a769413d89
Cinder requires temporary working space to convert images. This patch exposes cinder_volume_lv_size_gb to the user config file, so the user can decide how large the cinder volumes container should be based on available space and the size of images that will need to be converted. cinder_volume_lv_size_gb is used to override container_lvm_fssize in group_vars/cinder_volume. Simple enough but doesn't work because templated variables (or indirect variables) are not expanded when accessed via hostvars[] see: ansible/ansible#7844. In order to work around that, I have eliminated hostvars[] usage from the container creation mechanism. This may have positive speed implications as the limit of container creation parallelism is now forks rather than number of hosts. However it does make this change larger than a small bug fix. Also note that this patch makes use of delegate_to, so specific ansible versions must be used to avoid ansible/ansible#8705. Our requirements file currently specifies a version before this bug was introduced. There are two commits in this PR as one is the actual bugfix, the other is infrastructure changes required for that bugfix to work. Also only the bugfix may be needed if the upstream bugs are fixed. Closes-Bug: #1399427 Change-Id: I2b5c5e692d3d72b603fdd6298475cb76c52c66df |
||
---|---|---|
etc | ||
rpc_deployment | ||
scripts | ||
ssh | ||
tools | ||
.gitignore | ||
.gitreview | ||
Changelog.md | ||
CONTRIBUTING.md | ||
LICENSE | ||
README.rst | ||
requirements.txt |
Rackspace Private Cloud Version 9.0
- date
-
2014-09-25 09:00
- tags
-
rackspace, lxc, openstack, cloud, ansible
- category
-
*nix
License
Copyright 2014, Rackspace US, Inc.
Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with the License. You may obtain a copy of the License at:
http://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License.
Official Documentation
Comprehensive installation guides, including FAQs and release notes, can be found at http://docs.rackspace.com
Playbook Support
- OpenStack:
-
- keystone
- glance-api
- glance-registry
- cinder-api
- cinder-scheduler
- cinder-volume
- nova-api
- nova-api-ec2
- nova-api-metadata
- nova-api-os-compute
- nova-compute
- nova-conductor
- nova-scheduler
- heat-api
- heat-api-cfn
- heat-api-cloudwatch
- heat-engine
- horizon
- neutron-server
- neutron-dhcp-agent
- neutron-metadata-agent
- neutron-linuxbridge-agent
- Infrastructure:
-
- galera
- rabbitmq
- logstash
- elastic-search
- kibana
Assumptions
This repo assumes that you have setup the host servers that will be
running the OpenStack infrastructure with three bridged network devices
named: br-mgmt
, br-vxlan
,
br-vlan
. These bridges will be used throughout the
OpenStack infrastructure.
The repo also relies on configuration files found in the /etc directory of this repo. If you are running Ansible from an "unprivileged" host, you can place the contents of the /etc/ directory in your home folder; this would be in a directory similar to /home/<myusername>/rpc_deploy/. Once you have the file in place, you will have to enter the details of your environment in the rpc_user_config.yml file; please see the file for how this should look. After you have a bridged network and the files/directory in place, continue on to _Base Usage.
Base Usage
All commands must be executed from the rpc_deployment directory. From this directory you will have access to all of the playbooks, roles, and variables. It is recommended that you create an override file to contain any and all variables that you wish to override for the deployment. While the override file is is not required it will make life a bit easier.
All of the variables that you may wish to update are in the vars/ directory, however you should also be aware that services will pull in base group variables as found in inventory/group_vars.
All playbooks exist in the playbooks/
directory and are
grouped in different sub-directories.
All of the keys, tokens, and passwords are in the user_variables.yml file. This file contains no
preset passwords. To setup your keys, passwords, and tokens you will
need to either edit this file manually or use the script
pw-token-gen.py
. Example:
# Generate the tokens
scripts/pw-token-gen.py --file /etc/rpc_deploy/user_variables.yml
Example usage from the rpc_deployment directory in the ansible-rpc-lxc repository
# Run setup on all hosts:
ansible-playbook -e @vars/user_variables.yml playbooks/setup/host-setup.yml
# Run infrastructure on all hosts
ansible-playbook -e @vars/user_variables.yml playbooks/infrastructure/infrastructure-setup.yml
# Setup and configure openstack within your spec'd containers
ansible-playbook -e @vars/user_variables.yml playbooks/openstack/openstack-setup.yml
About Inventory
All things that Ansible cares about are located in inventory. In the Rackspace Private Cloud all inventory is dynamically generated using the previously mentioned configuration files. While this is a dynamically generated inventory, it is not 100% generated on every run. The inventory is saved in a file named rpc_inventory.json and is located in the directory where you've located your user configuration files. On every run a backup of the inventory json file is created in both the current working directory as well as the location where the user configuration files exist. The inventory json file is a living document and is intended to grow as the environment scales in infrastructure. This means that the inventory file will be appended to as you add more nodes and or change the container affinity from within the rpc_user_config.yml file. It is recommended that the base inventory file be backed up to a safe location upon the completion of a deployment operation. While the dynamic inventory processor has guards in it to ensure that the built inventory is not adversely effected by programmatic operations this does not guard against user error and/or catastrophic failure.
Scaling
If you are scaling the environment using the dynamically generated inventory you should know that the inventory was designed to generate new entries in inventory and not remove entries from inventory. These playbooks will build an environment to spec so if container affinity is changed and or a node is added or removed from an environment the user configuration file will need to be modified as well as the inventory json. For this reason it is recommended that should a physical node need replacing it should be renamed the same as the previous one. This will make things easier when rebuilding the environment. Additionally if a container is needing to be replaced it is better to simply remove the misbehaving container and rebuild it using the existing inventory.
Notes
- Library has an experimental Keystone
module which adds
keystone:
support to Ansible. - Library has an experimental Swift
module which adds
swift:
support to Ansible. - Library has an experimental LXC
module which adds
lxc:
support to Ansible.