#Docker

Run snapr in a container instead of installing the binary on the host.

#Docker

docker run -d \
  --name snapr \
  -p 8080:8080 \
  -e SNAPR_ADMIN_PASSWORD=changeme \
  -v $(pwd)/snapr.yaml:/etc/snapr/snapr.yaml:ro \
  -v $(pwd)/backups:/var/backups \
  ghcr.io/maximseshuk/snapr:latest

snapr looks for snapr.yaml in /etc/snapr/ by default inside the container.

#Docker Compose

The repo ships a dev stack with MinIO (S3), Postgres, SFTP, and WebDAV containers — useful for trying every storage backend locally. The stack only starts the storage targets; snapr itself runs from source via pnpm run dev (see examples/dev/README.md).

git clone https://github.com/maximseshuk/snapr.git
cd snapr
pnpm install
pnpm run stack:up    # start storage containers
pnpm run dev         # run snapr against ./examples/dev/snapr.yaml

The dev config binds the API on :47100 with admin / admin — for local testing only; never reuse these in production.

A minimal compose for production looks like this:

services:
  snapr:
    image: ghcr.io/maximseshuk/snapr:latest
    restart: unless-stopped
    ports:
      - '8080:8080'
    environment:
      SNAPR_ADMIN_PASSWORD: ${SNAPR_ADMIN_PASSWORD}
      SNAPR_JWT_SECRET: ${SNAPR_JWT_SECRET}
    volumes:
      - ./snapr.yaml:/etc/snapr/snapr.yaml:ro
      - ./backups:/var/backups

#Volumes & paths

Inside the container snapr runs as user snapr (uid 1001). Mount:

• the config file read-only at /etc/snapr/snapr.yaml
• any source path you want to back up — read-only is fine
• any local storage path — must be writable by uid 1001. If the host directory is owned by another user, either chown 1001 it or run the container with a matching --user.

#Environment variables

Use them with env: references in snapr.yaml:

server:
  secret: env:SNAPR_JWT_SECRET
  auth:
    password: env:SNAPR_ADMIN_PASSWORD

jobs:
  - sources:
      - type: postgresql
        password: env:PG_PASSWORD

See Configuration → Overview for the full env-ref behaviour.