Updating DB workflow
--------------------
1. wget desired tar.xz dump archives for tables for which recipes exist.
2. Make sure enough disk space exists for roughly the whole of UNCOMPRESSED
   data.
3. Run scripts/cassandra_dump_to_line_protocol.sh in the same dir.
4. Inspect ./notice for errors
5. rm tar.xz archives. Note to not needlessly download them again.


Recipes format
--------------
* two lines:
  - input (CSV dump) columns to select in the desired order
  - InfluxDB tags <" "> InfluxDB fields
* See examples.
* timestamp must be the last field on the first line
* second line should match the first line's fields (sans timestamp)