JorgeB

Try a different cable, and one as short as possible, initial connection, at least the ones I saw are only linking at 1.5gbps: Sep 12 10:59:21 Tower kernel: ata1: SATA link up 1.5 Gbps (SStatus 113 SControl 310) That already suggests a connection quality problem, later the disk dropped offline: Sep 15 07:14:24 Tower kernel: ata1: hard resetting link Sep 15 07:14:24 Tower kernel: ata1: SATA link up 1.5 Gbps (SStatus 113 SControl 310) Sep 15 07:14:54 Tower kernel: ata1.00: qc timeout (cmd 0xec) Sep 15 07:14:54 Tower kernel: ata1.00: failed to IDENTIFY (I/O error, err_mask=0x4) Sep 15 07:14:54 Tower kernel: ata1.00: revalidation failed (errno=-5) Sep 15 07:14:54 Tower kernel: ata1.00: disabled

Yep.

You can just unassign parity and start the array to "forget it" and stay with just parity2 as singe aprity, if you want to move parity2 to parity you'll need to re-sync it since it's not interchangeable.

It's up to you, preclear is not essential, but since you still have one parity disk and are still protected you can if you want.

The standard way, just assign a new disk to parity and sync it.

If the error was just the SMART warning it's likely a false positive, there are other SSDs with similar issues.

Did you try it already? Some info came to light that your problem might just be corrupt GPT tables, and if it's that it could be fixed without rebuilding the disks.

I would first test that drive with another power cable, if same then replace.

I had some time to take a look at this, if you still need help to repair the corrupt GPT partition post back, it's easily done with gdisk.

Sep 7 20:24:58 SERVER-LOKI emhttpd: error: ckmbr, 2136: No such file or directory (2): open: /dev/sdd1 This error suggests a problem with the GPT partition, after the problem happens please post output of: fdisk -l /dev/sdX

Parity disk is showing some reallocated sectors, but problem this time looks more like a connection issue, check/replace cables and re-sync parity.

Do you have the docker folders plugin installed? If yes try without it, if not try after booting in safe mode

Please post the diagnostics: Tools -> Diagnostics

Looks like a device problem, please post the diagnostics: Tools -> Diagnostics

It appears that the GPT corruption is causing an error that makes Unraid not even detect the exiting partition, I have no experience repairing GPT partitions, quick google search found this: http://www.rodsbooks.com/gdisk/repairing.html Another option would be to create a new partition on top of that one, it should work but I would recommend cloning the disk first with dd and try on the clone, if it works it might only mount with UD, not on the array/cache, but it would at least allow you to access the data. P.S. don't try to change the btrfs UUID, it won't help and can cause further problems.

There's a checkbox to allow starting the array without the cache assigned. According to UD there's no partition, post the output of: fdisk -l /dev/sde

There's a problem with the partition, unassign cache, start the array, install the Unassigned Devices plugin and see if there's an option to mount the disk there.

According to the manual it's on the OC page, then:

Don't move to slot 2, leave the ssd on the 1st slot and change total cache slots to 1.

Sorry, that was a typo, tools -> diagnostics is what I meant to write. Try this: -stop array -change cache slots to 1 -click on cache and set filesystem to "auto" -start array and post new diags

Was the cache a pool? Please post the Unraid diagnostics: Tools -Z Diagnostics

See the first couple of posts in the UD plugin support thread:

Syslog is after rebooting, so not much to see, try this and attach that one after it freezes, also attach complete diagnostics (even if after rebooting).

Yeah, probably a device problem, you can try formatting assigned as cache but would except same error.

Yep, it's what I'm doing also.