Red Hat Immersion 2023 Service Offerings and Labs - Deployment and App Management

Getting Started with OpenShift for Developers

Container Technology

• Easy to scale

• Lightweight

• Mature, suitable for cloud

• Container is the smallest compute unit

  • Package app and dependencies as unit
  • Created from base images
    • Layers: app, runtime (e.g. Java), OS
    • Image registry, versions
  • Immutable

Why Kubernetes (k8s)?

• History: application divided into microservices
  • Distributed deployment in network of services
  • Each microservice have their own data
  • Advantages:
    • Easier to scale specific services
    • Availability of services decoupled
• Open source software to manage container workloads across cluster of nodes like scale and integrate them

Understanding k8s Objects

• k8s has API
• Node: host machine of containers
  • Worker nodes
  • Master nodes - manage node activity
• Pod: group of 1 or more co-located containers
  • Smallest unit of scale
• Service: single endpoint for collection of replicated pods, like a load balancer
• Deployment: specific container runtime in terms of pods like name of app, image, replica
• Cluster: set of nodes
  • Master: api, etcd: persistent state of cluster, scheduler, controllers
    • etcd should be backed up to keep cluster state
    • Manage worker nodes like their locations
  • Nodes (kubelet)

Tools

• cri-o RHEL Container Runtime
  • Secure architecture
  • Optimized for k8s
  • Runs any OCI compliance image including docker
• Podman: manager

OpenShift Monitoring

• OpenShift has cluster monitoring
  • Metrics and their storage via Prometheus
  • Alerts via Prometheus Alertmanager
  • Metrics visualization via Grafana
• Observability
  • Tools: EFK:
    • Elasticsearch (ES) - search and analytics
    • Fluentd - logs and sends to ES
    • Kibana - web UI for ES
  • Users: Cluster admin can see logs and users can see logs for their projects
  • Logs can be forwarded like to Splunk

Dev vs Ops

• Usually OpenShift setup of manager nodes, persistent storage and controlling resources is done by ops. Dev can access projects and manage pods
• OpenShift gives many operators, access from Operator Hub:
  • Console
  • Monitoring
  • Auth
  • DNS
  • k8s control
  • DNS

OpenShift Features

1. Serverless

   • Based on KNative, Allows deployment functions

2. Service Mesh

   • Use case: Increased troubleshooting complexity with many containers. Need to trace requests, external services.
   • Service Mesh handles a common infrastructure layer with common policies and code for security, metrics, tracing, failure handling, traffic management - code as a service
     • Mesh can be configured
     • Function actions: secure, observe, connect, control
       • Tools: Jaegar, Kiali, Prometheus, Grafana

3. Platform Management and Development

   • Console
   • Helm charts to package, install, and update
   • OpenShift CLI
   • IDE plugins - ex. VS Code, IntelliJ
   • OpenShift Local - cluster for developer machines
   • DevSpaces - collaborative k8s web based IDE and developer stacks

Questions

• Master and worker locations?
  • Master and worker nodes can be on same host, only for small use cases, recommended to be separate
• How to manage different environment configurations?
  • Can do by same image and assemble image in DevOps with different configurations and secrets, database, and ConfigMaps

Workshop

See https://github.com/justunsix/openshift-k8s-tests

See Also

• Red Hat Immersion 2023 Service Offerings and Labs