atacontrol

[oliwiak82]

Hello,

How I can check status of my RAID0? I have upgraded freebsd FreeBSD 8.2 to 9.0 and before I was using [CMD=""]atacontrol list[/CMD] etc. to show status.

I have the following message in /var/log/messages:

Apr 11 19:41:51 xxx fsck: /dev/raid/r0s1f: INCORRECT BLOCK COUNT I=6691756 (2727392 should be 2688704) (CORRECTED)
Apr 11 19:41:51 xxx fsck: /dev/raid/r0s1f: Reclaimed: 0 directories, 0 files, 9664 fragments
Apr 11 19:41:51 xxx fsck: /dev/raid/r0s1f: 867497 files, 25763745 used, 9338659 free (68227 frags, 1158804 blocks, 0.2% fragmentation)
Apr 11 19:43:19 xxx fsck: /dev/raid/r0s1d: UNREF FILE I=306526  OWNER=mysql MODE=100600
Apr 11 19:43:19 xxx fsck: /dev/raid/r0s1d: SIZE=0 MTIME=Apr 11 16:02 2012  (CLEARED)
Apr 11 19:43:19 xxx fsck: /dev/raid/r0s1d: UNREF FILE I=306527  OWNER=mysql MODE=100600
Apr 11 19:43:19 xxx fsck: /dev/raid/r0s1d: SIZE=0 MTIME=Apr 11 16:02 2012  (CLEARED)
Apr 11 19:43:19 xxx fsck: /dev/raid/r0s1d: UNREF FILE I=306529  OWNER=mysql MODE=100600
Apr 11 19:43:19 xxx fsck: /dev/raid/r0s1d: SIZE=0 MTIME=Apr 11 16:02 2012  (CLEARED)
Apr 11 19:43:19 xxx fsck: /dev/raid/r0s1d: UNREF FILE I=306530  OWNER=mysql MODE=100600
Apr 11 19:43:19 xxx fsck: /dev/raid/r0s1d: SIZE=0 MTIME=Apr 11 16:02 2012  (CLEARED)
Apr 11 19:43:19 xxx fsck: /dev/raid/r0s1d: UNREF FILE I=306531  OWNER=mysql MODE=100600
Apr 11 19:43:19 xxx fsck: /dev/raid/r0s1d: SIZE=0 MTIME=Apr 11 16:02 2012  (CLEARED)
Apr 11 19:43:19 xxx fsck: /dev/raid/r0s1d: Reclaimed: 0 directories, 5 files, 0 fragments
Apr 11 19:43:19 xxx fsck: /dev/raid/r0s1d: 83608 files, 509822 used, 5583273 free (5649 frags, 697203 blocks, 0.1% fragmentation)

 

[Terri_Kennedy]

How I can check status of my RAID0? I have upgraded FreeBSD 8.2 to 9.0 and before I was using [CMD=""]atacontrol list[/CMD] etc. to show status.

I think your underlying disk devices changed from /dev/adX to adaX, and [CMD=""]atacontrol[/cmd] doesn't show those. [cmd=""]camcontrol devlist[/cmd] would provide somewhat similar output. [cmd=""]graid list[/cmd] or [cmd=""]graid status[/cmd] might provide more useful output.

None of those will tell you about underlying hardware problems until/if the drive vanishes. You might want to look at the sysutils/smartmontools port to query the internal performance information from the drive.

I have the following message in /var/log/messages:

Apr 11 19:41:51 xxx fsck: /dev/raid/r0s1f: INCORRECT BLOCK COUNT I=6691756 (2727392 should be 2688704) (CORRECTED)
...

Those indicate a problem with the logical filesystem, which may or may not be caused by an underlying drive hardware problem. Generally, these are caused by an unclean shutdown (power failure, kernel panic, someone pressing the reset button, etc.).

 

[mav@]

None of those will tell you about underlying hardware problems until/if the drive vanishes.

That is not exactly correct. graid has concept of "failed" devices. If write to device failed and volume has redundancy, device will be marked as "failed" and excluded from all I/O operations until replaced or manually removed from array and reinserted back. If read fails, graid redirects operation to other device and tries to fix original one by writing restored data over the corrupted area. If that write fails, device also marked as "failed". That status can be seen in graid list/status output.