Steps for Certify Disk Errors

A properly functioning disk should always pass certification. Any certification failure indicates a confirmed hardware issue — the only question is whether the cause is the disk itself, a communication problem, or (less commonly) a faulty enclosure.

Before You Begin: Check the SoftRAID Log

Launch SoftRAID and open the log: Utilities menu → SoftRAID Log

Scroll to the bottom to find entries from the failed certification. To filter the log, type certify in the search field in the upper right — this will show only lines containing that word.

The log entries will help you identify which of the three causes below applies to your situation.

Cause 1: Communication Problem

Communication errors can be caused by a loose cable, power fluctuation, the Mac going to sleep during certification, or bus interference from other connected devices.

How to identify:

• Check the SoftRAID log for the exact time of failure
• If multiple disks show errors with similar or identical timestamps, this strongly suggests a communication issue rather than individual disk failures

What to do:

• Set your Mac to never sleep during certification (System Settings → Energy)
• Ensure all cables are securely connected
• Avoid connecting other devices to the same Thunderbolt bus during certification
• Consider using a voltage regulator (e.g., APC R-1200) for stable power
• Resume the certification — SoftRAID will prompt you to continue from where it stopped

Cause 2: Faulty Disk

How to identify: Check the disk tile in SoftRAID for:

• SMART failure (shown in red) — the disk has already failed
• Predicted failure (shown in red) — the disk is failing, possibly DOA on a new drive
• I/O errors in the SoftRAID log

I/O errors come in two types — the sector offset tells you which:

Type A — Specific sector offset (non-zero): The error points to a specific location on the disk. This is a disk hardware failure. → Replace the disk.

Example log entry:

Dec 12 14:54:35 – SoftRAID Driver: A disk for the volume "ThunderBlade" (disk10)
encountered a read error (E00002CD). The disk (disk6, SoftRAID ID: 09B2D04656CF4B00)
was unable to read sectors. The error occurred at volume offset 4706153693184
(i/o block size 3145728). This disk should be replaced.

Type B — Sector offset 0: The error points to sector 0. This is more likely a communication error than a disk failure. → Check connections first (see Cause 1). Resume certification after verifying cables and bus stability.

Example log entry:

Jun 04 03:17:52 – SoftRAID Driver: A disk for the volume "ThunderBlade" (disk10)
encountered a read error (E00002CD). The disk (disk6, SoftRAID ID: 09B2D04656CF4B00)
was unable to read sectors. The error occurred at volume offset 0
(i/o block size 3145728). This disk should be replaced.

What to do:

• SMART failure → replace the disk immediately
• Predicted failure on a new disk → the disk is likely DOA; replace it
• Unreliable sectors → restart the certification; it may have been caused by a communication issue. However, if any sectors are subsequently marked as reallocated, replace the disk
• Type A I/O error → replace the disk
• Type B I/O error → investigate communication issues first; resume certification after checking connections

Cause 3: Faulty Enclosure

Enclosure failures are less common, but should be suspected when the pattern of failures doesn’t point to a single disk.

How to identify:

• Certification failures occur randomly on different disks across multiple certification attempts
• Two disks in adjacent slots fail consecutively (e.g., slots A+B or C+D) — this can indicate a shared power or signal issue in that section of the enclosure
• Disks are being ejected during certification

What to do:

• Replace the Thunderbolt cable first to rule out a cable issue
• Try different Thunderbolt ports on your Mac
• Temporarily disconnect other devices and monitors to isolate the Thunderbolt connection
• If failures persist across different disks and cables, contact OWC support — the enclosure may need service or replacement.

Summary