What should you focus on for a Kafka and message queues interview?
Can you prevent duplicate effects after a crash?
Start with delivery correctness. This topic appeared in 53 of 169 interviews. Preventing duplicate business effects after redelivery was the most asked question, appearing in 12 of 169 interviews.
Check whether you can explain what happens during a failure, not just name idempotency:
- What happens if a consumer crashes while handling a message?
- How would you stop a redelivered message from repeating a business effect?
- How would you handle failed events with retries and dead-letter queues?
- Can you explain acknowledgments, exactly-once processing, and transactional outboxes?
Can you handle late or out-of-order events?
Stream processing appeared in 47 of 169 interviews. Questions covered batch versus streaming, late data, and changes to data pipelines.
Start by explaining batch versus stream processing in plain words. Then check whether your answer covers what happens when events arrive out of order:
- How would you avoid applying a stale update without losing a valid change?
- How would you handle late data and use watermarks?
- How would you handle schema changes in a pipeline?
The topic also includes Spark, Flink, CDC, lake ingestion, and real-time analytics or ML pipelines.
Can you compare brokers and explain Kafka's parts?
Questions in these interviews asked candidates to compare Kafka with RabbitMQ and explain how Kafka works with an API. Check whether you can answer both the tool-choice question and the questions about how messages move:
- When would you use a queue rather than pub/sub, and which consumers receive each message?
- What is partitioning in Kafka, and what problems would arise without it?
- Can you explain topics, replication, consumer groups, rebalancing, and offsets?
Can you design workers and keep them running?
System design questions cover payments, notifications, chat, documents, and other asynchronous workflows. Choose one of these workflows and check your design:
- How would you route jobs to workers and control concurrency?
- How would you handle backpressure and scale workers as the queue backlog grows?
- What would you monitor, how would you plan capacity, and how would you recover from a failure?
Writing code was part of 2 of 169 interviews. Can you write a Python program to handle SNS or SQS failures? Can you write an AWS Glue PySpark job to clean data from Kinesis?
Who gets asked Kafka and Message Queues
The 20 most asked Kafka and Message Queues questions
Counted by interview, not by mention: a question asked twice in one conversation counts once.
How would you prevent duplicate business effects when a message is redelivered, including after a crash?
Asked in 12 of 169 interviews · last asked Oct 18 · Design · Senior
Show answer
I would use a stable payment ID as an idempotency key and enforce uniqueness in durable storage. For a local credit, I would save the credit and processed ID in one database transaction, then acknowledge the message. Redelivery after a crash would not repeat the credit. For external payments, use provider-supported idempotency or reconcile uncertain outcomes.
- Use a database uniqueness constraint, not only an in-memory check, to handle restarts and concurrent deliveries.
- Acknowledge the message or commit its Kafka offset only after the database transaction commits.
- For external payments with unknown outcomes, retry with the same idempotency key or reconcile with the provider.
- Enforce valid state transitions: late events must not repeat credits or move settled payments back to intermediate states.
What the interviewer checks: The interviewer is checking whether your design prevents repeated business effects across retries, concurrent deliveries, and crashes.
Other ways it was asked
- “If a service restarts and an external service resends the same message for a 50,000-peso payment, how would you prevent the recipient from receiving 100,000 pesos?”
- “If an external service resends the same 50,000-peso payment message after your service restarts, how would you prevent the recipient from receiving the payment twice?”
- “How would you design the system so that if a service restarts and an external service resends the same 50,000-peso credit message, the customer is not credited twice? How would you handle a late event indicating that the transaction is already settled or in an intermediate state?”
What are the differences between Kafka and RabbitMQ?
Asked in 4 of 169 interviews · last asked Sep 16 · Concept · Mid-level
Show answer
Kafka is a partitioned event log. RabbitMQ is mainly a message broker with queues and flexible routing. Kafka retains events for replay, while RabbitMQ queues normally remove messages after acknowledgment. Choose Kafka for event streams and replay; choose RabbitMQ for task distribution and routing messages to specific queues.
- Kafka consumers track offsets; reading an event does not delete it.
- Kafka preserves order within a partition, not across all partitions.
- RabbitMQ uses exchanges to route messages and supports acknowledgments and requeuing failed deliveries.
- RabbitMQ also supports streams with retention and replay, unlike its normal queue behavior.
What the interviewer checks: The interviewer checks whether you can compare message storage, delivery, and routing, then choose based on application needs.
Other ways it was asked
- “What is the difference between Kafka and RabbitMQ?”
- “What are the implementation differences between Kafka and RabbitMQ, and how do you decide which one to use?”
- “What is the key difference between Kafka and RabbitMQ?”
How does Kafka work, and how would you integrate it with an API?
Asked in 4 of 169 interviews · last asked Sep 10 · Concept · Mid-level
Show answer
Kafka stores events in replicated logs called topic partitions. Producers write events, and consumers read them while tracking their position with offsets. Within a consumer group, each partition has one consumer at a time. An API can publish an event and return HTTP 202 after Kafka confirms the write; workers then process it asynchronously.
- Brokers store partitions; replication helps Kafka survive broker failures.
- Ordering is guaranteed within a partition, not across an entire topic.
- Reading does not delete events; retention settings control how long Kafka keeps them.
- Handle duplicate processing safely; use a transactional outbox when database changes and event publishing must stay consistent.
What the interviewer checks: The interviewer checks whether you understand Kafka’s storage model and can connect an API to reliable asynchronous processing.
Other ways it was asked
- “How does Kafka work?”
- “How does Kafka work internally?”
- “What is Kafka?”
What is the difference between batch processing and stream processing?
Asked in 4 of 169 interviews · last asked Oct 19 · Concept · Mid-level
Show answer
Batch processing handles a fixed set of data in one job. Stream processing handles incoming data continuously, usually with lower delay. Use batch processing for daily reports or historical analysis, and stream processing for live alerts or fraud detection.
- Batch jobs process bounded data; streaming jobs usually process data with no defined end.
- Streaming may need to handle late events and often keeps state across events.
- Streaming can use small batches internally; it does not always process one event at a time.
- In Spark, use spark.read for batch input and spark.readStream for streaming input.
What the interviewer checks: The interviewer checks whether you understand data boundaries, processing delay, and when to use each approach.
Other ways it was asked
- “What are stream processing and batch processing, and when would you use each?”
- “What is the difference between batch processing and streaming?”
- “What is the difference between batch and stream processing, and how do you read data with Spark?”
How do you handle late or out-of-order events without applying stale updates or losing valid changes?
Asked in 4 of 169 interviews · last asked Oct 22 · Scenario · Senior
Show answer
I track event IDs and source versions for each payment or transfer. I apply updates and record processed events atomically, checking versions and allowed state transitions. I reject stale status updates, but preserve valid changes such as balance adjustments. If ordering is unclear, I buffer events or check the source system.
- Use event IDs for deduplication and source versions for ordering; arrival time does not show event order.
- Older full-state snapshots can be skipped; incremental changes may need buffering and replay in source order.
- After settlement, a stale “in progress” event must not move the transfer backward.
- Queue ordering by payment ID preserves arrival order, not necessarily the provider’s original event order.
What the interviewer checks: The interviewer checks whether you prevent stale state without confusing duplicate events with valid late changes.
Other ways it was asked
- “If a webhook provider can deliver events out of order, how would you design your processing so the final state is still correct and you don't drop legitimate updates?”
- “How would you handle a late-arriving event in a payment-processing workflow?”
- “An MQ consumer receives transfer events in this order: authorized, settled, then in progress. What should the final state be, and how would you handle the late event?”
What is the difference between a queue and messaging?
Asked in 3 of 169 interviews · last asked Jul 28 · Concept · Junior
Show answer
Messaging is the broad idea of sending messages between systems. A queue is one messaging pattern: workers share messages, with each message normally handled by one worker. A topic supports publish-subscribe messaging, where separate subscriptions can each receive the same message.
- Use queues for background jobs, sharing work, and handling traffic spikes.
- Use topics when several services need the same event.
- Kafka topics deliver independently to consumer groups; consumers within a group share partitions.
- Retries can cause duplicate delivery, so consumers may need to handle duplicates safely.
What the interviewer checks: The interviewer checks whether you distinguish messaging in general from work sharing and publish-subscribe patterns.
Other ways it was asked
- “What are messaging queues, and when would you use them?”
- “What is the difference between a topic and a queue in a messaging system?”
What is partitioning in Kafka, and what problems would arise without it?
Asked in 3 of 169 interviews · last asked Sep 4 · Concept · Mid-level
Show answer
Partitioning splits a Kafka topic into separate ordered logs that can be spread across brokers. This spreads storage and traffic and lets consumers in a group work in parallel. With only one partition, a topic is limited to one leader broker and one active consumer per group. Kafka guarantees order within each partition, not across the whole topic.
- A topic is a named stream of records; partitions hold parts of that stream.
- Partitions allow a topic to scale across brokers and consumers.
- With standard key-based partitioning, the same key goes to the same partition unless the partition count changes.
- Replication provides fault tolerance; partitioning alone does not.
What the interviewer checks: The interviewer checks whether you understand how partitions enable scaling and limit ordering guarantees.
Other ways it was asked
- “What are Kafka topics and partitions, and how can partitioning affect message ordering?”
- “What is partitioning in Kafka?”
- “Why does Kafka use partitions, and what would happen if a topic's data were not partitioned?”
How do you handle failed events in a message consumer, including retries and dead-letter queues?
Asked in 3 of 169 interviews · last asked Aug 20 · Scenario · Mid-level
Show answer
I retry temporary failures with a limit, exponential backoff, and jitter. I send invalid events or events that exhaust retries to a dead-letter queue with error details. I acknowledge messages or commit offsets only after successful processing or durable handoff. I make processing idempotent so redelivery does not repeat side effects.
- Do not retry permanent errors, such as invalid data.
- Monitor the dead-letter queue, fix the cause, and replay events safely.
- Retry topics avoid blocking consumption but can change processing order.
- In Kafka, do not commit past an earlier event unless it is processed or durably handed off.
What the interviewer checks: The interviewer checks whether you can recover from failures without losing events or repeating side effects.
Other ways it was asked
- “What architectural pattern would you use to handle errors in a Kafka-based system?”
- “What should a message consumer do with an event it cannot process?”
- “How do you handle events, and what do you do when a consumer fails?”
How would you scale workers to keep up with a growing queue backlog?
Asked in 3 of 169 interviews · last asked Oct 22 · Scenario · Senior
Show answer
I would scale workers based on queue depth, message age, and how fast messages arrive and finish. I would add enough capacity to handle new messages and clear the backlog within a target time. I would limit scaling to protect downstream systems, and scale down slowly when demand falls.
- Track backlog growth and oldest message age, not just worker CPU usage.
- Use arrival rate and measured worker throughput to estimate the workers needed.
- In a standard Kafka consumer group, consumers assigned partitions cannot outnumber partitions across subscribed topics.
- Check downstream limits and processing bottlenecks before adding more workers.
What the interviewer checks: The interviewer checks whether you can size worker capacity while respecting queue limits and downstream capacity.
Other ways it was asked
- “How do you ensure there are enough workers to take things off the queue?”
- “How would you handle a growing queue backlog when usage increases?”
- “How should worker instances scale as a queue grows?”
When would you use a queue versus pub/sub, and which consumers receive each message?
Asked in 2 of 169 interviews · last asked Aug 20 · Concept · Mid-level
Show answer
Use a queue to share work: each message goes to one consumer, not all consumers. Use pub/sub when several independent services need the same event: each subscription gets a copy. In Kafka, each consumer group gets the messages, while consumers within a group share the work.
- Queues fit background jobs where workers share the load.
- Pub/sub fits events needed by several independent services.
- Consumers sharing one subscription may split work rather than each receive every message.
- Retries can cause duplicate delivery; one consumer per message does not mean exactly-once processing.
What the interviewer checks: The interviewer is checking whether you understand shared work versus sending events to independent subscribers.
Other ways it was asked
- “With multiple consumers, when would you use a queue versus pub/sub, and do all consumers receive the same message in each pattern?”
- “When would you use the Pub/Sub pattern?”
When would you choose synchronous APIs versus asynchronous messaging, and how do failure modes influence that choice?
Asked in 2 of 169 interviews · last asked Oct 21 · Design · Senior
Show answer
Choose synchronous APIs when the caller needs an immediate result and can wait for the other system. Choose asynchronous messaging when work can finish later or must survive temporary consumer outages. Synchronous calls risk timeouts and cascading failures; messaging adds delayed results, possible duplicate delivery, and a queue backlog if consumers cannot keep up.
- Use HTTP/REST or gRPC for synchronous calls; Kafka or a message broker for asynchronous messaging.
- A timeout does not prove failure: the server may have completed the operation.
- Use bounded retries and idempotent operations; synchronous calls also need timeouts and often circuit breakers.
- For messaging, configure durable storage, acknowledge after processing, and handle duplicates and repeatedly failing messages.
What the interviewer checks: The interviewer checks whether you connect response needs and failure handling to the communication choice.
Other ways it was asked
- “When designing an interface between two systems, what factors determine whether it should be synchronous or asynchronous, and which tools and protocols would you use?”
- “When would you use synchronous REST communication versus asynchronous messaging between enterprise applications? Which failure modes influence that decision?”
What roles do Azure Service Bus and Azure Functions serve, and when would you use each?
Asked in 2 of 169 interviews · last asked Oct 16 · Design · Mid-level
Show answer
Azure Service Bus is a message broker that stores and delivers messages between services. Use it to separate services and buffer work until a receiver is ready. Azure Functions runs code when an event or timer triggers it. Use Functions for event-driven tasks, such as processing messages from Service Bus.
- Service Bus queues share work among receivers; topics and subscriptions send copies to separate subscribers.
- Functions runs application code; Service Bus transports and stores messages.
- They can work together: a Service Bus message can trigger a Function.
- Messages can be delivered again, so handlers should safely handle duplicates.
What the interviewer checks: The interviewer checks whether you can distinguish message delivery from code execution and explain when to combine them.
Other ways it was asked
- “What is the difference between Azure Service Bus and Azure Functions, and when would you use each?”
- “How can Azure Functions and Azure Service Bus be used in a backend application, and what purpose does each serve?”
- “What role does Azure Service Bus play in a backend architecture?”
How would you choose a message broker or queue, and when would you use alternatives to Kafka?
Asked in 2 of 169 interviews · last asked Sep 18 · Concept · Mid-level
Show answer
I choose based on message volume, latency, ordering, replay needs, delivery guarantees, and operating cost. Kafka fits event streams that need replay and several independent consumers. RabbitMQ fits task queues and flexible routing. A managed queue such as Amazon SQS fits background jobs when I want less infrastructure to manage.
- Check whether messages need long retention and replay, or mainly need processing once.
- Kafka guarantees ordering within a partition, not across the whole topic.
- Compare retries, duplicate handling, routing, and consumer scaling against the workload.
- Test realistic traffic and include hosting costs, team skills, and operational work.
What the interviewer checks: The interviewer is checking whether you can match system needs to messaging tools instead of choosing Kafka by default.
Other ways it was asked
- “Which messaging tools have you used, and how do you choose between them?”
- “What could you use instead of Kafka, and how would you know which option is better for your system than other queues?”
- “What criteria should guide the choice of a queue for your project?”
What problems does Azure Service Bus solve, and how would you use it in a web application?
Asked in 2 of 169 interviews · last asked Sep 18 · Tooling · Mid-level
Show answer
Azure Service Bus provides reliable messaging between parts of an application, so they do not need to run at the same time. It helps handle traffic spikes and move slow work out of web requests. In a web app, I would send an order message to a queue and let a background worker process it.
- Queues distribute work among workers; topics and subscriptions let several services receive the same event.
- Messages wait while workers are unavailable, reducing the need for services to be available together.
- Workers must handle duplicate messages safely because a message can be delivered more than once.
- Use retries for temporary failures and a dead-letter queue for messages that cannot be processed.
What the interviewer checks: The interviewer checks whether you can choose asynchronous messaging and explain reliable background processing in a web app.
Other ways it was asked
- “How would you use Azure Service Bus in a web app?”
- “What problems is Azure Service Bus suited to, and what kinds of tasks can it handle?”
What is the outbox pattern, and how does it prevent lost events after database commits?
Asked in 2 of 169 interviews · last asked Sep 18 · Concept · Mid-level
Show answer
The outbox pattern saves a business change and its event in the same database transaction. A separate process reads committed outbox events and publishes them to the message broker. If the application crashes after the commit, the event stays in the outbox for later delivery. Retries can create duplicates, so consumers must handle them safely.
- The business change and outbox event either both commit or both roll back.
- A publisher uses polling or change data capture to find committed events.
- Mark an event as sent only after the broker confirms publication.
- Delivery is usually at least once; use event IDs or idempotent processing to handle duplicates.
What the interviewer checks: The interviewer checks whether you can avoid lost events without assuming a database write and broker publish are atomic.
Other ways it was asked
- “What is the Outbox pattern, and what problem does it solve?”
- “What would you do if the application crashed after writing to the database but before publishing the event?”
How would you design a distributed task queue for background jobs?
Asked in 2 of 169 interviews · last asked Sep 1 · Design · Senior
Show answer
I would use an API to submit jobs, a durable message broker to hold them, and workers to process them. Workers record results in a database and acknowledge jobs only after success. I would retry failures with delays and move repeatedly failing jobs to a dead-letter queue. Jobs must be safe to repeat because delivery can happen more than once.
- Run more workers as the backlog grows, but limit concurrency to protect downstream services.
- Use job IDs to track status and prevent duplicate effects, such as sending the same email twice.
- Recover unfinished jobs when a worker crashes, using broker-specific acknowledgment or lease rules.
- Monitor queue depth, job age, failures, and worker health.
What the interviewer checks: The interviewer checks whether you can separate background work from requests and handle failures, duplicates, and growing load.
Other ways it was asked
- “How do you design a distributed task queue for background jobs?”
- “How would you design a simple event-driven background job for something like sending emails or processing long tasks? Just the high-level components.”
How would you design a real-time stream processing solution for IoT device data using PySpark and Azure Databricks, including aggregations, fault tolerance, and schema evolution handling?
Asked in 2 of 169 interviews · last asked Aug 11 · Design · Senior
Show answer
I would use PySpark Structured Streaming on Azure Databricks to read from Kafka or Azure Event Hubs. I would validate records, filter temperature anomalies, and calculate per-device averages over event-time windows. Delta tables would store readings and aggregates, while retry-safe writes send anomalies to Azure SQL. Durable checkpoints would support recovery.
- Use watermarks to limit aggregation state; events beyond the lateness limit may be dropped.
- Partition input for parallel reads and tune micro-batch intervals to balance latency and processing capacity.
- Keep raw events, explicitly approve additive Delta schema changes, and quarantine records with incompatible types.
- Store checkpoints in ADLS; use stable event IDs and SQL upserts because foreachBatch retries can duplicate writes.
What the interviewer checks: The interviewer checks whether you can combine streaming transformations with low latency, scaling, safe recovery, and controlled schema changes.
Other ways it was asked
- “You are tasked with designing a real time data processing pipeline using Azure Databricks and PySpark. The data comes from an IoT device stream with thousands of records per second, containing information such as device_id, temperature, and timestamp. The requirements are to identify devices with abnormal temperature readings, e.g. above a threshold, and store these anomalies in a separate Azure SQL database table for further analysis. How would you design this pipeline to ensure low latency, scalability, and fault tolerance? Please include any PySpark transformations or optimizations you would use.”
How would you design a high-volume real-time event processing system for reliability, scalability, and performance?
Asked in 2 of 169 interviews · last asked Jul 22 · Design · Senior
Show answer
I would put a replicated Kafka log between event ingestion and stream workers, then write results to suitable storage. I would size partitions and consumer groups for peak traffic, latency targets, and required ordering. I would use durable writes, retries, and idempotent processing. I would monitor lag and latency, apply backpressure, and test broker and worker failures.
- Use acks=all with suitable min.insync.replicas and replicas across failure zones; retain events long enough for replay.
- Partition by entity key for ordering; partition count limits active consumers in each consumer group.
- Commit offsets after successful processing; make writes idempotent, bound retries, and route failed events to a dead-letter topic.
- Batching and compression improve throughput but can add latency; load-test and scale consumers based on lag.
What the interviewer checks: The interviewer checks whether you can connect system components with clear scaling limits, failure handling, and performance trade-offs.
Other ways it was asked
- “How would you approach designing an end-to-end system for handling high-volume real-time event data like user actions or messages coming in rapidly?”
- “How would you approach end-to-end system design for handling high volume of basic event data like user actions or messages coming in rapidly? Walk through the key components you'd consider.”
- “Can you describe a system you built using Kafka or RabbitMQ to process large amounts of data or events in real time, including the reliability, scaling, and performance challenges you faced and how you solved them?”
How do you handle message exchange between microservices?
Asked once · last asked Oct 14 · Design · Senior
Show answer
I use direct API calls when a service needs an immediate reply, and a broker like Kafka or RabbitMQ for asynchronous messages. I agree on message formats, make consumers safe against duplicate processing, and retry temporary failures. I also monitor failures and pass trace IDs between services.
- Use events for facts that happened and commands to request work.
- Acknowledge messages or commit offsets after successful processing; expect possible redelivery.
- Use limited retries and a dead-letter queue for messages that keep failing.
- If a database update must produce a message, use a transactional outbox to avoid losing it.
What the interviewer checks: The interviewer checks whether you can choose a communication method and handle failures without losing work.
What are message brokers such as Kafka, RabbitMQ, and SQS, and what role do they play?
Asked once · last asked Jan 28 · Concept · Mid-level
Show answer
Message brokers are systems that receive messages from producers and make them available to consumers. They let services communicate without waiting for each other and buffer work when consumers are slow or unavailable. Kafka stores replayable event streams, RabbitMQ routes messages to queues, and Amazon SQS provides managed message queues.
- Producers and consumers can run and scale separately.
- Buffering helps handle traffic spikes and temporary consumer failures.
- Delivery guarantees vary; consumers may need to handle duplicate messages.
- Kafka retains messages for replay; queues usually remove them after acknowledgment or deletion.
What the interviewer checks: The interviewer is checking whether you understand asynchronous communication and the main differences between event streams and queues.
Message queues and event-driven architecture interview questions
Compare queues with pub/sub, commands with events, and synchronous with asynchronous communication, alongside questions on event sourcing. 25 questions.
Why use a message queue or broker instead of making an HTTP call for every message?
Asked once · last asked Jan 28 · Concept · Mid-level
What is event sourcing, and how could it help with concurrent updates and notifying other users?
Asked once · last asked Jan 28 · Concept · Senior
How do you understand the claim that Kafka is not a queue?
Asked once · last asked Mar 6 · Concept · Mid-level
Can you tell me the difference between a command and an event that might be fired?
Asked once · last asked Apr 2 · Concept · Mid-level
Can a command have multiple subscribers, or is there just a single subscriber?
Asked once · last asked Apr 2 · Concept · Mid-level
What are message queue systems used for, and what kinds of problems do they solve?
Asked once · last asked May 25 · Concept · Mid-level
What is the difference between synchronous and asynchronous integrations, and what technical approaches could be used for each?
Asked once · last asked Jun 9 · Concept · Mid-level
How would you decide whether to store groups in a database or represent them as Kafka events?
Asked once · last asked Jul 16 · Scenario · Mid-level
When would you use an API versus Kafka for service communication?
Asked once · last asked Jul 17 · Concept · Senior
How are domain events managed or captured?
Asked once · last asked Jul 19 · Concept · Senior
Why might you place a durable queue between a user-facing API and a background worker?
Asked once · last asked Aug 10 · Concept · Mid-level
If an API queues a request for processing, how should the user receive the response?
Asked once · last asked Sep 10 · Design · Senior
Show 8 more
What is the difference between event-driven architecture and event sourcing architecture?
Asked once · last asked Sep 18 · Concept · Senior
How can queues improve performance in a highly scalable, complex application?
Asked once · last asked Sep 23 · Design · Senior
How do message queues, Redis, and dead-letter queues (DLQs) support distributed systems?
Asked once · last asked Oct 15 · Concept · Senior
When would you recommend using an event bus in an application?
Asked once · last asked Oct 19 · Design · Senior
Other ways it was asked
- “Would you recommend using an event bus for this application?”
What is publish-subscribe, and when would you use it?
Asked once · last asked Oct 23 · Concept · Mid-level
What is event streaming?
Asked once · last asked Feb 18 · Concept · Junior
What is event-driven architecture?
Asked once · last asked Feb 22 · Concept · Junior
In an order-processing flow that uses a queue, what should the first and second messages represent?
Asked once · last asked Mar 19 · Design · Mid-level
Kafka vs RabbitMQ and cloud messaging interview questions
These questions compare Kafka, RabbitMQ, Redis, and cloud messaging tools, including their features and integration options. 33 questions.
What is RabbitMQ, and how does it work as a messaging system?
Asked once · last asked Oct 15 · Concept · Mid-level
Other ways it was asked
- “What does RabbitMQ do?”
How do delivery guarantees differ between Kafka, RabbitMQ, and Redis?
Asked once · last asked Jan 28 · Concept · Senior
Can Redis be used for publish-subscribe, or is it only for caching?
Asked once · last asked Jan 28 · Concept · Mid-level
How does message acknowledgment differ between RabbitMQ and Kafka?
Asked once · last asked Jan 28 · Concept · Senior
If the two Dataverse environments must communicate through a Service Bus broker, what are the pros and cons of using Azure Functions, Logic Apps, Power Automate, or another approach?
Asked once · last asked Mar 2 · Design · Senior
What is the difference between SQS, SNS, and EventBridge, and in what scenarios would you use each?
Asked once · last asked Mar 3 · Concept · Senior
Why would you choose RabbitMQ over Redis or Redis Pub/Sub for a messaging system?
Asked once · last asked Mar 19 · Concept · Mid-level
When should you use Azure Service Bus and Azure Functions with Dynamics 365, and what are the advantages of these solutions?
Asked once · last asked Apr 6 · Scenario · Senior
How long is data stored in Pub/Sub? What is its retention period?
Asked once · last asked Apr 30 · Concept · Mid-level
Why would the APIs read data from Pub/Sub instead of a database?
Asked once · last asked Apr 30 · Concept · Senior
In an architecture using Pub/Sub and APIs, how can a topic trigger an API, and what might the API do with the message?
Asked once · last asked Apr 30 · Scenario · Mid-level
How would you design the flow from publishing an event to consuming it with Azure Service Bus, including consumer counts and partitions?
Asked once · last asked Jun 18 · Design · Senior
Show 17 more
Would Kafka replace Redis in a chat system design?
Asked once · last asked Jul 28 · Design · Senior
What determines the number of partitions configured for an Azure Event Hub?
Asked once · last asked Aug 7 · Concept · Mid-level
How can you read the same message from Azure Event Hubs again after consuming it?
Asked once · last asked Aug 7 · Scenario · Mid-level
When publishing messages to Azure Service Bus, would you publish each event immediately or batch messages first, and why?
Asked once · last asked Sep 4 · Design · Senior
How would you route a request through Azure Service Bus when the system needs to send a request and receive a response?
Asked once · last asked Sep 18 · Design · Senior
Why would you use a message queue, and what factors would lead you to choose Kafka?
Asked once · last asked Sep 18 · Design · Mid-level
When would you choose Kafka over Amazon SQS or BullMQ for a distributed application?
Asked once · last asked Sep 21 · Concept · Senior
How do you choose between SQS and Kafka or Kinesis when handling communication between services?
Asked once · last asked Sep 25 · Design · Senior
Why would you use Amazon SQS between two services?
Asked once · last asked Sep 30 · Concept · Mid-level
What is the difference between Amazon SNS and Amazon SQS?
Asked once · last asked Sep 30 · Concept · Mid-level
What factors would you consider when selecting an event bus for this application?
Asked once · last asked Oct 19 · Design · Senior
What are AWS Lambda and queues, and how do they work?
Asked once · last asked Oct 23 · Concept · Junior
What messaging services are commonly used in software systems?
Asked once · last asked Feb 14 · Concept · Mid-level
What are common reasons to use RabbitMQ in an application?
Asked once · last asked Apr 27 · Concept · Mid-level
In AWS, which service represents a topic and which represents a queue: SNS or SQS?
Asked once · last asked Jun 9 · Concept · Junior
What is NSQ?
Asked once · last asked Jul 25 · Concept · Junior
How do you implement RabbitMQ?
Asked once · last asked Sep 16 · Tooling · Mid-level
Kafka architecture and consumer interview questions
Explore how Kafka records, topics, partitions, replication, consumer groups, rebalancing, and offsets work, plus Spring integration and throughput. 24 questions.
How does Kafka replicate messages to another node, and what protocol does it use?
Asked once · last asked Jul 22 · Concept · Senior
Can Kafka operate without ZooKeeper, and why?
Asked once · last asked Feb 18 · Concept · Mid-level
Why does Kafka use replication, and can a Kafka cluster operate without it?
Asked once · last asked Feb 18 · Concept · Mid-level
Other ways it was asked
- “What problems could occur in a Kafka cluster without data replication?”
How do you define a Kafka consumer in Spring, and which annotation do you use?
Asked once · last asked Jun 15 · Tooling · Mid-level
When would you use a Kafka consumer group, and what are its advantages?
Asked once · last asked Jun 15 · Concept · Mid-level
How does Kafka assign records to consumers in a consumer group?
Asked once · last asked Jun 15 · Concept · Senior
How do you control Kafka offset commits in Spring?
Asked once · last asked Jun 15 · Tooling · Senior
What would each message in a Kafka queue contain?
Asked once · last asked Jul 20 · Design · Mid-level
How does a Kafka consumer acknowledge message processing, and how are offsets committed?
Asked once · last asked Jul 22 · Concept · Mid-level
How does Kafka use offsets to determine where a consumer should resume after it goes down?
Asked once · last asked Jul 22 · Concept · Mid-level
What are partitions, consumers, consumer lag, and offsets in Kafka?
Asked once · last asked Sep 4 · Concept · Mid-level
How does Kafka handle such high throughput?
Asked once · last asked Sep 18 · Concept · Senior
Show 10 more
How does Kafka support event-driven architecture, and how do partitions, partition keys, consumer groups, ordering, and rebalancing work together?
Asked once · last asked Oct 20 · Concept · Senior
When is partition rebalancing needed in Kafka?
Asked once · last asked Oct 20 · Concept · Senior
Other ways it was asked
- “When do you need to do partition rebalancing?”
What are the components of Kafka?
Asked once · last asked Feb 18 · Concept · Junior
Other ways it was asked
- “What do the main components of Kafka do?”
What is a record in Kafka?
Asked once · last asked Jun 15 · Concept · Junior
Other ways it was asked
- “What is the structure of a Kafka record?”
What does an offset tracker do in a message-processing system?
Asked once · last asked Jan 28 · Concept · Mid-level
How would you explain Kafka?
Asked once · last asked Feb 18 · Concept · Junior
What is leader election in Kafka?
Asked once · last asked Feb 18 · Concept · Junior
What is replication in Kafka?
Asked once · last asked Feb 18 · Concept · Junior
What does availability mean in a Kafka cluster?
Asked once · last asked Feb 18 · Concept · Junior
In Kafka, what do producer and consumer mean?
Asked once · last asked Jul 21 · Concept · Junior
Message delivery guarantees and idempotency interview questions
Work through acknowledgments, retries, dead-letter queues, ordering, deduplication, exactly-once processing, transactional outboxes, and recovery from partial failures. 65 questions.
How should an order service handle a connection failure when publishing a message to Azure Service Bus?
Asked once · last asked Sep 4 · Scenario · Senior
If a message broker such as Kafka delivers messages more than once or out of order, how should the payment service handle them to avoid deducting payment more than once?
Asked once · last asked Jan 30 · Scenario · Senior
What is exactly-once processing in Kafka?
Asked once · last asked Feb 18 · Concept · Mid-level
How would you configure exactly-once processing in a Kafka cluster, and which property would you change?
Asked once · last asked Feb 18 · Tooling · Mid-level
How would you pass 50-millisecond audio chunks while preserving their order?
Asked once · last asked Feb 27 · Scenario · Senior
How is message ordering handled or guaranteed in Kafka?
Asked once · last asked Mar 6 · Concept · Senior
Suppose the Kafka consumer crashes before processing a message. How do you ensure the message is processed successfully?
Asked once · last asked Mar 6 · Scenario · Senior
What happens to a message in a RabbitMQ queue after a consumer reads it?
Asked once · last asked Mar 6 · Concept · Mid-level
How do you handle message acknowledgments and retry logic to ensure tasks are processed reliably?
Asked once · last asked Mar 17 · Scenario · Senior
Regarding exactly-once delivery, what steps would you implement?
Asked once · last asked Mar 17 · Scenario · Senior
How would you ensure delivery guarantees and handle notifications that fail to reach a user's device?
Asked once · last asked Mar 19 · Scenario · Senior
How would you ensure alerts are reliably delivered to human reviewers, even if the notification service experiences downtime or message loss?
Asked once · last asked Mar 27 · Scenario · Senior
All 65 Message delivery guarantees and idempotency questions
Background jobs and distributed task queue interview questions
Worker and task queue questions cover scheduling, batching, concurrency, routing, backpressure, autoscaling, and tools such as Sidekiq and Celery. 38 questions.
What problems could arise if a notification system sends a batch only when the queue contains more than 10 messages?
Asked once · last asked Mar 31 · Scenario · Mid-level
Other ways it was asked
- “What problem could occur if a notification system waits for a queue to exceed 10 messages before sending, when it receives only two messages per day?”
How would you run dynamic background jobs as Kubernetes Jobs or Rails processes, consume work from a queue, and handle long-running tasks?
Asked once · last asked Jul 24 · Design · Senior
Other ways it was asked
- “How do dynamic background jobs run as Kubernetes Jobs or Rails processes, pick up work from a queue, and handle long-running tasks?”
What data structure ensures fairness in checkout processing, and how do we prevent the payment system from overloading?
Asked once · last asked Oct 6 · Design · Mid-level
Other ways it was asked
- “What data structure would you use to ensure fair first-come-first-served processing for checkout payments while preventing the payment system from being overloaded?”
How do Bitrix agents compare with Laravel queues?
Asked once · last asked Oct 12 · Concept · Mid-level
Other ways it was asked
- “What capabilities do Laravel queues provide that Bitrix agents lack?”
How would you assess a backend service's request capacity and design its queues and concurrency?
Asked once · last asked Feb 23 · Design · Senior
How would you modify a notification system to batch messages instead of sending them immediately whenever dispatch is called?
Asked once · last asked Mar 31 · Design · Mid-level
In a batched notification system, how would you decide when to send the queued messages?
Asked once · last asked Mar 31 · Scenario · Mid-level
During intensive AI training cycles that spike database load and background jobs, how would you design backpressure so APIs stay responsive, using queues, rate limits, circuit breakers, or load shedding?
Asked once · last asked May 5 · Design · Senior
How would you implement queue prioritization and fairness so interactive requests aren't starved by large AI training jobs, and what metrics would you watch to validate it?
Asked once · last asked May 5 · Design · Senior
How would you send email to one million users in Laravel reliably, without the process being interrupted?
Asked once · last asked Jul 16 · Scenario · Senior
How would you process recipient batches and send email data in a Laravel bulk-email workflow?
Asked once · last asked Jul 16 · Scenario · Senior
How would you track batches in a Laravel bulk-email workflow so that later executions send emails only to recipients who have not yet received them?
Asked once · last asked Jul 16 · Scenario · Senior
Show 24 more
How would you coordinate Laravel commands and queued jobs to process millions of records and send email at scale?
Asked once · last asked Jul 16 · Design · Senior
How do you choose and configure a queue strategy for a Laravel application?
Asked once · last asked Jul 16 · Tooling · Senior
If a job keeps failing concurrently, how do you handle the failed jobs and retry them in production?
Asked once · last asked Jul 16 · Scenario · Senior
How would you configure BullMQ retries and backoff for a 24/7 Zoom recording pipeline encountering API rate limits, and what does fail-safe deletion mean in that architecture?
Asked once · last asked Aug 2 · Tooling · Senior
What happens to the system and the data if Redis crashes while using Sidekiq?
Asked once · last asked Aug 26 · Scenario · Senior
How would you handle jobs that hang or fail when a worker machine crashes?
Asked once · last asked Aug 26 · Scenario · Senior
Should a master job enqueue batches of 100 rows, with each batch then creating another 100 jobs, one per row?
Asked once · last asked Sep 2 · Design · Senior
In a batched CSV import, should the batch worker insert records before dispatching row-level jobs, or should it delegate insertion and hydration to the final worker layer?
Asked once · last asked Sep 2 · Design · Senior
When might inserting records in the batch worker before dispatching row-level jobs be a poor choice or fail?
Asked once · last asked Sep 2 · Scenario · Senior
How would you divide queues or topics for different image types or business scenarios?
Asked once · last asked Sep 18 · Design · Senior
How would you implement sending an email template to 1,000 leads every Monday?
Asked once · last asked Sep 28 · Scenario · Mid-level
Other ways it was asked
- “How would you handle sending an email and a WhatsApp message to 1,000 leads at the same time every Monday?”
Why would you use multiple named queues instead of a single queue in a background job system?
Asked once · last asked Sep 28 · Concept · Mid-level
How would you ensure that each lead is processed by only one worker?
Asked once · last asked Oct 1 · Design · Senior
What should you consider when horizontally scaling a message consumer that processes messages, uses a database, and calls an external API?
Asked once · last asked Oct 9 · Scenario · Senior
What is Celery, and what is it used for?
Asked once · last asked Feb 23 · Concept · Junior
What is Laravel's queue system?
Asked once · last asked Mar 12 · Concept · Junior
How would you determine the maximum queue size for a notification batching system?
Asked once · last asked Mar 31 · Scenario · Mid-level
Where should you run Laravel jobs, and when would you use the CLI, Tinker, or another execution method?
Asked once · last asked Jul 16 · Tooling · Mid-level
How does Celery work under the hood?
Asked once · last asked Aug 14 · Concept · Mid-level
Would you perform transaction matching directly in a worker or enqueue a separate job, and why?
Asked once · last asked Aug 26 · Design · Mid-level
Other ways it was asked
- “Does the worker do the transaction matching right away, or does it kick off a new job to handle that?”
How does the asynchronous job pattern work with multiple vendors?
Asked once · last asked Aug 26 · Design · Senior
What is a queue in a background job system?
Asked once · last asked Sep 28 · Concept · Junior
What is BullMQ used for?
Asked once · last asked Oct 7 · Concept · Junior
How do we implement background jobs?
Asked once · last asked Oct 14 · Design · Mid-level
Stream processing and Kafka data pipeline interview questions
Check your understanding of batch and streaming pipelines, Spark, Flink, CDC, schema changes, watermarks, late data, and real-time analytics. 61 questions.
Why is Schema Registry used with Kafka?
Asked once · last asked Feb 18 · Concept · Mid-level
Other ways it was asked
- “How do you use Schema Registry with Kafka?”
Why would you use DynamoDB Streams, and what benefits do they provide?
Asked once · last asked Jun 23 · Concept · Mid-level
Would you pass each event through the normalizer individually or process events in batches?
Asked once · last asked Jul 20 · Design · Mid-level
How would you design a simplified real-time data processing architecture using AWS services and Kafka Streams?
Asked once · last asked Sep 17 · Design · Senior
Other ways it was asked
- “What are the main components of a real-time data processing architecture using AWS and Kafka Streams, and how do they interact?”
What kinds of batch and real-time integrations can Apache NiFi support, and how would you implement them?
Asked once · last asked Feb 6 · Tooling · Mid-level
How do you create a Kafka Connect connector?
Asked once · last asked Feb 18 · Tooling · Mid-level
What audio chunk duration would you choose for speech analysis and translation, and which portion of the audio would you send to the model?
Asked once · last asked Feb 27 · Scenario · Senior
What is Kafka Streams used for, and how can you use it?
Asked once · last asked Mar 6 · Tooling · Mid-level
When migrating from Postgres to MySQL, how would you propagate CDC changes to keep the target system in sync?
Asked once · last asked Mar 20 · Design · Senior
What would you process with Flink versus Spark? Can you differentiate the two and explain how you would use each?
Asked once · last asked May 12 · Concept · Mid-level
Given Kafka data streaming into 40–50 PostgreSQL staging tables, how would you design near-real-time processing that transforms the data within 15 minutes and publishes it to downstream Kafka, without batch jobs or scheduled orchestration?
Asked once · last asked May 18 · Design · Senior
Other ways it was asked
- “Given a system that streams data from Kafka across 40–50 tables into PostgreSQL as a staging layer, how would you design near-real-time processing that transforms the data within 15 minutes and publishes it to a downstream Kafka layer, without relying on batch jobs or scheduled orchestration?”
What design considerations would you keep in mind when building a data integration pipeline that processes data from Kafka, applies business logic, and loads it into target tables or a database?
Asked once · last asked May 19 · Design · Senior
Event-driven system design interview questions
Design questions ask you to connect messaging with microservices, payments, notifications, chat, documents, IoT, and other high-volume workflows. 34 questions.
How would you design and implement a microservices architecture integrated with Kafka?
Asked once · last asked Sep 23 · Design · Senior
Other ways it was asked
- “What microservices architecture would you choose for a system integrated with Kafka?”
How would you design the system and choose the technology stack for the legacy ASP.NET Web Forms application with a SQL database and RabbitMQ?
Asked once · last asked Oct 19 · Design · Senior
Other ways it was asked
- “How would you design the system and choose the technology stack for the legacy ASP.NET Web Forms application with a SQL database, RabbitMQ, device messaging, and APIs, making it multi-tenant?”
How can a payment service notify a delivery service that an order has been paid so delivery can begin?
Asked once · last asked Feb 22 · Scenario · Mid-level
How would you implement messaging in PHP, configure the project, and choose design patterns?
Asked once · last asked Mar 9 · Design · Mid-level
In an asynchronous order workflow, when should the customer be told that an order has been purchased?
Asked once · last asked Mar 19 · Design · Senior
How would you design a notification service that can send 100,000 push notifications per minute across iOS, Android, and web, considering reliability, delivery guarantees, and failed deliveries?
Asked once · last asked Mar 19 · Design · Senior
How would you integrate a synchronous legacy system into an asynchronous microservice?
Asked once · last asked Apr 2 · Design · Senior
How would you architect cross-game sweepstakes credits earned and redeemed across games, using game, payment, and account events plus Kafka, Flink, Spark, Trino, ClickHouse, and Postgres to support near-real-time product analytics, auditable event trails for Compliance, and accurate daily reconciliations for Finance?
Asked once · last asked May 12 · Design · Senior
Other ways it was asked
- “A cross-game feature lets players earn sweepstakes credit in one game and redeem it in another. Product needs near-real-time analytics, Compliance needs an auditable event trail, and Finance needs accurate daily reconciliations. Given events from game, payment, and account services and Kafka, Flink, Spark, Trino, ClickHouse, and Postgres, what data platform architecture would you propose?”
How would you distribute one event to a very large number of users quickly and reliably without overloading a single service?
Asked once · last asked May 23 · Design · Senior
How would you use Kafka to integrate a billing system with other systems?
Asked once · last asked Jun 18 · Design · Senior
How would you design an event-driven solution, including its components and actors, and how would information flow through it?
Asked once · last asked Jun 24 · Design · Senior
If a recipient is offline when a message is sent, how would you handle delivery?
Asked once · last asked Jul 28 · Scenario · Senior
Show 21 more
Can you design a live high-volume activity feed and real-time notification engine using React, Node.js, TypeScript, and SQL that handles more than 10,000 writes per second at peak?
Asked once · last asked Jul 29 · Design · Senior
Walk me through the architecture of the webhook ingestion and fulfillment state machine end to end — from the incoming webhook to the final side effects.
Asked once · last asked Aug 11 · Design · Senior
Walk me through the chronological flow of user-generated actions through an event-processing system.
Asked once · last asked Aug 14 · Design · Senior
How would you design an e-commerce system in .NET with microservices, and would you use synchronous HTTP calls or a message broker like Azure Service Bus? How would you guarantee reliability if one service fails?
Asked once · last asked Sep 1 · Design · Senior
How would you design the end-to-end flow from a sensor sending a measurement to that measurement appearing on a Blazor dashboard?
Asked once · last asked Sep 1 · Design · Senior
How would you design an API, queue, and background workers for a document-processing workflow?
Asked once · last asked Sep 3 · Design · Mid-level
Other ways it was asked
- “How would you design a document-processing workflow in which requests are queued and workers process them?”
- “How would you design an API and choose a queue for a document-processing workflow handled by background workers?”
How would you design a document-processing workflow that supports multiple job types and workers?
Asked once · last asked Sep 3 · Design · Mid-level
Other ways it was asked
- “How would you design a workflow for processing a document when multiple types of jobs need to be performed on it?”
- “How would you design a pattern that supports multiple worker types processing the same queued document job?”
How would you design an architecture for automatically sending messages through email, WhatsApp, Signal, or LinkedIn, and what technologies would you choose?
Asked once · last asked Sep 8 · Design · Mid-level
How would you design an event-driven system, including its components and how they communicate?
Asked once · last asked Sep 10 · Behavioral · Senior
Other ways it was asked
- “Describe an event-driven system you designed, including its components and how they communicate.”
When Kafka is unavailable or lagging, what should your services do at request time to keep authorization safe without causing a major latency spike or outage?
Asked once · last asked Sep 16 · Scenario · Senior
How would you connect the frontend, backend, and services in a chat application that handles text, images, and video, and choose between REST, SignalR, and Service Bus?
Asked once · last asked Sep 18 · Design · Senior
How would you design an architecture to send WhatsApp messages to 10,000 leads under rate limits and track delivery in real time?
Asked once · last asked Oct 1 · Design · Senior
How would you design a real-time system for a cryptocurrency trading platform?
Asked once · last asked Oct 2 · Design · Senior
How would you use Kafka to handle the transfer of thousands of files within a short period?
Asked once · last asked Oct 7 · Scenario · Senior
Other ways it was asked
- “How would Kafka handle thousands of files arriving within a short period?”
How can you detect that a file has arrived in Amazon S3 before publishing an event to Kafka?
Asked once · last asked Oct 7 · Scenario · Senior
What is the best way to handle communication and failures between an order service and a payment service?
Asked once · last asked Oct 9 · Design · Senior
How would you approach load balancing and event-bus design for this application?
Asked once · last asked Oct 19 · Design · Senior
How would you design an event-driven pipeline using Kafka and Claude to handle growth?
Asked once · last asked Oct 20 · Design · Senior
How would you design the event-driven data flow so users are notified when a payment status changes, for example from pending to succeeded?
Asked once · last asked Oct 22 · Design · Senior
Why might you choose a distributed queue and a distributed data store for a production web crawler?
Asked once · last asked Oct 23 · Concept · Senior
How do you collect real-time data from sensors?
Asked once · last asked Sep 22 · Design · Mid-level
Other ways it was asked
- “How are you collecting real-time data from the sensors?”
- “What tools can be used to collect real-time data from sensors?”
Kafka monitoring, scaling, and troubleshooting interview questions
Prepare to discuss cluster setup, capacity, monitoring, incidents, disaster recovery, security, testing, Confluent tools, and infrastructure management. 32 questions.
How would you integrate Kafka with Dynatrace for monitoring?
Asked once · last asked Feb 18 · Tooling · Mid-level
Other ways it was asked
- “What method, such as a webhook, would you use to integrate Kafka with Dynatrace?”
Which Kafka metrics would you display on a Dynatrace dashboard?
Asked once · last asked Feb 18 · Tooling · Mid-level
Other ways it was asked
- “Which Kafka components would you monitor with Dynatrace?”
What approaches do you use to test backend systems, including APIs and Kafka?
Asked once · last asked Jan 27 · Tooling · Mid-level
Which Confluent components extend Apache Kafka, and what would you use them for?
Asked once · last asked Feb 18 · Concept · Mid-level
Other ways it was asked
- “What features and services does Confluent offer beyond Apache Kafka?”
- “What are the major differences between Confluent and Apache Kafka?”
- “Which Confluent components, such as Control Center or connectors, would you use in day-to-day Kafka operations, and what would you use them for?”
If a Kafka cluster has an issue and data is not being processed, how would you investigate and troubleshoot it?
Asked once · last asked Feb 18 · Scenario · Senior
How would you diagnose a Kafka broker outage, and what role do broker logs play in the investigation?
Asked once · last asked Feb 18 · Scenario · Mid-level
How would you investigate a ZooKeeper issue affecting Kafka, and where would you look for diagnostic information?
Asked once · last asked Feb 18 · Scenario · Mid-level
Is a Kafka cluster with two ZooKeeper nodes and four brokers an optimal setup, and what cost considerations would you evaluate?
Asked once · last asked Feb 18 · Scenario · Senior
When would you use a dedicated Kafka broker for disaster recovery?
Asked once · last asked Feb 18 · Scenario · Mid-level
How would you scale a Kafka cluster, including broker capacity, ZooKeeper nodes, and data redistribution?
Asked once · last asked Feb 18 · Scenario · Senior
Other ways it was asked
- “What challenges, particularly with data, would you face when scaling up an existing Kafka cluster?”
- “How would you scale a Kafka cluster by adding ZooKeeper nodes or brokers, and how would you decide how many to add?”
How would you rebalance data across three Kafka brokers with uneven data volumes, which component would you use, and what does two-way balancing mean?
Asked once · last asked Feb 18 · Scenario · Senior
What challenges would you anticipate when setting up a Kafka cluster from scratch?
Asked once · last asked Feb 18 · Behavioral · Mid-level
Other ways it was asked
- “What challenges did you face when setting up a Kafka cluster from scratch?”
Show 20 more
What is an access control list (ACL), and how is it used to control access in Kafka?
Asked once · last asked Feb 18 · Concept · Junior
How would you implement logging and monitoring for an Azure Functions integration using Azure Service Bus?
Asked once · last asked Mar 2 · Tooling · Senior
Monday morning, Compliance flags mismatches in Friday’s daily sweepstakes credit report against Finance’s reconciliation. The product dashboard and payment system disagree, and ClickHouse contains duplicate-looking user records. As the staff engineer escalation point, how would you handle the incident?
Asked once · last asked May 12 · Scenario · Senior
Other ways it was asked
- “On Monday morning, Compliance reports that Friday's daily sweepstakes credit report does not match Finance's reconciliation numbers. The product dashboard shows one number, the payment system shows another, and ClickHouse has duplicate-looking records for some users. As the staff engineer escalation point, how would you handle the incident?”
What would you monitor to detect lag, dropped messages, or slow consumers in a Kafka-based fan-out pipeline?
Asked once · last asked May 23 · Scenario · Senior
Other ways it was asked
- “What would you monitor to detect lag, dropped messages, or slow consumers in that Kafka-based fan-out pipeline?”
How would you design observability across regions so you can trace a single notification from publishing through fan-out to delivery confirmation?
Asked once · last asked May 23 · Design · Senior
How would you pass a trace identifier through the producer, broker, and consumers so logs and metrics can be correlated across services?
Asked once · last asked May 23 · Design · Senior
How would you detect and alert on partial delivery when only some consumer groups lag or fail, without flooding the system with noise?
Asked once · last asked May 23 · Design · Senior
If Kafka consumer lag suddenly increases while CPU remains normal, what would be the first thing you investigate?
Asked once · last asked Jul 16 · Scenario · Senior
How would you configure a Kafka cluster for disaster recovery rather than relying on a single instance?
Asked once · last asked Jul 22 · Design · Senior
If you created an Azure Service Bus using Terraform and a developer manually added queues and subscriptions, how would you detect those infrastructure changes?
Asked once · last asked Aug 4 · Scenario · Mid-level
What happens if an API writes 15,000 records per second to a topic that can handle 10,000 records per second?
Asked once · last asked Sep 10 · Scenario · Mid-level
How would you validate and debug in production that per-aggregate ordering is being preserved end to end across API, database, outbox, broker, and consumer, and what metrics and log fields would you require?
Asked once · last asked Sep 16 · Scenario · Senior
How would you monitor and respond to stuck messages or a growing dead-letter queue in Azure Service Bus?
Asked once · last asked Sep 18 · Scenario · Senior
How would you investigate a Kafka feed that appears healthy but is missing records downstream?
Asked once · last asked Sep 22 · Scenario · Senior
How would you address bottlenecks or performance issues with Amazon SQS, and what architectural changes might you consider?
Asked once · last asked Sep 25 · Scenario · Senior
An external system enqueues 1,000 messages, but your application processes only 997. How would you investigate what happened to the other three?
Asked once · last asked Oct 15 · Scenario · Senior
How would you mitigate the risk of exposing sensitive data in Kafka topics without encryption?
Asked once · last asked Oct 19 · Scenario · Senior
Assume the outbox backlog grows to millions of rows. How would you design storage, retention, and publisher workers so the outbox queries stay fast and don't overload the primary database?
Asked once · last asked Oct 22 · Design · Senior
How do you use message brokers? Do you only read and write messages, or do you also configure and scale them?
Asked once · last asked Oct 23 · Design · Senior
What Kafka mechanisms can support disaster recovery?
Asked once · last asked Jul 22 · Concept · Mid-level
Related interview questions
- Software Engineer interview questions
- Data Engineer interview questions
- AI Engineer interview questions
- DevOps Engineer interview questions
- Solutions Architect interview questions
- Machine Learning Engineer interview questions
- Data Structures and Algorithms interview questions
- OOP and Design Patterns interview questions
- System Design interview questions
- Microservices interview questions
- REST API interview questions
- Caching and Redis interview questions
- Operating Systems interview questions
- Computer Networks interview questions
- Git interview questions
How this list is made
Every question here was asked by a real interviewer in a real job interview that a candidate ran MeetAssist in. After the call, the interviewer's questions are pulled from the transcript. Practice sessions, tests, exams and anything that was not a job interview are dropped.
Questions are rewritten so they make sense on their own: names of people, employers and clients are removed, and speech-to-text mistakes are fixed. Small talk, logistics and questions about one candidate's own résumé are left out. Rephrasings of the same question are merged, and the count is the number of different interviews it came up in.
The interviews come from candidates who use MeetAssist, so they lean towards software, data and AI roles: the shares describe these interviews, not every Kafka and Message Queues interview. Answers to the most asked questions are written and checked by an editor; a "common mistake" is shown only when reviews of real answers in these interviews show it, with the count.
169 interviews, 104 candidates, Jan 2026 to Oct 2026. Updated every month with new interviews.
Or add the Chrome extension