Core Objects

Understanding Kubernetes object model, namespaces, pods, labels, and selectors

Overview

Core objects are the fundamental building blocks of Kubernetes. This section covers the essential objects you'll work with every day: the Kubernetes object model, namespaces, pods, labels, and selectors.

Study Path

1. Object Model - How Kubernetes objects work
2. Namespaces - Resource isolation and organization
3. Pods - The smallest deployable unit
4. Labels & Selectors - Identifying and grouping objects

Key Concepts Overview

Concept	Purpose	Docker Equivalent
Object Model	Declarative configuration	N/A
Namespace	Resource isolation	Docker context
Pod	Smallest deployable unit	docker run
Labels	Identify objects	Docker labels
Selectors	Group and select objects	N/A

Relationship Between Objects

┌─────────────────────────────────────────────────────────────────┐
│                      Namespace (ml-apps)                        │
│                                                               │
│  ┌───────────────────────────────────────────────────────────┐ │
│  │                  Deployment (model-api)                   │ │
│  │  Labels: app=ml-model, version=v2                        │ │
│  │  ┌─────────────┐  ┌─────────────┐  ┌─────────────┐      │ │
│  │  │    Pod     │  │    Pod     │  │    Pod     │      │ │
│  │  │  replica 1 │  │  replica 2 │  │  replica 3 │      │ │
│  │  │ Labels:    │  │ Labels:    │  │ Labels:    │      │ │
│  │  │ app=model  │  │ app=model  │  │ app=model  │      │ │
│  │  │ ┌─────────┐│  │ ┌─────────┐│  │ ┌─────────┐│      │ │
│  │  │ │Container││  │ │Container││  │ │Container││      │ │
│  │  │ │: ML     ││  │ │: ML     ││  │ │: ML     ││      │ │
│  │  │ │  Model  ││  │ │  Model  ││  │ │  Model  ││      │ │
│  │  │ └─────────┘│  │ └─────────┘│  │ └─────────┘│      │ │
│  │  └─────────────┘  └─────────────┘  └─────────────┘      │ │
│  └───────────────────────────────────────────────────────────┘ │
│                                                               │
│  ┌───────────────────────────────────────────────────────────┐ │
│  │               Service (model-api-svc)                     │ │
│  │  Selector: app=ml-model                                   │ │
│  │  Load Balancer → Routes traffic to all 3 pods            │ │
│  └───────────────────────────────────────────────────────────┘ │
└─────────────────────────────────────────────────────────────────┘

Quick Reference

Common Commands

# Namespaces
kubectl get namespaces
kubectl create namespace <name>
kubectl get pods -n <namespace>

# Pods
kubectl get pods
kubectl describe pod <name>
kubectl logs <pod-name>
kubectl exec -it <pod-name> -- bash

# Labels & Selectors
kubectl get pods -l app=ml-model
kubectl get pods --show-labels
kubectl label pod <name> key=value

# All resources in namespace
kubectl get all -n <namespace>

Imperative vs Declarative

Imperative (Do this):

kubectl create deployment nginx --image=nginx
kubectl scale deployment nginx --replicas=3

Declarative (Make it so):

kubectl apply -f deployment.yaml
# Kubernetes makes actual state match desired state

Best Practice: Always use declarative (YAML) for production.

Object YAML Structure

All Kubernetes objects follow this pattern:

apiVersion: apps/v1           # API version
kind: Deployment              # Type of object
metadata:                     # Object metadata
  name: nginx-deployment
  labels:
    app: nginx
spec:                         # Desired state
  replicas: 3
  selector:
    matchLabels:
      app: nginx
  template:                   # Pod template
    metadata:
      labels:
        app: nginx
    spec:
      containers:
      - name: nginx
        image: nginx:1.25
        ports:
        - containerPort: 80

Next Steps

1. Learn Object Model: Object Model
2. Understand Namespaces: Namespaces
3. Study Pods: Pods
4. Master Labels & Selectors: Labels & Selectors
5. Practice: Lab 02: First Deployment

---

Continue Learning:

• Object Model
• Namespaces
• Pods
• Labels & Selectors

Practice: Lab 02: First Deployment

Return to: Overview | K8s for MLOps