Catastrophical Error!

igor92128 · Posted: Tue Mar 08, 2005 4:38 am Post subject: Catastrophical Error!

I've been using my headless Gentoo Box without many problems for the past month or so. I couldn't login today so I decided to hookup the monitor and see what was going on. Here's the output:

donjuan · l33t Joined: 11 May 2004 Posts: 760 Location: At Uni

CRC, Cyclic Redundancy Check. That means that a file is corrupt. In this case it's your kernel. Recompile and try again. Hopefully it's not hardware.
_________________
Command-line ACCEPT_KEYWORDS is considered harmful, use the package.* files.

The Stage 1 on 3 Install

igor92128 · Posted: Tue Mar 08, 2005 4:59 am Post subject:

But how could the kernel just become corrupt when I'm running the machine? It was turned on when this happened, but I didn't see what was on the screen because there isn't one.

Igor

Archangel1 · Veteran Joined: 21 Apr 2004 Posts: 1212 Location: Work

The kernel is read into RAM and doesn't leave. The kernel on your hdd hence isn't used and can be happily modified without you noticing anything until you try to reboot to it.
The hdd probably lost a bit somewhere. Best hope the whole drive's not going down.
_________________
What are you, stupid?

donjuan · l33t Joined: 11 May 2004 Posts: 760 Location: At Uni

Sorry, assumed that you were installing....kinda skipped the intro.

There's probably several reasons that this could have happened. The one that comes to mind is that your hard drive's dying, but I hope that's not the case. Something happened before or during the last reboot, but I'm not sure what.
_________________
Command-line ACCEPT_KEYWORDS is considered harmful, use the package.* files.

The Stage 1 on 3 Install

igor92128 · Posted: Tue Mar 08, 2005 5:18 am Post subject:

Ok, so I was about to boot off the LiveCD and I realized that I put the wrong CD in the tray. After the CD boot failed, everything booted up fine and there was no CRC error. Maybe I have a tempermental kernel and I should recompile it anyway?

Igor

donjuan · l33t Joined: 11 May 2004 Posts: 760 Location: At Uni

I'd say just perform a check of the file system and if that's fine then you should be OK. You could also try rebooting a few more times to see if the same thing happens. This could have just been a result of corrupted cache or memory, which I've seen happen several times after a power failure (not with these results though).
_________________
Command-line ACCEPT_KEYWORDS is considered harmful, use the package.* files.

The Stage 1 on 3 Install

igor92128 · Posted: Tue Mar 08, 2005 5:37 am Post subject:

Ok I found the problem. I've been running Memtest for a few minutes and there are already 89 errors, so one of the DIMMs must be bad. But I just got new memory recently so this was the last thing on my mind. This probably explains what Archangel said about the kernel being loaded into memory. I'll never get generic memory again!

Thanks,
Igor

donjuan · l33t Joined: 11 May 2004 Posts: 760 Location: At Uni

Glad you figured it out. Good thinking.
_________________
Command-line ACCEPT_KEYWORDS is considered harmful, use the package.* files.

The Stage 1 on 3 Install

gtr-xu1 · Posted: Tue Mar 08, 2005 10:16 am Post subject:

You running two or more different DIMMS? Make sure to always run memtest with only 1 DIMM at a time, becuase sometimes (especially generic) wont play together nicely. If you get no errors when testing DIMMS by themselves then just swap say bank 0's ram with bank 1's ram etc. Im a tech I see this quite often when we mix ram

Hope that helps