Confluent Certified Developer for Apache Kafka (CCDAK) · 15% practice allocation

Confluent CCDAK: Application Observability Practice Questions

5 free sample questions from the Application Observability domain, with answers and explanations. The full CCDAK bank has 58 questions across 4 domains. Questions are scenario-based and written from the official Confluent exam guide.

1.

Kafka clients (producers, consumers, Streams) expose runtime metrics primarily through which mechanism?

  • A. JMX (Java Management Extensions) metrics✓ Correct
  • B. DNS records
  • C. S3 access logs
  • D. The Schema Registry API
Explanation

Kafka clients expose metrics via JMX, which monitoring systems scrape (often through exporters) for throughput, latency, and lag. The Schema Registry API (D) serves schemas, S3 logs (C) and DNS (B) are unrelated to client metrics.

2.

A producer's record-error-rate metric is rising. What is an appropriate first observability-driven response?

  • A. Investigate the errors (via logs and broker/ISR status) since records may be failing to be produced, then remediate✓ Correct
  • B. Ignore it because producers always retry
  • C. Immediately delete the topic
  • D. Reduce the number of partitions
Explanation

A rising producer error rate means produce attempts are failing; investigate via logs and broker/ISR health, then remediate (e.g., configuration, capacity, or availability issues). Ignoring it (B) risks data loss, deleting the topic (C) is destructive, and reducing partitions (D) doesn't address the errors.

3.

Which metric helps determine whether a consumer is committing offsets successfully and not repeatedly reprocessing the same records?

  • A. The number of brokers
  • B. Commit rate / commit latency metrics (and stable committed offsets advancing over time)✓ Correct
  • C. The Schema Registry's port number
  • D. The producer's compression type
Explanation

Offset commit metrics (rate, latency) and advancing committed offsets show that a consumer is committing progress and not stuck reprocessing. Compression type (D), broker count (A), and Schema Registry port (C) don't reflect commit health.

4.

Which producer metric would you watch to detect that records are taking too long to be acknowledged?

  • A. Consumer group count
  • B. Request latency (e.g., request-latency-avg/max)✓ Correct
  • C. The number of topics in the cluster
  • D. The broker's hostname
Explanation

Producer request latency metrics reveal how long produce requests take to be acknowledged; rising latency signals broker or network issues. Topic count (C), consumer-group count (A), and hostname (D) don't measure produce acknowledgment time.

5.

A Kafka Streams application intermittently fails to process records and logs deserialization exceptions. What observability practice helps diagnose this FASTEST?

  • A. Structured logging of the error with context (topic, partition, offset) plus metrics/alerts on error rates✓ Correct
  • B. Only checking the broker disk usage
  • C. Ignoring the errors since Streams retries automatically
  • D. Turning off all logging to reduce noise
Explanation

Structured logs capturing the topic/partition/offset of failing records, plus metrics and alerts on error rates, pinpoint the problem quickly. Disabling logging (D) removes the diagnostic trail, disk usage (B) is unrelated to deserialization, and ignoring errors (C) risks data loss and doesn't diagnose the cause.

Test yourself across every CCDAK domain

The free CCDAK practice test mixes about 20 questions weighted like the exam. The full 58-question bank adds instant grading, flashcards and a timed mock exam.

Study this domain

Application Observability study guide (CCDAK) →

Other CCDAK domains

← All Confluent Certified Developer for Apache Kafka study resources