Why doesn't the Apache Kafka consumer use the Log4j2 root logger?

Brad Urani :

I have this config:

pom.xml

<dependency>
  <groupId>org.apache.logging.log4j</groupId>
  <artifactId>log4j-api</artifactId>
  <version>2.11.1</version>
</dependency>
<dependency>
  <groupId>org.apache.logging.log4j</groupId>
  <artifactId>log4j-core</artifactId>
  <version>2.11.1</version>
</dependency>

<dependency>
  <groupId>org.apache.kafka</groupId>
  <artifactId>kafka-clients</artifactId>
  <version>1.0.1</version>
</dependency>

resources/log4j2.properties

name=PropertiesConfig
appenders = console
appender.console.type = Console
appender.console.name = STDOUT
appender.console.layout.type = PatternLayout
appender.console.layout.pattern = [%-5level] %d{yyyy-MM-dd HH:mm:ss.SSS} [%t] %c{1} - %msg%n
rootLogger.level = ERROR
rootLogger.appenderRefs = stdout
rootLogger.appenderRef.stdout.ref = STDOUT

When I start my Kafka consumer using org.apache.kafka.clients.consumer.KafkaConsumer, it's logging at the INFO level. Any new logger I create using LogManager.getLogger(name) honors the configuration and logs at level ERROR, so I know the log4j2.properties file is being loaded and used. Kafka seems to be ignoring it though. In the logs, when the Kafka consumer is starting up, I see this line a few time:

Aug 20, 2018 11:03:37 PM org.apache.kafka.common.utils.LogContext$KafkaLogger info

Any help appreciated!

Thilo :

Kafka logs via SLF4J. If you want to use Log4j2, you have to configure the two to work together (by adding the SLF4J implementation provided by Log4j2):

<dependency>
    <groupId>org.apache.logging.log4j</groupId>
    <artifactId>log4j-slf4j-impl</artifactId>
    <version>2.11.1</version>
</dependency>

Without this, log output from SLF4J will use some other backend (make sure you remove that one, you should not have any other slf4j-impl on your classpath) and ignore your Log4j2 configuration.