New: parallel query support in SQL (part 2/2)

[lvca]

In the case where most or all the database is loaded in RAM, it would be useful to execute the query on multiple threads. Certain queries like full scan or index lookups could run in parallel, one per bucket.

Also some aggregation could be cpu intensive and could be performed in parallel.

• [x] Parallel query for native select (Implemented in 9c39aba50, see #1296)
• [ ] Parallel query for SQL queries
• [ ] Auto determine when to use parallel query (i.e. multi buckets, more than 10K records, etc)