Flink异步io关联Hbase

主程序

csharp 复制代码
    public static void main(String[] args) throws Exception {
        //1.获取流执行环境
        StreamExecutionEnvironment env = StreamExecutionEnvironment.getExecutionEnvironment();
        env.setParallelism(1);
        SimpleDateFormat formatter = new SimpleDateFormat("yyyy-MM-dd HH:mm:ss");

        //设置动态参数
        ParameterTool propertiesargs = ParameterTool.fromArgs(args);
        String fileName = propertiesargs.get("CephConfPath");
        //从hdfs获取动态参数配置文件
        org.apache.hadoop.conf.Configuration conf = new org.apache.hadoop.conf.Configuration();
        FileSystem fs = FileSystem.get(URI.create(fileName), conf);
        fs.open(new org.apache.hadoop.fs.Path(fileName));
        ParameterTool propertiesFile = ParameterTool.fromPropertiesFile(fs.open(new org.apache.hadoop.fs.Path(fileName)).getWrappedStream());
        // 注册给环境变量(HBASE使用)
        env.getConfig().setGlobalJobParameters(propertiesFile);
        new CephConfig(propertiesFile);

        //2.设置CK&状态后端
        env.setStateBackend(new FsStateBackend(FSSTATEBACKEND));
        env.enableCheckpointing(10000);// 每 ** ms 开始一次 checkpoint
        env.getCheckpointConfig().setCheckpointingMode(CheckpointingMode.EXACTLY_ONCE);// 设置模式为精确一次
        env.getCheckpointConfig().setCheckpointTimeout(100000);// Checkpoint 必须在** ms内完成,否则就会被抛弃
        env.getCheckpointConfig().setMaxConcurrentCheckpoints(2);// 同一时间只允许一个 checkpoint 进行
        env.getCheckpointConfig().setMinPauseBetweenCheckpoints(3000);// 确认 checkpoints 之间的时间会进行 ** ms
        env.getCheckpointConfig().setTolerableCheckpointFailureNumber(5);
        env.setRestartStrategy(RestartStrategies.fixedDelayRestart(3, Time.of(10,TimeUnit.SECONDS)));//重启策略:重启3次,间隔10s

        //3.从kafka中读取日志信息,将将每行数据转换为JavaBean对象 主流
        DataStreamSource<String> dataStream = env.addSource(KafkaUtils.getKafkaSource(KAFKA_SOURCE_TOPIC, KAFKA_SOURCE_GROUP));
        ............
        //8.读取HBase中user表,进行维度关联
        SingleOutputStreamOperator<CephAccessRecord> record = AsyncDataStream.unorderedWait(
                validDS,
                new DimAsyncFunction<CephAccessRecord>() {
                    @Override
                    public String getKey(CephAccessRecord record) {
                        return record.access_key;
                    }
                },
                60, TimeUnit.SECONDS);
        BucketAssigner<String, String> assigner = new DateTimeBucketAssigner<>("yyyy-MM-dd", ZoneId.of("Asia/Shanghai"));
        StreamingFileSink<String> fileSink = StreamingFileSink.<String>forRowFormat(
                new Path(HDFS_FILE_PATH),
                new SimpleStringEncoder<>("UTF-8"))
                .withRollingPolicy(
                        DefaultRollingPolicy.builder()
                                .withRolloverInterval(TimeUnit.DAYS.toMillis(1))//至少包含 20 分钟的数据
                                .withInactivityInterval(TimeUnit.DAYS.toMillis(1 ))//最近 20 分钟没有收到新的数据
                                .withMaxPartSize(1024 * 1024 * 1024)//文件大小已达到 1 GB
                                .build())
                .withBucketAssigner(assigner)
                .build();

        // 将record-->过滤上传数据-->转换成jsonstring-->写入到hdfs
//        allDataDS.filter(log->log.event_type.equals("upload")).map(line->JSON.toJSONString(line)).addSink(fileSink);
        dataStream.map(line->JSON.toJSONString(line)).addSink(fileSink);

        //10.流环境执行
        env.execute();

异步关联程序

java 复制代码
package com.data.ceph.function;

import org.apache.commons.beanutils.BeanUtils;
import org.apache.flink.configuration.Configuration;
import org.apache.flink.streaming.api.functions.async.ResultFuture;
import org.apache.flink.streaming.api.functions.async.RichAsyncFunction;
import org.apache.hadoop.hbase.*;
import org.apache.hadoop.hbase.client.*;
import org.apache.hadoop.hbase.security.User;
import org.apache.hadoop.hbase.util.Bytes;
import org.apache.hadoop.security.UserGroupInformation;

import java.util.Collections;
import java.util.Map;

public abstract class DimAsyncFunction<T> extends RichAsyncFunction<T, T> implements DimAsyncJoinFunction<T> {

    private org.apache.hadoop.hbase.client.Connection connection = null;
    private ResultScanner rs = null;
    private Table table = null;

    @Override
    public void open(Configuration parameters) throws Exception {
        //不启用安全认证
        System.setProperty("zookeeper.sasl.client", "false");
        Map<String, String> stringStringMap = getRuntimeContext().getExecutionConfig().getGlobalJobParameters().toMap();
        String hbase = stringStringMap.get("hbase_zookeeper_quorum");
        org.apache.hadoop.conf.Configuration hconf = HBaseConfiguration.create();
        hconf.set(HConstants.ZOOKEEPER_QUORUM, "172.16.23.37,172.16.23.38,172.16.23.39");
//        hconf.set(HConstants.ZOOKEEPER_QUORUM, hbase);
        hconf.set(HConstants.ZOOKEEPER_CLIENT_PORT, "2181");
        hconf.set(HConstants.ZOOKEEPER_ZNODE_PARENT, "/hbase");

        //指定用户名为hbase的用户去访问hbase服务
        UserGroupInformation userGroupInformation = UserGroupInformation.createRemoteUser("hive");
        connection = ConnectionFactory.createConnection(hconf, User.create(userGroupInformation));
        table = connection.getTable(TableName.valueOf("cloud:user_info"));
    }


    @Override
    public void asyncInvoke(T input, ResultFuture<T> resultFuture) throws Exception {
        Get get = new Get(Bytes.toBytes(getKey(input)));
        Result rs = table.get(get);
        for (Cell cell : rs.rawCells()) {
            String column = Bytes.toString(cell.getQualifierArray(), cell.getQualifierOffset(), cell.getQualifierLength());
            String value = Bytes.toString(cell.getValueArray(), cell.getValueOffset(), cell.getValueLength());
            BeanUtils.setProperty(input, column, value);
        }
        resultFuture.complete(Collections.singletonList(input));
    }
    @Override
    public void close() throws Exception {
        if (rs != null) rs.close();
        if (table != null) table.close();
        if (connection != null) connection.close();
    }
    @Override
    public void timeout(T input, ResultFuture<T> resultFuture) throws Exception {
        System.out.println("TimeOut:" + input);
    }
}
相关推荐
huaqianzkh36 分钟前
了解Hadoop:大数据处理的核心框架
大数据·hadoop·分布式
Kika写代码1 小时前
【Hadoop】【hdfs】【大数据技术基础】实验三 HDFS 基础编程实验
大数据·hadoop·hdfs
我的K84091 小时前
Flink整合Hive、Mysql、Hbase、Kafka
hive·mysql·flink
okmacong3 小时前
2024.11.12_大数据的诞生以及解决的问题
大数据
Java资深爱好者5 小时前
数据湖与数据仓库的区别
大数据·数据仓库·spark
heromps5 小时前
hadoop报错找不到主类
大数据·hadoop·eclipse
未 顾7 小时前
day12:版本控制器
大数据·elasticsearch·搜索引擎
CherishTaoTao7 小时前
Git别名设置
大数据·git
Dreams°1238 小时前
【大数据测试HBase数据库 — 详细教程(含实例与监控调优)】
大数据·功能测试·单元测试