Robert, very nice directions, thank you!
Bernd, you might want to review http://linux.die.net/man/8/hdparm first. Depending upon what file system you are using, you might also be able to figure out in advance or after the fact, what file or directory is using the data stored at that disk sector. After you have finished you should run a full disk self-test (scan) to be sure that there are no other unreadable sectors.
On Jul 21, 2014, at 1:41 AM, email@example.com wrote:
The disk won't either re-allocate or "repair" the sector until it is
written to. If the sector can be written to, then the count goes back down
I clear pending sectors using hdparm. You will need a live Linux CD with
it. Knoppix has it, as does Clonezilla, or ALT rescue, among others.
Be warned that you will lose whatever data, hence file, is in that sector,
but then, it's really already lost anyway.
1) Boot off one of the live CDs and become root.
2) Figure out which drive name Linux gives this drive. Likely it will be
/dev/sda . Use fdisk -l or smartctl --scan to be sure.
3) Run the following command: hdparm --read-sector 121893751 /dev/sda
(assuming the drive is /dev/sda; if not substitute the drive's name). Did
you get an error? If yes, that is good (well, kind of). Continue to the
4) Edit the command you just ran (use the up arrow). Don't retype! You
only get one chance to get it right. Again, assuming the drive is
/dev/sda, edit the command to read:
hdparm --write-sector 121893751 --yes-i-know-what-i-am-doing /dev/sda
5) Did that succeed? Yes, then try step 3 again. Did it succeed and give
you nothing but zero's? If yes, now recheck SMART. The sector should be
6) Now wait 5 minutes to an hour and run step 3 again (or just run the
SMART short test). Did it come back? No is good. You should be done. If
yes, you have a medium error. That's bad.
a week ago on two consecutive days the syslog reported a defect block on
the internal HDD (attached).
Of course smartd noticed about the SMART attribute Current Pending
Sector Count: 1
Now I receive a smartd warning after every selftest of the HDD, because
the number of failed self tests increased (attached).
What I read about reallocating the defect block under Windows 7 I must
admit I did not right understand. As far as I understood the disk
reallocates the block after some time. After a week this did not happen
I tried the tool that came with smartmontools for windows
(fix_badsectors.cmd with dd.exe). But that corrupted the boot partition
(system not found at boot) and didn't change the SMART attributes. So I
recovered the boot partition.
Please, can someone give me advice with dealing with that problem? Thank
you very much.
E-Mail ist virenfrei.
Von AVG uberpruft - www.avg.de
Version: 2014.0.4716 / Virendatenbank: 3986/7892 - Ausgabedatum: 21.07.2014