Q11: Kubernetes Control Plane Architecture: How do Control Plane components and Worker Node agents interact? - DevOps & Kubernetes (50 Qs)

High-Level Architecture

Kubernetes follows a master-worker distributed architecture. The Control Plane maintains cluster state, evaluates scheduling decisions, and orchestrates reconciliation loops, while Worker Nodes execute the containerized workloads.

scss

+-------------------------------------------------------------------------------+
|                             KUBERNETES CONTROL PLANE                          |
|                                                                               |
|    +-------------+      +-------------------+      +---------------------+    |
|    |    etcd     |<---->|  kube-apiserver   |<---->|    kube-scheduler   |    |
|    | (Raft Store)|      | (Central Gateway) |      | (Node Placement)    |    |
|    +-------------+      +-------------------+      +---------------------+    |
|                                   ^                                           |
|                                   |                +---------------------+    |
|                                   +--------------->| kube-controller-mgr |    |
|                                                    | (Reconcile Loops)   |    |
|                                                    +---------------------+    |
+-----------------------------------|-------------------------------------------+
                                    | HTTPS (Port 6443)
      +-----------------------------+-----------------------------+
      |                                                           |
      v                                                           v
+-------------------------------+           +-------------------------------+
|          WORKER NODE 1        |           |          WORKER NODE 2        |
|  +-------------------------+  |           |  +-------------------------+  |
|  |        kubelet          |  |           |  |        kubelet          |  |
|  | (Node Agent & SyncLoop) |  |           |  | (Node Agent & SyncLoop) |  |
|  +-------------------------+  |           |  +-------------------------+  |
|  |       kube-proxy        |  |           |  |       kube-proxy        |  |
|  | (iptables/IPVS Rules)   |  |           |  | (iptables/IPVS Rules)   |  |
|  +-------------------------+  |           |  +-------------------------+  |
|  | Container Runtime (CRI) |  |           |  | Container Runtime (CRI) |  |
|  | (containerd / CRI-O)    |  |           |  | (containerd / CRI-O)    |  |
|  +-------------------------+  |           |  +-------------------------+  |
|  |   [Pod A]     [Pod B]   |  |           |  |   [Pod C]     [Pod D]   |  |
+-------------------------------+           +-------------------------------+

Component Roles & Responsibilities

Component	Location	Role in Cluster
`kube-apiserver`	Control Plane	The only component that reads/writes directly to `etcd`. Exposes the REST API and handles AuthN, AuthZ, and Admission.
`etcd`	Control Plane	Consistent, highly-available key-value store holding the complete declarative cluster state using the Raft consensus algorithm.
`kube-scheduler`	Control Plane	Watches for newly created unscheduled Pods (`nodeName == ""`) and assigns them to optimal nodes based on resource filtering and scoring.
`kube-controller-manager`	Control Plane	Runs core reconciliation loops (DeploymentController, ReplicaSetController, NodeController, EndpointsController).
`kubelet`	Worker Node	Primary node agent. Watches API Server for Pods assigned to its node and instructs CRI to create/start/stop containers.
`kube-proxy`	Worker Node	Maintains network routing rules (iptables or IPVS) on each node to implement Service abstraction and load balancing.
Container Runtime	Worker Node	Software executing the containers (containerd, CRI-O) implementing the Kubernetes Container Runtime Interface (CRI).