Memory issues and (maybe?) slow runs about rev HOT 7 CLOSED

@nickwg03, I don't really think it's a code bug, but I would recommend three things:

1. Set "memory_utilization_limit" in the "execution_control" block to 0.2 (default is 0.4)
2. Add more nodes per job. Typically i'll run CONUS (500k sites) across 10 nodes. It looks like you're running 400k sites per node.
3. Target the sites that the jobs are failing at and run only those sites (small subset of project points) in a separate gen test.

Let me know if y'all have time to work these issues or need any help.

from rev.

@WilliamsTravis can you try these and let us know how it goes?

from rev.

Sure thing, I'll give this a shot in a few minutes.

from rev.

@WilliamsTravis, I figure you're smart enough to figure this out, but I would definitely recommend testing some of these in separate run directories with only a single year.

from rev.

Yes, good idea.

from rev.

@WilliamsTravis let us know how it went and if we can chalk this "bug" up to not using enough nodes.

from rev.

Closing this as it seemed to just be a larger spatial extent than we had run previously. More nodes seemed to work fine.

from rev.