DevOps Comparisons
Compare tools, platforms, and concepts with clear tradeoffs and real-world understanding.
Docker vs Kubernetes
Compare container packaging and runtime workflows with full container orchestration across clusters.
Interview angle
A strong short answer is: Docker is for packaging and running containers, Kubernetes is for orchestrating and operating them at scale.
ECS vs EKS
Compare AWS-native container orchestration with managed Kubernetes on AWS.
Interview angle
The clean answer is: ECS is simpler and AWS-native, EKS is Kubernetes-based, more flexible, and more portable.
Docker Compose vs Kubernetes
Compare local multi-container application composition with production-grade orchestration.
Interview angle
Docker Compose is for local multi-container workflows. Kubernetes is for operating distributed containerized systems in production.
Helm vs Kustomize
Compare Kubernetes application packaging with native manifest customization.
Interview angle
A strong short answer: Helm is packaging plus templating. Kustomize is customization with overlays and patches.
Deployment vs StatefulSet
Compare stateless workload management with stable identity-based stateful workload management in Kubernetes.
Interview angle
Easy interview rule: Deployment is for stateless workloads. StatefulSet is for stable, stateful workloads.
Service vs Ingress
Compare stable service access inside Kubernetes with HTTP and HTTPS traffic routing into the cluster.
Interview angle
A strong short answer: Service gives stable access to pods. Ingress controls external HTTP and HTTPS routing into the cluster.
ConfigMap vs Secret
Compare non-sensitive configuration management with sensitive data storage in Kubernetes.
Interview angle
Good interview answer: ConfigMap is for normal configuration. Secret is for sensitive values, but it still needs strong security controls.
Readiness Probe vs Liveness Probe
Compare traffic eligibility checks with health checks that trigger container restarts in Kubernetes.
Interview angle
The classic short answer is: readiness controls traffic, liveness controls restarts.
DaemonSet vs Deployment
Compare one-pod-per-node scheduling with normal replica-based application deployment in Kubernetes.
Interview angle
The clean answer is: DaemonSet is for node coverage, Deployment is for normal application replicas.
Recreate vs Rolling Update
Compare all-at-once application replacement with gradual controlled rollout during deployment.
Interview angle
Strong short answer: Recreate is a full replacement with possible downtime. Rolling Update is a gradual rollout with better availability.
GitHub Actions vs Jenkins
Compare GitHub-native CI/CD automation with a highly customizable self-hosted automation server.
Interview angle
A strong short answer: GitHub Actions is easier and GitHub-native, while Jenkins is more customizable but much heavier to operate.
GitHub Actions vs GitLab CI
Compare CI/CD systems built directly into GitHub and GitLab development ecosystems.
Interview angle
A practical interview answer is: both are strong, but in real teams the better choice usually follows the source-control platform already in use.
Terraform vs Ansible
Compare infrastructure provisioning with configuration management and automation.
Interview angle
A clear answer is: Terraform provisions infrastructure, Ansible configures what runs on top of it.
Terraform vs CloudFormation
Compare a multi-provider infrastructure-as-code tool with AWS-native stack-based provisioning.
Interview angle
A good short answer is: CloudFormation is AWS-native and stack-based, while Terraform is broader, more portable, and multi-provider.
CI vs CD
Compare continuous integration with continuous delivery or deployment in modern software workflows.
Interview angle
A clean interview answer is: CI validates code continuously, while CD automates getting validated code safely delivered toward production.
Blue-Green vs Canary Deployment
Compare full-environment switchovers with gradual percentage-based rollout strategies.
Interview angle
The sharp short answer is: blue-green is a traffic switch between two full versions, canary is a gradual rollout to a subset of users.
Artifact vs Container Image
Compare general build outputs with OCI-style packaged runtime units for containerized delivery.
Interview angle
A strong answer is: a container image is one kind of artifact, but the word artifact is broader than container images alone.
Packer vs Docker
Compare machine-image creation with container-image creation.
Interview angle
A clean short answer is: Packer builds machine images, Docker builds container images.
git merge vs git rebase
Compare preserving branch history with rewriting commits into a cleaner linear history.
Interview angle
The standard answer is: merge preserves history, rebase rewrites history to produce a cleaner linear timeline.
git fetch vs git pull
Compare downloading remote updates safely with downloading and integrating them immediately.
Interview angle
The classic short answer is: git pull = git fetch + integrate. Fetch is safer, pull is more convenient.
git reset vs git revert
Compare rewriting branch history with safely creating a new commit that undoes previous changes.
Interview angle
A strong answer is: reset rewrites history, revert safely records an undo commit in history.
Prometheus vs Datadog
Compare open-source metrics monitoring with a fully managed observability platform.
Interview angle
A strong short answer is: Prometheus is open-source and metrics-focused, while Datadog is a managed all-in-one observability platform.
Grafana vs Kibana
Compare flexible multi-source dashboards with Elasticsearch-focused analytics and exploration.
Interview angle
A clean short answer is: Grafana is broader for dashboards across data sources, while Kibana is deeper for Elasticsearch-based search and log analysis.
Loki vs Elasticsearch
Compare lightweight label-based log aggregation with full-text indexing and search analytics.
Interview angle
A strong short answer is: Loki indexes labels, Elasticsearch indexes full content.
CloudWatch vs Prometheus
Compare AWS-native managed monitoring with open-source metrics monitoring for cloud-native environments.
Interview angle
The short answer is: CloudWatch is AWS-native and managed, while Prometheus is open-source and cloud-native focused.
CloudTrail vs CloudWatch
Compare AWS API audit logging with infrastructure and application monitoring.
Interview angle
A very clean answer is: CloudTrail is audit, CloudWatch is monitoring.
Metrics vs Logs
Compare numerical time-series telemetry with detailed event records for monitoring and troubleshooting.
Interview angle
The classic answer is: metrics tell you that something is wrong, logs help you understand why.
Logs vs Traces
Compare event records with distributed request flow visibility across services.
Interview angle
A strong short answer is: logs are event details, traces are request journeys.
Black-box vs White-box Monitoring
Compare external behavior monitoring with internal instrumented system visibility.
Interview angle
The clean short answer is: black-box sees the system from outside, white-box sees it from inside.
Push vs Pull Monitoring
Compare telemetry collection models where the monitoring system scrapes targets versus targets sending data outward.
Interview angle
A good short answer is: pull means the monitoring system scrapes targets, push means targets send telemetry outward themselves.
Alerting vs SLO
Compare reactive operational notifications with reliability targets and service-level objectives.
Interview angle
A sharp short answer is: alerts are signals, SLOs are targets.
EC2 vs Lambda
Compare virtual machine-based compute with serverless function execution in AWS.
Interview angle
The clean short answer is: EC2 gives server control, Lambda gives serverless execution.
S3 vs EBS
Compare object storage with block storage in AWS.
Interview angle
A strong short answer is: S3 is object storage, EBS is block storage for EC2.
S3 vs EFS
Compare object storage with managed shared file storage in AWS.
Interview angle
The short answer is: S3 is object storage, EFS is shared file storage.
RDS vs DynamoDB
Compare managed relational databases with managed NoSQL key-value and document storage in AWS.
Interview angle
A strong short answer is: RDS is for relational SQL workloads, DynamoDB is for high-scale NoSQL access patterns.
ALB vs NLB
Compare Layer 7 application load balancing with Layer 4 network load balancing in AWS.
Interview angle
The short answer is: ALB is Layer 7 and smart for HTTP, NLB is Layer 4 and strong for raw network traffic.
SNS vs SQS
Compare pub/sub event broadcasting with durable queue-based asynchronous processing in AWS.
Interview angle
A strong short answer is: SNS broadcasts, SQS queues.
SQS vs EventBridge
Compare queue-based asynchronous messaging with event bus routing in AWS.
Interview angle
The short answer is: SQS is for queued work, EventBridge is for event routing.
IAM User vs IAM Role
Compare long-lived AWS identities with temporary assumable permission models.
Interview angle
A strong short answer is: users are long-lived identities, roles provide temporary credentials and are preferred for workloads.
CloudFront vs ALB
Compare global CDN edge delivery with regional application load balancing.
Interview angle
A clean short answer is: CloudFront is edge delivery, ALB is backend application routing.
Public Subnet vs Private Subnet
Compare internet-routable subnet placement with isolated internal subnet placement in AWS.
Interview angle
The short answer is: public subnets are internet-routable, private subnets are for internal workloads with no direct inbound internet exposure.
NAT Gateway vs Internet Gateway
Compare outbound internet access for private resources with direct internet routing for public subnet resources.
Interview angle
A strong short answer is: Internet Gateway gives public routing, NAT Gateway gives private subnets outbound internet access.
TCP vs UDP
Compare reliable connection-oriented transport with lightweight connectionless transport in networking.
Interview angle
The short answer is: TCP is reliable and ordered, UDP is fast and connectionless.
HTTP vs HTTPS
Compare plain web traffic with encrypted web traffic protected by TLS.
Interview angle
A clean short answer is: HTTP is plain text, HTTPS is encrypted HTTP over TLS.
Monolith vs Microservices
Compare single deployable application architecture with distributed service-based architecture.
Interview angle
A strong short answer is: monoliths optimize simplicity, microservices optimize service independence at the cost of complexity.
Horizontal vs Vertical Scaling
Compare scaling by adding more instances with scaling by increasing the size of one instance.
Interview angle
The short answer is: horizontal scaling adds more instances, vertical scaling makes one instance bigger.
Cache vs Database
Compare fast temporary access layers with durable source-of-truth storage.
Interview angle
A strong short answer is: cache accelerates access, database persists truth.
RabbitMQ vs Kafka
Compare traditional broker-based messaging with distributed event streaming.
Interview angle
A strong short answer is: RabbitMQ is for classic brokered messaging, Kafka is for event streaming and durable logs.
SQL vs NoSQL
Compare relational structured databases with non-relational flexible data models.
Interview angle
A clean short answer is: SQL is relational and structured, NoSQL is flexible and model-specific.
REST vs GraphQL
Compare resource-based API design with query-driven API data selection.
Interview angle
A strong short answer is: REST is resource-based and simpler operationally, GraphQL is query-based and more flexible for clients.
JWT vs Session Authentication
Compare stateless token-based authentication with server-managed session state.
Interview angle
A clean short answer is: JWT is stateless token auth, session auth keeps state on the server.
WAF vs Firewall
Compare application-layer web filtering with broader network-level traffic control.
Interview angle
A strong short answer is: firewall controls network traffic, WAF protects web traffic at the application layer.