We want to copy just 1 month of raw data from current tables to partitioned tables. 

We have ENVs for that, but:

1. job_explorer_rollup has to be re-categorized as daily_rollup (by the data deleting threshold)
2. copy migration has to compute min(threshold, min(unfinished rollup))

TODO: the same could be applied to data pruning, it's not needed to skip pruning completely, but take the minimum allowed date ^^