check disk for bad clusters on disk

[doctordruidphd]

Re: check disk for bad clusters on disk

sudo hdparm -tT /dev/sdb

/dev/sdb:
Timing cached reads: 6190 MB in 2.00 seconds = 3095.92 MB/sec
Timing buffered disk reads: 34 MB in 4.95 seconds = 6.87 MB/sec
linda@linda:~$ sudo hdparm -tT /dev/sdb1

/dev/sdb1:
Timing cached reads: 2 MB in 4.43 seconds = 462.74 kB/sec
Timing buffered disk reads: 28 MB in 4.09 seconds = 6.85 MB/sec

Whoa! That's interesting. Tells me right away that the hardware is OK.
Maybe try creating partitions with different file systems - ext4, fat, ntfs, and run the check again & see if that makes a difference.

Also, how are you mounting sdb1?

[lindaonline15]

Re: check disk for bad clusters on disk

Whoa! That's interesting. Tells me right away that the hardware is OK.
Maybe try creating partitions with different file systems - ext4, fat, ntfs, and run the check again & see if that makes a difference.

yea.. it really sounds that its getting interesting !
anyways the drive is having only 1 partition, (i formatted with gparted as 1 whole primary partition) so i just mount it normally from GUI by opening it in dolphin... (should it be any other way...?)
ok... I formatted this time using ext4, and still the same problem exists...
im really going to set all bits to zero now...
the command to do it should be this, right? :

dd if=/dev/zero of=/dev/sdX

[doctordruidphd]

Re: check disk for bad clusters on disk

One thing you might do first -- see what the mount command returns. The only thing I can think of that's left to try, other than a low-level format, is looking at what sort of parameters the drive is mounted with.

Also, you might try mounting it manually, as see what it gives you. To do that, unmount the drive, then go into a terminal and type (assuming this is a ext4 formatted partition):

sudo mkdir /media/usbdrive
sudo chmod -R a+rwx /media/usbdrive
sudo mount -t ext4 -o users,rw,exec /dev/sdx1 /media/usbdrive

check that it mounted with the mount command, then try reading and writing to /media/usbdrive.

You can unmount this with sudo umount /media/usbdrive

If that doesn't do anything, then I guess a low level format is the nest step.

I don't use dd very much, so I would go with dibl's suggestion:

dd if=/dev/zero of=/dev/sdx bs=512 count=1

where the 'x' in sdx is the appropriate device; and as dibl said, make "double damned sure" because that is absolutely it for any data on the drive.

Then you will have to set up the drive again with gparted, with a new msdos mbr and a new partition.

Edit: oops, just caught a mistake & fixed -- should be /dev/sdx1 in the mount command.

[lindaonline15]

Re: check disk for bad clusters on disk

ok... I tried dd and I got this:

sudo dd if=/dev/zero of=/dev/sdb
dd: writing to `/dev/sdb': Input/output error
282065+0 records in
282064+0 records out
144416768 bytes (144 MB) copied, 3334.61 s, 43.3 kB/s

it doesnt seem that it has been totally done, right? (considering my hard disk's size is 1TB)

[doctordruidphd]

Re: check disk for bad clusters on disk

I don't use dd, never really had a reason to. But I found the following with google:

from: http://www.linuxquestions.org/questi...ommand-362506/

dd if=<source> of=<target> bs=<byte size> ("USUALLY" some power of 2, and usually not less than 512 bytes (ie, 512, 1024, 2048, 4096, 8192, 16384, but can be any reasonable whole integer value.) skip= seek= conv=<conversion>

And for zeroing a drive: http://www.linuxquestions.org/linux/...rive_zero_fill

Case:
I have some hard drives that I had donated, on the condition that I zero-fill them. Zero-fill will erase all data on a hard drive, but this needs to be differentiated from a low-level format, which could damage newer (pretty much anything over 512MB) drives. The key is to find a way to wipe the data on a hard drive without destroying the current manufacturer's low-level format. I searched the web and found DOS utilities that sounded like low-level format utilities and usually were vendor specific, but then I ran into some information about using the simple Linux dd command to zero-fill.

Solution:
What you need is a bootable Linux, there are several available, booting a live Linux such as knoppix or even Tom's RTB, a bootable Linux that fits on floppy, but I happened to use my trusty Slackware Installation CD which, during the install, dumps you to a handy root shell for this. When you boot into a command shell, as root, simply pick the drive you want to fill, hda is primary master, hdb is primary slave, hdc is secondary master, sda is the first SCSI drive, etc, and use the command:
Code:

dd if=/dev/zero of=/dev/hda bs=1M

WARNING! This will destroy ALL data on the hard drive, all boot sector info and all data on all partitions. It will not however render the disk useless, you simply have a clean disk that only needs to have partitions created and a new boot sector installed, which will happen when you install any OS including Linux or Windows. This is a good way to clean up any bad partitions, viruses, botched installs or data that you don't want to be seen (OK, it is theoretically possible to forensically retrive data after using this method, but not likely and not doable by the average computer user) and it is simple to do with existing GNU/Linux!

In the above case, you would use /dev/sdb instead of /dev/hda

So try with one of the bs suffixes. Since you are just zeroing and will reformat over it, probably doesn't matter what size.

Edit: looks like the best link for learning dd is:http://en.wikipedia.org/wiki/Dd_%28Unix%29

[Snowhog]

Re: check disk for bad clusters on disk

Also, the very well written NEW-> The dd Command by Qqmike right here in KFN.