I85PREP
⌘ K
Dell Cloud Infrastructure Design study materials
Dell Cloud Infrastructure Design study resources

Dell Cloud Infrastructure Design practice questions

One question at a time. Build your Dell Cloud Infrastructure Design understanding.

First 50 of 190 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 1Cloud infrastructure design, alternatives and lifecycle trade-offs

In a DIY private cloud, which compute design choice MOST directly affects per-VM density and licensing cost on a hypervisor licensed by physical core?

  1. Choice of memory DIMM color
  2. Selection of CPU SKU (core count and frequency) Correct answer.
  3. Vendor of the bezel
  4. Length of patch cables

Explanation

When hypervisors or databases license per physical core (e.g., VCF per-core, Oracle), the SKU choice (high-core vs. high-frequency parts) drives both density and license spend. Architects size CPU based on workload type (frequency-bound vs. throughput-bound) and license model.

Question 2Cloud infrastructure design, alternatives and lifecycle trade-offs

Which Dell PowerEdge platform is purpose-built as an 8-GPU AI training node and is commonly chosen in DIY AI cloud designs?

  1. R660
  2. XE9680 Correct answer.
  3. T150
  4. VxRail E660

Explanation

PowerEdge XE9680 is a 6U server hosting eight NVIDIA H100/H200/B100 SXM GPUs interconnected by NVLink/NVSwitch - the standard DIY building block for large-model training in PowerEdge designs.

Question 3Cloud infrastructure design, alternatives and lifecycle trade-offs

In a DIY storage tier, which storage type is BEST suited to large unstructured data sets, with HTTP/S3 access patterns and high parallelism?

  1. Block storage on PowerStore
  2. File storage on PowerStore
  3. Object storage on Dell ObjectScale Correct answer.
  4. Tape library

Explanation

Object storage (Dell ObjectScale on Kubernetes, ECS) exposes HTTP/S3 APIs and is designed for massively parallel access to billions of unstructured objects (logs, media, backups, AI training data). Block and file are optimized for VMs and shared file workloads.

Question 4Cloud infrastructure design, alternatives and lifecycle trade-offs

Which leaf-spine network design property MOST helps deliver predictable east-west latency in a cloud fabric?

  1. Daisy-chained access switches
  2. Equal-cost multipath (ECMP) with consistent oversubscription per leaf Correct answer.
  3. Single core switch as a chokepoint
  4. Spanning Tree blocking on every link

Explanation

Leaf-spine fabrics use ECMP across spines so any leaf reaches any other leaf in a constant hop count, with predictable latency. Architects size oversubscription consistently per leaf to keep performance uniform.

Question 5Cloud infrastructure design, alternatives and lifecycle trade-offs

Which protocol pair is commonly used to build a modern overlay/underlay data-center fabric that supports multi-tenant Layer-2 extension across an IP underlay?

  1. VXLAN with BGP EVPN Correct answer.
  2. GRE with RIPv2
  3. IPSec with OSPFv1
  4. PPP with HDLC

Explanation

VXLAN encapsulates Ethernet frames in UDP for L2 over L3 transport, while BGP EVPN distributes MAC/IP reachability information as the control plane. The combination is the de-facto standard for multi-tenant cloud DC fabrics, including those built with Dell PowerSwitch.

Question 6Cloud infrastructure design, alternatives and lifecycle trade-offs

Which technology decouples the network control plane from the forwarding plane, enabling programmable cloud networks?

  1. Software-Defined Networking (SDN) Correct answer.
  2. MPLS
  3. VLAN trunking
  4. STP

Explanation

SDN separates the control plane (which decides) from the data plane (which forwards), exposing programmable APIs (often OpenFlow or NETCONF) to orchestrate the network. VMware NSX, Cisco ACI, and Dell SmartFabric Services are SDN/SDDC examples.

Question 7Cloud infrastructure design, alternatives and lifecycle trade-offs

An architect is sizing a Dell PowerStore-based block tier for a database. Which two performance dimensions should be modeled together to avoid unexpected throttling?

  1. Capacity and BTU/h
  2. IOPS and bandwidth (MB/s) Correct answer.
  3. Number of users and time zone
  4. Drive enclosure color and rack unit

Explanation

Storage arrays have IOPS ceilings and bandwidth ceilings that interact. A small-block OLTP workload may be IOPS-bound, while a large-block analytics scan is bandwidth-bound. Sizing both ensures the array meets the workload across mixed access patterns.

Question 8Cloud infrastructure design, alternatives and lifecycle trade-offs

Which RAID level provides striping with two parity blocks and is commonly chosen in DIY cloud designs to tolerate two simultaneous drive failures on capacity-tier media?

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

Explanation

RAID 6 stripes data with dual parity and tolerates two concurrent drive failures, an important property when rebuild times on large drives stretch into many hours. RAID 5 tolerates only one failure and increases the risk of a second failure during rebuild on large drives.

Question 9Cloud infrastructure design, alternatives and lifecycle trade-offs

Which storage protocol provides parallel, low-latency, RDMA-style access to flash arrays over an Ethernet fabric and is increasingly used in cloud DIY designs?

  1. iSCSI
  2. NVMe-oF (NVMe/TCP, NVMe/RoCE, NVMe/FC) Correct answer.
  3. SMB1
  4. FTP

Explanation

NVMe over Fabrics extends the NVMe command set across Ethernet (TCP, RoCEv2) or Fibre Channel, exposing parallel queues and low latency that match NVMe-class media. iSCSI works but is older and serializes queue handling differently.

Question 10Cloud infrastructure design, alternatives and lifecycle trade-offs

Which compute virtualization concept allows a single physical NIC to expose multiple isolated PCIe functions to VMs?

  1. NUMA
  2. SR-IOV Correct answer.
  3. TPM
  4. BIOS POST

Explanation

Single Root I/O Virtualization (SR-IOV) lets a NIC expose multiple Virtual Functions (VFs) directly to VMs, bypassing the hypervisor for line-rate packet performance. Architects choose SR-IOV for low-latency networking workloads.

Question 11Cloud infrastructure design, alternatives and lifecycle trade-offs

Which Dell software-defined storage platform pools direct-attached drives across PowerEdge servers and exposes block storage to compute nodes?

  1. PowerFlex Correct answer.
  2. PowerStore-T
  3. PowerScale
  4. Data Domain

Explanation

Dell PowerFlex is a software-defined block storage platform that pools NVMe/SSD across PowerEdge servers and presents low-latency block volumes to virtualized or bare-metal consumers. PowerStore-T is a unified array, PowerScale is scale-out NAS, and Data Domain is purpose-built backup.

Question 12Cloud infrastructure design, alternatives and lifecycle trade-offs

Which fabric design choice typically MAXIMIZES cluster east-west bandwidth in a high-performance cloud rack?

  1. 10 GbE access with 1 GbE uplinks
  2. 25/100 GbE leaf-spine with non-blocking spine bandwidth Correct answer.
  3. FastEthernet stack
  4. USB tethering

Explanation

Modern cloud designs use 25 GbE (or 100 GbE) leaf access with 100/400 GbE spine uplinks and non-blocking ratios to keep east-west traffic uncongested. Lower-speed designs become bottlenecks for HCI vMotion, vSAN, PowerFlex, and AI traffic.

Question 13Cloud infrastructure design, alternatives and lifecycle trade-offs

Which design pattern offloads network virtualization, security, and storage protocols from server CPUs onto dedicated silicon close to the NIC?

  1. BIOS update
  2. Smart NIC / DPU (e.g., NVIDIA BlueField, AMD Pensando) Correct answer.
  3. RAID battery backup unit
  4. Faster fans

Explanation

Smart NICs / DPUs (BlueField, Pensando) integrate ARM cores plus NIC ASIC to offload SDN, encryption, telemetry, and storage processing from the host CPU, reclaiming cycles for tenant workloads. They are increasingly used as the cloud-fabric pivot point.

Question 14Cloud infrastructure design, alternatives and lifecycle trade-offs

Which design choice BEST supports a zero-trust security model in a DIY private cloud?

  1. Single flat L2 broadcast domain for all tenants
  2. Microsegmentation with identity-based policies between every workload Correct answer.
  3. Trust by VLAN ID
  4. Disable all logging

Explanation

Zero trust assumes no implicit trust based on network location. Microsegmentation enforces identity- and context-aware policies workload-to-workload (e.g., VMware NSX, Dell SmartFabric, DPU-enforced policy), limiting lateral movement after a breach.

Question 15Cloud infrastructure design, alternatives and lifecycle trade-offs

Which Dell scale-out NAS platform is commonly chosen in DIY cloud designs for unstructured data analytics and AI training data lakes?

  1. PowerScale (formerly Isilon) Correct answer.
  2. PowerVault
  3. Unity XT
  4. PowerEdge T150

Explanation

PowerScale (formerly Isilon) provides scale-out OneFS NAS with parallel NFS/SMB/HDFS/S3 access across nodes - well suited to AI training, analytics, and media workflows. PowerVault is entry SAN/DAS, Unity XT is unified midrange, and T150 is a tower compute server.

Question 16Cloud infrastructure design, alternatives and lifecycle trade-offs

Which network feature allows a single port to be partitioned into multiple logical ports each with its own MAC and bandwidth allocation, useful in DIY hyperconverged designs to separate management, storage, and tenant traffic?

  1. NPAR (NIC Partitioning) Correct answer.
  2. STP
  3. Jumbo frames
  4. WoL

Explanation

NIC Partitioning (NPAR) divides a single physical port into multiple logical functions, each with bandwidth and traffic-class allocation, so a 25/100 GbE port can serve management, storage, vMotion, and tenant traffic with QoS.

Question 17Cloud infrastructure design, alternatives and lifecycle trade-offs

Which power and cooling consideration is MOST important in a high-density DIY GPU rack design?

  1. Color of the rack
  2. Direct Liquid Cooling (DLC) and rack-level power budgets above 30 kW Correct answer.
  3. Number of empty 1U slots
  4. Use of consumer-grade UPS

Explanation

AI-class racks routinely exceed 30 kW and require DLC or rear-door heat exchangers plus matched PDU/branch-circuit design. Architects work with facilities to validate kW per rack and cooling capacity before placing GPU servers.

Question 18Cloud infrastructure design, alternatives and lifecycle trade-offs

Which design metric expresses how efficiently a data center turns input electrical power into IT compute power?

  1. RTO
  2. PUE (Power Usage Effectiveness) Correct answer.
  3. MTBF
  4. RPO

Explanation

PUE = total facility power / IT equipment power. Lower (closer to 1.0) is better. Sustainable cloud designs minimize PUE through DLC, hot/cold aisle containment, free-air cooling, and right-sized power distribution.

Question 19Cloud infrastructure design, alternatives and lifecycle trade-offs

An architect must size compute hosts for a workload that requires 2 TB of usable RAM at peak with 25% headroom. How much physical RAM should the host pool provide as a minimum target?

  1. 1.5 TB
  2. 2 TB
  3. 2.5 TB Correct answer.
  4. 5 TB

Explanation

2 TB usable plus 25% headroom = 2 TB / (1 - 0.25) is incorrect; the correct calculation is 2 TB * 1.25 = 2.5 TB minimum physical capacity to satisfy peak demand and reserved headroom for HA, bursts, and overhead.

Question 20Cloud infrastructure design, alternatives and lifecycle trade-offs

Which design property MOST helps achieve N+1 redundancy at the cluster level for a DIY virtualization compute pool?

  1. All hosts at 100% memory utilization
  2. Reserve enough capacity to absorb the failure of one host without violating SLA Correct answer.
  3. Disable HA
  4. Run workloads only on weekdays

Explanation

N+1 sizing reserves capacity equivalent to one host in the cluster. If any single host fails, surviving hosts can absorb its workload while still meeting SLOs. Architects size CPU and memory to leave at least 1/N headroom plus normal burst overhead.

Question 21Cloud infrastructure design, alternatives and lifecycle trade-offs

Which choice BEST balances performance and resilience for a database tier on Dell PowerEdge with PERC controllers?

  1. RAID 0 across 24 NVMe drives
  2. RAID 10 across paired NVMe drives with battery-backed write cache Correct answer.
  3. JBOD with no controller
  4. Single drive without RAID

Explanation

RAID 10 mirrors and stripes for high performance and tolerates multiple non-paired drive failures. PERC controllers with battery- or super-cap-backed cache further improve write latency. RAID 0 has no redundancy and a single drive has no resiliency.

Question 22Cloud infrastructure design, alternatives and lifecycle trade-offs

Which design technique makes a stateless web tier resilient and easy to scale horizontally in a DIY cloud?

  1. Place a sticky session on every host
  2. Externalize state (sessions, cache) and front instances with a load balancer Correct answer.
  3. Run all instances on one VM
  4. Disable the load balancer to reduce a hop

Explanation

Stateless instances can be added or removed freely if state is externalized to a session store/cache and they sit behind a load balancer. This pattern enables auto-scaling and rolling upgrades with zero downtime.

Question 23Cloud infrastructure design, alternatives and lifecycle trade-offs

Which design control specifically protects against silent data corruption on long-term object data in a DIY archive tier?

  1. BIOS Secure Boot
  2. End-to-end checksums and periodic scrub/repair Correct answer.
  3. ECC RAM
  4. TPM

Explanation

Object stores like Dell ObjectScale and ECS apply end-to-end checksums on every object and periodic scrub/repair operations to detect and fix bit rot. ECC RAM addresses memory errors but not on-disk silent corruption.

Question 24Cloud infrastructure design, alternatives and lifecycle trade-offs

Which Dell tier and component pairing is MOST appropriate for a high-throughput backup target with global deduplication?

  1. PowerVault DAS
  2. PowerProtect Data Domain Correct answer.
  3. ObjectScale
  4. PowerEdge T150

Explanation

PowerProtect Data Domain is Dell's purpose-built backup appliance with variable-length, global dedup, replication, and air-gap features. Object stores work for archive/cloud-tier but Data Domain is the canonical backup target with the highest dedup ratios.

Question 25Cloud infrastructure design, alternatives and lifecycle trade-offs

Which Dell PowerSwitch design feature provides multi-chassis link aggregation between top-of-rack pairs without spanning-tree blocking?

  1. VLT (Virtual Link Trunking) Correct answer.
  2. RSTP
  3. PVST+
  4. MAC Spoofing

Explanation

Dell VLT lets a pair of switches act as one logical switch from a downstream device's perspective, providing active-active uplinks without STP blocking. It is widely used to dual-home servers/HCI nodes in DIY designs.

Question 26Cloud infrastructure design, alternatives and lifecycle trade-offs

Which statement BEST distinguishes Converged Infrastructure (CI) from Hyper-Converged Infrastructure (HCI)?

  1. CI keeps compute, storage, and network as discrete validated components; HCI collapses storage into the compute layer using software-defined storage Correct answer.
  2. CI is virtual; HCI is physical
  3. CI runs only Linux; HCI runs only Windows
  4. There is no difference

Explanation

The first choice captures the stated conventional distinction: converged infrastructure integrates relevant discrete components, while a hyperconverged design integrates software-defined storage with compute under its architecture. Actual product variants can be disaggregated; do not assume every branded offering removes all separate storage hardware or integration responsibility.

Question 27Cloud infrastructure design, alternatives and lifecycle trade-offs

Which Dell HCI appliance is jointly engineered with VMware and is the primary HCI building block for vSphere/vSAN customers?

  1. VxRail Correct answer.
  2. PowerVault
  3. Unity XT
  4. PowerEdge XR8000

Explanation

VxRail is the Dell + VMware co-engineered HCI appliance running vSphere, vSAN, and VxRail HCI System Software (VxRail Manager). It is the canonical HCI building block for vSphere customers and is also the foundation of Dell APEX Private/Hybrid Cloud.

Question 28Cloud infrastructure design, alternatives and lifecycle trade-offs

Which Dell HCI option enables independent scaling of compute and storage by using compute-only nodes plus external SDS?

  1. VxRail dynamic nodes with PowerStore/PowerMax/Unity/PowerFlex backend Correct answer.
  2. VxRail with no storage at all
  3. Cloud-in-a-Box only
  4. Tape-only design

Explanation

VxRail dynamic nodes are compute-only HCI nodes that consume external Dell storage (PowerStore, PowerMax, Unity, PowerFlex), letting customers scale compute and storage independently rather than buying both in fixed ratios.

Question 29Cloud infrastructure design, alternatives and lifecycle trade-offs

Which platform delivers a full VMware Cloud Foundation (VCF) stack - vSphere, vSAN, NSX, and Aria - as a turnkey integrated cloud on Dell hardware?

  1. Dell APEX Cloud Platform / VCF on VxRail Correct answer.
  2. PowerVault standalone
  3. Data Domain
  4. PowerEdge T350

Explanation

Dell APEX Cloud Platform for VMware (and VCF on VxRail) packages VMware Cloud Foundation - vSphere, vSAN, NSX, Aria - as an integrated, lifecycle-managed turnkey on PowerEdge / VxRail. It is Dell's primary VCF-based hybrid cloud building block.

Question 30Cloud infrastructure design, alternatives and lifecycle trade-offs

Which design decision is unique to HCI that does NOT apply to traditional CI?

  1. Choice of Active Directory schema
  2. Selecting cache vs. capacity drive ratios per HCI node and disk-group/storage-policy design Correct answer.
  3. Choice of VLAN ID for management
  4. Choice of patch panel

Explanation

HCI builds storage from drives in compute nodes, so per-node disk groups, cache:capacity ratios, and storage policies (e.g., vSAN FTT/FTM, PowerFlex protection domains) are the storage design surface. CI uses dedicated arrays where these decisions live in the array.

Question 31Cloud infrastructure design, alternatives and lifecycle trade-offs

What is the primary value proposition of a 'Cloud-in-a-Box' appliance compared to a DIY private cloud?

  1. Lower performance ceiling
  2. Pre-validated, lifecycle-managed turnkey delivery with simplified day-2 operations Correct answer.
  3. Custom firmware per node
  4. More cabling decisions

Explanation

Cloud-in-a-Box (e.g., APEX Cloud Platforms, Azure Stack HCI integrated systems) ships pre-integrated, pre-tested, and lifecycle-managed by the vendor. The customer trades some flexibility for faster time-to-value and simpler operations.

Question 32Cloud infrastructure design, alternatives and lifecycle trade-offs

Which scenario BEST justifies choosing CI (e.g., VxBlock) over HCI?

  1. Edge ROBO deployment with three nodes and minimal storage
  2. Large enterprise needing very large, independently scaled tier-1 SAN with deep array services and a separate compute farm Correct answer.
  3. Single-server dev workstation
  4. Remote office with satellite Internet only

Explanation

CI shines when customers need huge, independently scaled, feature-rich block/file SANs (PowerMax, PowerStore, PowerScale) plus a discrete compute farm. HCI's compute-collocated SDS suits more horizontally scaled, mid-tier or VDI workloads.

Question 33Cloud infrastructure design, alternatives and lifecycle trade-offs

Which feature in vSAN Express Storage Architecture (ESA) on VxRail is specifically optimized for NVMe TLC flash and reduces write amplification?

  1. RAID 6 on HDD
  2. Single-tier all-NVMe layout with adaptive log structure and erasure coding Correct answer.
  3. Cache tier on SATA HDD
  4. Floppy tier

Explanation

vSAN ESA replaces the legacy two-tier (cache + capacity) OSA model with a single-tier all-NVMe design, log-structured writes, and inline compression/erasure coding. It reduces write amplification and lifts performance on NVMe TLC drives in VxRail.

Question 34Cloud infrastructure design, alternatives and lifecycle trade-offs

Which fault-tolerance setting on a vSAN cluster determines how many simultaneous host failures the storage policy can tolerate?

  1. FTT (Failures To Tolerate) Correct answer.
  2. MTU
  3. PVID
  4. QoS

Explanation

FTT (Failures To Tolerate) is the vSAN storage-policy attribute defining how many simultaneous host/disk failures a VM's data can survive (FTT=1 needs 3 hosts, FTT=2 needs 5, etc.). Architects choose FTT and FTM (mirror vs. erasure code) per workload.

Question 35Cloud infrastructure design, alternatives and lifecycle trade-offs

Which minimum number of hosts is required for a vSAN cluster using FTT=1 with mirror (RAID 1) protection?

  1. 1
  2. 2
  3. 3 Correct answer.
  4. 5

Explanation

FTT=1 with mirroring requires at least 3 hosts: two for the data copies plus one for the witness. Architects must size cluster minimums into HCI designs because too-few-host clusters waste capacity or fail policy enforcement.

Question 36Cloud infrastructure design, alternatives and lifecycle trade-offs

Which element of a VxRail deployment is responsible for joint lifecycle management of firmware, drivers, and ESXi/vSAN versions?

  1. VxRail Manager / HCI System Software Correct answer.
  2. Putty SSH client
  3. iDRAC Group Manager only
  4. Operating system patch script

Explanation

VxRail HCI System Software (anchored by VxRail Manager) provides continuously-validated state lifecycle management - coordinating ESXi, vSAN, BIOS, and firmware updates as a single bundle. This is a key reason customers choose VxRail over DIY vSAN.

Question 37Cloud infrastructure design, alternatives and lifecycle trade-offs

Which HCI platform from Dell is positioned for Microsoft hybrid cloud customers and runs Azure Local (formerly Azure Stack HCI)?

  1. Dell APEX Cloud Platform for Microsoft Azure Correct answer.
  2. VxRail Standard
  3. PowerScale
  4. PowerProtect DD

Explanation

Dell APEX Cloud Platform for Microsoft Azure is purpose-built to run Azure Local (formerly Azure Stack HCI) with deep Azure Arc integration, lifecycle automation, and storage innovations. VxRail Standard targets vSphere/vSAN.

Question 38Cloud infrastructure design, alternatives and lifecycle trade-offs

Which Dell APEX Cloud Platform variant targets Red Hat OpenShift and uses PowerFlex as the storage layer?

  1. Dell APEX Cloud Platform for Red Hat OpenShift Correct answer.
  2. Dell APEX Cloud Platform for Mainframe
  3. VxRail E Series only
  4. PowerVault ME5

Explanation

Dell APEX Cloud Platform for Red Hat OpenShift runs OpenShift with PowerFlex as the integrated storage layer, providing a turnkey, jointly-supported on-premises Kubernetes platform. It targets containerized workloads with enterprise persistent storage.

Question 39Cloud infrastructure design, alternatives and lifecycle trade-offs

Which factor MOST influences the HCI vs. CI vs. DIY decision for a customer with strong skills in their network and storage teams and unique workload tuning needs?

  1. Marketing collateral preference
  2. Operational maturity, customization needs, and willingness to own integration Correct answer.
  3. Vendor logo color
  4. Frequency of vendor newsletters

Explanation

When a customer has strong silo skills and unique tuning needs, DIY (or CI) often delivers more knobs. HCI/Cloud-in-a-Box trades knobs for simpler operations. Architects assess operational maturity and customization needs to pick the right approach.

Question 40Cloud infrastructure design, alternatives and lifecycle trade-offs

Which design pattern allows a stretched HCI cluster to provide site-level resiliency between two data centers?

  1. Single-site cluster with one fault domain
  2. Stretched cluster with synchronous replication and a witness in a third location Correct answer.
  3. Tape archive only
  4. Periodic offline copy via USB

Explanation

vSAN/PowerFlex stretched clusters synchronously mirror data across two sites with a witness node in a third location to break tie votes. This delivers near-zero RPO and automated failover for site-level resilience within metro distances.

Question 41Cloud infrastructure design, alternatives and lifecycle trade-offs

Which workload type is a CLASSIC fit for HCI rather than CI?

  1. Mainframe DB2
  2. VDI / Virtual Desktop Infrastructure scale-out Correct answer.
  3. Tier-0 mainframe transaction processing
  4. Bulk-tape archival

Explanation

VDI is a textbook HCI workload: scale-out, predictable per-desktop resource profiles, and benefit from collocated SDS. Mainframe and tier-0 z/OS workloads do not run on HCI, and tape archival is a separate domain.

Question 42Cloud infrastructure design, alternatives and lifecycle trade-offs

Which Dell platform extends a hyperscaler-style SDS philosophy and supports both two-layer (compute + storage) and HCI deployment topologies?

  1. PowerFlex Correct answer.
  2. PowerVault
  3. Unity XT
  4. PowerEdge T150

Explanation

PowerFlex supports two-layer (storage-only and compute-only nodes) and HCI (combined) deployment models, making it flexible for varied workload mixes - a hallmark of the cloud-native SDS approach.

Question 43Cloud infrastructure design, alternatives and lifecycle trade-offs

Which of the following BEST describes the Dell ObjectScale platform as an HCI-style storage building block?

  1. A Kubernetes-native, S3-compatible scale-out object store deployable on Dell hardware Correct answer.
  2. A tape jukebox controller
  3. A NIC vendor
  4. A hypervisor

Explanation

ObjectScale is Dell's Kubernetes-native, S3/Swift-compatible scale-out object storage software. It can be deployed on validated Dell hardware to provide on-prem object storage that mirrors hyperscaler S3 patterns.

Question 44Cloud concepts, architecture and service models

Which best describes a HYBRID cloud per NIST SP 800-145?

  1. Two or more distinct cloud infrastructures bound by standardized or proprietary technology that enables data and application portability Correct answer.
  2. A single private cloud only
  3. Two virtual machines on the same host
  4. Public cloud only with no on-prem components

Explanation

NIST defines hybrid cloud as a composition of two or more distinct cloud infrastructures (private, community, public) that remain unique entities but are bound by technology enabling data and application portability (e.g., cloud bursting).

Question 45Cloud concepts, architecture and service models

Which design consideration MOST limits the use of public cloud bursting for an existing on-prem database?

  1. Color of the public-cloud console
  2. Data gravity - moving large datasets and meeting low-latency app-to-DB requirements between sites Correct answer.
  3. Number of HTTP cookies on the website
  4. Browser version of the architect

Explanation

Data gravity is the property that large datasets pull workloads toward them. Bursting compute to a public cloud while leaving the database on-prem creates expensive egress and high inter-site latency, often making bursting impractical without a hybrid storage strategy.

Question 46Cloud concepts, architecture and service models

Which DR site posture has the HIGHEST recovery cost AND the LOWEST RTO/RPO?

  1. Cold site
  2. Warm site
  3. Hot site (active-active or near-real-time replication) Correct answer.
  4. No site

Explanation

A hot site runs current workload copies and continuously replicated data, delivering near-zero RTO/RPO. It costs the most because the secondary capacity sits ready. Warm runs partial workload, cold has only space and connectivity.

Question 47Cloud concepts, architecture and service models

Which metric expresses 'the maximum tolerable amount of data that may be lost from the most recent backup or replica' in DR design?

  1. RPO (Recovery Point Objective) Correct answer.
  2. RTO (Recovery Time Objective)
  3. MTTR
  4. TCO

Explanation

RPO defines the maximum tolerable data loss measured in time between the disaster and the most recent recoverable copy. RTO is the time-to-recover. Both are essential DR design constraints.

Question 48Cloud concepts, architecture and service models

Which DR strategy uses the public cloud as a recovery target without running steady-state production workloads there?

  1. Disaster Recovery as a Service (DRaaS) / pilot light pattern Correct answer.
  2. Active-active in two on-prem sites
  3. Tape-only DR
  4. Single-site only

Explanation

DRaaS / pilot light keeps a minimal footprint (replicated data, dormant compute) in the public cloud, and scales out the cloud workloads only on failover. It cuts steady-state cost while delivering reasonable RTO.

Question 49Cloud concepts, architecture and service models

Which design consideration is unique to MULTICLOUD compared to single-cloud designs?

  1. Need for cross-cloud identity, networking, and policy parity, plus FinOps to compare costs across providers Correct answer.
  2. DNS no longer needed
  3. Compute and storage cease to matter
  4. BGP no longer used

Explanation

Multicloud demands consistent identity (federation), networking (transit, VPNs, peering), policy (compliance per region/provider), and FinOps to compare and optimize cost across providers. Single-cloud designs ignore the cross-provider seams.

Question 50Cloud concepts, architecture and service models

Which cost-optimization practice specifically applies to public-cloud workloads with predictable steady-state demand?

  1. Reserved Instances / Savings Plans / Committed Use Discounts Correct answer.
  2. Spot/preemptible-only purchases
  3. Pay-as-you-go on-demand only
  4. Bare-metal lab usage

Explanation

When workloads run steadily, Reserved Instances (AWS), Savings Plans (AWS), Reserved VMs (Azure), or Committed Use Discounts (GCP) reduce per-hour cost by 30-70% versus on-demand. Architects model commitment terms during business value analysis.