Differences

This shows you the differences between two versions of the page.

--- k8s:core:architecture [2026/05/31 02:44] – created phong2018
+++ k8s:core:architecture [2026/05/31 02:57] (current) – phong2018
@@ Line 1: / Line 1: @@
 ====== Cluster Architecture ======
-Kubernetes cluster architecture is based on a **distributed system design**.
+Kubernetes cluster architecture is the overall design of how Kubernetes is built.
-A cluster is divided into two main parts:
+A cluster is split into two main parts:
-  * Control Plane (manages the cluster)
+  * Control Plane (brain)
-  * Worker Nodes (run applications)
+  * Worker Nodes (workers)
-This separation allows Kubernetes to:
+---
-  * Scale workloads easily
-  * Recover from failures automatically
-  * Manage applications in a declarative way
-A typical flow:
+===== Simple flow =====
-  * User sends request (kubectl / API)
-  * Control Plane processes it
-  * Worker Nodes execute workloads (Pods)
-The architecture ensures:
-  * High availability
+User request
-  * Fault tolerance
+↓
-  * Automated workload scheduling
+API Server (Control Plane)
+↓
+Scheduler decides node
+↓
+Worker Node runs Pod
+---
+===== Real-world example =====
+You deploy a web app:
+  * You want 3 replicas of your app
+  * Kubernetes distributes them across multiple nodes
+  * If a node fails → pods move to another node
+---
+===== Key idea =====
+Cluster = a group of machines working together:
+  * Some decide (Control Plane)
+  * Some execute (Worker Nodes)