Celery Worker has transient connection issues with gcloud-sqlproxy pod on deploy

[Datamance]

Narrative:
As a developer, I would like for the alert channel to stay clean and meaningful. The fact that every deploy results in repeated connection errors as such:

OperationalError
could not connect to server: Connection refused
	Is the server running on host "production-gcloud-sqlproxy" (10.0.10.54) and accepting
	TCP/IP connections on port 5432?

both potentially obscures more critical issues and generally points to imprecision in the deployment process.

Acceptance Criteria:

• These errors no longer happen on deploy.

Implementation Notes:
The solution for this will probably dovetail with that for #1 given the fact that they are both transient deployment errors.