I have completed extensive diagnostics on my FLEX 8 (named: TB100) RAID 5 volume and would appreciate your guidance before performing any additional write operations.
SYSTEM CONFIGURATION
• Mac Studio (Apple Silicon)
• Current macOS
• Latest SoftRAID
• OWC ThunderBay Flex 8
• Volume: TB100
• RAID 5
• HFS+
• Six active RAID member disks plus one separate standalone SoftRAID disk
• Complete verified backup of TB100 is available
WHAT HAPPENED
While investigating another issue, I removed and reinserted one drive in the Flex 8 enclosure. After reinserting the drive, the RAID became degraded.
Initially there was confusion because macOS reassigned BSD device numbers after the drives were rediscovered (for example, what had previously been disk17 later became disk18). After tracing the drives by both serial number and SoftRAID ID, I confirmed that all RAID member disks are present.
CURRENT STATUS
SoftRAIDTool reports:
• Volume State: Out of Sync
• RAID Level: RAID 5
• All six RAID member disks are present.
• disk14 (Serial Number ZVT0XGWX) is the only member marked "Out of Sync."
• Total Volume I/O Errors: 0.
• The volume remains unmounted.
Current RAID members:
disk15
disk14 (Out of Sync)
disk16
disk8
disk17
disk19
DRIVE HEALTH
Every RAID member reports:
• SMART Status: Passed
• Reallocated Sectors: 0
• Pending (Probation) Sectors: 0
• Uncorrectable Sectors: 0
The out-of-sync drive (disk14 / Serial ZVT0XGWX) also passes SMART with no indication of media failure.
SoftRAID lifetime I/O error counters increased only after the drive removal/reinsertion event. These appear to be communication events rather than actual disk failures.
PROBLEM
SoftRAID reports that the volume needs rebuilding.
When I select Rebuild, SoftRAID displays:
"Your SoftRAID volume cannot be rebuilt unless it is mounted on the desktop. Do you want to mount this volume and rebuild it?"
When I click Rebuild:
• SoftRAID attempts to mount the volume.
• "Mounting..." is displayed.
• After a period of time the volume silently returns to Unmounted.
• The rebuild never begins.
• No additional error message is displayed.
ADDITIONAL DIAGNOSTICS
Running:
diskutil mount disk21
returns:
Volume on disk21 failed to mount.
If it has a partitioning scheme, use "diskutil mountDisk".
If you think the volume is supported but damaged, try the "readOnly" option.
Attempting a read-only mount produces the same result.
Running:
diskutil verifyVolume disk21
returns:
Error starting file system verification for disk21:
Unrecognized file system (-69846)
Because the RAID is currently Out of Sync, I have not attempted DiskWarrior or fsck_hfs.
QUESTIONS
Since:
• All RAID members are present.
• SMART is healthy on every drive.
• The RAID metadata appears intact.
• Only one member is marked Out of Sync.
• SoftRAID recognizes the RAID but cannot mount it in order to begin rebuilding.
Could you please advise:
1. Is there a supported procedure to force a RAID rebuild or parity resynchronization without requiring the filesystem to mount first?
2. Is there a supported method to clear the Out of Sync state?
3. Is there a RAID metadata repair procedure for this condition?
4. Is there a recommended diagnostic or recovery step before deleting and recreating the RAID from my backup?
I have attached the latest SoftRAID diagnostics, screenshots, and command output.
Note: Attaching a SoftRAID support file, I do not need all the description.
Just run Disk Warrior. Use the Preview to ensure you are happy before clicking the Replace button. A volume like this with your description is likely an easy repair for Disk Warrior.
All six RAID member disks present in OWC ThunderBay Flex 8
BSD device: disk21, recognized by macOS as Apple_HFS at 100TB
Key Findings from Diagnostics:
Five of six members (disk8, disk14, disk15, disk16, disk17) each show exactly 4 I/O errors and are flagged for replacement. disk19 (SN: ZVTBES2X) shows 0 I/O errors.
The identical error count of 4 across five drives suggests a single bus-level communication event from a drive removal/reinsertion rather than independent drive failures.
Total Volume I/O Errors: 0. All six members pass SMART with zero reallocated, pending, or uncorrectable sectors.
Only disk14 (SN: ZVT0XGWX) is marked Out of Sync.
A seventh drive (disk18, SN: ZVT21YGN) is present in the enclosure as a standalone SoftRAID disk, not a RAID member.
SoftRAIDTool version 8.6.1, driver version 8.6.
The Problem:
SoftRAID GUI prompts to mount and rebuild but the volume silently fails to mount and the rebuild never starts. diskutil returns error -69846 (Unrecognized file system) on disk21.
Questions for Support:
Is the pattern of exactly 4 I/O errors on five members consistent with a bus-level communication event, and can those error counters be cleared?
Is there a supported path to initiate a parity resync or rebuild without requiring the volume to mount first?
Can you examine the RAID metadata and advise whether a metadata repair is possible before we consider deleting and recreating the RAID from backup?
Full diagnostics including the .sr_supt file, SoftRAIDTool output, and diskutil data are attached as a zip archive.
I have 3 txt files that the form wouldnt allow me to attach. Seems strange to reject text files!!
@slevitus Because those text files are really .rtf. You need to "make plain text" first. But they are not going to be useful, the support file has everything I need 99% of the time. Note: a 1 pass certify is just a "zero disk" and it is not really rigorous. At least do 2 passes, but I would not restart the certify for that.
You need to repair the directory on TB100 first, with Disk Warrior, then clear the IO counters, then rebuild. the rebuild should start as soon as DW is done. And I note that you have several drives with >30,000 PoH. Plan on starting to replace those and use them for tertiary backups, so they do not contain critical irreplacable data.
