1
0
mirror of synced 2026-01-23 10:38:15 +00:00
xen-tools.xen-tools/bin/xen-create-image
steve 4890ba9c97 2006-08-18 13:16:55 by steve
Use hostname when installing system.
2006-08-18 13:16:55 +00:00

2046 lines
46 KiB
Perl
Executable File

#!/usr/bin/perl -w
=head1 NAME
xen-create-image - Easily create new Xen instances with networking and OpenSSH.
=cut
=head1 SYNOPSIS
Help Options:
--help Show the help information for this script.
--manual Read the manual, and examples, for this script.
--verbose Show useful debugging information.
--version Show the version number and exit.
Size / General options:
--accounts Copy all non-system accounts to the guest image
--boot Boot the new instance after creating it.
--cache Cache .deb files on the host when installing the new guest
with the debootstrap tool.
--force Force overwriting existing images.
This will remove existing images or LVM volumes which match
those which are liable to be used by the new invocation.
--fs Specify the filesystem type to use for the new guest.
Valid choices are 'ext3', 'reiserfs', or 'xfs'.
--image Specify whether to create "sparse" or "full" disk images.
Full images are mandatory when using LVM, so this setting
is ignored in that case.
--initrd Specify the initial ramdisk
If an image is specified it must exist.
--kernel Set the path to the kernel to use for domU.
This image must exist on the host system.
--memory Setup the amount of memory allocated to the new instance.
--passwd Ask for a root password during setup.
This is done interactively.
--role Run a specific role command post-install.
These role scripts are discussed in the manpage later.
--size Set the size of the primary disk image.
--swap Set the size of the swap partition.
--noswap Do not create a swap partition.
When this option is used the system will not have a swap
entry added to its /etc/fstab file either.
--ide Use IDE names for virtual devices (hda not sda)
Installation options:
--arch Pass the given architecture to debootstrap or rpm strap.
This argument is ignored if you install with a different
installation type.
--copy Install a new image by copying a directory recursively,
this is assumed to be a previously installed system.
This is much faster than installing from scratch.
--dist Specify the distribution you wish to install.
--debootstrap Use debootstrap to install the guest system.
--dist should be used to specify a distribution debootstrap
understands.
--mirror Setup the mirror to use when installing via debootstrap.
--rpmstrap Use rpmstrap to install the guest distribution.
--dist should be used to specify a distribution rpmstrap
understands.
--tar Install the new image by untarring the specified file.
Similar to --copy this is signficantly faster than
installing via rpmstrap or debootstrap.
--template Specify which template file to use when creating the
Xen configuration file.
Networking options:
--dhcp The guest will fetch its networking details via DHCP.
--gateway Setup the network gateway for the new instance.
--ip Setup the IP address of the machine, multiple IPs
are allowed. When specifying more than one IP the
first one is setup as the "system" IP, and the additional
ones are added as aliases.
Note that Xen 3.x supports a maximum of three IP addresses
per guest.
This option conflicts with --dhcp.
--netmask Setup the netmask for the new instance.
Mandatory options:
--dir Specify where the output images should go.
Subdirectories will be created for each guest
If you do not wish to use loopback images specify --lvm.
(These two options are mutually exclusive.)
--lvm Specify the volume group to save images within.
If you do not wish to use LVM specify --dir.
(These two options are mutually exclusive.)
--hostname Set the hostname of the new guest system.
Ideally this will be fully-qualified since several
of the hook scripts will expect to be able to parse
a domain name out of it for various purposes.
=cut
=head1 NOTES
This script is a simple wrapper around three external tools
=over 8
=item B<xt-install-image>
Install a new distribution.
=item B<xt-customize-image>
Run a collection of hook scripts to customise the freshly installed system.
=item B<xt-create-xen-config>
Create a configuration file in /etc/xen so that xm can create the new image.
=back
The result of invoking these three scripts, and some minor glue between
them is a simple means of creating new Xen guest domains.
=cut
=head1 DESCRIPTION
xen-create-image is a simple script which allows you to create new
Xen instances easily. The new image will be given two volumes. These
volumes will be stored upon the host as either loopback files, or
LVM logical volumes:
1. An image for the systems root disk.
2. An image for the systems swap device.
The new virtual installations will be configured with networking,
have OpenSSH installed upon it, and have most of its basic files
setup correctly.
=cut
=head1 CONFIGURATION
To reduce the length of the command line each of the supported options
may be specified inside a configuration file.
The global configuration file read for options is:
/etc/xen-tools/xen-tools.conf
The configuration file may contain comments which begin with the
hash '#' character. Otherwise the format is 'key = value'.
A sample configuration file would look like this:
=for example begin
#
# Output directory. Images are stored beneath this directory, one
# subdirectory per hostname.
#
dir = /home/xen
#
# LVM users should disable the 'dir' setting above, and instead
# specify the name of the volume group to use.
#
# lvm = myvolume
#
# Disk and Sizing options.
#
size = 2Gb # Disk image size.
image = full # Allocate the full disk size immediately.
memory = 128Mb # Memory size
swap = 128Mb # Swap size
fs = ext3 # use EXT3 filesystems
dist = sarge # Default distribution to install.
#
# Kernel options.
#
initrd = /boot/initrd.img-2.6.16-2-xen-686
kernel = /boot/vmlinuz-2.6.16-2-xen-686
#
# Networking options.
#
gateway = 192.168.1.1
netmask = 255.255.255.0
#
# Installation options.
#
# copy = /path/to/pristine/image
debootstrap = 1
# rpmstrap = 1
# tar = /path/to/img.tar
=for example end
Using this configuration file a new image may be created with the
following command:
xen-create-image --hostname=vm03.my.flat --ip=192.168.1.201
This makes use of loopback images stored beneath /home/xen and
will be installed via the debootstrap command.
=cut
=head1 XEN CONFIGURATION FILE
Once a new image has been created an appropriate configuration file
for Xen will be saved in the directory /etc/xen.
The configuration file is built up using the template file
/etc/xen-tools/xm.tmpl - which is a file processed via
the Text::Template perl module.
If you wish to modify the files which are generated please make your
changes to that input file.
Alternatively you can create multiple configuration files and
specify the one to use with the --template option.
=cut
=head1 LOOPBACK EXAMPLES
The following will create a 2Gb disk image, along with a 128Mb
swap file with Debian Sarge setup and running via DHCP.
xen-create-image --size=2Gb --swap=128Mb --dhcp \
--dir=/home/xen --hostname=vm01.my.flat
This next example sets up a host which has the name 'vm02.my.flat' and
IP address 192.168.1.200, with the gateway address of 192.168.1.1
xen-create-image --size=2Gb --swap=128Mb \
--ip=192.168.1.200 \
--netmask=255.255.255.0
--gateway=192.168.1.1 \
--dir=/home/xen --hostname=vm02.my.flat
The directory specified for the output will be used to store the volumes
which are produced. To avoid clutter each host will have its images
stored beneath the specified directory, named after the hostname.
For example the images created above will be stored as:
$dir/domains/vm01.my.flat/
$dir/domains/vm01.my.flat/disk.img
$dir/domains/vm01.my.flat/swap.img
$dir/domains/vm02.my.flat/
$dir/domains/vm02.my.flat/disk.img
$dir/domains/vm02.my.flat/swap.img
The '/domains/' subdirectory will be created if necessary.
=cut
=head1 LVM EXAMPLES
If you wish to use an LVM volume group instead of a pair of loopback
images as shown above you can instead use the --lvm argument to
specify one.
xen-create-image --size=2Gb --swap=128Mb --dhcp \
--lvm=myvolumegroup --hostname=vm01.my.flat
The given volume group will have two new logical volumes created within it:
${hostname}-swap
${hostname}-disk
The disk image may be mounted, as you would expect, with the following
command:
mkdir -p /mnt/foo
mount /dev/myvolumegroup/vm01.my.flat-disk /mnt/foo
=cut
=head1 INSTALLATION METHODS
The new guest images may be installed in several different ways:
1. Using the debootstrap command, which must be installed and present.
2. Using the rpmstrap command, which must be installed and present.
3. By copying an existing installation.
4. By untarring a file containing a previous installation.
These different methods can be selected by either the command line
arguments, or settings in the configuration file. Only one installation
method may be specified at a time; they are mutually-exclusive.
=cut
=head1 INSTALLATION SPEEDUPS
After performing your first installation you can customize it, or
use it untouched, as a new installation source. By doing this you'll
achieve a significant speedup, even above using the debootstrap caching
support.
There are two different ways you can use the initial image as source
for a new image:
1. By tarring it up and using the tar-file as an installation source.
2. By mounting the disk image of the first system and doing a literal copy.
Tarring up a pristine, or customised, image will allow you to install
with a command such as:
xen-create-image --size=2Gb --swap=128Mb --dhcp \
--lvm=myvolumegroup --hostname=vm01.my.flat \
--tar=/path/to/tar.file.tar
The advantage of the tarfile approach is that you'll not need to
keep a disk image mounted if you were to use the --copy argument
to create a new image using the old one as source:
xen-create-image --size=2Gb --swap=128Mb --dhcp \
--lvm=myvolumegroup --hostname=vm01.my.flat \
--copy=/path/to/copy/from
=cut
=head1 DEBOOTSTRAP CACHING
When installing new systems with the debootstrap tool there is
a fair amount of network overhead.
To minimize this the .deb files which are downloaded into the
new instance are cached by default upon the host, in the directory
/var/cache/apt/archives.
When a new image is created these packages are copied into the new
image - before the debootstrap process runs - this should help avoid
expensive network reading.
If you wish to clean the cache upon the host you may do so with
apt-get, as you'd expect:
apt-get clean
(This feature can be disabled with the command line flag --cache=no,
or by the matching setting in the configuration file.)
=cut
=head1 ROLES
Each supported distribution has a hook directory which has been
described already. If that directory contains the subdirectory 'role.d'
then roles support is available.
Currently there are some roles scripts included for the Debian
distribution, these are intended primarily as examples:
=over 8
=item builder
Setup the new virtual images with commonly used packages for rebuilding Debian packages from their source.
=item gdm
Install an X11 server, using VNC and GDM
=item minimal
Customise the generated images to remove some packages.
=item xdm
Install an X11 server, using VNC and XDM
=back
If you'd like to include your own role scripts you'll need to
create $dist.d/role.d/foo - then specify "--role=foo" when you're
creating your new instance.
=cut
=head1 THE SKELETON DIRECTORY
Any files present in the directory /etc/xen-tools/skel will be copied
across to each new guest image. The role of this directory is analogous
to the /etc/skel directory.
A typical use for this would be to copy a public key across to each
new system. You could do this by running:
=for example start
mkdir -p /etc/xen-tools/skel/root/.ssh
chmod -R 700 /etc/xen-tools/skel/root
cp /root/.ssh/id_rsa.pub /etc/xen-tools/skel/root/.ssh/authorized_keys2
chmod 644 /etc/xen-tools/skel/root/.ssh/authorized_keys2
=for example cut
=head1 AUTHOR
Steve
--
http://www.steve.org.uk/
$Id: xen-create-image,v 1.80 2006-08-18 13:16:55 steve Exp $
=cut
=head1 LICENSE
Copyright (c) 2005-2006 by Steve Kemp. All rights reserved.
This module is free software;
you can redistribute it and/or modify it under
the same terms as Perl itself.
The LICENSE file contains the full text of the license.
=cut
use strict;
use English;
use Env;
use File::Temp qw/ tempdir /;
use Getopt::Long;
use Pod::Usage;
#
# Configuration values read initially from the global configuration
# file, then optionally overridden by the command line.
#
my %CONFIG;
#
# Global variable containing the temporary file where our image
# is mounted for installation purposes.
#
# Why is this here?
#
# Well it makes sure that the magic "END" section can unmount it
# if there are errors.
#
#
my $MOUNT_POINT = undef;
#
# Release number.
#
my $RELEASE = '2.3';
#
# Setup default options.
#
setupDefaultOptions();
#
# Read the global configuration file if it exists.
#
if ( -e "/etc/xen-tools/xen-tools.conf" )
{
readConfigurationFile( "/etc/xen-tools/xen-tools.conf" );
}
#
# Parse the command line arguments.
#
parseCommandLineArguments();
#
# Check the environment - after parsing arguments.
#
# This is required so that the "--help" flag will work even if our support
# scripts are not installed.
#
checkSystem();
#
# Ensure we're started by root at this point. This is required
# to make sure we can create new LVM volumes, or mount loopback images.
#
testRootUser();
#
# Check our arguments
#
checkArguments();
#
# Make sure we have a log directory
#
mkdir( "/var/log/xen-tools", 0777 ) if ( ! -d "/var/log/xen-tools" );
#
# Trash any existing logfile.
open( TRASH, ">", "/var/log/xen-tools/$CONFIG{'hostname'}.log" );
print TRASH "";
close(TRASH);
#
# Check we have installed binaries.
#
checkFilesPresent();
#
# Show a summery of what we're going to do.
#
showSummery();
#
# Create and format the images if we're using loopback filesystems.
#
if ( $CONFIG{'dir'} )
{
testLoopbackModule();
createLoopbackImages();
}
elsif ( $CONFIG{'lvm'} )
{
#
# Create our LVM partitions.
#
createLVMBits();
}
else
{
# Can't happen: Neither loopback nor LVM selected
logprint( "Error: Neither loopback nor LVM is selected.\n");
exit;
}
#
# Mount the image.
#
mountImage();
#
# Call xt-install-image to do the install.
#
installSystem();
#
# Make sure the installation succeeded.
#
if ( ! -x $MOUNT_POINT . "/bin/ls" )
{
logprint( "System installation failed. Aborting\n");
exit;
}
#
# Call xt-customise-image to setup networking, and run distro-specific hooks.
#
runCustomisationHooks();
#
# Run the user specified role script, if any.
#
runRoleScript();
#
# Create the Xen configuration file.
#
runXenConfigCreation();
#
# Setup the password if the user wanted that.
#
setupRootPassword() if ( $CONFIG{'passwd'} );
#
# Report success.
#
logprint( "All done\n");
#
# If we're supposed to start the new instance do so - note here we
# have to unmount the image first.
#
if ( $CONFIG{'boot'} )
{
#
# Unmount.
#
runCommand( "umount $MOUNT_POINT" );
$MOUNT_POINT = undef;
#
# Start the image
#
my $pid = fork();
if ( $pid )
{
# Parent.
exit;
}
else
{
# Child.
system( "$CONFIG{'xm'} create $CONFIG{'hostname'}.cfg >/dev/null 2>/dev/null" );
}
}
#
# Finished.
#
exit;
=begin doc
Test that this system is fully setup for the new xen-create-image
script.
This means that the the companion scripts xt-* are present on the
host.
=end doc
=cut
sub checkSystem
{
my @required = qw ( /xt-customize-image xt-install-image xt-create-xen-config / );
foreach my $bin ( @required )
{
if ( ! -x "/usr/bin/" . $bin )
{
logprint("The script '$bin' was not found.\n");
logprint( "Aborting\n\n" );
exit;
}
}
}
=begin doc
Setup the default options we'd expect into our global CONFIG hash.
=end doc
=cut
sub setupDefaultOptions
{
#
# Paths and files.
#
$CONFIG{'dir'} = '';
$CONFIG{'xm'} = '/usr/sbin/xm';
$CONFIG{'kernel'} = '/boot/vmlinuz-2.6.16-1-xen-686';
$CONFIG{'initrd'} = '';
#
# Sizing options.
#
$CONFIG{'memory'} = '96Mb';
$CONFIG{'size'} = '2000Mb';
$CONFIG{'swap'} = '128M';
$CONFIG{'noswap'} = 0;
$CONFIG{'cache'} = 'yes';
$CONFIG{'image'} = 'sparse';
#
# Misc. options.
#
$CONFIG{'mirror'} = 'http://ftp.us.debian.org/debian';
$CONFIG{'arch'} = '';
$CONFIG{'dist'} = 'sarge';
$CONFIG{'fs'} = 'ext3';
$CONFIG{'force'} = 0;
$CONFIG{'template'} = '';
#
# Installation methods
#
$CONFIG{'rpmstrap'} = 0;
$CONFIG{'debootstrap'} = 0;
$CONFIG{'copy'} = '';
$CONFIG{'tar'} = '';
#
# The program to run to create a filesystem.
#
# NOTE: These commands end in a trailing slash. The last parameter is
# added as the loopback file/LVM volume to create the fs on....
#
# NOTE 2: Each of these scripts will "force" the creation of a new
# filesystem, even if it exists. This script must detect
# prior existance itself.
#
$CONFIG{'make_fs_ext3'} = '/sbin/mkfs.ext3 -F ';
$CONFIG{'make_fs_xfs'} = '/sbin/mkfs.xfs -f -d name=';
$CONFIG{'make_fs_reiserfs'} = '/sbin/mkfs.reiserfs -f -q ';
#
# Flags to pass to "mount" to mount our image.
#
# NOTE: Kinda redundent and may go away since '-t auto' should do
# the right thing.
#
$CONFIG{'mount_fs_ext3'} = '-t ext3';
$CONFIG{'mount_fs_xfs'} = '-t xfs';
$CONFIG{'mount_fs_reiserfs'} = '-t reiserfs';
}
=begin doc
read the global configuration file /etc/xen-tools/xen-tools.conf
=end doc
=cut
sub readConfigurationFile
{
my ($file) = ( @_ );
my $line = "";
open( FILE, "<", $file ) or die "Cannot read file '$file' - $!";
while (defined($line = <FILE>) )
{
chomp $line;
if ($line =~ s/\\$//)
{
$line .= <FILE>;
redo unless eof(FILE);
}
# Skip lines beginning with comments
next if ( $line =~ /^([ \t]*)\#/ );
# Skip blank lines
next if ( length( $line ) < 1 );
# Strip trailing comments.
if ( $line =~ /(.*)\#(.*)/ )
{
$line = $1;
}
# Find variable settings
if ( $line =~ /([^=]+)=([^\n]+)/ )
{
my $key = $1;
my $val = $2;
# Strip leading and trailing whitespace.
$key =~ s/^\s+//;
$key =~ s/\s+$//;
$val =~ s/^\s+//;
$val =~ s/\s+$//;
# Store value.
$CONFIG{ $key } = $val;
}
}
close( FILE );
}
=begin doc
Parse the command line arguments this script was given.
=end doc
=cut
sub parseCommandLineArguments
{
my $HELP = 0;
my $MANUAL = 0;
my $VERSION = 0;
#
# We record the installation method here because we want
# to ensure that we allow the method supplied upon the command line
# to overwrite the one we might have ready read from the configuration
# file.
#
my %install;
$install{'debootstrap'} = 0;
$install{'rpmstrap'} = 0;
$install{'copy'} = undef;
$install{'tar'} = undef;
$install{'dir'} = undef;
$install{'lvm'} = undef;
#
# Parse options.
#
GetOptions(
# Mandatory
"dist=s", \$CONFIG{'dist'},
# Size options.
"size=s", \$CONFIG{'size'},
"swap=s", \$CONFIG{'swap'},
"noswap", \$CONFIG{'noswap'},
"image=s", \$CONFIG{'image'},
"memory=s", \$CONFIG{'memory'},
# Locations
"dir=s", \$install{'dir'},
"kernel=s", \$CONFIG{'kernel'},
"initrd=s", \$CONFIG{'initrd'},
"mirror=s", \$CONFIG{'mirror'},
"lvm=s", \$install{'lvm'},
# Networking options
"dhcp", \$CONFIG{'dhcp'},
"gateway=s", \$CONFIG{'gateway'},
"hostname=s", \$CONFIG{'hostname'},
"ip=s@", \$CONFIG{'ip'},
"netmask=s", \$CONFIG{'netmask'},
"p2p=s", \$CONFIG{'p2p'},
# Exclusive
#
# NOTE: We set the local variable here, not the global.
#
"copy=s", \$install{'copy'},
"debootstrap", \$install{'debootstrap'},
"rpmstrap", \$install{'rpmstrap'},
"tar=s", \$install{'tar'},
# Misc. options
"accounts", \$CONFIG{'accounts'},
"arch=s", \$CONFIG{'arch'},
"fs=s", \$CONFIG{'fs'},
"boot", \$CONFIG{'boot'},
"cache=s", \$CONFIG{'cache'},
"ide", \$CONFIG{'ide'},
"passwd", \$CONFIG{'passwd'},
"role=s", \$CONFIG{'role'},
"force", \$CONFIG{'force'},
"template=s", \$CONFIG{'template'},
# Help options
"debug", \$CONFIG{'verbose'},
"help", \$HELP,
"manual", \$MANUAL,
"verbose", \$CONFIG{'verbose'},
"version", \$VERSION
);
pod2usage(1) if $HELP;
pod2usage(-verbose => 2 ) if $MANUAL;
if ( $VERSION )
{
my $REVISION = '$Revision: 1.80 $';
if ( $REVISION =~ /1.([0-9.]+) / )
{
$REVISION = $1;
}
logprint( "xen-create-image release $RELEASE - CVS: $REVISION\n" );
exit;
}
#
# If we have had one of the local installation methods specified,
# and *only* one of them the we'll reset the global option(s) which
# came from the configuration file.
#
# Count the number which wer supplied
#
my $count = 0;
foreach my $key ( qw/debootstrap rpmstrap copy tar/ )
{
if ( $install{$key} )
{
$count += 1;
}
}
#
# If we had exactly one specified then we can proceed.
#
if ( $count == 1 )
{
foreach my $key ( qw/debootstrap rpmstrap copy tar/ )
{
$CONFIG{$key} = $install{$key};
}
}
elsif ( $count > 1 )
{
my $err =<<E_O_ERROR;
Please specify an installation method using only *one* of the following
command line options:
--debootstrap = Install the system using the deboostrap command.
--rpmstrap = Install the system using the rpmstrap command.
--copy /path = Install by copying the specified directory recursively.
--untar file.tar = Install by untarring the given file.
(You could specify one of these options as the default in the configuration
file /etc/xen-tools/xen-tools.conf.)
E_O_ERROR
logprint( $err );
}
else
{
# count == 0;
#OK.
}
#
# Now do a similar thing so that the command line setting of
# '--lvm=x' and '--dir=x' override anything in the configuration
# file.
#
if ( $install{'dir'} )
{
$CONFIG{'lvm'} = undef;
$CONFIG{'dir'} = $install{'dir'};
}
if ( $install{'lvm'} )
{
$CONFIG{'dir'} = undef;
$CONFIG{'lvm'} = $install{'lvm'};
}
}
=begin doc
Make sure this script is being run by a user with UID 0.
=end doc
=cut
sub testRootUser
{
if ( $EFFECTIVE_USER_ID != 0 )
{
my $err =<<E_O_ROOT;
In order to use this script you must be running with root privileges.
(This is necessary to mount the disk images which are created.)
E_O_ROOT
logprint( $err );
exit;
}
}
=begin doc
Test that the command line arguments we were given make sense.
=end doc
=cut
sub checkArguments
{
#
# We require a distribution name.
#
if ( ! defined( $CONFIG{'dist'} ) )
{
logprint( "The '--dist' argument is mandatory\n" );
exit 1;
}
#
# We require a hostname.
#
if ( ! defined( $CONFIG{'hostname'} ) )
{
logprint( "The '--hostname' argument is mandatory.\n" );
exit 1;
}
#
# FAKE:
#
if ( $CONFIG{'dist'} eq 'fedora-core4' )
{
$CONFIG{'dist'} = 'stentz';
}
#
#
# Test that the distribution name we've been given
# to configure has a collection of hook scripts.
#
# If there are no scripts then we clearly cannot
# customise it!
#
my $dir = "/usr/lib/xen-tools/" . $CONFIG{'dist'} . ".d";
if ( ! -d $dir )
{
my $err =<<E_OR;
We're trying to configure an installation of $CONFIG{'dist'} in
$CONFIG{'dir'} - but there is no hook directory for us to use.
This means we don't know how to configure this installation.
We'd expect the hook directory to be : $dir
Aborting.
E_OR
logprint( $err );
exit 1;
}
#
# Image must be 'sparse' or 'full'.
#
if ( defined( $CONFIG{'image'} ) )
{
if ( ( $CONFIG{'image'} ne "sparse" ) &&
( $CONFIG{'image'} ne "full" ) )
{
logprint( "Image type must be 'sparse' or 'full'\n" );
exit;
}
}
#
# If using LVM images cannot be sparse
#
if ( $CONFIG{'lvm'} )
{
$CONFIG{'image'} = "full";
logprint( "Switching to 'full' images rather than 'sparse', since you're using LVM\n" );
}
#
# The kernel + initrd images should exist.
#
if ( defined( $CONFIG{'kernel'} ) &&
length( $CONFIG{'kernel'} ) &&
! -e $CONFIG{'kernel'} )
{
logprint( "The kernel image we're trying to use does not exist.\n" );
logprint( "The image is - $CONFIG{'kernel'}\n" );
logprint( "Aborting\n" );
exit;
}
if ( defined( $CONFIG{'initrd'} ) &&
length( $CONFIG{'initrd'} ) &&
! -e $CONFIG{'initrd'} )
{
logprint( "The initial ramdisk we're trying to use does not exist.\n" );
logprint( "The image is - $CONFIG{'initrd'}\n" );
logprint( "Aborting\n" );
exit;
}
#
# Make sure we have one, and only one, installation method.
#
my $count = 0;
foreach my $key ( qw/debootstrap rpmstrap copy tar/ )
{
if ( $CONFIG{$key} )
{
$count += 1;
}
}
if ( $count > 1 )
{
my $err =<<E_O_ERROR;
Please specify an installation method using one of the following
command line options:
--debootstrap = Install the system using the deboostrap command.
--rpmstrap = Install the system using the rpmstrap command.
--copy /path = Install by copying the specified directory recursively.
--untar file.tar = Install by untarring the given file.
(You could specify one of these options as the default in the configuration
file /etc/xen-tools/xen-tools.conf.)
E_O_ERROR
logprint( $err );
exit;
}
#
# Make sure that any specified template file exists.
#
if ( ( defined( $CONFIG{'template'} ) ) && length( $CONFIG{'template'} ) )
{
if ( ! -e $CONFIG{'template'} )
{
logprint( "The specified template file, $CONFIG{'template'}, does not exist.\n" );
exit 1;
}
}
}
=begin doc
Check that we have some required files present.
=end doc
=cut
sub checkFilesPresent
{
#
# Files we demand are present in all cases.
#
my @required = qw ( /usr/sbin/debootstrap /bin/dd /bin/mount /sbin/mkswap );
foreach my $file ( @required )
{
if ( ! -x $file )
{
logprint( "The following binary is required to run this tool\n" );
logprint( "\t$file\n");
exit;
}
}
#
# LVM specific binaries
#
if ( $CONFIG{'lvm'} )
{
my @lvm = qw ( /sbin/lvcreate );
foreach my $file ( @lvm )
{
if ( ! -x $file )
{
logprint( "The following binary is required to run this tool\n");
logprint( "\t$file\n" );
logprint( "(This is only required for LVM volumes, which you've selected)\n" );
exit;
}
}
}
}
=begin doc
Show the user a summery of what is going to be created for them
=end doc
=cut
sub showSummery
{
#
# Show the user what to expect.
#
logprint( "\nGeneral Infomation\n" );
logprint( "--------------------\n" );
logprint( "Hostname : $CONFIG{'hostname'}\n" );
logprint( "Distribution : $CONFIG{'dist'}\n" );
logprint( "Fileystem Type : $CONFIG{'fs'}\n" );
logprint( "\nSize Information\n" );
logprint( "----------------\n" );
logprint( "Image size : $CONFIG{'size'}\n") ;
logprint( "Swap size : $CONFIG{'swap'}\n" ) unless ( $CONFIG{'noswap' } );
logprint( "Image type : $CONFIG{'image'}\n" );
logprint( "Memory size : $CONFIG{'memory'}\n" );
logprint( "Kernel path : $CONFIG{'kernel'}\n" );
if ( defined( $CONFIG{'initrd'} ) &&
length( $CONFIG{'initrd'} ) )
{
logprint( "initrd path : $CONFIG{'initrd'}\n" );
}
logprint( "\nNetworking Information\n" );
logprint( "----------------------\n" );
#
# Show each IP address added.
#
my $ips = $CONFIG{'ip'};
my $count = 1;
foreach my $i ( @$ips )
{
logprint( "IP Address $count : $i\n" );
$count += 1;
}
$CONFIG{'dhcp'} && logprint( "IP Address : DHCP\n" );
$CONFIG{'netmask'} && logprint( "Netmask : $CONFIG{'netmask'}\n" );
$CONFIG{'gateway'} && logprint( "Gateway : $CONFIG{'gateway'}\n" );
$CONFIG{'p2p'} && logprint( "Point to Point : $CONFIG{'p2p'}\n" );
print "\n";
}
=begin doc
Test that the user has the "loop" module loaded and present,
this is just a warning useful to newcomers.
=end doc
=cut
sub testLoopbackModule
{
if ( -e "/proc/modules" )
{
my $modules = `cat /proc/modules`;
if ( $modules !~ m/loop/ )
{
logprint( "WARNING\n" );
logprint( "-------\n" );
logprint( "Loopback module not loaded and you're using loopback images\n" );
logprint( "Run the following to load the module:\n\n" );
logprint( "modprobe loop loop_max=255\n\n" );
}
}
}
=begin doc
Create the two images "swap.img" and "disk.img" in the directory
we've been given.
We also will call the filesystem creation routine to make sure we
have a valid filesystem.
=end doc
=cut
sub createLoopbackImages
{
#
# The two files we need to test.
#
my $disk = $CONFIG{'dir'} . '/domains/' . $CONFIG{'hostname'} . "/disk.img" ;
my $swap = $CONFIG{'dir'} . '/domains/' . $CONFIG{'hostname'} . "/swap.img" ;
#
# Make sure we have the relevant output directory.
#
if ( ! -d $CONFIG{'dir'} . "/domains/" )
{
mkdir $CONFIG{'dir'} . '/domains', 0777
|| die "Cannot create $CONFIG{'dir'}/domains - $!";
}
if ( ! -d $CONFIG{'dir'} . "/domains/" . $CONFIG{'hostname'} )
{
mkdir $CONFIG{'dir'}. '/domains/' . $CONFIG{'hostname'}, 0777
|| die "Cannot create $CONFIG{'dir'}/domains/$CONFIG{'hostname'} - $!" ;
}
#
# Only proceed overwritting if we have --force specified.
#
if ( ( -e $disk ) && ! $CONFIG{'force'} )
{
logprint( "The disk image already exists. Aborting.\n" );
logprint( "Specify '--force' to overwrite, or remove the following file\n" );
logprint( $disk . "\n" );
exit;
}
if ( ( -e $swap ) && ! $CONFIG{'force'} )
{
logprint( "The swap image already exists. Aborting.\n" );
logprint( "Specify '--force' to overwrite, or remove the following file\n" );
logprint( $swap . "\n" );
exit;
}
#
# Modify the sizes to something reasonable
#
my $disk_size = $CONFIG{'size'};
my $swap_size = $CONFIG{'swap'};
#
# Convert Gb -> Mb for the disk image size, and swap size.
#
if ( $disk_size =~ /^(\d+)Gb*$/i )
{
$disk_size = $1 * 1024 . "M";
}
if ( $swap_size =~ /^(\d+)Gb*$/i )
{
$swap_size = $1 * 1024 . "M";
}
#
# Final adjustments to sizing.
#
$disk_size =~ s/Mb*$/k/i;
if ( $swap_size =~ /^(\d+)Mb*$/i )
{
$swap_size = $1;
}
# Use dd to create the swap
#
unless( $CONFIG{'noswap'} )
{
logprint( "\nCreating swap image: $swap\n" );
my $swap_cmd = "/bin/dd if=/dev/zero of=$swap bs=1024k count=$swap_size";
runCommand( $swap_cmd );
logprint( "Done\n" );
}
#
# Use dd to create the disk image.
#
logprint( "\nCreating disk image: $disk\n" );
my $image_cmd;
if ( $CONFIG{'image'} eq "sparse" )
{
$CONFIG{'verbose'} && logprint( "Creating sparse image\n" );
$image_cmd = "/bin/dd if=/dev/zero of=$disk bs=$disk_size count=0 seek=1024";
}
else
{
$CONFIG{'verbose'} && logprint( "Creating full-sized image\n" );
$image_cmd = "/bin/dd if=/dev/zero of=$disk bs=$disk_size count=1024";
}
runCommand( $image_cmd );
logprint( "Done\n" );
#
# Finally create the filesystem + swap
#
createFilesystem( $disk );
createSwap( $swap ) unless( $CONFIG{'noswap'} );
}
=begin doc
This function is responsible for creating two new logical volumes within
a given LVM volume group.
=end doc
=cut
sub createLVMBits
{
# The two volumes we will need to use..
my $disk = $CONFIG{'hostname'} . "-disk" ;
my $swap = $CONFIG{'hostname'} . "-swap" ;
#
# Check whether the disk volume exists already, and if so abort
# unless '--force' is specified.
#
my $lvm_disk = "/dev/$CONFIG{'lvm'}/$CONFIG{'hostname'}-disk";
if ( -e $lvm_disk )
{
# Delete if forcing
if ( $CONFIG{'force'} )
{
runCommand( "lvremove --force $lvm_disk" );
}
else
{
logprint( "The LVM disk image already exists. Aborting.\n" );
logprint( "Specify '--force' to delete and recreate\n" );
exit;
}
}
#
# Check whether the swap volume exists already, and if so abort
# unless '--force' is specified.
#
my $lvm_swap = "/dev/$CONFIG{'lvm'}/$CONFIG{'hostname'}-swap";
if ( -e $lvm_swap )
{
# Delete if forcing
if ( $CONFIG{'force'} )
{
runCommand( "lvremove --force $lvm_swap" );
}
else
{
logprint( "The LVM swap image already exists. Aborting.\n" );
logprint( "Specify '--force' to delete and recreate\n" );
exit;
}
}
#
# Modify the sizes to something reasonable
#
my $disk_size = $CONFIG{'size'};
my $swap_size = $CONFIG{'swap'};
#
# Convert Gb -> Mb for the disk image size, and swap size.
#
if ( $disk_size =~ /^(\d+)Gb*$/i )
{
$disk_size = $1 * 1024 . "M";
}
if ( $swap_size =~ /^(\d+)Gb*$/i )
{
$swap_size = $1 * 1024 . "M";
}
#
# Final adjustments to sizing.
#
$disk_size =~ s/Mb*$/k/i;
if ( $swap_size =~ /^(\d+)Mb*$/i )
{
$swap_size = $1;
}
#
# The commands to create the volumes.
#
my $disk_cmd = "/sbin/lvcreate $CONFIG{'lvm'} -L $CONFIG{'size'}M -n $disk";
my $swap_cmd = "/sbin/lvcreate $CONFIG{'lvm'} -L $CONFIG{'swap'} -n $swap";
#
# Create the volumes
#
runCommand( $disk_cmd );
runCommand( $swap_cmd ) unless( $CONFIG{'noswap'} );
#
# Initialise the partitions with the relevant filesystem.
#
createFilesystem( "/dev/$CONFIG{'lvm'}/$CONFIG{'hostname'}-disk" );
createSwap( "/dev/$CONFIG{'lvm'}/$CONFIG{'hostname'}-swap" ) unless ( $CONFIG{'noswap'} );
}
=begin doc
Format the given image in the users choice of filesystem.
=end doc
=cut
sub createFilesystem
{
my( $image ) = ( @_ );
#
# We have the filesystem the user wanted, make sure that the
# binary exists.
#
my $command = $CONFIG{ "make_fs_" . $CONFIG{'fs'} };
#
# Split the command into "binary" + "args". Make sure that
# the binary exists and is executable.
#
if ( $command =~ /([^ ]+) (.*)$/ )
{
my $binary = $1;
my $args = $2;
if ( ! -x $binary )
{
logprint( "The binary '$binary' required to create the filesystem $CONFIG{'fs'} is missing\n" );
exit;
}
}
else
{
logprint( "The filesystem creation hash is bogus for filesystem : $CONFIG{'fs'}\n" );
exit;
}
#
# OK we have the command and the filesystem. Create it.
#
logprint( "\nCreating $CONFIG{'fs'} filesystem on $image\n" );
$command .= $image;
runCommand( $command );
logprint( "Done\n" );
}
=begin doc
Create the swap filesystem on the given device.
=end doc
=cut
sub createSwap
{
my ( $path ) = ( @_ );
runCommand( "/sbin/mkswap $path" );
}
=begin doc
Mount the loopback disk image into a temporary directory.
Alternatively mount the relevant LVM volume instead.
=end doc
=cut
sub mountImage
{
#
# Determine what we're to mount
#
my $image;
if ( $CONFIG{'lvm'} )
{
$image = "/dev/" . $CONFIG{'lvm'} . "/" . $CONFIG{'hostname'} . '-disk';
}
elsif ( $CONFIG{'dir'} )
{
$image = $CONFIG{'dir'} . '/domains/' . $CONFIG{'hostname'} . "/disk.img" ;
}
else
{
logprint( "I don't know what to mount!\n" );
logprint( "Please specify '--dir' or '--lvm'\n" );
exit;
}
#
# Create a temporary mount-point to use for the image/volume.
#
$MOUNT_POINT = tempdir( CLEANUP => 1 );
#
# Lookup the correct arguments to pass to mount.
#
my $mount_cmd;
my $mount_type = $CONFIG{'mount_fs_' . $CONFIG{'fs'} };
#
# LVM partition
#
if ( $CONFIG{'lvm'} )
{
$mount_cmd = "mount $mount_type $image $MOUNT_POINT";
}
else
{
$mount_cmd = "mount $mount_type -o loop $image $MOUNT_POINT";
}
runCommand( $mount_cmd );
}
=begin doc
Install the system, by invoking the xt-install-system script.
The script will be given the appropriate arguments from our environment.
=end doc
=cut
sub installSystem
{
logprint( "\nInstalling your system with " );
#
# Basic command
#
my $cmd = "/usr/bin/xt-install-image --hostname=$CONFIG{'hostname'} --location=$MOUNT_POINT --dist=$CONFIG{'dist'}";
#
# Add on the current cache setting
#
$cmd .= " --cache=$CONFIG{'cache'}" if length( $CONFIG{'cache'} );
#
# Installation method
#
if ( $CONFIG{'copy'} )
{
$cmd .= " --copy=$CONFIG{'copy'}";
logprint( "copy from $CONFIG{'copy'}\n" );
}
if ( $CONFIG{'debootstrap'} )
{
$cmd .= " --debootstrap";
$cmd .= " --mirror=$CONFIG{'mirror'}";
logprint( "debootstrap mirror $CONFIG{'mirror'}\n" );
}
if ( $CONFIG{'rpmstrap'} )
{
$cmd .= " --rpmstrap";
logprint( "rpmstrap\n");
}
if ( $CONFIG{'tar'} )
{
$cmd .= " --tar=$CONFIG{'tar'}";
logprint( "tarfile $CONFIG{'tar'}\n" );
}
#
# Propogate --verbose
#
if ( $CONFIG{'verbose'} )
{
$cmd .= " --verbose";
}
runCommand( $cmd );
logprint( "Done\n" );
}
=begin doc
Run the xt-customise-system script to customize our fresh installation.
Before we do this we must pass all the relevent options into our
environment.
=end doc
=cut
sub runCustomisationHooks
{
#
# Setup the environment for the child processes.
#
foreach my $key ( keys %CONFIG )
{
if ( defined( $CONFIG{$key} ) )
{
$ENV{$key} = $CONFIG{$key};
}
}
#
# Now update the environment for each defined IP address.
# these are handled specially since we use arrays.
#
# Remove the value we set above.
delete $ENV{'ip'};
#
# Setup a seperate ip$count value for each IP address.
#
my $ips = $CONFIG{'ip'};
my $count = 1;
foreach my $i ( @$ips )
{
$ENV{'ip' . $count } = $i;
$count += 1;
}
$ENV{'ip_count'} = ($count - 1);
#
# Now show the environment the children get
#
if ( $CONFIG{'verbose'} )
{
logprint( "Customization Script Environment:\n" );
logprint( "---------------------------------\n" );
foreach my $key ( sort keys %ENV )
{
logprint( "\t'" . $key . "' = '" . $ENV{$key} . "'\n" );
}
}
#
# Actually run the appropriate hooks
#
my $customize = "xt-customize-image --dist=$CONFIG{'dist'} --location=$MOUNT_POINT";
if ( $CONFIG{'verbose'} )
{
$customize .= " --verbose";
}
logprint( "\nRunning hooks\n" );
runCommand( $customize );
logprint( "Done\n" );
}
=begin doc
If the user specified a role for the new instance then execute it.
=end doc
=cut
sub runRoleScript
{
my $roleDir = '/usr/lib/xen-tools/' . $CONFIG{'dist'} . '.d/role.d/' ;
my $role = $CONFIG{'role'};
if ( !defined( $role ) )
{
logprint( "\nNo role script specified. Skipping\n" );
return;
}
my $file = $roleDir . $role;
if ( -x $file )
{
logprint( "\nRunning role script $file for role '$role'\n" );
}
else
{
logprint( "\nRole script not executable : $file for role '$role'\n" );
logprint( "Ignoring\n" );
return;
}
#
# Our environment is already setup because of the call to
# runCustomisationHooks. So we just need to run the script
# with the mountpoint.
#
runCommand( $file . " " . $MOUNT_POINT );
logprint( "Done\n" );
}
=begin doc
Create the Xen configuration file.
Note that we don't need to do any setup for the environment since
we did this already before running the hook scripts.
=end doc
=cut
sub runXenConfigCreation
{
my $command = '/usr/bin/xt-create-xen-config --output=/etc/xen';
if ( ( defined( $CONFIG{'template'} ) ) &&
( -e $CONFIG{'template'} ) )
{
$command .= " --template=" . $CONFIG{'template'};
}
logprint( "\nCreating Xen configuration file\n" );
runCommand( $command );
logprint( "Done\n" );
}
=begin doc
chroot() into the new system and setup the password.
=end doc
=cut
sub setupRootPassword
{
logprint( "Setting up root password\n" );
if ( -x $MOUNT_POINT . "/usr/bin/passwd" )
{
system( "chroot $MOUNT_POINT /usr/bin/passwd" );
}
else
{
logprint( "/usr/bin/passwd on the new system doesn't exist...\n" );
}
}
=begin doc
Print the given string both to our screen, and to the logfile.
=cut
sub logprint
{
my ( $text ) = (@_);
print $text;
#
# Log.
#
if ( $CONFIG{'hostname'} )
{
open( LOGFILE, ">>", "/var/log/xen-tools/$CONFIG{'hostname'}.log" )
or return;
print LOGFILE $text;
close( LOGFILE );
}
}
=begin doc
A utility method to run a system command. We will capture the return
value and exit if the command files.
When running verbosely we will also display any command output once
it has finished.
=end doc
=cut
sub runCommand
{
my ( $cmd ) = (@_ );
#
# Header.
#
$CONFIG{'verbose'} && print "Executing : $cmd\n";
#
# Copy stderr to stdout, so we can see it, and make sure we log it.
#
$cmd .= " 2>&1 | tee --append /var/log/xen-tools/$CONFIG{'hostname'}.log";
#
# Run it.
#
my $output = `$cmd`;
$CONFIG{'verbose'} && print "Finished : $cmd\n";
if ( $? != 0 )
{
logprint( "Running command '$cmd' failed.\n" );
logpring( "Aborting\n" );
exit;
}
}
=begin doc
If we still have the temporary image mounted then make sure
it is unmounted before we terminate.
=end doc
=cut
sub END
{
if ( defined( $MOUNT_POINT ) )
{
#
# Run mount to see if this is still mounted.
#
my $mount = `/bin/mount`;
if ( $mount =~ /$MOUNT_POINT/)
{
runCommand( "umount $MOUNT_POINT" );
}
}
if ( ( defined( $CONFIG{'hostname'} ) ) &&
( -e "/var/log/xen-tools/$CONFIG{'hostname'}.log" ) )
{
print "\n\nLogfile produced at:\n";
print "\t /var/log/xen-tools/$CONFIG{'hostname'}.log\n";
}
}