Kubernetes
Open SourceProduction-Grade Container Orchestration.
Published 29 May 2026 · Last updated 27 September 2026
Scores
Popularity4/5
The standard container orchestration platform; required knowledge for most DevOps roles.
Learning Curve5/5
Pods, deployments, services, RBAC, and operators are a vast surface requiring significant experience.
Flexibility5/5
CRDs, operators, and the plugin ecosystem provide virtually unlimited extensibility.
Performance5/5
Handles millions of requests per second at scale; cluster overhead is a one-time cost.
Portability5/5
Open standard; cluster workloads run on any cloud or on-prem Kubernetes environment.
About Kubernetes
Kubernetes (K8s) is an open-source container orchestration system, originally designed at Google and now maintained by the Cloud Native Computing Foundation (CNCF). It runs containerized workloads across a cluster of machines and continuously works to keep the cluster in the declared state: if a container crashes or a node disappears, Kubernetes reschedules the work elsewhere.
The basic unit is the Pod, one or more containers that share networking and storage and are scheduled together. Deployments manage replicated Pods and roll out new versions gradually, rolling back when health checks fail; StatefulSets, DaemonSets, and Jobs cover databases, per-node agents, and batch work. Services give Pods stable DNS names and load-balanced addresses, while Ingress and the newer Gateway API route external traffic. ConfigMaps and Secrets inject configuration without baking it into images, and Persistent Volumes decouple storage from a Pod's lifecycle.
Scaling is built in: the Horizontal Pod Autoscaler adjusts replica counts from CPU, memory, or custom metrics, and cluster autoscalers add and remove nodes. Custom Resource Definitions and operators extend the API so that databases, message queues, and certificates can be managed as native objects, which is the base of a large ecosystem of Helm charts and GitOps tools such as Argo CD and Flux.
Kubernetes is free software. Most teams run it as a managed service (GKE, EKS, AKS, or offerings from DigitalOcean, Hetzner, and others) that handles the control plane, or self-host it with kubeadm or lighter distributions such as k3s. The project ships about three minor releases a year and supports each for roughly a year.
Key Features
- Automated rollouts and rollbacks
- Service discovery, load balancing, and the Gateway API
- Self-healing: restarts and reschedules failed workloads
- Horizontal Pod Autoscaler and cluster autoscaling
- Secret and configuration management
- Persistent Volumes and storage classes
- Jobs and CronJobs for batch work
- Custom Resource Definitions and operators
Pros
- Industry standard for container orchestration
- Runs on every major cloud, on-premises, and at the edge
- Declarative configuration fits GitOps workflows
- Operators and Helm charts package complex software as native objects
- Managed services remove most control-plane work
Cons
- Steep learning curve
- Complex to set up and operate when self-hosted
- Resource overhead is high for small deployments
- Overkill for a single simple application
- Frequent minor releases mean regular cluster upgrades
Kubernetes Pricing
Open SourceTech Stacks with Kubernetes
Advanced API (Go)
ProjectA high-performance API stack for advanced engineers. Go handles concurrency, PostgreSQL + Redis back the data layer, Prometheus + Grafana provide observability, and Kubernetes can orchestrate containers as an optional addition.
MLOps Pipeline
ProjectProduction-grade ML infrastructure. PyTorch for model training, Apache Airflow (or Dagster or Prefect) for orchestration, dbt for feature transformations, and Snowflake as the data warehouse, with Docker as an optional containerization addition.
Backend
Programming
Databases
Hosting
Kubernetes DevOps Pipeline
InfrastructureA self-managed DevOps platform for teams that want the whole path from commit to production in one place. Bring your own containerized app; GitHub Actions or GitLab CI/CD tests and builds it, Kubernetes deploys and runs it, PostgreSQL is available for its data, and Grafana and Sentry close the observability loop.
Tools Related to Kubernetes
Works well with Kubernetes(6)
GKE (Google Kubernetes Engine) is Google's managed Kubernetes service — Kubernetes was created at Google and GCP is its native cloud; GKE is the reference managed Kubernetes implementation.
Azure AKS is the managed Kubernetes service on Azure — Azure integrates Kubernetes with Azure AD, Azure Monitor, and Container Registry out of the box.
AWS EKS is the managed Kubernetes service on AWS — Kubernetes clusters on AWS consume EC2, ELB, EBS, and IAM; the AWS cloud provider is built into Kubernetes core.
Kubernetes hosts self-hosted GitHub Actions runners via the Actions Runner Controller (ARC), and is a frequent deployment target for applications built and tested in GitHub Actions pipelines.
Kubernetes hosts GitLab CI/CD runners via the native Kubernetes executor (jobs run as pods) and is a primary deployment target reached through the GitLab Kubernetes agent (agentk) for GitOps workflows.
Grafana visualises Prometheus and other Kubernetes metrics — Grafana dashboards for cluster health (CPU, memory, pod status) are a standard part of any Kubernetes monitoring setup.
Integrates with Kubernetes(6)
Spring Cloud Kubernetes provides native Kubernetes ConfigMap, health checks, and service discovery integration.
GitLab has native Kubernetes integration via Auto DevOps — it manages cluster connections, deployments, and monitoring directly from merge request workflows without external CI tools.
Docker Desktop bundles a local Kubernetes cluster, and Docker is a common way to build the OCI container images Kubernetes schedules — but Kubernetes dropped built-in Docker Engine (dockershim) support in v1.24 and runs on any CRI-compliant runtime (containerd, CRI-O), so it no longer requires Docker specifically.
Datadog is the leading APM and logging platform for Kubernetes — the Datadog Agent DaemonSet instruments all pods automatically; traces, logs, and metrics are correlated in Datadog dashboards.
Prometheus is the default Kubernetes metrics system — the kube-state-metrics and node-exporter components expose cluster health metrics; the Prometheus Operator CRD is the standard Kubernetes install method.
OpenSandbox ships a Kubernetes operator and Helm charts: in production each sandbox runs as a pod managed through custom resources, with resource pools for fast batch creation.