Status

Current state: Under Discussion

Discussion thread: here [Change the link from the KIP proposal email archive to your own email thread]

JIRA: here

Please keep the discussion on the mailing list rather than commenting on the wiki (wiki discussions get unwieldy fast).

Motivation

Currently, while building a new Topology through the KStreams DSL the processors are automatically named.

The genarated names are prefixed depending of the operation (i.e KSTREAM-SOURCE, KSTREAM-FILTER, KSTREAM-MAP, etc).  Moreover, they are added with a suffix like "-0000000000" to guarantee their uniqueness.

To debug/understand a topology it is possible to display the processor lineage with the method Topology#describe(). However, a complex topology with dozens of operations can be hard to understand if the processor names are not relevant.

So allowing users to be able to set more meaningful names can help to resolve complexity of some developed topologies. For example, a processor name could describe the business rule performed by a map() operation.

In addition, the generated names have a few disadvantages to guarantee topology compatibilities. In fact, adding a new operator, using a third-library doing some optimization to remove some operators or upgrading to a new KafkaStreams version with internal API changes may changed suffix indexing for a large amount of the processor names. This will in turn change the internal state store names, as well as internal topic names as well.

Public Interfaces

First, we propose to add one new class Described that can be used to provide the optional parameters when processing each record (i.e to set the processor name). The objective to create a new class is to keep consistent with the overall API design.

package org.apache.kafka.streams.kstream;

import java.util.Objects;

/**
 * This class is used to provide the optional parameters when processing each record.
 */
public class Described {

    protected final String processorName;

    protected Described(final String processorName) {
        this.processorName = processorName;
    }

    protected Described(final Described described) {
        this(described.getProcessorName());
    }

    /**
     * Create a Described instance with provided processor name.
     * @param processorName      name to use to describe the {@link org.apache.kafka.streams.processor.Processor}.
     * @return  A new {@link Described} instance configured withName processorName
     */
    public static Described withName(final String processorName) {
        return new Described(processorName);
    }

    @Override
    public boolean equals(Object o) {
        if (this == o) return true;
        if (o == null || getClass() != o.getClass()) return false;
        Described described = (Described) o;
        return Objects.equals(processorName, described.processorName);
    }

    @Override
    public int hashCode() {
        return Objects.hash(processorName);
    }

    public String getProcessorName() {
        return processorName;
    }
}

This class will be then used by a new method as(Processed)  added to KStream and KTable classes : 

/**
 * Describe the record-by-record operation previously made to obtain this stream.
 * This is not an operation. This method can be used to customized a processor name return
 *
 * <p>
 * For example you can used this method to customize the name of the last generated processors applied to this stream.
 * The given processor name will display in the description of the {@code Topology} built.
 * <pre>{@code
 * KStream<String, String> inputStream = builder.stream("topic");
 * KStream<String, String> outputStream = inputStream.map(new KeyValueMapper<String, String, KeyValue<String, String>> {
 *     KeyValue<String, String> apply(String key, String value) {
 *         return new KeyValue<>(key, value.toUpperCase());
 *     }
 * });
 * outputStream.as(Described.withName("MAP-VALUE-TO-UPPERCASE"))
 * }</pre>
 * </p>
 * @param described the options to describe the processor operation.
 * @return a {@code KStream}
 */
KStream<K, V> as(Described described);

Stateless operations

Stateless operations like stream(), map(), filter(), generate a single underlaying processor. The method depicted above will directly change the processor name.

Statefull operations

Stateful operations like table(), join() are translated to multiple processors (aka. nodes). The given name should be used to describe the first processor and as a prefix for all subsequent processors.

The method branch() results in multiple streams that can be described with the used of the new added method.

However to be able to custome the name of the first translated processor we propose to overload this method with a one accepting a Described instance in argument.

Sink operations

The Sink operations returning Void can't use the method depicted above.

Currently, KStream<K, V> interface defines the following methods : 

• void foreach(final ForeachAction<? super K, ? super V> action); 
• void process(final ProcessorSupplier<? super K, ? super V> processorSupplier, final String... stateStoreNames);

We propose to overload each of this method to accept a Described argument.

Then, for those  three methods we propose to add a new method withProcessorName() to classes Produced and Printed in order to minimize the number of overloaded methods.

• void print(final Printed<K, V> printed)
• void to(final String topic, final Produced<K, V> produced)
• void to(final TopicNameExtractor<K, V> topicExtractor, final Produced<K, V> produced)
  

Example for the Produced class : 

/**
 * Create an instance of {@link Produced} with provided processor name.
 *
 * @param processorName the processor name to be used. If {@code null} a default processor name will be generated
 * @param <K>         key type
 * @param <V>         value type
 * @return a new instance of {@link Produced}
 */
public static <K, V> Produced<K, V> with(final String processorName) {
    return new Produced<>(null, null, null, null, processorName);
}

/**
 * Configure the instance of {@link Produced} with a key {@link Serde}.
 *
 * @param processorName the processor name to be used. If {@code null} a default processor name will be generated
 * @return this
 */
public Produced<K, V> withProcessorName(final String processorName) {
    this.processorName = processorName;
    return this;
}

Then we propose to add new methods to the existing classes : Consumed, Produced, Joined and Printed in order to be able to define custom processor names.

Proposed Changes

• Implement the new class Described and update the KStream/KStreamImpl, KTable/KTableImpl methods that add stateless processor (i.e map(), mapValue(), filter()...).
• Update the Consumed, Produced and Joined classes to be able to set a processor name.
• The processor names specified by developer will be used in place of the static processor prefix. Statics prefixes will still be used if no custom processor name are specified.

Below is an application example : 

final StreamsBuilder builder = new StreamsBuilder();


KStream<String, String> stream = builder.stream("topic-input");
        stream.as(Described.withName("STREAM-FROM-TOPIC-INPUT"))
        .filter( (k, v) -> true ).as(Described.withName("FILTER-NULL-VALUE"))
        .map( (k, v) -> KeyValue.pair(k, v.toUpperCase())).as(Described.withName("MAP-TO-UPPERCASE"))
        .to("topic-output");

System.out.println(builder.build().describe());
---- (output)----
Topologies:
   Sub-topology: 0
    Source: STREAM-FROM-TOPIC-INPUT (topics: [topic-input])
      --> FILTER-NULL-VALUE
    Processor: FILTER-NULL-VALUE (stores: [])
      --> MAP-TO-UPPERCASE
      <-- STREAM-FROM-TOPIC-INPUT
    Processor: MAP-TO-UPPERCASE (stores: [])
      --> KSTREAM-SINK-0000000003
      <-- FILTER-NULL-VALUE
    Sink: KSTREAM-SINK-0000000003 (topic: topic-output)
      <-- MAP-TO-UPPERCASE

Compatibility, Deprecation, and Migration Plan

No compatibility issues foreseen.

Rejected Alternatives

1. The first proposition was to overload all stateless methods to accept an instance of Described class. However this solution was resulting in modiying a large percentage of the existing KStream and KTable methods.