Add thread ID to ingest table

[gregorydlogan]

As of 942df1a when running multiple ingest workers we end up with a race condition where ingests get uploaded multiple times. This is likely caused by a misunderstanding of how SQLAlchemy works.

The logic, as it stands, looks like this:
for each recording that's not in FINISHED:
set its state to in-progress
process it
...

I suspect that multiple threads are setting the in-progress value at the same time and sqlalchemy is allowing all of them to commit at the same time because they're all committing the same value. We need to add a thread id as well to prevent this.