Kubernetes Configuration Files

Overview

• Kubernetes configuration files are essential for creating and configuring components in a Kubernetes cluster.
• Consists of three main parts: metadata, specification, and status.
• Files are written in YAML, which is strict about indentations.

Components of a Configuration File

1. Metadata

• Contains information about the component, such as its name.
• Key-value pairs are used as labels.

2. Specification

• Defines the desired configuration for the component.
• Attributes vary depending on the component:
  • Deployment: includes replicas, template (which further includes metadata and specifications for pods), image, ports, etc.
  • Service: includes ports, targetPort, selector for connecting to deployments or pods.

3. Status

• Automatically generated by Kubernetes.
• Reflects the actual state of the component.
• Used for self-healing by comparing with the desired state to fix discrepancies.

Key Concepts

API Version

• Different components may require different API versions.

Self-Healing

• Kubernetes ensures the actual state matches the desired state.
• Example: If a deployment specifies 2 replicas, Kubernetes ensures 2 replicas are running.

Etcd

• A key-value store used by Kubernetes to store cluster data, including status information.

YAML Structure

• YAML format is simple but requires correct indentation.
• Online validators can help with formatting large YAML files.

Best Practices

• Store configuration files with application code as part of "Infrastructure as Code."
• Can also be stored in a separate Git repository.

Working with Deployments and Services

• Deployments manage and configure pods.
• Services require:
  • A port where the service is accessible.
  • Target port specifying where to forward requests.
  • Connection is established through labels and selectors.

Managing Configuration Files

• kubectl apply: To create components from configuration files.
• kubectl delete: To remove components using configuration files.

Practical Commands

• kubectl get pods to see running replicas.
• kubectl describe service <service-name> to validate service endpoints.
• kubectl get pod -o wide for extended information including IP addresses.
• kubectl get deployment <deployment-name> -o yaml to see the full configuration including status.

Debugging and Updating

• Automatically generated information such as timestamps and IDs should be removed if reusing configuration files.

Conclusion

• Regularly use configuration files for deploying, updating, and managing Kubernetes components.
• Videos are part of a series, with new content released weekly.

Additional Resources

• Video series subscription for updates.
• Comments section for questions and clarifications.