Can anyone clarify the expected behavior for genesyscloud_messaging_provider_webchat state drift during automated deployment runs?
Running Terraform apply with provider v1.45. The deployment pipeline (GitHub Actions) consistently reports a diff on the webchat_config block, specifically the widget_type attribute, even though the API response matches the HCL definition. This causes unnecessary re-deployments in the staging environment.
# genesyscloud_messaging_provider_webchat.main will be updated in-place
~ resource "genesyscloud_messaging_provider_webchat" "main" {
id = "12345678-1234-1234-1234-123456789012"
~ webchat_config = [
~ {
~ widget_type = "standard" -> "advanced"
}
]
}
The API GET endpoint /api/v2/messaging/providers/webchat/{providerId} returns widget_type as standard. However, the provider seems to normalize or default to advanced in the state file during the refresh phase. Checked the JSON payload sent to the API; it explicitly sets standard.
Environment:
- TF Provider: 1.45.0
- GC CLI: 2.18.0
- Region: ap-southeast-2
Is this a known provider bug regarding enum mapping, or is there a specific attribute I need to pin to prevent this drift?