Install Cyborg from Source

this page last updated: 2019-03-24 20:33:24

Install Cyborg from Source¶

This section describes how to install and configure the Acceleration Service for Ubuntu 18.04.1 LTS from source code.

Prerequisites¶

Before you install and configure the Accelerator Life Cycle Management service, you must create a database, service credentials, and API endpoints.

To create the database, complete these steps:
- Use the database access client to connect to the database server as the root user:
```
$ mysql -u root -p
```
- Create the cyborg database:
```
CREATE DATABASE cyborg;
```
- Grant proper access to the cyborg database:
```
GRANT ALL PRIVILEGES ON cyborg.* TO 'cyborg'@'localhost' IDENTIFIED BY 'CYBORG_DBPASS';
```
  Replace CYBORG_DBPASS with a suitable password.
- Exit the database access client.
```
exit;
```
Source the admin credentials to gain access to admin-only CLI commands:
```
$ . admin-openrc
```

To create the service credentials, complete these steps:

Create the cyborg user:

$ openstack user create --domain default --password-prompt cyborg

Add the admin role to the cyborg user:

$ openstack role add --project service --user cyborg admin

Create the cyborg service entities:

$ openstack service create --name cyborg --description "Acceleration Service" accelerator

Create the Acceleration service API endpoints:

If cyborg-api service is deployed using uwsgi, create the following endpoints:

$ openstack endpoint create --region RegionOne \
  accelerator public http://<cyborg-ip>/accelerator/v1
$ openstack endpoint create --region RegionOne \
  accelerator internal http://<cyborg-ip>/accelerator/v1
$ openstack endpoint create --region RegionOne \
  accelerator admin http://<cyborg-ip>/accelerator/v1

Otherwise, if cyborg-api service is running as a python process, create the following endpoints:

$ openstack endpoint create --region RegionOne \
  accelerator public http://<cyborg-ip>:6666/v1
$ openstack endpoint create --region RegionOne \
  accelerator internal http://<cyborg-ip>:6666/v1
$ openstack endpoint create --region RegionOne \
  accelerator admin http://<cyborg-ip>:6666/v1

Note

URLs (publicurl, internalurl and adminurl) may be different depending on your environment.

Install and Configure¶

Create a folder which will hold all Cyborg components.
```
mkdir ~/cyborg
```

Clone the cyborg git repository to the management server.

cd ~/cyborg
git clone https://git.openstack.org/openstack/cyborg

Set up the cyborg config file

First, generate a sample configuration file, using tox

cd ~/cyborg/cyborg
tox -e genconfig

And make a copy of it for further modifications

cp -r ~/cyborg/cyborg/etc/cyborg /etc
cd /etc/cyborg
ln -s cyborg.conf.sample cyborg.conf

Edit cyborg.conf with your favorite editor. Below is an example which contains basic settings you likely need to configure.

[DEFAULT]
transport_url = rabbit://%RABBITMQ_USER%:%RABBITMQ_PASSWORD%@%OPENSTACK_HOST_IP%:5672/
use_syslog = False
state_path = /var/lib/cyborg
debug = True

...

[database]
connection = mysql+pymysql://%DATABASE_USER%:%DATABASE_PASSWORD%@%OPENSTACK_HOST_IP%/cyborg

...

[service_catalog]
cafile = /opt/stack/data/ca-bundle.pem
project_domain_id = default
user_domain_id = default
project_name = service
password = cyborg
username = cyborg
auth_url = http://%OPENSTACK_HOST_IP%/identity
auth_type = password

...

[placement]
project_domain_name = Default
project_name = service
user_domain_name = Default
password = password
username = placement
auth_url = http://%OPENSTACK_HOST_IP%/identity
auth_type = password
auth_section = keystone_authtoken

...

[keystone_authtoken]
memcached_servers = localhost:11211
signing_dir = /var/cache/cyborg/api
cafile = /opt/stack/data/ca-bundle.pem
project_domain_name = Default
project_name = service
user_domain_name = Default
password = cyborg
username = cyborg
auth_url = http://%OPENSTACK_HOST_IP%/identity

Install Cyborg packages.

cd ~/cyborg/cyborg
sudo python setup.py install

Create database tables for Cyborg.

cd /usr/local/bin
cyborg-dbsync --config-file /etc/cyborg/cyborg.conf upgrade

Install Cyborg API via WSGI api-uwsgi

Note

Cyborg-api service can also be run as a Python command that runs a web serve, which can be launched as follows with different Acceleration service API endpoints as mentioned in Prerequisites part. However, we would like to recommend you the uwsgi way since when a project provides a WSGI application the API service gains flexibility in terms of deployment, performance, configuration and scaling. BYW, if you choose devstack to deploy your acceleration service, uwsgi is a default choice.

cyborg-api –config-file=/etc/cyborg/cyborg.conf

Launch Cyborg Conductor, Cyborg Agent services. Open a separate terminal for each service since the console will be locked by a running process.
```
cyborg-conductor --config-file=/etc/cyborg/cyborg.conf
cyborg-agent --config-file=/etc/cyborg/cyborg.conf
```

this page last updated: 2019-03-24 20:33:24

Except where otherwise noted, this document is licensed under Creative Commons Attribution 3.0 License. See all OpenStack Legal Documents.