Run self-hosted Dingolytics today

Dingolytics is an open-source tool to collect and store analytical data on product usage and metrics — highly customizable, interoperable, and developer-friendly.

The beauty of open-source solutions is that you can control every piece of it: you control the data flow and infrastructure. But it can also be challenging to run complex software on your own.

To make it easier, we provide a recipe for Docker/Docker Compose-based deployment. You can find it in Dingolytics/selfhosted @ GitHub repository.

Essentially, the deployment boils down to two steps on your Linux or Unix-like server:

1. Creating a .env file with your configuration
2. Running docker compose up -d --build

Let's explore the specific configuration settings in detail.

Application version

While APPLICATION_SERVER_TAG and APPLICATION_FRONTEND_TAG can be set to latest, we recommend using specific tags as shown in the example for better version control and predictability.

Caddy proxy (HTTPS)

In the example localhost is used but for actual deployment you would probably want to specify web host and email for a Let's Encrypt certificate. We use Caddy web server in the example, so the following options are used in Caddyfile.

# Web host configuration in .env
WEB_HOST=my.realdomain.tld
WEB_HOST_LETSENCRYPT_EMAIL=me@realmail.tld

Please make sure you've added DNS records first to resolve the domain name to your server. If you prefer using Nginx or other HTTP/HTTPS server you could replace Caddy and ajust the configuration accordingly.

Vector configuration

Note the VECTOR_INGEST_URL setting, it should specify a base URL for ingesting data, e.g. VECTOR_INGEST_URL=https://my.realdomain.tld

Managed deployment

If you'd like we manage the deployment for you please contact us via team@dingolytics.com