The application consists of 3 processes:

• [twitter-stream]
• [load-data]
• [api]

• NODE_ENV : development
• LOGGER_LEVEL ('debug' | 'info'), default: info
• LOGGER_ENABLED ('true' | 'false'), default: true
• TWITTER_CONSUMER_KEY
• TWITTER_CONSUMER_SECRET
• TWITTER_ACCESS_TOKEN_KEY
• TWITTER_ACCESS_TOKEN_SECRET
• TWITTER_TRACK: comma separated keywords to track, eg. platzi,open source, etc
• RABBITMQ_URI:amqp://rabbitmq_service
• REDIS_URI:redis://redis_service
• REDIS_DATA_RETENTION_IN_MS:86400000
• PORT: 3000

Use the docker-compose file to be run.

docker-compose up -d

Listening on API twitter for keywords and sends the tweets to a RabbitMQ queue.

Listening on the RabbitMQ queue and saves the tweets to Redis.

The process is serving an API in GraphQL to return the tweets.

• http://localhost:3000

• Query to list tweets

query {
  get(key:"tweets"){
    text
    tweeter
    createdAt
  }
}