How to avoid meltdowns in the future
I had a nasty morning this morning.
I had bought a cheap PCI serial card to try to diagnose and/or fix the
problem documented in:
http://www.trilug.org/pipermail/trilug/Week-of-Mon-20020617/008541.html
Somehow, when I installed it, something happened to the network card -
probably related to a hardware bulletin from IBM documenting a problem.
Here's the problem: when openAFS starts up and can't find a network card,
it hangs, constantly recycling the message "AFS: lost contact
with..." where ... is one of the afs database servers. That means I can't
boot with a broken network card.
So, I tried booting to the rescue disk, then doing:
linux single root=/dev/hda3
which looked promising until it constantly recycled the error about not
being able to create ksymoops because it's a read-only file system. I had
to hard-reset to get out of that.
The only spare network card I had around was a 3Com, for which the driver
wasn't in my kernel, so I couldn't toss it in.
What I ended up doing was booting the rescue disk all the way to the
beginning of the installation procedure, then using Alt-F2 to get a
virtual console. I could then mount my normal root partition and edit
/etc/init.d/openafs-client to insert an exit 0 at the beginning. A reboot
then gave me access to the system, enough to build and install the 3Com
driver.
This seems silly. Is there an easy/accessible way, preferably through
LILO, to boot into single-user mode with few, if any, daemons running?
Thanks.
----------------------------------------------------------------------
Andrew J Perrin - http://www.unc.edu/~aperrin
Assistant Professor of Sociology, U of North Carolina, Chapel Hill
clists@perrin.socsci.unc.edu * andrew_perrin (at) unc.edu
--
To UNSUBSCRIBE, email to debian-user-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Reply to: