如何在yarn上配置log4j？

如何解决如何在yarn上配置log4j？

背景：纱线由cloudera管理器安装。我使用yarn运行我的xxx.jar。

我使用yarn运行我的xxx.jar。众所周知，syslog文件将包含与log4j相关的日志。 stdout文件将包含一些输出，例如System.out.println（...）。打击是放入stdout文件中的文件：

===============================================================
LogType:stdout
LogLastModifiedTime:Sun Oct 11 21:19:27 +0800 2020
LogLength:109238
LogContents:
log4j: Trying to find [container-log4j.properties] using context classloader sun.misc.Launcher$AppClassLoader@5c647e05.
log4j: Using URL [jar:file:/usr/lib/hadoop-yarn/hadoop-yarn-server-nodemanager-3.0.0-cdh6.3.2.jar!/container-log4j.properties] for automatic log4j configuration.
log4j: Reading configuration from URL jar:file:/usr/lib/hadoop-yarn/hadoop-yarn-server-nodemanager-3.0.0-cdh6.3.2.jar!/container-log4j.properties
log4j: Hierarchy threshold set to [ALL].
log4j: Parsing for [root] with value=[INFO,CLA,EventCounter].
log4j: Level token is [INFO].
log4j: Category root set to INFO
log4j: Parsing appender named "CLA".
log4j: Parsing layout options for "CLA".
log4j: Setting property [conversionPattern] to [%d{ISO8601} %p [%t] %c: %m%n].
log4j: End of parsing for "CLA".
log4j: Setting property [containerLogFile] to [syslog].
log4j: Setting property [totalLogFileSize] to [1048576].
log4j: Setting property [containerLogDir] to [/yarn/container-logs/application_1602420941906_0002/container_1602420941906_0002_01_000001].
log4j: setFile called: /yarn/container-logs/application_1602420941906_0002/container_1602420941906_0002_01_000001/syslog,true
log4j: setFile ended
log4j: Parsed "CLA" options.
log4j: Parsing appender named "EventCounter".
log4j: Parsed "EventCounter" options.
log4j: Parsing for [org.apache.hadoop.mapreduce.task.reduce] with value=[INFO,CLA].
log4j: Level token is [INFO].
log4j: Category org.apache.hadoop.mapreduce.task.reduce set to INFO
log4j: Parsing appender named "CLA".
log4j: Appender "CLA" was already parsed.
log4j: Handling log4j.additivity.org.apache.hadoop.mapreduce.task.reduce=[false]
log4j: Setting additivity for "org.apache.hadoop.mapreduce.task.reduce" to false
log4j: Parsing for [org.apache.hadoop.mapred.Merger] with value=[INFO,CLA].
log4j: Level token is [INFO].
log4j: Category org.apache.hadoop.mapred.Merger set to INFO
log4j: Parsing appender named "CLA".
log4j: Appender "CLA" was already parsed.
log4j: Handling log4j.additivity.org.apache.hadoop.mapred.Merger=[false]
log4j: Setting additivity for "org.apache.hadoop.mapred.Merger" to false
log4j: Finished configuring.
Launcher AM configuration loaded

从日志中，我们可以看到log4j属性来自jar:file:/usr/lib/hadoop-yarn/hadoop-yarn-server-nodemanager-3.0.0-cdh6.3.2.jar!/container-log4j.properties。我还下载了hadoop-yarn-server-nodemanager-3.0.0-cdh6.3.2.jar。 container-log4j.properties的内容是：

#
# Licensed under the Apache License,Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#   http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing,software
# distributed under the License is distributed on an "AS IS" BASIS,# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND,either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License. See accompanying LICENSE file.
#
# Define some default values that can be overridden by system properties
hadoop.root.logger=DEBUG,CLA
yarn.app.mapreduce.shuffle.logger=${hadoop.root.logger}

# Define the root logger to the system property "hadoop.root.logger".
log4j.rootLogger=${hadoop.root.logger},EventCounter

# Logging Threshold
log4j.threshold=ALL

#
# ContainerLog Appender
#

#Default values
yarn.app.container.log.dir=null
yarn.app.container.log.filesize=100

log4j.appender.CLA=org.apache.hadoop.yarn.ContainerLogAppender
log4j.appender.CLA.containerLogDir=${yarn.app.container.log.dir}
log4j.appender.CLA.containerLogFile=${hadoop.root.logfile}
log4j.appender.CLA.totalLogFileSize=${yarn.app.container.log.filesize}
log4j.appender.CLA.layout=org.apache.log4j.PatternLayout
log4j.appender.CLA.layout.ConversionPattern=%d{ISO8601} %p [%t] %c: %m%n

log4j.appender.CRLA=org.apache.hadoop.yarn.ContainerRollingLogAppender
log4j.appender.CRLA.containerLogDir=${yarn.app.container.log.dir}
log4j.appender.CRLA.containerLogFile=${hadoop.root.logfile}
log4j.appender.CRLA.maximumFileSize=${yarn.app.container.log.filesize}
log4j.appender.CRLA.maxBackupIndex=${yarn.app.container.log.backups}
log4j.appender.CRLA.layout=org.apache.log4j.PatternLayout
log4j.appender.CRLA.layout.ConversionPattern=%d{ISO8601} %p [%t] %c: %m%n

log4j.appender.shuffleCLA=org.apache.hadoop.yarn.ContainerLogAppender
log4j.appender.shuffleCLA.containerLogDir=${yarn.app.container.log.dir}
log4j.appender.shuffleCLA.containerLogFile=${yarn.app.mapreduce.shuffle.logfile}
log4j.appender.shuffleCLA.totalLogFileSize=${yarn.app.mapreduce.shuffle.log.filesize}
log4j.appender.shuffleCLA.layout=org.apache.log4j.PatternLayout
log4j.appender.shuffleCLA.layout.ConversionPattern=%d{ISO8601} %p [%t] %c: %m%n

log4j.appender.shuffleCRLA=org.apache.hadoop.yarn.ContainerRollingLogAppender
log4j.appender.shuffleCRLA.containerLogDir=${yarn.app.container.log.dir}
log4j.appender.shuffleCRLA.containerLogFile=${yarn.app.mapreduce.shuffle.logfile}
log4j.appender.shuffleCRLA.maximumFileSize=${yarn.app.mapreduce.shuffle.log.filesize}
log4j.appender.shuffleCRLA.maxBackupIndex=${yarn.app.mapreduce.shuffle.log.backups}
log4j.appender.shuffleCRLA.layout=org.apache.log4j.PatternLayout
log4j.appender.shuffleCRLA.layout.ConversionPattern=%d{ISO8601} %p [%t] %c: %m%n

################################################################################
# Shuffle Logger
#
log4j.logger.org.apache.hadoop.mapreduce.task.reduce=${yarn.app.mapreduce.shuffle.logger}
log4j.additivity.org.apache.hadoop.mapreduce.task.reduce=false
# Merger is used for both map-side and reduce-side spill merging. On the map
# side yarn.app.mapreduce.shuffle.logger == hadoop.root.logger
#
log4j.logger.org.apache.hadoop.mapred.Merger=${yarn.app.mapreduce.shuffle.logger}
log4j.additivity.org.apache.hadoop.mapred.Merger=false
#
# Event Counter Appender
# Sends counts of logging messages at different severity levels to Hadoop Metrics.
#
log4j.appender.EventCounter=org.apache.hadoop.log.metrics.EventCounter

因此，我认为与log4j相关的日志将保存在syslog文件中。但是在用yarn运行我的xxx.jar之后，系统日志中没有任何内容，我的xxx.jar也包含一些代码，例如：System.out.println("my demo");。运行我的xxx.jar后，我可以在stdout中找到输出“ my demo”。

所以我的问题是为什么不能打印与log4j相关的输出？是否需要其他配置？

解决方法

遇到完全相同的问题。 （我使用 EMR-5.33，Oozie 5.2.0）

如果将此属性设置为 0，日志将写入 syslog log4j.appender.CLA.totalLogFileSize 它从 yarn.app.container.log.filesize 配置中获取值。

log4j.appender.CLA.totalLogFileSize=${yarn.app.container.log.filesize}

有一张 Oozie 票：https://issues.apache.org/jira/browse/OOZIE-3600