#!/usr/bin/env bash source ../../dmp_env.sh begin_day=$(date -d "$ScheduleTime 15 days ago" +"%Y%m%d") begin_day02=$(date -d "$ScheduleTime 30 days ago" +"%Y%m%d") end_day=$(date -d "$ScheduleTime 1 days ago" +"%Y%m%d") dt_dash_today=$(date -d "$ScheduleTime 1 days ago" +"%Y/%m/%d") check_await "${TMP_EGGPLANTS_OUTPUT_PATH}/${dt_dash_today}" hadoop fs -test -e "${ODS_OTHER_DEVICE_DAILY}/${dt_dash_today}" if [ $? -ne 0 ];then hadoop fs -mkdir -p "${ODS_OTHER_DEVICE_DAILY}/${dt_dash_today}" fi OUTPUT_PATH01="${TMP_COM_BTOP_TIKTOKRV_GAID_PATH}/${dt_dash_today}/01" OUTPUT_PATH02="${TMP_COM_BTOP_TIKTOKRV_GAID_PATH}/${dt_dash_today}/02" OUTPUT_PATH03="${RTDMP_TMP_PACKAGE_NAME_PATH}/com_btop_tiktokrv/${dt_dash_today}/gaid" hadoop fs -rm -r "${OUTPUT_PATH01}" hadoop fs -rm -r "${OUTPUT_PATH02}" hadoop fs -rm -r "${OUTPUT_PATH03}" spark-submit --class mobvista.dmp.datasource.dm.BtopTiktokrvGaid \ --conf spark.yarn.executor.memoryOverhead=2048 \ --conf spark.default.parallelism=3000 \ --conf spark.sql.shuffle.partitions=3000 \ --conf spark.driver.maxResultSize=4g \ --conf spark.network.timeout=720s \ --master yarn --deploy-mode cluster --executor-memory 8g --driver-memory 10g --executor-cores 5 --num-executors 70 \ ../../${JAR} -begin_day ${begin_day} -begin_day02 ${begin_day02} -end_day ${end_day} -output01 ${OUTPUT_PATH01} -output02 ${OUTPUT_PATH02} -output03 ${OUTPUT_PATH03} -coalesce 200 if [[ $? -ne 0 ]]; then exit 255 fi hadoop distcp -m20 "${OUTPUT_PATH01}/*" "${TMP_EGGPLANTS_OUTPUT_PATH}/${dt_dash_today}/" hadoop distcp -m20 "${OUTPUT_PATH02}/*" "${ODS_OTHER_DEVICE_DAILY}/${dt_dash_today}/"