Skip to content

Docker

Jump to bottom
Dave Bauman edited this page Mar 7, 2022 · 4 revisions

Docker

Overview

Running Insights Explorer with Docker allows you to easily spin up an instance without dealing with dependencies or building the project. It isolates it to a container without exposing it to the rest of your system.

This is the recommended way to install and use Insights Explorer. Official Docker images are automatically build and hosted on GitHub.

Make sure you have Docker installed and set up on your machine before following these instructions. If you don't already have Docker installed, follow the official install instructions for Linux, macOS, or Windows here: https://docs.docker.com/install/#supported-platforms.

Images

The following Docker images are available:

Image Purpose
insights-explorer Main service; serves the frontend and GraphQL backend
insights-explorer-convertbot Convertbot: provides file format conversions
insights-explorer-slackbot Slackbot: connects to Slack to provide unfurling

Docker Compose

An example docker-compose.yml is included in the project root. It includes Insights Explorer and all dependencies, and is pre-configured to be as easy to run as possible.

⚠️ The provided docker-compose.yml is intended for dev/test usage and is not recommended for production use as-is. It doesn't provide data redundancy, doesn't use HTTPS, and has hard-coded credentials.

Configuration

While mostly pre-configured, there are a few environment variables that must be configured prior to running the docker-compose.yml.

  1. Provide a GitHub Service Account
  2. Provide a GitHub Organization
  3. Provide an OAuth provider; GitHub OAuth is free and easy

Usage

The docker-compose.yml provides several services, grouped into different profiles to make it easy to customize.

Launch all services using the profile all:

docker compose --profile all up

Containers can be started in detached mode by adding the -d flag:

docker compose --profile all up -d

Once the containers have started, you can access the following URLs:

All running containers can be stopped with Ctrl-C, or when detached:

docker compose --profile all down

This doesn't remove the data volumes (see below).

🎯 Tip: Use the same --profile flag when running docker compose down or you may get an error

Profiles

Services in the docker-compose.yml file are grouped into a few profiles. This makes it easy to run a subset of services without modifying the file.

  • all: Runs all services
  • iex: Runs just Insights Explorer services, without any dependencies
  • infra: Runs the infrastructure dependencies (PostgreSQL, Elasticsearch, Minio)
  • extras: Runs optional services (e.g. Kibana)

One scenario where this is useful is for local development on IEX itself:

docker compose --profile infra up -d
npm start

Configure the .env.development.local file to connect to the running services on localhost.

Multiple profiles can be run by specifying --profile twice:

docker compose --profile infra --profile extras up

In addition, there are single-service profiles: es, db, minio, kibana.

Volumes

Four persistent volumes are created:

  • iex_db_data: Contains the PostgreSQL database
  • iex_es_data: Contains Elasticsearch data
  • iex_kibana_data: Contains Kibana data
  • iex_minio_data: Contains the S3-compatible object storage

These volumes persist by default even after the containers are stopped. This maintains state if the containers are restarted later.

Use docker compose down --volumes to remove them for a clean start.

Clone this wiki locally