provider/openstack: Enable DHCP By Default

The openstack_networking_subnet_v2 resource was originally designed
to have DHCP disabled by default; however, a bug in the original
implementation caused DHCP to always be enabled and never be
disabled. This bug was fixed in #6052.

Recent discussions have shown that users prefer if DHCP is enabled
by default. This commit implements makes the change.

Data sources with non-computed configurations should be read during the
refresh walk. This test ensures that this remains true.

Earlier we had a bug where data resources would not yet removed from the
state during a destroy. This was fixed in cd0c4522, and this test will
hopefully make sure it stays fixed.

* provider/fastly: Add cache settings

Docs, tests, and implementation for Cache Settings support

provider/cloudstack: make replacing the ACL of a network update the network in place

provider/cloudstack: list network ACLs with VPC IDs

Closes #6100. Set the default value for the `description` field to be "Managed by Terraform".

Gives the user the ability to specify the controller type they would
like to connect their disk to.  Supported options are scsi and ide.

* 'master' of github.com:hashicorp/terraform:
  Update CHANGELOG.md
  Update CHANGELOG.md
  core: honor "destroy" diffs for data resources
  provider/azurerm: `azurerm_network_interface` diffs didn't match during apply
  providers/google: Don't fail deleting disks that don't exist.

Fixes #5924

providers/google: Don't fail deleting disks that don't exist.

provider/azurerm: `azurerm_network_interface` diffs didn't match during apply

`azurerm_network_interface`

As requested in #6803

```

```

* provider/aws: Update documentation to use t2.micro, provide Classic options

Previously the "planDestroy" pass would correctly produce a destroy diff,
but the "apply" pass would just ignore it and make a fresh diff, turning
it back into a "create" because data resources are always eager to
refresh.

Now we consider the previous diff when re-diffing during apply and so
we can preserve the plan to destroy and then ultimately actually "destroy"
the data resource (remove from the state) when we get to ReadDataApply.

This ensures that the state is left empty after "terraform destroy";
previously we would leave behind data resource states.

* provider/aws: Add support for Skipping Final Snapshot in RDS Cluster

When stage_name is not passed to the resource
aws_api_gateway_deployment a terraform apply will fail. This is
because the stage_name is required and not optional.