dbtool-cli: Automating Database Ops with Go

In the world of DevOps, “Toil” (manual, repetitive work) is the enemy. I found myself manually creating database users for developers and writing brittle shell scripts for backups that eventually filled up the disk space.

To solve this, I built dbtool-cli, a unified command-line interface written in Go (Golang) that treats database management as code.

The Problems Solved

1. Developer Bottlenecks: Developers waiting for Ops to create database users/schemas.
2. Storage Exhaustion: Standard cron jobs keeping backups indefinitely until the disk is full.
3. Context Switching: Remembering the syntax differences between CREATE USER in Postgres vs MariaDB.

Key Capabilities

⚡ Automated Provisioning (Productivity)

When a new project starts, dbtool-cli handles the setup in seconds. It abstracts the complexity of different database drivers:

• Modular Architecture: Designed with Go interfaces to support multiple drivers (PostgreSQL, MariaDB) seamlessly.
• Secure Defaults: Auto-generates users with strictly scoped privileges.

🛡️ Smart Backup & Retention (Health)

The tool implements a “Backup Health Check” logic directly in Go:

• Verification: Validates archive integrity before success confirmation.
• Retention Policy: Automatically applies rotation (e.g., “Keep 7 Daily, 4 Weekly”) to prevent storage exhaustion.

Technical Implementation

I chose Go (Golang) and the Cobra library for this project to ensure:

• Portability: Compiled into a single static binary that runs on Linux servers, macOS workstations, and Windows.
• Performance: High-speed execution compared to interpreted bash scripts.
• Maintainability: Strongly typed code that is easier to extend than complex shell scripts.

Impact & Results

• 30% Increase in Productivity: Developers can provision their own local environments instantly using the CLI.
• Zero Storage Incidents: Smart retention policies ensure disk usage remains constant and predictable.
• Standardized Workflows: Unified command syntax across all database types.

View the Source Code on GitHub