If you create a resource via the provider, then manually delete that resource via the LogicMonitor portal, the next run of a terraform plan fails due to "unexpected" 404 errors.
What I would expect to happen is that Terraform state is updated to match the state in the tenant (by removing the resource from state). A 404 is not unexpected in this scenario, it is indicative of a change that happened elsewhere that should be reconciled.
In this scenario, I set up the two account device groups incorrectly and opted to delete them manually to have TF recreate them. I had low expectations of the provider being able to handle the conversion from a "Normal" group to an "AWS" group and thought this would be easier. Turns out I was wrong.
logicmonitor_device_group.aws: Refreshing state... [id=1238]
logicmonitor_device_group.aws_test_us: Refreshing state... [id=1239]
logicmonitor_device_group.aws_test_us_accounts["bar"]: Refreshing state... [id=1240]
logicmonitor_device_group.aws_test_us_accounts["foo"]: Refreshing state... [id=1241]
│ Error: unexpected: [GET /device/groups/{id}][404] getDeviceGroupById default &{ErrorCode:1404 ErrorDetail:<nil> ErrorMessage:The hostgroup ID <1240> is not found.}
│ Error: unexpected: [GET /device/groups/{id}][404] getDeviceGroupById default &{ErrorCode:1404 ErrorDetail:<nil> ErrorMessage:The hostgroup ID <1241> is not found.}
Error: Terraform exited with code 1.
Error: Process completed with exit code 1.
I'm only experimenting at this point, but I needed to delete the state files completely in order to move forward since the provider was unable to handle the now missing device groups.
If you create a resource via the provider, then manually delete that resource via the LogicMonitor portal, the next run of a terraform plan fails due to "unexpected" 404 errors.
What I would expect to happen is that Terraform state is updated to match the state in the tenant (by removing the resource from state). A 404 is not unexpected in this scenario, it is indicative of a change that happened elsewhere that should be reconciled.
In this scenario, I set up the two account device groups incorrectly and opted to delete them manually to have TF recreate them. I had low expectations of the provider being able to handle the conversion from a "Normal" group to an "AWS" group and thought this would be easier. Turns out I was wrong.
I'm only experimenting at this point, but I needed to delete the state files completely in order to move forward since the provider was unable to handle the now missing device groups.