使用Python和pyathenajdbc与Athena连接

Question

我正在尝试使用python连接到AWS Athena.我正在尝试使用pyathenajdbc来完成此任务.我遇到的问题是获得连接.当我运行下面的代码时,我收到一条错误消息,指出它无法找到AthenaDriver.(java.lang.RuntimeException:未找到类com.amazonaws.athena.jdbc.AthenaDriver).我确实从AWS下载了这个文件,我确认它正在该目录中.

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"

ATHENA_JDBC_CLASSPATH = "/opt/amazon/athenajdbc/AthenaJDBC41-1.0.0.jar"
EXPORT_OUTFILE = "RSI_Export.txt"
EXPORT_OUTFILE_PATH = os.path.join(WORKINGDIR, EXPORT_OUTFILE)


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")

    conn = pyathenajdbc.connect(
        s3_staging_dir="s3://athena",
        access_key=AWS_ACCESS_KEY_ID,
        secret_key=AWS_SECRET_ACCESS_KEY,
        region_name="us-east-1",
        log_path=LOGDIR,
        driver_path=ATHENA_JDBC_CLASSPATH
    )
    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)

回溯(最近一次调用最后一次):文件"/usr/lib64/python2.7/runpy.py",第174行,在_run_module_as_main" main ",fname,loader,pkg_name)文件"/usr/lib64/python2.7/ runpy.py",第72行,在run_globals文件中的_run_code exec代码中"/ home/ec2-user/jason_testing/mdpbi/rsi/athena/main .py",第53行,在rtn = main()文件"/ home/ec2-user/jason_testing/mdpbi/rsi/athena/main .py",第39行,在主driver_path = athena_jdbc_driver_path文件中"/opt/mdpbi/Python_Envs/2.7.10/local/lib/python2.7/dist-packages/ pyathenajdbc/init .py",第65行,连接driver_path,**kwargs)文件"/opt/mdpbi/Python_Envs/2.7.10/local/lib/python2.7/dist-packages/pyathenajdbc/connection.py",第68行,在init jpype.JClass(ATHENA_DRIVER_CLASS_NAME)文件"/opt/mdpbi/Python_Envs/2.7.10/lib64/python2.7/dist-packages/jpype/_jclass.py",第55行,在JClass中引发_RUNTIMEEXCEPTION.PYEXC ("未找到类%s"%名称)

Answer 1

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)