SMART Reallocated, Pending, Uncorrectable and CRC Counters
Common attribute names
The following meanings are common on ATA HDDs. Attribute IDs, raw encoding, normalization and thresholds are vendor/model dependent; they are not universal definitions for every HDD, SSD or NVMe device.
| Common ID | Name | Interpretation |
|---|---|---|
| 05 / 5 | Reallocated Sector Count | Commonly counts sectors remapped to spare storage. Confirm the raw-value encoding for the drive. |
| C5 / 197 | Current Pending Sector Count | Commonly indicates unstable/unreadable sectors awaiting reassessment or a subsequent operation. It is not proof that a temporary spare has already replaced each sector. |
| C6 / 198 | Offline Uncorrectable | Commonly records uncorrectable sectors found during offline scanning. It does not by itself identify a platter, head or mechanical root cause. |
| C7 / 199 | UltraDMA CRC Error Count | Commonly records interface-transfer CRC errors. Investigate the cable/connectors and relevant link/backplane/controller path; a yellow indicator alone does not prove that replacing a cable fixes the fault. |
How the source explains pending sectors
The source associates difficult reads with high workload, vibration or unstable voltage, and describes retries, temporary replacement and later idle-time retesting. It mentions “fewer than three retries” under load, then a pending marker after more retries. These are the source’s explanatory assumptions; no model/firmware specification was supplied to establish a universal retry threshold or temporary-spare algorithm.
Depending on the drive and later operations, a suspect sector may become readable again or be remapped, and the pending/reallocated counters may change. Do not assume that all pending sectors are automatically repaired by leaving the disk idle. Offline data collection and self-tests have drive-specific scheduling/capabilities, and a change in one counter must be interpreted with the full report.
Monitor and decide
- Identify attributes by their ID/name, not “row 5” in an arbitrary tool’s display. Preserve the original raw values, normalized values, thresholds, error logs and self-test results for comparison.
- Track whether reallocated, pending or uncorrectable counts are new, stable or rising, and correlate them with host I/O errors and vendor diagnostics. A zero reallocated or offline-uncorrectable count does not prove that a drive is healthy or free of unreadable sectors.
- The source suggests concern when reallocated sectors increase by more than 5–10 per month. This is an unverified source heuristic, not an industry replacement threshold or vendor warranty rule. Follow the actual model’s diagnostics and replacement criteria, while preserving needed data.
Seagate’s English SMART explanation notes that its attributes/thresholds are proprietary and recommends its own diagnostics for its drives. The smartctl manual source explains the tool’s reporting options and vendor-dependent attribute interpretation.
Original case/reference: stored source.