Applies to: Pro, Enterprise, and Audit
Purpose
Read Performance Benchmark measures representative sequential-read throughput and random-read response after sanitization. The default benchmark is read-only.
DriveErase Operation
- A sequential sample of up to 1 GiB using 1 MiB blocks;
- 200 random-read samples using 4 KiB blocks;
- Calculation of sequential read throughput;
- Calculation of average random-read latency;
- An estimated random-read IOPS result;
- A SMART snapshot where available;
- Health score and grade calculation.
Interpretation
Use the benchmark to identify a device performing materially below an approved baseline or showing abnormal response times. Compare results only with devices of similar technology, connection, controller, and capacity.
Results can be influenced by:
- Controller and driver behavior;
- USB or SAS bridge overhead;
- Cache state;
- Device temperature;
- Power state;
- Thermal throttling;
- Virtual storage contention;
- Background firmware maintenance;
- The minimal privileged environment.
A high performance result does not prove complete media health. A lower result does not necessarily mean that sanitization failed.
Asset grading
Where performance is used in reuse grading, define documented thresholds by device class. Retain the benchmark result with the asset record and include other indicators such as SMART, bad blocks, age, endurance, error history, and warranty status.