streamline 'scheduler' map with klepto backend to collect results

[mmckerns]

It'd be a good strategy to use a database/disk backend (from klepto) to store results from scheduled runs submitted from a pyina pool. Adding a little code to do this seamlessly would be a good improvement.