Legacy Haven University

A broken boot drive stops work. A misbuilt RAID array destroys data. CompTIA A+ Core 1 (220-1201) Objective 3.4 forces you to compare storage hardware, pick the right interface, and choose a RAID level that matches the job. You also map those choices to Objective 5.2 symptoms: clicking platters, missing arrays, S.M.A.R.T. failures, and low IOPS. Students who treat storage as “the disk that holds Windows” fail scenario questions. Technicians who can explain the bus, the form factor, and the parity math pass them. Use this article as a study map while you follow a structured plan such as how to pass CompTIA A+ Core 1 (220-1201).

What Storage Actually Does Under the Hood

The operating system issues a read or write. The storage controller translates that request into physical operations. A spinning hard disk drive (HDD) moves an actuator arm across magnetic platters. A solid-state drive (SSD) flips charge in NAND flash cells. The controller reports completion through a communications interface: SATA, NVMe over PCIe, or SAS.

Speed is not one number. Throughput measures megabytes per second for large sequential files. IOPS (input/output operations per second) measures how many small random requests the drive finishes in a second. Boot times, browser cache, and database pages live in the IOPS world. Video files and disk images live in the throughput world. The exam expects you to match the media to the workload.

Hard Disk Drives: Magnetism, Spindles, and Form Factors

An HDD stores bits as magnetic polarity on spinning platters. A read/write head floats nanometers above the surface. The spindle motor sets rotation speed. Common A+ speeds:

• 5,400 RPM. Quiet laptop and archive drives. Lower IOPS. Lower heat.
• 7,200 RPM. Desktop and most NAS drives. The default exam answer for a standard workstation HDD.
• 10,000 RPM and 15,000 RPM. Older enterprise SAS disks. Rare in new builds. Higher vibration and power draw.

Form factor is the physical size of the enclosure:

• 3.5-inch. Desktop and server bays. Higher capacity per dollar. Needs a 3.5-inch cage or adapter in a small chassis.
• 2.5-inch. Laptops, thin desktops, and many NAS trays. Originally a laptop HDD size. Today you also find 2.5-inch SATA SSDs in the same bay.

HDD failure sounds mechanical. Grinding means bearings or heads scrape the platter. Rapid clicking means the actuator cannot find a track (the classic “click of death”). The OS may report “bootable device not found,” extended read times, or a S.M.A.R.T. failure. S.M.A.R.T. (Self-Monitoring, Analysis, and Reporting Technology) is firmware that tracks reallocated sectors, spin-up retries, and temperature. A S.M.A.R.T. warning is a replacement ticket, not a suggestion.

HDDs still win on bulk cost per terabyte. They lose on shock resistance, power, and random IOPS. Do not put the OS on a 5,400 RPM disk if the ticket asks for “best performance.”

Solid-State Drives: NAND, Controllers, and Why They Feel Instant

An SSD has no moving parts. NAND flash cells hold electrons. A controller chip spreads writes across cells (wear leveling), tracks bad blocks, and runs TRIM so the OS can mark unused pages for later erase. Random reads complete in microseconds. That is why Windows boots faster and why browsers feel snappy.

SSDs wear out from write cycles, not from spinning. Consumer TLC and QLC NAND offer high capacity at lower endurance. Enterprise drives advertise higher DWPD (drive writes per day). For A+, remember the practical rule: an SSD is the correct boot and application drive unless the question specifies cheap bulk storage.

SSD Communications Interfaces

The same NAND can crawl or fly depending on the bus.

SATA (Serial ATA). A 2.5-inch or 3.5-inch SSD that uses the SATA data cable and a 15-pin SATA power connector. SATA III tops out near 600 MB/s. The protocol was designed for spinning disks. Command queues are shallow. A SATA SSD still beats any HDD on IOPS, but it hits a ceiling that NVMe does not.

NVMe (Non-Volatile Memory Express). A protocol written for flash. NVMe talks directly to the CPU over PCIe lanes. It supports tens of thousands of command queues. Sequential reads commonly exceed 3,000–7,000 MB/s on current consumer drives. The exam phrase is “NVMe uses PCIe.” If a question asks for the fastest local storage for a workstation or laptop, answer NVMe.

PCIe (Peripheral Component Interconnect Express). The physical lane set. An add-in card SSD plugs into a PCIe slot. An M.2 NVMe module uses PCIe lanes that the motherboard routes to the M.2 socket. More lanes and a newer PCIe generation raise peak bandwidth. A+ does not require you to memorize every generation number. It does require you to know that NVMe rides PCIe and SATA does not.

SAS (Serial Attached SCSI). Dual-port enterprise protocol. Servers use SAS for multipath access and dense drive cages. SAS expanders fan one controller port out to many disks. You will see SAS on exam items about rack servers and hardware RAID cards. A SAS SSD or HDD does not drop into a desktop SATA port without an adapter or RAID card that speaks SAS.

SSD Form Factors

M.2. A stick that screws into a motherboard or laptop slot. Keying matters. An M.2 slot labeled “SATA/NVMe” may accept either protocol. An NVMe-only slot will not talk to an M.2 SATA module. Length codes such as 2280 mean 22 mm wide and 80 mm long. Most desktop boards use 2280. Confirm the standoff position before you force the screw.

mSATA. An older mini-SATA card used in some laptops and thin clients. Looks similar to a small card, not an M.2 stick. Do not confuse mSATA with M.2 on a parts list.

2.5-inch SATA SSD. Drop-in replacement for a laptop HDD. Uses the same SATA data and power cables. Easy upgrade path when the chassis has no M.2 slot.

Install order on a modern board: NVMe M.2 for the OS, optional SATA SSD or HDD for bulk data. Check BIOS/UEFI for the M.2 slot’s shared-lane warning. Some boards disable a SATA port when you populate a particular M.2 socket. That is a classic “missing drive in OS” cause after a hardware upgrade.

RAID: One Logical Volume, Several Physical Disks

RAID (Redundant Array of Independent Disks) combines two or more drives into one logical volume. The array can stripe for speed, mirror for safety, or store parity so the set survives a failed disk. Hardware RAID uses a dedicated controller card with cache and battery backup. Software RAID uses the OS (Windows Storage Spaces, Linux mdadm, motherboard “fake RAID” firmware).

RAID is not a backup. A ransomware payload, a bad RAID controller firmware flash, or a technician who deletes the wrong volume can wipe every member at once. Keep a separate backup. Say that sentence on the exam and on the ticket.

CompTIA 220-1201 lists RAID 0, 1, 5, 6, and 10. Memorize minimum drives, usable capacity, fault tolerance, and the workload that fits.

RAID 0 — Striping, No Safety

The controller splits each write across all member disks. Two 1 TB disks present 2 TB. Reads and writes scale with the number of disks. There is zero redundancy. One dead member destroys the entire volume.

Use RAID 0 for scratch video, scratch CAD caches, and temporary render disks where you can regenerate the data. Never put customer records or the only copy of a project on RAID 0.

RAID 1 — Mirroring

Every write hits two disks. Two 1 TB disks present 1 TB. One disk can fail and the array stays online. Rebuild is a straight copy onto a replacement disk.

Use RAID 1 for OS volumes on a small server, financial workstations, and any two-bay NAS that must survive a single disk failure. You pay 50 percent of raw capacity for that safety.

RAID 5 — Striping Plus Distributed Parity

Data and parity stripe across all disks. Three 1 TB disks present 2 TB. The array survives one disk failure. During rebuild the controller reads every remaining disk and recalculates the missing stripe.

RAID 5 needs a minimum of three disks. Write penalty exists because each write updates data and parity. Modern SSDs reduce that pain. Large HDD RAID 5 sets rebuild for hours. A second disk failure during rebuild kills the array. That is why many shops moved critical HDD arrays to RAID 6.

RAID 6 — Dual Parity

RAID 6 stores two independent parity blocks. Four 1 TB disks present 2 TB. The array survives two disk failures. Rebuild still stresses the remaining disks, but you keep a second safety net.

Use RAID 6 for large HDD capacity pools: backup targets, media archives, and departmental file servers. The exam wants the phrase “survives two drive failures.”

RAID 10 — Stripe of Mirrors (RAID 1+0)

The controller mirrors pairs, then stripes across those pairs. Four 1 TB disks present 2 TB. You can lose one disk in each mirrored pair. You cannot lose both disks in the same pair.

RAID 10 rebuilds fast because a failed disk copies from its mirror, not from parity math across the whole set. Random writes stay strong. Cost is high: you keep only half of raw capacity. Use RAID 10 for databases, virtual machine datastores, and busy application servers.

RAID Decision Table

When a performance-based question shows four disks and “need speed and redundancy,” pick RAID 10. When it shows three disks and “need capacity with one-disk fault tolerance,” pick RAID 5. When it shows two disks and “must keep working if one dies,” pick RAID 1.

Hardware RAID vs Software RAID

A hardware RAID card presents one virtual disk to the OS. The OS never sees the individual members unless you open the RAID utility. Cache and a BBU (battery backup unit) protect writes if power drops mid-stripe. Replace a failed disk from the controller BIOS or vendor tool. Watch the LED status indicators on the hot-swap bay. Amber or flashing patterns usually mean rebuild or failed member.

Software RAID lives in the OS. Windows Storage Spaces and Linux mdadm cost nothing extra. Performance depends on CPU. A motherboard RAID option ROM is often “fake RAID”: firmware assists at boot, then a driver takes over. Fake RAID is painful to migrate. Prefer true hardware RAID for servers and OS-native software RAID for simple desktop mirrors.

Hot-swap trays let you pull a failed disk without powering down. Confirm the chassis and controller support hot-swap before you yank a live SAS or SATA member. Pulling the wrong disk during a degraded RAID 5 rebuild is a career-limiting move. Label bays. Match serial numbers in the RAID utility.

Troubleshooting Drives and RAID (Objective 5.2)

Work the symptom, then the layer.

LED status indicators. Solid activity on one member and nothing on another points to a dropped disk. Many backplanes light amber for a predicted failure.

Grinding or clicking. Mechanical HDD. Power down. Do not run chkdsk for sport on a clicking disk. Image first if the data matters, then replace.

Bootable device not found. Wrong boot order after you added an NVMe drive. Dead SATA cable. RAID volume that dropped to “failed” and no longer looks like a boot disk. Enter UEFI and confirm the NVMe or RAID volume is first.

Data loss or corruption. Failing NAND, a RAID write hole after a power loss on a controller with no battery, or a user who initialized the wrong disk in Disk Management. Stop writes. Assess backups.

RAID failure / array missing. Controller lost the configuration. A member dropped and the array went offline because the level cannot survive that loss (RAID 0, or two dead disks in RAID 5). Import foreign configurations only when you know which set is current.

S.M.A.R.T. failure. Replace the disk. For RAID, mark the member failed and rebuild onto a known-good spare.

Extended read/write times and low IOPS. HDD in a random-I/O role. SATA SSD on a saturated controller. Thermal throttle on a bare M.2 drive with no heatsink. A degraded RAID 5 that is already rebuilding.

Missing drives in the OS. Disabled SATA port because an M.2 slot stole the lanes. Loose cable. RAID controller that owns the disks so Disk Management shows one volume, not four physical disks. That last case is normal, not a failure.

Document the array level, member serials, and firmware before you change anything. After repair, verify the volume mounts, run a scrub or consistency check if the vendor tool offers one, and confirm the backup job still targets the volume.

Exam and Bench Habits That Stick

Read the constraint first: budget, bay count, “must survive one failure,” “fastest boot.” Those four phrases pick the media and the RAID level for you.

Match interface to slot. An M.2 NVMe module in a SATA-only M.2 key will not run at NVMe speed. A SAS disk in a desktop SATA cable will not enumerate.

Never confuse RAID with backup. The exam writers love that trap.

When you study, build the comparison in your own words, then test yourself on minimum disk counts. Pair this hardware map with a full Core 1 study path at how to pass CompTIA A+ Core 1 (220-1201) so storage sits next to motherboards, RAM, and troubleshooting method instead of floating as isolated trivia.