Re: Spontaneous unclean reboots



On Sun, 21 Sep 2008 18:55:37 +0000, Curt wrote:

On 2008-09-21, Stefan Patric <not@xxxxxxxxxxxxxxx> wrote:

I would still fsck / just to make sure there are no errors, etc.
Thoroughness does have its advantages.

I'm going to do that from the Knoppix cd I have, although I'm not sure
how to do it? Do I chroot into the partition, then run fsck?

You don't have to chroot. Just boot into Knoppix and run fsck on the
appropriate unmounted partition.

Basically, I think the problem is Opera and not your system, X server
or otherwise. Your system ran okay until you started Opera, right?
Did you run a checksum on the Opera download before installing it?
Have you updated/upgraded your Etch install, so that it has the most
current libraries, kernel, etc?

I wish it was only Opera. The system has crashed three times today, or
maybe four, I'm losing count.

Okay. Definitely NOT Opera. It could either be hardware or OS.

At the console (without X running), I ran

memtest 128m --log (maybe a dumb move, as I only have 128mb of ram)

and the system rebooted uncleanly.

Try running memtest from Knoppix and not from your installed OS. Or
better still, go into BIOS and set it to do a memory scan before it loads
any part of the OS.


In X, I was running Firefox (Iceweasel) and the system rebooted
uncleanly.

More evidence of a hardware problem. Test the voltages on your power
supply leads to make sure they are outputting the proper voltages.

I did some work on the machine--cleaned all the dust out of it, took out
the winmodem that I don't use, went into the bios and disabled the
parallel port and 1 serial port that I don't use, reseated nvidia
geforce graphics card and jiggled wires.

Then I booted the machine, started X, played some bzflag, ate dinner,
and when I came back I saw the login prompt, but this time the system
had rebooted *cleanly*, while I was away eating. At 20:14. Strange.

It shouldn't have reboot at all, cleanly or otherwise.

[snip]

Problems like this sometimes take weeks or months to solve and rectify.
Just eliminate each possible problem one by one.

So far, I don't seem to be making any progress and things are
deteriorating, as in "getting worse."

If your RAM checks OK as well as the power supply, consider reinstalling
the OS on just / and not /home. Corrupted files, which won't show up on
a fsck, could be causing the reboots.

Have you check to see if the kernel was or is damaged. Or just install
one whose checksum is good. Of course, you could have corrupted
libraries, etc. or your hard drive could be slowly failing, which is a
good possibility, if the hard drive like the system is 7 years old.

It's the same stock debian kernel 2.6.18-4-686. I don't know why I got
a kernel panic. I don't know how I'd check for corrupted libraries,
either. smartctl finds nothing wrong with my hard drive.

Only way I know of the check for "bad" files is to compare them with
known good files by checking file size, checksum both, binary compare.

Well, assume the HD is okay, and reinstall the OS, then you know all your
files are good. Checksum the install CD to make sure it is okay before
installing. Worse that will happen is the same thing.

I had a similar problem with my old system, except it didn't reboot, but
just turned off. After months of trying things, discovered the CPU was
bad. Replaced it and haven't had any problems since.

Stef
.



Relevant Pages

  • Re: 4.10 wont boot
    ... >> I installed 4.10 and finished installing software then rebuilt the ... Reboot and I its complaining of a ... >> dirty file system and needs to manulally run fsck. ... The system was shutdown a gracefull as reboot command can be. ...
    (comp.unix.bsd.freebsd.misc)
  • Re: [Bug 14354] Re: ext4 increased intolerance to unclean shutdown?
    ... I went through this bug report and here are some immediately ... the power button on my laptop when it is doing nothing. ... by another reboot only to have X fail to start and NetWorkManager ... After another fsck and a reboot it finally comes up. ...
    (Linux-Kernel)
  • Re: The ext3 way of journalling
    ... If these reasons are good ones then the correct way to really handle this would be to do regular background scrubbing during runtime; ideally with metadata checksums so that you can actually detect all corruption. ... Leave some extra space in your LVM volume group (enough for 1 snapshot over the time it takes to do an FSCK). ... If it succeeds you can adjust the "last checked" date to the time when the snapshot was taken and if it fails you can schedule an FSCK at next reboot. ...
    (Linux-Kernel)
  • Help! Corrupted filesystem
    ... I'm running an IBM R40 laptop with three systems: Debian Sarge, Knopix 3.4 hd install, and WinXP. ... I unplugged it and completed the boot, doing an fsck because it hadn't been done in a while. ... I rebooted to Knoppix to try to back up the data, but K3b won't burn a CD, indicating that it seems to be a buffer underrun. ...
    (Debian-User)
  • Re: Printer not working after a short circuit
    ... > system will reboot when you leave the shell. ... Give root password for maintenance or type Control-D to ... Run fsck manually. ... > entire file system and prompted me as to whether or not to continue. ...
    (microsoft.public.windowsxp.hardware)