Previously, when the server pre and post deployments were run, if one
failed, it would not be automatically retried since the deployed file
for the deployment exists under /var/lib/heat-config/deployed.
This could be overridden by passing -e force=true to ansible-playbook,
however for deployments that have previously failed, I don't think this
should be required, and we should just retry them automatically.
This new behavior more closely matches the traditional behavior with
Heat, where a stack update would automatically attempt to reapply and
deployment that was in a failed state.
Change-Id: I18df88297abe7be50364233ed8f2514de6d3d434
Closes-Bug: #1756921
7b359af52a