ZFS Checksum Errors

[sechs]

Twice now, I've had one of the drives in a RAIDZ produce about fifty recoverable checksum errors during the weekly scrub. Both times, I cleared the error and ran a new scrub with no problems. There are no apparent issues reporting via SMART. No problems were found upon inspecting the hardware.

This is an HP Microserver running FreeNAS. It does not have ECC memory, but I would expect memory errors to either be silent or spray indiscriminately. The drive in question is a 3TB Seagate only a few months old.

What could be causing this, and should I be concerned for the data?

 

[Chewy509]

There are no apparent issues reporting via SMART.

Are there any soft read/write errors?

This is an HP Microserver running FreeNAS. It does not have ECC memory, but I would expect memory errors to either be silent or spray indiscriminately.

If it was bad memory, I expect there to be bad checksums on all disks...

The drive in question is a 3TB Seagate only a few months old. What could be causing this, and should I be concerned for the data?

Bad cable(s), bad power connection or bad/non-clean power in general, bad drive... I would record the current SMART values, and if it happens again compare the SMART values to see if something shows up.

 

[sechs]

I haven't noticed any issues. The only signs that I've found of any problem are these checksum errors.

This sounds like a wait and watch situation. Not my favourite.