[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
I broke my system!
- To: misc_(_at_)_openbsd_(_dot_)_org
- Subject: I broke my system!
- From: "J. Joseph Max Katz" <jkatz_(_at_)_mac_(_dot_)_edu>
- Date: Tue, 24 Sep 1996 17:07:56 -0500 (CDT)
Hello! I'm back!
X has locked up the console a few times on my system, so I was
forced to resort to the reset button and corrupting my filesystem.
Consequently I have also had a lot of kernel panics about a bad filesystem
(dying on read/write/delete calls). I've been a good boy and running sync
and fsck on my own every so often, so nothing gets hurt too badly. (I'm
using a Beta I128 server from the XFree folks... possibly time to invest
in one from MetroLink...). The problem is I scrubbed /var/log/wtmp and
/var/run/utmp in the process. Maybe I have those locations reversed?
What happens is this: I can't see other logins except for any ftp
transfers or the 1st person to login on any console (including vt's).
(this involves all monitor-type programs-- finger, w, who, etc). If I
login on /dev/ttyC1 as jkatz and login as root a few seconds later on
/dev/ttyC0 and I do a "w" I only see jkatz. The same happens if Achilles
logs in on /dev/ttyp0 or /dev/ttyp2 and I've logged in on /dev/ttyC0
first. Also processes that start in /etc/rc and that are spawned by init
read that they started on the date Dec31 1969. Although I like to
reminiss, I don't think my box has been around that long (^: I had
something similar happen like this to my Linux box, so that's why I feel
utmp and wtmp are messed up.
On my system it currently reads like:
corinne# ls -al /var/run/*tmp
-rw-rw-r-- 1 root utmp 108 Sep 24 16:59 /var/run/utmp
corinne# ls -al /var/log/*tmp
lrwxr-xr-x 1 root wheel 13 Sep 21 11:49 /var/log/utmp -> /var/run/utmp
-rw-r--r-- 1 root wheel 1908 Sep 24 16:59 /var/log/wtmp
Are they in the right place?
Other problems :(
users cannot change their password using passwd :( They get "password
file is busy" message, but there is no /etc/ptmp or ptmp anywhere on the
constant kernel death-- once-twice a day on heavy fs read/writes. I'm
about to compile a new kernel to fix this and implement 16 partitions on
one disk ;-)
Any advice or words of sorrow?
Many thanks in advance.
-Jon (J. Joseph Max Katz) jkatz_(_at_)_mac_(_dot_)_edu |\/|=|===========================+
+=======================================|\/|=| MacMurray College Frosh * |
|OpenBSD Hacker (i386 and VAX) * MIS and CS double major * Tenor Sax |
|Player * MSA Rep (Theater Association) * 'Daily Other' lackee (Thursday |
|Edition Editor) * OFFICIAL Alpha Phi Omega pledge * lab assistant, Mac |
|Hall, Every Mon, Wed, and Fri: 3-5pm * Space Left Intentionally Blank |
Say, you want to talk? VOICE: +1 217.479.7308 (aka x308)
SNAIL: Box 4454, MacMurray College, Jacksonville, Illinois 62650 USA
NOTE: If I'm not slaving over OpenBSD on my PC, it's the VAX real-soon(tm).
Visit your host, monkey.org