使用Python连接kerberos认证的Hive

# !/usr/bin/env python
# -*- coding: utf-8 -*-
# hive util from CDH with kerberos
from impala.dbapi import connect
from krbcontext import krbcontext
import pandas as pd
from config.conf import HiveConfig


class HiveClient:
    __instance = None

    @classmethod
    def instance(cls, *args, **kwargs):
        if cls.__instance:
            return cls.__instance
        else:
            try:
                hive_conf = HiveConfig.instance()
                with krbcontext(using_keytab=True, principal=hive_conf.principal, keytab_file=hive_conf.keytab):
                    conn = connect(host=hive_conf.host,
                                   port=hive_conf.port,
                                   auth_mechanism='GSSAPI',
                                   kerberos_service_name='hive')
                    return cls(conn)
            except Exception as e:
                raise e

    def __init__(self, conn):
        self.conn = conn

    # 执行
    def execute(self, sql):
        try:
            cur = self.conn.cursor()
            cur.execute(sql)
        except Exception as err:
            print(err)
            raise err

    # 查询
    def query(self, sql):
        cur = self.conn.cursor()
        res = None
        try:
            cur.execute(sql)
            res = cur.fetchall()
        except Exception as err:
            print("Phoenix 查询失败, %s" % err)
            raise err
        finally:
            return res

    # 查询输出DataFrame
    def query_to_df(self, sql):
        with self.conn.cursor() as cursor:
            cursor.execute(sql)
            columns = [col[0] for col in cursor.description]
            records = [dict(zip(columns, row)) for row in cursor.fetchall()]
            results = pd.DataFrame(records)
            results.columns = columns
            return results

    def close(self):
        self.conn.close()

if __name__ == '__main__':
    hive_client = HiveClient.instance()
    print(hive_client.query_to_df('SELECT * from table1 limit 10'))
    print(hive_client.execute('SELECT * from table2 limit 10'))