dataops-blue-green-deployment

What problem does it solve?

This Skill helps you deploy data platform changes without breaking production by enabling controlled traffic cutovers, backward-compatible migrations, and fast rollback paths.

Core Features & Use Cases

• Kubernetes Blue/Green: Run blue and green deployments in parallel and switch traffic via Service selector, including a smoke-test-first cutover script.
• Argo Rollouts Blue/Green: Use preview and active services with pre-promotion analysis and manual promotion to reduce release risk.
• dbt Blue/Green Schema Swap: Build models in a shadow schema, test them, then atomically swap schemas using a swap_schema macro pattern.
• Migration Compatibility Patterns: Execute backward-compatible DDL changes using expand/backfill/contract or provide view-shim compatibility for incompatible changes.
• Airflow DAG Versioning: Transition with versioned dag_ids, run both versions during validation, pause old DAGs for rollback, then clean up.
• Decision Guide: Choose blue/green vs canary based on risk tolerance, migration constraints, rollback speed, and infra cost.

Quick Start

Ask the agent to produce a blue/green plan for your next production release, including backward-compatible database migration steps, a Kubernetes (or Argo Rollouts) cutover procedure, and a dbt shadow-schema swap approach with smoke tests before promotion.