使用Python和pyathenajdbc与Athena连接

JDBC驱动程序需要Java 8.我当前正在运行Java 7.我能够在EC2实例上安装另一个版本的Java.

https://tecadmin.net/install-java-8-on-centos-rhel-and-fedora/#

我还必须在我的代码中设置java版本.通过这些更改,代码现在可以按预期运行.

from mdpbi.rsi.config import *
from mdpbi.tools.functions import mdpLog
from pkg_resources import resource_string
import argparse
import os
import pyathenajdbc
import sys

SCRIPT_NAME = "Athena_Export"


def get_arg_parser():
    """This function returns the argument parser object to be used with this script"""
    parser = argparse.ArgumentParser(description=__doc__, formatter_class=argparse.RawDescriptionHelpFormatter)

    return parser


def main():
    args = get_arg_parser().parse_args(sys.argv[1:])
    logger = mdpLog(SCRIPT_NAME, LOGDIR)

    SQL = resource_string("mdpbi.rsi.athena.resources", "athena.sql")

    os.environ["JAVA_HOME"] = "/opt/jdk1.8.0_121"
    os.environ["JRE_HOME"] = "/opt/jdk1.8.0_121/jre"
    os.environ["PATH"] = "/opt/jdk1.8.0_121/bin:/opt/jdk1.8.0_121/jre/bin"

    conn = pyathenajdbc.connect(
        s3_staging_dir="s3://mdpbi.data.rsi.out/",
        access_key=AWS_ACCESS_KEY_ID,
        secret_key=AWS_SECRET_ACCESS_KEY,
        schema_name="rsi",
        region_name="us-east-1"
    )
    try:
        with conn.cursor() as cursor:
            cursor.execute(SQL)
            logger.info(cursor.description)
            logger.info(cursor.fetchall())
    finally:
        conn.close()

    return 0


if __name__ == '__main__':
    rtn = main()
    sys.exit(rtn)