kafka-streams-graphviz

Requirements

Java toolchain (JVM, Gradle, etc.)
Graphviz toolchain
Kafka Streams app

How to use

Here, we show how this library can be used to render a graph for your Kafka Streams topology.

Build your Kafka Streams topology

First, one must have a Kafka Streams Topology available:

StreamsBuilder builder = new StreamsBuilder();

... // This is where one would use builder to define a topology

Topology topology = builder.build();

(see TestKafkaStreamsTopology for a more complete example)

Describe as graphviz graph

Using the Kafka Streams Topology one can then use this library to describe this as a graphviz-java Graph:

Topology topology = builder.build();

Graph graph = GraphvizTopology
    .describe(topology)
    .asGraphvizGraph();

Render as DOT, Image, ...

This graphviz-java Graph can then be rendered, for example, as a DOT file, or PNG file:

Graph g = GraphvizTopology
    .describe(topology)
    .asGraphvizGraph();

Graphviz
    .fromGraph(g)
    .render(Format.DOT)
    .toFile(new File("examples/ex1.dot"));

Graphviz
    .fromGraph(g)
    .render(Format.PNG)
    .toFile(new File("examples/ex1.png"));

For TestKafkaStreamsTopology this yields the following image:

Finetuning with graphviz-java

Further finetuning of the Graph can be done programmatically by extending the Graph, e.g.:

Graph g = GraphvizTopology
    .describe(topology)
    .asGraphvizGraph();

Graph graphWithExtraInfo = g.with(
    graph()
        .graphAttr().with(Rank.inSubgraph(Rank.RankType.SAME))
        .with(node("input-topic-1"), node("input-topic-2"), node("input-topic-3"))
);

Graphviz
    .fromGraph(graphWithExtraInfo)
    .render(Format.DOT)
    .toFile(new File("examples/ex2.dot"));

See graphviz-java for more details.