Should DB should be de-duplicated?

[alexchristy]

Since we'll probably be uploading a lot of duplicate hosts, would it make sense to de-dup the database?

I'm unsure how postgres manages duplicate data internally, so I'm unsure if space is an issue. But it may be annoying to list all the hosts in the DB and get 100 different IDs that point to the same thing

This would put some extra load on the API calls for uploading blueprints but could make things nicer for the user later.

Maybe we could also consider implementing a feature to allow users to manually find and remove duplicate blueprints.

From @Adamkadaban at #48