Computer not working

[41mag]

New computer is not working.  I'm getting the initial bios screen (Press [DEL] to enter setup), but not getting any further.  Happened a few times before, but cycling power had fixed it.  Not working at all now.  Could this be a problem with the ram?

[Ryan McGuire]

Does pressing DEL actually get you into the BIOS? If so, then the motherboard/CPU at least is good.

[41mag]

Does pressing DEL actually get you into the BIOS? If so, then the motherboard/CPU at least is good.

No, pressing delete does not get me anywhere.  Wouldn't the data for the display have to go through the processor to get sent to the display though?

[Ryan McGuire]

Wouldn't the data for the display have to go through the processor to get sent to the display though?

Yes, but that doesn't rule out that its malfunctioning.

If it were the ram's fault instead of the motherboard/cpu then I would suspect that you would get some kind of beep code or message on the screen. Completely locking up like it is does not sound like a RAM issue to me.

[Caleb]

At any rate, there are any number of RAM diagnostic tools that work off of a bootable floppy, so you should be able to test the ram to see if it is causing the problem.

[aod318]

Do you have support with the manufacturer?  If it's a Dell, for example, they will have their own troubleshooting methodology and will walk you through testing.
I use http://www.softpedia.com/get/Tweak/Memory-Tweak/Microsoft-Windows-Memory-Diagnostic.shtml as memory diagnostic. 

[K. Darien Freeheart]

The complete lockups like that indicate hardware problems. If it's a pre-built system, I'd return it. If it's a home built system, I'd say swapping the mobo might make sense.

It's powering on, so your board isn't short circuiting. You didn't mentioned beeps, so I'm assuming it's not actually getting to POST which indicates either a hardware issue with the motherboard or a firmware issue with the motherboard too.

If it were a RAM mismatch or a misseated card, you'd get beeps. Faulty processors do happen, but they're very very rare and it usually results in your system not even powering up.

Finally, if it's a self-built system, there's another potential culrpit. The power supply. I've had a TON of computer issues that fix themselves with a good power supply.

[41mag]

The complete lockups like that indicate hardware problems. If it's a pre-built system, I'd return it. If it's a home built system, I'd say swapping the mobo might make sense.

It's powering on, so your board isn't short circuiting. You didn't mentioned beeps, so I'm assuming it's not actually getting to POST which indicates either a hardware issue with the motherboard or a firmware issue with the motherboard too.

If it were a RAM mismatch or a misseated card, you'd get beeps. Faulty processors do happen, but they're very very rare and it usually results in your system not even powering up.

Finally, if it's a self-built system, there's another potential culrpit. The power supply. I've had a TON of computer issues that fix themselves with a good power supply.

This is a system I built myself.  No beeps, so I'm thinking ram is ok (tested by pulling one at a time, and finally pulling all--did beep then).  New motherboard on the way.  Current one is 6 months old.  I'll see if I can get it swapped with the manufacturer and update my old computer.

[41mag]

New motherboard is in. 

Computer boots ok until it get to detecting the LVM group.  It doesn't find the LVM and gives me a bunch of errors.     LVM groups are ok (currently mounted using a Fedora Core live cd). 

[K. Darien Freeheart]

I had a Debian install that gave me tons of errors during boot about not finding volumes... And then proceeded right to the password prompt. :S

Perhaps GRUB is being stupid and has your drive ordering backwards?

[41mag]

Thanks.  I'll try unplugging the other dvd drives.

The only change to the system though was the motherboard.  Currently detected as sda using the live cd.   

Code Select Expand


[root@localhost ~]# /sbin/sfdisk -l /dev/sda

Disk /dev/sda: 60801 cylinders, 255 heads, 63 sectors/track
Units = cylinders of 8225280 bytes, blocks of 1024 bytes, counting from 0

   Device Boot Start     End   #cyls    #blocks   Id  System
/dev/sda1   *      0+     24      25-    200781   83  Linux
/dev/sda2         25   60800   60776  488183220   8e  Linux LVM
/dev/sda3          0       -       0          0    0  Empty
/dev/sda4          0       -       0          0    0  Empty
[root@localhost ~]# /sbin/pvscan
File descriptor 3 left open
File descriptor 4 left open
File descriptor 12 left open
File descriptor 13 left open
  PV /dev/sda2   VG VolGroup00   lvm2 [465.56 GB / 32.00 MB free]
  Total: 1 [465.56 GB] / in use: 1 [465.56 GB] / in no VG: 0 [0   ]
[root@localhost ~]# lvscan
File descriptor 3 left open
File descriptor 4 left open
File descriptor 12 left open
File descriptor 13 left open
  ACTIVE            '/dev/VolGroup00/LogVol00' [20.00 GB] inherit
  ACTIVE            '/dev/VolGroup00/LogVol02' [441.53 GB] inherit
  ACTIVE            '/dev/VolGroup00/LogVol01' [4.00 GB] inherit

[41mag]

Rebooted with just the one hard disk.  Still not finding the LVM.

Reading all physical volumes.  This may take a while...
No  volume groups found
Volume group "VolGroup00" not found

[K. Darien Freeheart]

Weird.

It looks like a pretty standard ext2/3 /boot partition (with MBR) and then VolGroup00 for /

Fedora is detecting it, but the booted system (with one drive removed) isn't catching it. The ONLY thing I can think is that your other drive contain the volume group data but that doesn't seem to be the case.

I feel as if I'm missing something obvious, but I'm missing it nonetheless.

[error]

The initrd probably needs to be rebuilt. This isn't too hard to do.

Boot the Fedora CD in Rescue mode. It'll offer to mount your system, which you should let it do. It gets mounted under /mnt/sysimage.

Then chroot into it, find your kernel, and rebuild the initrd, like so:

Code Select Expand

chroot /mnt/sysimage /bin/bash
rm -f /etc/lvm/backup/*        # Get rid of the old LVM backup data so it doesn't corrupt anything again
ls /boot/initrd-*

Find the latest numbered initrd which will be something like: initrd-2.6.26.5-45.fc9.x86_64.img
The underlined part is the kernel version, which you need for the next command:

Code Select Expand

mkinitrd -v -f /boot/initrd-2.6.26.5-45.fc9.x86_64.img 2.6.26.5-45.fc9.x86_64

If your latest kernel version is different, change that command to match it (in both parts).

Once this completes, do:

Code Select Expand

exit
reboot

[41mag]

That seems to fix the LVM problem.  Still didn't mount root, but did find the LVM.  Went back and rebuilt initrd again and looked closely at it:

Code Select Expand


error opening /sys/block/*/:  No such file or directory
resolveDevice:  device spec expected