Cohesity Data Protection Associate Exam
ACTUAL EXAM 2026/2027 | Data
Protection Associate | Verified Q&A |
Pass Guaranteed - A+ Graded
Domain 1: Cohesity Architecture & Core Concepts (10 questions)
Q1: What is the minimum number of physical or virtual nodes required to form a production Cohesity
cluster?
A. 1 node
B. 2 nodes
C. 3 nodes [CORRECT]
D. 4 nodes
Correct Answer: C
Rationale: Cohesity clusters require a minimum of 3 nodes for production. Three nodes provide quorum
for distributed consensus, enable replication factor (RF=2 or 3), and allow for rolling upgrades without
downtime. Single node (A) is possible for non-production evaluation only. Two nodes (B) cannot
maintain quorum if one fails. Four nodes (D) is supported but not the minimum.
Q2: Which Cohesity snapshot technology enables space-efficient, copy-on-write snapshots that support
instant clones and live mounts?
A. ZFS snapshots
B. SnapTree [CORRECT]
C. LVM snapshots
D. RAID 10 mirroring
Correct Answer: B
,Rationale: Cohesity uses SnapTree, a proprietary copy-on-write (COW) snapshot structure that allows
efficient snapshots, instant clones, and live mounts with minimal space overhead. Distractor A – ZFS is
not the underlying snapshot mechanism. Distractor C – LVM snapshots are not used. Distractor D – RAID
10 provides redundancy but not snapshot functionality.
Q3: In a Cohesity cluster, what is the typical resource allocation for a Controller VM (CVM) per node?
A. 2 vCPU, 4 GB RAM
B. 4 vCPU, 8 GB RAM
C. 6 vCPU, 12 GB RAM [CORRECT]
D. 8 vCPU, 16 GB RAM
Correct Answer: C
Rationale: The standard CVM configuration is 6 vCPU and 12 GB RAM per node, providing sufficient
resources to handle I/O operations, metadata management, and data services. Option A is insufficient
for production workloads. Option B may work for small configurations but is not standard. Option D is
excessive for typical deployments and not the standard specification.
Q4: Which component of Cohesity's distributed file system (SpanFS) is responsible for maintaining
metadata across the cluster?
A. CVM (Controller VM)
B. Metadata Store (MStore) [CORRECT]
C. Data Plane
D. IO Engine
Correct Answer: B
Rationale: The Metadata Store (MStore) is a distributed metadata service that maintains file system
metadata, directory structures, and object locations across the cluster. While CVMs (A) host services,
MStore specifically handles metadata. The Data Plane (C) handles data movement, and the IO Engine (D)
processes I/O operations but does not maintain metadata persistence.
Q5: What is the purpose of Storage Policy-Based Management (SPBM) in Cohesity?
, A. To encrypt data at rest
B. To define storage characteristics and service levels for protection groups [CORRECT]
C. To manage network bandwidth between nodes
D. To schedule backup jobs
Correct Answer: B
Rationale: SPBM allows administrators to define storage policies that specify characteristics like
replication factor, erasure coding, deduplication, and compression settings. These policies are then
applied to protection groups. Encryption (A) is handled separately. Network bandwidth (C) is managed
by QoS settings. Job scheduling (D) is part of SLA policies, not SPBM.
Q6: (Select all that apply): Which replication factors are supported in a production Cohesity cluster?
Select all that apply.
A. RF=1 [CORRECT]
B. RF=2 [CORRECT]
C. RF=3 [CORRECT]
D. RF=4
Correct Answer Set: A, B, C
Rationale: Cohesity supports RF=1 (single copy, used in specific scenarios with erasure coding), RF=2
(two copies for standard redundancy), and RF=3 (three copies for critical data). RF=4 (D) is not a
standard supported replication factor as it provides diminishing returns on redundancy versus storage
overhead.
Q7: What happens to cluster operations if a single node fails in a 4-node Cohesity cluster configured
with RF=2?
A. All operations halt until the node is restored
B. The cluster continues operating with reduced performance but full data availability
[CORRECT]
C. Data becomes temporarily unavailable for 30 minutes
D. Automatic failover to a secondary cluster occurs
ACTUAL EXAM 2026/2027 | Data
Protection Associate | Verified Q&A |
Pass Guaranteed - A+ Graded
Domain 1: Cohesity Architecture & Core Concepts (10 questions)
Q1: What is the minimum number of physical or virtual nodes required to form a production Cohesity
cluster?
A. 1 node
B. 2 nodes
C. 3 nodes [CORRECT]
D. 4 nodes
Correct Answer: C
Rationale: Cohesity clusters require a minimum of 3 nodes for production. Three nodes provide quorum
for distributed consensus, enable replication factor (RF=2 or 3), and allow for rolling upgrades without
downtime. Single node (A) is possible for non-production evaluation only. Two nodes (B) cannot
maintain quorum if one fails. Four nodes (D) is supported but not the minimum.
Q2: Which Cohesity snapshot technology enables space-efficient, copy-on-write snapshots that support
instant clones and live mounts?
A. ZFS snapshots
B. SnapTree [CORRECT]
C. LVM snapshots
D. RAID 10 mirroring
Correct Answer: B
,Rationale: Cohesity uses SnapTree, a proprietary copy-on-write (COW) snapshot structure that allows
efficient snapshots, instant clones, and live mounts with minimal space overhead. Distractor A – ZFS is
not the underlying snapshot mechanism. Distractor C – LVM snapshots are not used. Distractor D – RAID
10 provides redundancy but not snapshot functionality.
Q3: In a Cohesity cluster, what is the typical resource allocation for a Controller VM (CVM) per node?
A. 2 vCPU, 4 GB RAM
B. 4 vCPU, 8 GB RAM
C. 6 vCPU, 12 GB RAM [CORRECT]
D. 8 vCPU, 16 GB RAM
Correct Answer: C
Rationale: The standard CVM configuration is 6 vCPU and 12 GB RAM per node, providing sufficient
resources to handle I/O operations, metadata management, and data services. Option A is insufficient
for production workloads. Option B may work for small configurations but is not standard. Option D is
excessive for typical deployments and not the standard specification.
Q4: Which component of Cohesity's distributed file system (SpanFS) is responsible for maintaining
metadata across the cluster?
A. CVM (Controller VM)
B. Metadata Store (MStore) [CORRECT]
C. Data Plane
D. IO Engine
Correct Answer: B
Rationale: The Metadata Store (MStore) is a distributed metadata service that maintains file system
metadata, directory structures, and object locations across the cluster. While CVMs (A) host services,
MStore specifically handles metadata. The Data Plane (C) handles data movement, and the IO Engine (D)
processes I/O operations but does not maintain metadata persistence.
Q5: What is the purpose of Storage Policy-Based Management (SPBM) in Cohesity?
, A. To encrypt data at rest
B. To define storage characteristics and service levels for protection groups [CORRECT]
C. To manage network bandwidth between nodes
D. To schedule backup jobs
Correct Answer: B
Rationale: SPBM allows administrators to define storage policies that specify characteristics like
replication factor, erasure coding, deduplication, and compression settings. These policies are then
applied to protection groups. Encryption (A) is handled separately. Network bandwidth (C) is managed
by QoS settings. Job scheduling (D) is part of SLA policies, not SPBM.
Q6: (Select all that apply): Which replication factors are supported in a production Cohesity cluster?
Select all that apply.
A. RF=1 [CORRECT]
B. RF=2 [CORRECT]
C. RF=3 [CORRECT]
D. RF=4
Correct Answer Set: A, B, C
Rationale: Cohesity supports RF=1 (single copy, used in specific scenarios with erasure coding), RF=2
(two copies for standard redundancy), and RF=3 (three copies for critical data). RF=4 (D) is not a
standard supported replication factor as it provides diminishing returns on redundancy versus storage
overhead.
Q7: What happens to cluster operations if a single node fails in a 4-node Cohesity cluster configured
with RF=2?
A. All operations halt until the node is restored
B. The cluster continues operating with reduced performance but full data availability
[CORRECT]
C. Data becomes temporarily unavailable for 30 minutes
D. Automatic failover to a secondary cluster occurs
