JorgeB

Couple of more observations, regarding hardware, and obviously besides a fast device/pool, the only other thing you want is an high clock CPU with a fast single thread rating, Samba is single threaded, so e.g. a fast dual core CPU is better then a low clock 8 core Xeon. Also, and since 99% of the time my transfers to this server are in the other direction, from Unraid, not to Unraid, it's been some time since I've paid attention to write speed, fuse overhead appears to be more pronounced with v6.8.x compared to v6.7.x, did a quick test and currently I'm getting more like 700MB/s write speed to a user share, I remember getting around 800/900MB/s with v6.7.x. User share vs disk share

Try a different USB HUB? Powered on would be better.

Check filesystem on disk11: https://wiki.unraid.net/Check_Disk_Filesystems#Checking_and_fixing_drives_in_the_webGui Rebuilding a disk won't fix filesystem corruption.

Supermicro X11SSM-F, G4560, 6 SSD RAID5 cache pool

Delete network.cff and network-rules.cfg also on flash/config and reboot, if still nothing please post new diags.

Diags are after rebooting so we can't see what happened but disks look fine, so likely some connection/power issue. Yes, format is never part of a rebuild/replace. Check filesystem on the emulated disk, if it's fixable and it mounts correctly and data looks OK you can then rebuild on top, you can also re-sync parity at the same time, but suggest replacing/swapping cables (both SATA and power) on both disks first to rule that out.

If disabling power states didn't help it's likely a hardware/bios problem, either the board or the NVMe device.

Default is still usually not the correct setting needed for stability with Linux.

Yes, that is overclocking.

Pool wasn't created correctly, only one device is being used and there are multiple profiles, make sure you wipe both cache devices before formatting, if you get an error with blkdiscard on any of them wiping won't work, e.g., if the controller doesn't support trim.

After all data is backed up, stop array and type: blkdiscard /dev/sdX Replace X with both devices ID, one at a time, after that start the array and you'll have the option to format the pool.

There's corruption on the cache filesystem, with btrfs it's best to backup data, re-format pool and restore data, also a good idea to re-create docker image at the same time.

Please post the complete diagnostics: Tools -> Diagnostics

Don't know, but not the first time I see similar errors with AMD based systems, BIOS update might help.

It is, you might need to input the path manually.

Those look like an issue I've seen a couple of times before where drives on an LSI generate those errors after Unraid spins them up, if yes they are harmless, but you can temporarily disable spin down and see if they go away.

Power down and run chkdsk on your flash drive.

That's one of the reasons USB enclosures are not recommended, some use a smaller partition layout, and that won't be accepted by Unraid, you should be able to rebuild one disk at a time, assuming single parity, you can test first and make sure the emulated disk is mounting correctly, if yes rebuild on top with the disk connect via SATA.

Diags are just after rebooting, so not much to see, see here to enable syslog server/mirror, it might catch something more, still a couple of observations: Ryzen on Linux can lock up due to issues with c-states, make sure bios is up to date, then look for "Power Supply Idle Control" (or similar) and set it to "typical current idle" (or similar), or completely disable C-sates. More info here: https://forums.unraid.net/bug-reports/prereleases/670-rc1-system-hard-lock-r354/ This a big no-no, don't overclock the RAM, some Ryzen systems are known to be unstable and even corrupt data, respect max speed based on config:

You cache pool is at risk because of this bug, there are also instructions to correct it.

Can't open them, invalid zip.

When you posted the diags the disk was still disabled, but the emulated disk was empty, suggesting it was previously formatted, or it was already empty, that was I was asking, no rebuild had been done at that time, at least not on the diags posted.

Problem with the Asmedia controller (both affected disks are connected there), possibly related to virtualization: Jan 29 22:11:37 Tower kernel: ahci 0000:26:00.0: Event logged [IO_PAGE_FAULT domain=0x0000 address=0x000007fffff80000 flags=0x0020] Jan 29 22:11:37 Tower kernel: ahci 0000:26:00.0: Event logged [IO_PAGE_FAULT domain=0x0000 address=0x000007fffff80180 flags=0x0020] Jan 29 22:11:37 Tower kernel: ahci 0000:26:00.0: Event logged [IO_PAGE_FAULT domain=0x0000 address=0x000007fffff80280 flags=0x0020] Jan 29 22:11:37 Tower kernel: ahci 0000:26:00.0: Event logged [IO_PAGE_FAULT domain=0x0000 address=0x000007fffff80380 flags=0x0020] Jan 29 22:11:37 Tower kernel: ahci 0000:26:00.0: Event logged [IO_PAGE_FAULT domain=0x0000 address=0x000007fffff80480 flags=0x0020] Jan 29 22:11:37 Tower kernel: ahci 0000:26:00.0: Event logged [IO_PAGE_FAULT domain=0x0000 address=0x000007fffff80580 flags=0x0020] Jan 29 22:11:37 Tower kernel: ahci 0000:26:00.0: Event logged [IO_PAGE_FAULT domain=0x0000 address=0x000007fffff80600 flags=0x0020] Jan 29 22:11:37 Tower kernel: ahci 0000:26:00.0: Event logged [IO_PAGE_FAULT domain=0x0000 address=0x000007fffff80700 flags=0x0020] Jan 29 22:11:37 Tower kernel: ahci 0000:26:00.0: Event logged [IO_PAGE_FAULT domain=0x0000 address=0x000007fffff80800 flags=0x0020] Jan 29 22:11:37 Tower kernel: ahci 0000:26:00.0: Event logged [IO_PAGE_FAULT domain=0x0000 address=0x000007fffff80900 flags=0x0020] Jan 29 22:11:37 Tower kernel: AMD-Vi: Event logged [IO_PAGE_FAULT device=26:00.0 domain=0x0000 address=0x000007fffff80980 flags=0x0020] Jan 29 22:11:37 Tower kernel: AMD-Vi: Event logged [IO_PAGE_FAULT device=26:00.0 domain=0x0000 address=0x000007fffff80a80 flags=0x0020] Jan 29 22:11:37 Tower kernel: AMD-Vi: Event logged [IO_PAGE_FAULT device=26:00.0 domain=0x0000 address=0x000007fffff80b80 flags=0x0020] Jan 29 22:11:37 Tower kernel: AMD-Vi: Event logged [IO_PAGE_FAULT device=26:00.0 domain=0x0000 address=0x000007fffff80c80 flags=0x0020] Jan 29 22:11:37 Tower kernel: AMD-Vi: Event logged [IO_PAGE_FAULT device=26:00.0 domain=0x0000 address=0x000007fffff80d80 flags=0x0020] Jan 29 22:11:37 Tower kernel: AMD-Vi: Event logged [IO_PAGE_FAULT device=26:00.0 domain=0x0000 address=0x000007fffff80e80 flags=0x0020] Jan 29 22:11:37 Tower kernel: AMD-Vi: Event logged [IO_PAGE_FAULT device=26:00.0 domain=0x0000 address=0x000007fffff80f80 flags=0x0020] Jan 29 22:11:37 Tower kernel: AMD-Vi: Event logged [IO_PAGE_FAULT device=26:00.0 domain=0x0000 address=0x000007fffff81000 flags=0x0020] Jan 29 22:11:37 Tower kernel: AMD-Vi: Event logged [IO_PAGE_FAULT device=26:00.0 domain=0x0000 address=0x000007fffff81180 flags=0x0020] Jan 29 22:11:37 Tower kernel: AMD-Vi: Event logged [IO_PAGE_FAULT device=26:00.0 domain=0x0000 address=0x000007fffff81280 flags=0x0020]

Diags are after rebooting, so not much to see, see here for enabling syslog server/mirror, it might catch something.

It's possible to mount it with the Unassigned Devices plugin and transfer the data to the array, to be part of the array it needs to be formatted by Unraid.