CAM status: Uncorrectable parity/CRC error

Hi,
Looking into the machine’s syslog, lots (hundreds) of occurrences of

2023-11-16T22:09:32	Notice	kernel	 (ada0:ahcich0:0:0:0): CAM status: Uncorrectable parity/CRC error
2023-11-16T22:09:32	Notice	kernel	 (ada0:ahcich0:0:0:0): WRITE_FPDMA_QUEUED. ACB: 61 90 b8 68 68 40 01 00 00 00 00 00
2023-11-16T22:09:32 Notice	kernel	 (ada0:ahcich0:0:0:0): Retrying command, 3 more tries remain

Probably should be worried ? :frowning:
Nitrowall bought in Feb 23 (but in service since May)

I should say that, inconsistently, from time to time, DHCP and the web GUI stop working, although devices that already have a working lease can still use the network (and the internet), but any new device or reconnecting device is dead in the water. Only a cold reboot of the Nitrowall fixes the issue… :frowning:

I can reliably trigger the above messages with fio e.g.

$ fio --rw=randrw --name=test --bs=128k --direct=1 --numjobs=1 --ioengine=posixaio --iodepth=32 --refill_buffers --size=512M
root@gw01:~ # dmesg | tail -3
(ada0:ahcich0:0:0:0): WRITE_FPDMA_QUEUED. ACB: 61 b8 a0 fc 0a 40 02 00 00 00 00 00
(ada0:ahcich0:0:0:0): CAM status: Uncorrectable parity/CRC error
(ada0:ahcich0:0:0:0): Retrying command, 3 more tries remain
$ smartctl -a /dev/ada0
...
199 UDMA_CRC_Error_Count    0x0000   100   100   050    Old_age   Offline      -       28683
1 Like