MRS Flink使用SQL-Client对接Hive

一、准备环境

1.根据产品文档安装Flink客户端；

2.将sql-client-defaults.yaml放入/opt/client/Flink/flink/conf中

3.将jaas.conf放入/opt/client/Flink/flink/conf中

Client { com.sun.security.auth.module.Krb5LoginModule required useKeyTab=false useTicketCache=true debug=false; };

4.添加sql-client.sh中添加在JVM_ARGS参数：

JVM_ARGS="-Djava.security.auth.login.config=/opt/client/Flink/flink/conf/jaas.conf $JVM_ARGS"

二、启动Flink集群

例如：yarn-session.sh -t ssl -d

三、启动SQL-Client

./sql-client.sh embedded -d ./../conf/sql-client-defaults.yaml

四、运行SQL

CREATE TABLE kafkaSourceTable (   order_id VARCHAR,   shop_id VARCHAR,   member_id VARCHAR,   trade_amt DOUBLE ) WITH (   'connector.type' = 'kafka',   'connector.version' = 'universal',   'connector.topic' = 'order_sql',   'connector.properties.bootstrap.servers' = '10.162.147.217:21005',   'connector.properties.zookeeper.connect' = '10.162.147.217:24002',   'connector.properties.group.id' = 'test-consumer-group',   'connector.startup-mode' = 'latest-offset',   'format.type' = 'json' );

CREATE TABLE kafkaSinkTable(shop_id VARCHAR, member_id VARCHAR) WITH (   'connector.type' = 'kafka',   'connector.version' = 'universal',   'connector.topic' = 'order_sql',   'connector.properties.bootstrap.servers' = '10.162.147.217:21005',   'connector.properties.zookeeper.connect' = '10.162.147.217:24002',   'update-mode' = 'append',   'format.type' = 'json' ); INSERT INTO   kafkaSinkTable SELECT   shop_id,   member_id FROM   kafkaSourceTable; SELECT   shop_id,   member_id FROM   kafkaSourceTable;

五、对接Hive

1）修改sql-client-defaults.yaml

catalogs:     - name: myhive      type: hive     hive-conf-dir: /opt/clienrc5/Hive/config     hive-version: 3.1.0

2）在/opt/clienrc5/Hive/config/hive-site.xml添加配置

<property> <name>hive.metastore.sasl.enabled</name> <value>true</value> </property>

3）启动sql-client

use catalog myhive;

SET table.sql-dialect=hive;

CREATE TABLE IF NOT EXISTS hive_dialect_tbl ( `id` int , `name` string , `age` int ) ROW FORMAT DELIMITED FIELDS TERMINATED BY ','; SET table.sql-dialect=default;

CREATE TABLE datagen ( `id` int , `name` string , `age` int ) WITH ( 'connector' = 'datagen', 'rows-per-second'='1' ); INSERT INTO hive_dialect_tbl SELECT * FROM datagen; select * from hive_dialect_tbl;