120+ OpenShift Interview Questions and Answers (2027)

122 OpenShift interview questions with answers for freshers, 2-5 years and seniors: oc CLI, Routes, SCCs, BuildConfig, Operators, OLM v1, OpenShift 5.0.

Sep 12, 2025 - 17:51
Nov 7, 2026 - 00:05
 0  559
120+ OpenShift Interview Questions and Answers (2027)

OpenShift interview questions test whether you can run applications on Red Hat OpenShift Container Platform: Kubernetes basics, projects, Routes, builds with BuildConfig and S2I, Security Context Constraints, Operators and OLM, MachineConfig, upgrades, monitoring and troubleshooting with the oc CLI. This page has 122 questions with direct answers, grouped by experience level.

The answers were checked against the Red Hat documentation in November 2026. OpenShift Container Platform 5.0 became generally available on 3 November 2026 and is built on Kubernetes 1.36. Most clusters in production still run 4.x releases, so the answers point out where 4.x and 5.0 differ. For the Kubernetes concepts underneath, see our 12 must-know Kubernetes architecture components.

Key takeaways

  • OpenShift 5.0 (GA 3 November 2026) uses Kubernetes 1.36 with CRI-O, installs RHCOS 10.2 by default and supports updating from 4.x. OpenShift 4.22 became GA on 9 June 2026.
  • From OpenShift 5, Red Hat plans a minor release every four months, each with 12 months of full support and 12 months of maintenance support. An Extended Life Cycle add-on is offered for every third release, starting with 5.2.
  • DeploymentConfig is still deprecated in 5.0. Use Kubernetes Deployments for new work.
  • BuildConfig still supports Docker, Source-to-Image and Custom builds. The Pipeline (Jenkins) build strategy is deprecated in favour of OpenShift Pipelines, which is based on Tekton.
  • OpenShift SDN was removed in 4.17. OVN-Kubernetes is the network plugin.
  • OLM v1 has been GA since 4.18 (ClusterCatalog and ClusterExtension). The older framework is now called OLM (Classic).
  • New clusters since 4.11 give authenticated users the restricted-v2 SCC, which drops all capabilities and sets the seccomp profile to runtime/default.

How to use this guide

Your experienceRead firstThen practise
Fresher or career switcherQuestions 1 to 40Create a project, deploy an app with oc new-app, expose it with a Route
2 to 5 yearsQuestions 41 to 86BuildConfigs, SCCs, quotas, Operators, monitoring, Pipelines and GitOps
Senior or platform engineerQuestions 87 to 108Upgrades, MachineConfig, etcd backup, disconnected installs, multi-cluster
EveryoneQuestions 109 to 122Scenario troubleshooting with oc

OpenShift interview questions for freshers

1. What is Red Hat OpenShift Container Platform?

OpenShift Container Platform is Red Hat's enterprise Kubernetes distribution. It adds an integrated installer and over-the-air updates, a web console, built-in image registry, Routes, builds, an Operator-based platform, monitoring and stricter default security on top of upstream Kubernetes. It runs on Red Hat Enterprise Linux CoreOS (RHCOS) nodes and uses CRI-O as the container runtime.

2. How is OpenShift different from upstream Kubernetes?

Kubernetes is the orchestration engine. OpenShift packages it with opinionated defaults and extra APIs: Projects, Routes, BuildConfigs, ImageStreams, Security Context Constraints, an OAuth server, cluster Operators that manage every platform component, and a supported update path. Pods cannot run as root by default, which surprises people moving from plain Kubernetes.

3. What is the current version of OpenShift?

OpenShift Container Platform 5.0 became generally available on 3 November 2026 and uses Kubernetes 1.36. The last 4.x minor release, 4.22, became GA on 9 June 2026. Always confirm the version a company runs, because the answers to upgrade and feature questions depend on it.

4. What is OKD?

OKD is the community distribution of Kubernetes that OpenShift is built from. It is free to use without a subscription but has no Red Hat support. Many people use OKD or a free developer sandbox to practise before working with OpenShift Container Platform.

5. What are the main components of an OpenShift cluster?

Control plane nodes run the Kubernetes API server, etcd, the controller manager, the scheduler and the OpenShift API server and OAuth server. Compute (worker) nodes run application pods. Infrastructure nodes are optional workers dedicated to the router, registry and monitoring. Cluster Operators run on top and manage each component's configuration and upgrades.

6. What is RHCOS?

Red Hat Enterprise Linux CoreOS is the immutable, container-focused operating system used for OpenShift nodes. The Machine Config Operator manages its configuration, and it is updated as part of cluster updates. RHCOS is required for control plane machines and for compute machines. In 5.0, new installs use RHCOS 10.2, and RHCOS 9.x is deprecated.

7. What is a Project in OpenShift?

A Project is a Kubernetes namespace with extra annotations and access controls. Users with permission can request projects with oc new-project, and administrators can apply a project template that adds default quotas, limit ranges and network policies to every new project.

oc new-project shop --display-name="Shop" --description="Online shop services"
oc project shop
oc get projects

8. What is the oc CLI?

oc is the OpenShift command line client. It includes all kubectl commands and adds OpenShift-specific ones such as oc login, oc new-project, oc new-app, oc start-build, oc expose and the oc adm administration commands.

9. How do you log in to a cluster with oc?

Use oc login with the API server URL, either with a username and password from a configured identity provider or with a token copied from the web console.

oc login https://api.cluster.example.com:6443 -u developer
oc login --token=sha256~xxxx --server=https://api.cluster.example.com:6443
oc whoami
oc whoami --show-console

10. What is the kubeadmin user?

kubeadmin is the temporary cluster administrator created by the installer. Red Hat recommends configuring an identity provider, creating another cluster administrator, and then deleting kubeadmin to improve security.

11. Which identity providers does OpenShift support?

The built-in OAuth server supports htpasswd, Keystone, LDAP, basic authentication, request header, GitHub or GitHub Enterprise, GitLab, Google and OpenID Connect providers. Enterprises usually use LDAP or OpenID Connect so users log in with corporate accounts.

12. What is a Route?

A Route exposes a Service at a hostname through the OpenShift router, which is HAProxy-based and managed by the Ingress Operator. Routes existed before the Kubernetes Ingress API and support TLS termination types, path-based routing and traffic splitting between services.

oc expose svc hello-openshift
oc get routes

13. What are the TLS termination types for Routes?

There are three. Edge terminates TLS at the router and sends plain HTTP to the pod. Passthrough sends encrypted traffic straight to the pod, which terminates TLS itself. Re-encrypt terminates TLS at the router and opens a new TLS connection to the pod. Path-based routing is not available with passthrough because the router cannot read the request.

oc create route edge --service=frontend --cert=tls.crt --key=tls.key \
  --ca-cert=ca.crt --hostname=www.example.com

14. What is the difference between a Route and an Ingress in OpenShift?

Both expose HTTP services. OpenShift supports the Kubernetes Ingress API and converts Ingress objects into Routes, with annotations such as route.openshift.io/termination for TLS behaviour. Routes offer OpenShift-specific features like weighted backends. OpenShift also ships a managed Gateway API implementation; in 5.0 the Gateway API CRDs are at version 1.5.1.

15. What is an ImageStream?

An ImageStream is an OpenShift object that tracks container images by tag, whether they live in the internal registry or an external one. Builds and deployments can watch an ImageStreamTag and react when it moves to a new image. You manage tags with oc tag and import external images with oc import-image.

oc import-image myapp:1.4 --from=quay.io/example/myapp:1.4 --confirm
oc tag myapp:test myapp:prod

16. What is the internal image registry?

OpenShift includes an image registry managed by the Image Registry Operator. Builds push images to it, and pods in the cluster pull from it using service account credentials. It needs persistent storage in production, and it can be exposed externally through a route if teams push images from outside.

17. What is Source-to-Image (S2I)?

S2I builds a runnable image from application source code by injecting the code into a builder image (for example a Node.js or Java builder) and running its assemble script. Developers do not need to write a Dockerfile. S2I is one of the BuildConfig build strategies.

18. How do you deploy an application from source with oc new-app?

oc new-app detects the language, or you name the builder image with the builder~repo syntax. It creates a BuildConfig, an ImageStream, a Deployment and a Service.

oc new-app nodejs~https://github.com/sclorg/nodejs-ex.git --name=web
oc logs -f bc/web
oc expose svc/web

19. What is a BuildConfig?

A BuildConfig defines how to build an image: the source (Git, Dockerfile, binary or image), the strategy, the output image and the triggers. Each run creates a Build object and a build pod.

20. Which build strategies does BuildConfig support?

Docker builds (from a Dockerfile), Source-to-Image builds and Custom builds are the three primary strategies, and Docker and S2I are enabled by default. Custom builds run with high privilege and are not available to normal users. The Pipeline build strategy, which ran Jenkins pipelines, is deprecated; Red Hat points users to OpenShift Pipelines, based on Tekton.

21. How do you start a build manually?

oc start-build web --follow
oc start-build web --from-dir=. --follow   # binary build from local files
oc get builds

A binary build streams local content to the builder. It cannot be started by triggers or from the web console, because the content has to come from your machine.

22. What build triggers are available?

Webhook triggers (GitHub, GitLab, Bitbucket and Generic), image change triggers that rebuild when the builder image updates, and configuration change triggers that run a build when the BuildConfig is created. Webhooks only act on push events and check that the branch matches the BuildConfig.

23. What is a DeploymentConfig, and should you still use it?

DeploymentConfig is the older OpenShift-specific workload object with lifecycle hooks and image change triggers. It has been deprecated for several releases and is still deprecated in 5.0. Use Kubernetes Deployments for new applications and plan to migrate existing DeploymentConfigs.

24. How do you scale an application?

oc scale deployment/web --replicas=4
oc autoscale deployment/web --min=2 --max=8 --cpu-percent=70

The second command creates a HorizontalPodAutoscaler. The pods must have CPU requests for percentage-based scaling to work.

25. How do you view logs and open a shell in a pod?

oc logs deployment/web
oc logs -f pod/web-6d9f7 -c web
oc rsh deployment/web
oc exec web-6d9f7 -- env

26. How do you set environment variables and mount config?

oc create configmap web-config --from-literal=LOG_LEVEL=info
oc set env deployment/web --from=configmap/web-config
oc create secret generic db --from-literal=password=change-me
oc set env deployment/web --from=secret/db

Changing environment variables triggers a new rollout of the Deployment.

27. What is a Template?

A Template is a parameterised list of objects. You process it with values to produce manifests. Templates are still supported, but many teams now use Helm charts or Kustomize instead.

oc process -f app-template.yaml -p NAME=web -p REPLICAS=2 | oc apply -f -

28. Does OpenShift support Helm?

Yes. The web console can install Helm charts from repositories, and Red Hat ships a supported Helm CLI. In 5.0 the console uses the Helm v4 SDK, the Helm v3 CLI is deprecated, and the Helm v3 CLI is not available from 5.1.

29. What is the OpenShift web console?

The web console is the browser UI for developers and administrators. It shows topology views of applications, logs, events, metrics, Operator installation through the software catalog, and cluster settings such as updates.

30. What are Operators?

An Operator is a controller that manages an application using custom resources, encoding the operational knowledge a human administrator would apply: install, upgrade, backup, scaling and recovery. OpenShift itself is run by cluster Operators, and add-on software such as databases or service meshes is installed through Operators.

31. What are cluster Operators?

Cluster Operators manage OpenShift's own components: the API server, authentication, ingress, image registry, monitoring, networking, the Machine Config Operator and others. Their health is the first thing to check when a cluster misbehaves.

oc get clusteroperators
oc describe clusteroperator authentication

32. What is the Cluster Version Operator?

The Cluster Version Operator (CVO) manages cluster updates. It reads the target release image, applies the manifests for every cluster Operator in order, and reports progress in the ClusterVersion object.

oc get clusterversion
oc adm upgrade

33. What are Security Context Constraints?

SCCs are OpenShift admission controls that decide what a pod may do: run as a specific user ID, use host networking or host paths, add Linux capabilities, run privileged, or use certain volume types. They predate Kubernetes Pod Security Admission, and OpenShift uses both.

34. Why do many Docker Hub images fail on OpenShift?

By default, pods run with a random, non-root user ID from a range assigned to the project. Images that expect to run as root, write to root-owned directories or bind to ports below 1024 fail. Fix the image (make writable paths group-writable for the root group and listen on a high port) rather than granting a more permissive SCC.

35. What is the difference between a Service and a Route?

A Service gives a stable virtual IP and DNS name to a set of pods inside the cluster. A Route publishes a Service outside the cluster at a hostname through the router. You usually need both for an external web application.

36. What storage options does OpenShift support?

OpenShift uses Kubernetes PersistentVolumes, PersistentVolumeClaims and StorageClasses with CSI drivers for cloud disks, vSphere, NFS-style file storage and others. Red Hat OpenShift Data Foundation adds software-defined block, file and object storage based on Ceph.

37. How do you check resource usage?

oc adm top nodes
oc adm top pods -n shop

The web console also shows CPU, memory and network graphs per pod, project and node from the built-in monitoring stack.

38. What is a ResourceQuota and a LimitRange?

A ResourceQuota caps the total CPU, memory, storage and object counts in a project. A LimitRange sets default and maximum requests and limits for each pod or container. Administrators often add both through the project request template.

39. How do you give a user access to a project?

oc adm policy add-role-to-user edit alice -n shop
oc adm policy add-role-to-group view qa-team -n shop
oc get rolebindings -n shop

Default cluster roles include admin, edit and view. Use oc adm policy add-cluster-role-to-user only for cluster-wide access.

40. What is the EX280 exam?

EX280 is the Red Hat Certified OpenShift Administrator exam. It tests creating, configuring and managing an OpenShift cluster and its applications in a hands-on format. Check Red Hat's exam page for the OpenShift version it currently covers.

OpenShift interview questions for 2 to 5 years of experience

41. Which SCCs exist by default?

The defaults include anyuid, hostaccess, hostmount-anyuid, hostnetwork, hostnetwork-v2, nested-container, node-exporter, nonroot, nonroot-v2, privileged, restricted and restricted-v2. Operators can install more. Red Hat warns not to modify the default SCCs, because upgrades can reset them and platform pods depend on them.

42. What is the difference between restricted and restricted-v2?

restricted-v2 adds stricter settings: all capabilities are dropped (only NET_BIND_SERVICE can be added), the seccomp profile defaults to runtime/default, and privilege escalation must be off. New installs since 4.11 give authenticated users restricted-v2. The old restricted SCC is only available to all users on clusters upgraded from 4.10 or earlier.

43. How do you let a workload use a specific SCC?

Grant the pod's service account the use verb on that SCC through RBAC, which scopes the permission. The older shortcut command does the same at the namespace level.

oc create serviceaccount legacy-app -n shop
oc adm policy add-scc-to-user anyuid -z legacy-app -n shop
oc set serviceaccount deployment/legacy-app legacy-app -n shop

Pick the least permissive SCC that works and document why the exception exists.

44. How do you find which SCC admitted a pod?

oc get pod web-6d9f7 -o jsonpath='{.metadata.annotations.openshift\.io/scc}{"\n"}'

45. How does OpenShift choose an SCC when several are allowed?

The admission controller sorts the SCCs the user and service account can use by priority (higher first), then by how restrictive they are, and uses the first one that validates the pod. Red Hat warns that giving default SCCs a priority above 0 can cause cluster instability.

46. How do SCCs relate to Pod Security Admission?

OpenShift runs both. Pod Security Admission applies the Kubernetes Pod Security Standards using namespace labels, and OpenShift synchronises those labels with the SCC permissions of the service accounts in each namespace. Highly privileged system namespaces with the openshift.io/run-level label skip SCC and pod security admission entirely, so never put workloads there.

47. What is OpenShift Pipelines?

OpenShift Pipelines is Red Hat's supported CI/CD solution based on Tekton. Pipelines are built from Tasks that run as pods, defined as Kubernetes resources, with triggers for Git events. It replaces the deprecated Jenkins Pipeline build strategy.

48. What is OpenShift GitOps?

OpenShift GitOps is Red Hat's supported distribution of Argo CD, installed through an Operator. It continuously syncs cluster state with manifests in Git and is used both for applications and for cluster configuration. See our Argo CD interview questions for deeper practice.

49. What is Builds for Red Hat OpenShift?

It is a separate Operator, based on the Shipwright project, that offers a newer build framework alongside BuildConfig. The 5.0 documentation lists "Builds using Shipwright" next to "Builds using BuildConfig". Some features, such as mounting shared entitlements in builds, are delivered through this Operator.

50. How do you speed up slow S2I builds?

Use incremental builds where the builder image supports them, cache dependencies with a local mirror (Maven, npm), use a smaller builder image, and avoid triggering a full rebuild on every commit when an image change is enough. Check the build pod's CPU and memory limits too, because build pods are subject to quotas.

51. How do you build from a private Git repository?

Create a source secret (SSH key or basic auth token), link it to the builder service account or reference it in the BuildConfig's sourceSecret, and use the matching repository URL format.

oc create secret generic git-ssh --from-file=ssh-privatekey=$HOME/.ssh/id_ed25519 \
  --type=kubernetes.io/ssh-auth
oc set build-secret --source bc/web git-ssh

52. How do you pull images from a private external registry?

Create a pull secret and link it to the service account that runs the pods, or add it to the cluster-wide global pull secret for registries all projects use.

oc create secret docker-registry quay-pull --docker-server=quay.io \
  --docker-username=robot --docker-password="$TOKEN"
oc secrets link default quay-pull --for=pull

53. What is OLM?

Operator Lifecycle Manager installs, updates and manages Operators. OLM (Classic) uses CatalogSources, Subscriptions, InstallPlans, ClusterServiceVersions (CSVs) and OperatorGroups. OLM v1, GA since 4.18, simplifies this to ClusterCatalog and ClusterExtension objects.

54. How do you install an Operator with OLM (Classic)?

Create an OperatorGroup for the target namespace (if one does not exist) and a Subscription that names the package, channel and catalog source. OLM creates an InstallPlan and then the CSV.

apiVersion: operators.coreos.com/v1alpha1
kind: Subscription
metadata:
  name: openshift-pipelines-operator
  namespace: openshift-operators
spec:
  channel: latest
  name: openshift-pipelines-operator-rh
  source: redhat-operators
  sourceNamespace: openshift-marketplace

Check the package's available channels in the catalog before using a channel name.

55. What changed with OLM v1?

OLM v1 installs extensions through a cluster-scoped ClusterExtension that references a package from a ClusterCatalog. In 5.0, OLM v1 uses its own Operator Controller service account with cluster administrator permissions, so the spec.serviceAccount field in ClusterExtension is optional and deprecated. Access to extension resources is controlled with RBAC and validating admission policies.

56. How do you check an Operator's status?

oc get subscription,installplan,csv -n openshift-operators
oc describe csv <csv-name> -n openshift-operators
oc get clusterextension          # OLM v1

57. What are Operator catalogs, and what format do they use?

Catalogs are images that list Operator packages, channels and bundles. Red Hat provides catalogs such as redhat-operators, certified-operators and community-operators. Catalogs now use the file-based catalog format; the SQLite database format is deprecated. Red Hat Marketplace was removed from 4.22.

58. How is monitoring provided in OpenShift?

The Cluster Monitoring Operator deploys Prometheus, Alertmanager, Thanos Querier and related components for platform metrics. Administrators can also enable monitoring for user-defined projects, so teams create their own ServiceMonitor, PodMonitor and PrometheusRule objects.

59. How do you enable monitoring for user-defined projects?

Set enableUserWorkload: true in the cluster-monitoring-config ConfigMap in the openshift-monitoring namespace.

apiVersion: v1
kind: ConfigMap
metadata:
  name: cluster-monitoring-config
  namespace: openshift-monitoring
data:
  config.yaml: |
    enableUserWorkload: true

Then grant developers the monitoring-edit role in their projects so they can create ServiceMonitors and alerting rules.

60. How does logging work in OpenShift?

Red Hat OpenShift Logging (6.x) uses the ClusterLogForwarder custom resource as the central configuration for which logs to collect and where to send them. The Loki Operator manages an in-cluster LokiStack for storage, and the Cluster Observability Operator provides log visualisation in the console. Logs can also be forwarded to external systems.

61. What is the OpenShift router, and how do you shard it?

The router is an HAProxy-based Ingress Controller managed by the Ingress Operator. You can create extra IngressController objects and use namespace or route label selectors to split routes between them, for example an internal and an external router. In 4.22.11 and later, an haproxyVersion field lets administrators pin HAProxy 2.8 before upgrading to 5.0.

62. How do you split traffic between two versions with a Route?

A Route can send weighted traffic to its main service and up to three alternate services listed in alternateBackends.

apiVersion: route.openshift.io/v1
kind: Route
metadata:
  name: web
spec:
  host: web.apps.example.com
  to:
    kind: Service
    name: web-v1
    weight: 90
  alternateBackends:
  - kind: Service
    name: web-v2
    weight: 10
  tls:
    termination: edge

63. What is OVN-Kubernetes?

OVN-Kubernetes is the network plugin in OpenShift. It provides pod networking, Services, NetworkPolicy, egress IPs and egress firewalls using Open Virtual Network and Open vSwitch. OpenShift SDN, the older plugin, was removed in 4.17.

64. How do you restrict traffic between projects?

Use Kubernetes NetworkPolicies. A common baseline is to deny all ingress by default and then allow traffic from the same namespace, from the router (the ingress namespace) and from monitoring. Administrators can add AdminNetworkPolicy objects for cluster-wide rules that project owners cannot override.

65. What are egress IPs and egress firewalls?

An egress IP gives traffic leaving a namespace a fixed source address, so external firewalls can allow it. An EgressFirewall restricts which external hosts or CIDRs pods in a namespace can reach. Both are features of OVN-Kubernetes.

66. What are user-defined networks?

User-defined networks let administrators give namespaces their own isolated pod networks instead of the single default cluster network, using UserDefinedNetwork and ClusterUserDefinedNetwork objects. They help with multi-tenancy and with virtual machines that need their own layer 2 network.

67. How do you expose a non-HTTP service outside the cluster?

Routes handle HTTP, HTTPS and TLS with SNI. For other protocols, use a Service of type LoadBalancer (on cloud platforms or with MetalLB on bare metal), a NodePort Service, or Gateway API listeners if your configuration supports the protocol.

68. How do you configure persistent storage for a database?

Create a PVC against a StorageClass that provides block storage, and run the database as a StatefulSet or through its Operator. Use ReadWriteOnce for single-writer databases, set backup procedures, and confirm the storage class supports snapshots and volume expansion if you need them.

69. What is OpenShift Data Foundation?

OpenShift Data Foundation is Red Hat's software-defined storage for OpenShift, based on Ceph and managed by Operators. It provides block, file and S3-compatible object storage from disks in the cluster or external Ceph clusters, plus disaster recovery features.

70. What is OpenShift Serverless?

OpenShift Serverless is Red Hat's distribution of Knative. Knative Serving runs stateless containers with request-based autoscaling including scale to zero and traffic splitting between revisions, and Knative Eventing connects event sources to services.

71. What is OpenShift Service Mesh?

OpenShift Service Mesh is Red Hat's supported Istio distribution, used for mTLS between services, traffic management and observability with Kiali. Version 3 follows upstream Istio more closely. In 5.0, Red Hat bases the cluster's Gateway API controller on Service Mesh 3.4 and Istio 1.30.

72. What is OpenShift Virtualization?

OpenShift Virtualization, based on the KubeVirt project, runs virtual machines alongside containers on the same cluster, managed with Kubernetes objects. Teams use it to move VM workloads onto OpenShift without containerising them first.

73. How do you perform a rolling update and rollback?

oc set image deployment/web web=quay.io/example/web:1.5
oc rollout status deployment/web
oc rollout history deployment/web
oc rollout undo deployment/web

74. How do you add health checks?

Set readiness, liveness and startup probes on the container, either in YAML or with oc set probe.

oc set probe deployment/web --readiness --get-url=http://:8080/ready --initial-delay-seconds=5
oc set probe deployment/web --liveness --get-url=http://:8080/healthz --period-seconds=10

75. How do you run a pod on specific nodes?

Label the nodes and use a nodeSelector or node affinity in the pod spec. Administrators can also set a default node selector for a project with the openshift.io/node-selector annotation, and use taints and tolerations to keep general workloads off dedicated nodes such as infra nodes.

76. What are infrastructure nodes?

Infrastructure nodes are worker nodes labelled and tainted to run only platform services such as the router, registry, monitoring and logging. Separating them keeps platform workloads stable and may reduce subscription costs, depending on your agreement with Red Hat.

77. How do you autoscale nodes?

Create a ClusterAutoscaler resource for cluster-wide limits and a MachineAutoscaler for each MachineSet you want to scale. When pods are pending for lack of capacity, the autoscaler adds machines to the MachineSet, and it removes underused nodes later.

78. What is a MachineSet?

A MachineSet, part of the Machine API, describes a group of machines with the same cloud instance type, image and zone. Scaling the MachineSet creates or deletes Machine objects, and the platform provisions or removes the underlying VMs. Control plane machines use a ControlPlaneMachineSet.

79. What is a MachineHealthCheck?

A MachineHealthCheck watches nodes in a MachineSet and deletes machines whose nodes stay unhealthy (for example NotReady) for a set time, so the MachineSet replaces them automatically.

80. How do you back up applications?

Use OADP (OpenShift API for Data Protection), Red Hat's Operator based on Velero. It backs up Kubernetes objects and persistent volume data to object storage and restores them to the same or another cluster.

81. How do you secure the internal registry?

Keep the default route private unless you need external pushes, use RBAC (registry-viewer, registry-editor) to control who can pull and push, and restrict which external registries pods may use with the cluster image configuration (allowedRegistries or blockedRegistries).

82. How do you mirror images for a cluster without internet access?

Use the oc-mirror plugin v2 to mirror release images, Operator catalogs and extra images into a local registry, then apply the generated ImageDigestMirrorSet and catalog objects. oc-mirror v1, oc adm release mirror and ImageContentSourcePolicy objects are deprecated.

83. What is a ServiceAccount used for in OpenShift?

Service accounts give pods an identity for API access, image pulls and SCC decisions. Each project has default and builder service accounts, among others. Create dedicated service accounts for applications that need extra permissions, and bind only the roles they need.

84. How do you limit how many projects a user can create?

Configure the project request template and limits in the cluster's project configuration, or remove the self-provisioner cluster role from system:authenticated:oauth so only administrators create projects. Many enterprises use a GitOps request process instead of self-service.

85. What is Red Hat Advanced Cluster Security?

Red Hat Advanced Cluster Security for Kubernetes, based on the StackRox project, adds image vulnerability scanning, deployment policy checks, runtime threat detection and network visualisation across clusters. It complements the platform's built-in SCCs and RBAC.

86. What is the MCP server for OpenShift?

It is a Model Context Protocol server that lets an AI assistant inspect a cluster. It became generally available in 5.0 after a Technology Preview in 4.22, runs read-only with RBAC enforced, and includes toolsets for core resources, network ingress and DNS troubleshooting, Service Mesh and Tekton.

OpenShift interview questions for senior engineers

87. How do OpenShift updates work?

The Cluster Version Operator downloads the target release image, updates cluster Operators in a set order, and then the Machine Config Operator updates and reboots nodes pool by pool. Red Hat's update service recommends safe paths through update channels.

oc adm upgrade                      # current version and recommended updates
oc adm upgrade channel stable-5.0
oc adm upgrade --to=<version>
oc get clusterversion -w

88. What update channels are available?

For 5.0 the channels are candidate-5.0, fast-5.0, stable-5.0, and eus-4.y channels for moving between EUS releases. Stable is the usual choice for production; fast receives a release as soon as Red Hat declares it GA, and candidate is for testing.

89. How do you prepare for a minor version update?

Read the release notes for removed APIs and features, check that all cluster Operators are healthy and not degraded, confirm installed Operators support the target version, take an etcd backup, check PodDisruptionBudgets that could block node drains, and run the update in a non-production cluster first. Red Hat recommends updating only to versions the update service suggests.

90. What is EUS, and how does the lifecycle work in OpenShift 5?

For 4.x, even-numbered releases had Extended Update Support, giving up to 24 months of support, with an optional additional term to 36 months, and EUS-to-EUS updates. For OpenShift 5, Red Hat forecasts a minor release every four months, each with 12 months of full support and 12 months of maintenance support. An optional Extended Life Cycle add-on, for every third release starting with 5.2, extends support to up to 60 months from GA.

91. Can you update from OpenShift 4 to 5.0?

Yes. Red Hat documents updating a cluster from 4.x to 5.0 and then moving nodes to RHCOS 10.2. All nodes must move to RHCOS 10.2 together, because mixed RHCOS 9.x and 10.2 clusters are not supported, and you must switch from the deprecated runC runtime to crun before updating the RHCOS version.

92. What is the Machine Config Operator?

The MCO manages the operating system configuration of RHCOS nodes: files, systemd units, kernel arguments, kubelet and CRI-O settings, and OS updates. It renders MachineConfig objects into a configuration per MachineConfigPool and the Machine Config Daemon on each node applies it, usually with a reboot.

93. What is a MachineConfigPool?

A MachineConfigPool groups nodes (by default master and worker) that receive the same rendered configuration. You can create custom pools, for example for infra or GPU nodes, and pause a pool to delay reboots during a busy period.

oc get mcp
oc get machineconfig
oc patch mcp worker --type merge -p '{"spec":{"paused":true}}'

94. How do you add a kernel argument or file to all workers?

Create a MachineConfig with the machineconfiguration.openshift.io/role: worker label. The MCO renders a new configuration and rolls it out node by node, draining and rebooting each one.

apiVersion: machineconfiguration.openshift.io/v1
kind: MachineConfig
metadata:
  name: 99-worker-kargs
  labels:
    machineconfiguration.openshift.io/role: worker
spec:
  kernelArguments:
  - audit=1

95. What does a degraded MachineConfigPool mean?

The Machine Config Daemon could not apply the configuration on at least one node, or it detected configuration drift (someone changed a managed file by hand). The node keeps running but cannot be updated until an administrator fixes it. oc get mcp shows a non-zero DEGRADEDMACHINECOUNT; the node's annotations and the daemon logs show the reason.

96. How do you back up etcd in OpenShift?

Run the cluster-backup.sh script on one control plane node through a debug pod. The script is maintained by the etcd Cluster Operator and wraps etcdctl snapshot save, saving the snapshot and static pod resources.

oc debug node/master-0
chroot /host
/usr/local/bin/cluster-backup.sh /home/core/assets/backup

Copy the backup off the node and keep backups from before every update.

97. How do you debug a node without SSH?

oc debug node/worker-1
chroot /host
journalctl -u kubelet --since "1 hour ago"
crictl ps -a
oc adm node-logs worker-1 -u crio

RHCOS nodes are meant to be managed through the API, so oc debug node is the supported way to inspect them.

98. What is must-gather?

oc adm must-gather collects cluster state, resource definitions, logs and configuration into a local directory for troubleshooting or a Red Hat support case. Some Operators provide their own must-gather images for component-specific data.

oc adm must-gather --dest-dir=./must-gather
oc adm inspect clusteroperator/ingress --dest-dir=./inspect

99. Which installation methods does OpenShift support?

Installer-provisioned infrastructure (the installer creates VMs, networks and load balancers), user-provisioned infrastructure, the Assisted Installer, the Agent-based Installer for disconnected and bare metal sites, single node OpenShift, and two node clusters. Managed offerings such as Red Hat OpenShift Service on AWS and Azure Red Hat OpenShift are run with the cloud provider.

100. What is single node OpenShift used for?

Single node OpenShift puts the control plane and workloads on one machine. It suits edge sites with limited hardware, such as retail stores or telecom sites. It has no high availability, so the application or a central system must handle node failure.

101. How do you design OpenShift for high availability?

Use three control plane nodes spread across failure domains, run compute nodes in at least three zones, keep infra components with multiple replicas, use PodDisruptionBudgets and topology spread constraints for applications, use storage that replicates across zones, and test etcd restore and cluster recovery regularly.

102. How do you manage many OpenShift clusters?

Red Hat Advanced Cluster Management provides cluster lifecycle, policy and application placement across clusters, often combined with OpenShift GitOps for configuration. For edge fleets, GitOps Zero Touch Provisioning uses PolicyGenerator resources; the older PolicyGenTemplate CR is deprecated in 5.0.

103. How do you control which images can run in the cluster?

Restrict registries through the cluster image configuration, require signed images with image policies, scan images in the pipeline and with Advanced Cluster Security, and enforce rules with admission policies such as ValidatingAdmissionPolicy. Combine this with RBAC so only pipelines can deploy to production projects.

104. How do you handle certificates for the API and default ingress?

Replace the default ingress certificate with one signed by your CA by creating a TLS secret in openshift-ingress and referencing it in the IngressController's defaultCertificate. Add a named certificate for the API server in the APIServer configuration. Many teams automate certificates with the cert-manager Operator for Red Hat OpenShift.

105. How do you configure cluster-wide proxy and trusted CAs?

Set httpProxy, httpsProxy, noProxy and trustedCA in the cluster Proxy object. Cluster Operators pick up the settings, and builds can trust extra CAs through the build configuration. Getting noProxy wrong is a common cause of failing Operators.

106. How do you migrate workloads from DeploymentConfig to Deployment?

Translate the strategy, triggers and hooks: image change triggers become an image trigger annotation or a pipeline step, lifecycle hooks become init containers or Jobs, and custom strategies become Deployment rolling settings or Argo Rollouts. Deploy the Deployment alongside the old object, switch the Service selector or Route, then remove the DeploymentConfig.

107. How do you approach capacity planning?

Measure requests versus actual usage per project, set quotas from that data, keep headroom on each node for system daemons and for one node failure per pool, and track trends in the monitoring stack. Use the cluster autoscaler with sensible maximums, and review overcommit settings per environment.

108. What are the main changes to tell an interviewer about OpenShift 5.0?

Kubernetes 1.36, RHCOS 10.2 as the default OS with RHCOS 9.x deprecated (removal planned for 5.3), runC deprecated in favour of crun, Helm v4 in the console with the Helm v3 CLI deprecated, Gateway API CRDs at 1.5.1, OLM v1 using its own service account, the MCP server for OpenShift reaching GA, and a new four-month release cadence with an Extended Life Cycle option.

Scenario-based OpenShift questions

109. A pod fails with "unable to validate against any security context constraint". What do you do?

The pod requests something its service account's SCCs do not allow, such as running as root, host paths or a privileged container. Read the full event to see which field failed. Change the deployment to run as non-root without extra capabilities if possible. If the workload truly needs more, create a dedicated service account and grant it the least permissive SCC that works.

110. An image runs on Docker but crashes on OpenShift with permission denied. Why?

OpenShift runs the container with a random UID that belongs to the root group (GID 0). The image writes to a directory owned by root and not writable by the group. Fix the Dockerfile with chgrp -R 0 /app && chmod -R g=u /app and avoid ports below 1024.

111. A Route returns "Application is not available". How do you troubleshoot it?

That page comes from the router when it has no healthy backend. Check that the Service exists and has endpoints, that its selector matches the pod labels, that pods are Ready, and that the Route's target port matches the Service port name or number.

oc get route web -o yaml
oc get svc web
oc get endpointslices -l kubernetes.io/service-name=web
oc get pods -l app=web

112. A build fails at the push step. What do you check?

Check the build logs (oc logs build/web-12), then whether the output ImageStream exists, whether the internal registry is healthy and has storage, and whether the builder service account can push. For an external registry, confirm the push secret is set on the BuildConfig.

113. A GitHub webhook does not start builds. What is wrong?

Check the webhook URL and secret from oc describe bc/web, that the cluster API is reachable from GitHub, that the event is a push to the branch named in the BuildConfig, and the delivery log in GitHub. OpenShift ignores event types other than push.

114. A cluster update is stuck partway. How do you investigate?

Run oc adm upgrade and oc get clusterversion to see the message, then oc get co to find a degraded or progressing Operator. If the MCO is updating, check oc get mcp; a node that will not drain is often blocked by a PodDisruptionBudget. Collect oc adm must-gather before changing anything and open a support case if a platform Operator is stuck.

115. A node drain hangs during an update. What do you do?

Find the pod that cannot be evicted, usually because of a PodDisruptionBudget with no headroom or a pod with local storage. Scale the application or fix its unhealthy replicas so the PDB allows an eviction. Talk to the owning team before overriding a PDB.

116. An Operator is stuck in "Installing" or "Pending". How do you fix it?

Check the Subscription and InstallPlan status (an InstallPlan may need manual approval), the CSV conditions, the OperatorGroup in the namespace, and whether the catalog source pod in openshift-marketplace is healthy. In disconnected clusters, missing mirrored images are a common cause.

117. Users cannot log in after an LDAP change. Where do you look?

Check the authentication cluster Operator status, the OAuth configuration (oc get oauth cluster -o yaml), the bind credentials secret and CA in openshift-config, and the oauth-openshift pod logs in openshift-authentication. Keep a working break-glass admin account so an identity provider outage does not lock everyone out.

118. The image registry Operator is degraded after install on bare metal. Why?

On platforms without default object storage, the registry starts with no storage configured. Configure a PVC (ReadWriteMany for more than one replica) or object storage, then set managementState to Managed in the registry configuration.

119. Pods in one project cannot reach a database in another project. What do you check?

Test with the Service DNS name (db.data.svc.cluster.local), then check NetworkPolicies in both projects, AdminNetworkPolicies, and any EgressFirewall. Confirm the database Service has endpoints. If the database is outside the cluster, check egress IP and external firewall rules.

120. A MachineConfigPool is degraded after someone edited a file on a node. How do you recover?

The MCO detected drift. Revert the manual change on the node (or apply it properly as a MachineConfig), then let the daemon reconcile. Read the node annotations and the machine-config-daemon logs to see which file differs. Make all node changes through MachineConfig objects in Git to avoid this.

121. Developers say deployments became slow after quotas were added. What happened?

Rollouts with surge need extra quota for new pods while old pods still run. If the quota is close to current usage, new pods are rejected with "exceeded quota" and the rollout waits. Increase the quota to allow surge, lower maxSurge, or right-size requests.

122. You must give a team cluster-admin-like rights only for its own projects. How?

Bind the admin cluster role with RoleBindings in each of the team's projects, not a ClusterRoleBinding. Add a custom ClusterRole if they need read access to cluster resources such as nodes. Use groups synced from LDAP or OIDC instead of individual users, and manage the bindings through GitOps.

OpenShift objects compared with Kubernetes

OpenShiftKubernetes equivalentStatus in 5.0
ProjectNamespaceSupported
RouteIngress or Gateway API HTTPRouteSupported; Gateway API also available (CRDs 1.5.1)
DeploymentConfigDeploymentDeprecated
BuildConfigNo core equivalent (Tekton or Shipwright)Supported; Pipeline strategy deprecated
ImageStreamNo core equivalentSupported
Security Context ConstraintsPod Security AdmissionSupported; both run together
TemplateHelm chart or KustomizeSupported
OLM v1 ClusterExtensionNo core equivalentGA since 4.18
ImageContentSourcePolicyNo core equivalentDeprecated; use ImageDigestMirrorSet

How to prepare for an OpenShift interview

Get hands-on time with a real cluster: a Red Hat developer sandbox, OpenShift Local on a laptop, or OKD. Practise the full path of an application (project, build, image stream, deployment, service, route, probes, quotas) using only oc. Then practise administrator tasks: adding an identity provider, granting roles and SCCs, installing an Operator, reading oc get co and oc get mcp, and taking an etcd backup. Our OpenShift administration interview questions and scenario-based OpenShift questions give more practice, and the Kubernetes scenario-based questions cover the troubleshooting underneath. For the operating system side, read how RHEL 10 integrates with Kubernetes and OpenShift, and use the 200+ DevOps interview questions guide for the wider DevOps round.

If you want guided practice, DevOps Training Institute runs classroom batches in Kharadi, Pune and live online classes across India. The OpenShift course covers administration and EX280-style tasks, and the CKA course covers the Kubernetes skills OpenShift builds on.

References

FAQs

What is the latest version of OpenShift?

OpenShift Container Platform 5.0 became generally available on 3 November 2026. It is based on Kubernetes 1.36 and uses RHCOS 10.2 by default. The previous release, 4.22, became GA on 9 June 2026, and clusters can update from 4.x to 5.0.

Is DeploymentConfig removed in OpenShift?

No, but it is deprecated, including in 5.0. Existing DeploymentConfigs still work. Use Kubernetes Deployments for new applications and plan to migrate the old objects.

Is BuildConfig still supported?

Yes. BuildConfig supports Docker, Source-to-Image and Custom builds in 5.0. Only the Pipeline build strategy, which ran Jenkins pipelines, is deprecated. Builds for Red Hat OpenShift, based on Shipwright, is offered alongside it.

What is the difference between an SCC and RBAC?

RBAC controls which API actions a user or service account can perform, such as creating Deployments. An SCC controls what a pod is allowed to do at runtime, such as which user ID it runs as or whether it can use host networking. Access to an SCC is itself granted through RBAC.

Which network plugin does OpenShift use?

OVN-Kubernetes. The older OpenShift SDN plugin was deprecated in 4.15 and 4.16 and removed in 4.17, so clusters must use OVN-Kubernetes before updating past 4.16.

What is the difference between OLM Classic and OLM v1?

OLM Classic manages Operators with CatalogSources, Subscriptions, InstallPlans, CSVs and OperatorGroups. OLM v1, GA since 4.18, uses ClusterCatalog and ClusterExtension objects and a simpler, cluster-scoped model.

Which certification should I take for OpenShift jobs?

EX280, the Red Hat Certified OpenShift Administrator exam, is the usual target for administrators. Kubernetes certifications such as CKA are also useful, since OpenShift interviews include plain Kubernetes questions. Employers value hands-on experience you can explain as much as the certificate.

How many OpenShift interview questions should I practise?

Quality matters more than count. Make sure you can answer the freshers' section without notes, explain SCCs, Routes and BuildConfig strategies in your own words, and talk through at least ten of the scenario questions using real oc commands.

What's Your Reaction?

Like Like 0
Dislike Dislike 0
Love Love 0
Funny Funny 0
Angry Angry 0
Sad Sad 0
Wow Wow 0
Mridul I am a passionate technology enthusiast with a strong focus on DevOps, Cloud Computing, and Cybersecurity. Through my blogs at DevOps Training Institute, I aim to simplify complex concepts and share practical insights for learners and professionals. My goal is to empower readers with knowledge, hands-on tips, and industry best practices to stay ahead in the ever-evolving world of DevOps.