README.md

Docker images for Bullet

Bullet is a "real-time query engine for very large data streams". More info in the official docs.
This repository contains the containerized versions of various components in Bullet's architecture.

Containerized Components

• Spark Backend
• Web Service
• UI

Yet to get containerized: Storm Backend.

Bullet Quick Start

This repo also contains a docker-compose.yaml to start the Bullet Quick Start application on Apache Spark Streaming.

Besides the Bullet components, the docker-compose.yaml also starts:

• kafka as Bullet PubSub (docker.io/bitnami/kafka)
• zookeeper for Kafka (docker.io/bitnami/zookeeper)
• kafdrop for Kafka monitoring/debugging (obsidiandynamics/kafdrop)
• spark with 1 master and 3 worker nodes as the streaming platform to run Bullet's Backend (docker.io/bitnami/spark)
• hadoop to provide hdfs for the checkpointing of Bullet's Spark Backend (big-data-europe/docker-hadoop)

Please note: The Bullet Spark Backend container is based on docker.io/bitnami/spark and also used to bring up the Apache Spark cluster itself to ensure all necessary jars are present.

Start All Containers

docker-compose up

Optionally, append --detach to run containers in the background.

Some containers might restart until the infrastructure containers (Spark, Kafka, Zookeeper, Hadoop) are fully running.

Stop All Containers

docker-compose down

Optionally, include cleanup parameters --rmi all --volumes --remove-orphans.