1

I had messages in /var/log/messages that seem to indicate that /dev/sdc is problematic (all besides the last two):

$ sudo grep sdc /var/log/messages
Apr 11 14:05:57 docker smartd[1070]: Device: /dev/sdc [SAT], 56 Currently unreadable (pending) sectors
Apr 11 14:05:57 docker smartd[1070]: Device: /dev/sdc [SAT], 56 Offline uncorrectable sectors
Apr 11 14:35:57 docker smartd[1070]: Device: /dev/sdc [SAT], 56 Currently unreadable (pending) sectors
Apr 11 14:35:57 docker smartd[1070]: Device: /dev/sdc [SAT], 56 Offline uncorrectable sectors
Apr 11 15:05:57 docker smartd[1070]: Device: /dev/sdc [SAT], 56 Currently unreadable (pending) sectors
Apr 11 15:05:57 docker smartd[1070]: Device: /dev/sdc [SAT], 56 Offline uncorrectable sectors
Apr 11 15:35:58 docker smartd[1070]: Device: /dev/sdc [SAT], 56 Currently unreadable (pending) sectors
Apr 11 15:35:58 docker smartd[1070]: Device: /dev/sdc [SAT], 56 Offline uncorrectable sectors
Apr 11 16:05:57 docker smartd[1070]: Device: /dev/sdc [SAT], 56 Currently unreadable (pending) sectors
Apr 11 16:05:57 docker smartd[1070]: Device: /dev/sdc [SAT], 56 Offline uncorrectable sectors
Apr 11 16:35:57 docker smartd[1070]: Device: /dev/sdc [SAT], 56 Currently unreadable (pending) sectors
Apr 11 16:35:57 docker smartd[1070]: Device: /dev/sdc [SAT], 56 Offline uncorrectable sectors
Apr 11 17:05:57 docker smartd[1070]: Device: /dev/sdc [SAT], 56 Currently unreadable (pending) sectors
Apr 11 17:05:57 docker smartd[1070]: Device: /dev/sdc [SAT], 56 Offline uncorrectable sectors
Apr 11 17:35:57 docker smartd[1070]: Device: /dev/sdc [SAT], No more Currently unreadable (pending) sectors, warning condition reset after 1 email
Apr 11 17:35:57 docker smartd[1070]: Device: /dev/sdc [SAT], No more Offline uncorrectable sectors, warning condition reset after 1 email

However, these last two lines above, and:

$ cat /proc/mdstat
Personalities : [raid10] 
md0 : active raid10 sdc2[2] sdd2[3] sdh2[7] sda2[0] sdg2[6] sdf2[5] sdb2[1] sde2[4]
      11712147456 blocks super 1.2 512K chunks 2 far-copies [8/8] [UUUUUUUU]
      bitmap: 8/88 pages [32KB], 65536KB chunk

unused devices: <none>

seem to indicate that /dev/sdc is not problematic anymore.

Is that correct?

boardrider
  • 889
  • 2
  • 15
  • 26

1 Answers1

3

Unfortunately not.

Your best bet is to (if necessary) install smartmontools. Using that you can run smartctl -a /dev/sdc

Warning: ignore that it might say “PASSED” that just means it’s successfully read the SMART data for the drive, not that it’s okay.

You want to be looking at offline uncorrectable, raw read errors and reallocated sectors.

Drives can have blips and be on the brink of failure however can be seen as okay to go back in the array, however you can expect another failure in that case. In the same breath, even a power outage can result in a raid wobble so might be nothing. Let us know how the smart data checking goes.

Timothy Frew
  • 582
  • 3
  • 7