How long does a RAID 5 rebuild take?

RAID 5 rebuild time depends on drive capacity and rebuild throughput. A 4 TB HDD rebuilding at 100 MB/s takes about 11 hours. A 16 TB HDD at 80 MB/s under load takes over 55 hours. NVMe SSDs rebuild in minutes due to 2,000–5,000 MB/s sequential throughput.

What is RAID rebuild throughput?

Rebuild throughput is the effective speed at which data is read from surviving drives and written to the replacement. It depends on drive type (HDD vs SSD), RAID level overhead, and concurrent I/O load. Typical values: 7200 RPM HDD 60–100 MB/s, SATA SSD 200–400 MB/s, NVMe SSD 1,000–4,000 MB/s.

What is URE risk during RAID rebuild?

URE (Uncorrectable Read Error) risk is the probability of encountering an unrecoverable disk sector during the rebuild process. Consumer HDDs (10^14 URE rate) in large arrays have very high URE risk — over 90% for 6+ drive RAID 5 with 8 TB drives. Enterprise HDDs (10^15) and SSDs (10^17) are significantly safer.

RAID Rebuild Time Calculator

The RAID Rebuild Time Calculator estimates how long your array will spend in a degraded, data-at-risk state after a drive failure. Enter the Number of Drives, Drive Capacity, Drive Type, and RAID Level, set the I/O Load During Rebuild to reflect production workload competition, and click Calculate Rebuild Time to see estimated rebuild duration, data-at-risk window, and Uncorrectable Read Error (URE) probability with recommended actions.

RAID Level, Drive Count, Capacity and Rebuild Read Speed

Estimated Rebuild Duration, URE Risk and Double-Fault Probability

Rebuild Duration

—

Effective Rebuild Speed

—

Data at Risk

—

URE Probability

—

RAID rebuild detail: drive capacity, base and effective rebuild speed, data read, duration and URE probability
Metric	Value	Notes
Drive capacity	—	Per failed drive
Base rebuild speed	—	Unloaded sequential read
Effective rebuild speed	—	After I/O load deduction
Total data read during rebuild	—	From surviving drives
Rebuild duration	—	Estimated wall-clock time
URE probability	—	Chance of unrecoverable read error

HDD, SSD and NVMe RAID Rebuild Speeds and Estimated Duration

HDD, SSD and NVMe rebuild speed and estimated rebuild duration reference
Drive Type	Base Read Speed	Rebuild Speed (30% load)	8 TB Drive Rebuild	URE Rate
7200 RPM HDD	150 MB/s	~105 MB/s	~21 hrs	10^14–10^15
10K RPM HDD	200 MB/s	~140 MB/s	~16 hrs	10^15
15K RPM HDD	250 MB/s	~175 MB/s	~13 hrs	10^15
SATA SSD	550 MB/s	~385 MB/s	~5.8 hrs	10^17
NVMe SSD	3,500 MB/s	~2,450 MB/s	~54 min	10^17

URE Rates, RAID 5 vs RAID 6 and the Rebuild Window Risk Window

RAID rebuild time is the duration during which a degraded RAID array reconstructs data onto a replacement drive after a disk failure. During a RAID rebuild, the array operates in degraded mode — a state of elevated risk where a second drive failure would cause irreversible data loss in RAID 5, RAID 50, and single-parity configurations. The rebuild process must read data from all surviving member drives to reconstruct parity and write the regenerated data to the new replacement drive, making the rebuild I/O intensive and placing maximum stress on the remaining disks at exactly the moment they are most likely to fail. Learn more: calculate RAID array usable capacity and fault tolerance.

RAID 5 rebuild time for large-capacity drives has become a critical risk factor in modern storage arrays. A 16 TB nearline SAS or SATA HDD in a RAID 5 array may take 24–72 hours to rebuild even at sustained sequential read speeds, because the effective rebuild throughput is reduced by production I/O load and by the physical limitations of spinning media. During this entire rebuild window, the array has zero fault tolerance — if any remaining drive fails or returns an unrecoverable read error (URE), all data in the array is lost. This vulnerability is why RAID 6 and RAID 10 are now recommended for drives larger than 4 TB. You might also need: estimate SSD write endurance during RAID rebuild.

Uncorrectable Read Errors (UREs) present the most underappreciated risk during RAID rebuilds. A standard enterprise HDD has a URE rate of approximately 1 bit error per 10^15 bits read. For a RAID 5 array with eight 8 TB drives, the rebuild process reads 56 TB of data — about 4.48 × 10^14 bits — giving a URE probability of roughly 30% per rebuild event. With consumer-grade drives rated at 10^14 bits, the same rebuild carries a 96% chance of encountering a URE. NVMe SSD and SATA SSD arrays have URE rates of 10^17 or better, making SSD RAID 5 significantly safer than HDD RAID 5 during rebuilds, even at large capacities. See also: calculate ZFS RAIDZ usable capacity.

URE Probability, Drive Failure Rates and Rebuild Assumptions

Rebuild times assume sequential read throughput. Random I/O workloads reduce effective rebuild speed by 30–60%.
RAID 6 provides a second parity drive, allowing one additional drive failure during rebuild without data loss. Always use RAID 6 for HDD arrays with drives larger than 4 TB.
Hot spares reduce rebuild start latency from hours (manual replacement) to seconds (automatic rebuild trigger).
URE probability formula: P ≈ 1 − (1 − 1/URE_rate)^(total_bits_read). For total bits read: surviving_drives × capacity_bytes × 8 bits.
Enterprise storage controllers throttle rebuild I/O to 20–30% of drive bandwidth to limit production impact, significantly extending rebuild time.
RAID 10 rebuilds only need to read from a single mirror partner drive, making it far faster and safer than RAID 5/6 for the same drive capacity.