搭配 Flink 使用 AWS Glue

Amazon EMR on EKS 搭配 Apache Flink 6.15.0 版和更新版本支援使用 AWS Glue Data Catalog 做為串流和批次 SQL 工作流程的中繼資料存放區。

您必須先建立名為 AWS default的 Glue 資料庫，做為 Flink SQL Catalog。此 Flink Catalog 會存放中繼資料，例如資料庫、資料表、分割區、檢視、函數，以及存取其他外部系統中資料所需的其他資訊。

aws glue create-database \
    --database-input "{\"Name\":\"default\"}" 

若要啟用 AWS Glue 支援，請使用FlinkDeployment規格。此範例規格使用 Python 指令碼來快速發行一些 Flink SQL 陳述式，以與 Glue AWS 目錄互動。

apiVersion: flink.apache.org/v1beta1
kind: FlinkDeployment
metadata:
  name: python-example
spec:
  flinkVersion: v1_17
  flinkConfiguration:
    taskmanager.numberOfTaskSlots: "1"
    aws.glue.enabled: "true"
  executionRoleArn: job-execution-role-arn;
  emrReleaseLabel: "emr-6.15.0-flink-latest"
  jobManager:
    highAvailabilityEnabled: false
    replicas: 1
    resource:
      memory: "2048m"
      cpu: 1
  taskManager:
    resource:
      memory: "2048m"
      cpu: 1
  job:
    jarURI: s3://<S3_bucket_with_your_script/pyflink-glue-script.py
    entryClass: "org.apache.flink.client.python.PythonDriver"
    args: ["-py", "/opt/flink/usrlib/pyflink-glue-script.py"] 
    parallelism: 1
    upgradeMode: stateless

以下是 PyFlink 指令碼可能看起來的範例。

import logging
import sys
from pyflink.datastream import StreamExecutionEnvironment
from pyflink.table import StreamTableEnvironment

def glue_demo():
    env = StreamExecutionEnvironment.get_execution_environment()
    t_env = StreamTableEnvironment.create(stream_execution_environment=env)
    t_env.execute_sql("""
          CREATE CATALOG glue_catalog WITH (
          'type' = 'hive',
          'default-database' = 'default',
          'hive-conf-dir' = '/glue/confs/hive/conf',
          'hadoop-conf-dir' = '/glue/confs/hadoop/conf'
          )
                      """)
    t_env.execute_sql("""
          USE CATALOG glue_catalog;
                      """)
    t_env.execute_sql("""
          DROP DATABASE IF EXISTS eks_flink_db CASCADE;
                      """)
    t_env.execute_sql("""
          CREATE DATABASE IF NOT EXISTS eks_flink_db WITH ('hive.database.location-uri'= 's3a://S3-bucket-to-store-metadata/flink/flink-glue-for-hive/warehouse/');
                      """)
    t_env.execute_sql("""
          USE eks_flink_db;
                  """)
    t_env.execute_sql("""
          CREATE TABLE IF NOT EXISTS eksglueorders (
            order_number BIGINT,
            price        DECIMAL(32,2),
            buyer        RO first_name STRING, last_name STRING,
            order_time   TIMESTAMP(3)
          ) WITH (
            'connector' = 'datagen'
          );
                      """)
    t_env.execute_sql("""
          CREATE TABLE IF NOT EXISTS eksdestglueorders (
            order_number BIGINT,
            price        DECIMAL(32,2),
            buyer        ROW first_name STRING, last_name STRING,
            order_time   TIMESTAMP(3)
          ) WITH (
            'connector' = 'filesystem',
            'path' = 's3://S3-bucket-to-store-metadata/flink/flink-glue-for-hive/warehouse/eksdestglueorders',
            'format' = 'json'
          );
                  """)
    t_env.execute_sql("""
          CREATE TABLE IF NOT EXISTS print_table (
            order_number BIGINT,
            price        DECIMAL(32,2),
            buyer        ROW first_name STRING, last_name STRING,
            order_time   TIMESTAMP(3)
          ) WITH (
            'connector' = 'print'
          );
                """)
    t_env.execute_sql("""
          EXECUTE STATEMENT SET
          BEGIN
          INSERT INTO eksdestglueorders SELECT * FROM  eksglueorders LIMIT 10;
          INSERT INTO print_table SELECT * FROM eksdestglueorders;
          END;
            """)


if __name__ == '__main__':
    logging.basicConfig(stream=sys.stdout, level=logging.INFO, format="%(message)s")
    glue_demo()