Perusal, Synthesis, Bliss

November 14, 2015: confirmed death of an external hard drive 2"1/2 Samsung G2 500 Gb (model: HM502JX)

This disk was sold with its own box (read further). Four or five years ago I had I/O errors with this disk when performing my backups. But I wanted to be sure that it is not possible to save it and get it to work again. I have used the following procedure. First I got the confirmation that there was some problems on the disk, with the badblocks utility:
$ sudo badblocks /dev/sdd1
[sudo] password for jscordia: 
61016492
61016516
61016517
[many more lines]
So I have re-formatted the disk in a “detailed” way:
$ sudo mke2fs -cc -t ext4 /dev/sdd1
[sudo] password for jscordia: 
mke2fs 1.42.12 (29-Aug-2014)
/dev/sdd1 contains a ext4 file system labelled ’SAMSUNGCROCODILE’
        last mounted on Wed Nov 11 11:07:35 2015
Proceed anyway? (y,n) y
Creating filesystem with 122096384 4k blocks and 30531584 inodes
Filesystem UUID: d05ec811-69da-4f5f-98fa-debfc46be1f9
Superblock backups stored on blocks: 
        32768, 98304, 163840, 229376, 294912, 819200, 884736, 1605632, 2654208, 
        4096000, 7962624, 11239424, 20480000, 23887872, 71663616, 78675968, 
        102400000
Testing with pattern 0xaa: done                                                 
Reading and comparing: done                                                 
Testing with pattern 0x55: done                                                 
Reading and comparing: done                                                 
Testing with pattern 0xff: done                                                 
Reading and comparing: done                                                 
Testing with pattern 0x00: done                                                 
Reading and comparing: done                                                 
Allocating group tables: done                            
Writing inode tables: done                            
Creating journal (32768 blocks): done
Writing superblocks and filesystem accounting information: done
No error on disk was signaled by mke2fs. After that, badblocks returned no error:
$ sudo badblocks /dev/sdd1
$
But as it already happened to me in the past, after some extensive write operation on the disk (I have saved more than 400 Gb of data on it with rdiff-backup), badblocks again returns erroneous sectors:
$ sudo badblocks /dev/sdc1
362807452
362807600
362807601
362807602
[many more lines]
We can remark that the sector numbers (or at least their order) are different from the ones reported initially by badblocks.
Conclusion: the disk is really dead. And the box cannot be used for another hard drive, because of some special connection inside.