Hey guys,
I have successfully been running a FreeNAS Server for a while now without any issues.
I recently switched to a Software RAID configuration (GEOM Mirror) with my Disks and have had some weird problems with my 1TB Software RAID.
My server would suddenly shutdown after I tried accessing smb or when it was told so by a cron job but
it halted at the end and did not switch off completely. I then had to go up to the server and reboot it manually.
The following error messages where printed (only) on the monitor:
When I try to manually stop the device with
the same error comes up in the cli.
I ran fsck on the mirror and it doesn't find any faults.
Here some additional information:
Software RAID 1 with a Samsung SpinPoint F1 1 TB and WDCaviarGreen 1TB.
I've been also getting smartd reports with errors from the WDCaviarGreen Disk. This may be at the root of the problem:
(I can post full smartd report, if needed)
I know that the Load_cylce_count parameter on the WD Disk is very high, since I've only recently used WIDLE3 to set the parking of the head to a higher value. Maybe the disk is broken.. (hope not!)
Further errors are logged about the mirror device, but which, after searching through the FreeNAS forum, I have found to ignore because they are falsely reported by the OS.
System Info:
FreeNAS 0.7.2 Sabanda (revision 6694)
OS Version FreeBSD 7.3-RELEASE-p6 (revision 199506)
Platform amd64-embedded on Intel(R) Atom(TM) CPU 330 @ 1.60GHz
RAM 2GB
I'd be glad for some help with resolving this problem.
Cheers
I have successfully been running a FreeNAS Server for a while now without any issues.
I recently switched to a Software RAID configuration (GEOM Mirror) with my Disks and have had some weird problems with my 1TB Software RAID.
My server would suddenly shutdown after I tried accessing smb or when it was told so by a cron job but
it halted at the end and did not switch off completely. I then had to go up to the server and reboot it manually.
The following error messages where printed (only) on the monitor:
Code:
gmirror: Cannot destroy device Hell (error=16)
geom: Could not unload module: device busy.
... (syncing buffer, inodes, etc.)
GEOM_MIRROR: Device Hell: provider mirror /Hell destroyed.
GEOM_MIRROR: Device Hell destroyed
When I try to manually stop the device with
Code:
geom mirror stop Hell
Code:
geom: Cannot destroy device Hell (error=16).
I ran fsck on the mirror and it doesn't find any faults.
Here some additional information:
Software RAID 1 with a Samsung SpinPoint F1 1 TB and WDCaviarGreen 1TB.
I've been also getting smartd reports with errors from the WDCaviarGreen Disk. This may be at the root of the problem:
Code:
Device: /dev/ad12, 4 Currently unreadable (pending) sectors
Device: /dev/ad12, 4 Offline uncorrectable sectors
(I can post full smartd report, if needed)
I know that the Load_cylce_count parameter on the WD Disk is very high, since I've only recently used WIDLE3 to set the parking of the head to a higher value. Maybe the disk is broken.. (hope not!)
Further errors are logged about the mirror device, but which, after searching through the FreeNAS forum, I have found to ignore because they are falsely reported by the OS.
Code:
kernel: GEOM: ad14: using the primary only -- recovery suggested.
kernel: GEOM: ad14: the secondary GPT table is corrupt or invalid.
kernel: GEOM: ad12: using the primary only -- recovery suggested.
kernel: GEOM: ad12: the secondary GPT table is corrupt or invalid.
System Info:
FreeNAS 0.7.2 Sabanda (revision 6694)
OS Version FreeBSD 7.3-RELEASE-p6 (revision 199506)
Platform amd64-embedded on Intel(R) Atom(TM) CPU 330 @ 1.60GHz
RAM 2GB
I'd be glad for some help with resolving this problem.
Cheers