I85PREP
⌘ K
DMCP study materials
DMCP study resources

DMCP practice questions

One question at a time. Build your DMCP 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 and Data Protection

In a RAID 5 array, how is data protection achieved across the member disks?

  1. Data is striped with no redundancy, relying solely on backups for recovery.
  2. Two independent parity blocks are stored per stripe on separate disks.
  3. A full mirror copy of every block is written to a dedicated parity disk.
  4. Parity is distributed across all member disks, allowing recovery from one disk failure. Correct answer.

Explanation

RAID 5 stripes data and parity across all member disks, so the loss of any single disk can be reconstructed from the surviving data and parity. Because parity rotates across disks, no single disk is a dedicated parity bottleneck. RAID 5 tolerates exactly one member disk failure.

Question 2Storage and Data Protection

A backup job writes a full backup every Sunday and incremental backups on other days. Which statement describes the data needed to restore Wednesday's state?

  1. Sunday's full backup plus Wednesday's incremental backup.
  2. Sunday's full backup plus Monday, Tuesday, and Wednesday incrementals. Correct answer.
  3. Only Wednesday's incremental backup.
  4. Tuesday's incremental backup plus Wednesday's incremental backup.

Explanation

Incremental backups capture only changes since the most recent backup of any type, so each incremental depends on the chain before it. Restoring Wednesday therefore requires the Sunday full plus every incremental through Wednesday. A differential scheme would instead need only the full plus the latest differential.

Question 3Storage and Data Protection

What is the primary purpose of the 3-2-1 backup rule?

  1. To keep three copies of data on two media types with one copy offsite. Correct answer.
  2. To verify backups three times per week against two test environments and one production host.
  3. To encrypt backups with three keys held by two administrators and one escrow service.
  4. To run three backup jobs per day across two servers with one retention policy.

Explanation

The 3-2-1 rule is a resilience guideline: maintain at least three copies of data, on two different media or storage types, with one copy stored offsite. This reduces the chance that a single failure, disaster, or ransomware event destroys every copy. It addresses copy count, media diversity, and geographic separation.

Question 4Storage and Data Protection

Which storage characteristic distinguishes block storage from file storage?

  1. Block storage exposes raw addressable blocks that a host formats with its own file system. Correct answer.
  2. Block storage always requires an SMB or NFS client to mount volumes.
  3. Block storage organizes data as a hierarchical directory tree with file-level permissions.
  4. Block storage stores objects with immutable keys and metadata in a flat namespace.

Explanation

Block storage presents raw logical blocks over protocols such as iSCSI or Fibre Channel, and the host operating system applies its own file system. File storage instead presents a shared directory hierarchy through protocols like NFS or SMB. Object storage uses a flat key-value namespace with metadata.

Question 5Storage and Data Protection

In a deduplication system, what does the term 'chunking' refer to?

  1. Encrypting data blocks before they are written to tape.
  2. Compressing each backup stream with a single global dictionary.
  3. Dividing data into variable or fixed segments so duplicate segments can be identified. Correct answer.
  4. Spreading writes evenly across the disks in a RAID group.

Explanation

Deduplication splits incoming data into chunks, either fixed-size or variable-length, and hashes them to detect segments already stored. Duplicate chunks are replaced by references rather than stored again. Variable-length chunking improves detection when insertions shift data boundaries.

Question 6Storage and Data Protection

What does an RPO of 15 minutes mean for a data protection design?

  1. At most 15 minutes of data loss is acceptable after a failure. Correct answer.
  2. Backups must be retained for at least 15 minutes before deletion.
  3. A restore test must be performed every 15 minutes.
  4. The system must recover service within 15 minutes of an outage.

Explanation

Recovery Point Objective defines the maximum tolerable amount of data loss, measured in time. An RPO of 15 minutes means recovery must restore data to a point no more than 15 minutes before the failure. Recovery Time Objective, by contrast, governs how long recovery may take.

Question 7Storage and Data Protection

Which characteristic is typical of LTO tape used for long-term archival?

  1. Random access performance comparable to solid-state drives.
  2. Sequential access with high capacity and a long media shelf life. Correct answer.
  3. Automatic replication to a secondary site on every write.
  4. Mandatory network connectivity for every read operation.

Explanation

LTO tape is a sequential-access medium offering high capacity per cartridge and long archival life when stored properly. Its strength is cost-effective long-term retention and offline copies, not low-latency random access. Reading requires locating the position on the tape, so restore times are longer than disk.

Question 8Storage and Data Protection

What is the role of a snapshot in a storage array?

  1. It permanently deletes older data to reclaim capacity.
  2. It compresses data at rest to reduce power consumption.
  3. It captures a point-in-time image of a volume that can be used for recovery or cloning. Correct answer.
  4. It replaces the need for any offsite backup copy.

Explanation

A snapshot records the state of a volume at a specific moment, typically using copy-on-write or redirect-on-write metadata. It enables fast rollback, cloning, and backup offload. Snapshots usually reside on the same array, so they do not substitute for an independent offsite copy.

Question 9Storage and Data Protection

A company needs to protect a 200 TB database with a nightly backup window of only two hours. Which approach best addresses the window constraint?

  1. Back up only the database log files and skip data files.
  2. Take a storage snapshot and back up from that point-in-time image. Correct answer.
  3. Perform a full backup each night and compress the output.
  4. Switch to weekly full backups with no incrementals.

Explanation

A snapshot creates an instant point-in-time image, allowing the backup to read a consistent copy while production continues. This removes the need to quiesce the database for the full backup duration and shortens the effective window. Compression reduces size but does not solve the time needed to read 200 TB.

Question 10Storage and Data Protection

How does erasure coding differ from traditional mirroring for data protection?

  1. It uses parity fragments so data can be rebuilt from a subset of fragments. Correct answer.
  2. It eliminates all redundancy and relies on replication at the application layer.
  3. It only works with tape media and cannot be used on disk arrays.
  4. It stores identical copies on two disks and requires no parity computation.

Explanation

Erasure coding divides data into fragments and computes parity fragments, so the original can be reconstructed from any sufficient subset. This provides protection with less capacity overhead than full mirroring. The tradeoff is higher compute cost during rebuilds and reads.

Question 11Storage and Data Protection

An administrator configures a backup policy with a grandfather-father-son rotation. What does the 'son' tier typically represent?

  1. Daily backups retained for a short period. Correct answer.
  2. Weekly full backups retained for several weeks.
  3. Annual archives written to immutable media.
  4. Monthly full backups retained for a year or more.

Explanation

In GFS rotation, the son tier holds the most frequent backups, usually daily, kept for a short retention period. The father tier is typically weekly and the grandfather tier monthly or longer. This tiering balances frequent recovery points against long-term retention cost.

Question 12Storage and Data Protection

Which scenario best justifies using continuous data protection (CDP) rather than nightly backups?

  1. A transactional system where losing more than a few minutes of writes is unacceptable. Correct answer.
  2. A file server whose content changes only during a monthly maintenance window.
  3. An archive of read-only reference documents accessed rarely.
  4. A test environment rebuilt from scratch each quarter.

Explanation

CDP captures every write or change, enabling recovery to nearly any point in time and supporting very low RPO. It is justified when the cost of losing even a few minutes of transactions is high. Static or rarely changing data does not need that granularity.

Question 13Storage and Data Protection

What is the main risk of relying on snapshots stored on the same array as the production data?

  1. Snapshots cannot be used to create clones for testing.
  2. Snapshots consume no capacity and therefore cannot be monitored.
  3. Snapshots always degrade read performance to unacceptable levels.
  4. A failure or loss of the array destroys both production data and its snapshots. Correct answer.

Explanation

Snapshots on the same array share the failure domain of the production volumes, so an array loss removes both. They provide fast local rollback but must be complemented by copies on independent media or offsite. This is why snapshot-only strategies are considered incomplete protection.

Question 14Storage and Data Protection

Which protocol is most appropriate for presenting a shared file system to both Windows and Linux clients from a network-attached storage device?

  1. SMB for Windows clients and NFS for Linux clients. Correct answer.
  2. NVMe over Fabrics with raw namespace access.
  3. iSCSI with multipath enabled.
  4. Fibre Channel over a switched fabric.

Explanation

NAS devices present file-level shares, commonly using SMB for Windows and NFS for Linux and UNIX clients. These protocols let multiple clients mount a shared directory hierarchy with permissions. Block protocols such as iSCSI or Fibre Channel present raw volumes instead.

Question 15Storage and Data Protection

A deduplication ratio of 10:1 is reported for a backup repository. What does this ratio indicate?

  1. The repository can hold ten times more disks than a standard array.
  2. Ten percent of the backups failed verification.
  3. Data is compressed to one-tenth of its size by the compression engine alone.
  4. Logical data protected is ten times the physical capacity consumed. Correct answer.

Explanation

A deduplication ratio compares the logical amount of protected data with the physical capacity actually used after deduplication and often compression. A 10:1 ratio means ten units of logical data occupy roughly one unit of physical space. It is a capacity-efficiency metric, not a reliability or compression-only figure.

Question 16Storage and Data Protection

Which practice best protects backup data against ransomware that has compromised production credentials?

  1. Keeping immutable or air-gapped copies that the compromised account cannot modify. Correct answer.
  2. Disabling backup encryption to simplify recovery.
  3. Storing backups on the same domain with the same administrator account.
  4. Relying on snapshots that the production account can delete.

Explanation

Immutable or air-gapped copies cannot be altered or deleted even with compromised production credentials, preserving a clean recovery source. Backup systems should also use separate credentials and networks. Snapshots and backups reachable by the compromised account can be encrypted or deleted by the attacker.

Question 17Storage and Data Protection

What is the purpose of a synthetic full backup?

  1. It writes data directly to tape without any intermediate storage.
  2. It reads all production data from the source each time it runs.
  3. It stores only changed blocks and never produces a full image.
  4. It combines previous backups into a new full image without re-reading the source. Correct answer.

Explanation

A synthetic full is assembled from existing full and incremental backups, producing a new full image without another full read of production. This reduces load on source systems and shortens the backup window. It relies on the integrity of the prior backup chain.

Question 18Storage and Data Protection

A storage administrator must ensure a LUN is accessible only to authorized hosts. Which mechanism is designed for this on a Fibre Channel SAN?

  1. Object bucket policies enforced by the gateway.
  2. Zoning on the fabric combined with LUN masking on the array. Correct answer.
  3. NFS export rules applied at the file server.
  4. SMB share permissions set on the volume.

Explanation

Fibre Channel SAN access control combines fabric zoning, which limits which ports can communicate, with LUN masking, which restricts which hosts see a given LUN. Together they prevent unauthorized hosts from accessing block volumes. File-share and object policies apply to different storage models.

Question 19Storage and Data Protection

Which statement accurately describes the difference between replication and backup?

  1. Replication preserves historical versions by default, while backup keeps only the latest copy.
  2. Replication and backup are identical because both copy data to a second location.
  3. Replication maintains a current copy at another site, while backup preserves point-in-time versions for recovery. Correct answer.
  4. Backup continuously overwrites the target, while replication retains every change.

Explanation

Replication keeps a secondary copy synchronized with production, which supports fast failover but propagates corruption or deletion. Backup captures point-in-time versions retained over time, enabling recovery from logical errors. A robust design uses both, since each covers different failure modes.

Question 20Storage and Data Protection

What does the term 'immutable backup' mean in a data protection context?

  1. The backup cannot be altered or deleted for a defined retention period. Correct answer.
  2. The backup is compressed so it cannot be modified.
  3. The backup is encrypted with a key that is never rotated.
  4. The backup is stored only in memory for fast recovery.

Explanation

Immutability enforces write-once behavior for a retention period, often through object lock or WORM media, so data cannot be changed or removed early. This protects recovery copies from tampering, including ransomware. It is a policy and media control, not a compression or encryption property.

Question 21Storage and Data Protection

A storage team observes that RAID rebuilds on large nearline SAS drives take many hours and stress the array. Which design change most directly reduces exposure during a rebuild?

  1. Convert the array to RAID 0 to improve rebuild speed.
  2. Disable patrol reads to reduce background load.
  3. Use a scheme with dual parity or erasure coding that tolerates an additional failure. Correct answer.
  4. Increase the stripe width while keeping single parity.

Explanation

Long rebuilds create a window where a second failure can cause data loss under single parity. Dual-parity RAID 6 or erasure coding tolerates an additional failure during rebuild, reducing that exposure. Wider single-parity stripes actually increase risk, and RAID 0 offers no protection.

Question 22Storage and Data Protection

An organization wants backups to be recoverable even if the primary site is destroyed and the backup catalog is lost. Which combination best satisfies this requirement?

  1. Replicated backups to a remote site with an independently maintained catalog. Correct answer.
  2. Snapshots on the primary array with a local catalog.
  3. A single tape copy stored in the same data center as production.
  4. Incremental backups stored only on the backup server's local disk.

Explanation

Recovering without the primary site requires copies at an independent location, and recovering without the original catalog requires that catalog metadata also be protected remotely. Replicating both data and catalog to a remote site satisfies both conditions. Local-only copies share the failure domain of the primary site.

Question 23Storage and Data Protection

A deduplication appliance ingests encrypted backup streams where each backup uses a unique encryption key. Why does this reduce deduplication effectiveness?

  1. Encryption forces all chunks to be fixed-length, which defeats hashing.
  2. Encryption prevents the appliance from reading metadata such as file names.
  3. Encrypted streams cannot be written to disk and must go directly to tape.
  4. Identical plaintext produces different ciphertext, so duplicate chunks are not detected. Correct answer.

Explanation

Deduplication compares chunk content, and unique keys make identical plaintext encrypt to different ciphertext, so matching fails. Deduplication effectiveness is preserved by deduplicating before encryption or by using consistent keys across streams. Encryption itself does not prevent disk writes or force fixed chunk sizes.

Question 24Storage and Data Protection

In a stretched cluster with synchronous replication between two sites, what is the primary constraint on achievable RPO and write latency?

  1. The round-trip network latency between sites, since writes must be acknowledged by both. Correct answer.
  2. The number of snapshots retained on the primary array.
  3. The deduplication ratio achieved at the secondary site.
  4. The capacity of the tape library at the secondary site.

Explanation

Synchronous replication requires each write to be committed at both sites before acknowledgment, so inter-site round-trip latency directly affects application write latency. Zero RPO is achieved at the cost of that latency, which limits how far apart the sites can be. Asynchronous replication relaxes this by decoupling acknowledgment from the remote commit.

Question 25Storage and Data Protection

A backup administrator must choose between forward and reverse incremental schemes. Which statement correctly captures a key tradeoff?

  1. Reverse incremental updates the full image each run, shifting more work to every backup but simplifying restores. Correct answer.
  2. Forward incremental requires synthesizing a new full each night, increasing nightly load.
  3. Forward incremental stores only one full image and no incrementals.
  4. Reverse incremental cannot be used with deduplication appliances.

Explanation

Reverse incremental applies changes into the full image on each run, so the latest full is always current and restores need only that image, at the cost of more work per backup. Forward incremental writes small incrementals quickly but requires periodic synthetic fulls and a longer restore chain. Each scheme trades backup cost against restore simplicity.

Question 26Storage and Data Protection

In a DMCP data protection design, which storage characteristic most directly determines the maximum achievable backup throughput from a production array?

  1. The number of configured spare drives in each RAID group
  2. The physical rack unit height of the array controllers
  3. The raw capacity of the array's disk shelves
  4. The sustained write bandwidth of the backup target and the read bandwidth of the source Correct answer.

Explanation

Backup throughput is bounded by the slowest link in the data path, which is typically the sustained read rate of the source and the sustained write rate of the target. Raw capacity, sparing policy, and physical dimensions do not govern how fast data can move during a window.

Question 27Storage and Data Protection

Which statement describes the function of a storage array write cache with battery or capacitor protection during a power loss event?

  1. It disables replication to the remote site until power is restored
  2. It preserves acknowledged writes that have not yet reached persistent media Correct answer.
  3. It flushes all dirty data to disk before acknowledging any host write
  4. It converts synchronous writes into asynchronous writes permanently

Explanation

A protected write cache allows the array to acknowledge host writes before they land on spinning media, and the battery or capacitor holds that dirty data long enough to destage it after power returns. Without protection, acknowledged writes could be lost, which would violate the durability guarantee presented to hosts.

Question 28Storage and Data Protection

A storage administrator is asked to present block storage to an ESXi cluster where multipathing must provide redundant paths. Which combination of elements is required on the host side?

  1. A single HBA with one target port and no multipath plugin
  2. An NFS export with Kerberos authentication enabled
  3. Two or more initiator ports, a multipath plugin, and matching zoning or masking on the array Correct answer.
  4. A CIFS share mounted by each host with a single network interface

Explanation

Multipathing requires multiple initiator ports, a host-side multipath driver or plugin, and array-side zoning or LUN masking that exposes the same LUN through more than one target port. Only then can the host detect and use redundant paths to the same device.

Question 29Storage and Data Protection

Which mechanism allows a storage array to reclaim capacity when a thin-provisioned LUN is deleted or trimmed by the host?

  1. LUN masking
  2. Host bus adapter zoning
  3. LUN snapshot rollback
  4. SCSI UNMAP or space reclamation Correct answer.

Explanation

SCSI UNMAP, sometimes called space reclamation or trim, lets a host tell the array that specific blocks are no longer in use so the array can return them to the free pool. Without it, thin-provisioned capacity stays consumed even after files are deleted.

Question 30Storage and Data Protection

An administrator must protect a 40 TB file server where daily change rates are under 2 percent and the recovery point objective is four hours. Which backup design most efficiently meets the requirement?

  1. Continuous synchronous replication to a second array with no snapshots
  2. Daily incremental backups with a weekly full and periodic synthetic fulls Correct answer.
  3. Monthly full backups retained for one year with no interim copies
  4. Weekly full backups with no incremental jobs

Explanation

With a low change rate, daily incrementals capture the four-hour RPO cheaply, while a weekly full plus synthetic fulls bound recovery time without re-reading the entire source. This balances window pressure against restore complexity.

Question 31Storage and Data Protection

A backup target is configured with deduplication and a retention policy of 90 daily restore points. Which factor most affects the deduplication ratio achieved over that retention period?

  1. The number of tape drives in the library
  2. The rate of unique data change between successive backups Correct answer.
  3. The number of media servers licensed
  4. The physical port speed of the backup network

Explanation

Deduplication gains come from repeated content, so a low change rate produces high ratios while a high change rate produces lower ones. Tape drives, port speed, and license counts influence performance or cost but not how much data is actually duplicated.

Question 32Storage and Data Protection

An organization replicates backups to a secondary site but discovers the secondary copy is unreadable because the encryption keys were never exported. Which design change prevents this outcome?

  1. Storing key material in a separate, escrowed location accessible during disaster recovery Correct answer.
  2. Disabling encryption on the secondary copy only
  3. Compressing the backup catalog before replication
  4. Increasing the replication frequency to every hour

Explanation

Recoverability depends on having the decryption keys available at the recovery site independently of the failed primary. Key escrow or a separate key management service at the secondary site ensures the replicated data can actually be restored.

Question 33Storage and Data Protection

A storage team is evaluating RAID 6 versus RAID 10 for a large pool of 16 TB nearline drives hosting backup data. Which trade-off should drive the decision?

  1. RAID 6 provides better write performance than RAID 10 in all workloads
  2. RAID 6 tolerates two drive failures but has slower rebuilds and higher write penalty; RAID 10 rebuilds faster but halves usable capacity Correct answer.
  3. RAID 10 tolerates two simultaneous drive failures in every configuration
  4. RAID 6 requires no parity calculation and therefore has no write penalty

Explanation

RAID 6 survives two drive failures but pays a parity write penalty and rebuilds large drives slowly, while RAID 10 mirrors data for fast rebuilds at the cost of 50 percent usable capacity. The choice depends on whether capacity efficiency or rebuild risk and write performance dominate.

Question 34Storage and Data Protection

What is the primary purpose of a backup catalog in a data protection environment?

  1. To index which backup copies exist, where they reside, and what they contain Correct answer.
  2. To store the actual backup payload data on tape
  3. To provide block-level replication between arrays
  4. To encrypt data in flight between the client and the media server

Explanation

The catalog is metadata that tracks each backup's identity, location, contents, and retention so restores can locate the right copy. Losing the catalog makes recovery far harder even when the payload data still exists.

Question 35Storage and Data Protection

A company must guarantee that a database can be recovered to any point within the last 24 hours, not just to the previous nightly backup. Which technology directly addresses this requirement?

  1. Journal-based continuous data protection with log shipping Correct answer.
  2. Nightly incremental backups with a two-week retention
  3. Array-based snapshots taken once per day
  4. Weekly full backups with monthly archival copies

Explanation

Journal-based CDP or log shipping captures every change so recovery can target an arbitrary point within the retention window. Nightly backups and daily snapshots only offer discrete recovery points and cannot satisfy an any-point requirement.

Question 36Storage and Data Protection

An administrator observes that a backup job's throughput drops sharply when the job switches from reading a few large files to reading millions of small files. Which explanation best accounts for this behavior?

  1. The network switch has negotiated a lower link speed automatically
  2. The backup target's deduplication ratio has fallen to zero
  3. The tape library has run out of cleaning cartridges
  4. Per-file metadata overhead and reduced sequential read efficiency dominate the workload Correct answer.

Explanation

Small-file workloads incur per-file open, close, and metadata operations and prevent large sequential reads, so effective throughput falls even on the same hardware. This is a workload characteristic rather than a fault in the network or target.

Question 37Storage and Data Protection

Which control most directly limits the blast radius if a malicious insider with backup operator rights attempts to delete retention-locked copies?

  1. Compressing the backup data before writing it
  2. Increasing the backup frequency to every 15 minutes
  3. Adding a second media server in the same data center
  4. Role separation with dual authorization and a compliance-mode object lock Correct answer.

Explanation

Separating duties so no single operator can both request and approve deletion, combined with compliance-mode immutability, prevents unilateral destruction of protected copies. Frequency, compression, and extra media servers do not stop an authorized insider from deleting data.

Question 38Storage and Data Protection

A storage architect must choose between file-level and image-level backup for a large virtualized estate. Which consideration most strongly favors image-level backup?

  1. It captures the entire virtual disk and guest configuration in one operation, simplifying full-system recovery Correct answer.
  2. It removes the need for a backup catalog
  3. It eliminates the need for any application-consistent quiescing
  4. It requires no backup storage because images are stored in memory

Explanation

Image-level backup captures the whole virtual disk plus configuration, which makes bare-metal or full-VM restores straightforward and avoids per-file traversal. Application consistency still requires quiescing, and images still consume storage and require a catalog.

Question 39Storage and Data Protection

An array uses synchronous replication between two sites 300 km apart. Users report that write latency has increased substantially. Which factor best explains this outcome?

  1. Synchronous replication writes to tape before acknowledging
  2. The remote array performs deduplication before acknowledging writes
  3. Synchronous replication compresses data only at the remote site
  4. Each write must be acknowledged by the remote site before it completes locally, so round-trip distance adds directly to latency Correct answer.

Explanation

Synchronous replication requires the remote acknowledgement before the host write is confirmed, so the inter-site round-trip time is added to every write. At 300 km the propagation delay alone makes this significant, which is why long-distance links usually favor asynchronous modes.

Question 40Storage and Data Protection

A data protection team measures an effective deduplication ratio of 4:1 on a repository that also stores compressed, already-deduplicated data from an upstream appliance. Which conclusion is most defensible?

  1. The downstream repository is losing data during ingest
  2. The downstream ratio is low because upstream processing already removed redundancy, so the combined pipeline should be evaluated end to end Correct answer.
  3. The upstream appliance is misconfigured and should be bypassed
  4. Compression always increases the deduplication ratio on the downstream target

Explanation

When an upstream appliance has already deduplicated and compressed the stream, little redundancy remains for the downstream target, so a low ratio there is expected rather than a fault. Capacity planning should assess the combined effect of the whole pipeline instead of judging each stage in isolation.

Question 41Servers and Compute Infrastructure

In a data center, which server form factor is designed to slide into a standardized 19-inch equipment rack and is measured in rack units (U)?

  1. A mainframe cabinet
  2. A blade enclosure
  3. A tower server
  4. A rack-mounted server Correct answer.

Explanation

Rack-mounted servers are built to fit the 19-inch EIA rack standard and their height is expressed in rack units, with 1U equal to 1.75 inches. This standardization allows dense, orderly placement of many compute nodes in a data center row.

Question 42Servers and Compute Infrastructure

A technician needs a server that shares power, cooling, and networking through a chassis backplane while allowing many compute cartridges in a small footprint. Which server architecture best matches this requirement?

  1. Desktop workstation
  2. Blade server system Correct answer.
  3. Rack-optimized 4U storage server
  4. Pedestal server

Explanation

Blade systems place multiple hot-swappable server blades inside a shared enclosure that distributes power and cooling and provides consolidated network and management connectivity via the midplane. This shared infrastructure is the defining characteristic of blade architecture.

Question 43Servers and Compute Infrastructure

Which component on a server motherboard is responsible for initializing hardware and handing control to the bootloader during startup?

  1. The redundant power supply
  2. The UEFI/BIOS firmware Correct answer.
  3. The network interface controller
  4. The baseboard management controller

Explanation

UEFI or legacy BIOS firmware performs power-on self-test, initializes memory and devices, and then locates and launches the bootloader. The BMC operates independently for out-of-band management and does not normally perform the OS boot handoff.

Question 44Servers and Compute Infrastructure

What is the primary purpose of a baseboard management controller (BMC) in a rack server?

  1. To provide out-of-band management independent of the host operating system Correct answer.
  2. To accelerate cryptographic operations for the host CPU
  3. To cache frequently accessed disk blocks for the storage controller
  4. To balance network traffic across multiple physical uplinks

Explanation

The BMC is a dedicated service processor that lets administrators power-cycle, monitor sensors, and open a remote console even when the host OS is down. This out-of-band capability is its core function.

Question 45Servers and Compute Infrastructure

A server is described as having dual redundant hot-swappable power supplies. What operational benefit does this configuration provide?

  1. It removes the need for a UPS in the data center
  2. It doubles the total wattage available to the CPU at all times
  3. It allows one supply to be replaced while the server keeps running on the other Correct answer.
  4. It automatically converts AC input to a higher DC voltage for the drives

Explanation

Redundant hot-swap supplies let the system continue operating on the surviving unit while a failed supply is replaced without powering down. Capacity is shared, not necessarily doubled, and UPS protection is still required.

Question 46Servers and Compute Infrastructure

Which memory technology includes an extra chip that detects and corrects single-bit errors, making it standard for enterprise servers?

  1. GDDR graphics memory
  2. ECC registered DIMMs Correct answer.
  3. Unbuffered non-ECC DIMMs
  4. Non-volatile NVDIMM persistent memory

Explanation

ECC memory uses additional parity bits and a memory controller feature to detect and correct single-bit errors, and registered DIMMs add a buffer for stability at high capacities. This combination is the enterprise server standard.

Question 47Servers and Compute Infrastructure

In a data center, what does the term 'east-west traffic' describe?

  1. Traffic between servers within the same data center Correct answer.
  2. Traffic between the data center and a public cloud provider
  3. Traffic between the data center and end users on the internet
  4. Traffic replicated to a disaster recovery site in another region

Explanation

East-west traffic refers to server-to-server communication inside the data center, such as application tiers calling each other or distributed systems replicating state. North-south traffic, by contrast, crosses the data center boundary to clients or other sites.

Question 48Servers and Compute Infrastructure

Which RAID level provides mirroring by writing identical data to two drives, offering redundancy with a 50 percent usable capacity in a two-drive array?

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

Explanation

RAID 1 mirrors data across paired disks so either disk can fail without data loss, at the cost of halving usable capacity. RAID 0 provides no redundancy, while RAID 5 and 6 use parity across three or more disks.

Question 49Servers and Compute Infrastructure

An administrator wants to reduce the physical core count licensed per host while still delivering many logical processors to virtual machines. Which CPU feature should be enabled and planned for?

  1. Demand-based switching of DIMM voltage
  2. Turbo boost
  3. Non-uniform memory access
  4. Simultaneous multithreading Correct answer.

Explanation

Simultaneous multithreading exposes two logical processors per physical core, which increases schedulable vCPUs without adding physical cores. Licensing and capacity planning must account for whether the hypervisor and vendor count logical or physical processors.

Question 50Servers and Compute Infrastructure

A workload requires very low latency access to a shared dataset across several nodes in a cluster. Which infrastructure choice most directly addresses this need?

  1. A storage area network with fibre channel
  2. Larger local boot drives on each node
  3. A higher-wattage redundant power supply
  4. A low-latency RDMA-capable network fabric between nodes Correct answer.

Explanation

Remote direct memory access over fabrics such as InfiniBand or RoCE lets nodes read and write remote memory with minimal CPU involvement and very low latency. This directly targets inter-node communication latency for shared workloads.