基于MRS-Flink组件开发问题：JdbcSink操作GaussDB(DWS)数据库报错No Route to Host

       业务需求：应用华为云平台FusionInsight Manager中的Flink组件配合业务程序实现源端数据向DWS数据库的实时同步。

        在MRS二次开发样例工程的基础上加入自己的业务代码：

public class ReadFromKafka {

    public static void main(String[] args) throws Exception {
        StreamExecutionEnvironment env = StreamExecutionEnvironment.getExecutionEnvironment();
        env.setParallelism(1);
        ParameterTool paraTool = ParameterTool.fromArgs(args);
        Properties properties = paraTool.getProperties();

        DataStream<String> messageStream =
                env.addSource(
                        new FlinkKafkaConsumer<>(
                                paraTool.get("topic"), new SimpleStringSchema(), properties));

        SingleOutputStreamOperator<String> operator = messageStream.
                rebalance()
                .map(
                        new MapFunction<String, String>() {
                            @Override
                            public String map(String s) throws Exception {
                                return "Flink says => " + s + System.getProperty("line.separator");
                            }
                        });

        operator.addSink(JdbcSink.
                sink("insert into test.test_fff(fff)  values (?)",
                        (preparedStatement, s) -> {
                            preparedStatement.setString(1, s);
                        },
                        JdbcExecutionOptions.builder()
                                .withBatchSize(10)
                                .withBatchIntervalMs(1000)
                                .withMaxRetries(3)
                                .build(),
                        new JdbcConnectionOptions.JdbcConnectionOptionsBuilder()
                                .withDriverName("org.postgresql.Driver") 
                                .withUrl("jdbc:postgresql://xx.xxx.xxx.xxx:xxxx/GS_DW")// 数据库Url地址
                                .withUsername("username")
                                .withPassword("password")
                                .build()
                ));

        env.execute();
    }
}

        加入需求业务后启动程序，消息发送程序实时发送编辑好的Json字符串，启动ReakFromKafka接收并消费消息。此时发生报错“java.net.NoRouteToHostException: No route to host (Host unreachable)”,程序终止。报错信息：Caused by: java.net.NoRouteToHostException: No route to host (Host unreachable)

        数据库连接经过验证没有问题。（将数据库写入代码从addSink方法中取出，直接在main方法中执行入库正常，测试代码及数据库结果如下）

        

public class ReadFromKafka3 {

    public static void main(String[] args) throws Exception {
        StreamExecutionEnvironment env = StreamExecutionEnvironment.getExecutionEnvironment();
        env.setParallelism(1);
        ParameterTool paraTool = ParameterTool.fromArgs(args);
        Properties properties = paraTool.getProperties();

        DataStream<String> messageStream =
                env.addSource(
                        new FlinkKafkaConsumer<>(
                                paraTool.get("topic"), new SimpleStringSchema(), properties));


        tryInsert(new Worker(1, "jjj", "123@123.com", 22, new BigDecimal("553.212")));

        SingleOutputStreamOperator<String> operator = messageStream.
                rebalance()
                .map(
                        new MapFunction<String, String>() {
                            @Override
                            public String map(String s) throws Exception {
                                return "Flink says => " + s + System.getProperty("line.separator");
                            }
                        });

        env.execute();
    }


    public static void tryInsert(Worker worker) {
        String sql = "insert into test.worker(id, name, email, age, salary) values (? ,? ,? ,? ,? )";

        Connection conn = null;
        PreparedStatement ps = null;

        try {
            conn =DriverManager.getConnection("jdbc:postgresql://xx.xxx.xxx.xxx:xxxx/GS_DW",
                    "username", "password");
            ps = conn.prepareStatement(sql);

            ps.setInt(1, worker.getInteger("id"));
            ps.setString(2, worker.getString("name"));
            ps.setString(3, worker.getString("email"));
            ps.setInt(4, worker.getInteger("age"));
            ps.setBigDecimal(5, worker.getBigDecimal("salary"));

            ps.executeUpdate();
            System.out.println(" insert finished! ");

        } catch (Exception e) {
            System.out.println(" insert error : " + e.toString() + " , " + e.getCause());
            System.out.println("e.getMessage() = " + e.getMessage() + e.getLocalizedMessage());
            e.printStackTrace();
        } finally {
            DruidUtil.closePsConn(ps, conn);
        }
    }
}

操作结果：

 故问题定位就是只要将数据库操作代码放入addSink方法中就会报如上错误。PS:flink算子入库方法改为使用JdbcSinkFunction(继承 RichSinkFunction类)后报错还是一样，说明不是业务逻辑的问题。

 经过各种测试及查找资料，最终定位问题是数据库Url的IP问题。由于Flink程序是在云平台上运行，故此时需要将"jdbc:postgresql://xx.xxx.xxx.xxx:xxxx/GS_DW"中的IP改为云端IP，找云平台运维人员拿到云端IP并修改后程序即可正常运行。