Parallel ingestions on the same machine

[stuarteberg]

We don't customize the port that neo4j uses, so presumably we'll run into conflicts of we attempt to export two different neuprint datasets on the same machine at the same time.

It can't really happen on the Janelia cluster if we require more than half of the slots on any given node. But if we ever reduce our slot/RAM requirements, then this may become an issue.

---

Update: The Janelia cluster recently added 64-core nodes, so if we only request 32 slots per job, we are at risk of having multiple ingestions run in parallel on the same machine!