Feat: Add Terraform module

[mvlassis]

Ref: #63

This PR creates a terraform/ directory that hosts the Terraform module for this charm. It follows the structure proposed in this spec.

To test the module:

• Clone the repository and switch to this branch.
• First run tox -e tflint to ensure that linting is correct
• Change to the terraform/ directory and run terraform init.
• Run terraform apply -var "channel=latest/edge" -var "model_name=kubeflow" --auto-approve and wait until the charm is Active and Idle.

[mvlassis]

Running terraform apply produces the following error:

unit-resource-dispatcher-0: 14:57:25 ERROR unit.resource-dispatcher/0.juju-log Uncaught exception while in charm code:
Traceback (most recent call last):
  File "/var/lib/juju/agents/unit-resource-dispatcher-0/charm/venv/lightkube/codecs.py", line 37, in _load_model
    module = importlib.import_module(f'lightkube.resources.{version.lower()}')
  File "/usr/lib/python3.8/importlib/__init__.py", line 127, in import_module
    return _bootstrap._gcd_import(name[level:], package, level)
  File "<frozen importlib._bootstrap>", line 1014, in _gcd_import
  File "<frozen importlib._bootstrap>", line 991, in _find_and_load
  File "<frozen importlib._bootstrap>", line 973, in _find_and_load_unlocked
ModuleNotFoundError: No module named 'lightkube.resources.metacontroller_v1alpha1'

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "./src/charm.py", line 284, in <module>
    main(ResourceDispatcherOperator)
  File "/var/lib/juju/agents/unit-resource-dispatcher-0/charm/venv/ops/main.py", line 548, in main
    manager.run()
  File "/var/lib/juju/agents/unit-resource-dispatcher-0/charm/venv/ops/main.py", line 527, in run
    self._emit()
  File "/var/lib/juju/agents/unit-resource-dispatcher-0/charm/venv/ops/main.py", line 513, in _emit
    self.framework.reemit()
  File "/var/lib/juju/agents/unit-resource-dispatcher-0/charm/venv/ops/framework.py", line 870, in reemit
    self._reemit()
  File "/var/lib/juju/agents/unit-resource-dispatcher-0/charm/venv/ops/framework.py", line 950, in _reemit
    custom_handler(event)
  File "./src/charm.py", line 171, in _on_install
    self._deploy_k8s_resources()
  File "./src/charm.py", line 161, in _deploy_k8s_resources
    self.k8s_resource_handler.apply()
  File "/var/lib/juju/agents/unit-resource-dispatcher-0/charm/venv/charmed_kubeflow_chisme/kubernetes/_kubernetes_resource_handler.py", line 324, in apply
    resources = self.render_manifests(force_recompute=False)
  File "/var/lib/juju/agents/unit-resource-dispatcher-0/charm/venv/charmed_kubeflow_chisme/kubernetes/_kubernetes_resource_handler.py", line 272, in render_manifests
    self._manifests = codecs.load_all_yaml(
  File "/var/lib/juju/agents/unit-resource-dispatcher-0/charm/venv/lightkube/codecs.py", line 105, in load_all_yaml
    return _flatten(yaml.safe_load_all(stream))
  File "/var/lib/juju/agents/unit-resource-dispatcher-0/charm/venv/lightkube/codecs.py", line 98, in _flatten
    res = from_dict(obj)
  File "/var/lib/juju/agents/unit-resource-dispatcher-0/charm/venv/lightkube/codecs.py", line 61, in from_dict
    model = _load_model(d['apiVersion'], d['kind'])
  File "/var/lib/juju/agents/unit-resource-dispatcher-0/charm/venv/lightkube/codecs.py", line 40, in _load_model
    raise LoadResourceError(f"{e}. If using a CRD, ensure you define a generic resource.")
lightkube.core.exceptions.LoadResourceError: No module named 'lightkube.resources.metacontroller_v1alpha1'. If using a CRD, ensure you define a generic resource.

Looking into it

[mvlassis]

Following this comment, the charm needs the resource definition for metacontroller_v1alpha1. Deploying metacontroller-operator and then running terraform apply now reaches an active state:

Model     Controller  Cloud/Region      Version  SLA          Timestamp
kubeflow  uk8sx       my-k8s/localhost  3.4.5    unsupported  15:07:58+03:00

App                      Version  Status  Scale  Charm                    Channel        Rev  Address         Exposed  Message
metacontroller-operator           active      1  metacontroller-operator  latest/stable  245  10.152.183.107  no       
resource-dispatcher               active      1  resource-dispatcher      latest/edge    181  10.152.183.173  no       

Unit                        Workload  Agent  Address      Ports  Message
metacontroller-operator/0*  active    idle   10.1.90.170         
resource-dispatcher/0*      active    idle   10.1.90.172 

[NohaIhab]

Thanks @mvlassis, I followed your instructions and the charm went to active. Left some comments.

In response to your comment, I think the charm going to error when metacontroller is not there is a bug, the status should be either blocked or waiting I believe. However, this is not related to your change. Can you file an issue about this?

[orfeas-k]

One comment

[orfeas-k]

Let's backport this once merged!