Kafkamion

Table of content

1. Architecture
2. Topics
3. Topic description
4. Agregation
5. Topic result
6. How to run

Architecture

Topics

1. Entity registration

{
    "type": "string",
    "driver_id": "string",
    "first_name": "string",
    "last_name": "string",
    "email": "string",
    "phone": "string"
}

or

{
    "type": "string",
    "truck_id": "string",
    "immatriculation": "string"
}

2. Time Registration time_registration

{
    "type": "string",
    "timestamp": "string",
    "driver_id": "string",
    "truck_id": "string"
}

3. Position

{
    "truck_id": "string",
    "latitude": "number",
    "longitude": "number",
    "timestamp": "string"
}

Topic description

Each Topic is defined as follow :

1. Topic 1 The type can be driver or truck and will be handled differently. The key is the driver_id or the truck_id and the value is the information of the driver or the truck. This topic is called entity_topic

2. Topic 2 This topic is used to store the time registration of the driver. The key is the type of the time registration and the value is the timestamp. This topic is called time_registration_topic

3. Topic 3 This topic is used to store the position of the driver. The key is the truck_id and the value is the position of the driver. This topic is called position_topic

Agregation

The aggregation is done by the driver_id between the first topic and the second. The aggregation is done by the truck_id for the second and the third topic.

Topic result

The result of the merge, called report, will be a json flat topic of our three producers topics. The topic is called report_topic

{
    "driver_id": "string",
    "first_name": "string",
    "last_name": "string",
    "email": "string",
    "phone": "string",
    "truck_id": "string",
    "immatriculation": "string",
    "start_time": "string",
    "end_time": "string",
    "rest_time": "string",
    "latitude_start": "number",
    "longitude_start": "number",
    "timestamp_start": "string",
    "latitude_end": "number",
    "longitude_end": "number",
    "timestamp_end": "string",
    "latitude_rest": "number",
    "longitude_rest": "number",
    "timestamp_rest": "string"
}

How to run

We use docker-compose to run our system. To run the stack, you need to run the following command:

1. Up the stack

docker-compose up -d

2. Connect to a bash container

docker exec --workdir /opt/kafka/bin/ -it <shaKafka> sh

Note: <shaKafka> is the sha of the kafka container. You can get it by running docker ps

3. Create the topics

/opt/kafka/bin/kafka-topics.sh --bootstrap-server localhost:9092 --create --topic position_topic --partitions 1 --replication-factor 1
/opt/kafka/bin/kafka-topics.sh --bootstrap-server localhost:9092 --create --topic time_registration_topic --partitions 1 --replication-factor 1
/opt/kafka/bin/kafka-topics.sh --bootstrap-server localhost:9092 --create --topic report_topic --partitions 1 --replication-factor 1
/opt/kafka/bin/kafka-topics.sh --bootstrap-server localhost:9092 --create --topic entity_topic --partitions 1 --replication-factor 1

4. Mock the producer

./kafka-console-producer.sh --bootstrap-server localhost:9092 --topic time_registration_topic

Note: You can replace time_registration_topic by position_topic or entity_topic