Drive Read Errors

Anton Yuzhaninov citrin at citrin.ru
Wed Oct 19 23:54:16 UTC 2016


On 10/17/16 13:02, Jason C. Wells wrote:
> I recently used smartmontools long test and discovered that the drive
> had read errors.
>
> In your experience, is there any value in trying to use drive by mapping
> bad sectors or partitioning around the errors?

1. There is a value if your plan to use this HDD for not important data 
and downtime is also tolerable.

2. FYI: Remapping of bad sectors nowadays is done by disk's firmware. To 
make it happen just write any data to a all bad sectors. If disk is 
clean just run:
dd if=/dev/zero of=/dev/adaX bs=256k
then rerun SMART slef-test (and or dd if=/dev/adaX of=/dev/null).

3. Many people suggest to throw drive with bad sectors away. Such disk 
will likely have new pending sectors in future and probability of 
complete fail is higher than for drive with good SMART, but they can 
work for years after bad sectors are remapped. All depends on 
requirements for reliability and budget your have. At home I've used HDD 
which had slef-test fail and Current_Pending_Sector > 0 in past (for 
temporary data).

See also:
https://www.backblaze.com/blog/what-smart-stats-indicate-hard-drive-failures/


More information about the freebsd-questions mailing list