如何使用 kafka 流加入主题

Question

Kafka stream I am trying to stream but I've some issues, it doesn't work. To get started, I have three connectors but I can't use my own keys. I need the keys to join them, right? How can I join with 2 or more keys? I try to replicate something like this: select * from (select a. * from users a inner join deps b on a.dep = b.dep and a.group = b.group ) a inner join user_afy on a.id = b.id

I want to save the data the internal join in a topic and use it for the external join. This is an example I have.

Connector Properties:

....
connector.class=io.confluent.connect.jdbc.JdbcSourceConnector
mode=timestamp
query=select id, user, dep,tal, group,time from users
numeric.mapping=best_fit
table.types=TABLE
topic=users  
// I try use this with 1 or more fields but not worked  
transforms=createKey, extractInt  
transforms.createKey.type=org.apache.kafka.connect.transforms.ValueToKey  
transforms.createKey.fields=dep, group  
transforms.extractInt.type=org.apache.kafka.connect.transforms.ExtractField$Key  
transforms.extractInt.field=dep, group  

standalone.properties

bootstrap.servers=localhost:9092  
key.converter=org.apache.kafka.connect.json.JsonConverter  
value.converter=org.apache.kafka.connect.json.JsonConverter  
key.converter.schemas.enable=false  
value.converter.schemas.enable=false  
offset.storage.file.filename=D:/tmp/connect.offsets  
plugin.path=D:/connector/lib  

Topics:

Topic users    
{"id":"0001", "user":"Alex", "dep":"ofi", "postal":170, group="ingen",time:"xxx"}    
{"id":"0002", "user":"Emy", "dep":"lab", "postal":170, group="itn",time:"xxx"}    
{"id":"0003", "user":"Lea", "dep":"lab", "postal":170, group="itn",time:"xxx"}    
{"id":"0004", "user":"Silva", "dep":"cent", "postal":170, group="ingen",time:"xxx"}    
{"id":"0005", "user":"Foxy", "dep":"cent", "postal":170, group="ete",time:"xxx"}    

topic user_afy
{"id":"0001", name="bask"}
{"id":"0001", name="Silf"}
{"id":"0002", name="BTT"}
{"id":"0005", name="butf"}


Topic deps  
{"id_dep":"1", "dep":"ofi", "sind"="worker", "group"="ingen."}  
{"id_dep":"2", "dep":"lab", "sind"="worker", "group"="iti."}  
{"id_dep":"3", "dep":"cent", "sind"="worker", "group"="etc."} 

My code is an example of the official website but I can't test it

public static void main(String[] Args) {
        Properties props = new Properties();
        props.put(......);

    final Serializer<JsonNode> jsonSerializer = new JsonSerializer();
        final Deserializer<JsonNode> jsonDeserializer = new JsonDeserializer();
        final Serde<JsonNode> jsonSerde = Serdes.serdeFrom(jsonSerializer, jsonDeserializer);
        final Consumed<String, JsonNode> consumed = Consumed.with(Serdes.String(), jsonSerde);
        StreamsBuilder builder = new StreamsBuilder();
        final KStream<String, JsonNode> left = builder.stream("user", consumed);
        KTable<String, JsonNode> right = builder.table("deps", consumed);
        KStream<String, String> joined = left.join(right,
            (leftValue, rightValue) -> "left=" + leftValue + ", right=" + rightValue,
            Joined.with(Serdes.String(), jsonSerde, jsonSerde)
        );
//Edit
       joined.foreach((k, v) -> {
          System.out.println("key="+k+ ", val=" + v);
       });

}

The output, how would it show? To create a new topic, is a hashmap preferable with the values ​​you want to save in json format? Later I will create custom Serdes

Answer 1

What do you mean by "I can't use my own keys?". In Kafka Streams you can always set a new key as required for your processing.

If you want to read data into a KTable, you cannot change the key in a straightforward way though. You would need to read the topic as KStream, set a new key, and convert the KStream into a KTable (cf Kafka Streams API: KStream to KTable ).

For multiple consecutive joins you can just "chain" the corresponding operations together.

builder.stream("topic-1").selectKey(...).to("table-topic-1");
KTable t1 = builder.table("table-topic-1");

KStream firstJoinResult = builder.stream(...).selectKey(...).join(t1, ...).

builder.stream("topic-2").selectKey(...).to("table-topic-2");
KTable t2 = builder.table("table-topic-2");

firstJoinResult.selectKey(...).join(t2, ...).to("result-topic");