Gentoo!

I've run into another problem now on my install.

on attempting to bootstrap the system:

# cd /usr/portage
# scripts/bootstrap.sh

It seems to be searching for the files from the net, as i don't have this box on a net connection yet i'm getting a lot of host not found etc..

Seems strange to me as i'm installing using the universal CD so all the files should be on the harddisk.

Would appriciate if someone would get back to me on this one asap so i can leave it bootstrapping over night.

cheers
 
Originally posted by robmiller
Quick random bump; am I ok to delete things in the /var/tmp/portage directory? It's chock-full of stuff and it's wasting my precious hard drive space :p

Yes just issue rm -r * command in that directory, also remove all files under /usr/portage/distfiles by issuing the command rm *!

Using the -r switch allows you to remove directories containing data!!

Please actually go in to those directies before issuing those commands.

so actually make sure you are in:

cd /var/tmp/portage/

rm -r *

and

cd /usr/portage/distfiles/

rm *
 
Last edited:
Originally posted by englishpremier
I've run into another problem now on my install.

on attempting to bootstrap the system:

# cd /usr/portage
# scripts/bootstrap.sh

It seems to be searching for the files from the net, as i don't have this box on a net connection yet i'm getting a lot of host not found etc..

Seems strange to me as i'm installing using the universal CD so all the files should be on the harddisk.

Would appriciate if someone would get back to me on this one asap so i can leave it bootstrapping over night.

cheers
Not sure how to tell you this ... but, you'll need to start again :p Let me explain; Gentoo is a 'meta-distro', each time you install it - it is a little different to the last time you installed it. The reason for this is all the induvidual apps that make the OS are downloaded one by one, and at the latest deemed stable versions. This means that when you install gentoo it is always up to date (regardless of even which livecd to have). This is how gentoo was supposed to be. It's only recently that the gentoo livecd's have included pre-compiled binaries on the system and this may confuse the install procedure a little.

Now, it's pointless to build the system using from scratch (stage1) using the pre-compiled binaries from the livecd, (which you're trying to do since you don't want gentoo connected to the net) because what you'll end up with is not the latest version but the version date in which the binaries were compiled and placed onto the cd (whenever 2004.2 was released). In short what I'm saying is that this is in effect what the stage3 tarball is and thats what you should be using rather than the stage1 So right back at "5.c. Alternative: Using a Stage from the LiveCD" of the handbook you want to grab the stage3 tarball from the cd :)

edit: a stage3 is what you'll like want anyway, after reading yout other thread that you have a k6 processor, installing from stage3 will proberbly save you more than 24 hours of compiling (no joke)
 
Last edited:
oh well better to learn the hard was than not learn at all right?

so i'll do a complete reinstall and start from stage3, doesn't take that long to get where i am already and allows me to become familiar with the commands.

Basically is it the case that i should not use any stage other than 3 unless using an internet connection on the install pc?
 
Originally posted by englishpremier
Basically is it the case that i should not use any stage other than 3 unless using an internet connection on the install pc?
Yes, plainly because you cannot compile a stage1 or stage2 from the cd since the cd doesn't contain the sources for the apps. It does contain pre-compiled binaries for the apps but theres no point stating from stage1 or stage2 with these since you'll lose all the benefits from building gentoo from those stages i.e compile time flags.
 
edit: was trying to install something i didn't need, ignore that last txt.

now the real problem is, i couldn't install grub so tried lilo instead.

/etc/lilo.conf

boot=/dev/hda
prompt
timeout=50
default=gentoo
vga=792
image=boot/kernel-2.4.26-gentoo-r6
label=gentoo
read-only
root=/dev/hda3

now when i try /sbin/lilo i get the following errors:

/proc/partitions does not make /dev directory structre
name change '/dev/ide/host0/bus1/target0/lun0/disc' /dev/hdc

kernel compiled with DEVFS_FS, but devfs=mount was omitted
structure does not reflect DEVFS_FS device names
fatal /boot/kernel-2.4.26-gentoo-r6 does not exist "no such file etc.."


Right if this can't be solved easily i'm quiting and will try again next month. :(
 
Last edited:
Multiple possible reason or combinations of:

1. A recompile the kernel and add "Automatically mount at boot" for devfs (see code listing 8. That will proberbly stop the "kernel compiled with DEVFS_FS, but devfs=mount was omitted"

2. Like I said in one of my earlier posts I recommend you changed your cdrom and hard-drive around otherwise you need to change all references to hda to hdb. So in your lilo.conf all the hda should be hdb and same goes for your fstab in section 8a

3. Its possible that you didn't copy the kernel over to /boot as described in section 7c Code listing 14
 
I started again and have been very very careful in following the instructions, although they are not always clear.

everything is fine up untill the first reboot, where i get

NOT FOUND ANY ACTIVE PARTITION IN HDD
DISK BOOT FALIURE

the following line were placed in my nano -w /boot/grub/grub.conf

default 0
timeout 5
splashimage=(hd0,0)/grub/splash.xpm.gz
title=Gentoo Linux 2.6.7-r6
root (hd0,0)
kernel /kernel-2.6.7-gentoo-r11 root=/dev/hda3 vga=788

my hdd partitions/layout are exactly as described in the manual.

I ran the following after configuring the grub.conf
cp /proc/mounts /etc/mtab
grub-install --root-directory=/boot /dev/hda

# exit
# umount /mnt/gentoo/boot /mnt/gentoo/proc /mnt/gentoo
# reboot

However i accidentally typed exit twice and had to login again as root in order to unmount and reboot.

I notcied that on emerging grub it comes up with some warnings says that that some of the things are global dependencies. is this normal.

I feel i did ok upuntill this root boot, is there anyway i can rescue this?
 
Boot off the livecd and type
Code:
fdisk -l /dev/hdb
(Assuming your IDE drive is still on IDE1 slave. It should look something like this
Code:
Command (m for help): p

Disk /dev/hda: 30.0 GB, 30005821440 bytes
240 heads, 63 sectors/track, 3876 cylinders
Units = cylinders of 15120 * 512 = 7741440 bytes

Device Boot    Start       End    Blocks   Id  System
/dev/hda1 [color=deeppink]*[/color]        1        14    105808+  83  Linux
/dev/hda2         15        81    506520   82  Linux swap
/dev/hda3         82      3876  28690200   83  Linux
If you /boot (hda1 in theabove example) parition doesn't have a * then you have missed out "Section 4c - Code Listing 8" where it tells you
We need to make this partition bootable. Type a to toggle the bootable flag on this partition. If you press p again, you will notice that an * is placed in the "Boot" column.

Secondly you haven't said whether you moved your IDE drive to primary master or left it on primary slave. May I remind you that if you left it on primary slave you need to change all references of hda to hdb. That will also incude the hda's you have in grub.conf. You see by not having your drive as primary master you cannot follow the docs exactly. This is why I recommend you swap the drives over or be 100% sure to type hdb instead of hda whenever you see it.
 
ok it worked. silly me did miss the make partition bootable bit. lol.

it booted ok but there were two errors.

/etc/hostname somthing about selecting a valid hostmane

netmount was not started.

What is more concerning though is that i can't login to it. I tried 'root' as the username and the password that i crated in the setup but i still can't get in.
 
fixed the password thing by using the cd remounting and chrooting then changing password.

How do i check /etc/hostname and /etc/hosts files? i tried changing to that directory (cd /etc/hostname) but got the "no such file or directory" message. same goes for trying to view its contents with 'ls' command.
 
nope i deffinatly did these. I guess this does need to be configured being as i want this gentoo box to act as a router.

My concern at the moment has moved onto getting a gui installed. i know one isn't needed but i want one
.
after mounting the cdrom and running "ls /mnt/cdrom and trying "export PKGDIR=" /mnt/cdrom" the prompt changes to a simple >.

installing a gui from the pagakes cd is it just a case of following the gentoo handbook and typing "emerge --kde (or gnome) or whatever gui i want. Thinking of fluxbox
 
If you've configured them post up the output of
Code:
cat /etc/hosts
cat /etc/hostname
Code:
export PKGDIR=" /mnt/cdrom"
works fine, was that line copied and pasted straight from the handbook or is it the line you actually typed, you may have made a typo. Emerging a gui-wise would be a case of
Code:
emerge --usepkg fluxbox
It should automatically resolve all it's dependencies.
 
ok cat /etc/hosts comes up with

127.0.0.1 localhost
192.168.0.1 linix.7clarendon linux
192.168.0.100 one.7clarendon one

cat /etc/hostname

comes up with no such file or directory.

if i want to install fluxbox do i still need to run

# export PKGDIR=" /mnt/cdrom" or is it just for stuff off the cd (which fluxox isn't on)
 
You missed out "Section 8b - Code Listing 8". Give yourself a hostname. I take it by looking at your /etc/hosts that this machine in called "linux"? In which case type
Code:
echo linux > /etc/hostname

Ok installing, the "emerge" command has many options, usually you will install applications simply by typing "emerge" followed by the name of the app. Example
Code:
emerge fluxbox
This will download all sources from the internet and install them. However if you want to install apps from the cdrom rather than downloading the sources you will need to use the --usepkg switch. Now the --usepkg switch automatically looks for the variable $PKGDIR. So if you want to install from the CD you must type
Code:
export PKGDIR=" /mnt/cdrom"
before using the --usepkg switch. Since that environmental variable is not set as default. You can however add that line to the .bashrc file found in your home directory and that wil prevent you needing to type it in the future.
 
I messed up with X11.

how do i completly uninstall and then reinstall X11?

whenever i try 'emerge unmerge x11 (or X11)' it doesn't do anything (i mean it doesn't uninstall anything). to make matters worse i tried to do 'emerge --usepkg x11' over the top of the other installation, which it didn't like at all and came up with some errors.

so i want to completly uninstall X11, reinstall it and then install some sort of GUI a nice easy one to install and setup but not any bloatwear like KDE or Gnome.
 
why do you need to uninstall and reinstall X? When you first install it the chances are good that it will require configuring before it'll work. But for your reference X under portage is not called X11 it's "xfree". So you'll do
Code:
emerge --usepkg xfree
You can view a "pretend" emerge and see exactly what will be installed if you use the -p switch i.e
Code:
emerge -p --usepkg xfree
it's a good habit to use the -p before emerging things. I think you need to take a good read of the portage docs to learn how to install, uninstall and update.
 
Back
Top Bottom