Docker Certified Associate (DCA) Exam 2026: Professional
Certification Practice Test — Orchestration, Image Management,
Installation, Networking, Security & Storage — 190 Multiple-Choice
Questions with Answers & Detailed Rationales
Question 1
A swarm manager node becomes unreachable due to a network partition. The cluster has
five manager nodes total. What is the minimum number of managers that must remain
operational for the swarm to continue processing service updates?
A. 1
B. 2
C. 3
D. 4
Correct Answer: C
Explanation: Raft quorum requires a majority (N/2 + 1). For five managers, at least three
must be available to maintain consensus and process cluster state changes.
Question 2
You are deploying a service across a swarm cluster where specific nodes have SSD storage
and others have HDD. You want database tasks to only schedule on SSD nodes. Which
mechanism accomplishes this requirement with minimal configuration?
,A. Resource reservations in the service definition
B. Node labels combined with placement constraints
C. Separate overlay networks for SSD and HDD nodes
D. Manager-only node designation for SSD hosts
Correct Answer: B
Explanation: Node labels applied to hosts (e.g., storage=ssd) with placement constraints (--
constraint 'node.labels.storage==ssd') direct tasks to appropriate nodes.
Question 3
A developer runs docker run -d nginx on a swarm manager node. What is the state of this
container in relation to the swarm?
A. It becomes a swarm service automatically.
B. It runs as a standalone container outside swarm orchestration.
C. It is rejected because swarm managers cannot run containers.
D. It is scheduled across all worker nodes.
Correct Answer: B
,Explanation: docker run creates standalone containers that are not managed by swarm
orchestration. Only docker service create produces swarm-managed tasks.
Question 4
You need to update a service's image from nginx:1.24 to nginx:1.25 with zero downtime. The
service has 6 replicas. Which command initiates a rolling update with a 2-task-at-a-time
parallelism?
A. docker service update --image nginx:1.25 --update-parallelism 2 my_web
B. docker stack deploy --image nginx:1.25 my_stack
C. docker service scale my_web=6 --image nginx:1.25
D. docker service create --force --image nginx:1.25 my_web
Correct Answer: A
Explanation: docker service update with --update-parallelism controls how many tasks
update simultaneously during a rolling deployment.
Question 5
What happens to running tasks when a worker node is drained using docker node update --
availability drain worker-01?
A. Tasks are immediately killed and not rescheduled.
, B. Tasks continue running until manually stopped.
C. Tasks are gracefully stopped and rescheduled on available nodes.
D. The node is removed from the swarm permanently.
Correct Answer: C
Explanation: Draining a node stops existing tasks and reschedules them on other available
nodes, while preventing new tasks from being assigned to the drained node.
Question 6
You are troubleshooting a service that shows 0/3 replicas running. The service logs show "no
suitable node." Which command provides the most relevant diagnostic information?
A. docker service logs my_service
B. docker service ps my_service --no-trunc
C. docker node ls
D. docker network inspect ingress
Correct Answer: B
Explanation: docker service ps --no-trunc shows task-level scheduling errors, including
placement constraint failures and rejection reasons.
Certification Practice Test — Orchestration, Image Management,
Installation, Networking, Security & Storage — 190 Multiple-Choice
Questions with Answers & Detailed Rationales
Question 1
A swarm manager node becomes unreachable due to a network partition. The cluster has
five manager nodes total. What is the minimum number of managers that must remain
operational for the swarm to continue processing service updates?
A. 1
B. 2
C. 3
D. 4
Correct Answer: C
Explanation: Raft quorum requires a majority (N/2 + 1). For five managers, at least three
must be available to maintain consensus and process cluster state changes.
Question 2
You are deploying a service across a swarm cluster where specific nodes have SSD storage
and others have HDD. You want database tasks to only schedule on SSD nodes. Which
mechanism accomplishes this requirement with minimal configuration?
,A. Resource reservations in the service definition
B. Node labels combined with placement constraints
C. Separate overlay networks for SSD and HDD nodes
D. Manager-only node designation for SSD hosts
Correct Answer: B
Explanation: Node labels applied to hosts (e.g., storage=ssd) with placement constraints (--
constraint 'node.labels.storage==ssd') direct tasks to appropriate nodes.
Question 3
A developer runs docker run -d nginx on a swarm manager node. What is the state of this
container in relation to the swarm?
A. It becomes a swarm service automatically.
B. It runs as a standalone container outside swarm orchestration.
C. It is rejected because swarm managers cannot run containers.
D. It is scheduled across all worker nodes.
Correct Answer: B
,Explanation: docker run creates standalone containers that are not managed by swarm
orchestration. Only docker service create produces swarm-managed tasks.
Question 4
You need to update a service's image from nginx:1.24 to nginx:1.25 with zero downtime. The
service has 6 replicas. Which command initiates a rolling update with a 2-task-at-a-time
parallelism?
A. docker service update --image nginx:1.25 --update-parallelism 2 my_web
B. docker stack deploy --image nginx:1.25 my_stack
C. docker service scale my_web=6 --image nginx:1.25
D. docker service create --force --image nginx:1.25 my_web
Correct Answer: A
Explanation: docker service update with --update-parallelism controls how many tasks
update simultaneously during a rolling deployment.
Question 5
What happens to running tasks when a worker node is drained using docker node update --
availability drain worker-01?
A. Tasks are immediately killed and not rescheduled.
, B. Tasks continue running until manually stopped.
C. Tasks are gracefully stopped and rescheduled on available nodes.
D. The node is removed from the swarm permanently.
Correct Answer: C
Explanation: Draining a node stops existing tasks and reschedules them on other available
nodes, while preventing new tasks from being assigned to the drained node.
Question 6
You are troubleshooting a service that shows 0/3 replicas running. The service logs show "no
suitable node." Which command provides the most relevant diagnostic information?
A. docker service logs my_service
B. docker service ps my_service --no-trunc
C. docker node ls
D. docker network inspect ingress
Correct Answer: B
Explanation: docker service ps --no-trunc shows task-level scheduling errors, including
placement constraint failures and rejection reasons.
