Hi ruiyun,
Is there any log4j configuration file or logback.xml in your Flink configuration directory? Yarn should check whether the file exists first and set relative log options. > On 26 Apr 2022, at 5:21 PM, ruiyun wan <wryinv...@gmail.com> wrote: > > Flink Version (1.13) > There is not jobmanager.log file when submit yarn-per-job with > sql-client.sh。(set execution.target = yarn-per-job;) There are not -Dlog.file > and -Dlog4j.configuration properties argument in the auto-generated > launch_container.sh on YARN cluster。e.g. the last line of launch_container.sh > is : > exec /bin/bash -c "$JAVA_HOME/bin/java -Xmx50331648 -Xms50331648 \ > -XX:MaxMetaspaceSize=134217728 \ > org.apache.flink.yarn.entrypoint.YarnJobClusterEntrypoint ..... > > However,When use yarn-session.sh create a session cluster,the last line of > auto generated launch_container.sh is > exec /bin/bash -c "$JAVA_HOME/bin/java -Xmx1530082096 -Xms1530082096 \ > -XX:MaxMetaspaceSize=268435456 > > -Dlog.file="/mnt/disk3/log/hadoop-yarn/containers/application_1640228329925_0008/container_e05_1640228329925_0008_01_000001/jobmanager.log" > \ > -Dlog4j.configuration=file:log4j.properties \ > -Dlog4j.configurationFile=file:log4j.properties \ > org.apache.flink.yarn.entrypoint.YarnSessionClusterEntrypoint ..... > which contained -Dlog.file and -Dlog4j.configuration > > The question is :how to configure sql-client.sh to produce jobmanager.log for > yarn-per-job?
