System Architecture & Deployment Quiz

System Architecture & Deployment Quiz

Test your system architecture knowledge across microservices, caching, and scalability.

1. 1. What is the primary difference between a monolithic and a microservices architecture?

   • A monolith uses containers; microservices do not
   • A monolith is deployed as a single unit; microservices are deployed as independent services
   • Microservices must share a single database; a monolith cannot
   • A monolith can only run on one server; microservices run on many

2. 2. Which HTTP method is conventionally used to UPDATE an existing resource in a REST API?

   • POST
   • GET
   • PUT
   • DELETE

3. 3. In one sentence, what does Docker do?

   • It provisions virtual machines on cloud providers
   • It packages an application and its dependencies into a portable container that runs consistently anywhere
   • It replaces the operating system kernel for better performance
   • It orchestrates clusters of servers across data centers

4. 4. What does CI/CD stand for?

   • Container Integration / Container Delivery
   • Continuous Inspection / Continuous Deployment
   • Continuous Integration / Continuous Delivery (or Deployment)
   • Code Integration / Code Distribution

5. 5. Why are environment variables preferred over hard-coded values for secrets in an application?

   • They make the application run faster
   • They keep sensitive values out of source code and allow configuration per environment
   • They are automatically encrypted at the OS level
   • They are required by the Docker runtime

6. 6. What is the main job of a load balancer?

   • Compressing HTTP responses before they reach the client
   • Distributing incoming network traffic across multiple servers
   • Caching database query results
   • Monitoring application logs in real time

7. 7. What is the difference between horizontal and vertical scaling?

   • Horizontal scaling adds more CPU/RAM to a single server; vertical scaling adds more servers
   • Horizontal scaling adds more servers; vertical scaling adds more CPU/RAM to an existing server
   • They are the same concept but used in different cloud providers
   • Horizontal scaling only applies to databases; vertical scaling only applies to web servers

8. 8. When Nginx is used as a reverse proxy, which traffic direction does it primarily handle?

   • Requests from internal services to the database
   • Outbound requests from the server to external APIs
   • Inbound client requests forwarded to backend application servers
   • Traffic between microservices inside a Kubernetes cluster

9. 9. What is the purpose of a health check endpoint (e.g., GET /health) in a web service?

   • To return the current CPU and memory usage to the end user
   • To allow load balancers and orchestrators to verify that the service is running correctly
   • To trigger a manual restart of the service
   • To expose internal API documentation

10. 10. In a deployment pipeline, what does a rollback accomplish?

    • It deletes all data written since the last release
    • It reverts the running application to a previously known-good version
    • It scales the number of instances back to zero
    • It resets environment variables to their default values

11. 11. The 12-Factor App methodology's 'Config' factor states that configuration should be stored in:

    • A config file committed alongside the source code
    • The environment
    • A dedicated configuration microservice
    • A relational database table

12. 12. In a blue/green deployment, what happens to the 'blue' environment after a successful switch to 'green'?

    • It is permanently deleted to save cost
    • It continues to receive 50 % of production traffic
    • It is kept idle and can serve as a rollback target
    • It is automatically upgraded to match the green environment

13. 13. What differentiates a canary deployment from a blue/green deployment?

    • Canary deployments require two identical environments; blue/green requires only one
    • Canary deployments gradually shift a small percentage of traffic to the new version before a full rollout
    • Canary deployments are only used for database migrations
    • Blue/green deployments are always faster than canary deployments

14. 14. In Kubernetes, what is the role of a Pod?

    • It is the control plane component that schedules containers onto nodes
    • It is the smallest deployable unit, wrapping one or more co-located containers that share network and storage
    • It is the Kubernetes equivalent of a virtual machine
    • It is the configuration manifest for a Kubernetes cluster

15. 15. What is the key difference between DNS-based and registry-based service discovery?

    • DNS-based discovery works only inside a single host; registry-based works across hosts
    • DNS-based discovery resolves service names to IPs via standard DNS; registry-based services query a dedicated registry (e.g., Consul) for live instance lists
    • Registry-based discovery is slower because it requires a round trip to the registry on every request
    • There is no practical difference: both achieve exactly the same result

16. 16. What problem does an API gateway pattern solve in a microservices architecture?

    • It replaces the need for a message broker between services
    • It provides a single entry point for clients, handling cross-cutting concerns like auth, rate limiting, and routing
    • It synchronises databases across services automatically
    • It eliminates the need for service health checks

17. 17. Why should microservices be designed to be stateless?

    • Stateless services run faster because they never write to disk
    • Stateless services can be freely scaled out and replaced without migrating session data
    • Stateless services do not need a load balancer
    • Kubernetes only supports stateless workloads

18. 18. In the circuit breaker pattern, what does the 'Half-Open' state represent?

    • The circuit is permanently disabled and traffic is rejected
    • The circuit allows a limited number of test requests through to see whether the downstream service has recovered
    • The circuit is passing all traffic normally after a successful recovery
    • The circuit is blocked but still logging requests for audit purposes

19. 19. How does the token bucket algorithm differ from the leaky bucket algorithm for rate limiting?

    • Token bucket enforces a strict constant output rate; leaky bucket allows bursts up to the bucket capacity
    • Token bucket allows bursts up to the bucket capacity; leaky bucket enforces a strict constant output rate
    • They are identical: just different names used by different vendors
    • Leaky bucket is only used for UDP traffic; token bucket is for HTTP

20. 20. What does 'immutable infrastructure' mean?

    • Servers are patched in place and never restarted
    • Servers are never modified after deployment; changes are applied by replacing instances with new ones built from updated images
    • Infrastructure configuration files are stored in a read-only S3 bucket
    • All servers in a cluster must have identical hardware specifications

21. 21. Why does a rolling update in Kubernetes use readiness probes before routing traffic to new pods?

    • To check that the container image was signed by a trusted registry
    • To ensure the new pod is fully initialised and ready to serve requests before old pods are removed
    • To throttle the deployment speed to avoid overwhelming the scheduler
    • To run database migrations inside the new pod before traffic is sent

22. 22. What is the expand-contract (parallel change) pattern for database migrations?

    • Run a full database backup before and after every schema change
    • First expand the schema to support both old and new shapes, deploy code that works with both, then contract the schema by removing the old shape
    • Scale the database cluster horizontally before applying a migration, then scale it back down afterwards
    • Use feature flags to hide new columns from users until the migration is complete

23. 23. In the context of distributed systems, what does API idempotency guarantee?

    • A request will always be processed in under 100 ms
    • Making the same request multiple times produces the same result as making it once
    • The API will automatically retry failed requests on behalf of the client
    • Each request is processed by exactly one server instance

24. 24. In the Saga pattern, what is the key difference between choreography and orchestration?

    • Choreography uses a central coordinator to call each service in order; orchestration uses events emitted by each service
    • Choreography has each service react to domain events and publish new ones; orchestration uses a central saga orchestrator to command each step
    • Choreography is only used for read operations; orchestration is only for write operations
    • They are equivalent implementations with different naming conventions

25. 25. The Strangler Fig pattern is used to:

    • Gradually replace a legacy monolith by routing specific features to new microservices until the monolith can be retired
    • Throttle traffic to a legacy system to force users onto a new platform faster
    • Inject faults into a legacy system to find hidden dependencies before migration
    • Wrap a legacy database with an ORM to decouple it from the application layer

26. 26. What is the primary goal of chaos engineering tools like Chaos Monkey and Gremlin?

    • To stress-test databases by generating millions of random write operations
    • To proactively inject failures into production systems to expose weaknesses before they cause unplanned outages
    • To simulate high network latency during load testing in staging environments only
    • To fuzz API inputs and detect security vulnerabilities in microservices

27. 27. What does a service mesh like Istio or Linkerd provide that a traditional API gateway does not?

    • A single external entry point for all client-facing HTTP traffic
    • Mutual TLS, traffic observability, and fine-grained routing between every service-to-service call inside the cluster
    • Automatic horizontal pod autoscaling based on network throughput
    • DNS resolution for services that span multiple Kubernetes namespaces

28. 28. In GitOps (using ArgoCD or Flux), what is the single source of truth for the desired cluster state?

    • The running containers in the Kubernetes cluster
    • A Git repository containing declarative infrastructure and application manifests
    • The CI pipeline's most recently published artifact
    • An operations runbook stored in a wiki

29. 29. In a multi-region active-active architecture, which consistency challenge must be explicitly addressed?

    • Ensuring all regions use the same container image digest
    • Handling write conflicts and replication lag when the same data can be modified concurrently in different regions
    • Routing all writes through a single primary region to avoid latency
    • Synchronising Kubernetes node clocks across regions via NTP

30. 30. Which three pillars make up full observability in a distributed system, and why are correlation IDs important?

    • Uptime, latency, and error rate; correlation IDs link SLO alerts to on-call schedules
    • Logs, metrics, and traces; correlation IDs allow a single request to be tracked end-to-end across all services and their telemetry
    • Dashboards, alerts, and runbooks; correlation IDs replace the need for distributed tracing
    • CPU, memory, and disk; correlation IDs map infrastructure metrics to application deployments