某某大肠_配置spark的thriftserver模块 可以启动一个thriftserver(相当于一个spark application) 客户端可以通过jdbc连接过来发起SQL。 和spark-shell一样可以配置参数。 启动后 就跟普通的spark shell差不多。 只是多了一个端口可以让客户端jdbc连接过来。 参考文献:https://jaceklaskowski.gitbooks.io/mastering-spark-sql/content/spark-sql-thrift-server.html [root@db05 ~]# su - tidb [tidb@db05 ~]$ tiup cluster display test-tidb [tidb@db05 ~]$ cd /data5/tidb-4.0-cluster/tidb-deploy/tispark-master-7077/ [tidb@db05 /data5/tidb-4.0-cluster/tidb-deploy/tispark-master-7077]$ vim ./conf/spark-defaults.cnf 添加:spark.driver.host 10.0.0.15 启动Thrift JDBC / ODBC服务器 [tidb@db05 /data5/tidb-4.0-cluster/tidb-deploy/tispark-master-7077]$ ./sbin/start-thriftserver.sh --hiveconf hive.server2.thrift.bind.host=localhost hive.server2.thrift.port=10000 [tidb@db05 /data5/tidb-4.0-cluster/tidb-deploy/tispark-master-7077]$ ./sbin/start-thriftserver.sh --hiveconf hive.server2.thrift.bind.host=10.0.0.15 hive.server2.thrift.port=10000 [root@db05 ~]# ps -ef | grep thriftserver [root@db05 ~]# netstat -lnp | grep 10000 [root@db05 ~]# lsof -i:10000 使用Beeline JDBC客户端连接到Spark Thrift服务器 [tidb@db05 /data5/tidb-4.0-cluster/tidb-deploy/tispark-master-7077]$ ./bin/beeline 使用以下connect命令连接到Spark Thrift Server beeline> !connect jdbc:hive2://localhost:10000 beeline> !connect jdbc:hive2://10.0.0.15:10000 Enter username for jdbc:hive2://localhost:10000: root Enter password for jdbc:hive2://localhost:10000: 0: jdbc:hive2://localhost:10000> show databases; +--------------------+ | Database | +--------------------+ | information_schema | | app01 | | atm | | mysql | | performance_schema | | sys | | sysbench | +--------------------+ 7 rows in set (0.00 sec)