Free D-PE-FN-01 practice test: 30 questions with answers

BOSS is a dedicated boot device that mirrors two M.2 modules, so the operating system can survive a single module failure while data stays on the main drives. The nearline SAS drive holds bulk data without a boot mirror, the RAID 0 stripe has no redundancy, and an optical drive is not a redundant boot device.

The backplane is the board in the drive cage that links each bay to the storage controller and distributes power and signals to the drives. Firmware storage, AC-to-DC conversion and airflow belong to the system board, power supplies and fans, so those options describe other components.

A host bus adapter passes each drive through to the operating system without hardware RAID, leaving redundancy to software. A PERC with a RAID 5 virtual disk applies hardware RAID, the BOSS card presents only a mirrored boot volume, and the LCD module reports status without handling storage.

High availability means services keep running through component failures, which redundancy such as dual power supplies and RAID supports. Scalability describes growing capacity, manageability describes remote monitoring and control, and density describes how many servers fit in a rack.

The leading letter in a PowerEdge model name identifies the form factor, so R denotes rack and T denotes tower. Socket counts, memory capacity and firmware branches are not encoded in that letter.

Blade servers slide into a shared chassis that provides power, cooling and networking, which gives high density in a small footprint. Towers and wall-mounted edge gateways each use their own enclosure, and desktop workstations are not built for dense shared infrastructure.

A SAN presents block-level storage to servers across a dedicated storage network such as Fibre Channel or iSCSI. DAS attaches to one server directly, NAS shares files over the ordinary LAN, and a tape library is not a block storage network.

NAS provides file-level shares over standard network protocols such as SMB and NFS, which suits shared project folders for mixed clients. A SAN delivers raw block volumes the team does not need, direct attached drives stay isolated per workstation, and a hypervisor datastore is not a general file service for workstations.

A hypervisor sits between the physical hardware and the guest operating systems, allocating CPU, memory and devices to each virtual machine. Firmware and boot settings belong to BIOS or UEFI, switch ports belong to network switches, and RAID metadata belongs to the storage controller.

An incoming write arrives at the network adapter, is staged in system memory and processed by the CPU, and then passes through the storage controller to the drives. The other sequences put drives, power or firmware in positions that the write path never uses.

Balanced, matched DIMMs let the memory controller use multiple channels in parallel, which improves bandwidth. Skipping the POST memory test, reducing heat per module and avoiding BIOS updates are not reasons for balanced population.

Redundant supplies are hot-swappable, so the failed unit can be replaced while the healthy supply carries the load. Shutting down, removing the only working supply, or waiting for a window would cause avoidable downtime or leave the server without power.

Direct liquid cooling circulates coolant through cold plates mounted on the hot components, which removes heat that air cannot handle at high density. Immersion submerges whole systems rather than using cold plates, and the other options rely on fans or passive heat pipes instead of cold plates.

Testing against a supported firmware and driver baseline, then rolling out in phases, limits the risk of a bad update reaching the whole fleet. Updating everything at once, skipping release notes and leaving components at mixed unknown versions all increase risk.

A graceful OS shutdown lets applications write out cached data before power is removed, while a forced power-off can lose data still held in memory. Pulling the cord is not a recommended first step, iDRAC keeps working after the host OS halts, and a power loss does not always trigger a RAID rebuild.

Out-of-band management uses the iDRAC controller, which runs independently of the host OS, so it can power-cycle the server and open a remote console. The OpenManage agent depends on the failed OS, local access is not remote, and the workload's own application cannot control the hardware.

UEFI adds Secure Boot and uses GPT partition tables, which support boot disks larger than the 2 TB limit of MBR-based legacy boot. UEFI is not a 16-bit compatibility layer, it does not restrict boot to optical media, and it still needs firmware updates.

OpenManage Server Administrator is an in-band agent installed on the host OS that reports hardware health from inside that OS. The iDRAC interface is out-of-band, OpenManage Enterprise is a central console rather than a host agent, and Lifecycle Controller is used mainly for provisioning and pre-boot tasks.

iDRAC comes in tiers, and higher licensed tiers unlock remote management features beyond the base set. Option A is wrong because not all features are free, while options B and C wrongly tie licensing to the operating system and to a single BIOS task.

The front LCD panel shows system status and error messages directly at the server. The iDRAC web interface, the OpenManage Enterprise console and virtual media all require a remote workstation or session.

Secure Boot checks digital signatures on UEFI boot components and stops untrusted code from loading. Intrusion detection, self-encrypting drives and lockdown mode address physical access, data at rest and configuration changes, not signature checks on boot code.

A TPM 2.0 is a hardware security module that protects cryptographic keys and records boot measurements used to verify system integrity. RAID parity, file system journaling and management traffic routing are the jobs of other components.

Role-based access control assigns each administrator only the privileges needed, so juniors can monitor and use the console while firmware and BIOS changes stay with senior staff. Full rights, a shared account and removing all remote access each fail the requirement in different ways.

Multi-factor authentication requires more than one kind of proof, such as a password and a one-time code. Role-based access control limits privileges, intrusion detection monitors the chassis, and lockdown mode blocks configuration changes rather than checking the sign-in.

A self-encrypting drive encrypts data inside the drive, so the stored contents stay unreadable without the key. A hot spare, an over-provisioned SSD and a larger cache address availability, endurance and speed rather than data protection at rest.

Direct attach copper twinaxial cables are a cost-effective choice for short SFP+ links inside a rack. Long-reach single-mode fiber is built for distances far beyond a rack, RJ11 is a telephone cable, and a serial null-modem cable is for console access rather than data networking.

Add-in adapters extend connectivity with more ports, higher speeds or additional paths for redundancy beyond the integrated ports. Removing switches, storing RAID metadata and replacing firmware are not functions of a network adapter.

NIC teaming bonds two or more ports into one logical interface, so traffic continues on the remaining ports if one fails. VLAN tagging separates traffic, PXE boots an OS over the network, and DHCP relay forwards address requests, none of which provides link failover.

LOM, short for LAN on Motherboard, describes network ports integrated into the system board. A host bus adapter handles storage connectivity, PERC is a RAID controller, and a riser carries expansion cards, so none of them is the term for integrated network ports.

An access port carries a single VLAN untagged to the attached device, which matches a server NIC that expects plain frames. A trunk carries multiple tagged VLANs, while routed and mirror ports serve different purposes for routing and monitoring.