Kubernetes Operations

Smallest deployable unit in Kubernetes; one or more containers sharing the same network namespace and storage volumes.

"Each pod gets its own IP address — the sidecar container in the pod can reach the main app container on localhost."

Workload resource that manages a set of identical replica pods; handles rolling updates and rollbacks declaratively.

"We updated the image tag in the Deployment manifest and kubectl apply'd it — Kubernetes rolled out the new version one pod at a time with no downtime."

Stable network endpoint that load-balances traffic across pods matching a label selector; provides a fixed DNS name regardless of pod IP changes.

"The Service named api-svc always routes to healthy backend pods — even as pods are replaced during a rolling update, the DNS name stays constant."

Kubernetes resource that routes external HTTP/S traffic to services based on hostname and path rules; requires an Ingress controller such as nginx or Traefik.

"We added an Ingress rule to route /api/* to the backend service and /* to the frontend service, both under the same domain."

Virtual cluster within a Kubernetes cluster for isolating resources between teams or environments; RBAC and resource quotas can be scoped per namespace.

"We deploy staging and production in separate namespaces on the same cluster, with resource quotas preventing either from starving the other."

API object that stores non-sensitive configuration data as key-value pairs; mounted into pods as environment variables or files.

"The app reads its database URL from a ConfigMap volume mount — changing the config no longer requires rebuilding the Docker image."

API object that stores sensitive data such as passwords and tokens; base64-encoded by default and can be encrypted at rest with KMS.

"The database password is stored in a Kubernetes Secret and injected as an environment variable — it never appears in the container image or ConfigMap."

A request for storage by a pod; abstracts the underlying storage implementation (EBS, NFS, local disk) from the application.

"The database pod uses a PersistentVolumeClaim so data survives pod restarts and rescheduling to a different node."

Workload resource for stateful applications; provides stable network identity, ordered pod management, and persistent storage per pod.

"We deploy Kafka with a StatefulSet so each broker gets a stable hostname like kafka-0, kafka-1 and retains its own PersistentVolumeClaim."

Ensures exactly one copy of a pod runs on every node (or a subset); used for log collectors, monitoring agents, and network plugins.

"The log-shipping DaemonSet ensures Fluent Bit runs on every node and ships container logs to our centralized logging platform."

Horizontal Pod Autoscaler; automatically scales the number of pod replicas based on CPU utilisation, memory, or custom metrics.

"The HPA scaled the API deployment from 3 to 12 replicas during the Black Friday traffic spike, then scaled back down overnight."

Minimum CPU and memory guaranteed to a container by the scheduler; used to decide which node a pod is placed on.

"Setting a CPU request of 250m ensures our pod is only scheduled on nodes with at least that capacity available."

Maximum CPU and memory a container is allowed to use; a container exceeding its memory limit is OOMKilled by the kernel.

"We set a memory limit of 512Mi — when a memory leak caused the app to exceed it, Kubernetes restarted the container automatically."

Periodic check that restarts a container if it fails; indicates whether the application is alive and not deadlocked.

"The liveness probe hits /healthz every 10 seconds — when the app deadlocked and stopped responding, Kubernetes restarted the container within 30 seconds."

Periodic check that removes a pod from the Service's endpoint list if it fails; indicates whether the app is ready to accept traffic.

"The readiness probe prevents traffic from reaching a pod during startup warming — the pod only joins the load balancer pool once the probe passes."