Appeals Consumer

Appeals Consumer is an Rails application responsible for processing intake data sent by VBMS. When VBMS handles an intake, they will pushlish a message containing the relevent data to a Apache Kafka platform. This application will consume that data through a topic subscription. Appeals Consumer is also responsible for transforming the data to be sent to the Caseflow application.

• Appeals Consumer
  • Developer Setup
    • Machine setup
    • Development walkthrough
  • Building the containers
  • Starting the containers
  • Kafka Broker instance

Developer Setup

Machine setup

• Mac
• Windows - TBD

Development walkthrough

Building the containers

Running make build will build the initial docker containers. This will:

• Copy all necessary files into the docker container.
• Install required dependencies.
• Create the database if it has not already been created.
• Run any pending migrations

Additionally, run make registry to upload the schema to the registry server.

Starting the containers

Running make up will start up the docker compose environment for development. Any changes to the local code base will be reflected in the docker container with some exceptions. New migrations will need to be migrated to the containerized Postgres instance. Additionally, any modifications to the container environment, such as modifying the Dockerfile or supporting system level dependencies will require a rebuild of the containers.

Appeals Consumer utuilizes Supervisor to run multiple processes in parallel. The Rails server runs along side the Karafka consumer server when you run make up. The logs for these processes can be found in the project's /log directory, /log/rails and /log/kafka.

Alternatively, the Karafka consumer server and rails server can be run separately by running make consumer or make rails.

Kafka Broker instance

The Appeals Consumer docker compose includes a Kafka broker. The broker im age is cp-kafka which is downloaded from confluent. cp-kafka is based off of librdkafka and can be configured with all of the same properties. The cp-kafka is configured via environmental variables in the docker-compose file using their documented variable conversion rules.

There is an instance of Confluent's Registry Server included. The Registry Server can validate schemas for selected topics against a provided AVRO.

For local development, Kafka-UI by Provectus is included. This is a tool to help manage a Kafka Cluster through a web UI. While running, the UI can be found at localhost:8080. In the UI you can