Warn if `num_workers == 1`

Currently num_workers can wind up being 1 if a machine has only a single CPU (like a minimal cloud instance). In these cases, it might be worth warning users that they only have 1 worker and so they won’t get the benefits of parallelism with Dask. Of course the messaging/communication may need a bit of thought to find the right balance.

[jcrist]

I initially was pro adding a warning here (this came up after debugging why a user wasn’t seeing parallelism in their code). After reading through the comments above though, I think I’m now against adding a warning at all. I could see the benefit of a debug log message though when the scheduler starts up. Something like:

logger = logging.getLogger("dask.local")
logger.debug("Using threaded scheduler with %d threads", num_workers)

at the start of dask.threaded.get/dask.multiprocessing.get/dask.local.get_sync (adjusting the message for each scheduler accordingly).

[GenevieveBuckley]

I think what you’re proposing sounds reasonable. Yes, let’s see if other people also want to weigh in.