update script

eda/smart_rank/tag3_start.sh

@@ -3,9 +3,9 @@
 # 新画像
 mysql -u st_user -paqpuBLYzEV7tML5RPsN1pntUzFy -h 172.16.40.158 -P 4000 -D jerry_test -e "delete from kafka_tag3_log where log_time < UNIX_TIMESTAMP(DATE_SUB(NOW(), INTERVAL 60 day))"
 
-mysql -u st_user -paqpuBLYzEV7tML5RPsN1pntUzFy -h 172.16.40.158 -P 4000 -D jerry_test -e "delete from user_tag3_portrait where date < DATE_SUB(CURDATE(), INTERVAL 1 WEEK)"
+mysql -u st_user -paqpuBLYzEV7tML5RPsN1pntUzFy -h 172.16.40.158 -P 4000 -D jerry_test -e "delete from user_tag3_portrait where date < DATE_SUB(CURDATE(), INTERVAL 15 day)"
 
-mysql -u st_user -paqpuBLYzEV7tML5RPsN1pntUzFy -h 172.16.40.158 -P 4000 -D jerry_test -e "delete from user_tag3_event_portrait where date < DATE_SUB(CURDATE(), INTERVAL 1 WEEK)"
+mysql -u st_user -paqpuBLYzEV7tML5RPsN1pntUzFy -h 172.16.40.158 -P 4000 -D jerry_test -e "delete from user_tag3_event_portrait where date < DATE_SUB(CURDATE(), INTERVAL 15 day)"
 
 
 /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.executorEnv.LD_LIBRARY_PATH="/opt/java/jdk1.8.0_181/jre/lib/amd64/server:/opt/cloudera/parcels/CDH-5.16.1-1.cdh5.16.1.p0.3/lib64" --conf spark.locality.wait=0 --archives /srv/apps/ftrl/bandits.zip --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/ffm-baseline_git/eda/smart_rank/tag3_update_user_portrait_offline.py

eda/smart_rank/tag3_update_user_portrait_offline.py

@@ -196,9 +196,9 @@ def consume_kafka():
 
         device_ids_lst_rdd = spark.sparkContext.parallelize(device_ids_lst, numSlices=1000)
 
-        # result = device_ids_lst_rdd.repartition(100).map(lambda x: update_tag3_user_portrait(x))
-        # # result.foreach(print)
-        # result.collect()
+        result = device_ids_lst_rdd.repartition(100).map(lambda x: update_tag3_user_portrait(x))
+        # result.foreach(print)
+        result.collect()
 
         result2 = device_ids_lst_rdd.repartition(100).map(lambda x: update_tag3_user_portrait_by_event(x))
         # result2.foreach(print)