I85PREP
⌘ K
Huawei HCIP-Storage study materials
Huawei HCIP-Storage study resources

Huawei HCIP-Storage practice questions

One question at a time. Build your Huawei HCIP-Storage understanding.

First 50 of 200 questions. Each one lists the choices, the correct answer, and an explanation. The checker at the top of this page saves progress in this browser when it is available.

Continue in the practice tool above to load more questions, 50 at a time.

Question 1Storage Fundamentals and Architectures

Which is the minimum number of physical disks required to build a RAID 5 array?

  1. Three Correct answer.
  2. Five
  3. Two
  4. Four

Explanation

RAID 5 uses block-level striping with distributed parity across all member disks. At least three disks are required so that data and parity can be distributed and the array can survive one disk failure. Two disks cannot provide distributed parity, while four or five disks work but are not the minimum.

Question 2Storage Fundamentals and Architectures

Which storage interface is designed for full-duplex, enterprise-class disk connectivity and commonly supports dual-port drives for high availability?

  1. IDE
  2. SAS Correct answer.
  3. USB 3.0
  4. SATA

Explanation

SAS, or Serial Attached SCSI, is an enterprise disk interface that supports full-duplex transfers, dual-port connectivity, and higher reliability features. SATA is commonly used for cost-sensitive desktop and nearline drives and does not provide the same enterprise dual-port model. USB and IDE are not enterprise disk-array interconnect standards in this context.

Question 3Storage Fundamentals and Architectures

In a SAN, what does a LUN identify to a host?

  1. An addressable logical storage unit Correct answer.
  2. A physical disk slot in an enclosure
  3. A management network subnet ID
  4. A Fibre Channel switch port ID

Explanation

A LUN is a logical unit number that identifies an addressable logical storage unit exported by a storage array to a host. The host uses it as a block device target, not as a physical slot, switch port, or management subnet. Multiple hosts may be mapped to the same LUN or to different LUNs depending on the array configuration.

Question 4Storage Fundamentals and Architectures

A write-heavy relational database needs the lowest possible RAID write penalty and can tolerate losing one disk per mirror pair. Which RAID level best fits?

  1. RAID 0
  2. RAID 5
  3. RAID 10 Correct answer.
  4. RAID 6

Explanation

RAID 10 mirrors data and then stripes across mirror pairs, so writes do not require parity calculation or read-modify-write operations. It can tolerate one disk failure per mirror pair and generally offers better write performance than parity RAID. RAID 5 and RAID 6 incur parity write penalties, and RAID 0 provides no redundancy.

Question 5Storage Fundamentals and Architectures

In a RAID 2.0+ storage system, what is the main purpose of dividing physical disks into small chunks and then building logical disks from those chunks?

  1. To route all I/O through one controller
  2. To eliminate parity and mirroring entirely
  3. To convert SAS disks into NVMe devices
  4. To spread data and speed rebuilds Correct answer.

Explanation

RAID 2.0+ virtualizes physical disks into fine-grained chunks and composes logical disks from chunks distributed across many physical disks. This improves load distribution and allows rebuilds to focus on used chunks rather than entire disks. It does not remove the need for data protection or change the physical media type.

Question 6Storage Fundamentals and Architectures

A storage pool using thin provisioning reports 70% subscribed capacity but only 40% physically used. Which operational practice best prevents unexpected pool exhaustion?

  1. Convert the pool to thick provisioning now
  2. Increase the host file system block size
  3. Disable all snapshot and clone features
  4. Monitor usage and set capacity alarms Correct answer.

Explanation

Thin provisioning can overcommit capacity, so the array may report high subscription while physical usage is still lower. Administrators should monitor both subscribed and physical capacity and configure thresholds or alarms before the pool runs out. Snapshots and clones are not the root cause by themselves, and changing the file system block size does not manage pool capacity.

Question 7Storage Fundamentals and Architectures

In Fibre Channel SAN zoning, which identifier does WWN zoning use to permit communication between specific initiators and targets?

  1. Domain controller ID
  2. Logical unit number
  3. IP address
  4. Port WWN Correct answer.

Explanation

WWN zoning uses World Wide Port Names, specifically the port WWN of each Fibre Channel node, to define which initiators and targets may communicate. It is independent of the switch port or physical location used by the device. Domain controller IDs, IP addresses, and LUNs are not the identifiers used for WWN zoning.

Question 8Storage Fundamentals and Architectures

Which statement accurately describes a copy-on-write snapshot?

  1. It requires the source volume to remain offline for its lifetime
  2. It permanently redirects every host write to the snapshot volume
  3. It copies a block only before overwrite Correct answer.
  4. It immediately makes a complete physical copy of the source volume

Explanation

A copy-on-write snapshot initially captures metadata rather than a full data copy. When a source block is about to be overwritten, the original block is copied to snapshot-reserved space so the snapshot can preserve the earlier image. The source volume remains online, and writes are not permanently redirected to the snapshot.

Question 9Storage Fundamentals and Architectures

Compared with three-way replication, erasure coding in a distributed storage system typically provides which trade-off?

  1. No data protection but faster write latency
  2. Stronger protection with no CPU or network overhead
  3. Lower raw capacity efficiency but simpler rebuilds
  4. Higher raw capacity efficiency but more complex rebuild and recovery Correct answer.

Explanation

Erasure coding stores data and parity fragments across nodes, so it can achieve a given fault tolerance with less raw capacity than full replication. The trade-off is that rebuild and recovery require reading multiple fragments and reconstructing data, which increases CPU and network overhead. Replication is simpler to rebuild but consumes more raw capacity.

Question 10Storage Fundamentals and Architectures

Which statement best describes NVMe over Fabrics compared with iSCSI?

  1. It extends NVMe queues and commands across a fabric Correct answer.
  2. It encapsulates SCSI commands in TCP/IP and relies on Ethernet lossless tuning
  3. It translates NVMe commands into Fibre Channel arbitrated loop frames
  4. It requires SATA drives to be exported directly to remote hosts

Explanation

NVMe over Fabrics carries the NVMe command and queue model over fabrics such as RDMA, Fibre Channel, or TCP. Unlike iSCSI, which transports SCSI commands over TCP/IP, NVMe-oF avoids a SCSI translation layer and preserves NVMe's efficient queue architecture. It does not export SATA drives directly or use arbitrated loop framing as its defining mechanism.

Question 11Storage Fundamentals and Architectures

A RAID 6 array is built from six equal-capacity SAS drives. How many simultaneous drive failures can the array normally survive while continuing to serve data?

  1. Two Correct answer.
  2. Three
  3. Zero
  4. One

Explanation

RAID 6 maintains two independent parity blocks per stripe, so it can reconstruct data after any two member drives fail. A third drive failure exceeds the available parity protection and causes data loss.

Question 12Storage Fundamentals and Architectures

Which protocol does an iSCSI initiator use to carry SCSI command descriptor blocks across an IP network?

  1. Fibre Channel Protocol (FCP)
  2. PCI Express Non-Volatile Memory express
  3. InfiniBand Reliable Connection
  4. Transmission Control Protocol (TCP) Correct answer.

Explanation

iSCSI encapsulates SCSI command descriptor blocks and data in TCP/IP packets, commonly using TCP port 3260. FCP is used on Fibre Channel fabrics, not as the transport for iSCSI.

Question 13Storage Fundamentals and Architectures

In a Fibre Channel switched fabric, which port type is used on an ISL between two fabric switches?

  1. E_Port Correct answer.
  2. F_Port
  3. N_Port
  4. NL_Port

Explanation

An E_Port is an expansion port that connects two Fibre Channel switches in an interswitch link. N_Ports are node ports, F_Ports are switch ports facing nodes, and NL_Ports are node loop ports.

Question 14Storage Fundamentals and Architectures

A storage administrator wants to prevent a host HBA from accessing a LUN even though fabric zoning permits the two WWPNs to communicate. Which array-side control should be configured?

  1. A larger RAID stripe size on the LUN
  2. A hard zone that includes every initiator in the fabric
  3. Write-back caching on the target port
  4. LUN masking that maps the LUN only to the approved host initiator Correct answer.

Explanation

LUN masking is enforced by the storage array and limits which host initiators can detect and access a given LUN. Zoning controls fabric-level communication, but it does not by itself authorize a host to access a LUN.

Question 15Storage Fundamentals and Architectures

Which statement correctly distinguishes a storage clone from a snapshot?

  1. A snapshot cannot be used to restore the source volume
  2. A snapshot consumes a full copy of the source volume at creation time
  3. A clone is a full independent copy of the source volume Correct answer.
  4. A clone is a point-in-time metadata mapping that shares blocks with the source

Explanation

A clone is a complete independent copy of the source volume, so it does not depend on the source after creation. A snapshot is a point-in-time logical image that typically shares unchanged blocks and can be used to restore the source.

Question 16Storage Fundamentals and Architectures

A LUN is configured with write-back caching and no battery backup. What happens to a host write before the array acknowledges it?

  1. The write is acknowledged after the cache is flushed to a remote replication site
  2. The write is acknowledged only after the parity disk is updated
  3. The write is acknowledged only after it is written to the back-end disk
  4. The write is acknowledged after it is stored in cache but before it is destaged to disk Correct answer.

Explanation

Write-back caching acknowledges host writes once they are stored in cache, then destages them to disk later. Without cache protection, a power failure can lose acknowledged writes that have not yet been destaged.

Question 17Storage Fundamentals and Architectures

An SSD datasheet states an endurance rating of 3 DWPD over a five-year warranty. What does DWPD measure?

  1. The number of drive replacements allowed per year under warranty
  2. The maximum number of power cycles the drive can endure per day
  3. The number of full drive capacity writes the drive can sustain per day over its warranty Correct answer.
  4. The number of data path resets the drive can handle per day

Explanation

DWPD means drive writes per day and expresses endurance as the number of full drive capacity writes the SSD can sustain each day for its warranty period. TBW is a related endurance specification expressed as total bytes written.

Question 18Storage Fundamentals and Architectures

A storage pool contains NVMe SSDs, SAS SSDs, and 10K RPM HDDs. What is the main benefit of enabling automated tiering across these media?

  1. It increases the raw capacity of the HDD tier by compression
  2. It reduces the number of RAID parity disks required in the pool
  3. It encrypts each tier with a different key automatically
  4. It relocates frequently accessed extents to faster media and colder extents to slower media Correct answer.

Explanation

Automated tiering monitors access patterns and moves hot extents to higher-performance media while demoting cold extents to lower-cost media. This improves cost and performance balance without changing RAID parity requirements or providing encryption.

Question 19Storage Fundamentals and Architectures

A LUN is presented through two controllers in an active-active array that uses ALUA. The host sees two paths, and one path reports an active/optimized state while the other reports active/non-optimized. What should the host multipathing software do with this information?

  1. Send all I/O only to the non-optimized path to balance controller load
  2. Prefer the optimized path for normal I/O and use the non-optimized path for failover or limited load Correct answer.
  3. Ignore ALUA states because they apply only to tape devices
  4. Disable the optimized path because ALUA requires a single active path

Explanation

ALUA reports per-path access states so hosts can prefer the controller that owns the LUN and avoid unnecessary cross-controller traffic. Non-optimized paths remain available for failover and, depending on policy, limited load sharing.

Question 20Storage Fundamentals and Architectures

During Fibre Channel fabric login, which service does an N_Port contact at the well-known address 0xFFFFFE to obtain an FC ID?

  1. Time Server
  2. Name Server
  3. Fabric Login Server Correct answer.
  4. Management Server

Explanation

FLOGI is sent to the Fabric Login Server at well-known address 0xFFFFFE, which assigns the N_Port an FC ID and negotiates service parameters. The Name Server, reached at 0xFFFFFC, holds name-service registrations rather than performing fabric login.

Question 21Storage Fundamentals and Architectures

In a storage array, which component is volatile and primarily used to buffer incoming host writes before they are destaged to persistent media?

  1. Solid-state drive
  2. SAS expansion port
  3. Tape library
  4. Controller cache Correct answer.

Explanation

Controller cache is volatile memory that temporarily holds host writes so the array can acknowledge them quickly before destaging data to disk or flash. Battery or flash protection is often used to preserve cached writes during a power loss.

Question 22Storage Fundamentals and Architectures

A RAID 0 array is created from four disks. If one disk fails, what is the normal outcome for the array?

  1. All data on the array becomes inaccessible. Correct answer.
  2. Only the data stripes on the failed disk are lost.
  3. It continues in degraded mode and rebuilds automatically.
  4. It converts to RAID 5 on the remaining disks.

Explanation

RAID 0 stripes data across all member disks without mirroring or parity, so no member disk holds a complete recoverable copy of the data. When one disk fails, the stripe is broken and the array's data cannot be accessed normally.

Question 23Storage Fundamentals and Architectures

Which LUN provisioning method allocates physical capacity equal to the full LUN size at creation time?

  1. Thin provisioning
  2. Automatic tiering
  3. Snapshot reserve
  4. Thick provisioning Correct answer.

Explanation

Thick provisioning reserves the full configured capacity for the LUN when it is created, so the storage pool must have that space available. Thin provisioning instead allocates physical extents on demand as data is written.

Question 24Storage Fundamentals and Architectures

Which statement accurately describes a RAID 10 array built from four disks?

  1. It uses distributed parity across all member disks.
  2. It mirrors pairs and stripes across those pairs. Correct answer.
  3. It dedicates one parity disk for the entire array.
  4. It stripes data without any mirroring or parity.

Explanation

RAID 10 combines mirroring and striping: disks are paired into mirrors, and data is striped across those mirrored pairs. It can normally survive one disk failure in each mirrored pair without losing data.

Question 25Storage Fundamentals and Architectures

Which remote replication mode acknowledges a host write only after the write has been committed at both the local and remote storage systems?

  1. Snapshot-based replication
  2. Synchronous replication Correct answer.
  3. Asynchronous replication
  4. Journal-based asynchronous replication

Explanation

Synchronous replication completes the host write only after the data is committed at both sites, which supports a near-zero recovery point objective but adds latency. Asynchronous methods acknowledge locally first and transfer data to the remote site later.

Question 26Storage Fundamentals and Architectures

What is the primary storage efficiency benefit of data deduplication?

  1. It spreads data across additional disks to improve sequential throughput.
  2. It encrypts repeated data blocks so that they cannot be read by unauthorized hosts.
  3. It compresses every block with a fixed hardware codec before writing it to media.
  4. It stores one copy of identical blocks and references it. Correct answer.

Explanation

Deduplication identifies identical data blocks and keeps a single physical copy, with metadata references pointing multiple logical locations to that copy. This reduces the physical capacity needed for redundant data such as repeated backups or virtual machine images.

Question 27Storage Fundamentals and Architectures

In a clustered SAN environment, what is the main purpose of SCSI-3 persistent reservations?

  1. To assign a worldwide port name to each host bus adapter in the fabric.
  2. To compress SCSI commands before they are transmitted over the SAN.
  3. To arbitrate exclusive access to a LUN among cluster nodes. Correct answer.
  4. To discover iSCSI targets by using the SendTargets command.

Explanation

SCSI-3 persistent reservations let cluster nodes coordinate and enforce exclusive or shared access to a LUN, which helps prevent simultaneous writes from multiple nodes. The reservation persists across sessions and is used by clustering software for quorum and failover control.

Question 28Storage Fundamentals and Architectures

What does N_Port ID Virtualization (NPIV) allow in a Fibre Channel SAN?

  1. It converts FCoE frames into native FC frames at the edge.
  2. It lets one HBA port register multiple WWPNs with the fabric. Correct answer.
  3. It encrypts ISL traffic between two fabric switches.
  4. It allows a switch to assign a single domain ID to multiple fabrics.

Explanation

NPIV lets a single physical N_Port register multiple virtual WWPNs with the fabric, so separate virtual machines or applications can each appear as an independent initiator. Zoning and LUN masking can then be applied per virtual WWPN.

Question 29Storage Fundamentals and Architectures

In an SSD, what does write amplification describe?

  1. The increase in read latency caused by garbage collection.
  2. The number of host writes that can be combined into one command.
  3. The reduction in usable capacity caused by overprovisioning.
  4. The ratio of physical flash writes to host-requested writes. Correct answer.

Explanation

Write amplification compares the amount of data physically written to flash with the amount requested by the host. Garbage collection and page-level programming can cause more flash writes than host writes, which affects endurance and performance.

Question 30Storage Fundamentals and Architectures

What is the main effect of wide striping a LUN across many disks in a storage pool?

  1. It restricts the LUN to a single RAID group for isolation.
  2. It duplicates every block on two disks for redundancy.
  3. It reserves a dedicated parity disk for each LUN.
  4. It distributes I/O across more spindles to improve performance. Correct answer.

Explanation

Wide striping spreads a LUN's extents across many disks or RAID groups, so I/O can be serviced in parallel by a larger set of spindles. This typically improves throughput and reduces hotspots compared with a narrow LUN confined to few disks.

Question 31Storage Fundamentals and Architectures

In a Fibre Channel SAN, what does a World Wide Port Name (WWPN) uniquely identify?

  1. A logical unit number presented to a host
  2. An active zone set configured on a switch
  3. The whole fabric and every switch joined to it
  4. A single port on a node or switch in the fabric Correct answer.

Explanation

A WWPN is a 64-bit identifier assigned to one Fibre Channel port and stays with that port; the related WWNN identifies the node as a whole. Zoning rules, LUN mapping, and NPIV virtual ports all rely on WWPNs to distinguish individual ports.

Question 32Storage Fundamentals and Architectures

A RAID 5 array is built from four 2 TB drives. What usable capacity does the array provide?

  1. 6 TB Correct answer.
  2. 8 TB
  3. 2 TB
  4. 4 TB

Explanation

RAID 5 distributes parity across all member drives and reserves the capacity of one drive for that parity. Usable capacity is therefore (n - 1) x drive capacity = (4 - 1) x 2 TB = 6 TB, while raw capacity is 8 TB.

Question 33Storage Fundamentals and Architectures

Which underlying network capability is required to carry Fibre Channel frames over Ethernet using FCoE?

  1. Lossless Ethernet with Priority Flow Control and DCB Correct answer.
  2. An InfiniBand fabric with RDMA queue pairs
  3. An arbitrated loop shared by all converged network adapters
  4. A routed TCP/IP path between each server and the target array

Explanation

FCoE encapsulates Fibre Channel frames directly in Ethernet and does not use TCP/IP, so the Ethernet segment must be lossless. Priority Flow Control and DCB provide that lossless behaviour by pausing selected traffic classes instead of dropping frames.

Question 34Storage Fundamentals and Architectures

An iSCSI initiator opens a discovery session to a target portal and requests the list of targets that the portal exports. Which iSCSI mechanism performs this request?

  1. SendTargets Correct answer.
  2. ARP
  3. CHAP
  4. iSNS

Explanation

SendTargets is the iSCSI discovery method in which an initiator asks a target portal which targets it exports and receives their names and addresses. iSNS is a separate discovery service based on a central database, and CHAP only authenticates identities.

Question 35Storage Fundamentals and Architectures

Mutual CHAP is enabled on an iSCSI session between a host and a storage array. Which authentication behaviour does this configuration enforce?

  1. Only the array verifies the host's credentials
  2. Only the host verifies the array's credentials
  3. The array trusts any initiator that presents the correct target IQN
  4. The host verifies the array, and the array verifies the host Correct answer.

Explanation

With mutual CHAP, each end of the iSCSI session challenges the other, so the host proves its identity to the array and the array proves its identity to the host. Ordinary one-way CHAP authenticates only the initiator.

Question 36Storage Fundamentals and Architectures

Which characteristic distinguishes a SAS drive from a SATA drive in an enterprise storage enclosure?

  1. A shared parallel bus with a single controller
  2. Dual-ported, full-duplex links and a deeper command queue Correct answer.
  3. Fixed 4 KB sectors, whereas SATA uses only 512-byte sectors
  4. A direct IP connection that bypasses the HBA

Explanation

SAS is a serial, dual-port interface that supports full-duplex transfer and a deep command queue, allowing a drive to be connected to two controllers for multipath redundancy. SATA is single-ported and half-duplex, so it cannot provide the same high-availability connectivity.

Question 37Storage Fundamentals and Architectures

A SAN administrator wants switch hardware to enforce that only devices plugged into specific physical switch ports can communicate, regardless of the WWPNs they use. Which zoning method should be used?

  1. LUN masking applied on the storage array
  2. Fabric login with a well-known address
  3. Soft zoning keyed to worldwide port names
  4. Hard zoning keyed to switch port IDs Correct answer.

Explanation

Hard zoning is enforced by the switch ASIC using the domain and port ID of the physical port a device is attached to, so a device moved to another port loses its permitted communication. Soft zoning filters by WWPN through the name server and does not bind access to a physical port.

Question 38Storage Fundamentals and Architectures

In a dual-controller storage array, what is the purpose of mirroring write cache between the two controllers?

  1. To compress cache data before it is destaged to disk
  2. To double the cache capacity available for read prefetch
  3. To synchronize the clocks of the two controllers
  4. To preserve acknowledged writes if a controller fails Correct answer.

Explanation

In a dual-controller array, the write cache is mirrored to the peer controller before the host I/O is acknowledged. If one controller fails, the surviving controller still has a copy of the uncommitted write and can destage it, preventing data loss.

Question 39Storage Fundamentals and Architectures

Two Fibre Channel switches are connected by an ISL. Which flow-control mechanism prevents a transmitter from overrunning the receive buffers of the adjacent switch on that link?

  1. End-to-end credit managed by the destination N_Port
  2. Resource allocation timeout (R_A_TOV)
  3. Buffer-to-buffer credit negotiated per link Correct answer.
  4. Exchange IDs assigned by the fabric login server

Explanation

Buffer-to-buffer credit is a per-link count of free receive buffers at the adjacent port; a port may transmit only while its neighbour has advertised credit, which prevents buffer overrun on that link. End-to-end credit operates across the fabric to the destination N_Port rather than protecting a single ISL's buffers.

Question 40Storage Fundamentals and Architectures

A storage pool uses 8+2 erasure coding across ten 10 TB drives. What usable capacity does the pool provide, and how many simultaneous drive failures can it survive?

  1. 50 TB usable; five drive failures
  2. 90 TB usable; one drive failure
  3. 80 TB usable; two drive failures Correct answer.
  4. 100 TB usable; one drive failure

Explanation

In 8+2 erasure coding, eight fragments hold data and two hold parity, so usable capacity is 8 x 10 TB = 80 TB out of 100 TB raw, and any two drives can fail without data loss. The scheme tolerates two failures at 1.25x overhead.

Question 41Huawei Storage Product Portfolio (OceanStor Block, File, Object)

Which Huawei OceanStor family is positioned as all-flash storage for mission-critical block workloads?

  1. OceanStor Dorado Correct answer.
  2. OceanStor 9000
  3. OceanStor Pacific
  4. OceanProtect

Explanation

OceanStor Dorado is Huawei's all-flash storage family, designed for high-performance and low-latency block workloads such as core databases. OceanStor Pacific is a scale-out distributed family for file, object, and analytics workloads, while OceanStor 9000 is a scale-out NAS platform. OceanProtect is backup storage, not a primary block array.

Question 42Huawei Storage Product Portfolio (OceanStor Block, File, Object)

In the OceanStor Pacific portfolio, which protocol is used for native object access?

  1. FC
  2. NFS
  3. SMB
  4. S3 Correct answer.

Explanation

OceanStor Pacific exposes object data through the S3 API, which is the standard interface for object storage applications. NFS and SMB are file-access protocols, while FC is a block-access transport. Pacific's multi-protocol design lets file and object services coexist, but S3 is the object protocol.

Question 43Huawei Storage Product Portfolio (OceanStor Block, File, Object)

Which Huawei storage feature provides array-level active-active access across two data centers?

  1. HyperReplication
  2. HyperSnap
  3. HyperClone
  4. HyperMetro Correct answer.

Explanation

HyperMetro is Huawei's active-active data-center solution: both arrays can serve I/O for the same LUN or file system, and a site failure does not require manual failover. HyperSnap creates point-in-time snapshots, HyperClone creates clones, and HyperReplication performs remote replication rather than active-active access.

Question 44Huawei Storage Product Portfolio (OceanStor Block, File, Object)

Huawei RAID-TP is designed to survive the simultaneous failure of how many member disks in a RAID group?

  1. Four
  2. One
  3. Three Correct answer.
  4. Two

Explanation

RAID-TP uses triple parity, so it can reconstruct data after up to three member disks in the same RAID group fail simultaneously. It is used in OceanStor Dorado to improve rebuild tolerance with large-capacity SSDs. RAID-TP does not imply tolerance of four disk failures.

Question 45Huawei Storage Product Portfolio (OceanStor Block, File, Object)

What is the main effect of adding storage nodes to an OceanStor Pacific scale-out cluster?

  1. All file systems must be recreated from backups
  2. Capacity and performance scale online Correct answer.
  3. Only capacity grows because performance is fixed
  4. The cluster becomes a dual-controller array

Explanation

OceanStor Pacific is a distributed scale-out architecture, so adding nodes increases both usable capacity and aggregate performance. Data and metadata are redistributed across the expanded cluster, and the expansion is performed online. It does not convert the system into a dual-controller array or require file systems to be recreated.

Question 46Huawei Storage Product Portfolio (OceanStor Block, File, Object)

A customer needs a cost-effective Huawei array for mixed block and file workloads, using SSDs for caching and HDDs for capacity. Which family best fits?

  1. OceanStor V6 hybrid flash Correct answer.
  2. OceanStor Pacific distributed storage
  3. OceanStor Dorado all-flash
  4. OceanStor 9000 scale-out NAS

Explanation

OceanStor V6 hybrid flash arrays combine SSD tiering or caching with HDD capacity, which suits mixed block and file workloads that need cost efficiency rather than an all-flash design. Dorado is all-flash, Pacific is scale-out distributed storage, and OceanStor 9000 is scale-out NAS focused on file services. The hybrid V6 family is the portfolio match for the stated requirement.

Question 47Huawei Storage Product Portfolio (OceanStor Block, File, Object)

What is the primary purpose of Huawei SmartMatrix in an OceanStor Dorado system?

  1. Providing controller redundancy and failover Correct answer.
  2. Scheduling backup jobs to OceanProtect
  3. Compressing data before writing to SSDs
  4. Storing object metadata for S3 buckets

Explanation

SmartMatrix is the controller-level reliability architecture in OceanStor Dorado, designed so that a controller failure does not interrupt access to storage. It provides redundant paths and failover between controllers rather than data compression, object metadata, or backup scheduling. Those functions belong to other features and products.

Question 48Huawei Storage Product Portfolio (OceanStor Block, File, Object)

A research team wants one stored data set on OceanStor Pacific to be readable as files by Linux HPC jobs and as objects by S3 applications. Which Pacific capability addresses this requirement?

  1. Multi-protocol unified namespace Correct answer.
  2. SmartTier automatic data tiering
  3. HyperMetro active-active replication
  4. RAID-TP triple-parity disk protection

Explanation

OceanStor Pacific can present the same scale-out namespace through file protocols such as NFS and SMB and through the S3 object API, avoiding a duplicate copy for each access method. RAID-TP and HyperMetro address disk fault tolerance and active-active availability, not multi-protocol data access. SmartTier is a tiering feature and does not provide unified file and object access.

Question 49Huawei Storage Product Portfolio (OceanStor Block, File, Object)

Which Dorado technology is specifically intended to accelerate I/O across the host, controller, and SSD path for flash workloads?

  1. HyperMetro
  2. SmartMatrix
  3. FlashLink Correct answer.
  4. SmartTier

Explanation

FlashLink is Huawei's flash-oriented I/O acceleration technology for OceanStor Dorado, optimizing the end-to-end path between hosts, controllers, and SSDs. SmartMatrix focuses on controller redundancy, HyperMetro on active-active disaster tolerance, and SmartTier on data tiering. FlashLink is the feature tied to flash I/O path acceleration.

Question 50Huawei Storage Product Portfolio (OceanStor Block, File, Object)

Which redundancy schemes are supported by OceanStor Pacific for protecting file and object data?

  1. Erasure coding and replicas Correct answer.
  2. Only three-way replication per object
  3. Only two-way mirroring across nodes
  4. Only RAID-TP within each node

Explanation

OceanStor Pacific supports erasure coding and multiple replica schemes to protect distributed file and object data, allowing redundancy to be chosen according to capacity efficiency and reliability requirements. It is not limited to mirroring, RAID-TP, or three-way replication. RAID-TP is a Dorado RAID scheme, not the Pacific data protection model.