Remove HRRR forecast flag

[jomey]

Follow up request for removing the HRRR forecast flag in USDA-ARS-NWRC/smrf/pull/205

Also removed a method call in run_awsm, which definition was commented out.

[scotthavens]

Look like an old weather_forecast_retrieval.hrrr import which was removed.

[jomey]

Look like an old weather_forecast_retrieval.hrrr import which was removed.

SMRF needs a new release to include the updates to the import of WFR v0.7.1, which is installed via pip.
You good with cutting a new 0.11.8?

[scotthavens]

Sure, v0.11.7 will work, all the others were pre releases for testing the deployment.

[jomey]

SMRF release adressed the import error.
However, the OS X job has an interesting one: https://travis-ci.com/github/USDA-ARS-NWRC/awsm/jobs/506653187
Have you seen that before?

ImportError: Failed to import test module: awsm

Lower down there is also this message:

ImportError: dlopen(/Users/travis/build/USDA-ARS-NWRC/awsm/.tox/py37/lib/python3.7/site-packages/smrf/utils/wind/wind_c.cpython-37m-darwin.so, 2): Symbol not found: _aligned_alloc

[jomey]

The other builds are failing with this:

No output has been received in the last 10m0s, this potentially indicates a stalled build or something wrong with the build itself.

I noticed this myself locally, where tests with iPySnobal seem to be hung or stuck. This is with the current master too and I first accepted this as a possible issue on my machine.
Are you able to run them locally?

[scotthavens]

Ugh. Never seen that before. It's complaining that the SMRF wind was compiled for 10.15, try xcode11.4 in Travis to see if that helps?

It appears to be some compilation issue with the C code for either the wind or pysnobal.

[jomey]

Ugh. Never seen that before. It's complaining that the SMRF wind was compiled for 10.15, try xcode11.4 in Travis to see if that helps?

It appears to be some compilation issue with the C code for either the wind or pysnobal.

Bumping XCode addressed this issue. Now there is only the mystery of the hanging iPySnobal

[scotthavens]

Seems to be running but they are all failing...

[jomey]

My next guess is that the tests are not ready for the TopoCalc updates. Might need an update for the Gold files first

[scotthavens]

Does this pass locally for you?

[jomey]

I see the same issues as Travis, where tests using iPySnobal don't finish. Then also the differences to gold files, which I attribute to TopoCalc changes. Getting same errors when running master on my local without these changes.

Should there be a PR first, updating the Gold files with latest of TopoCalc v0.5 ?

[scotthavens]

Yeah, let's do a PR for topocalc v0.5.

You could also try the version <0.5 as a test to see if it's attributed to topocalc.

[jomey]

Yeah, let's do a PR for topocalc v0.5.

Do you want to create the files on a 'true' gold host and push a branch?
I will then do the comparison with goldmeister and open the PR

[scotthavens]

Updated the gold but there is another issue with the threading in Lakes with HRRR. Looking into it now.

[jomey]

Updated the gold but there is another issue with the threading in Lakes with HRRR. Looking into it now.

It hangs for me at the step of Distributing thermal in SMRF. That is the last log entry I get

[scotthavens]

Me too but I've found with the threading, that isn't always where it crashes. Usually it's a dependent of thermal