AWS Interview Questions
Prepare for AWS interviews with 150 questions and detailed answers for .NET Full Stack Developers. Cover AWS fundamentals, application hosting, Lambda, databases, storage, security, networking, messaging, CI/CD, monitoring, distributed systems, cloud architecture and production troubleshooting.
1. AWS Fundamentals
01What is Amazon Web Services (AWS), and why do organizations use it for application development?
AWS is Amazon’s cloud platform offering compute, storage, databases, networking, security and developer services on demand. Teams use it to avoid managing physical data centres, scale capacity, deploy globally and pay for measured usage. A .NET solution might run APIs on ECS or App Runner, store files in S3 and use RDS for SQL Server.
02What is cloud computing? Explain IaaS, PaaS and SaaS with real-world AWS examples.
IaaS provides virtual machines and networks that you manage (EC2); PaaS provides a managed runtime/platform (Elastic Beanstalk or App Runner); SaaS is a finished application delivered over the internet. The key difference is how much infrastructure and software the customer must operate.
03What is the difference between public, private and hybrid cloud?
A public cloud uses a provider’s shared infrastructure with tenant isolation; a private cloud is dedicated to one organization; a hybrid cloud combines public cloud with private or on-premises systems. Hybrid is useful when data, legacy systems or compliance requirements mean some workloads must remain on-premises.
04What is an AWS Account, and why is it required?
An AWS account is the primary boundary for resources, billing, quotas and access controls. It provides an isolated environment in which services are created and usage is charged. Organizations commonly separate production, development and security workloads into multiple accounts.
05What is an AWS Region, and how does it differ from an Availability Zone?
A Region is a separate geographic area containing multiple Availability Zones. An Availability Zone is an isolated location within a Region, designed with independent power, cooling and networking. Deploying across AZs improves resilience to a single data-centre failure.
06What is an Amazon VPC, and why is it required?
A VPC is a logically isolated virtual network in which AWS resources run. You define its IP address ranges, subnets, route tables and network controls. It lets an application place public entry points separately from private APIs and databases.
07What is the difference between a public subnet and a private subnet in AWS?
A public subnet has a route to an Internet Gateway and can host resources that need direct internet reachability, subject to public IPs and security rules. A private subnet has no direct inbound internet route; outbound access may use a NAT Gateway. Databases and internal services normally belong in private subnets.
08What is the AWS Management Console, and how is it different from AWS CLI?
The AWS Management Console is a browser-based graphical interface for operating services. AWS CLI is a command-line tool suited to repeatable scripts, automation and terminal workflows. Both call AWS APIs, so permissions determine what either can do.
09What is AWS CloudFormation, and why is Infrastructure as Code important?
CloudFormation describes AWS infrastructure in templates and provisions it consistently. Infrastructure as Code makes environments repeatable, reviewable, version-controlled and easier to recreate. Templates can define resources, parameters, outputs and dependencies.
10What is AWS IAM, and how does it control access to AWS resources?
IAM manages identities and permissions for AWS resources. Policies specify allowed or denied actions on resources; users, groups and roles receive permissions through policy evaluation. Apply least privilege and prefer temporary role credentials over long-lived keys.
2. AWS Application Hosting and .NET Applications
11What is Amazon EC2, and how can you host an ASP.NET Core Web API on an EC2 instance?
EC2 provides virtual servers. To host an ASP.NET Core API, launch a suitable Windows or Linux instance, install the .NET hosting runtime if needed, deploy the published application, configure a service or web server such as IIS/Nginx, open only required ports, add TLS through a load balancer and configure monitoring.
12What is an EC2 instance type, and how do instance families differ?
EC2 instance families are optimized for different workloads: general purpose balances CPU and memory, compute optimized suits CPU-heavy work, memory optimized suits large in-memory workloads, and storage or accelerated-computing families target specialized needs. Choose based on measured utilization and test performance.
13What is an Amazon Machine Image (AMI)?
An AMI is a reusable image containing the operating system and configuration needed to launch EC2 instances. It can include software and settings, and is used by Auto Scaling and repeatable deployments. Keep images patched and versioned.
14How do you deploy an ASP.NET Core application to an Amazon EC2 instance running Windows or Linux?
Publish the .NET app for the target runtime, provision EC2, install the required .NET runtime/hosting bundle, copy the artifact, configure IIS on Windows or systemd plus Nginx on Linux, set environment variables and secrets, configure security groups and TLS, then test health and logs. Automate repeat deployments with CI/CD rather than manual copying.
15What is AWS Elastic Beanstalk, and how can you deploy a .NET application using it?
Elastic Beanstalk is a managed application deployment service that provisions resources such as EC2, load balancing and Auto Scaling. Package the .NET application in a supported deployment format, create an environment, deploy the version and configure environment properties. It reduces infrastructure setup while still allowing access to underlying resources.
16What is AWS App Runner, and how is it used to host containerized web applications?
App Runner is a managed service for building and running web applications and APIs from source code or container images. Connect a repository or ECR image, configure build/runtime settings, port, environment variables and instance sizing; App Runner manages deployment, TLS endpoint, scaling and health checks.
17What is Amazon ECS, and how does it help deploy containerized .NET applications?
ECS is AWS’s container orchestration service. Define a cluster, task definition with the .NET container image, IAM task roles, networking and resource limits, then run a service with desired task count and optional load balancer. ECS manages task placement and service health.
18What is Amazon ECR, and why is it used with Docker-based .NET applications?
ECR is a managed private container image registry. A CI pipeline builds the .NET Docker image, authenticates to ECR, tags and pushes it; ECS or App Runner then pulls the approved image. Use immutable tags/digests and scan images for vulnerabilities.
19What is an Application Load Balancer (ALB), and why is it useful for ASP.NET Core applications?
An Application Load Balancer distributes HTTP/HTTPS traffic to healthy targets such as EC2 instances or ECS tasks. It supports host/path routing, TLS termination and health checks. It allows multiple stateless .NET API instances to share traffic and recover from an unhealthy target.
20What is Amazon CloudFront, and how can it be used to deliver an Angular application?
CloudFront is a content delivery network that caches content at edge locations close to users. Build Angular into static files, store them in S3 or an origin, put CloudFront in front, configure HTTPS and SPA fallback routing, and invalidate or version assets during releases.
3. AWS Lambda and Serverless
21What is AWS Lambda, and how is it different from hosting an ASP.NET Core Web API on EC2?
Lambda runs code in response to events without requiring you to manage servers. EC2 gives persistent virtual-machine control and is suitable for continuously running services; Lambda is event-driven, automatically scales within quotas and charges by execution/resources. Lambda has runtime, duration and execution-environment constraints.
22What is serverless computing in AWS?
Serverless means AWS manages server provisioning and much of the scaling and availability work while code runs on demand or in response to events. It does not mean there are no servers. Developers still design for timeouts, retries, concurrency, permissions, observability and cost.
23What are Lambda triggers and event sources?
Triggers connect Lambda to events such as API Gateway requests, S3 object events, EventBridge schedules, SQS messages or DynamoDB streams. The event source invokes the function with an event payload. Configure permissions, batching, retry behavior and dead-letter handling for the source.
24What is Amazon API Gateway, and how can it expose AWS Lambda functions as REST APIs?
API Gateway creates and manages HTTP API endpoints and routes requests to Lambda or other backends. Configure routes, integrations, authorization, throttling, validation and stages; grant API Gateway permission to invoke the function. The Lambda handler processes the request and returns the expected response shape.
25What is the difference between AWS Lambda and Amazon ECS?
Lambda is suited to event-driven, short-lived functions with managed scaling; ECS runs containerized services or workers with more control over runtime, networking and long-running processes. Choose Lambda for discrete event handlers and ECS for continuously running APIs, workers or workloads needing container control.
4. AWS Database and Storage
26What is Amazon RDS, and how is it different from installing SQL Server on an EC2 instance?
RDS is a managed relational database service that handles tasks such as provisioning, backups and patching. SQL Server on EC2 gives more OS and instance control but you manage the database installation, backups, patching and high availability. Choose RDS to reduce operations where its engine features meet requirements.
27Which database engines are supported by Amazon RDS?
Amazon RDS supports Amazon Aurora, PostgreSQL, MySQL, MariaDB, Oracle and Microsoft SQL Server. Engine availability, versions, editions and features vary by Region and deployment option, so verify current support before designing a migration.
28What is Amazon RDS for SQL Server, and how can an ASP.NET Core application connect to it?
RDS for SQL Server is a managed SQL Server database. Create the DB instance in private subnets, allow inbound SQL traffic only from the API’s security group, configure credentials or supported authentication, store the connection string securely, and connect using Microsoft.Data.SqlClient or EF Core’s SQL Server provider.
29What is Amazon S3, and what types of data can it store?
S3 is object storage for unstructured data such as documents, images, videos, backups, logs and static website assets. Objects are stored in buckets and addressed by keys. It offers durability, lifecycle policies, versioning and access controls.
30What is the difference between Amazon S3, Amazon EBS and Amazon EFS?
S3 stores objects and is accessed through APIs; EBS provides block storage volumes typically attached to EC2; EFS provides a shared elastic file system that multiple Linux instances can mount. Choose object storage for files, block storage for server disks and shared file storage for POSIX-style access.
31What are the different Amazon S3 storage classes?
S3 classes include Standard, Intelligent-Tiering, Standard-IA, One Zone-IA, Glacier Instant Retrieval, Glacier Flexible Retrieval and Glacier Deep Archive. They trade access latency, availability, storage price and retrieval cost. Select by access frequency and retention requirements; confirm current class names and constraints.
32What is an S3 Bucket, and how is it different from an S3 Object?
A bucket is a top-level S3 container in a Region; an object is an individual stored item consisting of data, key and metadata. A bucket contains many objects, and permissions and lifecycle rules can be applied at bucket or object-related scope.
33What is an S3 pre-signed URL, and why is it used?
A pre-signed URL is a time-limited URL signed by an authorized AWS identity that grants a specific operation on an S3 object. It lets a client upload or download a private object without receiving AWS credentials. Keep expiry short and restrict method, key and content size where possible.
34How can you upload and download files from Amazon S3 using C# and the AWS SDK for .NET?
Use the AWS SDK for .NET, typically AWSSDK.S3. Create an IAmazonS3 client using the default credential chain or workload role, then call PutObjectAsync or TransferUtility for upload and GetObjectAsync/download methods for retrieval. Validate file type/size and grant only required S3 permissions.
35What is Amazon DynamoDB, and how does it differ from Amazon RDS?
DynamoDB is a managed NoSQL key-value/document database designed for low-latency access at scale. RDS is relational and supports SQL, joins and relational constraints. Choose DynamoDB for access patterns built around partition/sort keys and flexible items; choose RDS for relational queries and transactions.
5. AWS Security and Networking
36What is AWS IAM, and what is the difference between an IAM User, Group and Role?
An IAM user represents a long-lived identity, a group organizes IAM users for shared permissions, and a role is an assumable identity that provides temporary credentials. Applications and AWS services should generally use roles rather than permanent user access keys.
37What is an IAM Policy, and how does AWS evaluate permissions?
An IAM policy is a JSON document describing effects, actions, resources and optional conditions. AWS evaluates applicable identity and resource policies, boundaries and organization controls; an explicit Deny overrides an Allow. If no applicable Allow exists, access is denied.
38What is an IAM Role, and why should applications use roles instead of hardcoded AWS access keys?
An IAM role provides temporary credentials through an AWS security token service session. Attach a least-privilege role to EC2, ECS, Lambda or another supported workload so the SDK obtains credentials automatically. This avoids storing long-lived access keys in source code or configuration.
39What is AWS Secrets Manager, and why should .NET developers use it?
Secrets Manager stores and retrieves sensitive values such as database passwords and API credentials, supports encryption and can integrate with rotation workflows. A .NET application should access it through an IAM role and AWS SDK, cache secrets appropriately and avoid logging them.
40What is AWS Key Management Service (KMS)?
KMS creates and manages cryptographic keys used by AWS services and applications. It supports encryption operations, key policies, grants, auditing and integration with services such as S3, RDS and Secrets Manager. Control key access carefully because encryption security depends on permissions.
41What is an Internet Gateway in AWS?
An Internet Gateway attaches to a VPC and enables communication between resources with public addressing and the internet, when route tables and security controls permit it. It is horizontally scaled and highly available; it does not by itself make every subnet or instance public.
42What is a NAT Gateway, and why is it required for private subnets?
A NAT Gateway lets instances in a private subnet initiate outbound IPv4 connections to the internet while preventing unsolicited inbound internet connections. The private subnet routes internet-bound traffic to the NAT Gateway in a public subnet, which routes through an Internet Gateway. Use per-AZ NAT design for resilience.
43What is the difference between a Security Group and a Network ACL?
Security Groups are stateful virtual firewalls attached to network interfaces/resources; return traffic is automatically allowed. Network ACLs are stateless rules applied at subnet level and require explicit inbound and outbound rules. Security Groups are commonly the primary workload control; NACLs provide an additional subnet boundary.
44What is Amazon Cognito, and how can it provide authentication for an Angular application?
Cognito provides user directories and identity federation, including sign-up, sign-in and token issuance. An Angular app can use a Cognito User Pool with an OAuth/OIDC flow such as authorization code with PKCE, then send access tokens to an ASP.NET Core API that validates issuer, audience and signature.
6. Deployment and Monitoring
45What is AWS CodePipeline, and what is its role in CI/CD?
CodePipeline orchestrates CI/CD stages such as source retrieval, build, test and deployment. It connects services such as CodeBuild, CodeDeploy, S3, ECS and external source providers to automate release workflows and approvals.
46What is the difference between AWS CodeBuild and AWS CodeDeploy?
CodeBuild compiles code, runs tests and creates artifacts or container images in a managed build environment. CodeDeploy automates application deployments to supported targets such as EC2, Lambda and ECS. A pipeline often uses CodeBuild to produce an artifact and CodeDeploy to release it.
47What is AWS CodeCommit, and what alternatives can be used for source control in AWS-based projects?
CodeCommit was AWS’s managed Git repository service, but it is no longer available to new customers as of the service’s announced closure to new customers. Existing users should verify current service status and plan source control with GitHub, GitLab, Bitbucket or another supported Git provider integrated into AWS pipelines.
48What is Amazon CloudWatch, and how can it monitor an ASP.NET Core application?
CloudWatch collects AWS service metrics, logs and events and supports dashboards and alarms. For ASP.NET Core, send structured application logs to CloudWatch Logs, publish or collect metrics, create alarms for errors/latency/CPU and use Logs Insights to investigate request failures. Add correlation IDs and avoid logging secrets.
49What is AWS CloudTrail, and how is it different from CloudWatch?
CloudTrail records account activity and AWS API calls for governance, audit and investigation. CloudWatch focuses on operational metrics, logs and alarms. In short, CloudTrail answers who did what to AWS resources; CloudWatch helps determine how systems are operating.
50How do you troubleshoot an ASP.NET Core application hosted on EC2 or AWS App Runner when it stops responding?
Check the endpoint and load balancer health, instance/task status, security groups, routes, DNS and TLS first. Inspect application and platform logs in CloudWatch, CPU/memory and network metrics, deployment/configuration changes and downstream database health. Reproduce with request IDs, verify health endpoints and roll back a bad release if necessary.
1. AWS Application Hosting and Deployment
01Explain the complete process of deploying an Angular frontend and ASP.NET Core Web API separately on AWS.
Build Angular and publish its static output to S3 behind CloudFront (or host it on App Runner if appropriate); deploy the ASP.NET Core API separately to ECS, App Runner, Elastic Beanstalk or EC2 behind an ALB/API Gateway. Configure DNS, HTTPS, CORS, environment-specific API URLs, IAM roles, secrets, private database access, CI/CD and monitoring.
02How do you configure environment-specific settings such as Development, Staging and Production for a .NET application hosted on AWS?
Use separate AWS accounts or isolated resources for Development, Staging and Production. Set ASPNETCORE_ENVIRONMENT and application settings through App Runner/ECS task definitions, Elastic Beanstalk environment properties, Systems Manager Parameter Store or Secrets Manager. Keep secrets out of source control and promote the same immutable build artifact through environments.
03How do you deploy an ASP.NET Core application to EC2 using a Git-based deployment workflow?
Use GitHub/GitLab or another Git provider to trigger CodePipeline or a CI workflow. Build and test the .NET app, publish a versioned artifact, deploy through CodeDeploy/SSM or a deployment script to EC2, restart the service safely, run health checks and retain a rollback version. Prefer immutable AMIs or blue-green deployment for repeatability.
04How do you configure an Application Load Balancer to distribute traffic across multiple EC2 instances running .NET APIs?
Create an ALB in public subnets, register EC2 instances in a target group, and configure listeners for HTTPS with an ACM certificate. Allow ALB-to-instance traffic only on the application port, configure a health-check path, and attach the instances through an Auto Scaling Group. Verify target health before routing production traffic.
05How do you configure an Auto Scaling Group for an ASP.NET Core application hosted on EC2?
Create a launch template with AMI, instance type, IAM instance profile and security groups; create an Auto Scaling Group across multiple AZs and attach it to the ALB target group. Configure minimum/desired/maximum capacity and scaling policies based on CPU, request count or custom metrics. Ensure the API is stateless and externalizes sessions/files.
06What is the difference between scaling up and scaling out in AWS?
Scale up means using a larger instance with more CPU or memory. Scale out means adding more instances/tasks to share traffic. Scale out improves horizontal resilience but requires stateless services, shared data/cache and load balancing; choose based on bottleneck and cost.
07How do you configure health checks for an ASP.NET Core application behind an Application Load Balancer?
Implement an ASP.NET Core readiness endpoint, for example /health/ready, and configure it as the ALB target-group health check. Return success only when the instance can serve traffic; set suitable path, interval, timeout and healthy/unhealthy thresholds. Keep checks lightweight and avoid marking an instance unhealthy for a transient optional dependency.
08How do you containerize an ASP.NET Core application using Docker and deploy it to Amazon ECS?
Create a multi-stage Dockerfile using the .NET SDK image to restore/build/publish and a smaller ASP.NET runtime image to run the app. Push the image to ECR, define an ECS task with port, CPU/memory, environment and task role, then deploy an ECS service with Fargate or EC2 capacity and an ALB target group.
09How do you push Docker images from a CI/CD pipeline to Amazon ECR?
Build and test the image in CI, authenticate to ECR using the pipeline’s IAM role, tag it with a unique commit/build identifier, and push it to the repository. Update the ECS task definition to reference that image digest/tag and deploy the new service revision. Avoid relying only on mutable latest tags.
10When would you choose Amazon ECS with Fargate instead of EC2-based application hosting?
Choose Fargate when you want ECS to manage server capacity and prefer per-task compute without patching EC2 hosts. Choose EC2 launch type when you need host-level control, specialized hardware, custom agents or potentially better economics at steady high utilization. Compare operational overhead, capacity and pricing.
2. AWS Lambda and Background Processing
11How do you create and deploy an AWS Lambda function using C# and .NET?
Create a .NET Lambda project using the AWS Lambda .NET tooling/template, select a supported .NET runtime, implement the handler and package it. Create an execution role with minimum permissions, deploy through SAM/CDK/CloudFormation or tooling, configure memory, timeout, environment and triggers, then test and inspect CloudWatch logs.
12How do you implement dependency injection in a .NET AWS Lambda application?
Use the .NET Generic Host where suitable, register services in the DI container during startup and resolve them through the Lambda function’s host/service provider. Reuse the host/container across warm invocations when safe, but avoid storing request-specific mutable state in singleton services. Dispose resources appropriately.
13How do you expose an AWS Lambda function through Amazon API Gateway?
Create an API Gateway HTTP API or REST API, add a Lambda integration and routes, configure invoke permissions, authorization and CORS, then deploy a stage. The Lambda receives the request event and returns status code, headers and body in the integration’s expected format. Test auth, errors and throttling.
14How do you implement scheduled background jobs using Amazon EventBridge Scheduler and AWS Lambda?
Create an EventBridge Scheduler schedule with a cron or rate expression, select Lambda as target and assign a role that can invoke it. Configure retry policy and a dead-letter queue where appropriate. Make the handler idempotent because retries or overlapping executions can occur.
15How do you process files automatically when they are uploaded to an Amazon S3 bucket?
Configure an S3 ObjectCreated event notification to invoke Lambda directly or publish to SQS/EventBridge for more control. Grant the bucket permission to invoke the function, filter by prefix/suffix to avoid recursive triggers, and make processing idempotent. For large files, pass bucket/key rather than copying the object into the event.
16What is AWS Step Functions, and when would you use it to orchestrate multiple Lambda functions?
Step Functions orchestrates multi-step workflows using state machines. It can invoke Lambda or AWS service integrations, branch, wait, retry, catch errors and maintain execution history. Use it when a process has multiple stages, compensation, human approval or complex retry logic.
17How do you handle exceptions and retries in AWS Lambda?
Handle expected errors explicitly, emit structured logs and return failures when the event source should retry. Configure retry behavior according to the trigger: asynchronous invocation has retry settings, while SQS event source mappings use visibility timeout, batch behavior and DLQ/redrive policies. Make processing idempotent and bound retries.
18How do you manage environment variables and secrets in an AWS Lambda application?
Use environment variables for non-secret configuration and Secrets Manager or Parameter Store for sensitive values. Grant the Lambda execution role permission to read only required parameters/secrets, retrieve them at startup or with controlled caching, and encrypt using KMS. Never package credentials in the deployment artifact.
3. Messaging and Event-Driven Architecture
19What is Amazon SQS, and how does it help build loosely coupled .NET applications?
SQS is a managed message queue that buffers work between producers and consumers. A .NET API can enqueue a job and return quickly; worker services or Lambda consumers process it independently. This decouples components, smooths traffic spikes and allows retry and failure isolation.
20What is the difference between Amazon SQS Standard and FIFO queues?
Standard queues provide very high throughput and at-least-once delivery with best-effort ordering. FIFO queues preserve order within message groups and support deduplication features, with throughput constraints depending on configuration. Choose FIFO only when ordering/deduplication requirements justify it.
21What is the difference between Amazon SQS and Amazon SNS?
SQS is point-to-point queueing: consumers poll messages and each message is processed by a consumer. SNS is pub/sub: publishers send a message to a topic that fans out to multiple subscribers. SNS can deliver to SQS queues to combine fan-out with durable independent consumption.
22How do you publish and consume messages from Amazon SQS using C# and the AWS SDK for .NET?
Use AWSSDK.SQS and an IAM task/instance role. Call SendMessageAsync to publish; consumers call ReceiveMessageAsync, process the payload, then DeleteMessageAsync only after successful processing. Configure long polling, visibility timeout, batch size, cancellation and a DLQ; design for duplicate delivery.
23What is a Dead-Letter Queue (DLQ) in Amazon SQS, and how do you process failed messages?
A DLQ stores messages that repeatedly fail processing after the source queue’s maxReceiveCount. Configure a redrive policy, inspect message attributes and failure logs, fix the cause, then replay safely or discard according to business rules. Protect against poison-message loops and duplicate side effects.
24What is message visibility timeout in Amazon SQS, and how does it work?
When a consumer receives a message, it becomes invisible to other consumers for the visibility timeout. If processing succeeds, the consumer deletes it; if not deleted before timeout, it becomes visible again. Set timeout longer than normal processing and extend it for long jobs where needed.
25What is Amazon EventBridge, and how is it different from Amazon SNS?
EventBridge routes events from AWS services, custom applications and SaaS sources using event buses and rules with content-based filtering. SNS primarily distributes published messages to subscribers. Use EventBridge for event routing and filtering across services; use SNS for straightforward topic fan-out and notifications.
26What is Amazon Kinesis, and when would you use it instead of SQS?
Kinesis Data Streams is for ordered, partitioned, high-volume streaming data with multiple consumers and replay within retention. SQS is a work queue where messages are typically removed after processing. Choose Kinesis for continuous event streams/analytics and SQS for decoupled task processing.
4. Authentication and Security
27How do you implement user authentication in an ASP.NET Core Web API using Amazon Cognito?
Configure a Cognito User Pool and app client, use OIDC authorization code with PKCE for Angular, and obtain tokens. Configure ASP.NET Core JWT bearer authentication with the User Pool issuer and correct audience/client validation; use claims and policies for authorization. Validate token scopes/claims and enforce authorization server-side.
28What is the difference between OAuth 2.0 and OpenID Connect, and how does Amazon Cognito support them?
OAuth 2.0 is an authorization framework for delegated access; OpenID Connect adds an identity layer and ID tokens on top of OAuth 2.0. Cognito User Pools support OAuth/OIDC flows and issue tokens. Use access tokens to authorize API calls and ID tokens for client identity information, validating each for its intended purpose.
29How do you secure an Angular application and ASP.NET Core Web API using Amazon Cognito?
Angular authenticates users with Cognito using authorization code plus PKCE and stores tokens using a secure library strategy. It sends an access token over HTTPS to the API; ASP.NET Core validates signature, issuer, expiry and audience, then applies role/scope policies. Never rely on frontend route guards as API security.
30How can an ASP.NET Core application access AWS Secrets Manager without storing AWS credentials in configuration files?
Attach an IAM role to the compute environment (EC2 instance profile, ECS task role or Lambda execution role) and let the AWS SDK credential provider chain obtain temporary credentials. Grant only Secrets Manager read permissions for the required secret and optionally use VPC endpoints. No static AWS keys are needed in appsettings.
31How do you assign an IAM Role to an EC2 instance so that an ASP.NET Core application can access S3?
Create an IAM role with a trust policy for EC2, attach a narrowly scoped S3 policy, and associate it with the instance through an instance profile. The .NET SDK on the instance obtains temporary credentials automatically. Restrict the policy to the required bucket and object prefix.
32What is the difference between authentication and authorization in an AWS-hosted application?
Authentication verifies who the user or workload is, typically through Cognito/OIDC or IAM role credentials. Authorization decides what the authenticated identity may do, using API policies, claims and IAM permissions. Both layers are required; successful login does not grant unrestricted access.
33How do you secure Amazon S3 objects using IAM policies, bucket policies and pre-signed URLs?
Use IAM identity policies for workload access and bucket policies for resource-level controls; block public access unless explicitly required. Keep objects private and issue short-lived pre-signed URLs for narrowly scoped upload/download operations. Enable encryption, logging and conditions such as TLS and expected VPC endpoint where suitable.
34What is AWS Systems Manager Parameter Store, and how can you manage application configuration using it?
Parameter Store stores hierarchical configuration values and secure strings, supports versioning and integrates with IAM/KMS. A .NET app can retrieve parameters using the AWS SDK or configuration provider, using a workload role and path-scoped permissions. Use Secrets Manager when managed secret rotation is required.
5. Database and Storage Integration
35How do you configure Entity Framework Core to work with Amazon RDS for SQL Server?
Install Microsoft.EntityFrameworkCore.SqlServer, configure UseSqlServer with the RDS endpoint and database, and use secure credentials from Secrets Manager or supported integrated authentication. Place the API and RDS in connected VPC subnets, allow port 1433 only from the API security group, enable TLS and configure retry strategy carefully.
36What is database connection pooling, and why is it important when connecting .NET applications to Amazon RDS?
Connection pooling reuses established database connections rather than opening a new physical connection for every request. It reduces latency and connection overhead, but each app instance can create its own pool. Limit Max Pool Size, dispose connections/DbContexts promptly and account for total instances so RDS is not overwhelmed.
37How do you handle transient database failures in an ASP.NET Core application connected to Amazon RDS?
Use EF Core SQL Server retry execution strategy or a carefully configured Polly policy for transient network/failover errors, with bounded exponential backoff and jitter. Retry only transient operations and avoid blindly retrying non-idempotent transactions. Log failures and set command/connection timeouts appropriately.
38What is an Amazon RDS Security Group, and how does it control database connectivity?
An RDS security group is a virtual firewall controlling network traffic to the database. Configure inbound SQL Server port 1433 from the application’s security group, not from the whole internet. Ensure subnet routes, NACLs and DNS also permit the connection.
39How do you implement secure file upload and download functionality using Amazon S3 in ASP.NET Core?
Validate file size, extension and content, generate a safe object key, and upload through the API or preferably a short-lived pre-signed URL. Keep the bucket private, use encryption and least-privilege IAM, scan files where needed, store metadata in the database and authorize downloads before issuing pre-signed URLs.
40How do you generate pre-signed S3 URLs for uploading and downloading private files?
Use the AWS SDK’s GetPreSignedURL request with bucket, object key, HTTP verb and expiration to generate a signed URL. For upload, sign a PUT operation and optionally constrain headers/content type; for download, sign GET. Return it only after application authorization and keep expiry short.
6. CI/CD, Monitoring and Performance
41How do you create a CI/CD pipeline for an ASP.NET Core application using AWS CodePipeline and CodeBuild?
Connect a source provider to CodePipeline, use CodeBuild to restore, build, test and publish the .NET artifact, then deploy to ECS, Elastic Beanstalk, EC2 or Lambda. Store artifacts in S3, use IAM roles, environment approvals and automated health checks. Promote immutable artifacts across environments.
42What is the difference between a managed CodeBuild environment and a self-hosted build runner?
A managed CodeBuild environment is provisioned and maintained by AWS for each build; a self-hosted runner is infrastructure you operate and patch. Managed builds reduce maintenance and integrate with IAM, while self-hosted runners may be needed for custom tools, private network access or specialized hardware. Secure both and isolate build credentials.
43How do you deploy an Angular application and ASP.NET Core API using separate CI/CD pipelines on AWS?
Use separate pipelines or stages: build Angular and publish static files to S3/CloudFront; build the API container or package and deploy to ECS/App Runner/EC2. Configure environment-specific API endpoints, cache invalidation/versioned assets, CORS, secrets, approvals and independent rollback paths.
44How do you configure Amazon CloudWatch to monitor application logs, metrics and exceptions?
Send structured logs to CloudWatch Logs using an agent, container log driver or application integration; collect platform metrics and custom business metrics. Add correlation/trace IDs, capture exceptions with stack traces, create dashboards and alarms for error rate, latency and saturation. Set retention and redact sensitive data.
45What is distributed tracing, and how can you trace requests across multiple .NET microservices using AWS X-Ray?
Distributed tracing follows a request across services and dependencies using trace and span identifiers. Instrument .NET services with OpenTelemetry/AWS X-Ray SDK integrations, propagate trace context over HTTP and messaging, and send segments to X-Ray. Use service maps and traces to identify slow downstream calls.
46How do you troubleshoot high response times in an ASP.NET Core API hosted on Amazon EC2 or ECS?
Break latency down by API endpoint, dependency and infrastructure metrics. Check ALB target response time, CPU/memory, thread pool, GC, database waits, connection pools, network and external calls. Use CloudWatch, X-Ray and logs with correlation IDs, profile under load, then optimize the measured bottleneck.
47What is CloudWatch Logs Insights, and how can you query application logs?
CloudWatch Logs Insights is a query interface for searching and aggregating log events. Select log groups, filter by timestamp/request ID/error, parse structured JSON fields and use stats/group-by to find frequent errors or slow requests. Save useful queries and use them during incident response.
48How do you configure CloudWatch Alarms when CPU utilization, API errors or application latency exceeds a threshold?
Publish or select metrics such as EC2 CPU, ALB 5xx/latency, ECS task health and custom API error/latency metrics. Create CloudWatch alarms with thresholds, evaluation periods and missing-data behavior, then route notifications through SNS or incident integrations. Use percentiles and sustained thresholds to reduce alert noise.
49How do you configure Amazon API Gateway to implement API throttling, authentication and request validation?
Configure API Gateway throttling at account, stage, route or usage-plan level as appropriate; attach Cognito/JWT/IAM/Lambda authorizers; define request/response models or validation where supported; and enable access logs. Also enforce authorization and validation in the ASP.NET Core service because gateway controls are not a substitute.
50How do you implement distributed caching in an ASP.NET Core application using Amazon ElastiCache for Redis?
Use ElastiCache for Redis as a shared cache and configure ASP.NET Core IDistributedCache or StackExchange.Redis. Choose cluster/network/security settings, use TLS and authentication, define keys and TTLs, cache-aside frequently read data, invalidate on writes and implement graceful fallback if Redis is unavailable.
1. AWS Cloud Architecture
01How would you design a scalable, highly available architecture for an Angular application with ASP.NET Core Web APIs and Amazon RDS?
Serve Angular static assets through CloudFront with S3 as origin; route API traffic through an ALB or API Gateway to stateless ASP.NET Core services across at least two AZs on ECS/App Runner/EC2. Put RDS in private subnets with Multi-AZ where required, add Redis and SQS for cache/background work, and use IAM roles, Secrets Manager, WAF, autoscaling, health checks and CloudWatch/X-Ray. Define RTO/RPO and test failure scenarios.
02How would you decide between Amazon EC2, AWS Elastic Beanstalk, Amazon ECS, Amazon EKS and AWS App Runner for a production .NET application?
EC2 gives maximum OS/runtime control but requires host operations; Elastic Beanstalk simplifies common application deployments while exposing underlying resources; ECS provides container orchestration with EC2 or Fargate; EKS is Kubernetes for teams needing Kubernetes APIs/ecosystem; App Runner is simplest for supported web services and containers. Compare control, operational skills, networking, scaling, portability and cost.
03How would you migrate a large monolithic ASP.NET Core application to AWS microservices architecture?
First map bounded contexts and dependencies, then extract a low-risk capability behind a stable API or event contract. Use the strangler pattern to route selected traffic to new services, containerize and deploy incrementally, and introduce observability, service ownership and independent data boundaries. Avoid a big-bang rewrite; plan data migration, consistency, rollback and security.
04How would you design communication between multiple .NET microservices using synchronous and asynchronous communication on AWS?
Use synchronous HTTP/gRPC for immediate request-response needs and asynchronous messaging for decoupled workflows. In AWS, use ALB/API Gateway for HTTP, SQS for work queues, SNS/EventBridge for fan-out/routing and Step Functions for orchestration. Set timeouts, retries, idempotency, versioned contracts and tracing.
05How would you implement the API Gateway pattern using Amazon API Gateway and AWS services?
Place API Gateway in front of backend integrations or services to provide a managed API entry point. Define routes, stages, authorization, throttling, validation, logging and integrations to Lambda or HTTP backends. For internal service-to-service routing, evaluate whether an ALB or service mesh is more appropriate.
06What is the difference between Amazon CloudFront, Application Load Balancer and Amazon API Gateway? When would you use each?
CloudFront is an edge CDN for caching and global delivery; ALB distributes Layer 7 HTTP traffic among targets inside a Region; API Gateway manages API endpoints, authorization, throttling and API lifecycle. Use CloudFront for frontend/static acceleration, ALB for application load balancing and API Gateway for managed API exposure.
07How would you design a multi-region AWS architecture to support disaster recovery and high availability?
Deploy the application in a second Region with replicated artifacts, infrastructure-as-code, network connectivity, secrets/configuration and data replication or backups. Choose active-active or pilot-light/warm-standby based on RTO/RPO and cost; use Route 53 or Global Accelerator for failover. Regularly test failover and data consistency.
08What are Availability Zones, and how would you use them to improve application availability?
Availability Zones are isolated locations within a Region. Place load balancers and compute across multiple AZs, use an AZ-resilient database deployment and avoid single-AZ dependencies. Configure health checks and Auto Scaling so traffic shifts away from failed targets; test AZ failure assumptions.
09How would you implement load balancing for a distributed .NET application hosted across multiple EC2 instances or ECS tasks?
Register EC2 instances or ECS tasks in an ALB target group and distribute traffic across healthy targets. Run targets in multiple AZs, use health checks and Auto Scaling/service desired count, and ensure sessions and files are externalized. For global routing, add CloudFront, Route 53 or Global Accelerator as appropriate.
10How would you implement Circuit Breaker, Retry and Bulkhead patterns in an AWS-based .NET microservices application?
Retry handles transient failures with bounded exponential backoff and jitter; Circuit Breaker temporarily stops calls to a failing dependency; Bulkhead limits concurrent work/resources so one failure cannot exhaust the whole service. In .NET, use Polly/resilience handlers, set timeouts and avoid retrying non-idempotent operations without protection.
2. Advanced Security and Identity
11How would you implement end-to-end authentication and authorization across Angular, ASP.NET Core APIs and downstream AWS services?
Use Cognito or another OIDC provider for end-user login; Angular obtains tokens using authorization code with PKCE and sends access tokens to ASP.NET Core APIs. APIs validate JWTs and enforce scopes/roles/policies. For AWS service access, use IAM task/instance/Lambda roles and temporary credentials, keeping user identity separate from workload permissions.
12How would you implement IAM Role-based authentication between AWS services without storing long-lived access keys?
Configure trust relationships and attach least-privilege IAM roles to EC2, ECS tasks, Lambda or other supported workloads. The AWS SDK obtains temporary credentials from the environment’s role provider; use role assumption for cross-account access. Rotate/revoke by changing role policies rather than distributing static keys.
13What is the difference between an IAM Role, an IAM User and an IAM Identity Center user?
An IAM user is a long-lived IAM identity, an IAM role is assumed and provides temporary credentials, and an IAM Identity Center user is a workforce identity centrally managed for access to multiple AWS accounts and applications. Prefer Identity Center for human workforce access and roles for workloads.
14How would you securely connect an ASP.NET Core application running on ECS or EC2 to Amazon RDS using IAM database authentication where supported?
IAM database authentication is supported only for certain RDS engines/configurations; it is not generally available for RDS for SQL Server. Verify engine support first. Where supported, grant database IAM authentication, generate a short-lived auth token and connect over TLS; otherwise use Secrets Manager-managed credentials and network restrictions.
15How would you implement automatic secret rotation using AWS Secrets Manager without manually changing application configuration?
Configure Secrets Manager rotation with a rotation Lambda or supported managed rotation workflow, granting narrowly scoped permissions to update the secret and database/service. Applications should retrieve and cache secrets for a bounded period, refresh on authentication failure and reconnect safely. Ensure rotation is tested with connection pools and rollback procedures.
16How would you secure an AWS-hosted application using VPC endpoints, private subnets, Security Groups and network restrictions?
Place workloads in private subnets, expose only ALB/API entry points, use security groups with least-privilege rules and VPC endpoints for private access to supported AWS services. Use route tables, NACLs, DNS and endpoint policies carefully; monitor flow logs and avoid broad egress/inbound rules.
17How would you implement least-privilege IAM policies across development, testing and production environments?
Create separate roles per workload and environment, scope actions to required resources and use conditions/tags where appropriate. Avoid wildcard actions/resources, use permission boundaries or Organizations SCPs as guardrails, review Access Analyzer findings and CloudTrail activity, and test policies before production rollout.
18How would you protect an AWS-hosted Web API against DDoS attacks, SQL injection, token theft and unauthorized API access?
Use WAF managed/custom rules and rate limits, Shield protections where appropriate, TLS, secure headers and strict CORS. Validate inputs and use parameterized SQL/EF Core, validate JWT issuer/audience/signature, rotate secrets, enforce least-privilege IAM and patch dependencies. Add monitoring, abuse detection and DDoS-resistant architecture; no single control prevents every attack.
19What is AWS WAF, and how would you configure it to protect an Angular and ASP.NET Core application?
AWS WAF is a web application firewall that filters HTTP(S) requests using managed and custom rules. Associate it with CloudFront or supported regional resources such as ALB/API Gateway, enable managed rule groups, rate-based rules and logging, then tune in count mode before blocking. Protect both frontend delivery and API entry points.
20How would you implement secure cross-origin communication between an Angular frontend and an AWS-hosted ASP.NET Core API?
Configure Angular to call the API over HTTPS and allow only the exact frontend origins, methods and headers in ASP.NET Core CORS policy. Handle preflight requests, avoid wildcard origins with credentials, and use secure token handling. CORS is a browser policy, not authentication; the API must still validate tokens and authorize every operation.
3. Distributed Systems and Messaging
21How would you design an event-driven architecture using Amazon SQS, SNS, EventBridge and AWS Lambda?
Publish domain events to EventBridge for routing/filtering, SNS for topic fan-out, SQS for durable consumer queues and Lambda for event processing. Use event schemas, DLQs, retries, idempotency and tracing. Keep producers independent of consumers and choose each service based on routing, delivery and ordering requirements.
22How would you guarantee reliable message processing when a .NET consumer fails after receiving an SQS message?
With SQS, messages may be delivered more than once. A consumer should process idempotently, delete only after successful durable completion, use a visibility timeout longer than normal processing, and configure a DLQ. If a consumer crashes before deletion, the message becomes visible again and can be retried.
23Explain the difference between at-most-once, at-least-once and effectively-once message processing in AWS.
At-most-once means a message is delivered zero or one time, so loss is possible; at-least-once means it is retried until acknowledged but duplicates can occur; effectively-once is achieved through idempotency/deduplication and transactional state handling. Distributed systems rarely provide universal exactly-once side effects across independent services.
24How would you implement idempotency in a .NET microservice that consumes duplicate messages from Amazon SQS?
Give each business operation a stable idempotency key or message identifier and store processing status in a durable database with a uniqueness constraint. In one transaction, detect duplicates and commit the business change plus completion marker; acknowledge the SQS message only after commit. Make external side effects idempotent too.
25How would you implement the Transactional Outbox pattern using EF Core, Amazon RDS and Amazon SQS?
Write the business update and an outbox event record in the same RDS transaction using EF Core. A background publisher reads unpublished outbox rows and sends them to SQS, then marks them published. Use a unique event ID, retries and idempotent consumers because a send may succeed even if the publisher crashes before marking the row.
26What is the Saga pattern, and how would you implement distributed transactions across multiple .NET microservices using AWS Step Functions?
Saga coordinates a business transaction across services using local transactions and compensating actions rather than one distributed database transaction. Step Functions can orchestrate steps, retries, catches and compensation tasks; persist workflow state, make each step idempotent and define what happens when compensation also fails.
27How would you handle message ordering and concurrent message processing using Amazon SQS FIFO queues?
SQS FIFO uses MessageGroupId to preserve ordering within a group while allowing parallel processing across groups. Use MessageDeduplicationId or content-based deduplication to suppress duplicate sends within the deduplication interval. Avoid one global message group if throughput is important; configure consumers for group-level concurrency.
28How would you design a system that processes millions of events using Amazon Kinesis Data Streams?
Partition the stream using keys that distribute traffic, choose shard capacity/On-Demand mode based on throughput, and use producers with retries and sequence handling. Consumers use enhanced fan-out or shared throughput, checkpoint progress and handle retries/duplicates. Monitor iterator age, throttling and shard utilization; size retention and replay needs.
29How would you handle poison messages and repeated failures in an event-driven AWS application?
Configure a redrive policy and DLQ, capture exception details and message metadata, and alert on DLQ depth. Classify transient versus permanent errors, fix the cause and replay in a controlled way with idempotency. Do not endlessly retry malformed messages or silently discard business-critical events.
30How would you implement background processing for long-running workflows using AWS Step Functions and Lambda?
Model the workflow as a Step Functions state machine with Lambda or service integrations for each step. Use wait states, retries, catches, parallel/map states and compensation as required. For long-running work, persist state in the workflow rather than keeping a process alive; make steps idempotent and monitor execution failures.
4. Performance, Scalability and Database Architecture
31How would you optimize an ASP.NET Core API hosted on EC2 or ECS when CPU and memory utilization remain consistently high?
Start with CloudWatch metrics and profiling to determine whether CPU, memory, GC, thread pool, network or application code is responsible. Review container/instance sizing, hot paths, synchronous blocking, excessive allocations and query latency; use dotnet-counters/profilers and load tests. Scale horizontally only after checking downstream capacity.
32How would you identify and resolve database connection pool exhaustion in an ASP.NET Core application connected to Amazon RDS?
Inspect pool metrics, active connections, request concurrency and DbContext/connection lifetimes. Ensure connections are disposed promptly, avoid holding them during slow work, bound parallel database operations and tune pool size relative to total app instances and RDS limits. Investigate long transactions and leaks; add backpressure if needed.
33How would you optimize Amazon RDS for SQL Server performance when multiple APIs execute heavy queries simultaneously?
Use Performance Insights/Enhanced Monitoring and SQL Server execution plans to find expensive queries, waits, blocking and missing/inefficient indexes. Optimize queries and indexes, parameterize SQL, avoid N+1 EF queries, tune connection concurrency and consider instance/storage sizing. Separate read workloads or cache only when consistency requirements permit.
34What is the difference between Amazon RDS Multi-AZ deployments and Read Replicas, and when would you use each?
Multi-AZ provides a standby/failover deployment for availability and disaster resilience; it is not primarily a read-scaling feature. Read Replicas serve read traffic and can be promoted in supported scenarios, but replication lag and engine support apply. Choose Multi-AZ for availability and replicas for read scaling or recovery design.
35How would you implement distributed caching across multiple ASP.NET Core application instances using Amazon ElastiCache for Redis?
Use Redis as a shared cache for multiple API instances, with consistent serialization, key naming, TTLs and cache-aside patterns. Invalidate or update entries after writes, partition keys by tenant where needed, protect Redis with private networking/TLS/auth and configure timeouts/fallback. Monitor hit ratio, latency and memory eviction.
36How would you prevent cache stampede and stale cache problems in a distributed .NET application?
Prevent stampedes with per-key single-flight locks, request coalescing, randomized TTLs and background refresh. Reduce stale values through explicit invalidation, versioned keys, short TTLs for volatile data and event-driven updates. Define behavior during cache outages and avoid making Redis the only source of truth.
37How would you design an application that handles sudden traffic spikes without overloading Amazon RDS?
Use stateless autoscaled API instances, ALB, request rate limiting and SQS to buffer asynchronous work. Add Redis caching, efficient queries, connection limits and bounded worker concurrency; scale workers separately from APIs. Load test burst patterns and apply backpressure or graceful degradation when the database nears capacity.
38How would you implement read scaling for an application with heavy read workloads using Amazon RDS read replicas where supported?
Where the selected RDS engine supports read replicas, route read-only workloads to replicas and keep writes on the primary. Account for replication lag and eventual consistency; use primary reads for read-after-write requirements. For RDS SQL Server, verify the supported read replica/availability options for the edition and deployment rather than assuming all engines behave alike.
39How would you design file storage for a system that processes millions of documents using Amazon S3?
Store documents as S3 objects with key prefixes designed for access patterns, enable versioning/lifecycle policies and choose storage classes by retrieval needs. Use multipart upload for large files, event-driven processing via SQS/Lambda, metadata/indexes in a database and CloudFront or pre-signed URLs for access. Apply encryption, retention and least-privilege policies.
40How would you optimize an AWS-hosted application to reduce infrastructure costs without affecting performance and availability?
Measure cost by service and workload, then right-size EC2/ECS, use autoscaling and savings options for predictable usage, and move infrequently accessed objects to suitable S3 tiers. Reduce unnecessary logs/data transfer, optimize RDS and caching, schedule non-production shutdowns and use budgets. Validate every saving against latency, availability, backup and security objectives.
5. DevOps, Infrastructure and Production Troubleshooting
41How would you implement Infrastructure as Code using AWS CloudFormation or AWS CDK to provision an entire .NET application environment?
Define resources in CloudFormation templates or CDK constructs, parameterize environments and split reusable components into stacks/modules. Store templates in Git, validate with lint/security checks, deploy through CI using least-privilege roles and change sets, and protect state/secrets. Make deployments repeatable and avoid manual drift.
42How would you design a multi-stage CI/CD pipeline with automated testing, security scanning, approvals and production deployment using AWS tools?
Use stages for source, build/test, static analysis and dependency/container scanning, artifact publication, staging deployment, integration/smoke tests, approval and production rollout. Use CodePipeline with CodeBuild and CodeDeploy/ECS blue-green as appropriate, store secrets in Secrets Manager, use immutable artifacts and automate rollback/health verification.
43How would you implement Blue-Green deployment and Canary releases for ASP.NET Core applications using Amazon ECS or AWS CodeDeploy?
Blue-Green deploys a complete new task set/environment and shifts traffic after validation; Canary gradually sends a small percentage of traffic to the new version. ECS deployment strategies and CodeDeploy can manage traffic shifting with ALB target groups. Monitor errors/latency during rollout and keep database changes backward-compatible.
44How would you implement centralized logging and distributed tracing across multiple .NET microservices using Amazon CloudWatch and AWS X-Ray?
Emit structured JSON logs with correlation IDs and trace context from all services, collect them in CloudWatch Logs and instrument requests/dependencies with OpenTelemetry/X-Ray. Propagate trace context across HTTP and SQS, build service maps/dashboards and create actionable alarms. Redact sensitive fields and set retention/access controls.
45How would you troubleshoot intermittent HTTP 502, 503 and 504 errors in an ASP.NET Core API hosted on AWS?
Correlate 502/503/504 timestamps and request IDs across ALB access logs, target health, ECS/EC2 logs and CloudWatch/X-Ray traces. Check app crashes/startup, health checks, CPU/memory, port bindings, TLS/DNS, idle/request timeouts and database/downstream latency. Reproduce, isolate the failing hop and fix root cause before merely increasing timeouts.
46How would you diagnose memory leaks and application crashes in an ASP.NET Core application running on EC2, ECS or AWS App Runner?
Monitor memory/CPU, container restarts and process exit codes; collect .NET dumps or profiling data in a controlled environment. Inspect heap growth, retained objects, unbounded caches, event subscriptions, large allocations and undisposed streams/DbContexts. Reproduce under load, analyze with dotnet-dump/diagnostic tools, fix and verify stable memory after deployment.
47How would you implement disaster recovery for Amazon RDS, S3 and EC2/ECS-hosted applications?
Define RPO/RTO and choose RDS automated backups, point-in-time restore, Multi-AZ and cross-Region snapshots/replicas as appropriate. Enable S3 versioning, replication, Object Lock or lifecycle protections according to requirements. Recreate compute from IaC/AMI/container images in a secondary AZ/Region, replicate configuration/secrets and test recovery regularly.
48How would you design monitoring dashboards and alerting strategies for a production AWS application using CloudWatch, X-Ray and CloudTrail?
Build dashboards for availability, request volume, p95/p99 latency, 4xx/5xx, target health, CPU/memory, database connections/latency, queue depth and business metrics. Use CloudWatch alarms and SNS/on-call integrations, X-Ray traces for diagnosis and CloudTrail for audit events. Set SLO-based thresholds, runbooks, ownership and noise controls.
49How would you implement centralized configuration management and environment-specific settings using AWS Systems Manager Parameter Store and Secrets Manager?
Store non-secret settings in Parameter Store with path conventions and environment prefixes; store credentials and sensitive tokens in Secrets Manager. Grant workload roles scoped read access, use KMS encryption, cache with bounded refresh and separate values by environment/account. Version changes, audit access and test rotation/config refresh behavior.
50You have an Angular frontend, multiple ASP.NET Core APIs, SQL Server, background workers and 2,000 concurrent users. How would you design, deploy, secure, monitor and scale the entire solution on AWS?
Design Angular as static assets behind CloudFront/WAF and Route 53; expose ASP.NET Core APIs through ALB/API Gateway to stateless ECS Fargate services across multiple AZs. Use RDS SQL Server in private subnets with Multi-AZ and tuned connection limits, Redis for cache, SQS/EventBridge for background work, and S3 for documents. Secure users with Cognito/OIDC, workloads with IAM roles, secrets with Secrets Manager, and networks with private subnets/security groups. Build CI/CD and IaC pipelines with automated tests/scanning, monitor via CloudWatch/X-Ray, load-test beyond 2,000 concurrent users, and define autoscaling, rate limits, DR RPO/RTO, rollback and cost controls.