Skip to content
José da Cruz

Enterprise Architect, and Life Thinker

José da Cruz

Enterprise Architect, and Life Thinker

Feature Flags as a Deployment Strategy: Deploy Dark, Release When Ready | ★ Bonus Episode

josedacruz, May 23, 2026May 23, 2026

Most deployment strategies tie the moment of shipping code to the moment of releasing a feature. Feature flags break that assumption entirely.

With feature flags, you deploy code to production in an inactive state — hidden behind a conditional check. The code is live on the server, but no user ever sees it. There’s no new pod, no load balancer rule, no infrastructure event. Later — hours, days, or weeks after that deployment — you flip a flag in a dashboard. The feature appears instantly, for exactly the users you choose. No pipeline run. No engineer required.

This model is called “deploy dark”. It’s the mental shift from thinking about deployments as release events to thinking about deployments as safe, continuous code delivery — where the release is a separate, product-owned decision.

In this bonus episode, we cover:

  • Why deploy ≠ release — and why that distinction changes everything
  • The 4 types of feature flags (release, experiment, ops, permission) and when to use each
  • How flag evaluation works at runtime, and why it’s <1ms
  • The full lifecycle from dark deployment to flag cleanup
  • How feature flags compare to every strategy in the main series
  • LaunchDarkly and Unleash implementation patterns
  • The biggest risk nobody talks about: flag debt

⭐ This episode follows a different path than the series. The 6-part series is about infrastructure-level deployment control. Feature flags are about application-level release control. Both are essential. They’re often used together.


📺 The Deployment Strategies Series This Bonus Extends

  • #1 → Recreate Deployment
  • #2 → Rolling Update
  • #3 → Blue/Green Deployment
  • #4 → Canary Release
  • #5 → A/B Testing Deployment
  • #6 → Shadow Deployment
  • ★ Bonus → Feature Flags (you’re here)

If you haven’t watched the original series, start at Episode 1 — it builds progressively from the simplest strategy to the most advanced. This bonus episode stands on its own but rewards the context from the full series.

Related

architecture

Post navigation

Previous post
Next post

Leave a Reply Cancel reply

Your email address will not be published. Required fields are marked *

  • architecture (167)
  • artificial-intelligence (6)
  • books (2)
  • courses (4)
  • decision-frameworks (4)
  • finances (1)
  • java (8)
  • life-improvement (20)
  • metrics (3)
  • observability (2)
  • puzzles and challenges (3)
  • reviews (1)
  • security (8)
  • Soft Skills (1)
  • spring-boot (36)
  • Systems Thinking (3)
  • GitHub Weekly: AI Agent Tools Trending Right Now (November 2 – November 8, 2026)
  • Designing a Distributed Session Store: Why Sticky Sessions Break Under Load
  • Essential Reading: Release It! — Why Production Is a Different Game Than Development
  • GitHub Weekly: Top 10 Trending Repos (October 26 – November 1, 2026)
  • GitHub Weekly: Architecture & Infra Tools Worth a Look (October 19 – October 25, 2026)
©2026 José da Cruz | WordPress Theme by SuperbThemes