Kernel Panic, looks like a HD problem

Spanik · Guru Joined: 12 Dec 2003 Posts: 457 Location: Belgium

Since about a week my desktop starts to behave strange. At first sometimes the mouse and keyboard seemed to lock. Then I had a few times that Opera could not write the bookmarks. But the last 2 days I got Kernel Panics. I wrote down the message of the last crash:

"Kernel Panic - not syncing : <4> Reiserfs panic (device sdc3): vs-7042 entry_points_to_object: entry must be ready

Pid: 3378, comm: claws-mail Not tainted 3.0.6-gentoo"

It happened when I closed Claws and right now it refuses to start. Sdc3 is the / of the setup. This is an SSD that has already a few years work done.

Now I have already put in a new SSD in order to install a new Gentoo, but right now I really want to know if there is a way to tell if:
- the whole pc is slowly disintegrating: could be as this is about 12 years old now
- the SSD /dev/sdc is failing
- there is a simple filesystem error on /dev/sdc3

If the pc is failing then I won't spend the time doing a complete setup on it but I'm a bit in trouble with some legacy pci cards that won't be easy/cheap to replace. If it is the SSD that is failing so fast then I won't put one in for the OS anymore. If it's a filesystem error then maybe I can get it running agian long enough to build a new OS.

It is the first time I got one of these, so how do I start digging around what is wrong?
_________________
Expert in non-working solutions

NeddySeagoon · Posted: Thu Jan 22, 2015 8:38 pm Post subject:

Spanik,

Check your SSD SMART data with smartmontools
Post the output of

Spanik · Guru Joined: 12 Dec 2003 Posts: 457 Location: Belgium

Is this provided on the livedvd? Don't have smartctl installed and emerge doesn't work (one of the reasons I need to re-install Gentoo).
_________________
Expert in non-working solutions

NeddySeagoon · Posted: Sat Jan 24, 2015 12:14 pm Post subject:

Spanik,

I don't have the livedvd - if you, try it.

Its very difficult to break gentoo so badly that you need to reinstall. emerge not working isn't one of them.
A dead or dying hdd might be though.
_________________
Regards,

NeddySeagoon

Computer users fall into two groups:-
those that do backups
those that have never had a hard drive fail.

Spanik · Guru Joined: 12 Dec 2003 Posts: 457 Location: Belgium

I know, but this pc hasn't been updated for very long, I changed the profile but now it is in conflict with whatever is installed etc. Then there is the systemd thing going on. So I'd like to just switch to a new install on a new disk and keep this one around "for just in case". Bootable with the applications as they are now. So if needed I can go back. Had to do this already once for old files of Rezound that for one reason or another I cannot get working on series 3 kernels.

I'll download the livedvd and try it.
_________________
Expert in non-working solutions

NeddySeagoon · Posted: Sat Jan 24, 2015 12:34 pm Post subject:

Spanik,

Ah, thats different. Its often faster to reinstall than to update a very old system.
smartmontools, which provides smartctl is on System Rescue CD and that's a much smaller download.
_________________
Regards,

NeddySeagoon

Computer users fall into two groups:-
those that do backups
those that have never had a hard drive fail.

Spanik · Guru Joined: 12 Dec 2003 Posts: 457 Location: Belgium

I started the system and KDE has a disk utility that let you look at the smart info. Don't know how to interprete this and as I can't even open a browser anymore I'll just type those that are clearly errors over:

NeddySeagoon · Posted: Sat Jan 24, 2015 1:55 pm Post subject:

Spanik,

The values in VALUE WORST and THRESH are normalised. A parameter has failed if VALUE or WORST is less that or equal to THRESH.

RAW_VALUE is vendor specic. All the values are 32 bit numbers but there may be several bit packed raw vales in the same 32 bit number.
That means that large RAW_VALUE is not always a cause for concern.

ID 5 Reallocated_Sector_Ct is useful. It need not be zero.
ID 196 Reallocated_Event_Count is also useful. Again, it need not be zero.

There are no failures in your typed values.

Try replacing the SATA data cable
_________________
Regards,

NeddySeagoon

Computer users fall into two groups:-
those that do backups
those that have never had a hard drive fail.

Spanik · Guru Joined: 12 Dec 2003 Posts: 457 Location: Belgium

Ok, didn't knew that. Parameter 5 is not listed however 196 is:

[code]
196 Reallocation Count Normalized: N/A
Worst: N/A
Threshold: 0
Value: 0
[code]

There are quite a few parameters where Value = Threshold:
- Read Error Rate
- end-to-end -error
- Hardware ECC Recovered
- Soft ECC coreection
- Thermal Asperity Rate

Always both values are 0.

I'll power down and open the pc, check cables and re-seat the memory. And then boot with the live-dvd and try to see if a filesystem checks find anything.
_________________
Expert in non-working solutions

NeddySeagoon · Posted: Sat Jan 24, 2015 2:40 pm Post subject:

Spanik,

Don't do a fsck. It may well make things worse. Use smartctl to run the short test.
If that passes, try the long test.

It does not sound good if Value=Threshold but I've never seen the KDE output so I am not confident of what its telling.
Is the value the VALUE WORST THRESH value or the RAW_VALUE?
_________________
Regards,

NeddySeagoon

Computer users fall into two groups:-
those that do backups
those that have never had a hard drive fail.

Spanik · Guru Joined: 12 Dec 2003 Posts: 457 Location: Belgium

It is the raw value I suppose. The "worst" and "normalized" are given as N/A.

Download has finished. I'll see what that gives if smartctl is on it.
_________________
Expert in non-working solutions

Spanik · Guru Joined: 12 Dec 2003 Posts: 457 Location: Belgium

The LiveDVD doesn't include smartctl so I had to use the RescueCD. This gives this:

NeddySeagoon · Posted: Sun Jan 25, 2015 3:06 pm Post subject:

Spanik,

That looks OK. I suspect its a filesystem problem.

Spanik · Guru Joined: 12 Dec 2003 Posts: 457 Location: Belgium

OK, in that case I'm going to transfer everything needed and let a fsck run. Thanks for the help.
_________________
Expert in non-working solutions

Black · Posted: Wed Feb 04, 2015 2:04 pm Post subject:

If you have time, can you try booting with your old kernel and see if you get the same errors?

I was having issues recently after upgrading my kernel, but the drive works fine with an old kernel.