Kafka Streams App - 计数和总和

Question

我正在尝试从 KGroupedStream 创建一个 KTable 来存储每个键的值的总和。

 final StreamsBuilder builder = new StreamsBuilder();
 final KTable<String, Long> sum = builder.stream("streams-plaintext-input")
            .groupByKey()
            .aggregate(new Initializer<Long>() {
                @Override
                public Long apply() {
                    return Long.MIN_VALUE;
                }
            }, new Aggregator<String, Long, Long>() {
                @Override
                public Long apply(final String key, final Long value,final Long aggregate) {
                    aggregate += value;
                    return aggregate;
                }
            }, Materialized.<String, Long, KeyValueStore<Byte, byte[]>>as("counts-store"));

但我得到了错误：

The method aggregate(Initializer<VR>, Aggregator<? super Object,? super Object,VR>, Materialized<Object,VR,KeyValueStore<Bytes,byte[]>>) in the type KGroupedStream<Object,Object> is not applicable for the arguments (new Initializer<Long>(){}, new Aggregator<String,Long,Long>(){}, Materialized<String,Long,KeyValueStore<Byte,byte[]>>)

我见过的所有示例都将 Serde 作为第三个参数传递，但我已经尝试过这个并得到一个非常相似的错误（我认为这可能来自旧版本，因为它与当前的签名不匹配实施？）：

final StreamsBuilder builder = new StreamsBuilder();
    final KTable<String, Long> sum = builder.stream("streams-plaintext-input")
            .groupByKey()
            .aggregate(new Initializer<Long>() {
                @Override
                public Long apply() {
                    return Long.MIN_VALUE;
                }
            }, new Aggregator<String, Long, Long>() {
                @Override
                public Long apply(final String key, final Long value,final Long aggregate) {
                    aggregate += value;
                    return aggregate;
                }
            }, Serdes.Long());

错误：

The method aggregate(Initializer<VR>, Aggregator<? super Object,? super Object,VR>, Materialized<Object,VR,KeyValueStore<Bytes,byte[]>>) in the type KGroupedStream<Object,Object> is not applicable for the arguments (new Initializer<Long>(){}, new Aggregator<String,Long,Long>(){}, Serde<Long>)

我做错了什么？

使用Kafka版本：2.1.0

Answer 1

您的代码中存在一些问题：

1. 对于Materialized.as 而不是java.lang.Byte 你应该通过org.apache.kafka.common.utils.Bytes
2. 你不应该修改final变量：aggregate += value;
3. 你必须在StreamsBuilder::stream调用（builder.<String, Long>stream("streams-plaintext-input")）中添加键和值的类型

修改后应该大致如下：

KTable<String, Long> sum = builder.<String, Long>stream("streams-plaintext-input")
        .groupByKey()
        .aggregate(new Initializer<Long>() {
            @Override
            public Long apply() {
                return Long.MIN_VALUE;
            }
        }, new Aggregator<String, Long, Long>() {
            @Override
            public Long apply(final String key, final Long value,final Long aggregate) {
                return aggregate + value;
            }
        }, Materialized.<String, Long, KeyValueStore<Bytes, byte[]>>as("counts-store"));