背景

在某个场景中，需要从Kafka中获取数据，经过转换处理后，需要同时sink到多个输出源中(kafka、mysql、hologres)等。两次调用execute, 阿里云Flink vvr引擎报错：

java 复制代码

public static void main(String[] args) {
        final StreamExecutionEnvironment env = StreamExecutionEnvironment.getExecutionEnvironment();
        StreamTableEnvironment tEnv = StreamTableEnvironment.create(env);
        StreamStatementSet streamStatementSet = tEnv.createStatementSet();

        String s = LocalDateTimeUtils.getDateTime(System.currentTimeMillis());

        DataStream<String> dataStream = env.fromElements(s, LocalDateTimeUtils.getDateTime(System.currentTimeMillis()));

        tEnv.executeSql(KAFKA_TABLE_SQL);
        tEnv.executeSql(KAFKA_TABLE_SQL_1);


        Table table = tEnv.fromDataStream(dataStream);
        table.insertInto("kafka_sink").execute();
        table.insertInto("kafka_sink_1").execute();

        streamStatementSet.execute();
    }

java 复制代码

Caused by: org.apache.flink.util.FlinkRuntimeException: Cannot have more than one execute() or executeAsync() call in a single environment.
	at org.apache.flink.client.program.StreamContextEnvironment.validateAllowedExecution(StreamContextEnvironment.java:199) ~[flink-dist-1.15-vvr-6.0.7-1-SNAPSHOT.jar:1.15-vvr-6.0.7-1-SNAPSHOT]
	at org.apache.flink.client.program.StreamContextEnvironment.executeAsync(StreamContextEnvironment.java:187) ~[flink-dist-1.15-vvr-6.0.7-1-SNAPSHOT.jar:1.15-vvr-6.0.7-1-SNAPSHOT]
	at org.apache.flink.table.planner.delegation.DefaultExecutor.executeAsync(DefaultExecutor.java:110) ~[?:?]
	at org.apache.flink.table.api.internal.TableEnvironmentImpl.executeInternal(TableEnvironmentImpl.java:877) ~[flink-table-api-java-uber-1.15-vvr-6.0.7-1-SNAPSHOT.jar:1.15-vvr-6.0.7-1-SNAPSHOT]
	at org.apache.flink.table.api.internal.TableEnvironmentImpl.executeInternal(TableEnvironmentImpl.java:756) ~[flink-table-api-java-uber-1.15-vvr-6.0.7-1-SNAPSHOT.jar:1.15-vvr-6.0.7-1-SNAPSHOT]
	at org.apache.flink.table.api.internal.TableEnvironmentImpl.executeInternal(TableEnvironmentImpl.java:955) ~[flink-table-api-java-uber-1.15-vvr-6.0.7-1-SNAPSHOT.jar:1.15-vvr-6.0.7-1-SNAPSHOT]
	at org.apache.flink.table.api.internal.TablePipelineImpl.execute(TablePipelineImpl.java:57) ~[flink-table-api-java-uber-1.15-vvr-6.0.7-1-SNAPSHOT.jar:1.15-vvr-6.0.7-1-SNAPSHOT]

解决

使用 StreamStatementSet. 具体参考官网：

https://nightlies.apache.org/flink/flink-docs-release-1.15/zh/docs/dev/table/data_stream_api/#converting-between-datastream-and-table

改良后的代码：

java 复制代码

public static void main(String[] args) {
        final StreamExecutionEnvironment env = StreamExecutionEnvironment.getExecutionEnvironment();
        StreamTableEnvironment tEnv = StreamTableEnvironment.create(env);
        StreamStatementSet streamStatementSet = tEnv.createStatementSet();

        String s = LocalDateTimeUtils.getDateTime(System.currentTimeMillis());

        DataStream<String> dataStream = env.fromElements(s, LocalDateTimeUtils.getDateTime(System.currentTimeMillis()));

        tEnv.executeSql(KAFKA_TABLE_SQL);
        tEnv.executeSql(KAFKA_TABLE_SQL_1);


        Table table = tEnv.fromDataStream(dataStream);

        streamStatementSet.addInsert("kafka_sink", table);
        streamStatementSet.addInsert("kafka_sink_1", table);

        streamStatementSet.execute();
    }

Flink Table API/SQL 多分支sink

背景

解决