Clone this repository and cd into it.
Create a virtualenv and install prerequisites:
$ virtualenv -p $(which python3) venv $ . venv/bin/activate $ pip install -r requirements.txt
Then run the following commands:
$ vagrant up $ vagrant ssh-config
Based on the following template, create the inventory/hosts inventory file and fill it in with values for parameters HostName, Port and IdentityFile outputted by the last command:
cloudkitty-playground ansible_user=vagrant ansible_host=<IP-from-HostName> ansible_port=<number-from-Port> ansible_private_key_file=<path-from-IdentityFile>
|
Note
|
You will need to update the inventory each time you destroy the AIO VM with |
Confirm you are all set up by running:
$ ansible -i inventory/ all -m shell -a "hostname"
TODO: Find a method to generate Ansible inventories for Vagrant boxes automatically.
Next up, create this symlink (or move/copy the kolla subdirectory to /etc/) on your deploy host and deploy the OpenStack AIO with kolla-ansible:
$ sudo ln -s $(pwd)/kolla /etc/kolla $ kolla-genpwd $ kolla-ansible -vv -i inventory/ bootstrap-servers $ kolla-ansible -vv -i inventory/ deploy
|
Note
|
Config files for all the deployed OpenStack services can be found in the |
When the deploy is done, run these commands to create some basic OpenStack resources:
$ kolla-ansible -vv -i inventory/ post-deploy $ source /etc/kolla/admin-openrc.sh $ wget https://opendev.org/openstack/kolla-ansible/raw/branch/stable/train/tools/init-runonce $ EXT_NET_CIDR='172.20.20.0/24' EXT_NET_RANGE='start=172.20.20.100,end=172.20.20.200' EXT_NET_GATEWAY='172.20.20.1' bash init-runonce
Now you can create a test instance and SSH into it to confirm the environment is fully functional:
$ openstack server create --image cirros --flavor m1.tiny --key-name mykey --network demo-net demo1
$ eval $(openstack floating ip create public1 -f shell | grep ^floating_ip_address)
$ openstack server add floating ip demo1 ${floating_ip_address}
$ openstack server list
$ ssh cirros@${floating_ip_address}
|
Note
|
You may need to wait a few seconds before SSH on the instance becomes available. |
|
Note
|
The Horizon dashboard URL is http://10.10.10.254. |
To confirm CloudKitty is available, run this command on your deploy host:
$ openstack rating summary get
This should return an empty list.
For Cloudkitty testing and development it might be more convenient to work with CloudKitty installed in a virtualenv on the host rather than in containers.
$ vagrant ssh $ sudo yum install python3 python3-devel gcc git vim -y $ curl https://bootstrap.pypa.io/get-pip.py | sudo python3 - $ pip3 install virtualenv==20.2.1 $ git clone https://opendev.org/openstack/cloudkitty $ cd cloudkitty/ $ git checkout stable/train $ virtualenv -p $(which python3) venv $ . venv/bin/activate $ pip install --upgrade pip $ pip install -r requirements.txt $ pip install pymysql==0.10.1 python-memcached==1.51 # these modules are needed but not listed in requirements $ pip install Werkzeug==0.16.1 # the latest version (1.0.1) no longer supports the "DispatcherMiddleware()" method $ python setup.py install $ sudo mkdir -p /opt/cloudkitty/log $ sudo chown -R $USER /opt/cloudkitty/
To replace the cloudkitty_processor container follow these steps:
$ sudo cat /etc/kolla/cloudkitty-processor/cloudkitty.conf | tee /opt/cloudkitty/cloudkitty.conf $ sudo cat /etc/kolla/cloudkitty-processor/metrics.yml | tee /opt/cloudkitty/metrics.yml
|
Note
|
Sample |
Adjust paths for the following config parameters found in /opt/cloudkitty/cloudkitty.conf:
sed -i.$(date +%s).bak 's|\(^log_dir = \).*|\1/opt/cloudkitty/log|' /opt/cloudkitty/cloudkitty.conf #`log_dir = /opt/cloudkitty/log` in section `[DEFAULT]` sed -i.$(date +%s).bak 's|\(^lock_path = \).*|\1/opt/cloudkitty/tmp|' /opt/cloudkitty/cloudkitty.conf #`lock_path = /opt/cloudkitty/tmp` in section `[oslo_concurrency]` sed -i.$(date +%s).bak 's|\(^metrics_conf = \).*|\1/opt/cloudkitty/metrics.yml|' /opt/cloudkitty/cloudkitty.conf #`metrics_conf = /opt/cloudkitty/metrics.yml` in section `[collect]`
Also, remove or comment out log_file if present.
Finally, start the service:
$ sudo docker stop cloudkitty_processor $ . venv/bin/activate $ $(which cloudkitty-processor) --config-file /opt/cloudkitty/cloudkitty.conf
Processor logs will be written to /opt/cloudkitty/log/cloudkitty-processor.log.
To replace the cloudkitty_api container follow these steps:
$ sudo cat /etc/kolla/cloudkitty-api/cloudkitty.conf | tee /opt/cloudkitty/cloudkitty.conf $ sudo docker exec cloudkitty_api cat /etc/cloudkitty/api_paste.ini | tee /opt/cloudkitty/api_paste.ini
Adjust paths for the following config parameters found in /opt/cloudkitty/cloudkitty.conf:
sed -i.$(date +%s).bak 's|\(^log_dir = \).*|\1/opt/cloudkitty/log|' /opt/cloudkitty/cloudkitty.conf #`log_dir = /opt/cloudkitty/log` in section `[DEFAULT]` sed -i.$(date +%s).bak 's|\(^lock_path = \).*|\1/opt/cloudkitty/tmp|' /opt/cloudkitty/cloudkitty.conf #`lock_path = /opt/cloudkitty/tmp` in section `[oslo_concurrency]`
Also, remove or comment out log_file if present.
Finally, start the service:
$ sudo docker stop cloudkitty_api $ . venv/bin/activate $ $(which cloudkitty-api) --host 10.10.10.254 --port 8889 -- --config-file /opt/cloudkitty/cloudkitty.conf
|
Note
|
The port number and the host IP have to be passed through flags even if they are specified again in the |
API logs will be written to /opt/cloudkitty/log/cloudkitty-api.log.
Run these commands to check out what data CloudKitty has collected so far in its InfluxDB database:
$ sudo docker exec -it influxdb influx -host 10.10.10.254 -database cloudkitty > SHOW MEASUREMENTS > SELECT * FROM dataframes ORDER BY time DESC LIMIT 10