[Help] F4-423 TRAID md0 inactive after abnormal shutdown during reshape - bad blocks present

[thorid]

Device: TerraMaster F4-423

TOS Version: 5.1

Drive config: 4x WD Red SA500 4TB

Background:
I was expanding my TRAID storage pool from 2 drives to 4 drives by adding two new drives one at a time. The first drive addition completed successfully on 3/24. After the second drive sync completed on 3/25, I was not notified, and a reshape began automatically.
What happened:
Approximately 14 hours into the reshape (stuck at 56.9%), I noticed the TOS web UI was showing "No valid data" for the storage pool. Via SSH I confirmed the array was in a "clean, degraded, reshaping" state with sdd4 demoted to spare rebuilding. I received an abnormal shutdown notification approximately 4 hours ago which appears to have interrupted the process.

Current state:

md0 is now inactive and disassembled
All 4 drives are physically present and showing healthy
Superblock is intact and readable on sda4

mdadm --examine /dev/sda4 output:
Magic : a92b4efc
Version : 1.2
Array UUID : b04dbc30:a85ebf81:b8169337:ec1733d3
Raid Level : raid5
Raid Devices : 4
Array Size : 116897800044 (11148.24 GiB)
State : clean
Checksum : 7cebbea6 - correct
Reshape pos'n : 66482817748 (6340.30 GiB)
Delta Devices : 1 (3->4)
Bad Block Log : 512 entries available - bad blocks present
mdadm --detail /dev/md0 output:
Version :
Raid Level : raid0
Total Devices : 0
State : inactive

dmesg | grep md0 | tail -20 output:
md: md0: reshape done.
md: detected capacity change
md: resync of RAID array md0
md: resync done.
md: recovery of RAID array md0
md: md0: recovery interrupted.
md: reshape of RAID array md0

Questions:

• Is it safe to attempt reassembly of md0 given the bad blocks present on sda4?
  
  What is the recommended procedure to resume/complete the reshape after an abnormal shutdown?
  
  Should sda4 be tested for bad sectors before attempting reassembly?

All 4 drives show no failures and data is confirmed intact on the partitions. Looking for guidance on safe recovery steps before attempting anything further. During the reshape I still had access to my data, I did not interact with it, but it was still visible via the previously setup network drive.

[TMnight]

Based on the information you provided, the current array is a RAID 5, which was previously undergoing an expansion operation (from 3 disks to 4 disks).

During the expansion (reshape) process, the operation was interrupted before completion. Although the status shows as “clean,” this only indicates that the array metadata is consistent and does not mean the expansion has been completed.

Additionally, /dev/md0 is currently shown as inactive, indicating that the array has not been properly assembled or loaded. The check also reports bad blocks, which typically indicates that one of the disks in the array may have an issue. Since bad blocks have already been detected, please do not attempt to reassemble md0 directly. It is recommended to perform a complete bad block check on all disks first to identify and exclude any problematic disks before proceeding with further operations.

To help us further analyze and assist you with this issue, please export the system logs via the following path:

TOS Desktop → Support & Help → Problem Feedback

Please send the exported log files to: support(at)terra-master.com
.

Once we receive your logs, we will provide further analysis and guidance to help resolve the issue.

[thorid]

I emailed support previously but didn't get a reply. I have sent the requested logs and referenced this post.

[Phil]

After checking your system log, your drive sda has bad blocks.

It's recommended to power off your NAS, pull out sda--Serial Number: 24332B4A1C03 and remain the rest 3 disks in, then power on your NAS,
check if storage pool and volume is accessible. If yes, you might then try to insert sda to repair RAID, or format it in windows then insert it back.

If Volume is not accessible, please visit our official website online chat for further help.

[thorid]

I'm currently standing by on support chat as directed/scheduled, but no one is available.

[TMnight]

We sincerely apologize for not responding to your message in a timely manner.

Due to a recent upgrade and migration of our website’s online chat system, some messages may not have been received properly during the transition.

We kindly recommend clearing your browser cache, refreshing the page, and reloading www.terra-master.com
, then contacting us again via the online chat to ensure your message is delivered successfully.