Automate the execution mesh
When the installation and configuration of the automation platform is fully automated,
there is still one thing to do.
Install the execution nodes and instance groups your organizations might need to run their jobs.
The RedHat documentation tells us how to install a execution node and register this node into rhaap.
We will create a playbook that can do this for us, not only installing a execution node, but also recover
it when it is damaged or corrupted. This can be usefull after a reinstall of the rhaap controller.
This is the moment where the certificates change that rhaap uses.
Also when using your organizations certificates, the connectivity breaks, this play will resolve this for us.
NOTE This will not work with containerized installations, this has no option to download the installation tarball
from the api. Openshift installations will work.
This will probably be added in future itterations of the product.
Steps
The playbook will take the folowing steps to configure the automation mesh:
- It will address all nodes in its (separate) inventory and:
- check if it is reachable.
- check if the sofware is installed (when reachable).
- If the node is not reachable, it will start the deployment workflow on rhaap and waits for completion.
- It will then define the instance group.
- place the instance in that group.
- Install the sofware.
- run the installation playbook.
- check if the software is running on the host.
- check the availlability in rhaap.
The project: