Microservices Resilience and Fault Tolerance 4

Question 1

What is the purpose of a dead letter queue (DLQ) in message-driven resilience?

Accepted Answer

To capture messages that repeatedly fail processing for later inspection or reprocessing

Answer

To speed up message delivery

Answer

To compress message bodies

Answer

To encrypt the message broker

Question 2

In the Saga pattern, how are failures handled across distributed transactions?

Accepted Answer

Compensating transactions undo the effects of previously completed steps

Answer

Two-phase commit locks all services

Answer

The system rolls back the database automatically

Answer

All services share one ACID transaction

Question 3

What resilience benefit does asynchronous, event-driven communication provide over synchronous calls?

Accepted Answer

It decouples services so a temporary consumer outage doesn't immediately fail the producer

Answer

It guarantees instant responses

Answer

It removes the need for monitoring

Answer

It eliminates all latency

Question 4

What is chaos engineering primarily used to achieve?

Accepted Answer

Proactively injecting controlled failures to validate system resilience before real outages occur

Answer

Randomly deleting production databases

Answer

Disabling all logging in production

Answer

Generating random API keys

Question 5

Why is setting a maximum retry count important?

Accepted Answer

It bounds resource usage and prevents infinite retry loops against a permanently failing dependency

Answer

It guarantees the request will eventually succeed

Answer

It disables the circuit breaker

Answer

It forces synchronous processing

Question 6

What does 'backpressure' refer to in a streaming or message-processing system?

Accepted Answer

A mechanism for a slow consumer to signal producers to slow down to avoid overload

Answer

Increasing network bandwidth automatically

Answer

Encrypting the data stream

Answer

Duplicating messages for redundancy

Question 7

How does request hedging improve tail latency resilience?

Accepted Answer

By sending duplicate requests to multiple replicas and using the first response

Answer

By caching all responses forever

Answer

By disabling timeouts

Answer

By batching all requests into one