Understanding Kubernetes: Cloud‑Native Architecture, Deployment Scenarios, and Operational Practices

Kubernetes, abbreviated as K8s, is an open‑source platform for managing containerized applications across multiple hosts in a cloud environment, providing mechanisms for deployment, planning, updating, and maintenance.

k8s in enterprise application scenarios

Key characteristics of Kubernetes:

Portability: supports public, private, hybrid, and multi‑cloud environments.

Extensibility: modular, plug‑in capable, mountable, and composable.

Automation: automatic deployment, restart, replication, and scaling.

Automated operations platform

For small‑ and medium‑size enterprises, K8s can build an automated operations platform that streamlines application deployment, planning, updating, and maintenance. Large internet companies use it to manage massive server fleets, enabling automated monitoring, deployment, and operation of services.

Maximizing server resource utilization

Example: 200 concurrent requests on a server with 2 CPU and 4 GB RAM. Assuming static requests consume 2 MB each and dynamic requests 10 MB each, the total memory usage is 800 MB, supporting a theoretical QPS of 800. In practice, factors such as response time and CPU context switching reduce achievable QPS, often to around 300, indicating under‑utilization of resources.

Seamless service migration

Container images encapsulate applications, decoupling them from the underlying OS and allowing consistent environments from development through testing to production. Kubernetes manages these containers, enabling seamless migration across environments.

Evolution of service deployment models

Traditional physical machines bind applications to the host OS, making upgrades and rollbacks difficult. Virtual machines improve isolation but are resource‑heavy. Container deployment isolates processes, provides lightweight, fast deployment, and easy migration across clouds and OS versions.

Challenges in containerized deployments include horizontal scaling, container failure recovery, zero‑downtime updates, monitoring, scheduling, and data security—all of which can be addressed by Kubernetes.

Cloud architecture & cloud‑native concepts

Cloud is a collection of container‑based service clusters; Kubernetes manages those containers. Cloud service models (IaaS, PaaS, SaaS, Serverless) are described, highlighting how users consume resources without managing underlying infrastructure.

Cloud‑native principles

Containerization: all services run in containers.

Microservices: applications are split into independent services.

CI/CD: continuous integration and delivery.

DevOps: tight integration of development and operations.

Kubernetes architecture overview

Kubernetes originated from Google’s internal Borg system and is written in Go. It manages clusters composed of a master node and multiple worker nodes.

Master node components

API server – the gateway for all commands.

Scheduler – assigns workloads to nodes based on algorithms.

Controller manager – maintains the desired state of resources (CRUD).

etcd – stores resource objects and supports service discovery.

Worker node components

Docker – container runtime.

Kubelet – runs on each node, watches etcd for instructions and manages pods.

Kube‑proxy – provides load balancing across services.

Fluentd – collects logs.

Pod – the smallest deployable unit; may contain one or more containers.

Key architectural takeaways

Kubernetes manages containers indirectly via pods.

One master controls many nodes; the master does not store containers.

Pods reside on worker nodes and can host multiple containers.

Kubelet handles pod lifecycle; kube‑proxy handles load balancing.