefix.

      Google Cloud requires the trace field to follow the "projects/<my-trace-project>/traces/12345" format. <my-trace-project> uses the value of the quarkus.application.name configuration property.

Custom JSON fields with JsonProvider

The JsonProvider SPI allows you to dynamically inject additional JSON fields into every log record at formatting time. Each registered provider’s writeTo method is called once per log record, receiving a JsonLogGenerator and the ExtLogRecord being formatted.

Providers can inspect the log record (level, logger name, MDC values, message, etc.) to decide which fields, if any, to add. Any keys listed in quarkus.log.*.json.excluded-keys are still filtered, even when written by a provider.

CDI registration

In a Quarkus application, the preferred way to register a JsonProvider is as a CDI bean:

import jakarta.enterprise.context.ApplicationScoped;
import org.jboss.logmanager.ExtLogRecord;
import org.jboss.logmanager.Level;
import io.quarkus.logging.json.runtime.JsonProvider;
import io.quarkus.logging.json.runtime.JsonFormatter.JsonLogGenerator;

@ApplicationScoped
public class MyJsonProvider implements JsonProvider {

    @Override
    public void writeTo(JsonLogGenerator generator, ExtLogRecord record) throws Exception {
        // Add a static field to every log record
        generator.add("environment", "production");

        // Conditionally add a field based on the log level
        if (record.getLevel().intValue() >= Level.ERROR.intValue()) {
            generator.add("alert", "true");
        }

        // Add an MDC value as a top-level field
        String requestId = record.getMDC("requestId");
        if (requestId != null) {
            generator.add("requestId", requestId);
        }

        // Add a nested JSON object
        generator.startObject("metadata")
                .add("version", "1.0")
                .add("region", "us-east-1")
                .endObject();
    }
}
ServiceLoader registration

For providers that must work without CDI (for example, in libraries or during early startup), register them via the Java ServiceLoader mechanism. Create a file named META-INF/services/io.quarkus.logging.json.runtime.JsonProvider containing the fully qualified class name of each implementation:

com.example.MyJsonProvider

The implementation does not need to be a CDI bean:

import io.quarkus.logging.json.runtime.JsonProvider;
import io.quarkus.logging.json.runtime.JsonFormatter.JsonLogGenerator;
import org.jboss.logmanager.ExtLogRecord;

public class MyJsonProvider implements JsonProvider {

    @Override
    public void writeTo(JsonLogGenerator generator, ExtLogRecord record) throws Exception {
        generator.add("source", "my-library");
    }
}
When both CDI beans and ServiceLoader providers are present, all of them are called for each log record. ServiceLoader providers are called first, in the order they appear in the META-INF/services/ configuration file. CDI beans are called afterwards, but their relative order is not guaranteed.

Configuration

Configure the JSON logging extension using supported properties to customize its behavior.

Enabling pretty printing might cause certain processors and JSON parsers to fail.
Printing the details can be expensive as the values are retrieved from the caller. The details include the source class name, source file name, source method name, and source line number.

Log handlers

A log handler is a logging component that emits log events to a recipient. Quarkus includes several different log handlers: console, file, and syslog.

The featured examples use com.example as a logging category.

Console log handler

The console log handler is enabled by default, and it directs all log events to the application’s console, usually the system’s stdout.

  • A global configuration example:

    quarkus.log.console.format=%d{yyyy-MM-dd HH:mm:ss} %-5p [%c] (%t) %s%e%n

  • A per-category configuration example:

    quarkus.log.handler.console.my-console-handler.format=%d{yyyy-MM-dd HH:mm:ss} [com.example] %s%e%n

quarkus.log.category."com.example".handlers=my-console-handler
quarkus.log.category."com.example".use-parent-handlers=false

For details about its configuration, see the console logging configuration reference.

File log handler

To log events to a file on the application’s host, use the Quarkus file log handler. The file log handler is disabled by default, so you must enable it first.

The Quarkus file log handler supports log file rotation.

Log file rotation ensures efficient log management by preserving a specified number of backup files while keeping the primary log file up to date and manageable in size.

  • A global configuration example:

    quarkus.log.file.enabled=true
quarkus.log.file.path=application.log
quarkus.log.file.format=%d{yyyy-MM-dd HH:mm:ss} %-5p [%c] (%t) %s%e%n

  • A per-category configuration example:

    quarkus.log.handler.file.my-file-handler.enabled=true
quarkus.log.handler.file.my-file-handler.path=application.log
quarkus.log.handler.file.my-file-handler.format=%d{yyyy-MM-dd HH:mm:ss} [com.example] %s%e%n

quarkus.log.category."com.example".handlers=my-file-handler
quarkus.log.category."com.example".use-parent-handlers=false

For details about its configuration, see the file logging configuration reference.

Syslog log handler

The syslog handler in Quarkus follows the Syslog protocol, which is used to send log messages on UNIX-like systems. It uses the protocol defined in RFC 5424.

By default, the syslog handler is disabled. When enabled, it sends all log events to a syslog server, typically the local syslog server for the application.

  • A global configuration example:

    quarkus.log.syslog.enabled=true
quarkus.log.syslog.app-name=my-application
quarkus.log.syslog.format=%d{yyyy-MM-dd HH:mm:ss} %-5p [%c] (%t) %s%e%n

  • A per-category configuration example:

    quarkus.log.handler.syslog.my-syslog-handler.enabled=true
quarkus.log.handler.syslog.my-syslog-handler.app-name=my-application
quarkus.log.handler.syslog.my-syslog-handler.format=%d{yyyy-MM-dd HH:mm:ss} [com.example] %s%e%n

quarkus.log.category."com.example".handlers=my-syslog-handler
quarkus.log.category."com.example".use-parent-handlers=false

For details about its configuration, see the Syslog logging configuration reference.

Socket log handler

This handler sends logs to a socket. The socket log handler is disabled by default; enable it in your configuration. When enabled, it sends all log events to a socket, such as a Logstash server.

  • A global configuration example:

    quarkus.log.socket.enabled=true
quarkus.log.socket.endpoint=localhost:4560

Typically, this handler is used with the quarkus-logging-json extension to send logs in ECS format to an Elasticsearch instance. For an example configuration, see the Centralized log management guide.

Add a logging filter to your log handler

You can attach a filter to a log handler such as the console handler. The filter determines whether the handler logs a given log record.

To register a logging filter:

  1. Annotate a final class that implements java.util.logging.Filter with @io.quarkus.logging.LoggingFilter, and set the name property:

    An example of writing a filter:
    package com.example;

import io.quarkus.logging.LoggingFilter;
import java.util.logging.Filter;
import java.util.logging.LogRecord;

@LoggingFilter(name = "my-filter")
public final class TestFilter implements Filter {

    private final String part;

    public TestFilter(@ConfigProperty(name = "my-filter.part") String part) {
        this.part = part;
    }

    @Override
    public boolean isLoggable(LogRecord record) {
        return !record.getMessage().contains(part);
    }
}

    In this example, the filter excludes log records containing specific text from console logs. The specific text to filter on is not hard-coded; instead, it is read from the my-filter.part configuration property.

    An example of Configuring the filter in application.properties:
    my-filter.part=TEST

  2. Attach the filter to the corresponding handler using the filter configuration property, located in application.properties:

    quarkus.log.console.filter=my-filter

Examples of logging configurations

The following examples show how to configure logging in Quarkus:

Console DEBUG logging except for Quarkus logs (INFO), no color, shortened time, shortened category prefixes
quarkus.log.console.format=%d{HH:mm:ss} %-5p [%c{2.}] (%t) %s%e%n
quarkus.log.console.level=DEBUG
quarkus.console.color=false

quarkus.log.category."io.quarkus".level=INFO

If you add these properties on the command line, ensure the " character is escaped with a backslash. For example, -Dquarkus.log.category.\"io.quarkus\".level=DEBUG.
File TRACE logging configuration
quarkus.log.file.enabled=true
# Send output to a trace.log file under the /tmp directory
quarkus.log.file.path=/tmp/trace.log
quarkus.log.file.level=TRACE
quarkus.log.file.format=%d{HH:mm:ss} %-5p [%c{2.}] (%t) %s%e%n
# Set 2 categories (io.quarkus.smallrye.jwt, io.undertow.request.security) to TRACE level
quarkus.log.min-level=TRACE
quarkus.log.category."io.quarkus.smallrye.jwt".level=TRACE
quarkus.log.category."io.undertow.request.security".level=TRACE
Because this example does not change the root logger, the console log contains only INFO or higher level logs.
Named handlers attached to a category
# Send output to a trace.log file under the /tmp directory
quarkus.log.file.path=/tmp/trace.log
quarkus.log.console.format=%d{HH:mm:ss} %-5p [%c{2.}] (%t) %s%e%n
# Configure a named handler that logs to console
quarkus.log.handler.console."STRUCTURED_LOGGING".format=%e%n
# Configure a named handler that logs to file
quarkus.log.handler.file."STRUCTURED_LOGGING_FILE".enabled=true
quarkus.log.handler.file."STRUCTURED_LOGGING_FILE".format=%e%n
# Configure the category and link the two named handlers to it
quarkus.log.category."io.quarkus.category".level=INFO
quarkus.log.category."io.quarkus.category".handlers=STRUCTURED_LOGGING,STRUCTURED_LOGGING_FILE
Named handlers attached to the root logger
# configure a named file handler that sends the output to 'quarkus.log'
quarkus.log.handler.file.CONSOLE_MIRROR.enabled=true
quarkus.log.handler.file.CONSOLE_MIRROR.path=quarkus.log
# attach the handler to the root logger
quarkus.log.handlers=CONSOLE_MIRROR

Centralized log management

Use a centralized location to efficiently collect, store, and analyze log data from various components and application instances.

To send logs to a centralized tool such as Graylog, Logstash, or Fluentd, see the Quarkus Centralized log management guide.

OpenTelemetry logging

You can enable sending log entries from all appenders to OpenTelemetry Logging.

For details, see the Quarkus OpenTelemetry Logging guide.

Configure logging for @QuarkusTest

To enable logging for @QuarkusTest, it is necessary to set the java.util.logging.manager system property to org.jboss.logmanager.LogManager early in the startup process, or logging will not work as expected.

Configure it in your build tool as follows:

  • For Maven:

    Set the property in the Maven Surefire plugin configuration.

    Setting java.util.logging.manager in the Maven Surefire plugin configuration:
    <build>
  <plugins>
    <plugin>
      <artifactId>maven-surefire-plugin</artifactId>
      <version>${surefire-plugin.version}</version>
      <configuration>
        <systemPropertyVariables>
          <java.util.logging.manager>org.jboss.logmanager.LogManager</java.util.logging.manager> (1)
          <quarkus.log.level>DEBUG</quarkus.log.level> (2)
          <maven.home>${maven.home}</maven.home>
        </systemPropertyVariables>
      </configuration>
    </plugin>
  </plugins>
</build>
    1 Make sure the org.jboss.logmanager.LogManager is used.
    2 Enable debug logging for all logging categories.

  • For Gradle:

    Set the property in the build.gradle file.

    test {
    systemProperty "java.util.logging.manager", "org.jboss.logmanager.LogManager"
}

For more information, see the Running @QuarkusTest from an IDE section of the Testing your Application guide.

Use other logging APIs

Quarkus relies on the JBoss Logging library for logging. If your dependencies use other logging APIs, such as Apache Commons Logging, Log4j, or SLF4J, exclude those libraries and use a JBoss Logging adapter. This step is especially important for native executables, because the native build can fail with errors such as the following:

Caused by java.lang.ClassNotFoundException: org.apache.commons.logging.impl.LogFactoryImpl

The native executable does not include the logging implementation for those APIs. Use JBoss Logging adapters to resolve this issue.

The next section describes adapters for common open-source logging APIs.

Add a logging adapter to your application

For each logging API that is not JBoss Logging, add a logging adapter library to ensure that messages logged through these APIs are routed to the JBoss Log Manager backend. You can then verify whether the logs generated by the added library adhere to the same format as the other Quarkus logs.

This step is unnecessary for libraries that are dependencies of a Quarkus extension where the extension handles it automatically.

SLF4J

The SLF4J adapter is provided by the following dependency:

pom.xml
<dependency>
    <groupId>org.jboss.slf4j</groupId>
    <artifactId>slf4j-jboss-logmanager</artifactId>
</dependency>
build.gradle
implementation("org.jboss.slf4j:slf4j-jboss-logmanager")

Apache Commons Logging

The Commons Logging adapter is provided by the following dependency:

pom.xml
<dependency>
    <groupId>org.jboss.logging</groupId>
    <artifactId>commons-logging-jboss-logging</artifactId>
</dependency>
build.gradle
implementation("org.jboss.logging:commons-logging-jboss-logging")

Log4j 2

The Log4j 2 adapter is provided by the following dependency:

pom.xml
<dependency>
    <groupId>org.jboss.logmanager</groupId>
    <artifactId>log4j2-jboss-logmanager</artifactId>
</dependency>
build.gradle
implementation("org.jboss.logmanager:log4j2-jboss-logmanager")

Do not include any Log4j dependencies, as the log4j2-jboss-logmanager library contains everything needed to use Log4j as a logging implementation.

Use MDC to add contextual log information

Quarkus overrides the logging Mapped Diagnostic Context (MDC) to improve compatibility with its reactive core.

Add and read MDC data

To add data to the MDC and extract it in your log output:

  1. Use the MDC class to set the data.

    1. Add import org.jboss.logmanager.MDC;

    2. Set MDC.put(…​) as shown in the example below:

      An example with JBoss Logging and io.quarkus.logging.Log
      package me.sample;

import io.quarkus.logging.Log;
import jakarta.ws.rs.GET;
import jakarta.ws.rs.Path;
import org.jboss.logmanager.MDC;

import java.util.UUID;

@Path("/hello/jboss")
public class GreetingResourceJbossLogging {

    @GET
    @Path("/test")
    public String greeting() {
        MDC.put("request.id", UUID.randomUUID().toString());
        MDC.put("request.path", "/hello/test");
        Log.info("request received");
        return "hello world!";
    }
}

  2. Configure the log format to use %X{mdc-key}:

    quarkus.log.console.format=%d{HH:mm:ss} %-5p request.id=%X{request.id} request.path=%X{request.path} [%c{2.}] (%t) %s%n

    The resulting message contains the MDC data:

    08:48:13 INFO request.id=c37a3a36-b7f6-4492-83a1-de41dbc26fe2 request.path=/hello/test [me.sa.GreetingResourceJbossLogging] (executor-thread-1) request received

MDC and supported logging APIs

Based on your logging API, use one of the following MDC classes:

  • Log4j 2 - org.apache.logging.log4j.ThreadContext.put(key, value)

  • SLF4J - org.slf4j.MDC.put(key, value)

MDC propagation

In Quarkus, the MDC provider has a specific implementation for handling the reactive context, ensuring that MDC data is propagated during reactive and asynchronous processing.

As a result, you can still access the MDC data in various scenarios:

  • After asynchronous calls, for example, when a REST client returns a Uni.

  • In code submitted to org.eclipse.microprofile.context.ManagedExecutor.

  • In code executed with vertx.executeBlocking().

If applicable, Quarkus stores MDC data in a Vert.x duplicated context, which isolates processing for a single task or request.

Logging configuration reference

1. Some extensions might define customized default log levels for certain categories to reduce log noise by default. Setting the log level in configuration will override any extension-defined log levels.
2. By default, the configured category gets the same handlers attached as the one on the root logger.
3. Format sequences which examine caller information might affect performance

Quarkus is open. All dependencies of this project are available under the Apache Software License 2.0 or compatible license. CC by 3.0

This website was built with Roq on Quarkus, is hosted on GitHub Pages and is completely open source. If you want to make it better, fork the website and show us what you’ve got.

Navigation
Follow Us
Get Help
Languages
Quarkus is made of community projects
Copyright © Quarkus. All rights reserved. For details on our trademarks, please visit our Trademark Policy and Trademark List. Trademarks of third parties are owned by their respective holders and their mention here does not suggest any endorsement or association.