About the project

A robust IT infrastructure is inconceivable without a solid Disaster Recovery strategy. While virtual machine backups cover bare-metal or full-system restores, databases require a more granular approach: generating consistent, scheduled dumps.

To solve this, I utilize Databasys — a dedicated Open Source solution for centralized database backups, specifically targeting MySQL and PostgreSQL. The tool is distributed as a Docker container, making its deployment highly predictable and entirely independent of the host operating system.

Why Synology NAS?

The ideal location for storing database dumps is a dedicated Network Attached Storage (NAS) appliance, isolated from the primary compute nodes. Synology hardware fits this role perfectly due to its RAID support, Btrfs file system (with bit-rot protection), and native containerization capabilities.

Deploying Databasys directly on a Synology NAS establishes a pull-based backup scheme: the storage appliance autonomously retrieves dumps from production servers, minimizing overhead on the primary network.

🛠 Practical Guide: Deploying Databasys on Synology

The launch process consists of two straightforward steps.

Step 1: Installing Docker on Synology

In modern Synology DSM versions (7.2 and later), the Docker ecosystem has been revamped.

  1. Open the Package Center in the DSM web interface.
  2. Search for and install the official Container Manager package (formerly known as Docker).
  3. Use File Station to create a shared folder for configurations and dumps, e.g., /docker/databasys.

Step 2: Launching the Container

To maintain declarative infrastructure, using docker-compose is the best practice. Via Container Manager (Projects section) or SSH access to the NAS, create a docker-compose.yml file:

version: '3.8'

services:
  databasys:
    image: databasys/databasys:latest
    container_name: databasys_backup
    restart: unless-stopped
    environment:
      - DB_TYPE=postgres  # or mysql
      - DB_HOST=192.168.1.100
      - DB_PORT=5432
      - DB_USER=backup_user
      - DB_PASS=SecurePassword123
      - CRON_SCHEDULE=0 2 * * * # Run every night at 02:00 AM
      - RETENTION_DAYS=14       # Keep archives for 14 days
    volumes:
      - /volume1/docker/databasys/backups:/backups

Deploy the stack. Databasys will automatically connect to the specified database, generate a compressed dump, and store it on the secure Synology volume, while older archives are purged according to the retention policy.

Conclusion:

Implementing Databasys via Docker is an elegant and reliable way to handle database backups.

Autonomy: The container ships with all necessary dependencies and clients natively installed (pg_dump, mysqldump).

Security: Dumps are stored on an isolated NAS, protected from potential threats inside the primary network.

Open Source: The absence of licensing restrictions allows you to scale this solution across any number of databases.