kubernetes-deployment-patterns

What it does

This skill enables AI agents to generate Kubernetes deployment configurations using specific rollout strategies like RollingUpdate, Recreate, and Blue/Green. It ensures applications deploy safely with defined constraints for availability and resource management.

When to use it

• Zero-downtime updates: Implementing RollingUpdate patterns to replace pods gradually without interrupting service traffic.
• Database migrations: Using the Recreate pattern to ensure all instances stop before a deployment, preventing data inconsistency during schema changes.
• High-availability testing: Setting up Blue/Green deployments to switch between stable and new versions instantly for rapid validation.
• Resource-constrained environments: Configuring specific replica counts and update timeouts to prevent cluster overload during scaling events.

Key capabilities

• Generates kubectl apply commands for various deployment strategies.
• Defines maxSurge and maxUnavailable parameters for fine-grained control over rollout speed.
• Supports pre-stop hooks and post-start lifecycle actions within deployment manifests.
• Configures readiness probes to ensure new pods pass health checks before traffic is routed.

Example prompts

• "Create a Kubernetes deployment manifest using the RollingUpdate strategy with a maxSurge of 25% and a maxUnavailable of 0."
• "Generate a Recreate deployment configuration for a legacy application that requires zero concurrent instances during updates."
• "Write a Blue/Green deployment setup that routes traffic to the new version only after successful health checks pass."

Tips & gotchas

Ensure your Kubernetes cluster supports the specific deployment strategy you request, as some patterns require additional API server permissions or controller versions. Always test rollout parameters in a staging environment first to verify they align with your actual pod startup times and network latency.