A simple example demonstrating the usage of Kafka plugin API where consumer is started after the message producer.
To start the example you have to have Kafka broker running first.
Furthermore, the example assumes that each auto-created topic has
at least 3 partitions (num.partitions >= 3 in server.properties).
If you don't have Kafka installed locally, you can use the following
docker image in combination with a prepared server.properties config
file:
sudo docker create -p 2181:2181 -p 9092:9092 --name kafka --rm \
--env ADVERTISED_HOST=172.17.0.1 --env ADVERTISED_PORT=9092 spotify/kafka
KAFKA_VERSION=$(docker inspect -f '{{ .Config.Env }}' kafka | tr ' ' '\n' | grep KAFKA_VERSION | sed 's/^.*=//')
SCALA_VERSION=$(docker inspect -f '{{ .Config.Env }}' kafka | tr ' ' '\n' | grep SCALA_VERSION | sed 's/^.*=//')
sudo docker cp server.properties kafka:/opt/kafka_${SCALA_VERSION}-${KAFKA_VERSION}/config/server.properties
sudo docker start kafka
It will bring up Kafka broker listening on port 9092 for client communication.
To run the example, type:
go run main.go deps.go [-kafka-config <config-filepath>]
If kafka-config is unspecified, the application will try to search
for kafka.conf in the current working directory.
If the configuration file cannot be loaded or is not defined, default
configuration will be used.