Fixture data import and sync performance

[lnielsen]

Goal

Goal significantly improve data import speed for both records, vocabularies and fixtures

Tasks

• Fixture loading:
  • Build commands for dumping/loading database tables and indexes, so that we can dump and restore something like affiliation data much faster (basically for speed, avoid passing through the Python layer).
  • Tools for having an overview over how much data is loaded and the speed
  • Build tools for performing sync between new versions of the data (e.g. a new ROR dump).