CS Techie
Comments
courseAWS

AWS certification-style practice questions — SQS, SNS, Kinesis, Amazon MQ

SET 1

  1. You need an AWS service to decouple producers and consumers where occasional duplicate processing is acceptable and you need virtually unlimited throughput. Which queue type should you choose?
    A. SQS FIFO
    B. SQS Standard
    C. Amazon MQ (ActiveMQ)
    D. Kinesis Data Streams
    Correct: B
    Explanation: SQS Standard offers at-least-once delivery, best-effort ordering, and virtually unlimited throughput, making it suitable when duplicates are acceptable.

  2. Which SQS feature prevents a message from being processed by multiple consumers while one consumer is handling it?
    A. Dead-letter queue
    B. Visibility timeout
    C. Long polling
    D. Message deduplication
    Correct: B
    Explanation: Visibility timeout hides a received message for a specified interval to avoid concurrent processing.

  3. You must preserve strict, per-key ordering and prevent duplicate processing. Which queueing option is best?
    A. SQS Standard
    B. SNS topic
    C. SQS FIFO with MessageGroupId
    D. Amazon S3 Events
    Correct: C
    Explanation: SQS FIFO supports message groups (MessageGroupId) to maintain ordering per group and supports deduplication.

  4. Which SQS setting reduces empty ReceiveMessage API responses and lowers costs by waiting up to N seconds for messages?
    A. VisibilityTimeout
    B. MessageRetentionPeriod
    C. ReceiveMessageWaitTimeSeconds (long polling)
    D. DelaySeconds
    Correct: C
    Explanation: Long polling waits up to the configured seconds for messages, reducing empty responses and API calls.

  5. You have an SQS queue that receives occasional poison messages that always fail processing and block retries. Which combination is recommended to handle them?
    A. Increase VisibilityTimeout only
    B. Use a Dead-Letter Queue (DLQ) and set maxReceiveCount
    C. Switch to SQS Standard from FIFO
    D. Use S3 for message storage
    Correct: B
    Explanation: A DLQ with a maxReceiveCount isolates repeatedly failing messages for analysis without blocking processing.

  1. Which service provides a managed publish/subscribe mechanism that can fan out a single message to multiple subscribers (SQS, Lambda, HTTP endpoints)?
    A. Amazon Kinesis Data Streams
    B. Amazon SNS
    C. AWS Step Functions
    D. Amazon MQ
    Correct: B
    Explanation: SNS is a pub/sub service that can push messages to multiple endpoints including SQS, Lambda, and HTTP/S.

  2. You need to send notifications to mobile devices, SMS, and multiple microservices simultaneously. Which AWS service is the simplest fit?
    A. SQS
    B. SES
    C. SNS
    D. Kinesis Firehose
    Correct: C
    Explanation: SNS supports multiple protocols (push to mobile, SMS, HTTP endpoints) and fan-out.

  3. For guaranteeing message ordering when publishing to multiple subscribers, which statement is true?
    A. SNS ensures global ordering to all subscribers by default.
    B. SNS FIFO topics can be used to preserve ordering for FIFO-compatible subscribers.
    C. SNS always uses SQS FIFO under the hood.
    D. Ordering is guaranteed only when using S3 as an origin.
    Correct: B
    Explanation: SNS FIFO topics (where applicable) support ordering semantics; standard SNS topics do not guarantee order.

  4. Which SNS feature helps handle failed deliveries to HTTP/S endpoints by storing undelivered messages for analysis?
    A. Message attributes
    B. Delivery status logging and retry policies with DLQs
    C. Topic policies
    D. Cross-account topics
    Correct: B
    Explanation: SNS supports retries and delivery status; combined with a DLQ (when enabled) you can capture undelivered messages for later analysis.

  1. You must ingest high-throughput, ordered clickstream data from multiple producers and process it with custom stream consumers. Which service is the best fit?
    A. Kinesis Data Streams
    B. Kinesis Data Firehose
    C. SQS Standard
    D. SNS
    Correct: A
    Explanation: Kinesis Data Streams supports high-throughput ingestion, ordering per shard, and multiple custom consumers for stream processing.

  2. Which Kinesis service is fully managed for delivering streaming data to S3/Redshift/Elasticsearch and supports optional on-delivery transformations (Lambda) without you building consumers?
    A. Kinesis Data Streams
    B. Kinesis Data Analytics
    C. Kinesis Data Firehose
    D. Amazon MSK
    Correct: C
    Explanation: Firehose is a managed delivery service that writes to destinations (S3, Redshift, OpenSearch, Splunk) and can optionally transform records using Lambda.

  3. In Kinesis Data Streams, what does a “shard” represent?
    A. A storage bucket in S3
    B. A unit of capacity providing a read and write throughput limit and ordering domain
    C. An IAM role
    D. A CloudWatch alarm
    Correct: B
    Explanation: A shard is a throughput/ordering unit in Kinesis Data Streams. It sets write and read capacity and preserves ordering for records within the shard.

  4. Your application requires sub-second analytics using SQL on streaming data with windowed aggregations. Which Kinesis service should you consider?
    A. Kinesis Data Streams only
    B. Kinesis Data Analytics
    C. Kinesis Data Firehose
    D. Amazon Athena
    Correct: B
    Explanation: Kinesis Data Analytics runs SQL (or Flink-based) queries on streaming data for near-real-time analytics and aggregations.

  5. Which guarantee best describes default delivery semantics of Kinesis Data Streams to consumers?
    A. Exactly-once across all consumers
    B. At-most-once delivery
    C. At-least-once delivery (consumers may see duplicates)
    D. Fire-and-forget best-effort
    Correct: C
    Explanation: Kinesis Data Streams provides at-least-once delivery semantics; consumers must handle possible duplicates.

  6. You want an easy way to archive all raw streaming records before transformation and also later reprocess them. Which architecture is recommended with Kinesis?
    A. Use only SQS to store raw records
    B. Use Kinesis Data Firehose to write to S3 as a durable landing zone (optionally also deliver to analytics)
    C. Use SNS to fan-out to S3
    D. Use Amazon MQ for archival
    Correct: B
    Explanation: Firehose can persist streaming data to S3 for durable archival and downstream reprocessing.

  1. You must migrate a JMS-based on-premises application to AWS without changing application code that uses AMQP/JMS protocols. Which service fits best?
    A. SQS
    B. Amazon MQ (managed ActiveMQ/RabbitMQ)
    C. SNS
    D. Kinesis Data Streams
    Correct: B
    Explanation: Amazon MQ provides managed ActiveMQ or RabbitMQ brokers and supports JMS/AMQP/STOMP, enabling lift-and-shift with minimal code changes.

  2. How does Amazon MQ differ from SQS?
    A. Amazon MQ is serverless and offers unlimited throughput like SQS.
    B. Amazon MQ is a broker-based service supporting standard messaging protocols and features like transactions and durable topics; SQS is a fully managed queue service without broker semantics.
    C. Amazon MQ cannot persist messages.
    D. SQS supports JMS natively while Amazon MQ does not.
    Correct: B
    Explanation: Amazon MQ is broker-based (ActiveMQ/RabbitMQ) with broker semantics and protocol compatibility; SQS is a managed queue service with simple API semantics.

  3. Which use case is Amazon MQ most appropriate for?
    A. High-throughput streaming analytics requiring partitioned shards
    B. Lift-and-shift of an enterprise JMS application that requires transactions and message brokers
    C. Simple fan-out notification to mobile devices
    D. Static website hosting
    Correct: B
    Explanation: Amazon MQ suits migrations of broker-dependent enterprise apps that rely on JMS or AMQP protocols, transactions, and advanced broker features.

  4. Which statement about Amazon MQ scaling is true?
    A. Amazon MQ scales automatically without any instance sizing decisions.
    B. You provision broker instances or clusters; scaling involves changing instance types, storage, or cluster size (managed but broker-like).
    C. Amazon MQ shards like Kinesis.
    D. Amazon MQ always uses SQS under the hood for scaling.
    Correct: B
    Explanation: Amazon MQ is managed but you provision broker instances/clusters. Scaling is performed via instance sizing, clustering, or storage adjustments.

  5. You require guaranteed in-order delivery and exactly-once processing across distributed consumers with JMS transactions. Which design best meets this with minimal code change?
    A. Use SQS Standard + Lambda consumers
    B. Use Kinesis Data Streams with custom idempotency logic
    C. Use Amazon MQ (ActiveMQ) with JMS and broker transactions
    D. Use SNS + SQS fan-out
    Correct: C
    Explanation: Amazon MQ supports JMS and broker-level transactions, preserving in-order and transactional semantics expected by JMS-based applications.

SET2 

​Scenario-Based Practice Questions (SQS, SNS, Kinesis, Amazon MQ)

Easy (foundational scenarios)

  1. A retail web app needs to decouple order intake from order processing. Occasional duplicate processing is acceptable, and bursty traffic must be buffered with virtually unlimited throughput. Which service?
    A. Amazon MQ
    B. SQS Standard
    C. SQS FIFO
    D. Kinesis Data Streams

  2. You must push a single “order created” event to multiple microservices and also send SMS alerts to customers. Which design fits best?
    A. SQS queue per consumer
    B. SNS topic with SQS subscriptions and SMS protocol
    C. Kinesis Data Streams with Lambda consumers and SMS
    D. Amazon MQ topic with SMS bridge

  3. Your Lambda function processes SQS messages but sometimes times out. You need to prevent the same message from being processed by multiple Lambdas at once. Which setting do you tune first?
    A. Message attributes
    B. Visibility timeout
    C. DelaySeconds
    D. Redrive policy

  4. You need strict per-customer ordering and deduplication for invoice events. Throughput is modest. Which option?
    A. SQS Standard
    B. SNS Standard
    C. SQS FIFO with MessageGroupId=customerId
    D. Kinesis Firehose

  5. A team wants to send 3 MB payloads through SQS. What is the correct pattern?
    A. Increase SQS max message size to 3 MB
    B. Use Amazon MQ
    C. Put payload in S3 and send pointer in SQS message
    D. Use SNS with large payloads

Medium (design trade-offs and integrations)

  1. A gaming platform needs to ingest high-throughput, ordered clickstream data and allow multiple downstream consumer applications to process the same stream differently. Which service?
    A. SQS Standard
    B. SNS
    C. Kinesis Data Streams
    D. Amazon MQ

  2. You run a legacy app using JMS with durable subscriptions and transactions. You want minimal code changes during migration to AWS. Which is best?
    A. SQS FIFO
    B. SNS FIFO
    C. Amazon MQ (ActiveMQ)
    D. Kinesis Data Streams

  3. A producer publishes to SNS; three microservices subscribe via SQS. One consumer is down and you don’t want to block others. What happens?
    A. SNS blocks all deliveries until all subscribers confirm
    B. Only the down consumer misses messages; others receive them independently
    C. SNS retries synchronously until all succeed
    D. SNS routes to Kinesis as fallback

  4. Your SQS-based worker sometimes takes longer than expected. Messages reappear and get double-processed. What’s the most appropriate fix?
    A. Increase message retention
    B. Increase visibility timeout and extend it during processing
    C. Use FIFO queue
    D. Decrease batch size to 1

  5. You need near-real-time delivery of streaming records into S3 and Redshift with optional Lambda transformation, without managing consumer code. Which service?
    A. Kinesis Data Streams
    B. Kinesis Data Firehose
    C. SQS FIFO
    D. Amazon MQ

  6. You have a multi-tenant system with per-tenant strict ordering and high throughput. Which SQS FIFO design scales best?
    A. Single FIFO queue, single MessageGroupId for all tenants
    B. One FIFO queue per tenant
    C. Single FIFO queue, use MessageGroupId=tenantId
    D. Use SQS Standard with client-side ordering

  7. Your compliance team requires message encryption at rest and private access from VPC to the queue endpoint. What should you use?
    A. SQS with SSE-KMS + VPC endpoints (PrivateLink)
    B. SNS with HTTPS only
    C. Kinesis with server-side encryption only
    D. Amazon MQ without VPC integration

Hard (complex, multi-constraint scenarios)

  1. You have an IoT telemetry pipeline with 200K messages/sec, ordered per device, multiple analytics consumers, and replay needs. Minimal custom ops. Which combination is best?
    A. SQS Standard + Lambda + DLQ
    B. Kinesis Data Streams (sharded per device hash) + Firehose to S3 + Analytics for SQL
    C. SNS to multiple SQS queues + Athena
    D. Amazon MQ cluster + EBS-optimized brokers

  2. An e-commerce checkout requires exactly-once order placement semantics with transactional guarantees, integrating a JMS-based payment gateway with minimal code change. Which is most suitable?
    A. SQS Standard + dedupe keys
    B. Amazon MQ (ActiveMQ) with JMS transactions
    C. Kinesis + idempotent consumer
    D. SNS FIFO + SQS FIFO

  3. EventBridge routes “OrderCreated” events to an SQS queue feeding a Lambda. Some events perpetually fail due to malformed payloads. You need isolation and later reprocessing. Which design?
    A. Increase Lambda timeout and retries
    B. Configure a DLQ for the SQS queue with maxReceiveCount, and create a reprocessor
    C. Turn off visibility timeout
    D. Use SNS instead of SQS

  4. A global API runs in us-east-1 and eu-west-1 with ALBs. You need transport-level acceleration, static IPs for partner allowlists, and sub-second failover between regions. Which front door?
    A. Route 53 latency-based routing
    B. CloudFront
    C. AWS Global Accelerator
    D. Direct Connect Gateway

  5. Your SQS consumer fleet scales out on spikes, causing downstream RDS contention. You must protect RDS without losing messages. Best approach?
    A. Increase SQS message retention and throttle consumers via concurrency limits
    B. Use SNS instead of SQS
    C. Decrease queue visibility timeout
    D. Disable Auto Scaling

  6. You must broadcast ordered events to multiple services and maintain ordering end-to-end for subscribers that care about sequence. Which is best?
    A. SNS Standard + SQS Standard
    B. SNS FIFO topic + SQS FIFO subscriptions (per ordered stream/group)
    C. Kinesis Firehose
    D. SQS FIFO only

  7. You need to process large (50 MB) messages with exactly-once semantics and transactional coordination with legacy JMS services. Which design meets the constraint with least refactor?
    A. SQS + S3 pointer pattern + idempotency keys
    B. Kinesis + S3 pointer pattern
    C. Amazon MQ (ActiveMQ) with BlobMessage or binary payload + transactions
    D. SNS with message splitting

  8. You run Kinesis Data Streams with 20 shards. One hot partition key causes throttling while others are underused. What’s the most appropriate mitigation?
    A. Add DLQ
    B. Increase shard count and implement partition-key sharding (e.g., key+suffix) to distribute load
    C. Switch to SQS Standard
    D. Lower producer throughput globally


Answer Key with Brief Rationales

  1. B — SQS Standard offers at-least-once, best-effort ordering, and virtually unlimited throughput.

  2. B — SNS supports multi-protocol fan-out including SQS and SMS.

  3. B — Visibility timeout prevents concurrent processing while a consumer works.

  4. C — SQS FIFO with MessageGroupId ensures per-key ordering and dedupe.

  5. C — Store payloads in S3 and send pointers through SQS.

  6. C — Kinesis Data Streams supports high-throughput, ordered, multi-consumer processing.

  7. C — Amazon MQ provides JMS compatibility and transactions for lift-and-shift.

  8. B — SNS delivers independently to each subscription; one subscriber down doesn’t block others.

  9. B — Increase and dynamically extend visibility timeout to cover long processing.

  10. B — Firehose delivers to S3/Redshift with optional Lambda transform, no custom consumer code.

  11. C — One FIFO queue using MessageGroupId=tenantId preserves per-tenant order and scales.

  12. A — SQS supports SSE-KMS and VPC endpoints for private access.

  13. B — KDS for high-throughput with ordering; Firehose to S3 for durable landing; Analytics for SQL.

  14. B — Amazon MQ with JMS transactions supports exactly-once-like semantics for JMS apps.

  15. B — DLQ isolates poison messages for later reprocessing.

  16. C — Global Accelerator provides static IPs, transport acceleration, and fast failover.

  17. A — Use retention as buffer and throttle/limit consumer concurrency to protect RDS.

  18. B — SNS FIFO with SQS FIFO preserves ordering for subscribers that need it.

  19. C — Amazon MQ supports large messages (via broker features) and JMS transactions with minimal refactor.

  20. B — Mitigate hot keys by increasing shards and spreading the key (key sharding).