issues

This tool needs to be simplified down to its original purpose, which is simply to import data in multiple threads for speed. Like mk_parallel_dump, the code has proven to be buggy and hard to work with.

The roadmap is mostly in the issue list linked above, but there are also some embedded TODO comments in the tool.

• Simplify the functionality.
• Simplify and test the code.
• Build a smarter algorithm for choosing concurrency, if possible.