Bug#211137: Install report - a bunch of problems
Package: installation-reports
INSTALL REPORT
Debian-installer-version: 09-15-2003 http://gluck.debian.org/cdimage/testing/netinst/i386/sarge-i386-netinst.iso
uname -a: 2.4.20-1-386
Date: 9-15-03 10:30
Method: How did you install? What did you boot off? If network
install, from where? Proxied?
Booting from CD. I used the "linux" method.
Machine: Thinkpad T30
Processor: 1.8GHz Pentium4
Memory: 768MB
Root Device: IDE
Root Size/partition table: Feel free to paste the full partition
table, with notes on which partitions are mounted where.
Output of lspci:
Base System Installation Checklist:
Initial boot worked: [O]
Configure network HW: [E]
Config network: [O]
Detect CD: [O]
Load installer modules: [O]
Detect hard drives: [O]
Partition hard drives: [E]
Create file systems: [O]
Mount partitions: [O]
Install base system: [E]
Install boot loader: [ ]
Reboot: [ ]
[O] = OK, [E] = Error (please elaborate below), [ ] = didn't try it
Comments/Problems:
When I try to autodetect the ethernet, it identifies as eepro100,
which is correct. However, the eepro100 wasn't found. It also
correctly detected the built-in wireless as orinoco_pci, also not
found.
I had to select e100 by hand to bring up the ethernet during the
install. After choosing e100, the next screen gives
"[!] Error running ethdetect"
An error or warning message was logged while running ethdetect
However, switching over to another console and looking at
/var/log/syslog and /var/log/messages, I didn't find anything that
made sense to cause that error. There were startup messages for e100,
but nothing out of the ordinary.
Next, after I loaded e100, I set up dhcp. This worked fine. However,
after dhcp came back and successfully configured, the installer put
the cursor on "Configure a network via PPP". This implies that
something still needs to be done to configure the network, but the
installer already has a working net connection. I think it should set
the cursor at "Choose mirror to install from".
I was able to get to "Detect HW (full)". This doesn't succeed because
the eepro100 and orinoco_pci modules aren't available. Also, I can't
get into partitioning without these modules.
To progress further, I linked e100 to eepro100 and orinoco_pci. This
makes the complaints about missing drivers shut up, despite errors
from modprobe -v while loading (since e100 is already loaded).
In "partition", I choose the only disk available (only one is
installed). The drive is fresh, with no partition map. When I hit
return, the screen first goes black. Looking on VC4, there's a
message from the installer: "Using default fdisk configuration".
Nothing happens until I hit return, when I see:
Partitioning error
Failed to partition the disc /dev/ide/host0/bus0/target0/lun0/disc
If I try cfdisk at the shell in VC2, I see that it says:
No partition table or unknown signature on partition table
Do you wish to start with a zero table [y/N]?
Thinking that perhaps I'm just not seeing the text, I switched back to
VC1, entered partitioning again, and typed a Y. This rewarded me with
blue and some black lines that seem like text should be there. So,
apparently, cfdisk is running, but not displaying correctly.
I ran cfdisk in the shell and partitioned the disk to my satisfaction,
then switched back to the installer.
Next I tried installing the base system. This appeared to go fine
until debootstrap exited with an error. In debootstrap.log, I find:
id: unknown group name: 0
[: 176: 0: unexpected operator
ln: /target/usr/bin/awk: File exists
On VC3, I see a bunch of dpkg errors like:
Err http://debootstrap.invalid sarge/main binutils 2.4.90.0.4-0.1
Could not resolve 'debootstrap.invalid'
for several packages.
Some more weirdness. On VC1, during the base install, I got:
info: Trying to enable linux framebuffer.
/lib/modules/2.4.20-1-386/kernel/drivers/video/vesafb.o: init_module:
No such device or address
And then, the main screen is showing up on VC4 with red background,
but still accepting input on VC1.
Reply to: