This page is a structured Kubernetes terms reference. The idea is not to turn every concept into a deep technical chapter, but to give you a practical map of the terms you keep seeing while learning and working with Kubernetes.
To make the page easier to scan, the terms are grouped into beginner, intermediate, and advanced levels, and the quick index below lets you jump directly to the term you want.
Quick Index
This glossary includes 55 terms. Use the level links above for broad navigation, or jump directly to a specific concept from the index cards below.
Beginner
Core concepts you should understand before touching real workloads.
Intermediate
Concepts that matter once you start operating real applications.
Advanced
Control-plane, networking, security, and operational concepts.
Beginner Kubernetes Concepts
Core concepts you should understand before touching real workloads.
1.Docker and Kubernetes
Docker is usually used to build images and run containers, while Kubernetes is used to manage them at scale. In short, Docker is closer to packaging and runtime, while Kubernetes is closer to operations and orchestration.
2.Cluster
A cluster is the group of machines that work together under Kubernetes. You can think of it like coordinating the same system across multiple branches instead of a single office.
3.Node
A node is a single machine inside the cluster. It can be a physical server or a virtual machine.
4.Pod
A Pod is the smallest deployable unit in Kubernetes. Most of the time one application container runs inside one Pod, but multiple tightly related containers can also share the same Pod.
5.Namespace
A namespace creates logical separation inside the same cluster. It is like giving different teams separate rooms inside the same building.
6.Deployment
A Deployment defines how Pods should be created and how many should keep running. It is the standard way to say, “I want three copies of this service online.”
7.Replicas and Scaling
Replicas describe how many copies of the same application should run at the same time. Scaling is the act of increasing or decreasing that number when traffic changes.
8.Service
A Service gives you a stable access point for Pods. Pods can change over time, but the Service name stays the same, which makes application-to-application communication easier.
9.Load Balancing
Load balancing means distributing incoming requests across multiple Pods. In real life, it is like sending customers to several cashiers instead of only one.
10.kubectl
kubectl is the command-line tool used to manage a Kubernetes cluster. You can think of it as the remote control that talks to the cluster.
11.YAML Manifest
A YAML manifest is the file where you describe what Kubernetes should run and how it should behave. Instead of typing one-off commands, you declare the desired state in a repeatable file.
12.ConfigMap
A ConfigMap is used to store non-sensitive configuration outside application code. It helps you avoid hardcoding settings directly into the app.
13.Secret
A Secret is used for sensitive data such as passwords, tokens, and API keys. It is conceptually similar to a ConfigMap, but intended for confidential values.
Intermediate Kubernetes Concepts
Concepts that matter once you start operating real applications.
14.Rolling Update
A rolling update moves an application to a new version gradually instead of shutting everything down at once. That reduces visible downtime for users.
15.Rollback
A rollback means returning to the previous working version when a new release causes problems. It is the practical way to recover from a broken deployment quickly.
16.Ingress
Ingress routes incoming HTTP and HTTPS traffic from outside the cluster to the correct internal Services. It acts like a reception desk that decides where visitors should go.
17.Resource Requests and Limits
Requests and limits define how much CPU and memory a Pod can ask for or consume. This helps prevent one application from taking over the whole machine.
18.Scheduler
The scheduler decides which Pod should run on which node. In other words, it is the planner that spreads workloads across available machines.
19.Liveness Probe
A liveness probe checks whether an application is still healthy or whether it is stuck. If the app is no longer responding correctly, Kubernetes can restart it.
20.Readiness Probe
A readiness probe checks whether an application is ready to receive traffic. An app may be running, but still not ready because dependencies such as the database are not fully available yet.
21.Persistent Volume (PV)
A Persistent Volume is durable storage for data that should survive Pod deletion. It is used when losing the data together with the Pod would be unacceptable.
22.Persistent Volume Claim (PVC)
A Persistent Volume Claim is the application’s storage request. It is how the workload says, “I need this much persistent disk space.”
23.StatefulSet
A StatefulSet is used for workloads that need stable identity. It is especially important for databases and similar systems where each instance should stay distinguishable.
24.DaemonSet
A DaemonSet is used when you want a Pod to run on every node in the cluster. A typical example is a log collector or monitoring agent.
25.Job
A Job is for work that should run once and then finish. Examples include a report generator or a one-time data migration task.
26.CronJob
A CronJob is a Job that runs on a schedule. A common use case is taking backups every night at a fixed hour.
27.Multi-Container Pod
A multi-container Pod runs more than one container inside the same Pod. Usually one container is the main application and another one supports it.
28.Sidecar Pattern
The sidecar pattern means adding a supporting container next to the main application container. Typical examples are log shippers, proxies, or sync helpers.
29.Service Discovery
Service discovery means Services inside the cluster can find one another by name. Even if IP addresses change, the stable Service name keeps communication simple.
30.Horizontal Pod Autoscaler (HPA)
An HPA automatically increases or decreases the number of Pods based on traffic or resource usage. It is similar to opening extra cashier lines during busy hours.
31.Init Container
An init container runs before the main application starts. It is useful for preparation steps such as downloading config files or setting up required directories.
Advanced Kubernetes Concepts
Control-plane, networking, security, and operational concepts.
32.Control Plane
The control plane is the brain of the cluster. Components such as the API server, scheduler, and controller manager live there and handle overall system management.
33.API Server
The API server is the front door of Kubernetes. Most commands and requests reach the cluster through it, including requests sent by kubectl.
34.etcd
etcd is the main data store for cluster state. Information such as which Pod is running where and which Deployments exist is stored there.
35.Controller
A controller keeps the current state of the system aligned with the desired state. If you say, “there should be three Pods,” a controller works to make sure that becomes true.
36.Controller Manager
The controller manager runs and coordinates multiple controllers. It plays a central role in keeping the cluster in order.
37.kubelet
kubelet runs on every node and is responsible for making sure Pods actually run there. It acts like the local executor between the node and the control plane.
38.Container Runtime
The container runtime is the engine that actually runs containers. Tools such as containerd and CRI-O belong to this layer.
39.CNI
CNI is the networking layer that gives Pods connectivity. Without it, applications inside the cluster would not be able to talk to each other properly.
40.NetworkPolicy
A NetworkPolicy restricts which Pods are allowed to talk to which other Pods. It is how you stop every service from being open to everything else by default.
41.RBAC
RBAC is the authorization model that defines who can do what inside the cluster. It is how you avoid giving every developer full admin rights.
42.Taints and Tolerations
Taints and tolerations let you protect special-purpose nodes and allow only suitable workloads to land there. They are essential when not every node should accept every Pod.
43.Affinity and Anti-Affinity
Affinity and anti-affinity define which Pods should be placed close together or far apart. For example, you may want to avoid placing all copies of the same app on one machine.
44.Pod Disruption Budget
A Pod Disruption Budget prevents too many Pods from disappearing at the same time during maintenance or upgrades. It helps keep the system available while changes happen.
45.Helm
Helm makes Kubernetes applications easier to package, configure, and reuse. You can think of it as a package manager for Kubernetes applications.
46.Operator Pattern
The operator pattern brings complex operational knowledge into Kubernetes itself. It is often used to automate expert tasks such as database setup, backups, and recovery.
47.Custom Resource Definition (CRD)
A CRD allows you to add your own resource types to Kubernetes. If the built-in objects are not enough, you can define your own custom ones.
48.Observability
Observability is the broader practice of understanding system behavior through logs, metrics, and tracing. It helps answer questions such as why an app is slow or where a request is getting stuck.
49.Logging and Monitoring
Logging collects event records, while monitoring tracks health and performance over time. One tells you what happened; the other tells you how the system is doing.
50.Service Mesh
A service mesh adds a layer that makes service-to-service communication more controlled and observable. It becomes especially useful for security, traffic management, and tracing.
51.GitOps
GitOps is the approach of treating Git as the source of truth for cluster configuration. In that model, the answer to “what should run in the cluster?” lives in a repository.
52.Blue/Green and Canary Deployment
These are controlled rollout strategies that avoid exposing a new version to everyone at once. They reduce risk and make it easier to catch problems early.
53.Security Hardening
Security hardening is the practice of tightening the security posture of the cluster and the workloads. This includes reducing permissions, scanning images, and handling secrets carefully.
54.Multi-Cluster
Multi-cluster means managing more than one Kubernetes cluster together. It is common when working across regions, environments, or stricter availability requirements.
55.Disaster Recovery
Disaster recovery is the plan for restoring the system after cluster or data loss. Backups, restore procedures, and recovery paths for critical components all belong here.