Proof that lack of backups causes disk failure
Ben Pfaff
blp@cs.stanford.edu
08 Oct 2001 17:16:52 -0700
As many of you know, I do nightly tape backups, and had for the
last 4 years or so. Or, at least, I did until I moved out here,
because my tape drive hasn't been shipped to me yet. I figured
that the chances of anything going wrong in the meantime were
small.
Ha ha--this afternoon I got a console message that there was an
unrecoverable error accessing /dev/hda1. Goddamn. I reboot
(can't even shut down cleanly because /dev/hda1 has been forcibly
remounted read-only), fsck -c, fsck -c again to make sure, fsck
-f a final time. So now I have a sh*tload of bad blocks.
Normal reaction: Back the disk the hell up, call up Compaq and
order a warranty replacement. But I'm going to wait until I
actually have a tape drive to do most of it. I'm still going to
do the "back the hell up" part by copying across the network to a
machine I admin at MSU, all the way across the country. Let's
see how this works out.
While I'm waiting, I'm monitoring the situation by adding a
nightly cron job that runs `badblocks' and compares to a known
list to make sure that new bad blocks aren't creeping in
slowly...
--
"Whoever you are -- SGI, SCO, HP, or even Microsoft -- most of the
smart people on the planet work somewhere else."
--Eric S. Raymond