[Help] TOS7 Bug – F4-423 with D5-300C TRAID Pool Not Imported After Reboot

[digi999]

Hi All

its a long shot but I wondered if there was a TOS / Linux expert in the house. The TLDR is my F4-423 with an attached D5-300C (split raid pools obviously) don't work after installing the latest TOS7 (as well as updates) its like the raid pool on the D5-300C is gone?

Ive attached the ChatGPT bug report 9I spent 3 hours with it trying to get it to work!) , does this mean anything to anyone ? lol

Hardware

Model: F4-423
TOS Version: TOS 7
System Disk: ADATA SX8200 NVMe SSD
External Enclosure: D5-300C
Connection: Official Terramaster USB-C cable
External Pool: TRAID RAID5
Drives: 5 × 14TB WD drives
Internal Pool: 4 × 8TB Seagate drives
Upgrade path: TOS6 → TOS7

Issue 1 (resolved)

After upgrading to TOS7, the NAS booted into the USB bootloader instead of the NVMe system disk.

This was resolved by changing the BIOS boot order to prefer:

NVME: UEFI OS (ADATA SX8200PNP)

over:

USB Device: UEFI General UDisk 5.00

This matches your published support article regarding TOS7 upgrades on 423-series devices with NVMe system disks.
Issue 2 (current problem)

After fixing the boot order issue, TOS7 loads successfully.

However, the external D5-300C pool always appears as:

RAID Corrupted
Missing Required Disks

Storage Manager reports the pool as failed after every reboot.

The pool itself is healthy.
Evidence

All five disks are detected:

sde
sdf
sdg
sdh
sdi

RAID metadata is intact:

mdadm --examine /dev/sd[e-i]4

shows:

Raid Level : raid5
Raid Devices : 5
State : clean
Array State : AAAAA
Events : identical on all disks

Manual assembly works perfectly:

mdadm --assemble --scan

Immediately restores:

md2 : active raid5
[UUUUU]

LVM detects the volume:

VG vg1
LV lv2
50.89TB

Btrfs filesystem is intact:

Label: TOS_VOL_20250516
FS bytes used 36.23TiB

I can manually mount the filesystem and access all data.

Examples:

Plex-TV
Plex-Movies
Plex-Docs
4K-Plex
Backups

The pool contains approximately 37TB of data and all data is accessible.
Logs

After reboot:

md2 is absent

However all five disks are already present.

Running:

mdadm --assemble --scan

immediately recreates:

md2 [UUUUU]

Storage Manager then returns to:

Normal

journalctl output

The logs indicate that Storage Manager itself assembles md2 and then immediately stops it.

Example:

storagemanager:
mdadm -A /dev/md2 ...
mdadm: /dev/md2 has been started with 5 drives.
storagemanager:
mdadm --stop /dev/md2
mdadm: stopped /dev/md2

Later messages show:

the available Pool(...) will be deleted because the disk(USB Disk_sdf...) is occupied

This behaviour repeats after every reboot.

The RAID is healthy, but TOS7 appears to stop md2 automatically.
Summary

The issue appears to be:

TOS7 correctly detects the D5-300C enclosure
md2 can be assembled successfully
vg1 and lv2 are healthy
Btrfs filesystem is healthy
Storage Manager subsequently stops md2
Pool 2 is then shown as corrupted

The workaround is currently:

mdadm --assemble --scan
vgchange -ay

which immediately restores the pool.

Could you please advise whether this is a known issue with:

F4-423
TOS7
NVMe system disk
D5-300C USB enclosure
TRAID pools

or whether there is an updated package or patch available?

[EriChan]

To better assist you in confirming whether there is an abnormality in the TOS system, please send the system report via email to our technical support mailbox [email protected], and also include the forum link and your forum ID.