vmbuilder create ubuntu-kvm directory even when using --raw

[Simon Huerlimann]

When calling vmbuilder with --raw, it still creates a directory <distro>-<hypervisor> in the current working directory. All it contains is a file 'run.sh'. Content is similar to this:

#!/bin/sh
exec kvm -m 2048 -smp 1 -drive file=kvm-web03.demo "$@"

ProblemType: Bug
DistroRelease: Ubuntu 10.04
Package: python-vm-builder 0.12.3-0ubuntu1
ProcVersionSignature: Ubuntu 2.6.32-21.32-server 2.6.32.11+drm33.2
Uname: Linux 2.6.32-21-server x86_64
Architecture: amd64
Date: Sat Apr 24 13:48:24 2010
InstallationMedia: Ubuntu-Server 10.04 "Lucid Lynx" - Alpha amd64 (20100410)
PackageArchitecture: all
ProcEnviron:
 LANG=en_US.UTF-8
 SHELL=/bin/bash
SourcePackage: vm-builder

[Simon Huerlimann (huerlisi)]

This is especially tricky as it won't let you create two raw vms (without specifying --destdir) in a row from the same directory. This is the output for the second run:

2010-04-25 06:21:52,984 ERROR : ubuntu-kvm already exists
Traceback (most recent call last):
  File "/usr/bin/vmbuilder", line 24, in <module>
    cli.main()
  File "/usr/lib/python2.6/dist-packages/VMBuilder/contrib/cli.py", line 81, in main
    raise VMBuilderUserError('%s already exists' % destdir)
VMBuilder.exception.VMBuilderUserError: ubuntu-kvm already exists

[Chuck Short (zulcss)]

Thank you for taking the time to report this bug and helping to make Ubuntu better. Please answer these questions:
1. Is this reproducible?
2. If so, what specific steps should we take to recreate this bug? Be as detailed as possible.
This will help us to find and resolve the problem.

[Simon Huerlimann (huerlisi)]

Yes, it is reproducible.

Preconditions:
* an LVM group named leo02
* PWD were you can write...

The following steps should do it, assuming a LVM group leo02:
sudo lvcreate /dev/leo02 --name kvm-test01 --size 5G
sudo vmbuilder kvm ubuntu --addpkg=openssh-server --addpkg=unattended-upgrades, --addpkg=acpid --addpkg=puppet --domain=example.com --bridge=virbr0 --gw=10.0.0.1 --suite=lucid --libvirt=qemu:///system --ip=10.0.1.1 --hostname=test01 --raw=/dev/leo02/kvm-test01
# $PWD/ubuntu-kvm gets created, containing file run.sh

sudo lvcreate /dev/leo02 --name kvm-test02 --size 5G
sudo vmbuilder kvm ubuntu --addpkg=openssh-server --addpkg=unattended-upgrades, --addpkg=acpid --addpkg=puppet --domain=example.com --bridge=virbr0 --gw=10.0.0.1 --suite=lucid --libvirt=qemu:///system --ip=10.0.1.2 --hostname=test02 --raw=/dev/leo02/kvm-test02
# Raises the following error:
2010-04-29 15:17:27,354 ERROR : ubuntu-kvm already exists
Traceback (most recent call last):
  File "/usr/bin/vmbuilder", line 24, in <module>
    cli.main()
  File "/usr/lib/python2.6/dist-packages/VMBuilder/contrib/cli.py", line 81, in main
    raise VMBuilderUserError('%s already exists' % destdir)
VMBuilder.exception.VMBuilderUserError: ubuntu-kvm already exists

This is with python-vm-builder 0.12.3-0ubuntu1

[Whit Blauvelt (whit-launchpad)]

Not sure if this is a related bug, or just a lack of documentation, but:

vmbuilder kvm ubuntu --suite lucid --flavour virtual --arch amd64 -o --libvirt qemu:///system

produces an ubuntu-kvm directory with, in this instance:

-rw-r--r-- 1 root root 396754944 2010-08-08 21:20 tmpEs8Mq4.qcow2
-rwx---r-x 1 root root 67 2010-08-08 21:21 run.sh

where run.sh consists of:

#!/bin/sh

exec kvm -m 128 -smp 1 -drive file=tmpEs8Mq4.qcow2 "$@"

and trying to run run.sh (which AFAIK is an undocumented file here) produces:

pci_add_option_rom: failed to find romfile "pxe-rtl8139.bin"
init kbd.
^CCould not open SDL display

(that last line after a control-C)

Is run.sh something that should be a final product? Or has vmbuilder got itself into an error state and quit in the middle somewhere, leaving that tmp*.qcow2 file and the run.sh which can't run it, and that would disappear if vmbuilder completed its job without error?