| 1 |
On 1/13/06, matthew.garman@×××××.com <matthew.garman@×××××.com> wrote: |
| 2 |
> |
| 3 |
> I keep getting hard drive errors in my kernel log/dmesg that have me |
| 4 |
> worried. From /var/log/kernel/current: |
| 5 |
> |
| 6 |
> Jan 13 11:42:31 [kernel] hda: dma_intr: status=0x59 { DriveReady SeekComplete DataRequest Error } |
| 7 |
> - Last output repeated 7 times - |
| 8 |
> Jan 13 11:42:39 [kernel] hda: dma_intr: error=0x40 { UncorrectableError }, LBAsect=206696214, high=12, low=5369622, sector=206695927 |
| 9 |
> Jan 13 11:42:39 [kernel] ide: failed opcode was: unknown |
| 10 |
> Jan 13 11:42:40 [kernel] hda: dma_intr: status=0x59 { DriveReady SeekComplete DataRequest Error } |
| 11 |
|
| 12 |
These mean the blocks are corrupt, and cannot be read. Whatever was |
| 13 |
on those blocks is now lost. |
| 14 |
|
| 15 |
> On the drive. Apparently, an error was found (details below). I'm |
| 16 |
> not sure if this drive is actually dying, though, as the following |
| 17 |
> article (by the smartmontools author) suggests that one or two |
| 18 |
> errors on a drive is nothing to worry about. Also, the SMART |
| 19 |
> overall-health self-assessment test comes back as PASSED. |
| 20 |
|
| 21 |
I was able to resurrect a drive with a similar problem with: |
| 22 |
|
| 23 |
dd if=/dev/zero of=/dev/hda bs=32k |
| 24 |
|
| 25 |
!DANGER! the above command will destroy all data on the drive...but by |
| 26 |
writing to those sectors you can cause the drive to remap them to |
| 27 |
sectors reserved for that purpose. |
| 28 |
|
| 29 |
You can then check that the drive is working with: |
| 30 |
|
| 31 |
dd if=/dev/hda of=/dev/null bs=32k |
| 32 |
|
| 33 |
If either command fails, then it is time to replace the drive. In my |
| 34 |
case, that drive was still working perfectly 18 months later when I |
| 35 |
sold it to someone else. |
| 36 |
|
| 37 |
In any case, time to make sure you have a good backup. |
| 38 |
|
| 39 |
-Richard |
| 40 |
|
| 41 |
-- |
| 42 |
gentoo-user@g.o mailing list |
Archives