Knowledge + Practice

CCNA Configure and Manage vSphere Storage Questions

75 of 78 questions · Page 1/2 · Configure and Manage vSphere Storage · Answers revealed

Practice these questions Domain overview All questions

1

MCQhard

Refer to the exhibit. An administrator runs vmkfstools -Ph on a VSAN datastore. The output shows VMFS version 3.61 and a capacity of 2.0 TB. Which statement is true about this datastore?

A.The datastore is a VMFS datastore that has been upgraded to version 3.61.

B.The datastore is a Virtual Volumes datastore, which uses VMFS 3.61.

C.The datastore is an NFS datastore mounted as VMFS 3.61.

D.The datastore is a vSAN datastore, and the VMFS version shown is for compatibility with VMFS tools.

AnswerD

vSAN datastores appear as VMFS 3.61 to allow tools like vmkfstools to read them.

Why this answer

The exhibit shows a VSAN datastore displayed as VMFS version 3.61. vSAN datastores are presented as VMFS version 3.61 to maintain compatibility with VMFS tools. The actual vSAN object store is not formatted with VMFS, but vmkfstools reports this version for compatibility.

Practice this question →

2

MCQeasy

An administrator needs to expand a VMFS5 datastore that is currently 2 TB in size. The underlying LUN is 5 TB. What is the maximum size the datastore can be expanded to without involving additional LUNs?

A.64 TB

B.5 TB

C.4 TB

D.2 TB

AnswerB

Correct: The datastore can use the entire 5 TB LUN.

Why this answer

VMFS5 supports datastores up to 64 TB, but the underlying LUN size limits the datastore expansion. The LUN is 5 TB, so the datastore can be expanded to the full LUN size of 5 TB. Option A (64 TB) is incorrect because it exceeds the LUN capacity.

Option C (4 TB) is incorrect because it is less than the available LUN size. Option D (2 TB) is incorrect because it is the current size, not the maximum expansion.

Practice this question →

3

MCQmedium

An administrator runs vmkfstools on a VMFS datastore and receives the output shown. The datastore is backed by a single LUN from a SAN array. What is the most likely explanation for the multiple extents shown?

A.The LUN was presented to the host with multiple paths, and each path is treated as a separate extent.

B.The datastore was expanded by adding an extent from the same LUN.

C.The datastore is a VMFS6 datastore with three logical volumes aggregated into one namespace.

D.The virtual machine has a large VMDK file that spans multiple extents due to VMFS6 sub-block allocation.

AnswerB

When expanding a VMFS datastore, the system can add extents from the same LUN, leading to multiple extents as shown.

Why this answer

When a VMFS datastore is expanded using the same LUN, vmkfstools can show multiple extents because the additional space is added as a separate extent on the same physical LUN. This is a common practice to grow a datastore without reprovisioning a new LUN, and the output reflects the original and expanded extents both pointing to the same LUN identifier.

Exam trap

The trap here is that candidates confuse multipathing (multiple paths to a LUN) with multiple extents, or assume that VMFS6's sub-block allocation creates extents for VMDK files, when in fact extents are a datastore-level construct, not a virtual disk feature.

How to eliminate wrong answers

Option A is wrong because multiple paths to the same LUN are managed by the host's multipathing plugin (e.g., NMP or SATP) and do not create separate extents; vmkfstools would show a single extent with multiple paths. Option C is wrong because VMFS6 does not aggregate multiple logical volumes into one namespace; it uses a single volume with sub-blocks and block sizes, and extents are not logical volumes. Option D is wrong because a large VMDK file spanning extents is a characteristic of VMFS2/3's extent-based file system, not VMFS6, and vmkfstools output showing multiple extents refers to the datastore's extents, not the virtual disk's allocation.

Practice this question →

4

MCQeasy

What is the maximum number of paths that vSphere supports for a single storage device?

A.8

B.256

C.32

D.4

AnswerC

vSphere supports up to 32 paths per device.

Why this answer

vSphere supports up to 32 paths per storage device. 4 paths is common for some arrays, 8 is typical but not the maximum, and 256 is too high.

Practice this question →

5

MCQhard

An administrator attempts to expand a VMFS datastore by increasing the LUN size, but the expand option is grayed out. What could be the reason?

A.The datastore is using block size 2 MB

B.The LUN is shared with another datastore

C.The host is not rescanned

D.The datastore is on a RDM

AnswerB

A LUN can only be used by one datastore; sharing prevents expansion.

Why this answer

If the LUN is shared with another datastore, it cannot be expanded because a LUN can belong to only one datastore. Block size does not affect expansion; RDMs are not used for VMFS; rescan is unrelated to the grayed-out option.

Practice this question →

6

Drag & Dropmedium

Place the steps to create a resource pool in a cluster.

Drag steps to the numbered slots on the right, or tap a step then tap a slot.

Steps

Order

1Step 1

2Step 2

3Step 3

4Step 4

Why this order

Creating a resource pool in a vSphere cluster involves first initiating the creation by right-clicking the cluster and selecting New Resource Pool, then providing a name, followed by configuring CPU and memory resources including shares, reservation, and limit, optionally enabling expandable reservation, and finally confirming by clicking OK. This sequence ensures all required parameters are set before finalizing the resource pool.

Practice this question →

7

MCQhard

An administrator manages a vSAN cluster with 5 ESXi hosts in a single failure domain. The cluster uses vSAN version 8 and is configured with a single disk group per host. A storage policy is applied to a group of VMs with 'Number of failures to tolerate = 1' and 'Primary level of failures = Host'. One host experiences a catastrophic hardware failure and is now marked as 'Absent' in the vSAN cluster. The administrator checks the vSAN health and finds that the affected host's disk group is completely lost. One of the VMs from that group previously had two replicas on different hosts and a witness component on a third host. The VM is still powered on, and the administrator sees that one replica was on the failed host. The other replica and witness are on surviving hosts. What is the current state of this VM regarding data accessibility?

A.The VM is accessible but performance is degraded because the disk group on the failed host is gone.

B.The VM is inaccessible because the lost replica cannot be rebuilt without a replacement host.

C.The VM is inaccessible because the witness component is now the only copy and cannot be used for reads.

D.The VM is fully accessible because it still has one replica and the witness.

AnswerD

The surviving replica and witness provide sufficient redundancy to maintain full data accessibility.

Why this answer

With FTT=1 and primary level of failures = Host, the VM's object has 2 replicas and 1 witness. The surviving replica and witness are sufficient to maintain full read/write access. vSAN can rebuild the lost replica on another host, but even before rebuild, the VM remains fully accessible because there is still a complete replica. Option D correctly states this.

Option A is incorrect because performance does not degrade; the VM uses the surviving replica. Option B is incorrect because the VM is accessible and can be rebuilt. Option C is incorrect because the witness is metadata; the VM has a full replica for reads.

Practice this question →

8

MCQhard

A company is designing a vSphere environment for a critical database application. The storage array supports both Fibre Channel (FC) and iSCSI. The application requires low latency and high IOPS. Which storage protocol and path policy should be recommended?

A.FC with Fixed path policy.

B.FC with Most Recently Used (MRU) path policy.

C.iSCSI with Round Robin path policy.

D.FC with Round Robin path policy.

AnswerD

FC is low latency; Round Robin balances I/O across paths.

Why this answer

FC provides lower latency and higher IOPS than iSCSI due to dedicated hardware and lower protocol overhead, making it ideal for critical database workloads. The Round Robin path policy is recommended for FC with active-active storage arrays because it distributes I/O across all available paths, maximizing throughput and load balancing, which aligns with the requirement for high IOPS.

Exam trap

The trap here is that candidates often assume MRU is the default for FC or that Fixed is sufficient, but the VCP-DCV exam tests the understanding that Round Robin is the recommended path policy for active-active arrays to achieve load balancing and high IOPS, especially for performance-sensitive workloads like databases.

How to eliminate wrong answers

Option A is wrong because the Fixed path policy uses a single preferred path and only switches on failure, which does not optimize for high IOPS or load balancing across multiple paths. Option B is wrong because the Most Recently Used (MRU) path policy is designed for active-passive arrays and can cause path thrashing or suboptimal performance in active-active environments, failing to meet low-latency and high-IOPS needs. Option C is wrong because iSCSI typically incurs higher latency and CPU overhead compared to FC due to TCP/IP processing, making it less suitable for a critical database application requiring low latency and high IOPS.

Practice this question →

9

MCQhard

Refer to the exhibit. An administrator configures a claim rule to use VMW_SATP_ALUA with VMW_PSP_RR and TPGS enabled. After applying the rule, the device shows one active and one standby path. However, the administrator notices that I/O is only sent to the active path. What is the most likely reason?

A.The PSP should be changed to VMW_PSP_MRU to use both paths.

B.The array is not properly configured to present both paths as active/optimized.

C.The claim rule should have used VMW_SATP_DEFAULT_AA instead.

D.The TPGS option should be disabled to force both paths active.

AnswerB

For RR to use multiple paths, the array must set both paths as active/optimized. One standby path will not receive I/O.

Why this answer

With ALUA, paths have asymmetric access states. The PSP RR distributes I/O across all active paths. If only one path is active and the other is standby, RR will only use the active path.

The standby path is not used for I/O. To utilize both paths, the array must present both paths as active/optimized. This is expected behavior for ALUA with one standby path.

Practice this question →

10

MCQeasy

An administrator notices that a VM with a 500 GB virtual disk stored on an NFS datastore is performing poorly during backup operations. The NFS datastore is mounted with default settings. Which change will most likely improve performance?

A.Move the VM to a cluster with more memory.

B.Enable jumbo frames on the ESXi host's VMkernel adapter for NFS.

C.Increase the size of the NFS datastore to 1 TB.

D.Configure the NFS datastore to use vStorage APIs for Array Integration (VAAI).

AnswerD

VAAI offloads storage operations to the array, improving performance.

Why this answer

VAAI for NFS offloads storage operations like hardware-assisted locking and full-file/clone operations to the NAS array, reducing CPU overhead on the ESXi host and improving performance during backup-intensive tasks. Since the NFS datastore is mounted with default settings, VAAI is not automatically enabled and must be explicitly configured to leverage array-based primitives.

Exam trap

The trap here is that candidates often assume jumbo frames (Option B) are the universal fix for NFS performance issues, but the question specifically points to backup operations, where VAAI offloads are the targeted solution for storage-level bottlenecks.

How to eliminate wrong answers

Option A is wrong because moving the VM to a cluster with more memory does not address the I/O bottleneck on the NFS datastore; backup performance is limited by storage latency and network throughput, not host memory. Option B is wrong because enabling jumbo frames on the VMkernel adapter for NFS can improve network efficiency for large transfers, but it does not directly optimize the storage operations (e.g., locking, cloning) that cause poor backup performance; the issue is likely due to software-based operations that VAAI can offload. Option C is wrong because increasing the NFS datastore size to 1 TB does not affect performance; it only provides more storage capacity, which does not resolve the underlying I/O or protocol overhead during backups.

Practice this question →

11

MCQeasy

A vSphere administrator has a VMFS6 datastore that is running low on space. The storage array has additional unallocated LUNs. The administrator attaches a new 2 TB LUN to the ESXi host. What is the correct procedure to increase the datastore capacity?

A.Use the 'Increase datastore capacity' wizard to add the LUN as an extent

B.Unmount the datastore, expand the VMFS partition, then remount

C.Recreate the datastore with a larger block size and restore from backup

D.Create a new datastore on the LUN and use Storage vMotion to migrate VMs

AnswerA

VMFS can be extended online by adding an extent.

Why this answer

The correct procedure is to use the 'Increase datastore capacity' wizard to add the new LUN as an extent to the existing VMFS6 datastore. VMFS6 supports online extension, allowing the datastore to be expanded without unmounting or disrupting running VMs. Option B is incorrect because unmounting and extending the partition is unnecessary; VMFS6 handles this dynamically.

Option C is incorrect because recreating the datastore with a larger block size is disruptive and not required—adding an extent achieves the capacity increase. Option D is incorrect because creating a new datastore and migrating VMs is unnecessary and adds complexity; the existing datastore can be extended seamlessly.

Practice this question →

12

MCQmedium

Refer to the exhibit. An administrator tries to enable vSAN on a host and receives this error. What is the most likely cause?

A.The vSAN cluster has insufficient hosts (needs at least 2).

B.The host does not have any disks installed.

C.The host has disks but they are not claimed for vSAN or not configured in a disk group.

D.The host's disks are all SSD and vSAN requires HDD for capacity tier.

AnswerC

vSAN disk groups must be created; just having disks is insufficient.

Why this answer

vSAN requires at least one disk group consisting of one or more capacity disks (and optionally a cache disk). The error indicates that no disk group is present or the disks are not properly claimed.

Practice this question →

13

MCQhard

An administrator is configuring persistent memory (PMem) for a critical database VM. The host has 512 GB of Intel Optane PMem. The VM must be able to vMotion while PMem is used. Which configuration meets this requirement?

A.Use NVDIMM labels and create a datastore for PMem

B.Add a vPMem device to the VM and ensure the destination host has compatible PMem

C.Configure the VM to use PMem in hardware passthrough mode

D.Emulate PMem using SSD-backed virtual disks

AnswerB

vPMem (virtual PMem) allows vMotion between hosts with PMem capacity.

Why this answer

vMotion with PMem is supported only when the PMem is presented as a virtual persistent memory (vPMem) device. Adding a vPMem device to the VM and ensuring the destination host has compatible PMem capacity allows the VM to vMotion while using PMem. Option A (NVDIMM labels and datastore) is incorrect because PMem cannot be used as a datastore for vMotion; vMotion requires direct vPMem attachment.

Option C (hardware passthrough mode) prevents vMotion because the VM has direct hardware access. Option D (SSD-backed virtual disks) does not use actual PMem, so it does not fulfill the requirement.

Practice this question →

14

MCQeasy

An administrator needs to expand a VMFS6 datastore that is currently 2 TB in size. The LUN presented to the ESXi host has been expanded to 3 TB from the storage array. Which command should the administrator use to extend the datastore?

A.esxcli storage vmfs extend -l naa.60050768018603e1b800000000000001

B.esxcli storage vmfs grow -l naa.60050768018603e1b800000000000001

C.vmkfstools --extend /vmfs/devices/disks/naa.60050768018603e1b800000000000001

D.vmkfstools --growfs /vmfs/devices/disks/naa.60050768018603e1b800000000000001

AnswerD

This command extends the VMFS datastore to the full LUN size.

Why this answer

The correct command is `vmkfstools --growfs` because it is the only VMware-supported method to expand a VMFS6 datastore after the underlying LUN has been resized. This command grows the file system to use the newly available unpartitioned space on the device, without requiring a new partition. The `--growfs` option specifically targets the VMFS file system, not the partition, which is essential for VMFS6 datastores.

Exam trap

The trap here is that candidates often confuse the `esxcli storage vmfs` commands with the `vmkfstools` commands, mistakenly thinking `esxcli storage vmfs grow` is valid, or they assume `vmkfstools --extend` applies to datastores when it is actually for VMDK files.

How to eliminate wrong answers

Option A is wrong because `esxcli storage vmfs extend` is used to extend a VMFS datastore by adding a new extent (i.e., a new LUN or partition), not to grow the existing file system into unpartitioned space on the same LUN. Option B is wrong because `esxcli storage vmfs grow` is not a valid command; the correct `esxcli` subcommand for growing a VMFS datastore into unpartitioned space is `esxcli storage vmfs growfs`. Option C is wrong because `vmkfstools --extend` is used to extend a VMDK file, not a VMFS datastore; it operates on virtual disk files, not on the underlying storage device.

Practice this question →

15

Multi-Selectmedium

Which TWO of the following are prerequisites for enabling jumbo frames on a vSphere host for storage traffic?

Select 2 answers

A.The virtual machine guest OS must support jumbo frames

B.The VMkernel port must be configured with MTU 9000

C.The storage array must support jumbo frames

D.The vCenter Server must be configured for jumbo frames

E.The physical switches must support jumbo frames

AnswersB, E

The VMkernel interface must have MTU set to 9000.

Why this answer

Jumbo frames require physical switch support and VMkernel port MTU configuration. The storage array, guest OS, and vCenter do not directly affect host-level jumbo frame operation.

Practice this question →

16

MCQhard

A vSphere administrator configures a VVol datastore on a Pure Storage array. The administrator creates a storage policy with a custom tag. After attaching the policy to a VM, the VM remains uncompliant. What is the most likely cause?

A.The storage array does not support the capability specified in the policy.

B.The storage array is not VASA-compliant.

C.The datastore cluster is not configured for Storage DRS.

D.The VVol datastore is using NFS protocol instead of iSCSI.

AnswerA

Correct: The policy tag must match a capability advertised by the array.

Why this answer

VVol storage policies rely on capabilities advertised by the array through VASA. If the array does not support the specific capability or tag used in the policy, the VM will be uncompliant. Option A is correct because the custom tag in the storage policy is not supported by the Pure Storage array.

Option B is incorrect: VVol datastores require VASA certification; if the array were not VASA-compliant, it could not host a VVol datastore at all. Option C is incorrect because Storage DRS is unrelated to VVol policy compliance. Option D is incorrect because VVol datastores can use either iSCSI or NFS protocols, and the protocol does not affect compliance.

Practice this question →

17

MCQhard

An administrator notices that a VM with high I/O demands is experiencing performance issues because other VMs on the same datastore are consuming too many I/O operations. Which feature can be used to guarantee a minimum I/O share to this VM?

A.Multipathing policy

B.VM storage policy

C.Storage DRS

D.Storage I/O Control shares

AnswerD

SIOC uses shares to allocate relative I/O priority to VMs.

Why this answer

Storage I/O Control (SIOC) allows setting shares, limits, and reservations to allocate I/O bandwidth per VM. Storage DRS balances I/O across datastores, not per-VM; multipathing selects paths; VM storage policies manage provisioning.

Practice this question →

18

MCQhard

Refer to the exhibit. An administrator observes high latency on a VMFS datastore. Based on the vdbench and esxtop output, which component is most likely causing the latency?

A.The virtual machine or application is causing high latency due to inefficient I/O patterns.

B.The storage array is overloaded, causing high queue depth (DAQLEN=32).

C.The VMFS datastore is fragmented, causing increased seek times.

D.The ESXi host's HBA queue depth is insufficient, causing queuing on the host.

AnswerA

GAVG is high (8.2 ms) indicating latency at the guest OS level, likely from application I/O or driver issues.

Why this answer

The esxtop output shows KAVG (kernel latency) of 2.5 ms and GAVG (guest latency) of 8.2 ms. The total latency (QAVG) is 10.7 ms, but the reported datastore latency is 12 ms. The high GAVG relative to KAVG indicates latency at the guest/application level, not at the storage array.

The vdbench test shows 12 ms average latency, which matches the guest-level latency. The issue is likely due to application I/O patterns or queue depth, not the storage backend.

Practice this question →

19

MCQhard

Refer to the exhibit. An administrator attempts to reclaim unused space on a VMFS datastore using esxcli storage vmfs unmap, but the command fails. Based on the logs and documentation, what is the most likely cause?

A.The reclaim-unit parameter is set too low, causing the array to reject the command.

B.The ESXi host does not have sufficient permissions to issue UNMAP commands.

C.The storage LUN is thick-provisioned and does not support UNMAP.

D.The VMFS datastore is not formatted with a supported block size.

AnswerC

The array documentation confirms UNMAP is only supported on thin-provisioned LUNs.

Why this answer

The vmkernel log shows the UNMAP command failed with ILLEGAL REQUEST, sense 0x5 0x20 0x0, which indicates an invalid command. The vendor documentation states that UNMAP is only supported when thin provisioning is enabled on the LUN. The LUN is likely thick-provisioned, so the array rejects the UNMAP command.

Practice this question →

20

Multi-Selectmedium

A storage administrator is planning a vSAN stretched cluster for a remote office. The cluster must support stretching across two sites for disaster avoidance. Which two prerequisites must be met? (Choose two.)

Select 2 answers

A.The network latency between sites must be less than 5 ms round-trip.

B.The witness host must be placed at the primary site.

C.Each site must have a minimum of two hosts.

D.All hosts must be configured with a single network adapter for vSAN traffic.

E.A vSAN stretched cluster requires a minimum of four fault domains.

AnswersA, C

Latency must be under 5 ms round-trip to ensure acceptable performance for vSAN synchronization traffic.

Why this answer

Correct options A and C: Each site must have a minimum of two hosts (C) and network latency between sites must be less than 5 ms round-trip (A). Option B is incorrect because the witness host must be placed at a third site, not the primary. Option D is incorrect because hosts can use multiple network adapters for vSAN traffic; a single adapter is not required.

Option E is incorrect because a vSAN stretched cluster typically uses two fault domains (one per site) plus a witness, not four.

Practice this question →

21

MCQeasy

Which storage feature in vSphere allows a VM to be migrated from one datastore to another without any downtime?

A.vMotion

B.Distributed Resource Scheduler (DRS)

C.Storage DRS

D.Storage vMotion

AnswerD

Storage vMotion migrates virtual disks with zero downtime.

Why this answer

Storage vMotion is the correct answer because it enables live migration of a virtual machine's virtual disk files from one datastore to another with zero downtime. It uses a mirrored copy mechanism where the source and destination datastores are synchronized before the VM is switched to the destination, ensuring continuous VM availability.

Exam trap

The trap here is confusing vMotion (host migration) with Storage vMotion (datastore migration), as both are live migration technologies but operate on entirely different resources.

How to eliminate wrong answers

Option A is wrong because vMotion migrates a running VM between ESXi hosts, not between datastores; it moves the VM's memory and CPU state, not its storage. Option B is wrong because Distributed Resource Scheduler (DRS) balances compute resources across hosts by recommending or initiating vMotion migrations, but it does not handle storage migrations. Option C is wrong because Storage DRS provides initial placement and ongoing load balancing of VMs across datastores, but it relies on Storage vMotion to perform the actual migration; Storage DRS itself does not execute the migration.

Practice this question →

22

MCQeasy

An administrator wants to use Storage DRS to balance space usage across datastores. What must be configured?

A.Host affinity rules

B.SIOC

C.Storage DRS automation level

D.Storage policy

AnswerC

The automation level determines how Storage DRS handles load balancing.

Why this answer

The storage DRS automation level (manual, partially automated, fully automated) controls whether migration recommendations are generated or applied automatically. SIOC, policies, and host affinity are not required for space balancing.

Practice this question →

23

MCQhard

Refer to the exhibit. What is the most likely reason the second path is in standby state?

A.The host has not successfully authenticated to the storage on that path.

B.The path is not properly zoned to the storage.

C.The storage array reports that path as non-optimized.

D.The multipathing policy is set to Fixed (VMW_PSP_FIXED).

AnswerC

ALUA arrays report non-optimized paths as standby.

Why this answer

The SATP is ALUA, which uses asymmetric states. 'Standby' indicates a non-optimized path. Zoning issues or authentication errors would result in a dead path, not standby. The PSP is round-robin, so it is not Fixed.

Practice this question →

24

MCQhard

Refer to the exhibit. An administrator configures Storage DRS for a datastore cluster. The cluster has datastores with varying performance. Which statement about Storage DRS behavior is correct?

A.Storage DRS will automatically migrate VMs when utilization exceeds 80%.

B.Storage DRS will only balance based on space, ignoring I/O metrics.

C.Storage DRS will balance datastores every 8 hours regardless of thresholds.

D.Storage DRS will generate recommendations but will not perform migrations automatically.

AnswerD

Manual automation means recommendations only.

Why this answer

With manual automation, Storage DRS only generates recommendations and does not automatically migrate VMs. The administrator must review and apply recommendations manually.

Practice this question →

25

MCQmedium

An administrator needs to attach a SAN LUN to a VM for a clustered application that requires SCSI-3 persistent reservations. The VM will run on two hosts in a cluster. Which storage option should be used?

A.Create a virtual mode RDM (Raw Device Mapping).

B.Create a physical mode RDM (Raw Device Mapping).

C.Create a thick-provisioned eager zeroed VMDK on VMFS.

D.Create a VVol representing the LUN.

AnswerB

Correct: Physical mode RDM supports SCSI-3 persistent reservations and cluster applications.

Why this answer

Physical mode RDM (Raw Device Mapping) supports SCSI-3 persistent reservations and allows sharing the LUN between VMs in a cluster. Virtual compatibility mode (virtual RDM) does not support persistent reservations, so option A is incorrect. Option C (thick-provisioned eager zeroed VMDK on VMFS) does not support persistent reservations because VMDKs are managed by VMFS and cannot pass through SCSI-3 commands.

Option D (VVol representing the LUN) may not support persistent reservations in the same direct manner as a physical RDM, and typically requires additional configuration. Therefore, physical mode RDM is the correct choice.

Practice this question →

26

MCQmedium

A vSphere cluster has multiple storage arrays with different capabilities. The administrator wants to automatically place VMs on datastores that match their storage policy. What is required?

A.VASA provider

B.Storage DRS only

C.Storage I/O Control

D.VM storage policy and Storage DRS

AnswerD

Storage DRS uses VM storage policies for automated placement based on capabilities.

Why this answer

Storage DRS uses VM storage policies to automatically place VMs on datastores that satisfy the policy requirements. SIOC is for I/O control; VASA provides capabilities but does not automate placement without Storage DRS.

Practice this question →

27

MCQmedium

Refer to the exhibit. What does this error indicate?

A.The path is misconfigured.

B.The storage device is not connected.

C.The target LUN is not available.

D.The host's HBA is faulty.

AnswerC

SCSI sense D:0x2 (Not Ready) indicates LUN not accessible.

Why this answer

SCSI sense key D:0x2 indicates 'Not Ready', meaning the LUN is not available. Host bus adapter failure or path misconfiguration would produce different sense codes; no path would show a different state.

Practice this question →

28

Multi-Selectmedium

Which TWO statements are true regarding vSAN deduplication and compression?

Select 2 answers

A.They are enabled at the vSAN cluster level

B.They are disabled by default on new vSAN clusters

C.They can only be used with RAID-1 mirroring policies

D.They require an all-flash vSAN configuration

E.Compression can be enabled independently of deduplication

AnswersB, D

They must be explicitly enabled.

Why this answer

vSAN deduplication and compression are disabled by default on new vSAN clusters (option B is correct). They require an all-flash vSAN configuration, meaning all capacity devices must be SSDs (option D is correct). Option A is incorrect because deduplication and compression are enabled at the disk group level, not cluster level.

Option C is incorrect because they can also be used with RAID-5/6 erasure coding policies, not just RAID-1 mirroring. Option E is incorrect because compression cannot be enabled independently; if compression is enabled, deduplication is also enabled.

Practice this question →

29

Multi-Selectmedium

A vSphere administrator is planning to configure VMware vSAN. Which TWO requirements must be met for a vSAN cluster?

Select 2 answers

A.A dedicated 10 GbE network for vSAN traffic.

B.All hosts must have identical CPU models.

C.Each host must have at least one SSD and one HDD for capacity.

D.A minimum of 3 ESXi hosts in the cluster.

E.A shared storage array accessible by all hosts.

AnswersC, D

vSAN uses SSD for caching and HDD for capacity.

Why this answer

VSAN requires at least one SSD (or NVMe) for the cache tier and one HDD (or SSD) for the capacity tier on each host to form a disk group. This hybrid or all-flash configuration is fundamental to vSAN's storage architecture, where the cache tier accelerates writes and reads, while the capacity tier provides persistent storage.

Exam trap

The trap here is that candidates often confuse vSAN's minimum requirement of 3 hosts with the common misconception that vSAN can run with 2 hosts (which requires a witness host for quorum), or they mistakenly think a dedicated 10 GbE network is mandatory, when vSAN can function on 1 GbE with proper configuration.

Practice this question →

30

MCQhard

Refer to the exhibit. An administrator increased the MaxQueueDepth parameter for an NFS 4.1 datastore. Which effect does this change have on performance?

A.It reduces latency by limiting the number of queued I/Os.

B.It limits the maximum size of read and write operations.

C.It increases throughput by allowing more concurrent I/O operations.

D.It has no effect because NFS 4.1 does not support queue depth adjustments.

AnswerC

A higher queue depth allows more I/Os to be processed in parallel, improving throughput.

Why this answer

Increasing MaxQueueDepth allows more concurrent I/Os to be queued to the NFS server, which can improve throughput for workloads with high I/O concurrency but may increase latency if the NFS server is not capable of handling the increased queue depth.

Practice this question →

31

Multi-Selecteasy

Which TWO conditions require increasing the VMFS heap size on an ESXi host?

Select 2 answers

A.The host is connected to many VMFS datastores (e.g., 64)

B.A virtual machine has a VMDK larger than 2 TB

C.The host is part of a vSAN cluster

D.The host has more than 500 virtual machines

E.The host has many snapshots per virtual machine

AnswersA, D

Each datastore consumes heap resources.

Why this answer

VMFS heap is used to track open file descriptors. Increasing the number of VMFS datastores (option A) increases the number of file descriptors needed, as does a large number of virtual machines (option D). Option B (large VMDK) does not directly increase the file descriptor count; it may require large file support but not heap.

Option C (vSAN) does not use VMFS. Option E (many snapshots) increases file count, but the heap size is primarily adjusted for many datastores and many VMs, not snapshots per VM.

Practice this question →

32

MCQhard

A company is implementing VMware vVols with a VASA provider from a major storage vendor. The environment includes multiple ESXi hosts and a vCenter Server. After binding a virtual machine to a vVol, the administrator notices that the VM cannot power on. The storage status shows 'Not mounted'. What is the most likely cause?

A.The VASA provider is not registered on the ESXi host where the VM is being powered on

B.The vCenter Server is not connected to the VASA provider

C.The protocol endpoint is not reachable from the ESXi host

D.The storage policy does not include the correct capability

AnswerA

The VASA provider must be registered on the specific ESXi host to mount the vVol. Without registration, the storage system cannot present the vVol to that host.

Why this answer

vVols require storage container binding per host. If the VASA provider is not properly registered on that host, the vVol cannot be mounted. Option A is correct.

Option B (vCenter Server not connected) does not prevent direct mount if the host has the provider. Option C (protocol endpoint not reachable) would affect all VMs. Option D (storage policy) would affect placement, not power-on.

Practice this question →

33

MCQeasy

A company uses vSphere 8 with a vSAN cluster consisting of 4 ESXi hosts. Each host has a single disk group with one 400 GB NVMe cache SSD and two 2 TB capacity SSDs. The vSAN cluster is used to host a mix of production VMs and test VMs. During peak business hours, the vSAN performance service shows high write latency (over 30 ms) for several production VMs, while test VMs are not affected. The administrator examines the vSAN performance charts and notices that the disk groups are not balanced: one disk group shows over 80% used capacity and high congestion, while the others are below 40%. The administrator wants to improve performance without adding new hardware or changing the storage policy. The vSAN health service reports no errors. Which action should the administrator take first?

A.Manually evacuate data from the most utilized disk group.

B.Increase the number of disk groups on the busiest hosts.

C.Change the storage policy to use RAID-5 erasure coding to reduce capacity overhead.

D.Enable vSAN proactive rebalance.

AnswerD

Proactive rebalance automatically redistributes objects across disk groups to improve balance and performance.

Why this answer

Enabling vSAN proactive rebalance (Option D) automatically redistributes data across disk groups based on capacity and performance metrics, alleviating congestion on the overutilized disk group and reducing write latency. Option A (manual evacuation) is disruptive and not recommended as a first step. Option B (increasing disk groups) requires additional hardware.

Option C (changing to RAID-5) does not address disk group imbalance and may increase overhead.

Practice this question →

34

MCQmedium

Refer to the exhibit. What type of virtual disk is represented by this descriptor file?

A.Thin provisioned VMDK

B.Thick provisioned eager zeroed VMDK

C.Sparse disk with multiple extents

D.Physical mode RDM

AnswerB

Correct. The createType='vmfs' with a single flat.vmdk extent indicates a thick provisioned eager zeroed VMDK (the default thick type on VMFS).

Why this answer

The descriptor shows createType='vmfs' and a single extent with a 'flat.vmdk' file. This is characteristic of a thick provisioned virtual disk on VMFS. Specifically, thick provisioned eager zeroed VMDK files use the 'flat.vmdk' naming convention.

Option A is incorrect because a thin provisioned VMDK would have createType='vmfsThin'. Option C is incorrect because a sparse disk with multiple extents would list multiple extents in the descriptor. Option D is incorrect because a physical mode RDM would have createType='vmfsRaw' or 'vmfsRawDeviceMap' and reference a raw device mapping.

Practice this question →

35

MCQmedium

Refer to the exhibit. An administrator attempts to add a host to a vSAN cluster and receives this error. Which step should the administrator take to resolve the issue?

A.Reconfigure the vMotion interface (vmk1) to also carry vSAN traffic.

B.Add the host to the cluster without vSAN and enable vSAN later.

C.Create a new VMkernel adapter (vmk3) on the host's network and enable vSAN traffic.

D.Enable vSAN traffic on the existing vmk0 interface.

AnswerC

vSAN requires its own VMkernel interface with vSAN enabled.

Why this answer

vSAN requires a dedicated VMkernel interface with the vSAN service enabled. The host has management, vMotion, and provisioning interfaces but none with vSAN enabled. The administrator must create a new VMkernel adapter (e.g., vmk3) on a host network and enable the vSAN service on it.

Practice this question →

36

Drag & Dropmedium

Sequence the steps to configure a DRS rule that keeps two VMs on different hosts.

Drag steps to the numbered slots on the right, or tap a step then tap a slot.

Steps

Order

1Step 1

2Step 2

3Step 3

4Step 4

Why this order

The correct sequence to configure a DRS rule that keeps two VMs on different hosts is: first access the cluster configuration, then add a new DRS rule, then name the rule, then choose the 'Separate VMs' type to enforce host separation, and finally assign the two VMs. This order ensures that each step logically follows the previous one, adhering to the typical workflow in vSphere where the rule is created before its properties are defined and applied.

Practice this question →

37

MCQmedium

A company experiences high latency on a VM running a critical database. The storage is a VMFS datastore on a SAN. The administrator notices that other VMs on the same datastore are idle. What should the administrator configure to ensure the database VM gets sufficient storage I/O resources?

A.Configure Storage I/O Control and set the database VM's shares to High.

B.Set storage I/O limits on all other VMs to 100 IOPS.

C.Configure Storage I/O Control and set a latency threshold of 5 ms.

D.Use Storage DRS to migrate idle VMs to another datastore.

AnswerA

Correct: SIOC with high shares gives the database VM priority during congestion.

Why this answer

Storage I/O Control (SIOC) dynamically shares storage I/O resources among VMs on a datastore. By setting a higher share value for the database VM, it gets priority during congestion. Option B is incorrect because manually setting low I/O limits on all other VMs is inefficient and may lead to underutilization; SIOC with shares is adaptive.

Option C is incorrect because setting a latency threshold (e.g., 5 ms) is a configuration for SIOC to start congestion control, but it does not by itself ensure the database VM gets sufficient I/O; the key is adjusting shares. Option D is incorrect because Storage DRS (not compute DRS) migrates VMs for space and I/O load balancing, but it does not prioritize I/O for a specific VM; it balances across datastores over time.

Practice this question →

38

Multi-Selecteasy

Which TWO of the following are required for vSphere Storage DRS to function?

Select 2 answers

A.VM storage policies defined

B.VASA provider registered

C.vMotion enabled on the hosts

D.Multiple datastores in a datastore cluster

E.SIOC enabled on each datastore

AnswersC, D

vMotion is used to move VMs between datastores.

Why this answer

Storage DRS requires multiple datastores in a datastore cluster and vMotion to migrate VMs between them. SIOC, VASA, and storage policies are optional.

Practice this question →

39

MCQmedium

A vSphere administrator is observing that a VM with a 2 TB thin-provisioned virtual disk on a VMFS6 datastore is reporting 1.5 TB of used space inside the guest OS, but the datastore shows only 800 GB consumed by the VM. What is the most likely cause of this discrepancy?

A.The virtual disk needs to be defragmented to reclaim space.

B.The VM has a snapshot that is consolidating, reducing storage usage.

C.The virtual disk is thick-provisioned lazy zeroed, which delays allocation.

D.The virtual disk is thin-provisioned, so only the actual written blocks consume space on the datastore.

AnswerD

Thin provisioning allocates space as data is written.

Why this answer

The discrepancy is because the virtual disk is thin-provisioned. Thin provisioning means the virtual disk file (VMDK) on the datastore only occupies space for blocks that have been written to by the guest OS, not the full allocated size. The guest OS reports 1.5 TB of used space because it sees the logical file system usage, but the datastore only shows 800 GB consumed because that is the actual physical storage used by the written blocks.

Exam trap

The trap here is that candidates may confuse guest OS reported usage with datastore consumption, not realizing that thin provisioning only allocates storage for blocks actually written, leading them to incorrectly suspect snapshots or defragmentation issues.

How to eliminate wrong answers

Option A is wrong because defragmentation reorganizes data within the guest OS but does not reclaim space on the datastore for thin-provisioned disks; it may even increase fragmentation of the underlying VMDK. Option B is wrong because a consolidating snapshot would temporarily increase storage usage, not decrease it, and the scenario describes a lower datastore consumption, not higher. Option C is wrong because a thick-provisioned lazy zeroed disk allocates all space at creation (2 TB) and does not show only 800 GB consumed; it would show the full 2 TB allocated on the datastore regardless of guest usage.

Practice this question →

40

MCQeasy

An ESXi host is connected to an iSCSI storage array using software iSCSI initiator. The administrator has configured two NICs for iSCSI traffic. During setup, the administrator selects 'Round Robin' as the path policy for the storage device. What is the benefit of this path policy?

A.It uses a single path until failure, then switches to another

B.It minimizes latency by always using the path with lowest latency

C.It load balances I/O across all active paths

D.It provides the highest performance for all workloads

AnswerC

Round Robin alternates I/O requests among paths.

Why this answer

The Round Robin path policy distributes I/O requests evenly across all active paths to the storage device. This improves load balancing by preventing any single path from becoming a bottleneck, which is especially beneficial in environments with multiple NICs configured for iSCSI traffic.

Practice this question →

41

Multi-Selecthard

Which THREE actions can be performed on a VMFS datastore without unmounting it or putting the ESXi host into maintenance mode?

Select 3 answers

A.Remove the datastore from the host's inventory.

B.Upgrade the datastore from VMFS5 to VMFS6.

C.Add a new extent to the datastore from a different LUN.

D.Increase the size of the datastore by expanding an existing extent.

E.Rename the datastore.

AnswersA, D, E

Removing from inventory is possible if no VMs are registered on it.

Why this answer

Removing a datastore from the host's inventory simply detaches the datastore object from the ESXi host's configuration without affecting the underlying LUN or requiring the host to enter maintenance mode. This operation only modifies the host's metadata, not the storage device itself, so it can be performed while VMs are running on other datastores.

Exam trap

The trap here is that candidates often assume any storage-level operation requires maintenance mode or unmounting, but VMware specifically allows certain non-disruptive metadata changes like removal from inventory and renaming while the datastore remains fully accessible to running VMs.

Practice this question →

42

MCQeasy

An administrator needs to create a datastore cluster for a set of VMs that require high availability. The VMs should be automatically balanced across datastores based on I/O latency. Which feature must be enabled on the datastore cluster?

A.Storage I/O Control (SIOC) on each datastore.

B.vSphere HA for the datastore cluster.

C.Storage DRS without I/O metric, using space only.

D.Storage DRS with I/O metric enabled.

AnswerD

Correct: Storage DRS with I/O metric can balance VMs based on storage I/O latency.

Why this answer

Storage DRS with I/O metric enabled allows automatic balancing based on I/O latency. Option A is incorrect because SIOC is a per-datastore feature for controlling I/O shares and limits, not for balancing across datastores in a cluster. Option B is incorrect because vSphere HA provides high availability for VMs in case of host failure, not storage balancing.

Option C is incorrect because Storage DRS without the I/O metric uses only space capacity to balance VMs, not I/O latency. Option D is correct because Storage DRS with the I/O metric enables I/O latency-based balancing across datastores in the cluster.

Practice this question →

43

MCQeasy

What is the default block size of a VMFS5 datastore?

A.1 MB

B.8 MB

C.2 MB

D.4 MB

AnswerA

The default block size for VMFS5 is 1 MB.

Why this answer

VMFS5 uses a default block size of 1 MB. Larger block sizes are available but not default.

Practice this question →

44

MCQhard

A vSAN cluster uses a storage policy with Primary Failures to Tolerate (PFTT) = 1 and Failure Tolerance Method = RAID-1 (Mirroring). What is the minimum number of hosts required to support this policy if each host contributes one disk group?

A.4

B.3

C.6

D.2

AnswerB

Correct: vSAN requires 3 hosts for RAID-1 with PFTT=1 (2 data copies + 1 witness).

Why this answer

For a vSAN cluster with PFTT=1 and RAID-1 mirroring, the minimum number of hosts required is 3. This ensures two copies of data (mirrors) and one witness component. A standard cluster with 2 hosts cannot support PFTT=1 because only two copies exist without a witness.

Options 4 and 6 exceed the minimum, so option B (3) is correct.

Practice this question →

45

MCQeasy

An administrator is configuring multipathing for a Fibre Channel storage array. The administrator wants to maximize throughput by using all available paths. Which path selection policy should be chosen?

A.Fixed (VMW_PSP_FIXED)

B.Most Recently Used (VMW_PSP_MRU)

C.Vendor-specific (VMW_PSP_FIXED_AP)

D.Round Robin (VMW_PSP_RR)

AnswerD

Correct: Round Robin spreads I/O across all active paths, maximizing throughput.

Why this answer

Round Robin (VMW_PSP_RR) policy distributes I/O across all active paths, which maximizes throughput by utilizing all available paths. Option A (Fixed) uses a single preferred path and does not utilize all paths. Option B (MRU) uses one path until failure, then switches, so it does not use all paths simultaneously.

Option C (Vendor-specific) refers to a policy that may be provided by the storage vendor, but the given code VMW_PSP_FIXED_AP is actually a variant of the Fixed policy for Active/Passive arrays, so it also does not use all paths. Therefore, Round Robin is the correct choice.

Practice this question →

46

MCQmedium

A company runs a critical SQL Server VM on vSphere 7.0. The VM has a single 300 GB virtual disk on a VMFS6 datastore backed by a SAN with 8 Gbps Fibre Channel. The VM is configured with 16 vCPUs and 64 GB RAM. Recently, users have reported slow query performance. The administrator checks the datastore performance and sees average latency of 15 ms with peaks of 50 ms during business hours. The storage array has multiple paths to the ESXi host, and the current path policy is Fixed with a single active path. The administrator wants to improve storage performance with minimal cost. Which action should the administrator take first?

A.Change the path selection policy to Round Robin on the ESXi host.

B.Add a vSphere Flash Read Cache to the VM.

C.Upgrade the Fibre Channel infrastructure to 16 Gbps.

D.Convert the virtual disk to thin provisioning to reduce I/O.

AnswerA

Utilizes multiple paths, reducing latency.

Why this answer

The current Fixed path policy with a single active path underutilizes the available storage bandwidth, causing high latency during peak I/O. Changing to Round Robin (RR) distributes I/O across all available paths, reducing queue depth on any single path and lowering latency without any hardware cost. This is the most immediate and cost-effective fix for the observed performance issue.

Exam trap

The trap here is that candidates may assume hardware upgrades (like faster Fibre Channel) are the only solution to high latency, overlooking the fact that a misconfigured path policy can cause a single path to become a bottleneck even when multiple paths exist.

How to eliminate wrong answers

Option B is wrong because vSphere Flash Read Cache is deprecated in vSphere 7.0 and only accelerates read operations, not writes; the SQL Server workload likely involves significant write I/O, and adding it would not address the path-level bottleneck. Option C is wrong because upgrading the Fibre Channel infrastructure to 16 Gbps is a costly hardware change that does not fix the root cause—the single active path policy—and may not reduce latency if the bottleneck is path saturation rather than link speed. Option D is wrong because converting to thin provisioning does not improve I/O performance; it can actually increase latency due to on-demand allocation overhead and does not affect the path selection or queue depth issue.

Practice this question →

47

MCQmedium

Refer to the exhibit. A VM is deployed with this storage policy. What is the minimum number of hosts required for the VM's objects to be provisioned successfully?

A.4

B.2

C.1

D.3

AnswerD

With FTT=1, 2 data replicas and 1 witness require 3 hosts.

Why this answer

With FTT=1 and default RAID-1 mirroring, vSAN needs 2 data replicas plus 1 witness, requiring at least 3 hosts. A single host cannot tolerate failure; two hosts would work for FTT=0.

Practice this question →

48

Multi-Selectmedium

Which TWO vSAN components are always created for every virtual disk object in a vSAN datastore? (Choose two.)

Select 2 answers

A.RAID-5 parity component

B.Checksum component

C.Metadata component

D.Witness

E.Data component (replica)

AnswersC, E

Correct: Metadata (including namespace) is always created for each object.

Practice this question →

49

MCQmedium

An administrator is configuring a vSAN cluster with all-flash capacity. The cluster uses RAID-5 erasure coding with a policy of 'Number of failures to tolerate = 1' (PFTT=1). The administrator wants to minimize storage overhead while ensuring availability. Which vSAN storage policy setting should be used?

A.Set 'Primary level of failures to tolerate' to 1 and 'Primary level of failures to tolerate method' to 'RAID-1 (Mirroring)'

B.Set 'Primary level of failures to tolerate' to 2 and 'Primary level of failures to tolerate method' to 'RAID-6 (Erasure Coding)'

C.Set 'Primary level of failures to tolerate' to 1 and 'Primary level of failures to tolerate method' to 'RAID-5 (Erasure Coding)'

D.Set 'Primary level of failures to tolerate' to 0

AnswerC

RAID-5 with PFTT=1 provides efficient capacity usage and fault tolerance.

Why this answer

For all-flash vSAN, RAID-5/6 erasure coding is recommended. With PFTT=1, RAID-5 requires only 1 parity component, minimizing overhead. Option C is correct.

Option A (RAID-1 mirroring) would use more capacity. Option B (RAID-6) is for PFTT=2 and adds more overhead. Option D (no redundancy) is not fault tolerant.

Practice this question →

50

MCQhard

A company has a vSphere 7.0 cluster with 6 ESXi hosts connected to a Dell EMC PowerStore array using iSCSI. They are using VMFS6 datastores. After a recent firmware upgrade on the array, they notice that performance on one datastore has degraded significantly. The datastore is used by several high-I/O VMs. The administrator runs 'esxcli storage core path list' and sees that all paths to that datastore are active but with varying latency. The PSP is set to Round Robin with IOPS limit of 1000. The SATP is VMW_SATP_ALUA. The storage administrator confirms that the array is in active-active mode. What should the administrator do to improve performance?

A.Enable the array's QoS policy and configure the datastore for VAAI.

B.Change the PSP to Most Recently Used (MRU) to reduce path thrashing.

C.Increase the Round Robin IOPS limit to 10000.

D.Change the SATP to VMW_SATP_DEFAULT_AA and reconfigure the iSCSI initiator.

AnswerC

Increasing the IOPS limit reduces path switching frequency, optimizing throughput for high-I/O VMs.

Why this answer

The Round Robin PSP with an IOPS limit of 1000 is causing the ESXi host to switch paths too frequently, which can lead to higher latency due to path thrashing and suboptimal use of the array's active-active ALUA mode. Increasing the Round Robin IOPS limit to 10000 reduces the frequency of path switches, allowing each path to handle more I/Os before switching, which improves performance for high-I/O workloads on VMFS6 datastores.

Exam trap

The trap here is that candidates often assume a higher IOPS limit will cause more path switching and latency, when in fact the opposite is true—a low IOPS limit causes thrashing, and increasing it stabilizes performance on active-active arrays.

How to eliminate wrong answers

Option A is wrong because enabling QoS on the array does not directly address path switching behavior, and VAAI is a storage offload feature that does not resolve high latency from path thrashing. Option B is wrong because changing the PSP to MRU would use only one active path, negating the benefits of the active-active array and likely worsening performance for high-I/O VMs. Option D is wrong because VMW_SATP_DEFAULT_AA is a legacy SATP for active-active arrays without ALUA, and changing to it would break the ALUA path selection logic, potentially causing incorrect path states or failover issues.

Practice this question →

51

MCQhard

A vSphere administrator is planning a Storage DRS implementation for a datastore cluster containing multiple VMFS datastores. The VMs have various I/O patterns. Some VMs should always remain together on the same datastore. Which configuration should be used?

A.Create a VM-to-VM affinity rule in the datastore cluster.

B.Set individual datastore affinity rules on each host.

C.Apply Storage I/O Control with high shares to the VMs.

D.Create a VM-to-VM anti-affinity rule in the datastore cluster.

AnswerA

Correct: Affinity rule ensures VMs are placed on the same datastore.

Why this answer

Storage DRS affinity rules allow you to keep VMs together (VM-to-VM affinity) or separate (anti-affinity). To ensure VMs remain together, a VM-to-VM affinity rule is required. Option B is incorrect because Storage DRS does not support per-host datastore affinity rules; rules are at the datastore cluster level.

Option C is incorrect because Storage I/O Control manages I/O prioritization, not VM placement. Option D is incorrect because a VM-to-VM anti-affinity rule would separate the VMs, not keep them together.

Practice this question →

52

MCQmedium

An administrator has configured Storage DRS on a cluster of datastores. The cluster includes both NFS and VMFS datastores. The administrator notices that Storage DRS recommendations are not being generated for some datastores. What is the most likely cause?

A.NFS datastores are not supported by Storage DRS

B.The datastore space utilization is below the threshold

C.The datastores are part of a vSAN cluster

D.I/O metrics are not enabled for the Storage DRS cluster

AnswerA

Storage DRS only supports VMFS datastores and certain NFS arrays with VAAI.

Why this answer

Storage DRS does not support NFS datastores unless they are on a NAS array that supports VAAI for storage offload operations. In this scenario, the cluster includes both NFS and VMFS datastores, but Storage DRS recommendations are not generated for some datastores. The most likely cause is that those datastores are NFS and do not meet the requirements for Storage DRS.

Option B (space utilization below threshold) is incorrect because Storage DRS would still generate recommendations when space utilization is below the threshold, just perhaps less frequently. Option C (datastores part of a vSAN cluster) is incorrect because vSAN uses its own distributed resource scheduler, not Storage DRS, and would not be included in the cluster. Option D (I/O metrics not enabled) is incorrect because if I/O metrics were not enabled, recommendations based on I/O latency would not be generated, but recommendations based on space utilization would still be generated; however, the question states recommendations are not being generated for some datastores, which points to the NFS incompatibility.

Practice this question →

53

MCQeasy

An ESXi host has an NFS datastore mounted from a NAS array. The administrator wants to enable hardware acceleration (VAAI) for the datastore. Which requirement must be met?

A.The NAS array must present the LUN as an iSCSI target

B.The datastore must be formatted as VMFS6

C.The ESXi host must have a software FCoE adapter configured

D.The NAS array must support NFSv3 and VAAI primitives

AnswerD

VAAI for NFS requires NAS support for NFSv3/v4.1 and the relevant primitives.

Why this answer

VAAI for NFS requires the NAS array to support NFSv3 (or v4.1) and implement the VAAI primitives. Option D is correct. Option A (iSCSI) is block storage, not NFS.

Option B (VMFS6) is a local filesystem and not required for NFS VAAI. Option C (software FCoE) is unrelated to NFS.

Practice this question →

54

MCQmedium

Refer to the exhibit. An administrator checks the vSAN cluster health and sees the above output. The cluster has been running for months. What is the most likely cause of the degraded objects?

A.A disk failure occurred and components are being rebuilt

B.A storage policy was changed to increase redundancy

C.A disk group is full and needs capacity expansion

D.A network partition between hosts

AnswerA

Degraded objects with resync indicate repair after failure.

Why this answer

Degraded objects in vSAN indicate that some components are unavailable and the system is rebuilding from redundant copies. This is typical after a disk failure, where the cluster automatically initiates resynchronization to restore fault tolerance. The cluster has been running for months, so a recent disk failure is the most likely cause.

Option A correctly identifies this scenario. Option B (network partition) would typically show connectivity issues or 'Unreachable' objects, not degraded. Option C (disk group full) would show 'Absent' or 'No disk space' errors, not degraded.

Option D (policy change) would cause reconfiguration but not necessarily degrade objects; objects would be marked as 'Compliant' after reconfiguration.

Practice this question →

55

MCQhard

A vSphere environment uses an iSCSI storage array with multiple targets. The administrator configures CHAP authentication but some hosts fail to connect. The working hosts are configured with mutual CHAP, while the failing hosts use only one-way CHAP. What is the most likely reason for the failures?

A.The CHAP secret is not set on the failing hosts.

B.The storage array requires mutual CHAP, but the failing hosts are configured for one-way CHAP.

C.The iSCSI target name is incorrect on the failing hosts.

D.The storage array is configured to use only one-way CHAP.

AnswerB

Mutual CHAP requires both sides to authenticate.

Why this answer

Mutual CHAP requires the target to authenticate the initiator and the initiator to authenticate the target. If the storage array is configured to require mutual CHAP, hosts using only one-way CHAP will fail to complete the authentication handshake. The working hosts are configured with mutual CHAP, confirming the array enforces mutual authentication.

Exam trap

The trap here is that candidates assume CHAP configuration is binary (enabled or disabled) and overlook the distinction between one-way and mutual CHAP, leading them to incorrectly select a missing secret or target name issue.

How to eliminate wrong answers

Option A is wrong because the failing hosts are configured with one-way CHAP, which still requires a CHAP secret; the issue is not the absence of a secret but the authentication direction. Option C is wrong because an incorrect iSCSI target name would cause all hosts to fail, not just those using one-way CHAP, and the working hosts connect successfully. Option D is wrong because if the array used only one-way CHAP, the hosts configured with mutual CHAP would fail, but they are working, so the array must require mutual CHAP.

Practice this question →

56

Multi-Selecthard

Which THREE conditions must be met for Storage I/O Control (SIOC) to work effectively? (Choose three.)

Select 3 answers

A.The datastore must be VMFS5 or later.

B.Jumbo frames must be enabled on the storage network.

C.vCenter Server must be available.

D.All hosts accessing the datastore must be in the same vSphere cluster.

E.The datastore must be connected via Fibre Channel only.

AnswersA, C, D

Correct: SIOC is supported on VMFS5 and later versions.

Practice this question →

57

Multi-Selecteasy

Which TWO statements are true about datastore heartbeating in vSphere HA? (Choose two.)

Select 2 answers

A.At least two datastores should be selected for heartbeating to provide redundancy.

B.Datastore heartbeating is enabled by default when vSphere HA is configured.

C.Datastore heartbeating is used to detect host isolation.

D.Datastore heartbeating is only supported on VMFS datastores.

E.vSphere HA uses datastore heartbeating as the primary method for failure detection.

AnswersA, D

Correct: Two datastores are recommended to avoid a single point of failure.

Practice this question →

58

MCQmedium

An administrator needs to migrate a VM from a VMFS5 datastore to a vSAN datastore while preserving the storage policy. Which migration method should be used?

A.Cold migration

B.Clone

C.Storage vMotion

D.vMotion

AnswerC

Storage vMotion migrates disks between datastores with no downtime.

Why this answer

Storage vMotion moves VM disks between datastores while preserving the storage policy association. vMotion moves compute only; cold migration requires downtime; clone creates a copy.

Practice this question →

59

Multi-Selecthard

Which THREE components are required to configure Storage DRS affinity and anti-affinity rules?

Select 3 answers

A.A resource pool

B.Datastores

C.An ESXi host

D.Virtual machines

E.A datastore cluster

AnswersB, D, E

Datastores that are members of the datastore cluster are required for defining affinity/anti-affinity rules, as the rules specify relationships between VMs and datastores.

Why this answer

Storage DRS affinity and anti-affinity rules require a datastore cluster (option E) where the rules are defined, individual datastores (option B) that are members of the cluster and are associated with the rules, and virtual machines (option D) that are the targets of the rules. A resource pool (option A) is used for compute resource management and is not required for these rules. An ESXi host (option C) is also not a component; the rules operate at the datastore cluster level independently of hosts.

Practice this question →

60

MCQhard

Refer to the exhibit. An administrator runs the 'esxcli storage core device list' command and sees the output. The storage array is an active-passive array with ALUA. The current path selection policy is set to Fixed with no preferred path. What is the consequence of this configuration?

A.The I/O is load-balanced across all four paths

B.ALUA is not supported, so the host will use only one path

C.I/O may be sent to a non-optimized path, causing higher latency

D.The device cannot be used for VMDKs, only for RDM

AnswerC

Without preferred path, the host may use a path that is not in active state.

Why this answer

With an active-passive ALUA array, the Fixed path selection policy without a preferred path means the host may not switch to the optimized active path for a given LUN. Since no preferred path is set, the host can select any available path, potentially sending I/O through a non-optimized passive path. This results in higher latency as the array must redirect I/O through the partner controller.

Option C correctly identifies this consequence. Option A is incorrect because Fixed policy does not load-balance; it uses a single path unless it fails. Option B is wrong because ALUA is supported and the host can use multiple paths but only one is active.

Option D is irrelevant as VMDKs can be used on any path.

Practice this question →

61

Multi-Selecteasy

Which THREE of the following are true about vSphere Storage DRS?

Select 3 answers

A.Storage DRS always automatically migrates VMs.

B.Storage DRS works only with VMFS datastores.

C.Storage DRS can be configured with affinity rules.

D.Storage DRS can move virtual machine files between datastores.

E.Storage DRS requires vMotion.

AnswersC, D, E

Affinity rules allow keeping VMs on specific datastores.

Why this answer

Storage DRS can be configured with affinity rules to control which datastores a virtual machine's files should be placed on or kept together, ensuring performance or compliance requirements are met. This is a core feature that allows administrators to define VM-to-datastore affinity and anti-affinity rules within a datastore cluster.

Exam trap

The trap here is that candidates often assume Storage DRS requires vMotion for all operations, but while Storage DRS does use Storage vMotion for migrations, the requirement for vMotion (host-based) is a common confusion—Storage DRS actually relies on Storage vMotion, not host vMotion, though both are often enabled together.

Practice this question →

62

Multi-Selecthard

An administrator is configuring Storage DRS on a datastore cluster. Which THREE conditions must be met for Storage DRS to migrate virtual disks?

Select 3 answers

A.The source and destination datastores must be part of the same datastore cluster.

B.The ESXi host must have a valid Storage vMotion license.

C.The virtual disk must not be attached to a snapshot.

D.The storage array must support VAAI.

E.The virtual machine must be powered on.

AnswersA, B, E

Storage DRS operates within a datastore cluster.

Why this answer

Storage DRS operates within a single datastore cluster, which is a collection of datastores that share resources and are managed as a single entity. For Storage DRS to migrate virtual disks via Storage vMotion, both the source and destination datastores must be members of the same datastore cluster; cross-cluster migrations are not supported by Storage DRS.

Exam trap

The trap here is that candidates often assume Storage DRS requires VAAI for migrations, but VAAI is only an optimization for certain operations (like hardware-assisted locking or copy offload) and is not a prerequisite for Storage DRS to function.

Practice this question →

63

MCQmedium

An administrator notices that a VMFS datastore has only one optimized path. The exhibit shows the output of esxcli storage nmp path list for that datastore. What should the administrator do to fully utilize both storage controllers?

A.Rescan the storage adapters.

B.Change the PSP to Round Robin.

C.Enable the path failover policy.

D.Change the SATP to VMW_SATP_ALUA.

AnswerD

VMW_SATP_ALUA correctly recognizes optimized and non-optimized paths, allowing the PSP to use both paths for I/O.

Why this answer

The exhibit shows only one optimized path because the storage array uses Asymmetric Logical Unit Access (ALUA), which requires the VMW_SATP_ALUA SATP to properly recognize and manage both active/optimized and active/non-optimized paths. With the default SATP (e.g., VMW_SATP_DEFAULT_AA), the host treats all paths as active but fails to distinguish optimized from non-optimized paths, often resulting in only one path being used. Changing to VMW_SATP_ALUA allows the host to correctly identify the optimized path and actively use both storage controllers for load balancing.

Exam trap

The trap here is that candidates assume a Round Robin PSP alone will solve multipathing issues, but they overlook that the SATP must first correctly classify the path states for ALUA arrays before the PSP can distribute I/O across them.

How to eliminate wrong answers

Option A is wrong because rescanning storage adapters only discovers new LUNs or changes in path state; it does not alter the path selection policy or SATP, so it cannot enable the use of both controllers. Option B is wrong because changing the PSP to Round Robin would distribute I/O across all available paths, but if the SATP does not recognize the non-optimized path as active (or marks it as dead), Round Robin will still only use the single optimized path. Option C is wrong because there is no generic 'path failover policy' to enable; path failover is inherent to the NMP and SATP behavior, and the issue is about path classification, not failover activation.

Practice this question →

64

MCQhard

Refer to the exhibit. An administrator creates a storage policy with the rule: "Rule: provisioningType equals thick". When the policy is applied to a VM on this array, what will be the compliance status?

A.Not compliant - the array does not advertise thick provisioning capability.

B.Not compliant - the replication RPO is too high.

C.Compliant - the storage policy is applied regardless of capabilities.

D.Compliant - the array supports thin provisioning by default.

AnswerA

Correct: The array does not have a thick provisioning capability, so the VM is uncompliant.

Why this answer

The array only advertises thin provisioning and replication capabilities. The policy requires thick provisioning, which is not advertised. Therefore the VM will be marked as uncompliant because the array does not support the required capability.

Option B is incorrect because it does not match; the array supports thin only. Option C is incorrect because VVols do not automatically default to thick. Option D is incorrect because the array does not provide thick.

Practice this question →

65

MCQmedium

An administrator is troubleshooting performance issues on a VMFS datastore hosting SQL Server VMs. Storage I/O Control (SIOC) is enabled, but the administrator notices that congestion is detected but no throttling is applied. What could be the reason?

A.The SIOC congestion threshold is set too high

B.The datastore is deduplicated

C.SIOC is disabled on the datastore

D.The host has exceeded the maximum number of concurrent I/Os

AnswerA

The threshold determines when throttling starts; a high value may prevent activation.

Why this answer

SIOC (Storage I/O Control) throttles I/O from VMs when congestion is detected and the queue depth threshold is exceeded. If the congestion threshold is set too high, throttling may not be triggered even when congestion is detected, because the actual queue depth may not exceed the threshold. Option B (the datastore is deduplicated) is not relevant because deduplication does not affect SIOC throttling.

Option C (SIOC is disabled) would mean congestion would not be detected. Option D (host exceeded max concurrent I/Os) is a host-side issue, not directly related to SIOC threshold.

Practice this question →

66

MCQeasy

Which feature allows a VM to use storage directly from the host's local disks, pooled across a cluster?

A.VMFS

B.vFlash

C.NFS

D.vSAN

AnswerD

vSAN creates a distributed datastore from local disks.

Why this answer

vSAN pools local disks from multiple hosts into a shared datastore. VMFS is a filesystem, NFS is network storage, vFlash is caching.

Practice this question →

67

MCQmedium

A company has a vSphere cluster consisting of 8 ESXi hosts connected to a single Fibre Channel SAN array. They use VMFS6 datastores to store virtual machine files. The storage administrator has scheduled a firmware upgrade for the SAN array that requires a controller reboot. This will cause a temporary loss of connectivity to one LUN (datastore) for approximately 5 minutes. The datastore hosts 15 production VMs, including critical database servers. The cluster has sufficient spare capacity on other datastores, but the VMs are large (each about 200 GB). The vSphere administrator must ensure that these VMs remain available during the upgrade. The cluster has vSphere HA enabled with default settings. What should the administrator do to meet the requirement?

A.Use Storage vMotion to migrate all VMs on the affected datastores to a healthy datastore before the upgrade.

B.Configure a vSphere HA admission control policy to reserve resources in case of host failure.

C.Place all ESXi hosts into maintenance mode.

D.Enable Storage I/O Control on the affected datastore to manage I/O during the upgrade.

AnswerA

Storage vMotion allows live migration of VM files without downtime, keeping VMs available during the storage upgrade.

Why this answer

Storage vMotion (option A) allows live migration of virtual machines to a different datastore with zero downtime. Since the cluster has sufficient spare capacity on other datastores, migrating the VMs before the firmware upgrade ensures they remain available during the temporary LUN disconnection. Option B (HA admission control) is designed to reserve resources for host failures, not storage disruptions.

Option C (placing hosts into maintenance mode) would evacuate VMs to other hosts, but the VMs would still be on the affected datastore and would lose access during the upgrade. Option D (Storage I/O Control) manages I/O congestion but does not prevent VM downtime from a LUN disconnect.

Practice this question →

68

MCQmedium

A vSphere administrator is troubleshooting a VM that has been disconnected from its virtual disk. The VM's virtual disk file (vmdk) is still present on the datastore, but the VM cannot be powered on. Which step should the administrator take first to resolve the issue?

A.Create a new VM and attach the existing vmdk file.

B.Re-add the virtual disk from the datastore browser to the VM.

C.Use the vmkfstools command to re-register the virtual disk.

D.Perform a storage vMotion of the VM to another datastore.

AnswerC

vmkfstools -i can recreate the descriptor file from the flat vmdk.

Why this answer

When a VM is disconnected from its virtual disk but the .vmdk file remains on the datastore, the issue is often a corrupt or missing disk descriptor file. The vmkfstools command with the -fix option can repair the descriptor file, re-establishing the connection between the VM and its virtual disk without requiring VM recreation or manual re-attachment.

Exam trap

The trap here is that candidates assume the disk must be manually re-attached via the GUI (Option B), but vSphere requires the descriptor file to be intact for the disk to be recognized, and vmkfstools is the proper tool to repair it without data loss.

How to eliminate wrong answers

Option A is wrong because creating a new VM and attaching the existing .vmdk file is unnecessary and time-consuming; it does not address the underlying descriptor corruption and may cause additional configuration mismatches. Option B is wrong because re-adding the virtual disk from the datastore browser assumes the disk is properly formatted and recognized, but if the descriptor file is corrupt, the datastore browser may not list the disk correctly or the re-add operation will fail. Option D is wrong because performing a Storage vMotion moves the VM to another datastore but does not repair the disconnected disk; the disk would remain disconnected after the migration, and the VM still cannot power on.

Practice this question →

69

MCQhard

Refer to the exhibit. An administrator is troubleshooting performance issues on a vSphere host. The device is a Fibre Channel disk with a queue depth of 32. The esxtop output shows KAVG=2.5 ms and GAVG=8.2 ms. What is the most likely cause of the high GAVG?

A.The virtual machine's storage stack (e.g., VMkernel or guest OS) is introducing additional latency.

B.The device is an SSD, but the system is not using the correct scheduling policy.

C.The storage array is experiencing high contention, causing increased latency.

D.The ESXi host's HBA queue depth is too low, causing queuing.

AnswerA

High GAVG relative to KAVG suggests latency in the VMkernel or guest OS, possibly due to CPU constraints or driver issues.

Why this answer

GAVG (guest average latency) is the time taken from the guest OS perspective, including processing in the VM kernel. High GAVG with low KAVG indicates that the latency is introduced at the guest level, possibly due to high CPU usage, inefficient storage drivers, or application I/O patterns. The storage array is responding quickly (KAVG low), so the issue is not the array or the host HBA.

Practice this question →

70

MCQeasy

A company uses vSphere 7.0 with two ESXi hosts in a cluster. All virtual machines (VMs) are stored on a VMFS6 datastore backed by a single LUN from a mid-range SAN. The LUN is presented to both hosts and both hosts have identical multipathing configuration. Recently, an administrator noticed that one host shows the datastore as "Inactive" while the other host can access it normally. The administrator verifies that both hosts have connectivity to the SAN, the LUN is visible to both hosts, and the storage array reports the LUN as online. The administrator wants to restore access to the datastore on the affected host without disrupting VMs running on the other host. Which action should the administrator take?

A.Perform a storage rescan on the affected host.

B.Reboot the affected ESXi host.

C.Change the multipathing policy on the affected host to Round Robin.

D.Unmount the datastore from the affected host and then remount it.

AnswerA

Rescanning refreshes the list of paths and resolves transient path issues.

Why this answer

The datastore showing as 'Inactive' on one host while remaining accessible on the other indicates a path failure or a transient storage connectivity issue at the host level, not a permanent problem with the LUN or array. Performing a storage rescan on the affected host forces the ESXi host to re-evaluate all storage paths and re-register the datastore, restoring access without impacting VMs on the other host. This is the safest and least disruptive action because it does not require a reboot, multipathing policy change, or unmounting the datastore.

Exam trap

The trap here is that candidates may assume an 'Inactive' datastore requires a disruptive action like a reboot or unmount, when in fact a simple storage rescan is the standard VMware-recommended first step to re-establish path connectivity without affecting running VMs.

How to eliminate wrong answers

Option B is wrong because rebooting the affected host would cause unnecessary downtime for any VMs running on that host and is not required to fix a simple path visibility issue; a rescan is sufficient. Option C is wrong because changing the multipathing policy to Round Robin addresses load balancing across multiple active paths, but the problem here is that the datastore is 'Inactive' (all paths are dead or not claimed), not that the current policy is suboptimal; changing the policy will not re-establish connectivity. Option D is wrong because unmounting and remounting the datastore is a more invasive operation that could disrupt VMs on the affected host if any are present, and it does not address the underlying path issue; a rescan is the correct first step to re-discover the LUN.

Practice this question →

71

Matchingmedium

Match each vSphere object to its maximum supported size (vSphere 7).

Drag a concept onto its matching description — or click a concept then click the description.

Concepts

Matches

62 TB

256

6 TB

64

1024

Why these pairings

Correct matches: VM max vCPUs 256, VM max memory 6 TB, datastore max size 64 TB, cluster max hosts 64. Common confusions include lower values from previous versions.

Practice this question →

72

MCQhard

A vSphere administrator is deploying VMs on a vSAN cluster with 6 hosts. Each host has two disk groups, each with one cache SSD and four capacity SSDs. The administrator applies a storage policy using RAID 5 erasure coding with Number of failures to tolerate set to 1. After some time, the administrator notices that several VMs are showing compliance status as 'Non-compliant'. Investigating further, the administrator finds that on one host, the cache SSD of the first disk group has failed. The capacity SSDs in that disk group are still functional. The vSAN cluster still has sufficient overall capacity and the health service shows no other issues. What is the most likely reason for the VMs' non-compliance?

A.The failed cache device reduces the overall cache capacity, causing the policy to be violated.

B.The vSAN cluster has lost a fault domain due to the failed cache device.

C.The RAID 5 policy requires a minimum of 4 hosts with cache devices, and now only 5 are available.

D.The failed cache device makes the entire disk group unavailable, so components on that disk group are inaccessible.

AnswerD

A failed cache device causes the entire disk group to be non-operational, leading to loss of any VM components stored on it, which results in non-compliance with the storage policy.

Why this answer

A failed cache device renders the entire disk group non-operational; any VM components on that disk group become inaccessible, causing non-compliance. Option A is incorrect because cache capacity is not a compliance factor. Option B is incorrect because loss of a single cache device does not remove a fault domain.

Option C is incorrect because the number of hosts with cache devices is still sufficient.

Practice this question →

73

Multi-Selecthard

Which THREE of the following are characteristics of vSAN stretched clusters?

Select 3 answers

A.Provides site-level fault tolerance

B.Requires two fault domains (one per site)

C.Requires a witness host in a third location

D.Supports RAID 5/6 erasure coding

E.Uses a dedicated vSAN cluster separate from the main datastores

AnswersA, B, C

Stretched clusters survive failure of an entire site.

Why this answer

Stretched clusters require a witness host in a third location, provide site-level fault tolerance, and require two fault domains (one per site). RAID 5/6 is supported but not unique to stretched clusters; it is still a single vSAN cluster.

Practice this question →

74

MCQeasy

Refer to the exhibit. An administrator runs the 'vmkfstools -P -v 10' command on a datastore. What is the total capacity of the datastore?

A.2.5 TB

B.2 TB

C.3 TB

D.1 TB

AnswerC

1,048,576 + 2,097,152 = 3,145,728 blocks × 1 MB = 3,145,728 MB ≈ 3 TB.

Why this answer

The vmkfstools -P -v 10 command displays the datastore's extent information. The output shows two extents: one with 1,048,576 blocks and another with 2,097,152 blocks, both using a 1 MB block size. Total blocks = 1,048,576 + 2,097,152 = 3,145,728 blocks.

Since each block is 1 MB, total capacity = 3,145,728 MB = 3,072 GB = 3 TB. Therefore, option C (3 TB) is correct. Option A (2.5 TB) and Option B (2 TB) are incorrect; Option D (1 TB) matches only the smaller extent.

Practice this question →

75

Matchingmedium

Match each VMware acronym to its full name.

Drag a concept onto its matching description — or click a concept then click the description.

Concepts

Matches

Virtual Distributed Switch

Virtual Standard Switch

vSphere Installation Bundle

vSphere Storage APIs for Array Integration

vSphere APIs for Storage Awareness

Why these pairings

Common VMware acronyms and their expansions.

Practice this question →

Page 1 of 2 · 78 questions totalNext →

Ready to test yourself?

Try a timed practice session using only Configure and Manage vSphere Storage questions.

Start 20-question session