Kubernetes Cheatsheet

Show the installed kubectl client version without querying the cluster

Display the cluster control plane and core service endpoints

💡 Use this first when validating access to a newly configured cluster context.

List all nodes and their readiness status in the current cluster

List nodes with extra details like internal IPs, OS image, and container runtime

Show detailed capacity, conditions, labels, taints, and pod allocation for a node

💡 Check the Conditions and Allocated resources sections when investigating scheduling issues.

View live CPU and memory usage for nodes from Metrics Server

Print the merged kubeconfig with clusters, users, and contexts

Show the currently selected kubeconfig context

List all available kubeconfig contexts and highlight the active one

💡 Name contexts consistently by environment and region so switching is less error-prone.

Switch kubectl to a different cluster and user context

List pods in the current namespace with basic readiness and status info

List pods across every namespace in the cluster

List pods only in a specific namespace

Show pods with extra details like node placement and pod IPs

Inspect pod metadata, container states, events, probes, and mounts

💡 Describe is often faster than logs for spotting image pull, probe, and scheduling failures.

Fetch logs from the main container in a pod

Stream logs from a pod in real time

View logs from the previously terminated container instance in a restarting pod

💡 Essential for CrashLoopBackOff debugging because the current container may restart before you inspect it.

Open an interactive shell inside a pod using the default container

Open a shell inside a specific container in a multi-container pod

Delete a pod and let its controller recreate it if managed

💡 Deleting a pod is a safe way to force a restart when it belongs to a Deployment or StatefulSet.

Create a quick standalone pod directly from an image

Forward a local port to a pod port for direct debugging access

List deployments in the current namespace with desired and ready replica counts

Inspect a deployment strategy, pod template, rollout conditions, and events

Create a deployment from a single image directly from the CLI

Change the number of desired pod replicas for a deployment

Update a container image and trigger a rolling rollout

💡 Pair this with rollout status to confirm the new version becomes healthy before continuing.

Watch the rollout until the deployment successfully completes or fails

Show previous rollout revisions for a deployment

Roll back the deployment to the previous revision

Roll back the deployment to a specific stored revision

💡 Use rollout history first so you can target the exact known-good revision.

Delete a deployment and its managed ReplicaSets and pods

Create or update a deployment declaratively from a manifest file

💡 Use apply for GitOps-style workflows so the manifest stays the source of truth.

Open the live deployment manifest in your editor for an immediate change

List services in the current namespace with type and cluster IP

List services across all namespaces

Inspect service ports, selectors, endpoints, and load balancer details

Expose a deployment internally within the cluster using a ClusterIP service

Expose a deployment externally through a cloud load balancer

💡 Cloud environments may take time to provision an external IP; watch the EXTERNAL-IP field.

Expose a deployment on a static port on each cluster node

List ingress resources and their configured hosts and addresses

Inspect ingress rules, TLS configuration, and backend service routing

Show which pod IPs are currently backing each service

💡 Empty endpoints usually mean your service selector does not match any healthy pods.

Forward a local port to a service for testing without external exposure

Delete a service resource by name

List all namespaces in the cluster

Create a new namespace for an application, team, or environment

Delete a namespace and everything contained inside it

💡 Use with care because namespace deletion removes all contained workloads, services, and secrets.

List the main workload and service resources in a target namespace

Set the default namespace for the current context to reduce repeated -n flags

💡 This is a huge quality-of-life improvement for daily operations in a single namespace.

List pods across every namespace using the long-form namespace flag

View resource quota usage and hard limits for a namespace

Inspect a namespace including labels, annotations, quotas, and limit ranges

List ConfigMaps in the current namespace

Show the keys and metadata stored in a ConfigMap

Create a ConfigMap from one or more literal key-value pairs

Create a ConfigMap from a local file such as an env file or config artifact

💡 Use --dry-run=client -o yaml when you want to generate manifests instead of creating resources directly.

List secrets in the current namespace

Inspect secret metadata and available keys without printing decoded values

Create a generic secret from literal key-value pairs

Create an image pull secret for a private container registry

💡 Pair the secret with a service account so workloads automatically inherit private registry access.

Decode a specific secret key value from base64 for troubleshooting

List cluster-wide persistent volumes and their binding state

List persistent volume claims in the current namespace

List persistent volume claims in a specific namespace

Inspect a persistent volume including reclaim policy, capacity, and claim binding

Inspect a claim including requested size, access modes, and mounted workloads

💡 Describe the PVC when a pod is Pending because unbound claims often block scheduling.

Delete a persistent volume claim and release its bound volume according to policy

List available storage classes and identify the default provisioner

Inspect provisioner parameters, reclaim policy, and binding mode for a storage class

List cluster events in chronological order to surface the newest failures last

💡 Events often reveal the root cause faster than controller logs when pods are stuck Pending.

Filter namespace events down to failure-related entries only

Extract the events section from a pod description for quick CLI triage

Show logs generated within the last hour to narrow noisy output

Show only the most recent log lines from a pod

View CPU and memory usage for pods in the current namespace

View CPU and memory usage for pods in a specific namespace

💡 Compare usage against requests and limits to spot pods being throttled or under-provisioned.

List only pods that are still waiting to schedule or start

List pods whose last phase is Failed for cleanup or investigation

Safely evict workloads from a node before maintenance or termination

💡 Add --delete-emptydir-data only when you understand the impact on ephemeral workload data.

Mark a node unschedulable so new pods are not placed on it

Mark a node schedulable again after maintenance is complete