This removes the need for relying on Rails' "dependent" option for data
removal, which is _incredibly_ slow (even when using :delete_all) when
deleting large amounts of data. This also ensures data consistency is
enforced on DB level and not on application level (something Rails is
really bad at).
This commit also includes various migrations to add foreign keys to
tables that eventually point to "projects" to ensure no rows get
orphaned upon removing a project.
Not only tracking auto-cancelling in pipelines,
we'll also track this in jobs because pipelines
could be retried and the information would get lost
when this happened. Also erase auto-cancelling
relation for pipelines when they're retried.
