01-04-2026, 01:34 PM
Thanks for raising this. At this point there are several independent reports from Reddit and Level1Techs showing the same pattern on the N5 Pro. During heavy mixed I O or sustained write loads, the onboard JMB585 SATA controller appears to reset, which briefly drops all SATA disks. When that happens, any data being written can be corrupted before ZFS generates checksums, so ZFS only detects the problem after corruption is already committed.
This does not seem to be a single faulty unit, so until Minisforum makes a public statement or releases a firmware fix, anyone planning to use the N5 Pro as a main NAS should be cautious, especially if using ZFS for important data.
For users who already own the N5 Pro and want to test their unit safely:
1. Create a temporary pool with empty test disks only.
2. Run a long sequential write test, for example:
dd if=/dev/zero of=/pool/testfile bs=1M status=progress
3. Run a mixed workload test, for example:
fio --name=load --rw=randrw --bs=128k --size=10G --iodepth=32 --numjobs=1
4. Watch dmesg -w or journalctl -f for resets such as “link reset”, “SATA channel disabled”, or drives dropping then reappearing.
If the controller is unstable, you will see I O errors or SATA resets almost immediately under load.
I will update this thread as soon as Minisforum publishes a response or patch. In the meantime, users should keep verified backups and avoid placing critical data on a SATA pool until this is clarified.
This does not seem to be a single faulty unit, so until Minisforum makes a public statement or releases a firmware fix, anyone planning to use the N5 Pro as a main NAS should be cautious, especially if using ZFS for important data.
For users who already own the N5 Pro and want to test their unit safely:
1. Create a temporary pool with empty test disks only.
2. Run a long sequential write test, for example:
dd if=/dev/zero of=/pool/testfile bs=1M status=progress
3. Run a mixed workload test, for example:
fio --name=load --rw=randrw --bs=128k --size=10G --iodepth=32 --numjobs=1
4. Watch dmesg -w or journalctl -f for resets such as “link reset”, “SATA channel disabled”, or drives dropping then reappearing.
If the controller is unstable, you will see I O errors or SATA resets almost immediately under load.
I will update this thread as soon as Minisforum publishes a response or patch. In the meantime, users should keep verified backups and avoid placing critical data on a SATA pool until this is clarified.