1、org.apache.spark.shuffle.FetchFailedException: Failed to connect
当出现这个错误的时候,绝大部分原因是spark连接executor超时了,解决方案:
在spark-submit脚本中增加:--conf spark.core.connection.ack.wait.timeout=600s
2、spark 访问hbase时,出现Caused by: java.lang.UnsatisfiedLinkError: org.apache.hadoop.util.NativeCodeLoader.buildSupportsSnappy()Z
解决方案:在spark-env.sh中增加环境变量
export JAVA_LIBRARY_PATH=$JAVA_LIBRARY_PATH:$HADOOP_HOME/lib/native
export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:$HADOOP_HOME/lib/native
export SPARK_YARN_USER_ENV="JAVA_LIBRARY_PATH=$JAVA_LIBRARY_PATH,LD_LIBRARY_PATH=$LD_LIBRARY_PATH"