Is disk really bad if mkfs.ext4 -c -c reports errors?

[Gene Heskett]

On Wednesday 11 January 2017 06:47:51 Chris Green wrote:

> I have been reformatting a disk which had a few corrupted files on it
> with the following command:-
>
>     mkfs.ext4 -c -c /dev/sdb1
>
> (Yes, I know this destroys all the data, everything except the few
> corrupted files has been copied off the disk)
>
> The second pass of badblocks is reporting some errors.  Does this
> indicate that the disk is dying?
>
> The problem is that the documentation doesn't really tell one what the
> errors mean.  Do the reported errors get remapped using the bad blocks
> list or are they errors which *can't* be remapped because the bad
> blocks list is full?  Running SMART diagnostics tells me the disk
> drive has no errors.
>
> --
> Chris Green

You might try a different sata cable, particularly it its a red one 
tinged with a hot magenta coloring. Something in that color of that dye, 
turns the copper in the wires into red, oxidized copper powder in a few 
years.  I am  C.E.T., and its been a problem child for me ever since the 
Japanese took over the microphone cable business for CB radios in the 
early 1970's. This system is nearly a decade old and I've had to replace 
every red cable in it with some other color of jacket in that decade. 
I've several drives, and some of mine are black, some tan, but no red 
ones have survived. I have several terabyte drives, one of which has had 
a reallocated sector count stuck at 25 since I first ran the smartctl 
report on it at about 24,000 spinning hours. Its now just past the 
70,000 spinning hours mark, no additional reallocated sectors have 
occurred.

You might also check its model/serial numbers against a list that at 
least seagate has, download and install the latest firmware in the 
drive. I did that within the first 5k hours of service and gained 40 
megabytes a second on two of them I'd bought locally from Staples, with 
no loss of existing data. Its relatively safe, burn it to a cd, and 
reboot to the cd, it surveys the system and updates the firmware in only 
those drives that need it. In my case the updated drives included the 
one amanda uses for virtual tapes, IOW the drive all my backups were on. 
And thats the drive with 70k+ spinning hours on it right now. I run 
the -t long test on them about monthly, including the system drive while 
its live, no problems have been detected in all this time.

The best cable test? Put a tail on the messages log, and touch the cable, 
moving it an inch or so.  If the logfile gets spammed with drive resets, 
bad dog, replace it soonest, preferably last month.  With a non-red 
cable.

Cheers, Gene Heskett
-- 
"There are four boxes to be used in defense of liberty:
 soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
Genes Web page <http://geneslinuxbox.net:6309/gene>