Hello all:
My company builds Linux based web content filters using CentOS 4.3 as the base. We always apply updates to the new machines from the default repositories before deployment. Recently, the last two CentOS machines we built on Dell Dimension 1100s have failed, and the root file system seem to be getting corrupted. An e2fsck tries to recover the journal, but never repairs the machine enough to completely boot up. This only seems to be happening on the Dimension 1100s, as we have several other machines built with the same kernel version and software running on it that have not failed. The machines are Celeron-Ds, 2.4-2.8 Ghz, with either 256 or 512 MB of RAM and IDE hard disks.
Does anyone know of any issues which could be causing this? Any fixes?
On Thu, 27 Jul 2006, Russell Butturini wrote:
Hello all:
My company builds Linux based web content filters using CentOS 4.3 as the base. We always apply updates to the new machines from the default repositories before deployment. Recently, the last two CentOS machines we built on Dell Dimension 1100s have failed, and the root file system seem to be getting corrupted. An e2fsck tries to recover the journal, but never repairs the machine enough to completely boot up. This only seems to be happening on the Dimension 1100s, as we have several other machines built with the same kernel version and software running on it that have not failed. The machines are Celeron-Ds, 2.4-2.8 Ghz, with either 256 or 512 MB of RAM and IDE hard disks.
We've got a handul of 1100s (actually, the 1100n model, mostly because I love getting a printed copy of the GPL from Dell :-), with CentOS 4.3. No troubles to report.
Can you get to a recovery shell? If so, does 'smartctl -a /dev/hda' provide any clues?
-- Paul Heinlein heinlein@madboa.com
-
Paul Heinlein -
Russell Butturini