Quick Hard Drive Error Question

[PK1048]

On Aug 17, 2015, at 8:56, dweimer <dweimer at dweimer.net> wrote:

> I have an external SATA drive dock that I use to write backups to, I have steadily been getting more parity/CRC errors over the last couple of weeks.
> 
> Aug 17 00:14:14 freebsd kernel: (ada4:ahcich5:0:0:0): WRITE_FPDMA_QUEUED. ACB: 61 00 70 e7 53 40 85 00 00 01 00 00
> Aug 17 00:14:14 freebsd kernel: (ada4:ahcich5:0:0:0): CAM status: Uncorrectable parity/CRC error
> Aug 17 00:14:14 freebsd kernel: (ada4:ahcich5:0:0:0): Retrying command
> Aug 17 00:14:14 freebsd kernel: (ada4:ahcich5:0:0:0): READ_FPDMA_QUEUED. ACB: 60 70 d0 c6 1c 40 84 00 00 00 00 00
> Aug 17 00:14:14 freebsd kernel: (ada4:ahcich5:0:0:0): CAM status: Uncorrectable parity/CRC error
> Aug 17 00:14:14 freebsd kernel: (ada4:ahcich5:0:0:0): Retrying command
> 
> I have ran smart tests on the drives, they report no errors, hers the most recent summary line from the currently mounted drive.

Are there any port multipliers involved ? When I had 4 SATA drives behind a port multiplier I had these type of errors frequently under high load. I was using ZFS so dialed back the outstanding queue depth from 10 to 2 to reduce load and the errors went away. I have since gone to one port per drive (and went back to the default queue depth of 10) and have not had the problem recur…

But, I have seen a “finicky-ness” between eSATA ports and cables. Certain cables work on certain ports but not on others. Have you tried substituting the eSATA cable ? A different eSATA port perhaps ?