Cloudera Certified Developer for Apache Hadoop (CCDH) Practice
Exam 2026: Professional Certification Test — Hadoop, HDFS,
MapReduce, YARN, Data Processing & Cluster Optimization — 190
Multiple-Choice Questions with Answers & Detailed Rationales
Section 1: Hadoop Infrastructure & HDFS Architecture (Questions 1-38)
Question 1
Which HDFS daemon is responsible for storing the actual data blocks on the local file
system?
A. NameNode
B. DataNode
C. SecondaryNameNode
D. ResourceManager
Correct Answer: B
Explanation: The DataNode stores actual data blocks on local disk and serves read/write
requests from clients. NameNode manages metadata only.
Question 2
A client application needs to read a 500MB file from HDFS. The file was written with a block
size of 128MB. How many blocks comprise this file?
A. 2 blocks
,B. 3 blocks
C. 4 blocks
D. 5 blocks
Correct Answer: C
Explanation: 500MB divided by 128MB yields 3 full blocks (384MB) plus one partial block
(116MB), totaling 4 blocks.
Question 3
What is the primary function of the SecondaryNameNode in HDFS?
A. Provides hot standby for NameNode failover
B. Performs checkpointing of the namespace by merging fsimage and edit logs
C. Stores replicas of all data blocks for fault tolerance
D. Manages block allocation across the cluster
Correct Answer: B
Explanation: SecondaryNameNode periodically merges the namespace image with edit logs
to prevent unlimited log growth, reducing NameNode restart time.
Question 4
During HDFS write operations, what is the correct sequence of client actions?
A. Client requests block locations, receives DataNode list, writes directly to first DataNode
,B. Client writes all data to NameNode, which distributes blocks to DataNodes
C. Client writes to SecondaryNameNode, which forwards blocks to DataNodes
D. Client queries all DataNodes in parallel and writes to each directly
Correct Answer: A
Explanation: Client requests block locations from NameNode, receives a prioritized list of
DataNodes, then writes data directly to the first DataNode in the pipeline.
Question 5
A Hadoop cluster has a default replication factor of 3. If the cluster spans 2 racks, what is
Hadoop's default replica placement policy?
A. All 3 replicas on the same rack
B. 2 replicas on one rack, 1 replica on another rack
C. All 3 replicas distributed evenly across both racks
D. 1 replica on first rack, 2 replicas on second rack
Correct Answer: B
Explanation: Default placement puts first replica on local rack (or writer node), second on a
remote rack, third on the same remote rack for write performance.
Question 6
What happens when a DataNode fails to send a heartbeat to the NameNode within the
configured timeout period?
, A. The NameNode immediately deletes all data on that DataNode
B. The DataNode is marked as dead and its blocks are scheduled for re-replication
C. The NameNode promotes the SecondaryNameNode to active status
D. The cluster shuts down until the DataNode recovers
Correct Answer: B
Explanation: After heartbeat timeout, NameNode marks the DataNode dead and initiates re-
replication of its blocks to maintain replication factor.
Question 7
In a Hadoop cluster with rack awareness configured, what is the primary benefit for
MapReduce job execution?
A. Increased NameNode memory efficiency
B. Reduced network traffic by scheduling tasks on nodes with local data
C. Faster block replication across DataNodes
D. Automatic load balancing of reduce tasks
Correct Answer: B
Explanation: Rack awareness enables MapReduce to schedule map tasks on nodes
containing input data, minimizing cross-rack network transfer.
Question 8
Which command correctly uploads a local file named data.txt to an HDFS directory
/user/hadoop/input?
Exam 2026: Professional Certification Test — Hadoop, HDFS,
MapReduce, YARN, Data Processing & Cluster Optimization — 190
Multiple-Choice Questions with Answers & Detailed Rationales
Section 1: Hadoop Infrastructure & HDFS Architecture (Questions 1-38)
Question 1
Which HDFS daemon is responsible for storing the actual data blocks on the local file
system?
A. NameNode
B. DataNode
C. SecondaryNameNode
D. ResourceManager
Correct Answer: B
Explanation: The DataNode stores actual data blocks on local disk and serves read/write
requests from clients. NameNode manages metadata only.
Question 2
A client application needs to read a 500MB file from HDFS. The file was written with a block
size of 128MB. How many blocks comprise this file?
A. 2 blocks
,B. 3 blocks
C. 4 blocks
D. 5 blocks
Correct Answer: C
Explanation: 500MB divided by 128MB yields 3 full blocks (384MB) plus one partial block
(116MB), totaling 4 blocks.
Question 3
What is the primary function of the SecondaryNameNode in HDFS?
A. Provides hot standby for NameNode failover
B. Performs checkpointing of the namespace by merging fsimage and edit logs
C. Stores replicas of all data blocks for fault tolerance
D. Manages block allocation across the cluster
Correct Answer: B
Explanation: SecondaryNameNode periodically merges the namespace image with edit logs
to prevent unlimited log growth, reducing NameNode restart time.
Question 4
During HDFS write operations, what is the correct sequence of client actions?
A. Client requests block locations, receives DataNode list, writes directly to first DataNode
,B. Client writes all data to NameNode, which distributes blocks to DataNodes
C. Client writes to SecondaryNameNode, which forwards blocks to DataNodes
D. Client queries all DataNodes in parallel and writes to each directly
Correct Answer: A
Explanation: Client requests block locations from NameNode, receives a prioritized list of
DataNodes, then writes data directly to the first DataNode in the pipeline.
Question 5
A Hadoop cluster has a default replication factor of 3. If the cluster spans 2 racks, what is
Hadoop's default replica placement policy?
A. All 3 replicas on the same rack
B. 2 replicas on one rack, 1 replica on another rack
C. All 3 replicas distributed evenly across both racks
D. 1 replica on first rack, 2 replicas on second rack
Correct Answer: B
Explanation: Default placement puts first replica on local rack (or writer node), second on a
remote rack, third on the same remote rack for write performance.
Question 6
What happens when a DataNode fails to send a heartbeat to the NameNode within the
configured timeout period?
, A. The NameNode immediately deletes all data on that DataNode
B. The DataNode is marked as dead and its blocks are scheduled for re-replication
C. The NameNode promotes the SecondaryNameNode to active status
D. The cluster shuts down until the DataNode recovers
Correct Answer: B
Explanation: After heartbeat timeout, NameNode marks the DataNode dead and initiates re-
replication of its blocks to maintain replication factor.
Question 7
In a Hadoop cluster with rack awareness configured, what is the primary benefit for
MapReduce job execution?
A. Increased NameNode memory efficiency
B. Reduced network traffic by scheduling tasks on nodes with local data
C. Faster block replication across DataNodes
D. Automatic load balancing of reduce tasks
Correct Answer: B
Explanation: Rack awareness enables MapReduce to schedule map tasks on nodes
containing input data, minimizing cross-rack network transfer.
Question 8
Which command correctly uploads a local file named data.txt to an HDFS directory
/user/hadoop/input?
