Change optscale deploy to make it reusable on the same cluster

[Geraldo155]

Currently when you deploy optscale on a cluster delete it and re-deploy, the diworker pod fails complaining that some tables already exist, when they should not exist according to the way the cluster is structured, currently.