Apache Flink 与 Elasticsearch 的集成

Question

我正在尝试将 Flink 与 Elasticsearch 2.1.1 集成，我正在使用 maven 依赖项

     <dependency>
        <groupId>org.apache.flink</groupId>
        <artifactId>flink-connector-elasticsearch2_2.10</artifactId>
        <version>1.1-SNAPSHOT</version>
    </dependency>

这是Java代码，我正在从Kafka队列中读取事件（工作正常），但不知何故，事件没有发布在Elasticsearch中，也没有错误，在下面的代码中，如果我更改与ElasticSearch的端口，主机名，集群名称或索引名称相关的任何设置，那么我立即看到一个错误，但目前它没有显示任何错误，也没有任何新文档得到在 ElasticSearch 中创建

       StreamExecutionEnvironment env = StreamExecutionEnvironment.getExecutionEnvironment();
    // parse user parameters
    ParameterTool parameterTool = ParameterTool.fromArgs(args);
    DataStream<String> messageStream = env.addSource(new FlinkKafkaConsumer082<>(parameterTool.getRequired("topic"), new SimpleStringSchema(), parameterTool.getProperties()));
    messageStream.print();
    Map<String, String> config = new HashMap<>();
    config.put(ElasticsearchSink.CONFIG_KEY_BULK_FLUSH_MAX_ACTIONS, "1");
    config.put(ElasticsearchSink.CONFIG_KEY_BULK_FLUSH_INTERVAL_MS, "1");
    config.put("cluster.name", "FlinkDemo");
    List<InetSocketAddress> transports = new ArrayList<>();
    transports.add(new InetSocketAddress(InetAddress.getByName("localhost"), 9300));
    messageStream.addSink(new ElasticsearchSink<String>(config, transports, new TestElasticsearchSinkFunction()));
    env.execute();
}
private static class TestElasticsearchSinkFunction implements ElasticsearchSinkFunction<String> {
    private static final long serialVersionUID = 1L;
    public IndexRequest createIndexRequest(String element) {
        Map<String, Object> json = new HashMap<>();
        json.put("data", element);
        return Requests.indexRequest()
                .index("flink").id("hash"+element).source(json);
    }
    @Override
    public void process(String element, RuntimeContext ctx, RequestIndexer indexer) {
        indexer.add(createIndexRequest(element));
    }
}

Answer 1

我确实在本地机器上运行它并进行调试，但是，我唯一缺少的是正确配置日志记录，因为大多数弹性问题都在"log.warn"语句中描述。问题是 elasticsearch-2.2.1 客户端 API 中"BulkRequestHandler.java"中的异常，它抛出错误 -"org.elasticsearch.action.ActionRequestValidationException： 验证失败： 1： 缺少类型;"由于我已经创建了索引，但不是我觉得很奇怪的类型，因为它应该主要关注索引并默认创建类型。

Answer 2

我找到了一个非常好的例子 Flink & Elasticsearch Connector

第一个Maven依赖项：

<dependency>
  <groupId>org.apache.flink</groupId>
  <artifactId>flink-connector-elasticsearch2_2.10</artifactId>
  <version>1.1-SNAPSHOT</version>
</dependency>

第二个示例 Java 代码

public static void writeElastic(DataStream<String> input) {
    Map<String, String> config = new HashMap<>();
    // This instructs the sink to emit after every element, otherwise they would be buffered
    config.put("bulk.flush.max.actions", "1");
    config.put("cluster.name", "es_keira");
    try {
        // Add elasticsearch hosts on startup
        List<InetSocketAddress> transports = new ArrayList<>();
        transports.add(new InetSocketAddress("127.0.0.1", 9300)); // port is 9300 not 9200 for ES TransportClient
        ElasticsearchSinkFunction<String> indexLog = new ElasticsearchSinkFunction<String>() {
            public IndexRequest createIndexRequest(String element) {
                String[] logContent = element.trim().split("t");
                Map<String, String> esJson = new HashMap<>();
                esJson.put("IP", logContent[0]);
                esJson.put("info", logContent[1]);
                return Requests
                        .indexRequest()
                        .index("viper-test")
                        .type("viper-log")
                        .source(esJson);
            }
            @Override
            public void process(String element, RuntimeContext ctx, RequestIndexer indexer) {
                indexer.add(createIndexRequest(element));
            }
        };
        ElasticsearchSink esSink = new ElasticsearchSink(config, transports, indexLog);
        input.addSink(esSink);
    } catch (Exception e) {
        System.out.println(e);
    }
}