Country or region and language

    Choose your country or region and your language.

    Suggested for you

      All countries and regions

        Choose your character

        It shows in the top bar and on your profile. Change it any time.

        Interviews in Apache Kafka

        Work in the tool while you explain your decisions.

        Choose a role and level. Each interview has a new task, follow-up questions and a report that checks your work and your explanation.

        Speak or type. Sign in to save your results across devices.

        Timed assessment

        • 6 practical questions
        • 35 minutes
        • Mid level

        Questions are drawn from a bank of 18. Your report includes a skill breakdown and lessons to practice next.

        Practice interviews

        Junior tasks cover the fundamentals. Mid and senior tasks add debugging, changing requirements and review decisions.

        6 interviews

        • Data engineer
        • Junior
        • 22 min

        A replicated dispatch journal

        Create the assigned topic with the assigned partition count and three replicas per partition. Set min.insync.replicas to 2. Configure hub_writer with acks=all and enable.idempotence=true. These are local model settings; they do not establish a measured production service level. The assigned topic is hub_dispatch_71, the parcel key is parcel_71, the value is 271, and the partition count is 5.

        Start interview
        Skills and rubricReplication and acknowledgments
        • I create the assigned partition count with replication factor three and combine acks=all with minimum ISR two because acknowledgment must satisfy the configured in-sync durability boundary. I inspect replica assignments and the producer settings before accepting the configuration.
        • Complete the requested work and identify the source fact that makes it valid.
        • I distinguish all current in-sync replicas from the minimum ISR threshold because acks=all does not mean an arbitrary fixed count of acknowledgments. I test a reduced ISR rather than infer durability from replication factor alone.
        • Data engineer
        • Junior
        • 22 min

        The next record after a checked batch

        The assigned single-partition topic contains a batch of records. Join hub_group, process every supplied record with the local process operation, then commit the next offset after the processed batch. Keep the member active. Inspect the log end before choosing the count. The assigned topic is hub_dispatch_84, the parcel key is parcel_84, the value is 284, and the partition count is 1.

        Start interview
        Skills and rubricCommitted offsets
        • I process the records before committing the next offset because committing the last processed record’s own offset would replay that record. I compare processed evidence, consumer position and the committed next offset instead of treating a low lag as proof of completed work.
        • Complete the requested work and identify the source fact that makes it valid.
        • I separate fetch position from durable processing because polling a record does not prove its external effect committed. I also check the next-offset boundary so the first unprocessed record remains available after restart.
        • Data engineer
        • Mid
        • 30 min

        A repeated producer sequence

        Configure hub_writer for idempotence with acks=all. Write the assigned keyed value to partition 0 using sequence 0, then repeat that same sequence once as a simulated producer retry. Exactly one record should remain. Use the same producer identity for both sends. The assigned topic is hub_dispatch_97, the parcel key is parcel_97, the value is 297, and the partition count is 1.

        Start interview
        Skills and rubricProducer idempotence
        • I distinguish an identical producer retry from a new application send because reusing a business key does not by itself remove a second record. A new producer identity or sequence needs its own business deduplication rule.
        • I treat it as a separate producer operation because producer idempotence is scoped to its protocol identity and sequence. I use an explicit event ID and destination uniqueness rule if the business contract requires deduplication across producer restarts.
        • Use a counterexample and a readback to defend the result under changed inputs.
        • Data engineer
        • Mid
        • 30 min

        Two updates for the same parcel

        Send the assigned parcel’s created event followed by its signed event to partition 0 with the same parcel key. Keep their values and order exact. The topic has several partitions; do not claim a global ordering guarantee from this per-partition trace. The assigned topic is hub_dispatch_110, the parcel key is parcel_110, the value is 310, and the partition count is 4.

        Start interview
        Skills and rubricPartition ordering
        • I check both event values and their offsets because the right key on reversed updates still violates the required lifecycle. Explicit partition selection in this exercise makes the boundary visible; production key routing also depends on the partitioner and partition count.
        • I review key-to-partition routing and the cutover boundary because increasing partition count can route future records for a key differently from old ones. I use a controlled migration or application ordering contract instead of assuming cross-partition order.
        • Use a counterexample and a readback to defend the result under changed inputs.
        • Data engineer
        • Senior
        • 38 min

        A replacement worker after departure

        hub_group currently has member old_worker on the assigned topic. Join replacement_worker, then remove old_worker. Keep replacement_worker active and verify it owns every assigned partition. Do not create a second group to imitate a rebalance. The assigned topic is hub_dispatch_123, the parcel key is parcel_123, the value is 323, and the partition count is 5.

        Start interview
        Skills and rubricConsumer group rebalances
        • I coordinate revocation with pending work and use a durable destination idempotency rule because group ownership alone cannot cancel an already committed external effect. I commit only completed work and verify behavior at the handoff boundary.
        • I retain the original request, input identities, code or commands, checked output and run evidence because another engineer needs a reproducible result. I name the remaining assumption and the next check that could change it. The browser model cannot establish production performance or external system behavior.
        • Define the replay or recovery boundary and the evidence required before publication.
        • Data engineer
        • Senior
        • 38 min

        A returned replica outside the ISR

        The assigned replicated topic has broker 0 down. Restore broker 0, then run the local replicate operation so every partition returns to three in-sync replicas. Keep the existing records and replication factor. Broker up alone does not complete this teaching recovery. The assigned topic is hub_dispatch_136, the parcel key is parcel_136, the value is 336, and the partition count is 2.

        Start interview
        Skills and rubricReplica recovery
        • I keep the availability and potential data-loss decision explicit because unclean leader election can lose acknowledged records. I preserve the logs, identify the recoverable replicas and obtain the required recovery authority instead of treating a green leader as proof of complete data.
        • I retain the original request, input identities, code or commands, checked output and run evidence because another engineer needs a reproducible result. I name the remaining assumption and the next check that could change it. The browser model cannot establish production performance or external system behavior.
        • Define the replay or recovery boundary and the evidence required before publication.

        Interview Hub access

        The Interview Hub is included in All Access. Choose a plan or sign in to your account to continue.

        View plansSign in or join

        Your Work Gym account

        Save your progress and pick up on another device. Your browser's progress comes with you.

        or

        You can keep practicing without an account.

        Change your plan

        Checking the price

        Help and feedback

        Ask a question, tell us what is not working, or share an idea. We reply by email.

        What is it about?

        We use it only to reply to you.