Can bad cables cause file corruption on a disk? I'm assuming yes.
Can this happen with no messages in the message log?
I have a 6.1 box that has had files on the root partition get severly
corrupted. The first time I restored a few key files under /etc, and the
box was back in operation. A week later it happened again, and I'm still
not quite sure what got corrupted, but when booting, I get to the point
where init starts and
INIT: entering runlevel 3
INIT: cannot execute "/etc/rc.d/rc"
followed by a bunch of these
INIT: ld "1" respawning too fast: disabled for 5 minutes
INIT: ld "2" respawning too fast: disabled for 5 minutes
...
Even if I boot using 'linux single' it does this, including saying
runlevel 3.
Does anyone have any idea what might be corrupted here? /etc/rc.d/rc
exists and is executable. All the other scripts and directories under
/etc/rc.d look fine too. In fact I restored /etc from backups, and this
still happens.
If it turns out the disk is fine, I'd hate to have to rebuild this OS, so
it would be nice if I could fix this.
Now, what's most curious is there are no disk related errors in the
logs. The *only* errors, and few and far between, are a few
kernel: Unable to load Interpreter
and
dlerror: /some/lib.so:failed to map a segment from shared object: Cannot
allocate memory
So I'm very suspicious now. I have a seperate box that is getting the same
errors, and sometimes causes various daemons to die resulting in needing a
reboot, but fortunately no corruption.
How can I test this machine? I'll run memtest86 on it to check the RAM,
but how to I stress test the disks? I could compile the kernel ina loop
for a while. Anything else?
thanks
charles
_______________________________________________
Redhat-list mailing list
[EMAIL PROTECTED]
https://listman.redhat.com/mailman/listinfo/redhat-list
