Platform Engineering

Challenge between developers and operations team

• Developers: works in my laptop
• Operations: doesn't works in production

Solution: DevOps (ssh to the server and run your application)

Challenges of DevOps

• Complexity of the infrastructure
• AWS
• Azure
• GCP
• Kubernetes
• OVHcloud
• etc.

Solution: Platform Engineering

Internal Developer Platform (IDP)

• Platform tools and shared services
• Templates for developers
• Self-service tools
• Best defaults (Golden Path)

Benefits of Internal Developer Platform

• Developers can focus on writing code
• IT teams can focus on building the platform
• Consistency across all environments
• Unified way of working for all teams

Career Path:

• Developer
• Platform Engineer
• Infrastructure Engineer

Common Challenges in Software Development Teams

• Manual Kubernetes setup leads to technical debt and scaling issues.
• Managing multiple environments is challenging and inconsistent.
• Different teams using varied tools and workflows lack mutual support.
• Lack of self-service tools overwhelms IT teams with developer requests.
• Developer-created infrastructure can expose security vulnerabilities.

Platform Engineering Categories

• Infrastructure as Code (Terraform)
• Observability
  • Monitoring Metrics and Dashboards (Prometheus, Grafana)
  • Centralized Logging (Promtail, Loki)
  • Tracing (Jaeger)
  • Alerting (Alertmanager)
• Generic Deployment Helm Chart
• Centralized Ingress with SSL offloading (Ingress-Nginx)
• SSL Certificates Management (Cert-Manager)
• Role-Based Access Control
• GitOps Deployments (ArgoCD)
• Data Tools
  • Databases (PostgreSQL, Redis)
  • Message Brokers (Kafka, RabbitMQ)
• Secret Management (External-Secrets)
• DNS Management (ExternalDNS)
• Authentication and Authorization (Keycloak)
  • OIDC
  • Single Sign-On (SSO)
  • Single Log-Out
  • Identity Federation
• Kubernetes Policy Management (Kyverno)

  Two options for IDP:

  • Using Cloud-Specific Services
  • Using Kubernetes Clusters

  Cloud-Specific Services

  • Pros:
    • Reduced operational overhead
    • Seamless cloud integration
    • High availability
    • Built-in security
  • Cons:
    • Vendor lock-in
    • Limited customization
    • Higher costs
    • SLA and cloud policy dependency

  Inside the Kubernetes Cluster

  • Pros:
    • Greater control and customization
    • Consistent environments across clouds adn on-premises
    • Open-source tools
    • Avoid vendor lock-in
  • Cons:
    • Higher complexity
    • Requires Kubernetes knowledge
    • Scaling challenges
    • Manage security

Two categories of deployment

• Platform tools
• Developer Applications

Kubernetes Platform Architecture

• Mixed or Separate Kubernetes Cluster per environment (Dev, Test, Staging, Production)
• Mixed or Separate Kubernetes Cluster for Platform Tools and Developer Applications

GitOps

• The code repository is the single source of truth for the entire platform
• Track changes
  • Who made the change
  • When the change was made
  • What the change was
  • Why the change was made
  • Who approved the change
• In case of failure, the system can be rolled back to the last known working version
• ArgoCD is a popular GitOps tool for Kubernetes
  • Auto-Syncing
  • Auto-Pruning
  • Self-Healing

Kuberise.io

• an open-source Internal Developer Platform for Kubernetes
• GitOps-based deployment with ArgoCD
• DRY (Don't Repeat Yourself) approach
  • Same templates for all environments
  • Different values for each environment

Installation Steps

• Installation of Kubernetes cluster using IaC tools like Terraform or onprem tools
• Installation and Configuration of Kubernetes tools and microservices using GitOps and ArgoCD or by tools like Kuberise.io