#!/bin/bash # # # # # # # # # # # # # # # # # # # # # # # @file :3s_tracking_event_daily.sh # @author :wangjf # @revision:2019-04-30 16:20:25 # # # # # # # # # # # # # # # # # # # # # # source ../../dmp_env.sh today=${ScheduleTime:-$1} date=$(date +"%Y%m%d" -d "-1 day $today") before_date=$(date +"%Y%m%d" -d "-2 day $today") date_path=$(date +%Y/%m/%d -d "-1 day $today") expire_date=$(date +%Y%m%d -d "-2 day $today") expire_date_path=$(date +%Y/%m/%d -d "-2 day $today") PATH_3S_EVENT="s3://trackingcsv-3s/trackingcsv/event" INPUT_PATH_3S="${PATH_3S_EVENT}/$date_path/" check_await "${INPUT_PATH_3S}/23/45/_SUCCESS" output_path="s3://mob-emr-test/dataplatform/DataWareHouse/data/dwh/etl_3s_event_daily/${date_path}" info_output_path="s3://mob-emr-test/dataplatform/DataWareHouse/data/dev/event_info/${date_path}" spark-submit --class mobvista.dmp.datasource.datatory.TrackingEventDaily \ --conf spark.sql.shuffle.partitions=10 \ --conf spark.default.parallelism=10 \ --conf spark.sql.files.maxPartitionBytes=536870912 \ --conf spark.executor.extraJavaOptions="-XX:+UseG1GC" \ --files ${HIVE_SITE_PATH} \ --jars ${JARS} \ --master yarn --deploy-mode cluster --executor-memory 4g --driver-memory 4g --executor-cores 2 --num-executors 50 \ ../../${JAR} -date ${date} -output ${output_path} -info_output ${info_output_path} -before_date ${before_date} if [[ $? -ne 0 ]]; then exit 255 fi # 挂载分区 mount_partition "etl_3s_event_daily" "dt='${date}'" "${output_path}" common_mount_partition "dev" "event_info" "dt='${date}'" "${info_output_path}"