recommend-emr-dataprocess/zhangbo/50_delete_hdfs.sh

#!/bin/sh

export SPARK_HOME=/opt/apps/SPARK2/spark-2.4.8-hadoop3.2-1.0.8
export PATH=$SPARK_HOME/bin:$PATH
export HADOOP_CONF_DIR=/etc/taihao-apps/hadoop-conf
export JAVA_HOME=/usr/lib/jvm/java-1.8.0

DATE="$(date -d '9 days ago' +%Y%m%d)"
HADOOP="/opt/apps/HADOOP-COMMON/hadoop-common-current/bin/hadoop"

path="/dw/recommend/model/feature/user/dt=${DATE}"
$HADOOP fs -test -e ${path}
if [ $? -eq 0 ]; then
    echo "${path} 数据存在, 删除。"
    $HADOOP fs -rm -r -skipTrash ${path}
else
    echo "${path} 数据不存在"
fi

path="/dw/recommend/model/feature/video/dt=${DATE}"
$HADOOP fs -test -e ${path}
if [ $? -eq 0 ]; then
    echo "${path} 数据存在, 删除。"
    $HADOOP fs -rm -r -skipTrash ${path}
else
    echo "${path} 数据不存在"
fi

path="/dw/recommend/model/00_sample_data/dt=${DATE}"
$HADOOP fs -test -e ${path}
if [ $? -eq 0 ]; then
    echo "${path} 数据存在, 删除。"
    $HADOOP fs -rm -r -skipTrash ${path}
else
    echo "${path} 数据不存在"
fi



function delete_path() {
    if [ "$#" -ne 2 ]; then
        echo "Usage: delete_path <early> <path>"
        return 1
    fi
    early=$1
    path=$2
    date="$(date -d "${early} days ago" +%Y%m%d)"
    path_delete=${path}${date}
    $HADOOP fs -test -e ${path_delete}
    if [ $? -eq 0 ]; then
        echo "${path_delete} 数据存在, 删除。"
        if $HADOOP fs -rm -r -skipTrash "${path_delete}"; then
            echo "删除成功。"
        else
            echo "删除失败。"
        fi
    else
        echo "${path_delete} 数据不存在"
    fi
}

delete_path 7 /dw/recommend/model/11_str_data_v3/dt=
delete_path 7 /dw/recommend/model/12_ros_data_v3/dt=
delete_path 7 /dw/recommend/model/10_sample_data_v3/dt=
delete_path 3 /dw/recommend/model/09_feature/user/all/dt=
delete_path 3 /dw/recommend/model/09_feature/user/true/dt=
delete_path 3 /dw/recommend/model/09_feature/video/dt=