Robustness in transient network or server issue.

[nobuto-m]

Requirements

It would be nice if the provider is robust in transient network or server issues. For example, the following output is when charmhub.io gives 503 (I believe) even though other applications were processed properly in the same terraform apply.

I know that terraform apply can be run again after the failure. However, that comes with an unnecessary workaround in an automated deployment like in CI.

| Error: Client Error
|
|   with module.nova.juju_application.nova-api-mysql-router[0],
|   on modules/openstack-api/main.tf line 198, in resource "juju_application" "nova-api-mysql-router":
|  198: resource "juju_application" "nova-api-mysql-router" {
|
| Unable to create application, got error: resolving with preferred channel:
| unexpected content-type from server "text/html"
|
| Error configuring cloud
| Traceback (most recent call last):
|   File "/snap/openstack/661/lib/python3.12/site-packages/sunbeam/core/terraform.py", line 207, in apply
|     process = subprocess.run(
|               ^^^^^^^^^^^^^^^
|   File "/usr/lib/python3.12/subprocess.py", line 571, in run
|     raise CalledProcessError(retcode, process.args,
| subprocess.CalledProcessError: Command '['/snap/openstack/661/bin/terraform', 'apply', '-auto-approve', '-no-color']' returned non-zero exit status 1.

Urgency

It would save from a weird workaround

Notes & References

No response