Skip to content

Deployment

Introduction

Deployment is the process of delivering software from artifacts to running production environments. How you deploy directly impacts risk, downtime, and ability to respond to problems.

The CD Model treats deployment as a deliberate, controlled process with multiple strategies available depending on risk tolerance, infrastructure capabilities, and organizational maturity.

Deployment vs Release

Critical distinction:

  • Deployment (Stage 10): Code reaches production environment
  • Release (Stage 12): Features become available to users

Modern CD practices decouple these:

  • Deploy code to production (with features disabled via feature flags)
  • Release features gradually (enable via flags for specific user segments)

This decoupling enables:

  • Continuous deployment without continuous release
  • Risk reduction through gradual rollout
  • Instant feature disablement without redeployment
  • A/B testing and experimentation

Deployment in the CD Model

Stage 10: Production Deployment:

  • Deploy artifacts to production infrastructure
  • Choose deployment strategy based on risk and infrastructure
  • Execute deployment via Deploy Agents (Zone C)
  • Verify health checks and smoke tests

Stage 11: Live Monitoring:

  • Monitor application health in production
  • Track key metrics (errors, latency, throughput)
  • Automated rollback on threshold breaches
  • Manual rollback procedures available

Stage 12: Release Toggling:

  • Control feature exposure via feature flags
  • Gradual rollout to user segments
  • Instant disable if issues detected
  • Independent of deployment

Deployment Strategy Selection

Different situations call for different strategies:

Hot Deploy: Fast, simple, brief downtime acceptable

Rolling: Zero downtime, gradual instance updates

Blue-Green: Instant rollback, higher infrastructure cost

Canary: Gradual production validation, metrics-driven

Rings: Progressive user group rollout, organizational-scale

Feature Flags: Decouple deployment from release, runtime control

The right strategy depends on:

  • Risk tolerance
  • Infrastructure capabilities
  • Downtime requirements
  • Team maturity
  • Cost constraints

In This Section

Topic Description
Deployment Strategies Comprehensive explanation of deployment patterns with tradeoffs and use cases
Deployment Rings Progressive user group rollout pattern explained in depth
Feature Flags Feature flag lifecycle management from creation to cleanup
Rollback Procedures Emergency rollback execution guide with decision criteria and procedures
Incident Response Production incident response playbook with severity levels and templates

Next Steps

Quick Reference


Tutorials | How-to Guides | Explanation | Reference

You are here: Explanation — understanding-oriented discussion that clarifies concepts.