Switch charm to Bobcat release. Microk8s 1.28 will be the default for Bobcat. Also ensure images are being collected from ghcr.io/canonical Change-Id: I51edbefbe61e2450323bc54c6f0271ce435a80fd
1.4 KiB
glance-k8s
Developing
Create and activate a virtualenv with the development requirements:
virtualenv -p python3 venv
source venv/bin/activate
pip install -r requirements-dev.txt
Code overview
Get familiarise with Charmed Operator Framework and Sunbeam documentation.
glance-k8s charm uses the ops_sunbeam library and extends OSBaseOperatorAPICharm from the library.
glance-k8s charm consumes database relation to connect to database, identity-service to register the service endpoints to keystone and ingress-internal/ingress-public relation to get exposed over internal and public networks.
Intended use case
glance-k8s charm deploys and configures OpenStack Image service on a kubernetes based environment. It should be possible to use local storage or ceph based storage as backend.
Roadmap
TODO
Testing
The Python operator framework includes a very nice harness for testing operator behaviour without full deployment. Run tests using command:
tox -e py3
Deployment
This project uses tox for building and managing. To build the charm run:
tox -e build
To deploy the local test instance:
juju deploy ./glance-k8s_ubuntu-20.04-amd64.charm --trust --resource glance-image=ghcr.io/canonical/glance-api:2023.2