Open emlys opened 10 months ago
Another missing piece from the InVEST side of things is what's a reasonable timeout for super large rasters. I think we cap it to some fixed number of seconds, regardless of the data size. Would it be worth making the timeout a function of the source raster size and the kernel size?
Related: https://github.com/natcap/pygeoprocessing/issues/251
From this forum thread: https://community.naturalcapitalproject.org/t/errors-and-very-long-period-of-running-urban-cooling-model/4298
This could happen if no results are available to
get
from the queue within the default timeout of 60 seconds. We should provide an error message explaining what timed out.