Great! Let's break this down.
Based on your earlier messages, here's how I
can structure answers for you across key categories:
✅ Terraform
1. What Terraform modules have you written?
Answer:
"I've written reusable Terraform modules for provisioning EC2 instances, VPC
setup, EKS cluster creation, and S3 buckets with versioning and lifecycle
policies. For example, in the EKS module, I parameterized the VPC ID, subnet
IDs, and node group configurations to be reused across environments."
2. Difference between count and for_each?
Answer:
count is used for simple numeric iteration and indexing.
for_each is used when you want to iterate over a map or a set of
strings, especially when you need to reference specific elements by
key.
count is index-based; for_each is key-based.
3. Difference between local-exec and remote-exec?
Answer:
local-exec runs commands on the machine where Terraform is
running.
remote-exec runs commands on the resource being provisioned (like an
EC2 instance) using SSH.
4. State locking in Terraform?
Answer:
State locking prevents concurrent modifications of the state file. I’ve
implemented state locking using a DynamoDB table in AWS. This ensures
that only one terraform apply can happen at a time.
5. EC2 deleted manually – how to restore?
Answer:
Run terraform plan. Terraform will detect that the EC2 instance is missing
and show it as a resource to be recreated. Then run terraform apply to
restore it.
✅ Kubernetes + ArgoCD
6. How is your application deployed to AWS EKS?
Answer:
"We package our applications using Helm charts and deploy them to EKS via
ArgoCD. ArgoCD syncs changes from Git repos to the cluster. Our CI pipeline
builds Docker images, pushes them to ECR, and updates Helm values in Git."
7. How is ArgoCD installed and configured?
Answer:
"We install ArgoCD in the argocd namespace using the official Helm chart or
YAML manifests. It’s configured with access to the EKS cluster via its in-
cluster Kubernetes service account."
8. How do you log in to the ArgoCD UI?
Answer:
"Either using the admin password (from secret) or via SSO if configured. For
DevOps teams, we map users via RBAC using ArgoCD’s RBAC policies file."
9. What is the difference between ReplicaSet and StatefulSet?
Answer:
ReplicaSet is stateless and used in Deployments.
StatefulSet is used for stateful apps that need persistent identity and
storage (like databases). It ensures stable pod names and ordered
deployment.
✅ CI/CD – Jenkins
10. What have you done in Jenkins?
Answer:
"I’ve written declarative pipelines for CI/CD processes involving code build,
test, Docker image creation, and deployment to Kubernetes. I also manage
shared libraries, credential binding, and plugins like SonarQube, Artifactory,
and Slack."
11. Deployment strategy?
Answer:
"We use rolling updates for most apps, and canary deployment for
critical services. We control rollout using Helm and Kubernetes deployment
strategies."
✅ Debugging Kubernetes Issues
12. Application shows 503 error – troubleshooting steps?
Answer:
Check pod status with kubectl get pods
Describe pod and service: kubectl describe pod, kubectl describe svc
Check logs: kubectl logs
Confirm ingress rules and readiness probes
Verify DNS and service endpoints
13. Pod-to-pod communication issue – what to check?
Answer:
NetworkPolicy blocking traffic
DNS resolution (nslookup or dig)
Service or endpoint not created
Pod IPs not reachable (CNI plugin issue)
✅ Other Quick Hits
14. Git: Pull vs Fetch?
git fetch: downloads changes but doesn’t merge
git pull: fetch + merge automatically
15. Jenkins pipeline not up – what could be the reason?
JVM crashed, disk full, memory issue
Misconfigured plugins or corrupted job configs
Restart Jenkins and check logs in /var/log/jenkins/jenkins.log
16. How to secure Jenkins?
Enable matrix-based security
Integrate with LDAP or SSO
Use credentials binding
Backup config.xml and credentials.xml
17. Observability tools used?
Prometheus + Grafana for metrics
ELK stack for logs
Jaeger or OpenTelemetry for tracing
Would you like a PDF or document version of this for revision? Or should we
go deeper into any one section, like ArgoCD, Terraform, or Jenkins pipelines?
