It’s never a good sign when ZFS starts giving you uncorrectable I/O failure errors.

kernel: WARNING: Pool 'parity0' has encountered an uncorrectable I/O failure and has been suspended.
kernel: WARNING: Pool 'parity0' has encountered an uncorrectable I/O failure and has been suspended.
kernel: WARNING: Pool 'parity0' has encountered an uncorrectable I/O failure and has been suspended.
kernel: WARNING: Pool 'parity0' has encountered an uncorrectable I/O failure and has been suspended.
kernel: scsi_io_completion_action: 16 callbacks suppressed
kernel: sd 3:0:0:0: [sdc] tag#24 FAILED Result: hostbyte=DID_BAD_TARGET driverbyte=DRIVER_OK
kernel: sd 3:0:0:0: [sdc] tag#24 CDB: ATA command pass through(16) 85 06 20 00 00 00 00 00 00 00 00 00 00 0
0 e5 00
rem kernel: sd 3:0:0:0: [sdc] tag#26 FAILED Result: hostbyte=DID_BAD_TARGET driverbyte=DRIVER_OK
kernel: sd 3:0:0:0: [sdc] tag#26 CDB: ATA command pass through(16) 85 06 20 00 00 00 00 00 00 00 00 00 00 0
0 e5 00

Currently, it seems like the drive is fine until you actually try to write to it, at which point ZFS locks up and dmesg starts spewing DID_BAD_TARGET errors, which are apparently the mark of a failing disk. I’m leaving it alone for now, in hopes that its internal firmware can quietly do some recovery (re-shingling?).

Don’t worry, though - this is part of a redundant LizardFS pool, so no data has been lost. Yet.

It seems that this error might also just show up sometimes with SMR disks? Which is slightly reassuring, but not really since I never got these issues in over a year of usage so far.

Update (2019-03-16)

Seems we’re good. After letting it sit for an hour or two, I re-imported the ZFS pool and started the LizardFS chunkserver. No errors yet.