Salix Forums

A few days ago when tried to open my VLSOHO-7 box I got kernel panic.
Trying another box Slackel-OpenBox same, Kernel panic.

Both partitions are on a relatively new sata drive. Booting into Knoppix
it showed that in those above mentioned partitions most of the / directories
were gone. Of the 19 or so only 7 were left. The same is true for my Salix-Fluxbox
on same drive, -gone.

Other partitons on the sata drive were unaffected. I checked all partitions with
badblock and no bad blocks were detected.

In my 30 odd years on computers I never experienced such a loss of directories
and have no clue how this could have happened.

Had it to do with the passing of the asteroid or the crash of the meteor in Russia at that time?    

Anyone having had such an experience?

I posted this in another forum and had sofar one reply of a person saying the same happened
to him at about the same time in his slackware-13.37 box.

There may have been a misunderstanding, or at least I hope so, as I replied to your post on the Slackel forum. I apologize in advance for any misunderstandings on my part:

While I do run the same distros which you lost, Slackel Openbox and Salix Fluxbox 13.37, I have yet to lose any data on either of those.

With modern hard drives, badblocks rarely if ever actually reports any bad blocks found. Instead, the point of running badblocks is to alert the drive itself to the presence of any sectors which are about to fail, so the drive can swap those out for spare sectors.

Actually, badblocks should be run on an entire drive, rather than on individual partitions. Also, for best results, it should be allowed to completely wipe the drive using its option -w. Before doing that, of course, you should have a better reason to suspect hardware issues with the drive:

So for in-depth testing, the GSmartControl utility can definitively report on recent failures, as well as test the drive's current condition.

SpinRite is a commercial, proprietary utility which can sometimes be more thorough than the F/LOSS tools I mentioned above. SpinRite also can often accomplish more non-destructively than can badblocks.

Jayseye, it was not you who reported it even so yes your post read like it did happen.
It was in another forum not slackel's.

smartctl -A, does not report any errors on the drive in question.

Just a thought, but do you have children or pets, etc., etc.

knome wrote:Just a thought, but do you have children or pets, etc., etc.

No children, no warm blooded animals except my dog but he can't reach the keyboard.

Before this wipe-out happened I tried to open the forum of cups. Firefox gave me a warning
that this site is insecure. I tried Midori and there was no warning. I opened the forum of cups
and did not stay long since it was somewhat confusing to me.

Of the 5 partitions I have on this sata drive 2 were not affected, one my /usr/local and one the
Tiny Core Linux box. The partition with VL7 was only partially cleaned, that is more directories
were left than on the Slackel-Openbox and Salix-Fluxbox.

I have reformatted the wiped partitons, reinstalled Slackel-OpenBox and made a copy of
Salix-14-Xfce4 to the partition where Salix-Fluxbox was. Sofar both boxes work w/o a problem.

Atip wrote:smartctl -A, does not report any errors on the drive in question.

At least run a long test before checking the output (if you don't run it on a regular basis): smartctl -t long /dev/sd...
This will take some time to finish.

PS: last time I had a badly corrupted ext4 file system it was a USB hard disk failure

thenktor wrote:At least run a long test before checking the output...

Agreed and, even more to the point in this case, the error log would confirm whether a drive error occurred at the time when files were lost. As a lazy Slacker I really like GSmartControl.

thenktor wrote:

Atip wrote:smartctl -A, does not report any errors on the drive in question.

At least run a long test before checking the output (if you don't run it on a regular basis): smartctl -t long /dev/sd...
This will take some time to finish.

smartctl -t long /dev/sdb

SMART Extended Self-test Log Version: 1 (1 sectors)
Num Test_Description Status Remaining LifeTime(hours)
LBA_of_first_error
# 1 Short offline Completed without error 00% 2209
# -
# 2 Extended offline Completed without error 00% 2209

The 2 partitions which were wiped were xfs and one less damaged ext4. Other xfs and ext4
on the drive were not affected.

Well, your SMART output looks good. Last thing you can do is using the hard disk manufacturer's test tool to check your drive. But I don't know if they test more than SMART already does.
Also doing a memory test is never a bad idea if there are problems of any kind

The drive currently tests OK, because any bad sectors would likely have been swapped out by running badblocks -n, which was done after re-installing.

However, if the data loss had been caused by a drive problem, then looking at the history, via the SMART error log, would confirm that. It would also prove that software, including Salix, did not cause the data loss.

Of course I also agree about memtest86+.