Skip to main content

​
Overview

Unraid’s parity-protected array allows for single-drive replacement without data loss. This guide covers:
  • Identifying failed drives
  • Pre-replacement preparation
  • Physical drive replacement
  • Array rebuild process
  • Post-rebuild verification
Do not replace multiple drives simultaneously. Unraid can recover from a single drive failure with dual parity, but replacing multiple data drives at once requires careful planning. Always replace and rebuild one drive at a time.

​
Identifying Drive Failures

​
Dashboard Indicators

Unraid Main Dashboard will show:
  • Red “X” icon on failed drive
  • Array status: “Degraded” (yellow warning)
  • Notification in system log

​
SMART Data Review

1

Access SMART Info

Unraid UI → Main → [Select Drive] → SMART Info
2

Check Critical Attributes

Red flags (immediate replacement required):
  • Reallocated Sectors Count: > 0
  • Current Pending Sector Count: > 0
  • Offline Uncorrectable Sectors: > 0
  • SMART overall-health self-assessment: FAILED
Yellow flags (monitor closely, plan replacement):
  • Drive temperature: > 55°C sustained
  • Power-On Hours: > 40,000 hours (5+ years)
  • Spin Retry Count: > 0
3

Document Drive Info

Record before replacement:
  • Drive serial number (from SMART or physical label)
  • Drive position (slot number in array)
  • Drive role (data disk, parity1, parity2)
  • Filesystem type (XFS, BTRFS)

​
Pre-Replacement Checklist

1

Verify Parity is Valid

Critical: Do not replace a drive if parity is invalid.Unraid UI → Main → Array OperationParity status should show:
  • Parity is valid (green checkmark)
  • No sync/check in progress
If parity is invalid:
  1. Run parity check first: Main → Parity Check → Start
  2. Wait for completion (may take 8-24 hours)
  3. Review results — if errors found, investigate before proceeding
2

Stop All Docker Containers

Minimize disk activity during replacement.Docker → [Select All] → Stop
3

Stop Array (If Drive is Still Detected)

Unraid UI → Main → Array Operation → Stop
Only stop the array if the drive is still detected by Unraid. If the drive has completely failed and disappeared, proceed directly to physical replacement.
4

Prepare Replacement Drive

Requirements:
  • Replacement drive must be same size or larger than failed drive
  • CMR drive (not SMR) for array drives
  • NAS-rated (WD Red, IronWolf, etc.)
Do NOT pre-format — Unraid handles formatting during rebuild.

​
Physical Drive Replacement

Shutdown the server before opening the chassis unless hot-swap bays are confirmed working. The 4U chassis used in this build may support hot-swap, but safe shutdown is recommended for first-time replacements.
1

Power Down Server

Unraid UI → Main → ShutdownWait for complete shutdown (fans stop, activity lights off).
2

Identify Failed Drive Physically

Match drive serial number from SMART data to physical label.Location reference:
  • 4U 12-bay chassis: drives are numbered left-to-right, top-to-bottom
  • Unraid assigns sdX device IDs — cross-reference in Tools → System Devices
3

Remove Failed Drive

  • Pull drive tray release latch
  • Slide tray out gently
  • Unscrew drive from tray (4 screws)
  • Keep failed drive for RMA if under warranty
4

Install Replacement Drive

  • Mount new drive in tray (4 screws, same orientation)
  • Slide tray back into same slot
  • Ensure tray locks securely
5

Power On Server

Press power button and wait for POST.

​
Array Rebuild Process

1

Boot Unraid and Review Main Page

Unraid UI → MainNew drive should appear as “Unassigned Device” or “New Device”.
2

Assign Replacement Drive

Main → [Failed Drive Slot] → Device → [Select New Drive]Example:
  • If Disk 4 failed, select the new drive from dropdown under Disk 4
Verify you selected the correct slot. Assigning the wrong drive will wipe it.
3

Apply Configuration

Click Apply at the bottom of the Main page.Unraid will prompt:
“Are you sure you want to start the array? This will rebuild the disabled disk.”
Confirm and click Start.
4

Monitor Rebuild Progress

Main → Array OperationShows:
  • Rebuild progress percentage
  • Estimated time remaining
  • Current speed (MB/s)
Expected duration:
  • 4TB drive: 6-10 hours
  • 12TB drive: 18-30 hours
Rebuild runs in background — server remains usable but slower.
5

Do Not Interrupt Rebuild

Do not stop the array or power off during rebuild. Interruption forces restart from beginning.
Safe to:
  • Access shares (read/write)
  • Start Docker containers (after rebuild > 10%)
  • Use Plex (performance may be degraded)
Avoid:
  • Stopping array
  • Rebooting server
  • Replacing additional drives

​
Post-Rebuild Verification

1

Verify Rebuild Completed

Main → Array OperationStatus should show:
  • “Reconstruct finished” (green)
  • Array status: Normal (green checkmark)
  • No disabled drives
2

Run Parity Check

Main → Parity Check → Correcting → StartThis verifies parity is consistent after rebuild.Expected result: 0 errorsIf errors are found:
  • Review Unraid logs: Tools → System Log
  • Check SMART data on all drives
  • Consider running a second parity check
3

Check SMART Data on New Drive

Main → [New Drive] → SMART InfoVerify:
  • SMART overall-health: PASSED
  • Reallocated Sectors: 0
  • Temperature: < 50°C
4

Update Drive Inventory

Document replacement in homelab-docs repo:
  • Date replaced
  • Old drive serial number
  • New drive serial number and model
  • RMA status (if applicable)
5

Resume Normal Operations

  • Restart Docker containers: Docker → [Select All] → Start
  • Verify media shares accessible
  • Test Plex playback
  • Confirm backup jobs resume (check Healthchecks.io)

​
Special Cases

​
Replacing a Parity Drive

Simpler than data drive replacement — no rebuild required if parity drive fails.
1

Stop Array and Replace Drive

Same physical replacement process as data drive.
2

Assign New Parity Drive

Main → Parity → [Select New Drive]
3

Start Array - Parity Will Sync

Unraid will build new parity from scratch using all data drives.This takes just as long as a rebuild (18-30 hours for 12TB array).
4

Verify Parity Sync Completes

Main → Array Operation → “Parity sync finished”

​
Dual Drive Failure (Critical)

Data loss scenario: If two data drives fail simultaneously and you only have dual parity, one drive’s data is unrecoverable.
If dual parity is intact:
  1. Replace drives one at a time
  2. Rebuild first drive completely
  3. Then replace and rebuild second drive
If parity is also compromised:
  • Stop immediately — do not start array
  • Consult Unraid forums or professional data recovery
  • Consider restoring from Synology ABB (Tier 3 backup)

​
Upgrading Drive Size

Replace smaller drive with larger drive to expand array capacity.
1

Replace Drive Normally

Follow standard replacement procedure.
2

Rebuild Completes at Original Size

Unraid will only use the same capacity as the old drive initially.
3

Expand Array (Optional)

After rebuild completes:Tools → New Config → Retain All Drives → ApplyStart array — Unraid will recognize the larger capacity and expand.

​
Prevention & Monitoring

​
Schedule Regular Parity Checks

Recommended: Monthly parity check (non-correcting) Settings → Scheduler → Parity Check
  • Day of month: 1 (first of month)
  • Time: 02:00 (overnight)
  • Type: Non-correcting (read-only check)

​
Monitor SMART Weekly

Add to weekly operations checklist:
  • Review SMART data for all drives
  • Check for temperature spikes
  • Note Power-On Hours for aging drives

​
Replace Drives Proactively

Don’t wait for failure. Replace if:
  • Reallocated sectors appear (even if count is low)
  • Drive is 5+ years old (> 40,000 hours)
  • Drive has been through 1+ rebuilds already (stress accumulates)

​
Maintain Hot Spares

Current spares:
  • 1x WD Red Plus 4TB (for ZFS mirror pool)
  • 2x Seagate IronWolf 6TB (for array expansion or emergency replacement)
Before using a spare: Run SMART extended test to verify health.

​
Troubleshooting

​
Drive Not Detected After Replacement

Check:
  • SATA cable seated firmly (both ends)
  • Power cable connected
  • Drive shows in BIOS/UEFI
  • Different SATA port if available

​
Rebuild Extremely Slow (< 50 MB/s)

Causes:
  • Docker containers running heavy I/O
  • Plex transcoding active
  • Another parity check running
Fix:
  • Stop Docker containers temporarily
  • Pause Plex activity
  • Check Settings → Disk Settings → Tunable (md_write_method) — should be auto

​
Parity Errors After Rebuild

Not necessarily a problem — small number of errors (< 10) can occur during rebuild if array was accessed. Action:
  1. Run second correcting parity check
  2. If errors persist, review SMART on all drives
  3. Consider testing drives individually with badblocks

​
New Drive Fails SMART Immediately

DOA (Dead on Arrival) — return for RMA. Before RMA:
  • Test in different SATA port
  • Test with different cable
  • Document SMART output and serial number for RMA claim

Build docs developers (and LLMs) love

Get started for freeTalk to us