Deploys Kubernetes on various providers. Deployments are for evaluation/experimental purpose (not production).

Behind the scenes this project uses:

• Kubespray (https://github.com/kubernetes-sigs/kubespray): It uses decent default values but this should be reasonnably customizable (through roles and vars in the configuration file)
• EnOSlib (https://gitlab.inria.fr/discovery/enoslib) for the framework part.

Install the project:

virtualenv -p python3 venv
source venv/bin/activate
pip install enos-kubernetes

Configure the Grid5000 REST API access:

echo '
username: MYLOGIN
password: MYPASSWORD
' > ~/.python-grid5000.yaml

Get a sample configuration file at:

wget https://gitlab.inria.fr/msimonin/enos-kubernetes/raw/master/conf.yaml

Deploy on g5k:

ek deploy g5k

Deploy on g5k using virtual machines:

ek deploy vmong5k

Build a base image on g5k:

ek build g5k

Build a base image on vmong5k with an alternative cluster:

ek build vmong5k --cluster=chetemi

This also can be used from python directly using the provided API:

# pseudo-code to deploy to g5k
from enos_kubernetes import tasks

...
tasks.g5k(...)
tasks.inventory(...)
tasks.prepare(...)
...