1、pyspark增加config设置

java heap错误增加内存

spark = (
    SparkSession.builder.config("spark.hadoop.hive.exec.dynamic.partition", "true")
    .config("spark.hadoop.hive.exec.dynamic.partition.mode", "nonstrict")
    .config("spark.driver.memory", "32g") \
   .config("spark.executor.memory", "64g") \
   .config("spark.executor.cores","6")
    .enableHiveSupport()
    .getOrCreate()
)
sc = spark.sparkContext

*****限制_success生成文件

spark = SparkSession \
    .builder \
  .config("spark.hadoop.mapreduce.fileoutputcommitter.marksuccessfuljobs","false") \
    .enableHiveSupport() \
    .getOrCreate()
sc = spark.sparkContext

2、spark-submit 参数

参考：https://www.cnblogs.com/weiweifeng/p/8073553.html

nohup spark-submit     --class com.tcl.video.search.recommend.SparkQueryApplication      --executor-memory 6G     --num-executors 6     --master yarn /tmp/java-applic****-spark.jar >> /tmp/java-applicati****ta-spark.log 2>&1 &

spark-submit     --class com.tcl.video.search.recommend.SparkQueryApplication      --executor-memory 6G     --num-executors 6     --master yarn *******st.py --date 20211117

查看集群执行log:

yarn logs -applicationId applic******id值**0_77969