update

task.sh

 source /srv/envs/esmm/bin/activate
 /opt/spark/bin/spark-submit --master yarn --deploy-mode client --queue root.strategy --driver-memory 16g --executor-memory 1g --executor-cores 1 --num-executors 70 --conf spark.default.parallelism=100 --conf spark.storage.memoryFraction=0.5 --conf spark.shuffle.memoryFraction=0.3  --conf spark.locality.wait=0 --jars /srv/apps/tispark-core-2.1-SNAPSHOT-jar-with-dependencies.jar,/srv/apps/spark-connector_2.11-1.9.0-rc2.jar,/srv/apps/mysql-connector-java-5.1.38.jar /srv/apps/meta_base_code/task/conent_detail_page_grayscale_ctr.py
 /opt/spark/bin/spark-submit --master yarn --deploy-mode client --queue root.strategy --driver-memory 16g --executor-memory 1g --executor-cores 1 --num-executors 70 --conf spark.default.parallelism=100 --conf spark.storage.memoryFraction=0.5 --conf spark.shuffle.memoryFraction=0.3  --conf spark.locality.wait=0 --jars /srv/apps/tispark-core-2.1-SNAPSHOT-jar-with-dependencies.jar,/srv/apps/spark-connector_2.11-1.9.0-rc2.jar,/srv/apps/mysql-connector-java-5.1.38.jar /srv/apps/meta_base_code/task/recommend_strategy_d.py
-#/opt/spark/bin/spark-submit --master yarn --deploy-mode client --queue root.strategy --driver-memory 16g --executor-memory 1g --executor-cores 1 --num-executors 70 --conf spark.default.parallelism=100 --conf spark.storage.memoryFraction=0.5 --conf spark.shuffle.memoryFraction=0.3  --conf spark.locality.wait=0 --jars /srv/apps/tispark-core-2.1-SNAPSHOT-jar-with-dependencies.jar,/srv/apps/spark-connector_2.11-1.9.0-rc2.jar,/srv/apps/mysql-connector-java-5.1.38.jar /srv/apps/meta_base_code/task/search_strategy_d.py
+/opt/spark/bin/spark-submit --master yarn --deploy-mode client --queue root.strategy --driver-memory 16g --executor-memory 1g --executor-cores 1 --num-executors 70 --conf spark.default.parallelism=100 --conf spark.storage.memoryFraction=0.5 --conf spark.shuffle.memoryFraction=0.3  --conf spark.locality.wait=0 --jars /srv/apps/tispark-core-2.1-SNAPSHOT-jar-with-dependencies.jar,/srv/apps/spark-connector_2.11-1.9.0-rc2.jar,/srv/apps/mysql-connector-java-5.1.38.jar /srv/apps/meta_base_code/task/search_strategy_d.py
 /opt/spark/bin/spark-submit --master yarn --deploy-mode client --queue root.strategy --driver-memory 16g --executor-memory 2g --executor-cores 4 --num-executors 70 --conf spark.default.parallelism=100 --conf spark.storage.memoryFraction=0.5 --conf spark.shuffle.memoryFraction=0.3  --conf spark.locality.wait=0 --jars /srv/apps/tispark-core-2.1-SNAPSHOT-jar-with-dependencies.jar,/srv/apps/spark-connector_2.11-1.9.0-rc2.jar,/srv/apps/mysql-connector-java-5.1.38.jar /srv/apps/meta_base_code/task/core_indicators_monitoring.py
 

task/core_indicators_monitoring.py

@@ -60,7 +60,7 @@ spark.sql("CREATE TEMPORARY FUNCTION is_json AS 'com.gmei.hive.common.udf.UDFJso
 spark.sql("CREATE TEMPORARY FUNCTION arrayMerge AS 'com.gmei.hive.common.udf.UDFArryMerge'")
 
 task_list = []
-task_days = 5
+task_days = 10
 
 for t in range(0, task_days):
     day_num = 0 - t

task/search_meigou_ctr.py

@@ -165,17 +165,16 @@ for t in range(0, task_days):
     print("-------------------------------")
 
     sql_spam_pv_device_id = """
-SELECT T.DEVICE_ID, --设备ID
+    select T.DEVICE_ID,T.CARD_ID,if(NVL(C.EXPOSURE,0) <> 0 ,cast((NVL(T.CLICK_NUM,0)/NVL(C.EXPOSURE,0)) as decimal(18,5)) , 0) as search_ctr from
+(SELECT T.DEVICE_ID, --设备ID
        T.CARD_ID, --卡片ID
-       SUM(T.CLICK_NUM) AS CLICK_NUM, --点击次数
-       C.EXPOSURE as EXPOSURE  --曝光次数
-       if(NVL(C.EXPOSURE,0) <> 0 ,cast((NVL(T.CLICK_NUM,0)/NVL(C.EXPOSURE,0)) as decimal(18,5)) , 0) as search_ctr
+       SUM(T.CLICK_NUM) AS CLICK_NUM --点击次数
   FROM ML.ML_C_ET_CK_CLICK_DIMEN_D T
  WHERE T.PARTITION_DAY = '{partition_day}'
    AND T.PAGE_CODE = 'search_result_welfare'
    AND T.ACTION IN ('goto_welfare_detail','search_result_welfare_click_item')
  GROUP BY T.DEVICE_ID,
-          T.CARD_ID
+          T.CARD_ID) T
 
     left join 
  (SELECT T.DEVICE_ID as DEVICE_ID, --设备ID

task/search_strategy_d.py

@@ -72,7 +72,7 @@ spark.sql("CREATE TEMPORARY FUNCTION is_json AS 'com.gmei.hive.common.udf.UDFJso
 spark.sql("CREATE TEMPORARY FUNCTION arrayMerge AS 'com.gmei.hive.common.udf.UDFArryMerge'")
 
 task_list = []
-task_days = 90
+task_days = 5
 
 for t in range(0, task_days):
     day_num = 0 - t