100+ Free CCOAK Practice Questions

Pass your Confluent Certified Operator for Apache Kafka (CCOAK) exam on the first try — instant access, no signup required.

✓ No registration✓ No credit card✓ No hidden fees✓ Start practicing immediately

~70-80% Pass Rate

100+ Questions

100% Free

1 / 10

Question 1

Score: 0/0

A Kafka cluster is experiencing high request latency. The num.io.threads pool is saturated. Which is the most appropriate first response?

Increase num.io.threads incrementally and monitor handler utilization

Restart all brokers immediately

Reduce replication factor

Decrease min.insync.replicas

to track

2026 Statistics

Key Facts: CCOAK Exam

~60

Exam Questions

Confluent

90 min

Exam Duration

Confluent

$150

Exam Fee

Confluent

2 years

Validity

Confluent

Production

Ops Focus

Confluent

PSI Online

Delivery

Confluent

The CCOAK exam has approximately 60 multiple-choice, matching, and list-order questions in 90 minutes. It focuses on production-operations topics: capacity planning, JVM/OS tuning, broker performance, security operations, incident response, backup/DR, upgrades, and Kubernetes operations. Exam fee is $150 USD, delivered via PSI online proctored. Certification is valid for 2 years. Must be completed within 12 months from date of purchase.

Sample CCOAK Practice Questions

Try these sample questions to test your CCOAK exam readiness. Each question includes a detailed explanation. Start the interactive quiz above for the full 100+ question experience with AI tutoring.

1A Kafka cluster is experiencing high request latency. The num.io.threads pool is saturated. Which is the most appropriate first response?

A.Increase num.io.threads incrementally and monitor handler utilization

B.Restart all brokers immediately

C.Reduce replication factor

D.Decrease min.insync.replicas

Explanation: When I/O handler threads are saturated (low IdlePercent), the first tuning step is to increase num.io.threads. Restarting blindly or reducing replication/ISR settings damages durability and does not address the CPU bottleneck. Correlate the change with broker CPU to avoid over-provisioning.

2Which metric is the clearest early warning that a broker's I/O handler threads are saturated?

A.RequestHandlerAvgIdlePercent below 0.3

B.BytesInPerSec rising

C.MessagesInPerSec decreasing

D.UnderReplicatedPartitions = 0

Explanation: RequestHandlerAvgIdlePercent measures how much time the handler threads are idle. A sustained value below 0.3 means handlers are over 70% busy and requests are queuing. This precedes latency regressions and is a primary signal for capacity scaling.

3A production Kafka cluster is running G1GC. Which JVM option is most relevant to reducing GC pause times?

A.-XX:MaxGCPauseMillis=20

B.-XX:+UseSerialGC

C.-Xmx64g

D.-XX:+UseConcMarkSweepGC

Explanation: -XX:MaxGCPauseMillis=20 (Confluent's recommended value) sets the G1GC target pause goal. Kafka is latency-sensitive so low pause times are essential. Oversizing heap (-Xmx too big) actually hurts performance by leaving less OS page cache; CMS is deprecated.

4What is the typical recommended heap size range for a Kafka broker in production?

A.1-2 GB

B.6-8 GB with the rest of RAM reserved for the page cache

C.64 GB minimum

D.As much heap as the machine has

Explanation: Kafka brokers typically use a 6-8 GB heap; the rest of system memory should be available to the OS page cache because Kafka's read path depends on cached log segments. Over-sized heaps starve the page cache and increase GC pause time.

5Which Linux kernel setting is typically increased on Kafka brokers to support many open files and network connections?

A.vm.swappiness

B.net.ipv4.ip_local_port_range

C.nofile and nproc ulimits

D.vm.overcommit_memory

Explanation: Kafka brokers keep many file descriptors open (log segments, sockets). Production deployments raise nofile (open files) and nproc (processes/threads) ulimits, typically to 100000+. vm.swappiness=1 is also commonly set to minimize swap use.

6Which is the recommended approach to reduce cross-AZ network costs in a multi-AZ Kafka cluster?

A.Reduce replication factor to 1

B.Use KIP-392 fetch-from-follower with a rack-aware replica selector

C.Disable consumers in other AZs

D.Increase partition count

Explanation: KIP-392 allows consumers to fetch from a follower in the same rack/AZ instead of the leader, dramatically reducing cross-AZ data egress. Set replica.selector.class=org.apache.kafka.common.replica.RackAwareReplicaSelector on the broker and client.rack on consumers.

7A broker has crashed and is offline. Which Kafka JMX metric should immediately alert you to this?

A.MessagesInPerSec

B.OfflinePartitionsCount > 0 or UnderReplicatedPartitions spike

C.RequestQueueSize = 0

D.BytesOutPerSec declining slightly

Explanation: When a broker fails, partitions it led will momentarily show OfflinePartitionsCount > 0 until a new leader is elected, and replicas on it will show UnderReplicatedPartitions until it comes back or is reassigned. These should be page-alerted.

8What happens to active-producer traffic during a rolling upgrade if acks=all and min.insync.replicas=2 on a three-replica topic?

A.Traffic stops

B.Traffic continues—one broker down still leaves 2 in-sync replicas meeting min.insync.replicas

C.acks=all upgrades are impossible

D.Data is duplicated

Explanation: With replication factor 3 and min.insync.replicas=2, one broker can be offline while producers with acks=all still write successfully (2 replicas in-sync). This is the core reason for choosing RF=3 and min.insync.replicas=2 in production—it supports maintenance without producer errors.

9Which tool automates partition rebalancing and broker workload optimization in a Kafka cluster?

A.kafkacat

B.Cruise Control with goals like ReplicaDistributionGoal and LeaderBytesInDistributionGoal

C.kafka-topics.sh

D.Schema Registry

Explanation: Cruise Control (open-source from LinkedIn, integrated with Confluent Platform) monitors cluster state and applies analyzer goals to generate and execute reassignment plans. Goals include replica/leader distribution, disk usage, and network bandwidth balance.

10Which mechanism in MirrorMaker 2 translates consumer group offsets across clusters for failover?

A.mm2-status topic

B.MirrorCheckpointConnector emitting offset translations to the __consumer_offsets on the target

C.mm2-configs

D.MirrorSourceConnector heartbeats only

Explanation: MirrorCheckpointConnector writes consumer group offset checkpoints from the source cluster to the target cluster, translating offsets via the offset-sync topic. In failover, the target cluster's consumers can resume from approximately the same logical position.

About the CCOAK Exam

The Confluent Certified Operator for Apache Kafka (CCOAK) exam validates production-operations expertise for running Apache Kafka at scale. It tests capacity planning, performance tuning (JVM, OS, broker threads), incident response, backup and disaster recovery (MirrorMaker 2, Cluster Linking), security operations, upgrades (rolling, ZK-to-KRaft), Kubernetes operations (Confluent for Kubernetes / Strimzi), and advanced troubleshooting.

Questions

60 scored questions

Time Limit

90 minutes

Passing Score

Scaled (not publicly disclosed)

Exam Fee

$150 USD (Confluent / PSI Online Proctored)

CCOAK Exam Content Outline

~25%

Performance & Tuning

Capacity planning (broker count, partition limits), JVM tuning (G1GC, heap sizing), OS tuning (ulimits, swappiness), broker threads (num.io.threads, num.network.threads)

~20%

Incident Response & Troubleshooting

Broker failures, ISR shrinks, unclean leader elections, controller failover, error codes (NOT_LEADER_FOR_PARTITION, CORRUPT_MESSAGE), consumer lag remediation

~15%

Backup & Disaster Recovery

MirrorMaker 2 with MirrorSource/Checkpoint/Heartbeat connectors, Cluster Linking, multi-AZ stretched clusters, rack-aware replica placement, snapshot strategies

~15%

Upgrades & Migration

Rolling upgrades with inter.broker.protocol.version, ZK-to-KRaft migration, version compatibility matrix, controlled.shutdown

~15%

Security Operations

TLS certificate rotation, SCRAM credential management, SASL_PLAIN to SASL_SCRAM migration, ACLs, audit logging, Confluent RBAC

~10%

Monitoring & Observability

JMX Exporter, Prometheus + Grafana, Confluent Control Center, Cruise Control, Health+, quota metrics

How to Pass the CCOAK Exam

What You Need to Know

Passing score: Scaled (not publicly disclosed)
Exam length: 60 questions
Time limit: 90 minutes
Exam fee: $150 USD

Keys to Passing

Complete 500+ practice questions
Score 80%+ consistently before scheduling
Focus on highest-weighted sections
Use our AI tutor for tough concepts

CCOAK Study Tips from Top Performers

1Prioritize performance tuning: JVM heap (6-8 GB), G1GC with -XX:MaxGCPauseMillis=20, OS ulimits (nofile), swappiness

2Learn disaster recovery patterns: Cluster Linking (byte-identical) vs MirrorMaker 2 (offset translation)

3Understand incident response playbooks: URP spikes, ISR shrinks, controller failover, split-brain

4Study the ZK-to-KRaft migration procedure in detail—bridge release, migration=true, rolling steps

5Know monitoring stack integrations: jmx_exporter, Prometheus, Grafana, Confluent Health+

6Practice partition reassignment with --throttle to avoid client impact

7Review Confluent for Kubernetes (CFK) CRDs: KafkaCluster, Connect, SchemaRegistry, ControlCenter

Frequently Asked Questions

What is the CCOAK exam?

The Confluent Certified Operator for Apache Kafka (CCOAK) is Confluent's production-operations-focused certification. It validates the ability to deploy, tune, monitor, and operate Kafka clusters reliably at scale, including incident response and disaster recovery.

How does CCOAK differ from CCAAK?

CCAAK focuses on administrator fundamentals (configuring and supporting clusters), while CCOAK emphasizes production operations at scale (performance tuning, capacity planning, DR, multi-cluster, upgrades). CCOAK is geared toward SREs and senior Kafka operators.

How many questions are on the CCOAK exam?

The CCOAK exam has approximately 60 questions in 90 minutes, delivered via PSI online proctoring. Question types include multiple-choice, matching, and list-order. Candidates have 12 months from purchase to complete the exam.

Are there prerequisites for the CCOAK exam?

There are no formal prerequisites. Confluent strongly recommends significant hands-on production experience operating Kafka, including multi-AZ deployments, monitoring with JMX/Prometheus, rolling upgrades, disaster recovery testing, and incident response.

How should I prepare for the CCOAK exam?

Plan 50-80 hours of study over 6-10 weeks. Stand up a multi-broker Kafka cluster with TLS and SASL, practice rolling upgrades, run MirrorMaker 2 replication, use Cruise Control for rebalancing, simulate broker failures, and tune JVM/OS. Read Confluent's operations guide and complete 100+ practice questions.

What jobs can I get with CCOAK certification?

CCOAK supports senior roles including Senior Kafka Engineer, Site Reliability Engineer (SRE), Platform Architect for streaming, Principal DevOps Engineer, and Kafka Specialist. It signals you can own Kafka in production at scale.