adn_tencent_adx_package.sh

#!/usr/bin/env bash

source ../dmp_env.sh


today=$(date -d "$ScheduleTime 1 days ago" +"%Y/%m/%d")
dt_today=$(date -d "$ScheduleTime 1 days ago" +"%Y%m%d")
umount_time=$(date -d "$ScheduleTime 7 days ago" +"%Y%m%d")
umount_date_path=$(date -d "$ScheduleTime 7 days ago" +"%Y/%m/%d")

INPUT_ADX_REQ_PATH="${ADN_ADX_REQ_ORG}/${today}"
INPUT_ADN_MIDWAY_BACKEND_PATH="${TOUTIAO_LAUNCH_PATH}/${today}"
OUTPUT_DIM_ADN_ADX_PKG_PATH="${DIM_ADN_ADX_PKG}/${today}"
UMOUNT_DIM_ADN_ADX_PKG_PATH="${DIM_ADN_ADX_PKG}/${umount_date_path}"


OUTPUT_ODS_ADX_TMP_PATH="${ODS_ADN_ADX_REQ_TMP}/${today}"
UMOUNT_OUTPUT_ODS_ADX_TMP_PATH="${ODS_ADN_ADX_REQ_TMP}/${umount_date_path}"

check_await "${INPUT_ADX_REQ_PATH}/virginia/23/_SUCCESS"
check_await "${INPUT_ADN_MIDWAY_BACKEND_PATH}/virginia/23/_SUCCESS"


mount_partition "dim_adn_adx_package" "dt='${dt_today}'" "${OUTPUT_DIM_ADN_ADX_PKG_PATH}"

hadoop fs -rm -r "${OUTPUT_DIM_ADN_ADX_PKG_PATH}/"
hadoop fs -rm -r "${OUTPUT_ODS_ADX_TMP_PATH}/"


spark-submit --class mobvista.dmp.datasource.adn_adx.AdnTecentAdxDataMidWay \
 --conf spark.yarn.executor.memoryOverhead=2048  \
 --conf spark.network.timeout=720s \
 --conf spark.sql.shuffle.partitions=1000 \
 --conf spark.default.parallelism=1000 \
 --conf spark.sql.autoBroadcastJoinThreshold=31457280 \
 --conf spark.executor.extraJavaOptions="-XX:+UseG1GC" \
 --files ${HIVE_SITE_PATH} \
 --jars s3://mob-emr-test/dataplatform/DataWareHouse/offline/myjar/hive-hcatalog-core-2.3.3.jar,s3://mob-emr-test/dataplatform/DataWareHouse/offline/myjar/json-serde-1.3.7-jar-with-dependencies.jar \
 --master yarn --deploy-mode cluster --name  AdnTecentAdxDataMidWay --executor-memory 10g --driver-memory 4g  --executor-cores 5  --num-executors 100 \
 ../${JAR}  -outputadxtmp ${OUTPUT_ODS_ADX_TMP_PATH} -dimadxpkg ${OUTPUT_DIM_ADN_ADX_PKG_PATH} \
 -coalesce 60 \
 -today ${dt_today}


if [ $? -ne 0 ];then
  exit 255
fi



mount_partition "ods_adn_adx_req_tmp" "dt='${dt_today}'" "${OUTPUT_ODS_ADX_TMP_PATH}"

unmount_partition "ods_adn_adx_req_tmp" "dt='${umount_time}'" "${UMOUNT_OUTPUT_ODS_ADX_TMP_PATH}"
unmount_partition "dim_adn_adx_package" "dt='${umount_time}'" "${UMOUNT_DIM_ADN_ADX_PKG_PATH}"

hadoop fs -touchz ${OUTPUT_ODS_ADX_TMP_PATH}/_SUCCESS
hadoop fs -touchz ${OUTPUT_DIM_ADN_ADX_PKG_PATH}/_SUCCESS