Re: Brocken Raid & LUKS

[Phil Turmel <philip@turmel.org>]

On 02/21/2013 10:38 AM, Phil Turmel wrote:
> On 02/21/2013 10:30 AM, stone@heisl.org wrote:
> 
>> dd if=/dev/zero bs=4096 count=1 seek=1073006628 of=/dev/sdc1
>> dd: `/dev/sdc1': cannot seek: Invalid argument
>> 0+0 records in
>> 0+0 records out
>> 0 bytes (0 B) copied, 0,000493485 s, 0,0 kB/s
>>
>> is there a problem with the bs parameter?
>> shoud i try dd if=/dev/zero bs=512 count=8 seek=1073006628 of=/dev/sdc1
>> ?
> 
> How did you get 1073006628?  That is around the 4T mark?
> 
> Please show the badblocks output file.

I'm going to guess you didn't specify the block size when you used
badblocks.  It defaults to 1024.  If so, dd needs "bs=1024"

It is likely that your 48 errors are really 12 errors, four sequential
"blocks" for each.  Your drives are advanced format, so they really have
4k sectors, and that should have been specified to badblocks.

If so, you need to fix the sequential blocks together, or the drive will
fail to perform read-modify-write.

You probably need:

dd if=/dev/zero bs=1024 count=4 seek=1073006628 of=/dev/sdc1

But recheck everything carefully.  You can't undo whatever dd does.

Phil