Is there way to get filename for specific LBA?

[Ross]

On Wed, Aug 31, 2011 at 4:11 PM,  <perryh at pluto.rain.com> wrote:
> Ross <basarevych at gmail.com> wrote:
>
>> Aug 31 05:13:24 da kernel: ad6: WARNING - READ_DMA UDMA ICRC error
>> (retrying request) LBA=107491647
>
> That message is reporting a problem in communication between the
> drive and the controller (or, perhaps, between the controller and
> main memory), not a problem reading the media, so the LBA is likely
> not all that useful (esp. since, if you got no other messages, the
> retry succeeded so no data was lost).
>
> What does
>
>  egrep 'ad[0-9]|ata' /var/run/dmesg.boot
>
> report?
>

atapci0: <Intel ICH7 SATA300 controller> port
0x20b8-0x20bf,0x20cc-0x20cf,0x20b0-0x20b7,0x20c8-0x20cb,0x20a0-0x20af
mem 0xe0284000-0xe02843ff irq 19 at device 31.2 on pci0
atapci0: [ITHREAD]
ata2: <ATA channel 0> on atapci0
ata2: [ITHREAD]
ata3: <ATA channel 1> on atapci0
ata3: [ITHREAD]
ad4: 238475MB <Seagate ST9250315AS 0001SDM1> at ata2-master UDMA100 SATA
ad6: 476940MB <Seagate ST9500325AS 0001SDM1> at ata3-master UDMA100 SATA
Trying to mount root from ufs:/dev/ad6s1a


>> #  dd if=/dev/ad6 of=/dev/null bs=1m seek=107491647 count=1
>> dd: /dev/null: Inappropriate ioctl for device
>>
>> Another question: why does it fail?
>
> seek= applies to the output file, so it tried to do a seek on
> /dev/null :)  You probably wanted skip= (or iseek=).
>

Thank you :) I have read the man page now.

smartd also reports this:

Aug 31 10:41:04 da smartd[886]: Device: /dev/ad6, Failed SMART usage
Attribute: 184 End-to-End_Error.

I found this explanation: http://kb.acronis.com/content/9119

So disk is dying? Or is it cable. I have no physical access to the
server at the moment.
But still, is there a way to get the filename for LBA?