A set of rake tasks to build packages and install OpenStack.
-
Integrates with Kytoon to create groups of servers for installation.
-
Build packages and create a package repo.
-
Install and configure packages on nodes (via Puppet).
-
Tasks to configure keystone data, load glance images, etc.
-
Run various test suites (Torpedo, Tempest, Nova Smoke Tests)
Useful for developmental and functional testing.
-Kytoon: https://github.com/dprince/kytoon
Available tasks:
rake fedora:create_rpm_repo # Create an RPM repo.
rake fedora:fill_cache # Upload packages to the cache URL.
rake glance:build_packages # Build Glance packages.
rake glance:install_source # Install local Glance source code into the g...
rake kytoon:create # Create a new group of cloud servers
rake kytoon:delete # Delete a cloud server group
rake kytoon:gateway_ip # Print the VPN gateway IP address
rake kytoon:list # List existing cloud server groups.
rake kytoon:show # Print information for a cloud server group
rake keystone:build_packages # Build Keystone packages.
rake keystone:configure # Configure keystone
rake nova:build_packages # Build Nova packages.
rake nova:install_source # Install local Nova source code into the group.
rake nova:smoke_tests # Run the nova smoke tests.
rake nova:tail_logs # Tail nova logs.
rake puppet:install # Install and configure packages on clients w...
rake ssh # SSH into the most recently created VPN gate...
rake stackattack:install # Install stack attack and dependencies on SE...
rake swift:build_packages # Build Swift packages.
rake tail_logs # Tail nova, glance, keystone logs.
rake tempest # Install and run Tempest.
rake torpedo # Install and run Torpedo: Fast Openstack tests
rake usage # Print help and usage information
rake xen:install_plugins # Install plugins into the XenServer dom0.
# Before using Firestack w/ Libvirt you need to have an image and a libvirt
# XML dom file on disk that will be used to clone new VM's for each
# Firestack group.
LIBVIRT_XML_FILE=/path/to/libvirt/dom/xml
set -x
# install rubygems, rubygem-json, and Git, make, gcc, etc
for X in rubygems ruby-devel rubygem-bundler git make gcc; do
rpm -q $X &> /dev/null || yum install -q -y $X
done
# Some older RHEL/Fedora distro don't have a bundler package
if ! gem list | grep bundler &> /dev/null; then
gem install --no-rdoc --no-ri bundler
fi
git clone git://github.com/dprince/firestack.git
cd firestack
bundle install
# Configure the server group XML for Kytoon
cat >> config/server_group_libvirt.json <<EOF_CAT
{
"name": "Fedora",
"servers": [
{
"hostname": "nova1",
"memory": "4",
"gateway": "true",
"original_xml": "$LIBVIRT_XML_FILE",
"create_cow": "true"
}
]
}
EOF_CAT
# Configure kytoon.conf
# NOTE: This config assumes you have configured libvirt to run with your
# username. Alternately you can have Kytoon use sudo via the libvirt_use_sudo.
#
# See this puppet module if you are interested in quickly configuring worker
# nodes to allow a 'smokestack' user to make use of libvirt:
# https://github.com/dprince/smokestack-puppet/blob/master/modules/smokestack/manifests/libvirt.pp
cat >> ~/.kytoon.conf <<EOF_CAT
# The default group type.
# Set to one of: openstack, libvirt, xenserver
group_type: libvirt
# Libvirt settings
# Whether commands to create local group should use sudo
libvirt_use_sudo: False
EOF_CAT
# Generate an ssh keypair if you don't already have one
if [ ! -f ~/.ssh/id_rsa ]; then
ssh-keygen -q -t rsa -f ~/.ssh/id_rsa -N ""
fi
# Now you are ready to run the example_libvirt.bash
# From here on out just run this directly!
bash example_libvirt.bashTypically you'll want to create a runner script that creates a new group, builds packages, installs them, etc. See example_libvirt.bash as an example. The following commands are commonly used:
#create a group
rake kytoon:create SERVER_GROUP_JSON="config/server_group_fedora.json"
#build Nova packages
rake nova:build_packages \
RPM_PACKAGER_URL="git://github.com/redhat-openstack/openstack-nova.git" \
SOURCE_URL="git://github.com/openstack/nova.git" \
SOURCE_BRANCH="master"
#create an RPM repo
rake fedora:create_rpm_repo
#install/configure packages with Puppet
rake puppet:install SOURCE_URL="git://github.com/redhat-openstack/openstack-puppet.git" PUPPET_CONFIG="single_node_mysql"
#configure keystone
rake keystone:configure
#load glance images
rake glance:load_images
#Run test suites
rake torpedo
rake tempest
See the example bash script in this directory for detailed example using libvirt.