1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
#!/usr/bin/env bash
source ../../dmp_env.sh
dt_today=$(date -d "$ScheduleTime 1 days ago" +"%Y%m%d")
dt_dash_today=$(date -d "$ScheduleTime 1 days ago" +"%Y/%m/%d")
update=$(date -d "$ScheduleTime 90 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}/${dt_dash_today}"
fi
OUTPUT_PATH01="${Canglan_Package_Names_Result}/${dt_dash_today}/01"
OUTPUT_PATH02="${Canglan_Package_Names_Result}/${dt_dash_today}/02"
hadoop fs -rm -r "${OUTPUT_PATH01}"
hadoop fs -rm -r "${OUTPUT_PATH02}"
spark-submit --class mobvista.dmp.datasource.dm.CanglanPackageNames \
--conf spark.yarn.executor.memoryOverhead=4096 \
--conf spark.default.parallelism=3000 \
--conf spark.sql.shuffle.partitions=3000 \
--conf spark.network.timeout=720s \
--master yarn --deploy-mode cluster --executor-memory 6g --driver-memory 4g --executor-cores 2 --num-executors 220 \
../../${JAR} -dt_today ${dt_today} -update ${update} -Canglan_Package_Names ${Canglan_Package_Names} -output01 ${OUTPUT_PATH01} -output02 ${OUTPUT_PATH02} -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}/"